diff --git a/ATTRIBUTION.md b/ATTRIBUTION.md index dc3412843..1bf3fa163 100644 --- a/ATTRIBUTION.md +++ b/ATTRIBUTION.md @@ -11,6 +11,8 @@ Generated-in-code art is original and has no file to list: the UI icon set (`cra | `assets/hyphenation/en-us.pat` | DesignCraft contributors (generated) | trained from the Moby list by DesignCraft's patgen | Public Domain (CC0-1.0) | Our own Liang hyphenation patterns (no TeX patterns) | | `assets/hyphenation/en-us.dic` | Grady Ward (Moby Hyphenator II word list); converted by DesignCraft contributors | https://www.gutenberg.org/ebooks/3204 (file `mhyph.txt`, SHA-256 `eeb30474c86b8af3469035ec1a0913e35905325ca885290db9dfed9881e230ac`), via DesignCraft | Public Domain | ~165k words with break points; drives spelling and hyphenation | | `assets/spelling/en-extra.txt` | WordCraft contributors | original word list written for this repository (common modern words the Moby list lacks); no other dictionary copied | MIT OR Apache-2.0 | ~220 lowercase English base forms supplementing spelling (#520) | +| `assets/hyphenation/pl.pat` | Hanna Kołodziejska, Bogusław Jackowski, Marek Ryćko (Polish TeX hyphenation patterns, 1987–1995); hyph-utf8 package | https://github.com/hyphenation/tex-hyphen (`hyph-utf8/tex/generic/hyph-utf8/patterns/tex/hyph-pl.tex`, SHA-256 `83ddbbd21bc06a44bc4f4a4c49c38782e809cff5f026b764c044daa29e818843`), bundled unmodified | MIT (the file is offered under MIT or Knuth's public-domain terms; WordCraft uses MIT, whose notice is in the file's header) | 4,053 Liang patterns and 20 exception words; Polish hyphenation | +| `assets/hyphenation/pt.pat` | Pedro J. de Rezende, J. Joao Dias Almeida, Leonardo Araujo, Aline Benevides (Portuguese TeX hyphenation patterns, 1987–2024); hyph-utf8 package | https://github.com/hyphenation/tex-hyphen (`hyph-utf8/tex/generic/hyph-utf8/patterns/tex/hyph-pt.tex`, SHA-256 `c110a39e501db372d1d755e1c64464321a2efa7d8e9617f66001000389b941f2`), bundled unmodified | BSD 3-clause (the file's licence header names it and carries the notice) | 426 Liang patterns and 2 exception words; Portuguese hyphenation | | `assets/app-icon/wordcraft-1024.png` | WordCraft contributors | original artwork drawn in SVG (`assets/app-icon/wordcraft.svg`), rendered by `packaging/icons.sh` | MIT OR Apache-2.0 | Placeholder app icon until the WordCraft mascot drawing lands | | `packaging/macos/dmg/background.svg` | @XusBadia | original work for this repository, derived from the WordCraft app icon (`assets/app-icon/wordcraft.svg`, referenced via ``, not copied); text is outlined Inter / JetBrains Mono glyph paths | MIT OR Apache-2.0, like the code ([`LICENSE-MIT`](LICENSE-MIT), [`LICENSE-APACHE`](LICENSE-APACHE)) | WordCraft macOS DMG window background (2026-10-08); used for the Finder window of the macOS DMG, rendered to `background.tiff` by `packaging/macos/dmg/generate.py` | | `packaging/macos/dmg/background.tiff` | @XusBadia | Rendered from `packaging/macos/dmg/background.svg` by `packaging/macos/dmg/generate.py` | MIT OR Apache-2.0, like the code ([`LICENSE-MIT`](LICENSE-MIT), [`LICENSE-APACHE`](LICENSE-APACHE)) | 1x + 2x HiDPI TIFF for the macOS DMG window; includes the app icon (see its row) and LittleCMS's built-in sRGB profile ("No copyright, use freely") | @@ -63,6 +65,12 @@ Generated-in-code art is original and has no file to list: the UI icon set (`cra | craft-fonts (optional build input, not files in this repo): BIZ UDPGothic, BIZ UDMincho, Shippori Mincho, Noto Sans CJK SC, Noto Sans Arabic | see https://github.com/storytold/craft-fonts/blob/main/ATTRIBUTION.md | embedded only when built with `CRAFT_FONTS_DIR` (all official releases) | OFL-1.1 | Japanese and Chinese UI, Japanese document fallback | | `crates/ui-egui/src/i18n/uk.tsv` | @dmatviichuk with Codex | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Ukrainian interface catalog; spelling checked with external VESUM and ULIF dictionaries, neither bundled | | `crates/ui-egui/src/i18n/et.tsv` | @henno with Codex | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Estonian interface catalog | +| `crates/ui-egui/src/i18n/pl.tsv` | Konrad Witkowski with Claude (AI agent) | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Polish interface catalog; not yet reviewed by a second native speaker | | `crates/ui-egui/src/i18n/pt-br.tsv` | @edurbs with ZCode | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Brazilian Portuguese interface catalog; spelling checked with the external pt_BR Hunspell dictionary, not bundled | +| `crates/ui-egui/src/i18n/cs.tsv` | @DanielKnourek with Antigravity for VS Code | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Czech interface catalog; review suggestions by @Losminek | +| `crates/ui-egui/src/i18n/de.tsv` | @arekown with Kimi | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | German interface catalog | | `crates/ui-egui/src/i18n/ru.tsv` | Dmitrii Khlapov with Claude | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Russian interface catalog; reviewed by a native speaker | | `crates/ui-egui/src/i18n/tr.tsv` | @PbyLoyth with Codex and Claude Code | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Turkish interface catalog | +| `crates/ui-egui/src/i18n/bg.tsv` | Dimitar Klaturov (@bulgariamitko) with Claude Code | Original translation from the English interface labels in this repository; no other product localisation or other catalog used | MIT OR Apache-2.0 | Bulgarian interface catalog | +| `crates/ui-egui/src/i18n/hu.tsv` | @Skaywalker | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | Hungarian interface catalog | +| `crates/ui-egui/src/i18n/fr.tsv` | @pechnikov2026-star | Original translation from the English interface labels in this repository; no other product localisation copied | MIT OR Apache-2.0 | French interface catalog | diff --git a/Cargo.lock b/Cargo.lock index 01aa626c4..6902a9119 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -241,6 +241,7 @@ dependencies = [ "parking_lot", "percent-encoding", "windows-sys 0.60.2", + "wl-clipboard-rs", "x11rb", ] @@ -1331,6 +1332,12 @@ version = "0.1.14" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "aedcfb3409746eddb02b9e19ebda1c3394f759a152e48ee875a0844d1b955484" +[[package]] +name = "fixedbitset" +version = "0.5.7" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "1d674e81391d1e1ab681a28d99df07927c6d4aa5b027d7da16ba32d1d21ecd99" + [[package]] name = "flate2" version = "1.1.10" @@ -1368,6 +1375,12 @@ version = "1.0.7" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "3f9eec918d3f24069decb9af1554cad7c880e2da24a9afd88aca000531ab82c1" +[[package]] +name = "foldhash" +version = "0.1.5" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "d9c4f5dac5e15c24eb999c26181a6ca40b39fe946cbe4c263c7209467bc83af2" + [[package]] name = "foldhash" version = "0.2.0" @@ -1555,7 +1568,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "ed4a1bb24121291d27230c1b1b44e07d6a9b28cefdb32fe1581dfb84e14f940a" dependencies = [ "bytemuck", - "foldhash", + "foldhash 0.2.0", "hashbrown 0.17.1", "log", "peniko", @@ -1571,7 +1584,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "282a26c1e23de04bdab3e34a21b6f877a479b96737792516b9b8b8f69b6661be" dependencies = [ "bytemuck", - "foldhash", + "foldhash 0.2.0", "hashbrown 0.17.1", "log", "peniko", @@ -1719,6 +1732,15 @@ dependencies = [ "smallvec", ] +[[package]] +name = "hashbrown" +version = "0.15.5" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "9229cfe53dfd69f0609a49f65461bd93001ea1ef889cd5529dd176593f5338a1" +dependencies = [ + "foldhash 0.1.5", +] + [[package]] name = "hashbrown" version = "0.16.1" @@ -1727,7 +1749,7 @@ checksum = "841d1cc9bed7f9236f321df977030373f4a4163ae1a7dbfe1a51a2c1a51d9100" dependencies = [ "allocator-api2", "equivalent", - "foldhash", + "foldhash 0.2.0", ] [[package]] @@ -1736,7 +1758,7 @@ version = "0.17.1" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "ed5909b6e89a2db4456e54cd5f673791d7eca6732202bbf2a9cc504fe2f9b84a" dependencies = [ - "foldhash", + "foldhash 0.2.0", ] [[package]] @@ -2415,6 +2437,15 @@ version = "0.2.0" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "2bf50223579dc7cdcfb3bfcacf7069ff68243f8c363f62ffa99cf000a6b9c451" +[[package]] +name = "nom" +version = "8.0.0" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "df9761775871bdef83bee530e60050f7e54b1105350d6884eb0fb4f46c2f9405" +dependencies = [ + "memchr", +] + [[package]] name = "num-traits" version = "0.2.19" @@ -2812,6 +2843,16 @@ dependencies = [ "pin-project-lite", ] +[[package]] +name = "os_pipe" +version = "1.2.3" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "7d8fae84b431384b68627d0f9b3b1245fcf9f46f6c0e3dc902e9dce64edd1967" +dependencies = [ + "libc", + "windows-sys 0.52.0", +] + [[package]] name = "parking" version = "2.2.1" @@ -2872,6 +2913,17 @@ version = "2.3.2" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220" +[[package]] +name = "petgraph" +version = "0.8.3" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "8701b58ea97060d5e5b155d383a69952a60943f0e6dfe30b04c287beb0b27455" +dependencies = [ + "fixedbitset", + "hashbrown 0.15.5", + "indexmap", +] + [[package]] name = "phf" version = "0.13.1" @@ -3811,7 +3863,10 @@ version = "0.3.2" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "8eab9a99a024a169fe8a903cf9d4a3b3601109bcc13bd9e3c6fff259138626c4" dependencies = [ + "js-sys", "libc", + "wasm-bindgen", + "web-sys", ] [[package]] @@ -4012,6 +4067,17 @@ dependencies = [ "once_cell", ] +[[package]] +name = "tree_magic_mini" +version = "3.2.2" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "b8765b90061cba6c22b5831f675da109ae5561588290f9fa2317adab2714d5a6" +dependencies = [ + "memchr", + "nom", + "petgraph", +] + [[package]] name = "ttf-parser" version = "0.25.1" @@ -5012,11 +5078,30 @@ version = "0.57.1" source = "registry+https://github.com/rust-lang/crates.io-index" checksum = "1ebf944e87a7c253233ad6766e082e3cd714b5d03812acc24c318f549614536e" +[[package]] +name = "wl-clipboard-rs" +version = "0.9.4" +source = "registry+https://github.com/rust-lang/crates.io-index" +checksum = "4d7888ccd4896447b2d14d3a9350a85df2aeb6f181e2e7a31349d104ac46cac1" +dependencies = [ + "libc", + "log", + "os_pipe", + "rustix 1.1.5", + "thiserror 2.0.21", + "tree_magic_mini", + "wayland-backend", + "wayland-client", + "wayland-protocols", + "wayland-protocols-wlr", +] + [[package]] name = "wordcraft" -version = "0.5.0" +version = "0.6.0" dependencies = [ "arboard", + "clipboard-win", "craft-print", "dispatch2", "eframe", @@ -5026,22 +5111,26 @@ dependencies = [ "log", "objc2 0.6.5", "objc2-app-kit 0.3.2", + "open", "pollster 0.4.0", "rfd", "serde_json", "windows", + "winit", "winresource", + "wl-clipboard-rs", "wordcraft-control-key", "wordcraft-doc", "wordcraft-engine", "wordcraft-fonts", + "wordcraft-proof", "wordcraft-ui-egui", "zbus", ] [[package]] name = "wordcraft-cli" -version = "0.5.0" +version = "0.6.0" dependencies = [ "serde_json", "wordcraft-doc", @@ -5053,11 +5142,11 @@ dependencies = [ [[package]] name = "wordcraft-control-key" -version = "0.5.0" +version = "0.6.0" [[package]] name = "wordcraft-doc" -version = "0.5.0" +version = "0.6.0" dependencies = [ "proptest", "serde", @@ -5069,10 +5158,12 @@ dependencies = [ [[package]] name = "wordcraft-docbin" -version = "0.5.0" +version = "0.6.0" dependencies = [ "cfb", + "image", "log", + "miniz_oxide 0.8.9", "proptest", "serde_json", "thiserror 2.0.21", @@ -5082,7 +5173,7 @@ dependencies = [ [[package]] name = "wordcraft-docx" -version = "0.5.0" +version = "0.6.0" dependencies = [ "aes", "base64", @@ -5104,7 +5195,7 @@ dependencies = [ [[package]] name = "wordcraft-engine" -version = "0.5.0" +version = "0.6.0" dependencies = [ "image", "js-sys", @@ -5130,7 +5221,7 @@ dependencies = [ [[package]] name = "wordcraft-fonts" -version = "0.5.0" +version = "0.6.0" dependencies = [ "harfrust 0.13.3", "kurbo", @@ -5143,7 +5234,7 @@ dependencies = [ [[package]] name = "wordcraft-formats" -version = "0.5.0" +version = "0.6.0" dependencies = [ "image", "log", @@ -5156,14 +5247,14 @@ dependencies = [ [[package]] name = "wordcraft-geom" -version = "0.5.0" +version = "0.6.0" dependencies = [ "serde", ] [[package]] name = "wordcraft-layout" -version = "0.5.0" +version = "0.6.0" dependencies = [ "image", "js-sys", @@ -5182,7 +5273,7 @@ dependencies = [ [[package]] name = "wordcraft-mcp" -version = "0.5.0" +version = "0.6.0" dependencies = [ "log", "serde_json", @@ -5194,7 +5285,7 @@ dependencies = [ [[package]] name = "wordcraft-metafile" -version = "0.5.0" +version = "0.6.0" dependencies = [ "kurbo", "thiserror 2.0.21", @@ -5202,7 +5293,7 @@ dependencies = [ [[package]] name = "wordcraft-pdf" -version = "0.5.0" +version = "0.6.0" dependencies = [ "hayro-cmap", "hayro-interpret", @@ -5222,7 +5313,7 @@ dependencies = [ [[package]] name = "wordcraft-proof" -version = "0.5.0" +version = "0.6.0" dependencies = [ "log", "miniz_oxide 0.8.9", @@ -5231,7 +5322,7 @@ dependencies = [ [[package]] name = "wordcraft-render" -version = "0.5.0" +version = "0.6.0" dependencies = [ "image", "log", @@ -5245,7 +5336,7 @@ dependencies = [ [[package]] name = "wordcraft-ui-egui" -version = "0.5.0" +version = "0.6.0" dependencies = [ "craft-ui", "egui", @@ -5264,12 +5355,13 @@ dependencies = [ "wordcraft-layout", "wordcraft-render", "wordcraft-zotero", + "zbus", "zip", ] [[package]] name = "wordcraft-web" -version = "0.5.0" +version = "0.6.0" dependencies = [ "eframe", "egui", @@ -5288,7 +5380,7 @@ dependencies = [ [[package]] name = "wordcraft-zotero" -version = "0.5.0" +version = "0.6.0" dependencies = [ "log", "proptest", @@ -5401,7 +5493,7 @@ checksum = "9440ea3e5aeabb0ac63af70daf835274065238cdd0cec83418f417eae38bacee" [[package]] name = "xtask" -version = "0.5.0" +version = "0.6.0" dependencies = [ "serde_json", ] diff --git a/Cargo.toml b/Cargo.toml index 438cbc9ee..3ca7c3cb1 100644 --- a/Cargo.toml +++ b/Cargo.toml @@ -4,7 +4,7 @@ members = ["crates/*", "apps/*", "xtask"] default-members = ["crates/*", "apps/wordcraft", "apps/wordcraft-cli", "xtask"] [workspace.package] -version = "0.5.0" +version = "0.6.0" edition = "2024" license = "MIT OR Apache-2.0" rust-version = "1.90" @@ -55,6 +55,8 @@ image = { version = "0.25", default-features = false, features = ["png", "jpeg", krilla = { version = "0.8", features = ["pdf"] } quick-xml = "0.41" zip = { version = "3", default-features = false, features = ["deflate"] } +# zlib streams (compressed metafile pictures in `.doc` drawings); already in the lock as a dependency of other crates. +miniz_oxide = "0.8" cfb = "0.15" # Password-protected (agile-encrypted) Word documents: pure-Rust RustCrypto primitives. aes = "0.9" @@ -71,7 +73,7 @@ egui_extras = { version = "0.36", default-features = false } eframe = { version = "0.36", default-features = false } egui_kittest = "0.36" rfd = "0.17" -arboard = { version = "3.6", default-features = false } +arboard = { version = "3.6", default-features = false, features = ["wayland-data-control"] } [workspace.lints.rust] unsafe_code = "forbid" diff --git a/NOTICE b/NOTICE index 4bc9002c8..8d09e43bd 100644 --- a/NOTICE +++ b/NOTICE @@ -23,6 +23,35 @@ list (author, source, license) is in ATTRIBUTION.md. - English word list (assets/hyphenation/en-us.dic) — derived from the Moby Hyphenator II list by Grady Ward, public domain (Project Gutenberg eBook #3204). The hyphenation patterns (assets/hyphenation/en-us.pat) were trained from it and are released as CC0-1.0. +- Polish hyphenation patterns (assets/hyphenation/pl.pat) — hyph-pl.tex from the hyph-utf8 + package, Copyright (C) 1987-1995 Hanna Kołodziejska, Bogusław Jackowski, Marek Ryćko, bundled + unmodified and used under the MIT License; the copyright and permission notice is in the file. +- Portuguese hyphenation patterns (assets/hyphenation/pt.pat) — hyph-pt.tex from the hyph-utf8 + package, bundled unmodified and used under the BSD 3-Clause License: + + Copyright (C) 1987, 1994, 1996, 2015 Pedro J. de Rezende, 1996, 2015 J. Joao Dias Almeida, + 2024 Leonardo Araujo and Aline Benevides + + Redistribution and use in source and binary forms, with or without modification, are + permitted provided that the following conditions are met: + * Redistributions of source code must retain the above copyright notice, this list of + conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright notice, this list of + conditions and the following disclaimer in the documentation and/or other materials + provided with the distribution. + * Neither the name of the University of Campinas, of the University of Minho nor the names + of its contributors may be used to endorse or promote products derived from this software + without specific prior written permission. + + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS + OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL PEDRO + J. DE REZENDE OR J.JOAO DIAS ALMEIDA BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, + EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE + GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING + NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED + OF THE POSSIBILITY OF SUCH DAMAGE. - All UI icons are drawn in code (crates/ui-egui/src/icons.rs) and are original work, MIT OR Apache-2.0. Rust dependencies (egui, eframe, vello_cpu, skrifa, harfrust, krilla, quick-xml, zip, image, …) diff --git a/README.md b/README.md index d273d490a..2929018bb 100644 --- a/README.md +++ b/README.md @@ -115,9 +115,9 @@ Every screenshot below is WordCraft itself, rendered offscreen by its own UI tes | **Pages** | Margins, orientation, size, columns, page/column/section breaks, headers and footers (first page, odd/even), page numbers, watermark, page borders, line numbers, vertical alignment, drop caps, automatic hyphenation | | **Objects** | Pictures (resize, crop, recolour, brightness/contrast, transparency, background removal, picture styles, rotate), shapes, text boxes, floating position with text wrapping (square, top and bottom, behind or in front of text) | | **References** | Table of contents, footnotes and endnotes, citations and bibliography (APA, MLA, Chicago, IEEE), captions, table of figures, cross-references, index, table of authorities | -| **Review** | Spelling and grammar with suggestions, thesaurus, word count, comments in margin balloons or a pane, track changes, accept/reject, compare documents, restrict editing, accessibility checker, document inspector | +| **Review** | Spelling and grammar with suggestions (English and Polish, by each run's language), thesaurus, word count, comments in margin balloons or a pane, track changes, accept/reject, compare documents, restrict editing, accessibility checker, document inspector | | **Mailings** | Mail merge from CSV, merge fields, address block, greeting line, rules, preview, finish to a document; envelopes and labels | -| **View** | Print layout, web layout, draft, read mode, focus, zoom, one/multiple pages, page width, Navigation pane, rulers, gridlines, dark mode; interface in English, 简体中文, 繁體中文, 日本語, Português (Brasil), Español, Українська, Српски / Srpski, Eesti, Türkçe or Русский (follows the system language by default) | +| **View** | Print layout, web layout, draft, read mode, focus, zoom, one/multiple pages, page width, Navigation pane, rulers, gridlines, dark mode; interface in English, 简体中文, 繁體中文, 日本語, Português (Brasil), Español, Українська, Српски / Srpski, Eesti, Türkçe, Русский, Italiano, Deutsch, Polski, Čeština, Magyar, Français or Български (follows the system language by default) | | **Files** | .docx read/write (opens in Word), Word 97-2003 .doc import, PDF export, .odt, .rtf, .html, .md, .tex, .txt import/export, page images | The honest picture, area by area, is in [ROADMAP.md](ROADMAP.md), the @@ -151,6 +151,7 @@ WordCraft was designed to be driven by people *and* by AI agents. ```sh claude mcp add wordcraft -- wordcraft-cli mcp # headless documents +# confine the agent to folders: wordcraft-cli mcp --automation-read-root DIR --automation-write-root DIR wordcraft --control 7981 & # or drive the running app… claude mcp add wordcraft-app -- wordcraft-cli mcp --connect 127.0.0.1:7981 ``` @@ -234,6 +235,10 @@ Remove-Item Env:WGPU_BACKEND # restore the default for later An explicit `gl` override can bring the driver crash back on affected systems. The macOS, Linux and web backend defaults are unchanged. +The **32-bit** build uses OpenGL even when a DirectX 12 graphics card is present. On 32-bit +Windows, creating a DirectX 12 graphics pipeline fails immediately and the app exits before the +window appears. `WGPU_BACKEND=vulkan` (or `dx12`) still overrides that. + **If the app quits right after its window appears with "Parent device is lost" in the log:** WordCraft compiles DirectX 12 shaders with FXC, the compiler built into Windows, so it needs no `dxcompiler.dll`/`dxil.dll` and ignores any copy another program put on `PATH` (#603). @@ -267,17 +272,23 @@ instead, without setting anything. The log file records the choice (lines starti RISC-V (riscv64): `wordcraft--linux-riscv64.tar.gz`, a tarball cross-compiled for glibc 2.39+ (Ubuntu 24.04, Debian 13 and newer). +**Wayland:** on a Wayland desktop the app runs through XWayland when it is available, because +winit (the windowing library) supports dragging files into the window only on X11. To use the +native Wayland backend instead (for example for sharper text with fractional scaling, at the cost +of file drag-and-drop), start it with `WORDCRAFT_WAYLAND=1`: + +```sh +WORDCRAFT_WAYLAND=1 wordcraft +``` + **Arch Linux (AUR).** Community members maintain three AUR packages; the project doesn't publish to the AUR itself. Install one of them: - | Package | What it installs | Choose it when | |---|---|---| | [`wordcraft-bin`](https://aur.archlinux.org/packages/wordcraft-bin) | The release tarball above, prebuilt (x86_64, aarch64) | You want the latest release without compiling (quickest) | | [`wordcraft`](https://aur.archlinux.org/packages/wordcraft) | The latest release, built from source with `cargo` | You prefer to build releases yourself | | [`wordcraft-git`](https://aur.archlinux.org/packages/wordcraft-git) | The current `main` branch, built from source | You want unreleased changes and accept an occasional broken build | - For example `yay -S wordcraft-bin` or `paru -S wordcraft-bin`. - ### FreeBSD | Build | File | @@ -339,8 +350,10 @@ Copyright (c) 2026 ArtCraft Team and the WordCraft contributors. Required notice Bundled fonts, icons, images and other assets keep their own open licenses; each one is listed with its author, source and license in [ATTRIBUTION.md](ATTRIBUTION.md). -The spelling dictionary and hyphenation come from Grady Ward's public-domain Moby Hyphenator II -word list. The sample documents and templates are original text written for WordCraft. +The English spelling dictionary and hyphenation come from Grady Ward's public-domain Moby +Hyphenator II word list. Polish spelling reads the Hunspell `pl_PL` dictionary installed on the system (not bundled) and +Polish hyphenation the TeX patterns by Hanna Kołodziejska, Bogusław Jackowski and Marek Ryćko (MIT +option). The sample documents and templates are original text written for WordCraft. The ArtCraft name, wordmark and logos in [`docs/brand/`](docs/brand/) are trademarks of the ArtCraft Team and are not covered by this license. They may be used only unmodified, and only as diff --git a/ROADMAP.md b/ROADMAP.md index 8249f2bef..2b8cf998a 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -2,7 +2,7 @@ **Stage: alpha** · next: beta, ~15 points (60% → 75% ready for real work) and ~150–230 h away -> **Last reviewed:** 2026-10-10 · **Last updated:** 2026-10-11 · **Change:** trivial (Animation effects setting landed; previously trivial: every measurement follows the Options unit, #755; previously trivial: Insert › SmartArt landed, #497: catalog 429/451 → 437/458 (95.4%); previously trivial: Turkish interface landed; previously trivial: formatting revisions landed, #41; previously trivial: Draw tab ink, #307, and Draw Table and Eraser, #303, landed: catalog 402/434; previously minor: readiness table with hours per audience; full number back to the additive weighted sum, 60%; mainstream and essentials numbers added; alpha gate checked; previously major: full re-measure against Word for Mac 16.113.4; restructured to the craftrules progress-docs standard) · **Target:** Microsoft Word (Microsoft 365) for Mac 16.113.4 +> **Last reviewed:** 2026-10-11 · **Last updated:** 2026-10-11 · **Change:** trivial (drags and Shift+arrows over table cells select whole cells; previously trivial: table selection areas, insert buttons, move handle and double-click AutoFit on the page, Select Column: catalog 484/499; previously trivial: independent complex-script formatting preserved across DOCX open/save; previously trivial: Turkish interface landed; previously trivial: formatting revisions landed, #41; previously trivial: Draw tab ink, #307, and Draw Table and Eraser, #303, landed: catalog 402/434; previously minor: readiness table with hours per audience; full number back to the additive weighted sum, 60%; mainstream and essentials numbers added; alpha gate checked; previously major: full re-measure against Word for Mac 16.113.4; restructured to the craftrules progress-docs standard) · **Target:** Microsoft Word (Microsoft 365) for Mac 16.113.4 > **Printing update:** 2026-10-12 · direct desktop printing through craft-print; details below. @@ -16,7 +16,7 @@ agent control. This page is the summary; the assessment is | Number | Value | Kind | |---|---|---| -| Ribbon/menu catalog coverage | **483 / 498 (97.0%)** | measured (`cargo xtask parity` → [`docs/parity-checklist.md`](docs/parity-checklist.md)) | +| Ribbon/menu catalog coverage | **485 / 500 (97.0%)** | measured (`cargo xtask parity` → [`docs/parity-checklist.md`](docs/parity-checklist.md)) | | **Feature breadth** (weighted, incl. dialog options, styles, languages beyond the ribbon) | **~80%** | estimated | | **Ready for real work** (full target) | **~60%** (55–63%) | estimated, additive weighted sum over the dimensions | | **Mainstream practitioner** | **~55%** | estimated: weekly-work depth × discounts for interaction (×0.93), stability (×0.90), file exchange (×0.90) | @@ -60,7 +60,7 @@ the two partial ones (real-world .docx, printing) complete, roughly. | File formats | 60% | 185–300 | [file-format-parity.md](docs/file-format-parity.md) | | Layout and pagination fidelity | 55% | 90–145 | [layout-parity.md](docs/layout-parity.md) | | Hardware | 40% | 20–40 | [hardware-parity.md](docs/hardware-parity.md) | -| Localization | 25% | 80–130 + native review | [localization-parity.md](docs/localization-parity.md) | +| Localization | 27% | 75–120 + native review | [localization-parity.md](docs/localization-parity.md) | | Performance | 75% | 10–20 | [layout-parity.md](docs/layout-parity.md#performance) | | Stability | 50% | 25–45 | [gaps.md](docs/gaps.md) (#3) | | Platforms | 95% (beyond Word on Linux, BSD, web) | 5–10 | this page | @@ -87,7 +87,7 @@ the two partial ones (real-world .docx, printing) complete, roughly. | Pictures, shapes, text boxes, WordArt | 60% | 20–30 | | Backstage, printing, options | 58% | 12–20 | | Right-to-left and complex scripts | 50% | 15–25 | -| Proofing (spelling, grammar, thesaurus, languages) | 35% | 25–45 | +| Proofing (spelling, grammar, thesaurus, languages) | 40% | 20–40 | | Accessibility | 35% | 15–25 | | East Asian typography | 25% | 15–25 | | Charts, SmartArt, Draw/ink, 3D models, icons | 3% | 70–110 | @@ -104,16 +104,18 @@ Weights and evidence per area: [docs/target-app-parity.md](docs/target-app-parit | Spanish | es | partial | ~95% | | Hindi | hi | none | 0% | | Arabic | ar | none (document shaping and bidi work; no mirrored UI) | 0% | -| French | fr | none | 0% | +| French | fr | partial (interface) | ~95% | | Portuguese | pt-br (pt-PT none) | partial | ~95% | | Indonesian | id | none | 0% | | Japanese | ja | partial | ~95% | -| German | de | none (PR #203 open) | 0% | +| German | de | partial (#574) | ~95% | | Korean | ko | none | 0% | | Vietnamese | vi | none | 0% | -Also shipped: Traditional Chinese (`zh-hant`), Ukrainian (`uk`), Serbian (`sr`, `sr-latn`, #250) and Russian (`ru`), all partial (~95%). Estonian (`et`) and Turkish (`tr`) now have complete 1,307-entry interface catalogs with -system-locale selection and a saved language choice; Estonian and Turkish proofing are not included. Word ships +Also shipped: Traditional Chinese (`zh-hant`), Ukrainian (`uk`), Serbian (`sr`, `sr-latn`, #250), Russian (`ru`) and Polish (`pl`), all partial (~95%). Bulgarian (`bg`) covers everything the interface audit finds (2,419 entries). Estonian (`et`), Italian (`it`, #594) and Czech (`cs`) now have complete interface catalogs, and Turkish (`tr`) a complete 2,325-entry one (every key of the Russian catalog), all with +system-locale selection and a saved language choice (Italian also has 51 strings none of the others had yet); Estonian, Turkish, Italian and Czech proofing are not included. Hungarian (`hu`) has a +complete 2,419-entry catalog, also without proofing. +Proofing: English, Polish and Portuguese (spelling, grammar, hyphenation), chosen per run from its language. Word ships 30 interface languages and proofing for ~50. Detail: [docs/localization-parity.md](docs/localization-parity.md). ## Upcoming @@ -132,9 +134,53 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm ## Progress log | Date | What landed | +|---|---| +| 2026-10-11 | Right-clicking a table cell offers the table's commands, and as in Word they follow the selection: with the caret in a cell Insert, Delete Cells, Split Cells, Text Direction and Table Properties…; with several cells Merge Cells instead of Split; with whole columns or rows Delete Columns or Rows, Merge Cells and Distribute Rows and Columns; with the whole table Delete Table, Distribute, AutoFit and Insert Caption…. A right-click beside selected cells moves the caret there instead of keeping them | +| 2026-10-11 | Drags, Shift+arrows and Shift+click over table cells select whole cells, as in Word: from one cell into another the rectangle between them (bold, alignment and Delete leave the other cells alone), into or out of a table whole rows; extending again goes on from where the selection started | +| 2026-10-11 | More of the table mouse: arrows above a column, left of a row and along a cell's left edge select it (a drag selects more; a selected column is edited cell by cell, numbering it makes one list), "+" buttons beside the borders insert a row or column, the move handle past the top-left corner selects the table or drags it to float where it's dropped, and a double-click on a column border fits the column; Select Column in the ribbon; `table.selectAt`, `table.insertAt`, `table.moveDrag`, `table.fitColumn` and `table.selectColumn` for agents | +| 2026-10-12 | AutoCorrect keys its built-in English replacement list by language (#835): Polish, Croatian, Bosnian and Serbian `i` ("and") is no longer turned into `I`; English and untagged text unchanged | +| 2026-10-12 | Ribbon groups fit above their labels: View › Show splits its four check boxes into two columns (Hide White Space, #645, had pushed the fourth onto the "Show" label), Developer › Controls loses the gaps that ran into its label, and Layout › Paragraph (#654) no longer trips a debug assertion and sits 2.5 pt higher; a test draws every tab in every language and fails on any group taller than the ribbon | +| 2026-10-12 | Window caption: a non-Latin file name is drawn with a document fallback face (or the CJK UI font), instead of `?` | +| 2026-10-12 | 32-bit Windows starts on OpenGL. DirectX 12 graphics pipeline creation returns "The parameter is incorrect" and the app exited before the window appeared | +| 2026-10-12 | New documents follow the system region (#838): A4 and centimetres outside the US (Letter in the CLDR Letter territories, inches in the US), lengths written with the region's decimal separator (`2,54 cm`) and typed with either; an explicit Measurement units choice wins; `file.options {"region"}` for agents | +| 2026-10-12 | Paste from other programs keeps their formatting (#687): Ctrl+V and Home › Paste read the system clipboard's HTML (and RTF on Windows and Wayland); Merge Formatting takes the destination's font and paragraph and keeps bold, italic, underline and strikethrough; Paste Special lists what the system clipboard holds (HTML, RTF, picture, text). Linux: pictures copied in a browser or Spectacle paste on KDE and other Wayland desktops (data-control clipboard), and a paste that finds nothing usable names what the clipboard holds instead of "empty" (#833) | +| 2026-10-12 | Undo history list (#701): a ▾ beside the Quick Access Toolbar's Undo lists recent steps newest first; hovering highlights what a click undoes ("Undo 3 Actions"), and a click undoes them in one go. `edit.undoTo {count}` and `edit.undoHistory` (newest first, up to 100) for CLI and MCP | +| 2026-10-12 | Start screen (#766): launched with nothing to open, WordCraft starts on File › Home (Blank document, templates, Recent) like Word, desktop and web; Esc leaves a blank document; File › Options › General › Start-up turns it off | +| 2026-10-12 | Word 365 cover pages (#743): a freeform inside a group keeps its shape (it was drawn as a rectangle); text boxes keep their own internal margins and Resize shape to fit text, so a tight title box no longer loses its second line; both survive saving. The missing fills in the report were theme-coloured shapes, fixed by #611 | +| 2026-10-12 | Alt+Q moves to "Tell me" command search, as in Word for Windows (#819); not on macOS, where Option+Q types a character | +| 2026-10-12 | PDF export of long documents with many internal links (a linked table of contents) is linear: NIH's 67-page protocol template 10.6 s → 0.9 s | +| 2026-10-12 | Keytips: Word for Windows' tab letters (File F, Design G, References S, Review R, View W, Help Y) | +| 2026-10-12 | Table-of-contents links look like Word's: plain text, not blue and underlined (Word ignores the Hyperlink style on TOC entry links; measured in 9 cases) | +| 2026-10-12 | Line heights like Word's: spaces and tabs never make a line taller (13 cases measured in Word for Windows); tables of contents with big tab runs were a third taller | +| 2026-10-12 | `.doc` import: tab stops (centred titles, right-aligned dates, leaders) and sections ending at their section marks (continuous breaks, footers); EPA .doc templates now paginate like Word | +| 2026-10-12 | DOCX headings numbered like Word when a heading style gives only a list level (6.4.5, not 6.15) | +| 2026-10-12 | A list label makes its line taller above the baseline only, as in Word (6 cases measured in Word for Windows): Symbol bullets no longer add their deeper descent, so bulleted lists keep Word's height and page breaks | +| 2026-10-12 | Hidden end-of-cell marks (`w:hideMark`) read, written and laid out like Word (9 cases measured in Word for Windows): a cell's empty last paragraph takes no room, so rows of such empty cells shrink to their borders | +| 2026-10-12 | Text frames (`w:framePr`) in headers, footers, cells and text boxes: placed apart with text wrapping around them, and kept on save (a framed footer page number took a line of its own and was lost on save) | +| 2026-10-12 | List indents like Word's when a numbered style has indents of its own: the style giving the list (and styles based on it) beat the level's indents, and a paragraph with its list switched off drops the styles' indents (27 cases measured in Word for Windows); NIH's Heading 3s now sit at their 0.9" | +| 2026-10-12 | Keep with next like Word (20 cases measured in Word for Windows): a kept paragraph moves to the next page only when it would fit on this one by itself, otherwise it breaks as usual; chains of kept paragraphs move together | +| 2026-10-12 | An empty paragraph that only carries a continuous section break takes no room, as in Word (10 cases measured in Word for Windows): text after it follows right on and balanced columns don't count it | +| 2026-10-12 | Floating objects positioned relative to their paragraph start above its space before, as in Word (10 cases measured in Word for Windows) | +| 2026-10-12 | Unequal columns: text after a column break is broken for the column it lands in (NIH's SARP form had a paragraph a word or two wide) | +| 2026-10-12 | Space before at the top of a page like Word after page breaks (32 cases measured): a paragraph ending with a page break keeps its mark with it, the next paragraph gets no space before; page break before drops it in Word 2013 mode | +| 2026-10-12 | Automatic paragraph spacing drops between list items, as in Word (16 cases measured in Word for Windows): bulleted items that ask for automatic space before and after now touch, as in NIH's monitoring plan | +| 2026-10-12 | Rendering like Word for Windows, found by comparing real .gov documents page by page: text is kerned only where Kerning for fonts asks for it (line breaks now match Word's), and automatic paragraph spacing (`w:beforeAutospacing`/`w:afterAutospacing`, common in pasted web text) is 14 pt, dropped at the document's and cells' edges | +| 2026-10-12 | Keyboard context menu (#821): Shift+F10 opens the page's right-click menu below the caret; `ui.contextMenu` | | 2026-10-11 | Hide White Space (#645): `view.hideWhiteSpace` (View › Show checkbox, or double-click the gap between pages) collapses margins, headers and footers between pages to a thin line in Print Layout; screen only, page count unchanged | | 2026-10-11 | Fonts: Palatino Linotype and Book Antiqua fall back to Palatino (macOS) or TeX Gyre Pagella / P052 / URW Palladio (Linux) before generic serif (#670) | |---|---| +| 2026-10-12 | AutoCorrect capitals (#544): Correct TWo INitial CApitals (with exceptions), accidental cAPS LOCK, first letter of table cells; each undoable and switchable | +| 2026-10-11 | Turkish (`tr`) interface completed (#774): 2,325 entries, every key of the Russian catalog (equation gallery, built-in style names, watermarks, paper sizes, preview tiles); the interface audit checks Turkish too | +| 2026-10-12 | Navigation headings truncate within the resizable pane, with an ellipsis for long titles and full text on hover (#759) | +| 2026-10-11 | Reading Mode: existing View ribbon button now opens a read-only, viewport-column layout with animated horizontal paging, transient display controls, heading navigation (Navigation pane) and resize anchoring; document layout and export unchanged | +| 2026-10-11 | DOCX open/save keeps inherited complex-script size, bold and italic independent of Latin formatting; explicit values are retained, and font commands still format both scripts. Covered by synthetic round trips and a local GUI save/reopen check. | +| 2026-10-11 | Czech (`cs`) interface catalog: 2,021 entries (the full zh-hans key set), Czech locale detection and persisted language choice | +| 2026-10-12 | Portuguese proofing: pt-BR spelling against the installed Hunspell `pt_BR` dictionary (VERO; read at runtime, not bundled) — the affix engine grew `FLAG UTF-8`, 128 flag bits, continuation classes and `FORBIDDENWORD` to read it — suggestions tuned for Portuguese (missing accents first), conservative grammar rules with Portuguese messages, TeX Portuguese hyphenation patterns (BSD 3-clause, attributed); Review › Language offers Português (Brasil); a pt-br interface writes new documents in pt-BR | +| 2026-10-11 | Polish: interface (`pl`, the thirteenth language, every catalog key) and proofing — spelling against the installed Hunspell `pl_PL` dictionary (read at runtime, not bundled) through a new Hunspell-style affix engine, suggestions tuned for Polish, Polish grammar rules with Polish messages, TeX Polish hyphenation patterns (MIT option); proofing language per run (`w:lang`) for squiggles, F7, suggestions and hyphenation, languages without tools left unmarked; Review › Language menu and the status bar's language; a Polish interface writes new documents in Polish | +| 2026-10-11 | German (`de`) interface (#574): 2,021 translated entries covering the full zh-hans key set, locale and saved-preference tests, bundled-font checks; no German proofing resources | +| 2026-10-12 | Sorting and table numbers follow the text's language (#836, #837): Table › Sort, Home › Sort, the index (with its letter headings), the bibliography and the Table of Authorities use the language's alphabet (Polish `Łódź` between `Kraków` and `Zakopane`, Swedish `å ä ö` after `z`, German `ä` with `a`); table sort and formulas read `1,5` and `1 250,00` as Polish decimals and `=SUM` writes `5,75` | +| 2026-10-11 | `.doc` import keeps pictures and drawings (#551): inline and floating pictures (PNG, JPEG, DIB, EMF, WMF), shapes, lines, groups and text boxes with position and wrap, from [MS-ODRAW] | +| 2026-10-11 | Tables resize with the mouse (#217): drag a column border (the neighbour makes up the difference; Shift moves one column), a row's bottom border, or the handle past the bottom-right corner to scale the table; a dashed guide while dragging, one undo step per drag; `table.resizeDrag` for agents | | 2026-10-11 | Animation effects setting: File › Options › Accessibility › Animation effects (Use system setting, On, Off). With the system's animations off (Windows' Animation effects, macOS' Reduce motion, the Linux/BSD desktop's setting, a browser's reduced motion) the page no longer smooth-scrolls: wheel notches and Ctrl+wheel zoom land at once, in the document and in every list and pane, and menus and buttons don't fade; `ui.animations` for agents | | 2026-10-11 | Linux/BSD: Open, Save As, Export and Insert Picture work without a file chooser portal or zenity (labwc, sway…): WordCraft shows its own file browser and says how to get the system dialog; File › Open › Browse in WordCraft… is always there (#576) | | 2026-10-11 | Every measurement follows File › Options › Advanced › Measurement units (#755): the Line Numbers "From text" distance and the Hyphenation zone, still fixed in inches after #485, now convert like every other dialog | @@ -176,17 +222,23 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm | 2026-10-11 | Move tracking (#400): `w:moveFrom`/`w:moveTo` and their named ranges kept in DOCX, a tracked cut and paste of a paragraph or sentence records a move, accept/reject at either end resolves both, green double strikethrough/underline and "Moved from/to" balloons | | 2026-10-11 | Footnote and Endnote dialog from the References › Footnotes launcher: options per section or document (below text, end of section, number format, start at, restart each section or page), custom marks, convert footnotes ↔ endnotes; kept in DOCX (#385) | | 2026-10-11 | Shape Format › Text: text direction, Align Text and linked text boxes (Create Link / Break Link), read and written in DOCX (#369) | +| 2026-10-11 | Quick Access Toolbar KeyTips in the customized button order (Alt/F10 then a number), disabled-command checks and a saved File › Options switch (`ui.keytips`), #620 | | 2026-10-11 | Mail-merge recipients from Excel (.xlsx/.xlsm) and OpenDocument (.ods) spreadsheets, with Select Table to pick the sheet (#334) | | 2026-10-11 | Insert › Icons (#309): a library of 61 original icons drawn in code, a searchable picker and Icon Color | | 2026-10-11 | Home › Font › Enclose Characters (#297): circle, square, triangle or diamond around a character, saved as Word's `EQ \o\ac` field | | 2026-10-11 | Phonetic Guide (#288): ruby text over base text, `w:ruby` round trip | | 2026-10-12 | Custom XML backing stores preserved as opaque data and related properties through edited DOCX saves; content-control markup is handled by the existing reader/writer and #532 bibliography I/O remains separate; binding evaluation and readiness unchanged | +| 2026-10-12 | Imported TOCs keep actual result boundaries, nested page fields, formatting and locks through DOCX saves without relying on TOC styles; existing engine-managed styled TOCs remain refreshable | | 2026-10-11 | Customize Ribbon and the Quick Access Toolbar: custom tabs and groups, hide and reorder tabs, add or remove toolbar commands (right-click any button, or the toolbar's … menu); kept between runs (#379, #376) | | 2026-10-11 | Customize Keyboard: assign, remove and reset keyboard shortcuts; custom keys override built-ins and are kept between runs (#368) | | 2026-10-12 | Native saves use exclusively created sibling temporary files, synchronize before replacement and clean up on failure; existing temporary-name files/links and original files on failed saves are protected. Synthetic CLI and focused tests on macOS; no new Word parity or readiness claim. | | 2026-10-11 | Field codes like Word's field keys (#339): show codes (Alt+F9, Shift+F9), edit, unlink, lock and step through fields | | 2026-10-11 | Word's common keyboard shortcuts (#330): 101 → 125 of ~250 | +| 2026-10-12 | Bulgarian (`bg`) interface: 2,419 catalog entries (every key of the reference catalog, the equation gallery and everything the interface audit finds), `bg-BG` system locales and a saved language choice; locale, catalog and bundled-font tests; no Bulgarian proofing resources | +| 2026-10-11 | Hungarian (`hu`) interface: all 2,419 catalog entries, built-in style names translated, `hu-HU` system locales and a saved language choice; locale, catalog and bundled-font tests; no Hungarian proofing resources | | 2026-10-11 | Turkish (`tr`) interface: all 1,307 catalog entries, `tr-TR`/`tr-CY` system locales and a saved language choice; command search keeps Turkish dotted and dotless I apart (`italik` finds İtalik); locale, catalog, search and bundled-font tests; no Turkish proofing resources | +| 2026-10-11 | Italian (`it`) interface (#594): 2,399 entries, feature names as Italian Word users know them, locale and saved-preference tests, bundled-font checks; no Italian proofing resources | +| 2026-10-12 | Native drawing-level Word ink preserved through DOCX edits and repeated saves, including its InkML part and optional VML picture fallback; native ink rendering and editing remain missing | | 2026-10-11 | Citations and bibliography in all 12 of Word's styles: Harvard, ISO 690 (author-date and numerical), Turabian, GB/T 7714, GOST (name and title sort) and SIST02 join APA, MLA, Chicago and IEEE; numbered styles count in order of first citation (#383) | | 2026-10-11 | Interface size setting (#475): File › Options › General scales the interface 80–200 % on top of the system display scale, kept between runs; `ui.interfaceSize` for agents | | 2026-10-11 | DOCX paragraph defaults: preserve omission separately from explicit empty defaults, with repeated-save and default-edit regressions | @@ -196,7 +248,7 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm | 2026-10-10 | Draw tab: pen, pencil and highlighter ink, stroke eraser, Select, Review › Hide Ink; ink saved to DOCX as DrawingML freeforms (#307); catalog 399/431 (92.6%) | | 2026-10-10 | Draw Table and Eraser (#303): a pen draws one-cell tables and splits cells along drawn lines; the eraser merges the cells beside a border. Catalog 393/431 (91.2%) | | 2026-10-10 | DOCX charts, SmartArt and OLE objects survive open and save with their parts (#319) | -| 2026-10-10 | Full re-measure against Word for Mac 16.113.4 and progress docs to the craftrules standard (this page, `docs/target-app-parity.md`, `gaps.md`, `roadmap.md`, `architecture.md`, format/layout/UI/hardware/localization parity; `docs/parity.md` → `docs/parity-checklist.md`). Landed the same day (~80 PRs): equations with OMML and an Equation tab (#191); Word 97-2003 `.doc` import (#145); right-to-left and Persian text (#207); Zotero integration (#189); Read Aloud player (#190); text boxes and floating objects editable (#46) and placed like Word (#136); floating tables (#137); track-changes fixes (#125, #244); typing parity with Word (#204); hidden text (#173); save prompts and AutoSave rules (#151); LaTeX import/export (#11); keytips and mini-toolbar (#43); Spanish, Ukrainian and Brazilian Portuguese interfaces; system theme (#249); Linux file dialogs no longer freeze (#246); Chinese UI font on Windows (#248); every installed font weight (#239); cell text direction (#245); Paste Special (#235), Advanced Find (#234), Column Selection (#237), Style Inspector (#236), custom table styles (#256), View gridlines (#243), zoom buttons (#232), inertial touchpad scrolling (#252), File Info fields kept (#268), ¶ in mixed-direction paragraphs (#278), Serbian interface (#250), table Height/Width boxes, AutoFit Contents and Table Properties (#44), AutoSave switch greyed out with the reason where it can't save (.md/.txt/.html, unsaved, browser) (#196, #176), OpenGL fallback on Windows PCs without a DirectX 12 driver (#316), Columns, Symbol and Field dialogs (#321), Tabs and Borders and Shading dialogs (#320); free rotation and flips for pictures, shapes and text boxes with a rotation handle (#332); v0.4.0 released | +| 2026-10-10 | Full re-measure against Word for Mac 16.113.4 and progress docs to the craftrules standard (this page, `docs/target-app-parity.md`, `gaps.md`, `roadmap.md`, `architecture.md`, format/layout/UI/hardware/localization parity; `docs/parity.md` → `docs/parity-checklist.md`). Landed the same day (~80 PRs): equations with OMML and an Equation tab (#191); Word 97-2003 `.doc` import (#145); right-to-left and Persian text (#207); Zotero integration (#189); Read Aloud player (#190); text boxes and floating objects editable (#46) and placed like Word (#136); floating tables (#137); track-changes fixes (#125, #244); typing parity with Word (#204); hidden text (#173); save prompts and AutoSave rules (#151); LaTeX import/export (#11); keytips and mini-toolbar (#43); Spanish, Ukrainian and Brazilian Portuguese interfaces; system theme (#249); Linux file dialogs no longer freeze (#246); Chinese UI font on Windows (#248); every installed font weight (#239); cell text direction (#245); Paste Special (#235), Advanced Find (#234), Column Selection (#237), Style Inspector (#236), custom table styles (#256), View gridlines (#243), zoom buttons (#232), inertial touchpad scrolling (#252), File Info fields kept (#268), ¶ in mixed-direction paragraphs (#278), Serbian interface (#250), table Height/Width boxes, AutoFit Contents and Table Properties (#44), AutoSave switch greyed out with the reason where it can't save (.md/.txt/.html, unsaved, browser) (#196, #176), OpenGL fallback on Windows PCs without a DirectX 12 driver (#316), Columns, Symbol and Field dialogs (#321), Tabs and Borders and Shading dialogs (#320); free rotation and flips for pictures, shapes and text boxes with a rotation handle (#332); v0.4.0 released; file associations on Windows and Linux with a Make-default button in Options (#295) | | 2026-10-09 | Interface languages follow the system or Options (#12; zh-hans, zh-hant, ja); TOC page numbers as an updatable field (#52); rotating log file (#17); window geometry remembered (#38) | | 2026-10-08 | v0.2.0 and v0.3.0 released; Flatpak and AppImage auto-update (#22) | | 2026-10-07 | v0.1.0 released; contributor credits in About (#6) | @@ -207,6 +259,24 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm | Date | Change | Summary | |---|---|---| +| 2026-10-11 | trivial | Progress log: the table context menu | +| 2026-10-11 | trivial | Progress log: drags and Shift+arrows over table cells select whole cells | +| 2026-10-11 | trivial | Progress log: table selection areas, insert buttons, move handle and double-click AutoFit; catalog 484/499 | +| 2026-10-12 | trivial | Progress log: Undo history list (#701) | +| 2026-10-12 | trivial | Progress log: Start screen at launch (#766) | +| 2026-10-12 | trivial | Progress log: Alt+Q command search (#819); log table header fixed | +| 2026-10-12 | trivial | Progress log: ribbon groups fit above their labels | +| 2026-10-11 | trivial | Progress log and shipped languages: Turkish interface completed, 2,325 entries (#774) | +| 2026-10-12 | trivial | Progress log: Navigation heading truncation (#759) | +| 2026-10-11 | trivial | Progress log: Reading Mode implementation | +| 2026-10-12 | trivial | Progress log: new documents follow the system region (#838) | +| 2026-10-11 | trivial | DOCX complex-script formatting inheritance fix and round-trip coverage | +| 2026-10-11 | minor | Czech (`cs`) interface catalog added | +| 2026-10-12 | trivial | Progress log: AutoCorrect replacement list follows the text's language (#835) | +| 2026-10-11 | trivial | Progress log: German interface (#574) | +| 2026-10-12 | trivial | Progress log: language-aware sorting and table numbers (#836, #837) | +| 2026-10-12 | trivial | Progress log: Word 365 cover pages (#743) | +| 2026-10-11 | trivial | Progress log: tables resize with the mouse (#217) | | 2026-10-11 | trivial | Progress log: Animation effects setting | | 2026-10-11 | trivial | Progress log: built-in file browser when the system file dialog is unavailable (#576) | | 2026-10-11 | trivial | Progress log: security review fixes (temp files, exported link schemes) | @@ -230,17 +300,22 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm | 2026-10-11 | trivial | Progress log: document grid (#391) | | 2026-10-11 | trivial | Progress log: text direction, Align Text and linked text boxes (#369) | | 2026-10-10 | trivial | Progress: Drawing Canvas (#344): insert, shapes/pictures/text boxes inside, Fit/Expand/Scale, DOCX `wpc:wpc`, turns and flips as a whole; catalog 404/434 | +| 2026-10-11 | trivial | Quick Access Toolbar KeyTips and persisted enable/disable setting (#620) | | 2026-10-11 | trivial | Progress log: mail-merge recipients from .xlsx/.xlsm/.ods (#334) | | 2026-10-11 | trivial | Insert › Icons landed (#309): catalog 402/434 → 403/434 (92.9%) | | 2026-10-11 | trivial | Enclose Characters (#297): catalog 403/434 (92.9%) | | 2026-10-12 | minor | Progress log: opaque custom XML store preservation, with synthetic package regressions and explicit limits | +| 2026-10-12 | minor | Preserve imported TOC ranges and retain #538 refresh/undo/save coverage on current main; readiness unchanged | | 2026-10-11 | trivial | Progress log: Customize Ribbon and the Quick Access Toolbar (#379) | | 2026-10-11 | trivial | Omitted `w:pPrDefault` stays omitted across saves; explicit empty/default properties and later spacing edits remain writable. Synthetic preservation checks only; readiness estimates unchanged | | 2026-10-11 | trivial | Progress log: Customize Keyboard (#368) | | 2026-10-12 | trivial | Native saves protect existing temporary-name entries and originals on pre-replacement failure; synthetic macOS evidence, explicit durability/platform limits; readiness unchanged | | 2026-10-11 | trivial | Word's common keyboard shortcuts (#330): 125 of ~250 | | 2026-10-11 | trivial | Collapsed duplicate rows left by parallel merges | +| 2026-10-12 | trivial | Bulgarian interface (`bg`): progress log and shipped languages | +| 2026-10-11 | trivial | Hungarian interface (`hu`): progress log and shipped languages | | 2026-10-11 | trivial | Turkish interface (`tr`): progress log and shipped languages | +| 2026-10-12 | minor | Native drawing-level ink and its picture fallback survive DOCX edits and repeated saves; adapted to the current graphic writer; rendering and editing InkML remain missing | | 2026-10-11 | trivial | Progress log: eight more bibliography styles (#383) | | 2026-10-10 | trivial | Progress log: Draw tab ink (#307) | | 2026-10-10 | trivial | Draw Table and Eraser (#303): catalog 393/431 (91.2%) | @@ -250,6 +325,7 @@ Ranked, to beta (detail and remaining milestones in [docs/roadmap.md](docs/roadm | 2026-10-10 | trivial | Progress log: Windows OpenGL fallback without a DirectX 12 driver (#316) | | 2026-10-10 | minor | Password-protected .docx (#281): files encrypted with a password (agile encryption) open and save; File › Info › Encrypt with Password | | 2026-10-10 | minor | Relanded clipboard pane, table styles editor, Manage Styles, Asian typography (kinsoku), Group/Ungroup and Shape Effects (#264 #263 #270 #271 #267 #280): catalog 390/431 (90.5%) | +| 2026-10-10 | minor | Polish interface and proofing landed: Localization 25% → 27%, Proofing 35% → 40% | | 2026-10-10 | minor | Merged main: catalog 384/430 (89.3%), 431 commands; Paste Special, Advanced Find, Column Selection, Style Inspector, custom table styles, gridlines, zoom buttons, touchpad scrolling, File Info, Serbian landed; mainstream 54→55% | | 2026-10-10 | minor | Readiness table with hours per audience; full number restored to the additive weighted sum (~47% → ~60%): method aligned with the standard, no new evidence; beta distance back to ~15 points and ~150–230 h | | 2026-10-10 | minor | Mainstream practitioner (~54%) and essentials user (~63%) numbers added; full number recomputed 60% → ~47% with the standard's discounts; beta distance now ~28 points and ~450–700 h; stage stays alpha | diff --git a/apps/wordcraft-cli/src/automation_args.rs b/apps/wordcraft-cli/src/automation_args.rs new file mode 100644 index 000000000..60fa0e30a --- /dev/null +++ b/apps/wordcraft-cli/src/automation_args.rs @@ -0,0 +1,90 @@ +//! Parse `--automation-read-root` / `--automation-write-root` before the server starts. A malformed +//! flag must not silently remove a root and leave the server unconfined. + +use std::path::PathBuf; + +pub struct AutomationArgs { + pub read_root: Option, + pub write_root: Option, + read_given: bool, + write_given: bool, +} + +impl AutomationArgs { + pub fn new(read_root: Option, write_root: Option) -> Self { + Self { read_root, write_root, read_given: false, write_given: false } + } + + /// Returns whether this argument is a root flag. A second occurrence of a flag is an error, + /// including when the two spellings (`--flag DIR`, `--flag=DIR`) are mixed. + pub fn consume(&mut self, arg: &str, args: &mut impl Iterator) -> Result { + let (flag, inline) = arg.split_once('=').map_or((arg, None), |(flag, value)| (flag, Some(value))); + let (root, given) = match flag { + "--automation-read-root" => (&mut self.read_root, &mut self.read_given), + "--automation-write-root" => (&mut self.write_root, &mut self.write_given), + _ => return Ok(false), + }; + if *given { + return Err(format!("{flag} given twice")); + } + let value = match inline { + Some(value) => value.to_owned(), + None => { + let value = args.next().ok_or_else(|| format!("{flag} needs a folder"))?; + if value.starts_with('-') { + return Err(format!("{flag} needs a folder; use {flag}= for a folder starting with '-'")); + } + value + } + }; + if value.is_empty() { + return Err(format!("{flag} needs a non-empty folder")); + } + *root = Some(value.into()); + *given = true; + Ok(true) + } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn roots_are_taken_once_and_other_arguments_are_left() { + let mut roots = AutomationArgs::new(None, None); + let mut args = ["a folder".to_owned(), "report.docx".to_owned()].into_iter(); + assert!(roots.consume("--automation-read-root", &mut args).unwrap()); + assert!(roots.consume("--automation-write-root=out", &mut args).unwrap()); + assert!(!roots.consume("--author", &mut args).unwrap()); + assert_eq!(roots.read_root, Some("a folder".into())); + assert_eq!(roots.write_root, Some("out".into())); + assert_eq!(args.next().as_deref(), Some("report.docx")); + assert!(roots.consume("--automation-read-root=other", &mut args).is_err()); + assert_eq!(roots.read_root, Some("a folder".into())); + } + + #[test] + fn invalid_operands_leave_roots_unchanged() { + for flag in ["--automation-read-root", "--automation-write-root"] { + for operand in [None, Some(""), Some("--version"), Some("--connect"), Some("-folder")] { + let mut roots = AutomationArgs::new(Some("read".into()), Some("write".into())); + let mut args = operand.map(str::to_owned).into_iter(); + assert!(roots.consume(flag, &mut args).is_err(), "{flag} {operand:?}"); + assert_eq!(roots.read_root, Some("read".into())); + assert_eq!(roots.write_root, Some("write".into())); + } + let mut roots = AutomationArgs::new(None, None); + assert!(roots.consume(&format!("{flag}="), &mut std::iter::empty()).is_err()); + } + } + + #[test] + fn explicit_dash_folder_and_relative_dash_folder_are_valid() { + let mut roots = AutomationArgs::new(None, None); + assert!(roots.consume("--automation-read-root=-folder", &mut std::iter::empty()).unwrap()); + assert!(roots.consume("--automation-write-root", &mut ["./-folder".to_owned()].into_iter()).unwrap()); + assert_eq!(roots.read_root, Some("-folder".into())); + assert_eq!(roots.write_root, Some("./-folder".into())); + } +} diff --git a/apps/wordcraft-cli/src/main.rs b/apps/wordcraft-cli/src/main.rs index d3355ec5e..a0a2d5f23 100644 --- a/apps/wordcraft-cli/src/main.rs +++ b/apps/wordcraft-cli/src/main.rs @@ -1,6 +1,8 @@ //! `wordcraft-cli`: WordCraft from the command line. #![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)] +mod automation_args; + use std::fmt; use std::io::Write; use std::process::ExitCode; @@ -25,8 +27,14 @@ USAGE: who tracked changes and comments are by) wordcraft-cli commands [--json] list every command wordcraft-cli parity [--markdown] feature-catalog parity - wordcraft-cli mcp [--connect HOST:PORT | --author NAME] - MCP server on stdio (headless, or bridged to the app) + wordcraft-cli mcp [--connect HOST:PORT | --author NAME] [--automation-read-root DIR] [--automation-write-root DIR] + MCP server on stdio (headless, or bridged to the app). + --automation-read-root / --automation-write-root confine + the files the agent's commands read (open, insert text + or picture, mail-merge data, templates…) and write (save, + export, merge output…) to DIR, links followed; a root left + out grants none of its access. They apply to the headless + server and are an error with --connect. wordcraft-cli zotero [--cmd 'id={json}' …] [--save OUT] [--trace] [--port P] run a Zotero command on a document (Zotero must be running): addEditCitation, addEditBibliography, @@ -76,6 +84,26 @@ fn validate_options(command: &str, args: &[String], specs: &[OptionSpec]) -> Res Ok(()) } +/// Take `--automation-read-root` / `--automation-write-root` (each `--flag DIR` or `--flag=DIR`, +/// once) out of `args`, and open the roots, which must be existing folders. Returns the arguments +/// that are left. A malformed root flag is an error, never a root silently left out. +fn automation_roots( + command: &str, + args: &[String], +) -> Result<(Option>, Vec), String> { + let mut found = automation_args::AutomationArgs::new(None, None); + let mut left = Vec::new(); + let mut it = args.iter().cloned(); + while let Some(arg) = it.next() { + if !found.consume(&arg, &mut it).map_err(|e| format!("{command}: {e}"))? { + left.push(arg); + } + } + let roots = wordcraft_engine::file_access::AutomationRoots::new(found.read_root.as_deref(), found.write_root.as_deref()) + .map_err(|e| format!("{command}: {e}"))?; + Ok((roots, left)) +} + const NO_OPTIONS: &[OptionSpec] = &[]; const CONVERT_OPTIONS: &[OptionSpec] = &[OptionSpec { name: "--markup", takes_value: false }]; const RENDER_OPTIONS: &[OptionSpec] = &[OptionSpec { name: "--page", takes_value: true }, OptionSpec { name: "--scale", takes_value: true }]; @@ -265,9 +293,17 @@ fn run(args: &[String]) -> Result<(), Failure> { } } "mcp" => { + let (roots, rest) = automation_roots("mcp", &rest)?; validate_options("mcp", &rest, MCP_OPTIONS)?; let author = arg_value(&rest, "--author"); - let backend: Box = match arg_value(&rest, "--connect") { + let connect = arg_value(&rest, "--connect"); + if roots.is_some() && connect.is_some() { + return Err( + "mcp: --automation-read-root and --automation-write-root confine the headless server; a running app is not confined by them" + .into(), + ); + } + let backend: Box = match connect { // The app's user name is the person's own setting; don't rewrite it from a flag. Some(_) if author.is_some() => { return Err("mcp: --author applies to the headless server; with --connect, run file.setAuthor".into()); @@ -281,6 +317,13 @@ fn run(args: &[String]) -> Result<(), Failure> { } None => { let mut headless = wordcraft_mcp::Headless::default(); + if let Some(roots) = roots { + // Every thread of this process, not only the server's: nothing reaches past the roots. + wordcraft_engine::file_access::confine_process(roots.clone()); + let show = |root: Option<&std::path::Path>| root.map_or_else(|| "none".to_string(), |r| r.display().to_string()); + eprintln!("wordcraft-cli: files confined: read root {}, write root {}", show(roots.read_root()), show(roots.write_root())); + headless = headless.with_automation_roots(Some(roots)); + } if let Some(name) = author { set_author(&mut headless.session, &name)?; } @@ -428,4 +471,58 @@ mod tests { // The app's user name is the person's own; a bridge flag must not rewrite it. assert!(run(&args(&["mcp", "--connect", "127.0.0.1:9", "--author", "Claude"])).is_err_and(|e| e.to_string().contains("--author"))); } + + #[test] + fn automation_root_flags_are_taken_out_of_the_mcp_arguments() { + let base = std::env::temp_dir().join(format!("wc-cli-roots-{}", std::process::id())); + let _ = std::fs::remove_dir_all(&base); + std::fs::create_dir_all(base.join("in")).unwrap(); + std::fs::create_dir_all(base.join("out")).unwrap(); + let (read, write) = (base.join("in").to_string_lossy().into_owned(), base.join("out").to_string_lossy().into_owned()); + + let (roots, left) = + automation_roots("mcp", &args(&["--automation-read-root", &read, "--author", "Claude", &format!("--automation-write-root={write}")])) + .unwrap(); + let roots = roots.unwrap(); + assert_eq!(left, args(&["--author", "Claude"])); + assert!(roots.read_root().is_some() && roots.write_root().is_some()); + assert!(validate_options("mcp", &left, MCP_OPTIONS).is_ok()); + + // A root left out grants none of its access. + let (roots, _) = automation_roots("mcp", &args(&["--automation-read-root", &read])).unwrap(); + let roots = roots.unwrap(); + assert!(roots.write_root().is_none()); + assert!( + roots.check(&format!("{write}/x.docx"), wordcraft_engine::file_access::Access::Write).unwrap_err().contains("write authority is absent") + ); + + // No flags: nothing is confined and the arguments pass through. + let (roots, left) = automation_roots("mcp", &args(&["--connect", "127.0.0.1:9"])).unwrap(); + assert!(roots.is_none()); + assert_eq!(left, args(&["--connect", "127.0.0.1:9"])); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn automation_root_flags_reject_bad_operands_and_missing_folders() { + for bad in [ + args(&["--automation-read-root"]), + args(&["--automation-read-root", "--author"]), + args(&["--automation-write-root="]), + args(&["--automation-read-root", "a", "--automation-read-root=b"]), + ] { + assert!(automation_roots("mcp", &bad).is_err(), "{bad:?}"); + } + let e = automation_roots("mcp", &args(&["--automation-read-root", "/definitely/not/a/folder"])).err().unwrap(); + assert!(e.starts_with("mcp: cannot open the automation read root"), "{e}"); + // Through the command line: an error, not a server that quietly runs unconfined. + assert!(run(&args(&["mcp", "--automation-write-root", "/definitely/not/a/folder"])).is_err()); + } + + #[test] + fn automation_roots_are_an_error_with_connect() { + let dir = std::env::temp_dir().to_string_lossy().into_owned(); + let e = run(&args(&["mcp", "--connect", "127.0.0.1:9", "--automation-read-root", &dir])).unwrap_err().to_string(); + assert!(e.contains("--automation-read-root") && e.contains("headless"), "{e}"); + } } diff --git a/apps/wordcraft-web/src/web.rs b/apps/wordcraft-web/src/web.rs index bc5e47987..30ab58ef6 100644 --- a/apps/wordcraft-web/src/web.rs +++ b/apps/wordcraft-web/src/web.rs @@ -1,7 +1,7 @@ //! The browser shell: web `Services`, drag-and-drop, the unsaved-changes guard and the eframe //! web runner. -use std::cell::Cell; +use std::cell::{Cell, RefCell}; use std::rc::Rc; use wasm_bindgen::JsCast as _; @@ -27,7 +27,8 @@ pub fn start() { return; }; let dirty = Rc::new(Cell::new(false)); - if let Err(e) = guard_unload(dirty.clone()) { + let saves = Rc::new(PickerSaves::default()); + if let Err(e) = guard_unload(dirty.clone(), saves.clone()) { log::error!("no unsaved-changes guard: {e}"); } let mut options = eframe::WebOptions::default(); @@ -50,10 +51,20 @@ pub fn start() { if let Err(e) = listen_for_pasted_pictures(inbox.clone(), cc.egui_ctx.clone()) { log::error!("pasting pictures is unavailable: {e}"); } - let doc = if query().contains("sample") { wordcraft_engine::sample::sample_document() } else { wordcraft_doc::Document::new() }; - let mut app = WordApp::new(Session::new(doc), services(inbox.clone(), cc.egui_ctx.clone(), dirty)); + // The browser's locale picks the new document's paper, the units and the + // decimal separator (#838). + let region = wordcraft_ui_egui::i18n::system_region(); + let sample = query().contains("sample"); + let doc = if sample { wordcraft_engine::sample::sample_document() } else { wordcraft_engine::blank_document(®ion) }; + let mut session = Session::new(doc); + session.system_region = region; + let mut app = WordApp::new(session, services(inbox.clone(), cc.egui_ctx.clone(), dirty, saves.clone())); app.autosave = false; - Ok(Box::new(WebShell { app, inbox })) + // No document to show: start on File › Home, as the desktop app does (#766). + if !sample { + app.start_screen(); + } + Ok(Box::new(WebShell { app, inbox, saves })) }), ) .await; @@ -75,6 +86,38 @@ fn query() -> String { struct WebShell { app: WordApp, inbox: Inbox, + saves: Rc, +} + +/// Saves the browser's save picker finishes after `download` has returned (#728). The app marks +/// the document saved as soon as `download` returns, so a cancelled or failed picker save must +/// mark it unsaved again, and the `beforeunload` guard keeps warning while one is in flight. +#[derive(Default)] +struct PickerSaves { + /// Picker saves started and not yet written and closed. + pending: Cell, + /// Picker saves that didn't write the file: (name, error, or `None` when cancelled). + unsaved: RefCell)>>, +} + +impl WebShell { + /// A picker save that didn't happen leaves the document with unsaved changes. + fn settle_picker_saves(&mut self) { + let unsaved = std::mem::take(&mut *self.saves.unsaved.borrow_mut()); + for (name, error) in unsaved { + // Only a save in a format that keeps everything marked the document saved. + if wordcraft_ui_egui::keeps_everything(&name) { + self.app.session.dirty = true; + } + if let Some(error) = error { + let msg = wordcraft_ui_egui::i18n::fmt( + wordcraft_ui_egui::i18n::t("Couldn't download {name}: {error}"), + &[("name", &name), ("error", &error)], + ); + self.app.status(msg); + } + } + } } impl eframe::App for WebShell { @@ -94,6 +137,7 @@ impl eframe::App for WebShell { } }); } + self.settle_picker_saves(); self.app.logic(ctx); } @@ -153,8 +197,9 @@ fn listen_for_pasted_pictures(inbox: Inbox, ctx: egui::Context) -> Result<(), St } /// `dirty` is the flag the `beforeunload` guard reads ([`guard_unload`]). -fn services(inbox: Inbox, ctx: egui::Context, dirty: Rc>) -> Services { +fn services(inbox: Inbox, ctx: egui::Context, dirty: Rc>, saves: Rc) -> Services { let open_inbox = inbox.clone(); + let save_ctx = ctx.clone(); Services { open_async: Some(Box::new(move |purpose: &str| { let inbox = open_inbox.clone(); @@ -179,7 +224,7 @@ fn services(inbox: Inbox, ctx: egui::Context, dirty: Rc>) -> Services })), // Exports ask for a name; the browser decides where the download goes. pick_save: Some(Box::new(|name: &str| Some(name.to_string()))), - download: Some(Box::new(download)), + download: Some(Box::new(move |path: &str, bytes: &[u8]| download(&saves, &save_ctx, path, bytes))), print: Some(Box::new(print_pdf)), inbox: Some(inbox), on_dirty: Some(Box::new(move |d| dirty.set(d))), @@ -200,10 +245,11 @@ fn animations_enabled() -> Box Option> { /// Closing, reloading or leaving the tab with unsaved changes asks first. Browsers show their own /// confirmation (a page can only ask for it), so there is no Save button: Cancel, then File › Save. -fn guard_unload(dirty: Rc>) -> Result<(), String> { +fn guard_unload(dirty: Rc>, saves: Rc) -> Result<(), String> { let window = web_sys::window().ok_or("no window")?; let on_unload = wasm_bindgen::closure::Closure::::new(move |e: web_sys::BeforeUnloadEvent| { - if dirty.get() { + // A save still in the browser's picker hasn't written anything yet. + if dirty.get() || saves.pending.get() > 0 { e.prevent_default(); // Older browsers ask only when `returnValue` is set. e.set_return_value("unsaved changes"); @@ -215,20 +261,145 @@ fn guard_unload(dirty: Rc>) -> Result<(), String> { Ok(()) } -/// Trigger a browser download of `bytes` named after the last component of `path`. -fn download(path: &str, bytes: &[u8]) -> Result<(), String> { - let js = |e: wasm_bindgen::JsValue| format!("{e:?}"); +/// Trigger a browser save dialog or download of `bytes` named after the last component of `path`. +fn download(saves: &Rc, ctx: &egui::Context, path: &str, bytes: &[u8]) -> Result<(), String> { let name = std::path::Path::new(path).file_name().map(|n| n.to_string_lossy().to_string()).unwrap_or_else(|| "document".into()); let window = web_sys::window().ok_or("no window")?; + + // On browsers supporting the File System Access API (Chromium: Chrome, Edge, etc.), + // `showSaveFilePicker` lets the user choose the save location and filename natively. + if let Ok(picker_val) = js_sys::Reflect::get(&window, &wasm_bindgen::JsValue::from_str("showSaveFilePicker")) + && let Ok(picker_fn) = picker_val.dyn_into::() + { + match trigger_save_file_picker(saves, ctx, &window, &picker_fn, &name, bytes) { + Ok(()) => return Ok(()), + Err(e) => { + log::warn!("showSaveFilePicker failed: {e}; falling back to anchor download"); + } + } + } + + // Fallback for browsers without showSaveFilePicker (e.g. Firefox, Safari): + trigger_anchor_download(&window, &name, bytes) +} + +fn trigger_save_file_picker( + saves: &Rc, + ctx: &egui::Context, + window: &web_sys::Window, + picker_fn: &js_sys::Function, + name: &str, + bytes: &[u8], +) -> Result<(), String> { + let js = |e: wasm_bindgen::JsValue| format!("{e:?}"); + let ext = std::path::Path::new(name).extension().and_then(|e| e.to_str()).unwrap_or("docx").to_ascii_lowercase(); + let dot_ext = format!(".{ext}"); + let mime = mime_for(name); + + let accept_obj = js_sys::Object::new(); + let exts_arr = js_sys::Array::new(); + exts_arr.push(&wasm_bindgen::JsValue::from_str(&dot_ext)); + let _ = js_sys::Reflect::set(&accept_obj, &wasm_bindgen::JsValue::from_str(mime), &exts_arr); + + // Our own neutral description: the format's name, as its extension says it. + let type_desc = match ext.as_str() { + "htm" => "HTML".to_string(), + "txt" => "Text".to_string(), + e if !e.is_empty() && e.len() <= 8 && e.chars().all(|c| c.is_ascii_alphanumeric()) => e.to_ascii_uppercase(), + _ => "Document".to_string(), + }; + + let type_obj = js_sys::Object::new(); + let _ = js_sys::Reflect::set(&type_obj, &wasm_bindgen::JsValue::from_str("description"), &wasm_bindgen::JsValue::from_str(&type_desc)); + let _ = js_sys::Reflect::set(&type_obj, &wasm_bindgen::JsValue::from_str("accept"), &accept_obj); + + let types_arr = js_sys::Array::new(); + types_arr.push(&type_obj); + + let picker_opts = js_sys::Object::new(); + let _ = js_sys::Reflect::set(&picker_opts, &wasm_bindgen::JsValue::from_str("suggestedName"), &wasm_bindgen::JsValue::from_str(name)); + let _ = js_sys::Reflect::set(&picker_opts, &wasm_bindgen::JsValue::from_str("types"), &types_arr); + + // Call showSaveFilePicker synchronously while transient user activation is active: + let prom_val = picker_fn.call1(window, &picker_opts).map_err(js)?; + let prom: js_sys::Promise = prom_val.dyn_into().map_err(|_| "not a promise".to_string())?; + + let bytes_vec = bytes.to_vec(); + let saves = saves.clone(); + let ctx = ctx.clone(); + let name = name.to_string(); + saves.pending.set(saves.pending.get().saturating_add(1)); + wasm_bindgen_futures::spawn_local(async move { + // The document counts as saved only once the file is written and closed. + let unsaved = match wasm_bindgen_futures::JsFuture::from(prom).await { + Ok(handle_val) => match write_file_handle(&handle_val, &bytes_vec).await { + Ok(()) => { + log::info!("saved {name} through the save picker"); + None + } + Err(e) => { + log::error!("couldn't write {name}: {e}"); + Some(Some(e)) + } + }, + Err(e) => { + let err_name = + js_sys::Reflect::get(&e, &wasm_bindgen::JsValue::from_str("name")).ok().and_then(|v| v.as_string()).unwrap_or_default(); + if err_name == "AbortError" { + log::debug!("save picker cancelled"); + Some(None) + } else { + log::error!("showSaveFilePicker error: {e:?}"); + Some(Some(format!("{e:?}"))) + } + } + }; + if let Some(error) = unsaved { + saves.unsaved.borrow_mut().push((name, error)); + } + saves.pending.set(saves.pending.get().saturating_sub(1)); + ctx.request_repaint(); + }); + Ok(()) +} + +async fn write_file_handle(handle: &wasm_bindgen::JsValue, bytes: &[u8]) -> Result<(), String> { + let js = |e: wasm_bindgen::JsValue| format!("{e:?}"); + let create_fn: js_sys::Function = js_sys::Reflect::get(handle, &wasm_bindgen::JsValue::from_str("createWritable")) + .map_err(js)? + .dyn_into() + .map_err(|_| "createWritable is not a function".to_string())?; + let stream_prom: js_sys::Promise = create_fn.call0(handle).map_err(js)?.dyn_into().map_err(|_| "not a promise".to_string())?; + let stream = wasm_bindgen_futures::JsFuture::from(stream_prom).await.map_err(js)?; + + let write_fn: js_sys::Function = js_sys::Reflect::get(&stream, &wasm_bindgen::JsValue::from_str("write")) + .map_err(js)? + .dyn_into() + .map_err(|_| "write is not a function".to_string())?; + let uint8 = js_sys::Uint8Array::from(bytes); + let write_prom: js_sys::Promise = write_fn.call1(&stream, &uint8).map_err(js)?.dyn_into().map_err(|_| "not a promise".to_string())?; + wasm_bindgen_futures::JsFuture::from(write_prom).await.map_err(js)?; + + let close_fn: js_sys::Function = js_sys::Reflect::get(&stream, &wasm_bindgen::JsValue::from_str("close")) + .map_err(js)? + .dyn_into() + .map_err(|_| "close is not a function".to_string())?; + let close_prom: js_sys::Promise = close_fn.call0(&stream).map_err(js)?.dyn_into().map_err(|_| "not a promise".to_string())?; + wasm_bindgen_futures::JsFuture::from(close_prom).await.map_err(js)?; + Ok(()) +} + +fn trigger_anchor_download(window: &web_sys::Window, name: &str, bytes: &[u8]) -> Result<(), String> { + let js = |e: wasm_bindgen::JsValue| format!("{e:?}"); let document = window.document().ok_or("no document")?; let parts = js_sys::Array::of1(&js_sys::Uint8Array::from(bytes)); let opts = web_sys::BlobPropertyBag::new(); - opts.set_type(mime_for(&name)); + opts.set_type(mime_for(name)); let blob = web_sys::Blob::new_with_u8_array_sequence_and_options(&parts, &opts).map_err(js)?; let url = web_sys::Url::create_object_url_with_blob(&blob).map_err(js)?; - let a: web_sys::HtmlAnchorElement = document.create_element("a").map_err(js)?.dyn_into().map_err(|_| "not an anchor")?; + let a: web_sys::HtmlAnchorElement = document.create_element("a").map_err(js)?.dyn_into().map_err(|_| "not an anchor".to_string())?; a.set_href(&url); - a.set_download(&name); + a.set_download(name); a.style().set_property("display", "none").map_err(js)?; let body = document.body().ok_or("no body")?; body.append_child(&a).map_err(js)?; diff --git a/apps/wordcraft/Cargo.toml b/apps/wordcraft/Cargo.toml index 6452ce553..7e79f24bc 100644 --- a/apps/wordcraft/Cargo.toml +++ b/apps/wordcraft/Cargo.toml @@ -13,6 +13,7 @@ workspace = true [dependencies] wordcraft-doc = { workspace = true } wordcraft-engine = { workspace = true } +wordcraft-proof = { workspace = true } # Fonts the user added (#582) load from the settings folder at start-up. wordcraft-fonts = { workspace = true } wordcraft-ui-egui = { workspace = true } @@ -43,16 +44,27 @@ objc2 = "0.6" dispatch2 = "0.3" [target.'cfg(all(unix, not(target_os = "macos")))'.dependencies] +winit = { version = "0.30", default-features = false, features = ["x11"] } # The desktop's animation setting on Linux and the BSDs, read from the XDG settings portal over # D-Bus. Already in the build through AccessKit, at this version; this adds only its blocking API. zbus = { version = "5", default-features = false, features = ["async-io", "blocking-api"] } +# Wayland's clipboard through the data-control protocol (#687, #833): the RTF and other picture +# types that arboard doesn't read, and the formats offered. Already in the build through arboard's +# `wayland-data-control` feature, at this version. +wl-clipboard-rs = "0.9" [target.'cfg(target_os = "windows")'.dependencies] +# The upstream default-apps button opens Windows Settings; printing uses craft-print instead. +open = "5" # The system's Animation effects setting (Settings › Accessibility › Visual effects), read through # the Windows Runtime's UISettings: a safe API, where the Win32 call needs unsafe, which our # workspace forbids. Already in the build through wgpu and AccessKit, at this version; this adds -# only the `UI_ViewManagement` bindings. -windows = { version = "0.62", features = ["UI_ViewManagement"] } +# only the `UI_ViewManagement` bindings, and `System_UserProfile` for the home region new +# documents take their paper and units from (#838). +windows = { version = "0.62", features = ["UI_ViewManagement", "System_UserProfile"] } +# The clipboard's "Rich Text Format" and the names of its formats (#687, #833), which arboard +# doesn't read. Already in the build through arboard, at this version. +clipboard-win = { version = "5.4", features = ["std"] } [target.'cfg(not(target_arch = "wasm32"))'.dependencies] craft-print = { git = "https://github.com/storytold/craft-libs", rev = "d27646a5f7c8b4fc85d9dd3213b8c88b58e0088d" } diff --git a/apps/wordcraft/src/clipboard.rs b/apps/wordcraft/src/clipboard.rs new file mode 100644 index 000000000..ecd284679 --- /dev/null +++ b/apps/wordcraft/src/clipboard.rs @@ -0,0 +1,164 @@ +//! The system clipboard beyond egui's text (#45, #584, #687, #833): text for the ribbon's Paste, +//! another program's formatted copy (HTML, and RTF where the platform offers it), pictures and +//! copied files, and the names of the formats there for a paste that finds nothing it can use. +//! +//! On Linux, arboard talks to Wayland compositors through the data-control protocol (KDE Plasma, +//! wlroots compositors); without it, a native Wayland window only sees what the compositor copies +//! over to Xwayland, which KWin does only while an X11 window has focus — so pictures copied in a +//! browser or Spectacle never arrived (#833). Compositors without the protocol fall back to X11. +//! +//! Everything read here comes from other programs: sizes are capped as it is read. + +use wordcraft_engine::cmd::paste::MAX_PASTE_BYTES; +use wordcraft_ui_egui::paste_picture::ClipboardPicture; + +fn open() -> Option { + arboard::Clipboard::new().map_err(|e| log::warn!("clipboard: {e}")).ok() +} + +/// The clipboard's text. +pub fn text() -> Option { + open()?.get_text().ok() +} + +/// The clipboard's HTML (Windows: the fragment of its "HTML Format"), if not too large. +pub fn html() -> Option { + open()?.get().html().ok().filter(|h| h.len() <= MAX_PASTE_BYTES) +} + +/// The clipboard's RTF: Windows ("Rich Text Format") and Wayland (`text/rtf`); not offered by +/// arboard, so read here. +pub fn rtf() -> Option { + let bytes = platform::rtf()?; + if bytes.len() > MAX_PASTE_BYTES { + return None; + } + // RTF is 7-bit by the spec; stray 8-bit bytes become replacement characters. + let s = String::from_utf8_lossy(&bytes).trim_end_matches('\0').to_string(); + s.trim_start().starts_with("{\\rtf").then_some(s) +} + +/// The names of the formats on the clipboard, where the platform lists them (Wayland, Windows). +pub fn formats() -> Vec { + platform::formats() +} + +/// The picture on the clipboard: copied files, else image pixels, else (Wayland) an encoded +/// picture in a format arboard doesn't decode. A picture that is there but can't be read says +/// why (#833). The UI checks the sizes before inserting anything. +pub fn picture() -> Option { + let mut clipboard = match arboard::Clipboard::new() { + Ok(c) => c, + Err(e) => { + log::warn!("clipboard: {e}"); + return Some(ClipboardPicture::Unreadable(format!("the system clipboard couldn't be opened ({e})"))); + } + }; + if let Ok(files) = clipboard.get().file_list() + && !files.is_empty() + { + return Some(ClipboardPicture::Files(files)); + } + match clipboard.get_image() { + Ok(image) => Some(ClipboardPicture::Pixels { width: image.width, height: image.height, rgba: image.bytes.into_owned() }), + Err(arboard::Error::ContentNotAvailable) => platform::encoded_picture().map(ClipboardPicture::Encoded), + Err(arboard::Error::ConversionFailure) => Some( + platform::encoded_picture() + .map(ClipboardPicture::Encoded) + .unwrap_or_else(|| ClipboardPicture::Unreadable("WordCraft can't decode its format".into())), + ), + Err(e) => { + log::warn!("clipboard picture: {e}"); + platform::encoded_picture().map(ClipboardPicture::Encoded).or(Some(ClipboardPicture::Unreadable(e.to_string()))) + } + } +} + +/// Picture types read straight from Wayland when arboard (PNG only) finds none, best first. +#[cfg(all(unix, not(any(target_os = "macos", target_os = "android", target_os = "emscripten"))))] +const PICTURE_MIMES: [&str; 6] = ["image/png", "image/jpeg", "image/gif", "image/webp", "image/bmp", "image/x-bmp"]; + +#[cfg(all(unix, not(any(target_os = "macos", target_os = "android", target_os = "emscripten"))))] +mod platform { + use std::io::Read; + + use wl_clipboard_rs::paste::{ClipboardType, MimeType, Seat, get_contents, get_mime_types_ordered}; + + fn wayland() -> bool { + std::env::var_os("WAYLAND_DISPLAY").is_some() + } + + /// One MIME type's bytes, at most `cap`; `None` when absent or larger. + fn read(mime: &str, cap: u64) -> Option> { + let (pipe, _) = get_contents(ClipboardType::Regular, Seat::Unspecified, MimeType::Specific(mime)).ok()?; + let mut out = Vec::new(); + pipe.take(cap.saturating_add(1)).read_to_end(&mut out).ok()?; + (out.len() as u64 <= cap).then_some(out) + } + + pub fn formats() -> Vec { + if !wayland() { + return Vec::new(); + } + get_mime_types_ordered(ClipboardType::Regular, Seat::Unspecified).unwrap_or_default() + } + + pub fn rtf() -> Option> { + if !wayland() { + return None; + } + let offered = formats(); + let mime = ["text/rtf", "application/rtf", "text/richtext"].into_iter().find(|m| offered.iter().any(|o| o == m))?; + read(mime, super::MAX_PASTE_BYTES as u64) + } + + pub fn encoded_picture() -> Option> { + if !wayland() { + return None; + } + let offered = formats(); + let mime = super::PICTURE_MIMES.into_iter().find(|m| offered.iter().any(|o| o == m))?; + read(mime, wordcraft_ui_egui::paste_picture::MAX_FILE_BYTES) + } +} + +#[cfg(target_os = "windows")] +mod platform { + use clipboard_win::{Clipboard, EnumFormats, raw}; + + pub fn formats() -> Vec { + let Ok(_open) = Clipboard::new_attempts(10) else { return Vec::new() }; + // Standard formats are named too (CF_BITMAP, CF_UNICODETEXT…). + EnumFormats::new().take(64).filter_map(raw::format_name_big).collect() + } + + pub fn rtf() -> Option> { + let format = raw::register_format("Rich Text Format")?; + let _open = Clipboard::new_attempts(10).ok()?; + if raw::size(format.get()).is_none_or(|n| n.get() > super::MAX_PASTE_BYTES) { + return None; + } + let mut out = Vec::new(); + raw::get_vec(format.get(), &mut out).ok()?; + Some(out) + } + + pub fn encoded_picture() -> Option> { + None + } +} + +#[cfg(not(any(target_os = "windows", all(unix, not(any(target_os = "macos", target_os = "android", target_os = "emscripten"))))))] +mod platform { + pub fn formats() -> Vec { + Vec::new() + } + + pub fn rtf() -> Option> { + None + } + + pub fn encoded_picture() -> Option> { + None + } +} diff --git a/apps/wordcraft/src/default_apps.rs b/apps/wordcraft/src/default_apps.rs new file mode 100644 index 000000000..8f07b2170 --- /dev/null +++ b/apps/wordcraft/src/default_apps.rs @@ -0,0 +1,181 @@ +//! File › Options › General › Make WordCraft the default for Word documents (#295). +//! +//! The installers register WordCraft for Word, RTF and ODT files without taking over the +//! defaults (`packaging/windows/wordcraft.wxs`, the `.desktop` file's `MimeType=`, `Info.plist`). +//! This is the button that makes it the default when the user asks: +//! - **Windows** doesn't let apps set defaults: it opens Settings › Default apps on WordCraft's +//! page (`ms-settings:defaultapps?registeredAppMachine=WordCraft`; Windows 10 ignores the query +//! and shows the Default apps list), where the user picks WordCraft per type. +//! - **Linux and BSD**: `xdg-mime default ai.storyteller.wordcraft.desktop ` for +//! [`WORD_MIME_TYPES`], which writes the user's `mimeapps.list`. Only when the desktop entry is +//! installed (deb, rpm, Flatpak; a bare AppImage has none), and not from inside the Flatpak +//! sandbox, which can't change the host's defaults. +//! - **macOS** has no API for it here: the Options page explains Finder's Get Info › Change All. +//! +//! [`WORD_MIME_TYPES`]: self::WORD_MIME_TYPES + +use wordcraft_ui_egui::DefaultApp; + +/// The `Services::make_default_app` hook. +pub type Hook = Box Result>; + +/// The hook for this platform; `None` where there is no button (macOS). +pub fn hook() -> Option { + #[cfg(target_os = "windows")] + let hook: Option = Some(Box::new(windows::open_default_apps)); + #[cfg(all(unix, not(target_os = "macos")))] + let hook: Option = Some(Box::new(xdg::make_default)); + #[cfg(not(any(target_os = "windows", all(unix, not(target_os = "macos")))))] + let hook: Option = None; + hook +} + +/// The desktop entry the Linux packages install, and the Flatpak's app id. +#[cfg(any(all(unix, not(target_os = "macos")), test))] +const DESKTOP_ID: &str = "ai.storyteller.wordcraft.desktop"; + +/// The types the button makes WordCraft the default for on Linux and BSD: Word documents and +/// templates (also macro-enabled and legacy), RTF and ODT. The same set the Windows installer +/// lists under WordCraft's Capabilities; Markdown and plain text are only offered under Open With. +/// Every one is in the `.desktop` file's `MimeType=` (a test checks). +#[cfg(any(all(unix, not(target_os = "macos")), test))] +const WORD_MIME_TYPES: &[&str] = &[ + "application/vnd.openxmlformats-officedocument.wordprocessingml.document", // .docx + "application/vnd.ms-word.document.macroEnabled.12", // .docm + "application/vnd.openxmlformats-officedocument.wordprocessingml.template", // .dotx + "application/vnd.ms-word.template.macroEnabled.12", // .dotm + "application/msword", // .doc + "application/msword-template", // .dot + "application/rtf", // .rtf + "text/rtf", // .rtf (older name) + "application/vnd.oasis.opendocument.text", // .odt +]; + +/// The arguments for `xdg-mime` that make `desktop_id` the default for [`WORD_MIME_TYPES`]: one +/// call sets them all. +#[cfg(any(all(unix, not(target_os = "macos")), test))] +fn xdg_mime_args(desktop_id: &str) -> Vec { + ["default", desktop_id].into_iter().chain(WORD_MIME_TYPES.iter().copied()).map(str::to_string).collect() +} + +/// The `applications` folders the desktop looks for entries in, from `XDG_DATA_HOME` (default +/// `~/.local/share`) and `XDG_DATA_DIRS` (default `/usr/local/share:/usr/share`). Relative paths +/// are ignored, as the XDG Base Directory spec says. +#[cfg(any(all(unix, not(target_os = "macos")), test))] +fn application_dirs(data_home: Option<&str>, data_dirs: Option<&str>, home: Option<&str>) -> Vec { + use std::path::{Path, PathBuf}; + let absolute = |s: &&str| Path::new(s).is_absolute(); + let home_dir = data_home + .filter(|s| !s.is_empty()) + .filter(absolute) + .map(PathBuf::from) + .or_else(|| home.filter(|s| !s.is_empty()).filter(absolute).map(|h| Path::new(h).join(".local/share"))); + let system = data_dirs.filter(|s| !s.is_empty()).unwrap_or("/usr/local/share:/usr/share"); + home_dir + .into_iter() + .chain(system.split(':').filter(|s| !s.is_empty()).filter(absolute).map(PathBuf::from)) + .map(|d| d.join("applications")) + .collect() +} + +#[cfg(target_os = "windows")] +mod windows { + use wordcraft_ui_egui::DefaultApp; + + /// Settings › Default apps, on the page of the app registered as `WordCraft` under + /// `HKLM\Software\RegisteredApplications` (the MSI writes it). + const SETTINGS_URI: &str = "ms-settings:defaultapps?registeredAppMachine=WordCraft"; + + pub(super) fn open_default_apps() -> Result { + open::that_detached(SETTINGS_URI).map_err(|e| format!("couldn't open Settings: {e}"))?; + Ok(DefaultApp::SettingsOpened) + } +} + +#[cfg(all(unix, not(target_os = "macos")))] +mod xdg { + use std::process::Command; + + use wordcraft_ui_egui::DefaultApp; + + use super::{DESKTOP_ID, application_dirs, xdg_mime_args}; + + /// The longest error text from `xdg-mime` shown in the status bar. + const MAX_ERROR: usize = 200; + + pub(super) fn make_default() -> Result { + if std::env::var_os("FLATPAK_ID").is_some() || std::path::Path::new("/.flatpak-info").exists() { + return Err( + "the Flatpak sandbox can't change default apps; choose WordCraft under Open With in your file manager or system settings".into() + ); + } + let var = |k: &str| std::env::var(k).ok(); + let dirs = application_dirs(var("XDG_DATA_HOME").as_deref(), var("XDG_DATA_DIRS").as_deref(), var("HOME").as_deref()); + if !dirs.iter().any(|d| d.join(DESKTOP_ID).is_file()) { + return Err(format!("{DESKTOP_ID} isn't installed (an AppImage needs desktop integration first)")); + } + let out = Command::new("xdg-mime").args(xdg_mime_args(DESKTOP_ID)).output().map_err(|e| match e.kind() { + std::io::ErrorKind::NotFound => "xdg-mime isn't installed (it comes with xdg-utils)".to_string(), + _ => format!("couldn't run xdg-mime: {e}"), + })?; + if out.status.success() { + return Ok(DefaultApp::MadeDefault); + } + let stderr = String::from_utf8_lossy(&out.stderr); + let first = stderr.lines().map(str::trim).find(|l| !l.is_empty()).unwrap_or(""); + let message: String = first.chars().take(MAX_ERROR).collect(); + Err(if message.is_empty() { format!("xdg-mime failed ({})", out.status) } else { format!("xdg-mime failed: {message}") }) + } +} + +#[cfg(test)] +mod tests { + use std::path::PathBuf; + + use super::*; + + #[test] + fn xdg_mime_sets_every_word_type_in_one_call() { + let args = xdg_mime_args(DESKTOP_ID); + assert_eq!(args.first().map(String::as_str), Some("default")); + assert_eq!(args.get(1).map(String::as_str), Some("ai.storyteller.wordcraft.desktop")); + assert_eq!(&args[2..], WORD_MIME_TYPES); + for t in [ + "application/vnd.openxmlformats-officedocument.wordprocessingml.document", + "application/msword", + "application/rtf", + "application/vnd.oasis.opendocument.text", + ] { + assert!(args.iter().any(|a| a == t), "{t}"); + } + assert!(!args.iter().any(|a| a == "text/plain" || a == "text/markdown"), "plain text and Markdown are never claimed"); + } + + /// Every type the button claims is one the desktop entry declares, or desktops ignore it. + #[test] + fn the_desktop_entry_declares_every_claimed_type() { + let entry = include_str!("../../../packaging/linux/ai.storyteller.wordcraft.desktop"); + let declared: Vec<&str> = + entry.lines().find_map(|l| l.strip_prefix("MimeType=")).map(|v| v.split(';').filter(|s| !s.is_empty()).collect()).unwrap_or_default(); + for t in WORD_MIME_TYPES { + assert!(declared.contains(t), "{t} missing from MimeType="); + } + assert!(entry.lines().any(|l| l == "Exec=wordcraft %F"), "files are passed on the command line"); + } + + #[test] + fn application_dirs_follow_the_xdg_spec() { + let p = |s: &str| PathBuf::from(s); + assert_eq!( + application_dirs(None, None, Some("/home/a")), + [p("/home/a/.local/share/applications"), p("/usr/local/share/applications"), p("/usr/share/applications")] + ); + assert_eq!( + application_dirs(Some("/data"), Some("/opt/share::relative:/usr/share"), Some("/home/a")), + [p("/data/applications"), p("/opt/share/applications"), p("/usr/share/applications")] + ); + // Empty or relative values fall back to the defaults; no home at all is fine. + assert_eq!(application_dirs(Some(""), Some(""), None), [p("/usr/local/share/applications"), p("/usr/share/applications")]); + assert_eq!(application_dirs(Some("rel"), None, Some("also-rel")), [p("/usr/local/share/applications"), p("/usr/share/applications")]); + } +} diff --git a/apps/wordcraft/src/graphics.rs b/apps/wordcraft/src/graphics.rs index 3ef515797..564b6ed29 100644 --- a/apps/wordcraft/src/graphics.rs +++ b/apps/wordcraft/src/graphics.rs @@ -20,18 +20,28 @@ pub struct Dx12Adapter { /// The backends to initialize on Windows. /// /// An explicit `backend_override` (normally `Backends::from_env()`, i.e. `WGPU_BACKEND`) always -/// wins. Otherwise DirectX 12 only, as long as `dx12_adapters` (what a DirectX 12-only probe found) -/// contains a usable hardware GPU. Without one, OpenGL is the way to the GPU: graphics chips whose -/// driver has no DirectX 12 support (Intel 4th generation on Windows 11, #316) still have an -/// OpenGL driver, and GL can't reach the AMD driver crash described in [`configure_backends`], -/// because an AMD GPU with a driver shows up as a DirectX 12 adapter and keeps us on DirectX 12. -/// When DirectX 12 offers only WARP, it stays enabled next to GL as a last resort for machines -/// without an OpenGL driver either (virtual machines): wgpu prefers any GPU over a software -/// adapter, so it's only picked when GL has nothing better. -pub fn choose_backends(backend_override: Option, dx12_adapters: &[Dx12Adapter]) -> Backends { +/// wins. Otherwise DirectX 12 only, as long as `dx12_graphics` is set and `dx12_adapters` (what a +/// DirectX 12-only probe found) contains a usable hardware GPU. Without one, OpenGL is the way to +/// the GPU: graphics chips whose driver has no DirectX 12 support (Intel 4th generation on Windows +/// 11, #316) still have an OpenGL driver, and GL can't reach the AMD driver crash described in +/// [`configure_backends`], because an AMD GPU with a driver shows up as a DirectX 12 adapter and +/// keeps us on DirectX 12. When DirectX 12 offers only WARP, it stays enabled next to GL as a last +/// resort for machines without an OpenGL driver either (virtual machines): wgpu prefers any GPU +/// over a software adapter, so it's only picked when GL has nothing better. +/// +/// `dx12_graphics` is false on 32-bit Windows. wgpu creates the graphics pipeline through +/// `ID3D12Device2::CreatePipelineState`, and that stream is aligned to 8 bytes. The 32-bit runtime +/// expects pointer alignment (4), so creation returns `E_INVALIDARG` ("The parameter is incorrect") +/// and the process exits before the first frame. Compute pipelines use the older API and succeed, +/// which is why startup gets as far as the egui pipeline and then quits. OpenGL does not use that +/// stream. +pub fn choose_backends(backend_override: Option, dx12_adapters: &[Dx12Adapter], dx12_graphics: bool) -> Backends { if let Some(backends) = backend_override { return backends; } + if !dx12_graphics { + return Backends::GL; + } let hardware = |a: &&Dx12Adapter| a.device_type != DeviceType::Cpu; if dx12_adapters.iter().filter(hardware).any(|a| a.meets_limits) { Backends::DX12 @@ -94,9 +104,16 @@ pub fn configure(options: &mut NativeOptions, backend_override: Option backends } None => { - let adapters = probe_dx12(options); - let backends = choose_backends(None, &adapters); - log::info!("graphics: {} DirectX 12 adapter(s) {adapters:?}, backends {backends:?}", adapters.len()); + // 32-bit: skip the DirectX 12 probe. A capable GPU would be selected, then + // CreatePipelineState fails and the app exits with no window. + let dx12_graphics = cfg!(not(target_arch = "x86")); + let adapters = if dx12_graphics { probe_dx12(options) } else { Vec::new() }; + let backends = choose_backends(None, &adapters, dx12_graphics); + if dx12_graphics { + log::info!("graphics: {} DirectX 12 adapter(s) {adapters:?}, backends {backends:?}", adapters.len()); + } else { + log::info!("graphics: 32-bit Windows uses OpenGL; DirectX 12 graphics pipeline creation fails"); + } backends } }; @@ -157,7 +174,7 @@ mod tests { }; assert!(eframe_default.instance_descriptor.backends.contains(Backends::GL)); let gpu = [adapter(DeviceType::IntegratedGpu, true)]; - assert_eq!(configured_backends(choose_backends(None, &gpu)), Backends::DX12); + assert_eq!(configured_backends(choose_backends(None, &gpu, true)), Backends::DX12); } #[test] @@ -165,8 +182,8 @@ mod tests { let warp_only = [adapter(DeviceType::Cpu, true)]; for backends in [Backends::VULKAN, Backends::GL, Backends::DX12, Backends::VULKAN | Backends::DX12, Backends::empty()] { // Whatever the probe would have found, the override wins (the probe doesn't run then). - assert_eq!(choose_backends(Some(backends), &[]), backends); - assert_eq!(choose_backends(Some(backends), &warp_only), backends); + assert_eq!(choose_backends(Some(backends), &[], true), backends); + assert_eq!(choose_backends(Some(backends), &warp_only, true), backends); assert_eq!(configured_backends(backends), backends); } } @@ -192,13 +209,23 @@ mod tests { fn gl_when_dx12_has_no_usable_gpu() { // #316: no DirectX 12 driver, so DirectX 12 offers only WARP (or nothing) and lags. let warp_only = [adapter(DeviceType::Cpu, true)]; - assert_eq!(choose_backends(None, &warp_only), Backends::GL | Backends::DX12); - assert_eq!(choose_backends(None, &[]), Backends::GL); + assert_eq!(choose_backends(None, &warp_only, true), Backends::GL | Backends::DX12); + assert_eq!(choose_backends(None, &[], true), Backends::GL); // A GPU that can't create eframe's device: GL alone, so it isn't picked over the GL one. let weak_gpu = [adapter(DeviceType::IntegratedGpu, false), adapter(DeviceType::Cpu, true)]; - assert_eq!(choose_backends(None, &weak_gpu), Backends::GL); + assert_eq!(choose_backends(None, &weak_gpu, true), Backends::GL); // Any usable GPU keeps DirectX 12 only, next to WARP or a weaker GPU. let mixed = [adapter(DeviceType::Cpu, true), adapter(DeviceType::IntegratedGpu, false), adapter(DeviceType::DiscreteGpu, true)]; - assert_eq!(choose_backends(None, &mixed), Backends::DX12); + assert_eq!(choose_backends(None, &mixed, true), Backends::DX12); + } + + /// 32-bit Windows: a usable DirectX 12 GPU still starts on OpenGL. `WGPU_BACKEND` still wins. + #[test] + fn x86_uses_opengl_even_with_a_usable_gpu() { + let gpu = [adapter(DeviceType::IntegratedGpu, true)]; + assert_eq!(choose_backends(None, &gpu, false), Backends::GL); + assert_eq!(choose_backends(Some(Backends::DX12), &gpu, false), Backends::DX12); + assert_eq!(choose_backends(Some(Backends::VULKAN), &[], false), Backends::VULKAN); + assert_eq!(configured_backends(choose_backends(None, &gpu, false)), Backends::GL); } } diff --git a/apps/wordcraft/src/main.rs b/apps/wordcraft/src/main.rs index 5bdfedaad..988b4f67b 100644 --- a/apps/wordcraft/src/main.rs +++ b/apps/wordcraft/src/main.rs @@ -12,17 +12,19 @@ mod animations; #[cfg(target_os = "macos")] mod apple_events; +mod clipboard; mod control_server; +mod default_apps; mod file_dialogs; #[cfg(any(target_os = "windows", test))] mod graphics; mod logging; mod print; +mod region; use wordcraft_engine::Session; use wordcraft_ui_egui::{ Services, UiState, WordApp, - paste_picture::ClipboardPicture, window_geometry::{WindowGeometry, take_rescue}, }; @@ -128,6 +130,21 @@ fn log_dir() -> Option { Some(prefs_path()?.parent()?.join("logs")) } +/// `custom-dictionary.txt` in the settings folder: the words Add to Dictionary kept (#40). +fn user_dict_path() -> Option { + wordcraft_control_key::settings_dir().map(|b| b.join(wordcraft_engine::Session::USER_DICT_FILE)) +} + +fn load_user_dictionary(app: &mut WordApp) { + if !prefs_enabled() { + return; + } + if let Some(p) = user_dict_path() { + wordcraft_proof::load_user_dictionary(&p); + app.session.user_dict_path = Some(p); + } +} + /// Runs without preferences (`WORDCRAFT_NO_PREFS`, agents' test runs) neither read nor write them. fn prefs_enabled() -> bool { std::env::var_os("WORDCRAFT_NO_PREFS").is_none() @@ -182,33 +199,25 @@ fn save_prefs(app: &WordApp) { } /// Native file dialogs, shown without blocking the window (`file_dialogs`, #94), File › Print -/// through the system spooler (`print`), pictures/text on the system clipboard (#45), and -/// the system's animation setting (`animations`). +/// through the system spooler (`print`), the system clipboard's text, formatted copies, +/// pictures and format names (`clipboard`, #45, #584, #687, #833), the system's animation setting +/// (`animations`), and making WordCraft the default app for Word documents (`default_apps`, #295). fn services(file_dialog: file_dialogs::Hook, ctx: &egui::Context) -> Services { Services { file_dialog: Some(file_dialog), printers: Some(print::discover()), print_native: Some(print::submit()), - clipboard_picture: Some(Box::new(clipboard_picture)), + clipboard_picture: Some(Box::new(clipboard::picture)), + make_default_app: default_apps::hook(), animations_enabled: Some(animations::hook(ctx)), - clipboard_text: Some(Box::new(|| arboard::Clipboard::new().ok()?.get_text().ok())), + clipboard_text: Some(Box::new(clipboard::text)), + clipboard_html: Some(Box::new(clipboard::html)), + clipboard_rtf: Some(Box::new(clipboard::rtf)), + clipboard_formats: Some(Box::new(clipboard::formats)), ..Default::default() } } -/// The picture on the system clipboard (#45): copied files, else image pixels. The UI checks the -/// sizes before inserting anything. -fn clipboard_picture() -> Option { - let mut clipboard = arboard::Clipboard::new().map_err(|e| log::warn!("clipboard: {e}")).ok()?; - if let Ok(files) = clipboard.get().file_list() - && !files.is_empty() - { - return Some(ClipboardPicture::Files(files)); - } - let image = clipboard.get_image().ok()?; - Some(ClipboardPicture::Pixels { width: image.width, height: image.height, rgba: image.bytes.into_owned() }) -} - /// Window, Dock and taskbar icon. fn app_icon() -> Option { #[cfg(target_os = "macos")] @@ -272,8 +281,8 @@ fn main() -> eframe::Result { options.viewport = options.viewport.with_icon(icon); } // Before eframe creates the wgpu instance: default Windows to DirectX 12 only, or OpenGL when - // DirectX 12 has no GPU driver, and compile DirectX 12 shaders with FXC, which needs no - // DLL of ours (#603; see graphics.rs). + // DirectX 12 has no GPU driver or this is a 32-bit build, and compile DirectX 12 shaders with + // FXC, which needs no DLL of ours (#603; see graphics.rs). #[cfg(target_os = "windows")] graphics::configure(&mut options, eframe::wgpu::Backends::from_env()); load_user_fonts(); @@ -281,6 +290,16 @@ fn main() -> eframe::Result { if let Some(window) = restored { options.viewport = window.apply(options.viewport); } + #[cfg(all(unix, not(target_os = "macos")))] + // winit has no file drag-and-drop on Wayland (only on X11), so dropping files from the file + // manager showed a "no" cursor. Run through XWayland when it's there; WORDCRAFT_WAYLAND=1 + // keeps the native Wayland backend. + if std::env::var_os("DISPLAY").is_some() && std::env::var_os("WORDCRAFT_WAYLAND").is_none() { + options.event_loop_builder = Some(Box::new(|b| { + use winit::platform::x11::EventLoopBuilderExtX11; + b.with_x11(); + })); + } // Registered before the event loop starts, so it catches the Finder event that launched us as // well as later ones. Lives until the event loop returns; the app creator only borrows it. #[cfg(target_os = "macos")] @@ -291,10 +310,23 @@ fn main() -> eframe::Result { "WordCraft", options, Box::new(move |cc| { - let doc = if sample { wordcraft_engine::sample::sample_document() } else { wordcraft_doc::Document::new() }; + // The system region picks the new document's paper, the units and the decimal + // separator (#838). + let region = region::system(); + let doc = if sample { wordcraft_engine::sample::sample_document() } else { wordcraft_engine::blank_document(®ion) }; let (file_dialog, dialogs) = file_dialogs::hook(&cc.egui_ctx); - let mut app = WordApp::new(Session::new(doc), services(file_dialog, &cc.egui_ctx)); + let mut session = Session::new(doc); + session.system_region = region; + let mut app = WordApp::new(session, services(file_dialog, &cc.egui_ctx)); load_prefs(&mut app); + load_user_dictionary(&mut app); + // A region chosen in File › Options replaces the system's for the first blank + // document too (#838). + if !sample && app.session.prefs.region.is_some() { + let doc = app.session.blank_document(); + app.session.set_document(doc); + } + app.watch_desktop_theme(&cc.egui_ctx); app.ui.window = restored; app.integrated_titlebar = cfg!(target_os = "macos"); let mut key_file = None; @@ -304,6 +336,12 @@ fn main() -> eframe::Result { app = app.with_control(rx); key_file = file; } + // Nothing to open (no files, no sample, no control channel driving the editor): start + // on File › Home, as Word does (#766), unless the Start screen option is off. Files + // macOS hands over later open on top of it and close it (`apple_events`). + if files.is_empty() && !sample && control_port.is_none() { + app.start_screen(); + } for f in files { if let Err(e) = app.run("file.open", serde_json::json!({"path": f})) { log::warn!("{f}: {e}"); diff --git a/apps/wordcraft/src/region.rs b/apps/wordcraft/src/region.rs new file mode 100644 index 000000000..e58a9276d --- /dev/null +++ b/apps/wordcraft/src/region.rs @@ -0,0 +1,30 @@ +//! The system region new documents follow (#838): the paper File › New starts on, the default +//! measurement unit and the decimal separator. +//! +//! - Windows: the preferred language with the home location from Settings › Time & language › +//! Region (`GlobalizationPreferences.HomeGeographicRegion`, a safe Windows Runtime API), since +//! the language list alone says `en-US` for an English interface in Poland. +//! - Elsewhere: the locale the UI crate reads ([`wordcraft_ui_egui::i18n::system_region`]). + +/// The system region. +pub fn system() -> wordcraft_engine::Region { + // Runs without preferences (agents' and screenshot runs) behave the same on every machine: + // en-US unless `WORDCRAFT_REGION` says otherwise. + if std::env::var_os("WORDCRAFT_NO_PREFS").is_some() && std::env::var_os("WORDCRAFT_REGION").is_none() { + return wordcraft_engine::Region::default(); + } + let region = wordcraft_ui_egui::i18n::system_region(); + #[cfg(target_os = "windows")] + { + if std::env::var_os("WORDCRAFT_REGION").is_none() { + return match windows::System::UserProfile::GlobalizationPreferences::HomeGeographicRegion() { + Ok(home) => region.with_territory(&home.to_string()), + Err(e) => { + log::debug!("system home region: {e}"); + region + } + }; + } + } + region +} diff --git a/assets/hyphenation/pl.pat b/assets/hyphenation/pl.pat new file mode 100644 index 000000000..1a3c6c1d3 --- /dev/null +++ b/assets/hyphenation/pl.pat @@ -0,0 +1,4154 @@ +% title: Hyphenation patterns for Polish +% copyright: Copyright (C) 1987-1995 Hanna Kołodziejska, Bogusław Jackowski, Marek Ryćko +% notice: This file is part of the hyph-utf8 package. +% See http://www.hyphenation.org/tex for more information. +% language: +% name: Polish +% tag: pl +% version: 3.0a, Wednesday, May 17th, 1995 +% authors: +% - +% name: Hanna Kołodziejska +% - +% name: Bogusław Jackowski +% contact: b_jackowski (at) gust.org.pl +% - +% name: Marek Ryćko +% contact: marek (at) do.com.pl +% licence: +% - This file is made available under any of the following licences: +% - +% name: MIT +% url: https://opensource.org/licenses/MIT +% text: > +% Permission is hereby granted, free of charge, to any person +% obtaining a copy of this software and associated documentation +% files (the “Software”), to deal in the Software without +% restriction, including without limitation the rights to use, +% copy, modify, merge, publish, distribute, sublicense, and/or sell +% copies of the Software, and to permit persons to whom the +% Software is furnished to do so, subject to the following +% conditions: +% +% The above copyright notice and this permission notice shall be +% included in all copies or substantial portions of the Software. +% +% THE SOFTWARE IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, +% EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES +% OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND +% NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT +% HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, +% WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +% FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR +% OTHER DEALINGS IN THE SOFTWARE. +% - text: > +% This macro file belongs to the public domain +% under the conditions specified by the author of TeX: +% +% “Macro files like PLAIN.TEX should not be changed in any way, +% except with respect to preloaded fonts, +% unless the changes are authorized by the authors of the macros.” +% +% Donald E. Knuth +% hyphenmins: +% typesetting: +% left: 2 +% right: 2 +% changes: +% - The first version of the patterns was developed +% by Hanna Kołodziejska (1987). +% - The adaptation to the LeX format (see below) and extensive modification +% were done by Bogusław Jackowski & Marek Ryćko (1987--1989). +% - The hyphenation rules were further improved and adapted to the +% TeX 3.x requirements by Hanna Kołodziejska (1991). +% - Lone-standing version (3.0a) of patterns was prepared (under pressure +% from LaTeX users) by Bogusław Jackowski and Marek Ryćko, following +% Mariusz Olko's suggestions, 1995. +% - The LeX format mentioned above was the first version of the adaptation +% of TeX to the Polish language. The next version is called MeX. +% texlive: +% encoding: qx +% babelname: polish +% legacy_patterns: plhyph.tex +% message: Polish hyphenation patterns +% description: |- +% Hyphenation patterns for Polish in QX and UTF-8 encodings. +% These patterns are also used by Polish TeX formats MeX and LaMeX. +% +\patterns{ +.ćć8 +.ćł8 +.ćń8 +.ćś8 +.ćź8 +.ćż8 +.ć8 +.ćb8 +.ćc8 +.ćd8 +.ćf8 +.ćg8 +.ćh8 +.ćj8 +.ćk8 +.ćl8 +.ćm8 +.ćn8 +.ćp8 +.ćr8 +.ćs8 +.ćt8 +.ćv8 +.ćw8 +.ćwier2ć3 +.ćx8 +.ćz8 +.łć8 +.łł8 +.łń8 +.łś8 +.łź8 +.łż8 +.ł8 +.łb8 +.łc8 +.łd8 +.łf8 +.łg8 +.łh8 +.łj8 +.łk8 +.łl8 +.łm8 +.łn8 +.łp8 +.łr8 +.łs8 +.łt8 +.łv8 +.łw8 +.łx8 +.łz8 +.ńć8 +.ńł8 +.ńń8 +.ńś8 +.ńź8 +.ńż8 +.ń8 +.ńb8 +.ńc8 +.ńd8 +.ńf8 +.ńg8 +.ńh8 +.ńj8 +.ńk8 +.ńl8 +.ńm8 +.ńn8 +.ńp8 +.ńr8 +.ńs8 +.ńt8 +.ńv8 +.ńw8 +.ńx8 +.ńz8 +.ść8 +.śł8 +.śń8 +.śś8 +.śź8 +.śż8 +.ś8 +.śb8 +.śc8 +.śd8 +.śf8 +.śg8 +.śh8 +.śj8 +.śk8 +.śl8 +.śm8 +.śn8 +.śp8 +.śró2d5 +.śródr2 +.śr8 +.śs8 +.śt8 +.śv8 +.św8 +.światło3w2 +.śx8 +.śz8 +.źć8 +.źł8 +.źń8 +.źś8 +.źź8 +.źż8 +.ź8 +.źb8 +.źc8 +.źdź8 +.źd8 +.źf8 +.źg8 +.źh8 +.źj8 +.źk8 +.źl8 +.źm8 +.źn8 +.źp8 +.źr8 +.źs8 +.źt8 +.źv8 +.źw8 +.źx8 +.źz8 +.żć8 +.żł8 +.żń8 +.żś8 +.żź8 +.żż8 +.ż8 +.żb8 +.żc8 +.żd8 +.żf8 +.żg8 +.żh8 +.żj8 +.żk8 +.żl8 +.żm8 +.żn8 +.żp8 +.żr8 +.żs8 +.żt8 +.żv8 +.żw8 +.żx8 +.żz8 +.a2b2s3t +.a2d3 +.ad4a +.ad4e +.ad4i +.ad4o +.ad4u +.ad4y +.ad5apt +.ad5iu +.ad5op +.ad5or +.ae3ro +.aeroa2 +.aeroe2 +.aeroi2 +.aeroo2 +.aerou2 +.antya2 +.antye2 +.antyi2 +.antyo2 +.antyu2 +.arcy3ł2 +.arcy3b2 +.arcy3bz2 +.arcy3k2 +.arcy3m2 +.arcya2 +.arcye2 +.arcyi2 +.arcyo2 +.arcyu2 +.au3g2 +.au3k2 +.au3t2 +.auto3ch2 +.autoa2 +.autoe2 +.autoi2 +.autoo2 +.autotran2s3 +.autou2 +.bć8 +.bł8 +.bń8 +.bś8 +.bź8 +.bż8 +.b8 +.bb8 +.bc8 +.bd8 +.be2z3 +.be3z4an +.be3z4ec +.be3z4ik +.bezch2 +.bezm2 +.bezo2 +.bezo2b1j +.bezw2 +.bezzw2 +.bf8 +.bg8 +.bh8 +.bj8 +.bk8 +.bl8 +.bm8 +.bn8 +.bp8 +.br8 +.brz8 +.bs8 +.bt8 +.bv8 +.bw8 +.bx8 +.bz8 +.cć8 +.cł8 +.cń8 +.cś8 +.cź8 +.cż8 +.c8 +.cało3ś2 +.cało3k2 +.cb8 +.cc8 +.cd8 +.cf8 +.cg8 +.ch8 +.chrz8 +.cienko3w2 +.ciepło3kr2 +.cj8 +.ck8 +.cl8 +.cm8 +.cn8 +.cp8 +.cr8 +.cs8 +.ct8 +.cv8 +.cw8 +.cx8 +.cz8 +.czarno3k2 +.czk8 +.cztere2ch3 +.czterechse2t3 +.cztero3ś2 +.czwó2r3 +.czwó3r4ą +.czwó3r4ę +.czwó3r4a +.czwó3r4e +.czwó3r4o +.dć8 +.dł8 +.długo3tr2 +.długo3w2 +.dń8 +.dś8 +.dź8 +.dż8 +.d8 +.daleko3w2 +.db8 +.dc8 +.dd8 +.de2z3 +.de3z4a3bil +.de3z4a3wu +.de3z4el +.de3z4er +.de3z4y +.deza2 +.dezo2 +.df8 +.dg8 +.dh8 +.dj8 +.dk8 +.dl8 +.dm8 +.dn8 +.do3ć2 +.do3ł2 +.do3ś2 +.do3ź2 +.do3ż2 +.do3b2 +.do3c2 +.do3d2 +.do3f2 +.do3g2 +.do3h2 +.do3k2 +.do3l2 +.do3m2 +.do3p2 +.do3r2 +.do3s2 +.do3t2 +.do3w2 +.do3z2 +.do4ł3k +.do4k3t +.do4l3n +.do4m3k +.do4r3s +.do4w3c +.do5m4k2n +.dobr2 +.dobrz2 +.doch2 +.docz2 +.dodź2 +.dodż2 +.dodz2 +.dogrz2 +.dopch2 +.doprz2 +.dorż2 +.dorz2 +.dosch2 +.dosm2 +.dosz2 +.dotk2 +.dotr2 +.dp8 +.dr8 +.drogo3w2 +.drz8 +.ds8 +.dt8 +.dv8 +.dwó2j3 +.dwó3j4ą +.dwó3j4ę +.dwó3j4a +.dwó3j4e +.dwó3j4o +.dw8 +.dx8 +.dy2s3 +.dy2z3 +.dy3s4e +.dy3s4o +.dy3s4ta +.dy3s4y +.dy3sz +.dy3z4e +.dyzu2 +.dz8 +.dziesięcio3ś2 +.dziewięćse2t3 +.dziewię2ć3 +.dziewięcio3ś2 +.e2k2s3 +.e2m3e2s5ze2t +.e2s1e2s1ma +.e2s1ha +.e2s1t +.egoa2 +.egoe2 +.egoi2 +.egoo2 +.egou2 +.eks4y +.elektroa2 +.elektroe2 +.elektroi2 +.elektroo2 +.elektrou2 +.fć8 +.fł8 +.fń8 +.fś8 +.fź8 +.fż8 +.f8 +.fb8 +.fc8 +.fd8 +.ff8 +.fg8 +.fh8 +.fj8 +.fk8 +.fl8 +.fm8 +.fn8 +.fp8 +.fr8 +.fs8 +.ft8 +.fv8 +.fw8 +.fx8 +.fz8 +.gć8 +.gł8 +.gń8 +.gś8 +.gź8 +.gż8 +.g8 +.gb8 +.gc8 +.gd8 +.ge2o3 +.gf8 +.gg8 +.gh8 +.gj8 +.gk8 +.gl8 +.gm8 +.gn8 +.go2u3 +.gp8 +.gr8 +.grubo3w2 +.grz8 +.gs8 +.gt8 +.gv8 +.gw8 +.gx8 +.gz8 +.hć8 +.hł8 +.hń8 +.hś8 +.hź8 +.hż8 +.h8 +.hb8 +.hc8 +.hd8 +.hf8 +.hg8 +.hh8 +.hipe2r3 +.hipe3r4o +.hipera2 +.hipere2 +.hj8 +.hk8 +.hl8 +.hm8 +.hn8 +.hp8 +.hr8 +.hs8 +.ht8 +.hv8 +.hw8 +.hx8 +.hz8 +.i2n3 +.i2s3l +.i3n4ic +.i3n4o +.i3n4u +.i4n5o2k +.in4f3lan +.ino3w2 +.izoa2 +.izoe2 +.izoi2 +.izoo2 +.izou2 +.jć8 +.jł8 +.jń8 +.jś8 +.jź8 +.jż8 +.j8 +.jadło3w2 +.jb8 +.jc8 +.jd8 +.jf8 +.jg8 +.jh8 +.jj8 +.jk8 +.jl8 +.jm8 +.jn8 +.jp8 +.jr8 +.js8 +.jt8 +.jv8 +.jw8 +.jx8 +.jz8 +.kć8 +.kł8 +.kń8 +.kś8 +.kź8 +.kż8 +.k8 +.kb8 +.kc8 +.kd8 +.kf8 +.kg8 +.kh8 +.kilkuse2t3 +.kilkuseto2 +.kj8 +.kk8 +.kl8 +.km8 +.kn8 +.koło3w2 +.kon2t2r3 +.kon3tr4a +.kon3tr4e +.kon3tr4o3l +.kon3tr4o3w +.kon3tr4y +.kon4tr5a2gi +.kon4tr5a2se +.kon4tr5a2sy +.kon4tr5a2ta +.kon4tr5adm +.kon4tr5akc +.kon4tr5alt +.kon4tr5arg +.kontro2 +.kontru2 +.kp8 +.krótko3tr2 +.krótko3w2 +.kr8 +.kro2ć3 +.krz8 +.ks8 +.kt8 +.kv8 +.kw8 +.kx8 +.kz8 +.lć8 +.lł8 +.lń8 +.lś8 +.lź8 +.lż8 +.l8 +.lb8 +.lc8 +.ld8 +.lf8 +.lg8 +.lh8 +.lj8 +.lk8 +.ll8 +.lm8 +.ln8 +.lp8 +.lr8 +.ls8 +.lt8 +.ludo3w2 +.lv8 +.lw8 +.lx8 +.lz8 +.mć8 +.mł8 +.mń8 +.mś8 +.mź8 +.mż8 +.m8 +.mb8 +.mc8 +.md8 +.mf8 +.mg8 +.mh8 +.mili3amp +.mj8 +.mk8 +.ml8 +.mm8 +.mn8 +.możno3w2 +.mp8 +.mr8 +.ms8 +.mt8 +.mv8 +.mw8 +.mx8 +.mz8 +.nć8 +.nł8 +.nń8 +.nś8 +.nź8 +.nż8 +.n8 +.na2d2 +.na2j +.na3ć2 +.na3ł2 +.na3ś2 +.na3ź2 +.na3ż2 +.na3b2 +.na3c2 +.na3dą +.na3dę +.na3dź2 +.na3d4łub +.na3d4ir +.na3d4much +.na3d4ręcz +.na3d4r2w +.na3d4repcz +.na3d4rept +.na3d4ruk +.na3d4rz +.na3d4worn +.na3daj +.na3de +.na3do +.na3dy +.na3dzi +.na3f2 +.na3g2 +.na3h2 +.na3ją +.na3ję +.na3jazd +.na3je +.na3k2 +.na3l2 +.na3m2 +.na3p2 +.na3r2 +.na3s2 +.na3t2 +.na3u2 +.na3w2 +.na3z2 +.na4d3o2b2ł +.na4d3o2bojcz +.na4d3o2bowi +.na4d3o2brot +.na4d3o2drz +.na4d3o2kien +.na4d3olbrz +.na4d5rzą +.na4d5rzę +.na4d5rzecz +.na4d5rzy +.na4d5ziem +.na4f3c +.na4f3t +.na4j3e2f +.na4j3e2g +.na4j3e2k2s +.na4j3e2ko +.na4j3e2n +.na4j3e2r +.na4j3e2s +.na4j3e2w +.na4j3emf +.na4j3eu +.na4r3c +.na4r3d +.na4r3k +.na4r3r +.na4r3t +.nabrz2 +.nach2 +.nacz2 +.nadśrod5ziem +.nad3ć2 +.nad3ł2 +.nad3ś2 +.nad3b2 +.nad3c2 +.nad3d2 +.nad3e2tat +.nad3f2 +.nad3g2 +.nad3h2 +.nad3i2 +.nad3j2 +.nad3k2 +.nad3l2 +.nad3m2 +.nad3n2 +.nad3p2 +.nad3r2 +.nad3s2 +.nad3t2 +.nad3u2 +.nad3w2 +.nad5ż2 +.nad5zó +.nad5z2mys +.nad5zo +.nad5zwycz +.nadch2 +.nadcz2 +.naddź2 +.nade3ć2 +.nade3ł2 +.nade3ś2 +.nade3ź2 +.nade3ż2 +.nade3b2 +.nade3c2 +.nade3d2 +.nade3f2 +.nade3g2 +.nade3h2 +.nade3k2 +.nade3l2 +.nade3m2 +.nade3p2 +.nade3r2 +.nade3s2 +.nade3t2 +.nade3w2 +.nade3z2 +.nade4p3c +.nade4p3n +.nade4p3t +.nadech2 +.nadecz2 +.nadedź2 +.nadedż2 +.nadedz2 +.naderż2 +.naderz2 +.nadesz2 +.nadsz2 +.nadtr2 +.nadz2 +.nagrz2 +.naj3ć2 +.naj3ł2 +.naj3ś2 +.naj3ź2 +.naj3ż2 +.naj3akt +.naj3au +.naj3b2 +.naj3c2 +.naj3d2 +.naj3f2 +.naj3g2 +.naj3h2 +.naj3i2 +.naj3k2 +.naj3l2 +.naj3m2 +.naj3o2 +.naj3o2ć2 +.naj3o2ł2 +.naj3o2ś2 +.naj3o2ź2 +.naj3o2ż2 +.naj3o2b2 +.naj3o2c2 +.naj3o2d2 +.naj3o2f2 +.naj3o2g2 +.naj3o2h2 +.naj3o2k2 +.naj3o2l2 +.naj3o2m2 +.naj3o2p2 +.naj3o2r2 +.naj3o2s2 +.naj3o2t2 +.naj3o2w2 +.naj3o2z2 +.naj3p2 +.naj3r2 +.naj3ro2z3 +.naj3s2 +.naj3t2 +.naj3u2 +.naj3w2 +.naj3z2 +.najbe2z3 +.najbezw2 +.najch2 +.najcz2 +.najdź2 +.najdż2 +.najdo3ć2 +.najdo3ł2 +.najdo3ś2 +.najdo3ź2 +.najdo3ż2 +.najdo3b2 +.najdo3c2 +.najdo3d2 +.najdo3f2 +.najdo3g2 +.najdo3h2 +.najdo3k2 +.najdo3l2 +.najdo3m2 +.najdo3p2 +.najdo3r2 +.najdo3s2 +.najdo3t2 +.najdo3w2 +.najdo3z2 +.najdoch2 +.najdocz2 +.najdodź2 +.najdodż2 +.najdodz2 +.najdorz2 +.najdosz2 +.najdotk2 +.najdz2 +.najkr2 +.najob3ć2 +.najob3ł2 +.najob3ś2 +.najob3ź2 +.najob3ż2 +.najob3c2 +.najob3d2 +.najob3f2 +.najob3g2 +.najob3h2 +.najob3j2 +.najob3k2 +.najob3l2 +.najob3m2 +.najob3n2 +.najob3p2 +.najob3s2 +.najob3t2 +.najob3w2 +.najobch2 +.najobcz2 +.najobdź2 +.najobdż2 +.najobdz2 +.najobrz2 +.najobsz2 +.najoch2 +.najocz2 +.najodź2 +.najod3ć2 +.najod3ś2 +.najod3c2 +.najod3d2 +.najod3f2 +.najod3g2 +.najod3h2 +.najod3j2 +.najod3k2 +.najod3l2 +.najod3m2 +.najod3n2 +.najod3p2 +.najod3s2 +.najod3t2 +.najod3w2 +.najod5ż2 +.najodch2 +.najodcz2 +.najoddź2 +.najoddż2 +.najoddz2 +.najodsz2 +.najodz2 +.najorz2 +.najosz2 +.najro3z4u +.najrz2 +.najsm2 +.najsz2 +.najtk2 +.najtr2 +.najucz2 +.najzw2 +.nakr2 +.napo2d2 +.napo3ć2 +.napo3ł2 +.napo3ś2 +.napo3ź2 +.napo3ż2 +.napo3b2 +.napo3c2 +.napo3f2 +.napo3g2 +.napo3h2 +.napo3k2 +.napo3l2 +.napo3m2 +.napo3p2 +.napo3r2 +.napo3s2 +.napo3t2 +.napo3w2 +.napo3z2 +.napo4m3p +.napoch2 +.napocz2 +.napodź2 +.napodż2 +.napod3d +.napomk2 +.naporz2 +.naposz2 +.naprz2 +.narż2 +.naro2z3 +.narz2 +.nasm2 +.nasz2 +.natch2 +.natk2 +.naz3m2 +.nazw2 +.nb8 +.nc8 +.nd8 +.ne2o3 +.nf8 +.ng8 +.nh8 +.nie3ć2 +.nie3ł2 +.nie3ś2 +.nie3ź2 +.nie3ż2 +.nie3b2 +.nie3c2 +.nie3d2 +.nie3f2 +.nie3g2 +.nie3h2 +.nie3k2 +.nie3l2 +.nie3m2 +.nie3p2 +.nie3r2 +.nie3s2 +.nie3t2 +.nie3u2 +.nie3w2 +.nie3z2 +.nie4c3c +.nie4c3k +.nie4dź3 +.nie4m3c +.nie4m3k +.niech2 +.niecz2 +.niedż2 +.niedo3ć2 +.niedo3ł2 +.niedo3ś2 +.niedo3ź2 +.niedo3ż2 +.niedo3b2 +.niedo3c2 +.niedo3d2 +.niedo3f2 +.niedo3g2 +.niedo3h2 +.niedo3k2 +.niedo3l2 +.niedo3m2 +.niedo3p2 +.niedo3r2 +.niedo3s2 +.niedo3t2 +.niedo3w2 +.niedo3z2 +.niedobrz2 +.niedoch2 +.niedocz2 +.niedodź2 +.niedodż2 +.niedodz2 +.niedokr2 +.niedomk2 +.niedopch2 +.niedorz2 +.niedosz2 +.niedotk2 +.niedz2 +.nieoć2 +.nieoł2 +.nieoś2 +.nieoź2 +.nieoż2 +.nieo2 +.nieob2 +.nieob3ć2 +.nieob3ś2 +.nieob3ź2 +.nieob3ż2 +.nieob3c2 +.nieob3d2 +.nieob3f2 +.nieob3g2 +.nieob3h2 +.nieob3j2 +.nieob3k2 +.nieob3m2 +.nieob3p2 +.nieob3s2 +.nieob3w2 +.nieobch2 +.nieobcz2 +.nieobdź2 +.nieobdż2 +.nieobdz2 +.nieobsz2 +.nieoc2 +.nieoch2 +.nieocz2 +.nieodź2 +.nieod2 +.nieod3ć2 +.nieod3ł2 +.nieod3ś2 +.nieod3c2 +.nieod3d2 +.nieod3f2 +.nieod3g2 +.nieod3h2 +.nieod3j2 +.nieod3k2 +.nieod3l2 +.nieod3n2 +.nieod3p2 +.nieod3s2 +.nieod3t2 +.nieod3wr +.nieod5ż2 +.nieodch2 +.nieodcz2 +.nieoddź2 +.nieoddż2 +.nieoddz2 +.nieodsz2 +.nieodw2 +.nieodz2 +.nieof2 +.nieog2 +.nieoh2 +.nieok2 +.nieol2 +.nieom2 +.nieop2 +.nieor2 +.nieorz2 +.nieos2 +.nieosz2 +.nieot2 +.nieow2 +.nieoz2 +.niepo2d2 +.niepo3ć2 +.niepo3ł2 +.niepo3ś2 +.niepo3ź2 +.niepo3ż2 +.niepo3b2 +.niepo3c2 +.niepo3dź2 +.niepo3d4łu +.niepo3d4much +.niepo3d4ręcz +.niepo3d4raż +.niepo3d4rap +.niepo3d4repcz +.niepo3d4rept +.niepo3d4waj +.niepo3d4woj +.niepo3do +.niepo3du +.niepo3dz2 +.niepo3f2 +.niepo3g2 +.niepo3h2 +.niepo3k2 +.niepo3l2 +.niepo3m2 +.niepo3p2 +.niepo3r2 +.niepo3s2 +.niepo3t2 +.niepo3w2 +.niepo3z2 +.niepo4d3o2choc +.niepo4d3o2strz +.niepoch2 +.niepocz2 +.niepod3ć2 +.niepod3ł2 +.niepod3ś2 +.niepod3b2 +.niepod3c2 +.niepod3d2 +.niepod3f2 +.niepod3g2 +.niepod3h2 +.niepod3j2 +.niepod3k2 +.niepod3l2 +.niepod3m2 +.niepod3n2 +.niepod3p2 +.niepod3r2 +.niepod3s2 +.niepod3t2 +.niepod3w2 +.niepod5ż +.niepodch2 +.niepodcz2 +.niepoddź2 +.niepoddż2 +.niepodsm2 +.niepodsz2 +.nieporz2 +.nieposm2 +.nieposz2 +.nieprzełk2 +.nieprze2d2 +.nieprze3ć2 +.nieprze3ł2 +.nieprze3ś2 +.nieprze3ź2 +.nieprze3ż2 +.nieprze3b2 +.nieprze3brz2 +.nieprze3c2 +.nieprze3dź2 +.nieprze3d4łuż +.nieprze3d4much +.nieprze3d4ramat +.nieprze3d4ruk +.nieprze3d4ryl +.nieprze3d4rz2 +.nieprze3d4um +.nieprze3dy +.nieprze3dz2 +.nieprze3e2k2s3 +.nieprze3f2 +.nieprze3g2 +.nieprze3h2 +.nieprze3k2 +.nieprze3l2 +.nieprze3m2 +.nieprze3n2 +.nieprze3p2 +.nieprze3r2 +.nieprze3s2 +.nieprze3t2 +.nieprze3w2 +.nieprze3z2 +.nieprze4d5łużyc +.nieprze4d5ż2 +.nieprze4d5z2a +.nieprze4d5zg2 +.nieprze4d5zim +.nieprze4d5zj +.nieprze4d5zl +.nieprze4d5zw2r +.nieprze4d5zwoj +.nieprzech2 +.nieprzecz2 +.nieprzed3ć2 +.nieprzed3ł2 +.nieprzed3ś2 +.nieprzed3c2 +.nieprzed3d2 +.nieprzed3f2 +.nieprzed3g2 +.nieprzed3h2 +.nieprzed3i2 +.nieprzed3j2 +.nieprzed3k2 +.nieprzed3l2 +.nieprzed3m2 +.nieprzed3n2 +.nieprzed3p2 +.nieprzed3r2 +.nieprzed3s2 +.nieprzed3sz2 +.nieprzed3t2 +.nieprzed3u2 +.nieprzed3w2 +.nieprzedch2 +.nieprzedcz2 +.nieprzeddź2 +.nieprzeddż2 +.nieprzeddz2 +.nieprzegrz2 +.nieprzekl2 +.nieprzekr2 +.nieprzepch2 +.nieprzerż2 +.nieprzerz2 +.nieprzesch2 +.nieprzesm2 +.nieprzesz2 +.nieprzetk2 +.nieprzetr2 +.niero2z3 +.niero3z4e +.niero3z4u +.nierozś2 +.nierozbrz2 +.nieroze3r2 +.nierozm2 +.nieroztr2 +.nierz2 +.niesu2b3 +.niesu3b4ie +.niesz2 +.nietk2 +.nietr2 +.nieucz2 +.nieuw2 +.niewy3ć2 +.niewy3ł2 +.niewy3ś2 +.niewy3ź2 +.niewy3ż2 +.niewy3b2 +.niewy3c2 +.niewy3d2 +.niewy3f2 +.niewy3g2 +.niewy3h2 +.niewy3k2 +.niewy3l2 +.niewy3m2 +.niewy3p2 +.niewy3r2 +.niewy3s2 +.niewy3t2 +.niewy3w2 +.niewy3z2 +.niewybrz2 +.niewych2 +.niewycz2 +.niewydź2 +.niewydż2 +.niewydz2 +.niewyrz2 +.niewysz2 +.niewytk2 +.niewytr2 +.niezw2 +.nj8 +.nk8 +.nl8 +.nm8 +.nn8 +.np8 +.nr8 +.ns8 +.nt8 +.nv8 +.nw8 +.nx8 +.nz8 +.oć2 +.oś2 +.ośmio3ś2 +.oź2 +.oż2 +.o2b2 +.o2d2 +.o2t3chł +.o3b4łą +.o3b4łę +.o3b4łoc +.o3b4luzg +.o3b4rać +.o3b4raso +.o3b4roń +.o3b4ron +.o3b4ryź +.o3b4ryz +.o3b4rz2 +.o3be +.o3bi +.o3d4iu +.o3d4ręt +.o3d4rap +.o3d4robin +.o3d4rut +.o3d4rwi +.o3d4rzeć +.o3d4rzw +.o3d6zia +.o3d6zie +.o3de +.o3l2śn +.o4b5łocz +.o4b5rzą +.o4b5rzęd +.o4b5rzez +.o4b5rzuc +.o4b5rzut +.o4b5rzyn +.o4d7ziar +.o4d7ziem +.oa3z +.ob3ć2 +.ob3ł2 +.ob3ś2 +.ob3ź2 +.ob3ż2 +.ob3c2 +.ob3d2 +.ob3f2 +.ob3g2 +.ob3h2 +.ob3j2 +.ob3k2 +.ob3l2 +.ob3m2 +.ob3n2 +.ob3o2strz +.ob3p2 +.ob3r +.ob3s2 +.ob3t2 +.ob3u2m2 +.ob3w2 +.obch2 +.obcz2 +.obdź2 +.obdż2 +.obdz2 +.obe3ć2 +.obe3ł2 +.obe3ś2 +.obe3ź2 +.obe3ż2 +.obe3b2 +.obe3c2 +.obe3d2 +.obe3f2 +.obe3g2 +.obe3h2 +.obe3k2 +.obe3l2 +.obe3m2 +.obe3p2 +.obe3r2 +.obe3r3t +.obe3s2 +.obe3t2 +.obe3w2 +.obe3z2 +.obe4c3n +.obe4z3w +.obech2 +.obecz2 +.obedź2 +.obedż2 +.obedz2 +.oberż2 +.ober3m +.oberz2 +.obesch2 +.obesz2 +.obetk2 +.obi3b2 +.obsz2 +.oc2 +.och2 +.ochrz2 +.ocz2 +.odź2 +.od3ć2 +.od3ś2 +.od3au +.od3b2 +.od3c2 +.od3d2 +.od3f2 +.od3g2 +.od3h2 +.od3i2 +.od3i2zo +.od3j2 +.od3k2 +.od3l2 +.od3m2 +.od3n2 +.od3o2s +.od3p2 +.od3r2 +.od3s2 +.od3t2 +.od3u2cz +.od3u2m2 +.od3w2 +.od5ż2 +.od5z2 +.odbe2z3 +.odch2 +.odcz2 +.oddź2 +.oddż2 +.oddz2 +.ode3ć2 +.ode3ł2 +.ode3ś2 +.ode3ź2 +.ode3ż2 +.ode3b2 +.ode3c2 +.ode3d2 +.ode3f2 +.ode3g2 +.ode3h2 +.ode3k2 +.ode3l2 +.ode3m2 +.ode3mk2 +.ode3p2 +.ode3r2 +.ode3s2 +.ode3t2 +.ode3w2 +.ode3z2 +.odech2 +.odecz2 +.odedź2 +.odedż2 +.odedz2 +.odepch2 +.oderż2 +.oderz2 +.odesz2 +.odetch2 +.odetk2 +.odkrz2 +.odrz2 +.odsz2 +.of2 +.ogólno3k2 +.og2 +.ognio3tr2 +.oh2 +.ok2 +.oka3m2 +.okr2 +.ole2o3 +.om2 +.op2 +.opch2 +.or2ż2 +.or2tę +.or2z2 +.os2 +.osie2m3 +.osiemse2t3 +.osz2 +.ot2 +.ow2 +.oz2 +.pć8 +.pł8 +.płasko3w2 +.pń8 +.półk2 +.półkr2 +.półm2 +.póło2 +.półob3r +.półom2d +.półprzy3m2k +.pó2ł3 +.pó3ł4ą +.pó3ł4ę +.pó3ł4ecz +.pó3ł4y +.pś8 +.pź8 +.pż8 +.p8 +.pb8 +.pc8 +.pch8 +.pd8 +.pełno3kr2 +.pe2r3 +.pe3c2k +.pe3r4e +.pe3r4i +.pe3r4o +.pe3r4u +.pe3r4y +.pe4r5i2n +.pee2se2l +.pepee2r +.pepee2s +.peze2t1pee2r +.pf8 +.pg8 +.ph8 +.pięćse2t3 +.pię2ć3 +.pięcio3ś2 +.pierwo3w2 +.piono3w2 +.pj8 +.pk8 +.pl8 +.pm8 +.pn8 +.połk2 +.po2d2 +.po3ć2 +.po3ł2 +.po3ś2 +.po3ź2 +.po3ż2 +.po3b2 +.po3c2 +.po3dą +.po3dę +.po3dź2 +.po3d4łu +.po3d4much +.po3d4naw +.po3d4ręcz +.po3d4rętw +.po3d4róż +.po3d4r2wi +.po3d4raż +.po3d4rap +.po3d4repcz +.po3d4rept +.po3d4roż +.po3d4robó +.po3d4roba +.po3d4robo +.po3d4roby +.po3d4rocz +.po3d4ruzg +.po3d4ryg +.po3d4rze +.po3d4wójn +.po3d4wór +.po3d4waj +.po3d4woi +.po3d4woj +.po3d4worz +.po3da +.po3de +.po3dej +.po3diu +.po3do +.po3du +.po3dy +.po3dz2 +.po3e2k2s3 +.po3f2 +.po3g2 +.po3h2 +.po3k2 +.po3l2 +.po3m2 +.po3p2 +.po3rż +.po3r2 +.po3s2 +.po3t2 +.po3w2 +.po3z2 +.po4ń3c +.po4cz3d +.po4cz3t +.po4d3ów +.po4d3e4k2s3 +.po4d3o2bóz +.po4d3o2biad +.po4d3o2bojcz +.po4d3o2braz +.po4d3o2choc +.po4d3o2dm +.po4d3o2f +.po4d3o2g +.po4d3o2kien +.po4d3o2kn +.po4d3o2kręg +.po4d3o2kres +.po4d3o2piecz +.po4d3o2ryw +.po4d3o2siniak +.po4d3o2strz +.po4d3obsz +.po4d3odd +.po4d3olbrz +.po4d3u2cz +.po4d3u2dz +.po4d3u2pa +.po4d3u2ral +.po4d3u2sta +.po4d3u2szcz +.po4d5ręczn +.po4d5zakr +.po4d5zam +.po4d5zast +.po4d5zbi +.po4d5ze +.po4d5zielenią +.po4d5zielenić +.po4d5zielenię +.po4d5zielenił +.po4d5zielenic +.po4d5zielenien +.po4d5zielenil +.po4d5zielenim +.po4d5zielenio +.po4d5zielenis +.po4d5ziem +.po4d5ziom +.po4d5zw2r +.po4l3s +.po4m3p +.po4r3c +.po4r3f +.po4r3n +.po4r3t +.po4st3d +.po4st3f +.po4st3g +.po4st3h +.po4st3i2 +.po4st3k +.po4st3l +.po4st3m +.po4st3p +.po4st3rom +.po4st3s +.po5d4uszczyn +.po5r4tę +.pobr2 +.pobrz2 +.poch2 +.pochrz2 +.pocz2 +.pod3ć2 +.pod3ł2 +.pod3ś2 +.pod3śró2d5 +.pod3alp +.pod3b2 +.pod3c2 +.pod3d2 +.pod3f2 +.pod3g2 +.pod3h2 +.pod3i2n +.pod3j2 +.pod3k2 +.pod3l2 +.pod3m2 +.pod3n2 +.pod3p2 +.pod3r2 +.pod3s2 +.pod3t2 +.pod3w2 +.pod5ż2 +.podch2 +.podcz2 +.poddź2 +.poddż2 +.pode3ć2 +.pode3ł2 +.pode3ś2 +.pode3ź2 +.pode3ż2 +.pode3b2 +.pode3c2 +.pode3d2 +.pode3f2 +.pode3g2 +.pode3h2 +.pode3k2 +.pode3l2 +.pode3m2 +.pode3p2 +.pode3r2 +.pode3s2 +.pode3t2 +.pode3tk2 +.pode3w2 +.pode3z2 +.podech2 +.podecz2 +.podedź2 +.podedż2 +.podedz2 +.podepch2 +.poderż2 +.poderz2 +.podesch2 +.podesz2 +.podro2z3 +.podsm2 +.podsz2 +.pogrz2 +.pokl2 +.pokr2 +.pom4pk +.pomk2 +.pona2d2 +.pona3ć2 +.pona3ł2 +.pona3ś2 +.pona3ź2 +.pona3ż2 +.pona3b2 +.pona3c2 +.pona3cz2 +.pona3dź2 +.pona3do +.pona3f2 +.pona3g2 +.pona3h2 +.pona3k2 +.pona3l2 +.pona3m2 +.pona3p2 +.pona3r2 +.pona3s2 +.pona3t2 +.pona3w2 +.pona3z2 +.pona4f3t +.ponabrz2 +.ponach2 +.ponad3ć2 +.ponad3ś2 +.ponad3c2 +.ponad3ch2 +.ponad3cz2 +.ponad3dź2 +.ponad3f2 +.ponad3g2 +.ponad3h2 +.ponad3j2 +.ponad3k2 +.ponad3l2 +.ponad3p2 +.ponad3s2 +.ponad3t2 +.ponadz2 +.ponarz2 +.ponasm2 +.ponasz2 +.ponaz3m2 +.ponazw2 +.ponie3k2 +.ponie3w2 +.popch2 +.popo3w2 +.poprz2 +.por4t1w +.por4tf +.por4tm +.poro2z3 +.poro3z4u +.porz2 +.posch2 +.posm2 +.posz2 +.potk2 +.potr2 +.poz4m2 +.poza3u2 +.pozw2 +.pp8 +.pr8 +.pra3s2 +.pra3w2nu +.pra3w2z +.prapra3w2nu +.predy2s3po +.prz8 +.przełk2 +.prze2d2 +.prze3ć2 +.prze3ł2 +.prze3ś2 +.prze3ź2 +.prze3ż2 +.prze3b2 +.prze3c2 +.prze3dą +.prze3dę +.prze3dź2 +.prze3d4łuż +.prze3d4much +.prze3d4o3br +.prze3d4o3st +.prze3d4o3zo +.prze3d4ramat +.prze3d4ruk +.prze3d4ryl +.prze3d4rz2 +.prze3d4um +.prze3dy +.prze3dz2 +.prze3e2k2s3 +.prze3f2 +.prze3g2 +.prze3h2 +.prze3k2 +.prze3l2 +.prze3m2 +.prze3n2 +.prze3p2 +.prze3r2 +.prze3s2 +.prze3t2 +.prze3u2 +.prze3w2 +.prze3z2 +.prze4d5łużyc +.prze4d5ż2 +.prze4d5o4stat +.prze4d5za +.prze4d5zg2 +.prze4d5zim +.prze4d5zj +.prze4d5zl +.prze4d5zw2r +.prze4d5zwoj +.przebr2 +.przebrz2 +.przech2 +.przechrz2 +.przeci2w3 +.przeci3w4ie +.przeciwa2 +.przeciww2 +.przecz2 +.przed3ć2 +.przed3ł2 +.przed3ś2 +.przed3a2gon +.przed3a2kc +.przed3alp +.przed3b2 +.przed3c2 +.przed3d2 +.przed3e2gz +.przed3e2mer +.przed3f2 +.przed3g2 +.przed3h2 +.przed3i2 +.przed3j2 +.przed3k2 +.przed3l2 +.przed3m2 +.przed3n2 +.przed3o2 +.przed3p2 +.przed3r2 +.przed3s2 +.przed3się3w2 +.przed3sz2 +.przed3t2 +.przed3u2 +.przed3w2 +.przedch2 +.przedcz2 +.przeddź2 +.przeddż2 +.przeddz2 +.przedgrz2 +.przedy2s3ku +.przegrz2 +.przekl2 +.przekr2 +.przemk2 +.przepch2 +.przerż2 +.przerz2 +.przesch2 +.przesm2 +.przesz2 +.przetk2 +.przetr2 +.przetran2s3 +.przy3ć2 +.przy3ł2 +.przy3ś2 +.przy3ź2 +.przy3ż2 +.przy3b2 +.przy3c2 +.przy3d2 +.przy3f2 +.przy3g2 +.przy3h2 +.przy3k2 +.przy3l2 +.przy3m2 +.przy3p2 +.przy3r2 +.przy3s2 +.przy3t2 +.przy3w2 +.przy3z2 +.przybr2 +.przych2 +.przycz2 +.przydź2 +.przydż2 +.przydz2 +.przygrz2 +.przymk2 +.przyoz2 +.przypch2 +.przyrż2 +.przyrz2 +.przysch2 +.przysz2 +.przytk2 +.ps8 +.pt8 +.pv8 +.pw8 +.px8 +.pz8 +.rć8 +.rł8 +.rń8 +.rś8 +.rź8 +.rż8 +.r8 +.rb8 +.rc8 +.rd8 +.retran2s3 +.rf8 +.rg8 +.rh8 +.rj8 +.rk8 +.rl8 +.rm8 +.rn8 +.ro2z3 +.ro3z4a +.ro3z4e +.ro3z4e3ć2 +.ro3z4e3ł2 +.ro3z4e3ś2 +.ro3z4e3ź2 +.ro3z4e3ż2 +.ro3z4e3b2 +.ro3z4e3c2 +.ro3z4e3d2 +.ro3z4e3f2 +.ro3z4e3g2 +.ro3z4e3h2 +.ro3z4e3k2 +.ro3z4e3l2 +.ro3z4e3m2 +.ro3z4e3p2 +.ro3z4e3r2 +.ro3z4e3s2 +.ro3z4e3t2 +.ro3z4e3w2 +.ro3z4e3z2 +.ro3z4ej +.ro3z4u +.ro4z5a2gi +.ro4z5a2nie +.ro4z5e2mo +.ro4z5e4g3z +.ro4z5e4n3t +.rozś2 +.rozbrz2 +.rozd2 +.rozech2 +.rozecz2 +.rozedź2 +.rozedż2 +.rozedz2 +.rozepch2 +.rozerż2 +.rozerz2 +.rozesch2 +.rozesz2 +.rozi2 +.rozm2 +.rozo2 +.rozpo3w2 +.rozt2 +.roztr2 +.rozw2 +.rp8 +.rr8 +.rs8 +.rt8 +.rv8 +.rw8 +.rx8 +.rz8 +.sć8 +.sł8 +.sń8 +.sś8 +.sź8 +.sż8 +.s8 +.samo3ch2 +.samo3k2 +.samo3p2 +.samo3w2 +.samoro2z3 +.sb8 +.sc8 +.sch8 +.sd8 +.sf8 +.sg8 +.sh8 +.siede2m3 +.siedemse2t3 +.siedmio3ś2 +.sj8 +.ską2d5że +.sk8 +.skl8 +.skr8 +.sl8 +.sm8 +.sn8 +.sobo3w2 +.spó2ł3 +.sp8 +.spo2d2 +.spo3ć2 +.spo3ł2 +.spo3ś2 +.spo3ź2 +.spo3ż2 +.spo3b2 +.spo3c2 +.spo3dz2 +.spo3f2 +.spo3g2 +.spo3h2 +.spo3k2 +.spo3l2 +.spo3m2 +.spo3p2 +.spo3r2 +.spo3s2 +.spo3t2 +.spo3w2 +.spo3z2 +.spo4r3n +.spo4r3t +.spoch2 +.spocz2 +.spodź2 +.spodż2 +.spod3d +.sporz2 +.sposz2 +.sr8 +.ss8 +.st8 +.stere2o3 +.stereoa2 +.stereoe2 +.stereoi2 +.stereoo2 +.stereou2 +.su2b3 +.su3b4ie +.su3b4otn +.supe2r3 +.supe3r4at +.supe3r4io +.supe4r5a2tr +.super5z2b +.supere2 +.supero2d1rzut +.sv8 +.sw8 +.sx8 +.sz8 +.sześćse2t3 +.sześcio3ś2 +.sze2ś2ć3 +.sze2s3 +.tć8 +.tł8 +.tń8 +.tś8 +.tź8 +.tż8 +.t8 +.ta2o3 +.ta2r7zan +.tb8 +.tc8 +.tch8 +.td8 +.te2o3 +.tf8 +.tg8 +.th8 +.tj8 +.tk8 +.tl8 +.tm8 +.tn8 +.toa3 +.tp8 +.tró2j3 +.tró3j4ą +.tró3j4ę +.tró3j4ecz +.tr8 +.tran2s3 +.tran3s4e +.tran3s4ie +.tran3s4y +.tran3sz +.tran4s5eu +.transa2 +.transo2 +.trz8 +.trze2ch3 +.trzechse2t3 +.ts8 +.tt8 +.tv8 +.tw8 +.tx8 +.tysią2c3 +.tysią3c4a +.tysią3c4e +.tysią3cz +.tysią4c5zł +.tz8 +.uć2 +.uś2 +.u3ł2 +.u3ź2 +.u3ż2 +.u3b2 +.u3c2 +.u3d2 +.u3f2 +.u3g2 +.u3h2 +.u3k2 +.u3l2 +.u3m2 +.u3n2 +.u3p2 +.u3r2 +.u3s2 +.u3t2 +.u3w2 +.u3z2 +.u4d3k +.u4f3n +.u4k3lej +.u4l3s +.u4l3t +.u4m3br +.u4n3c +.u4n3d +.u4p3p2s +.u4r3s +.u4st3n +.u4stc +.u4stk +.u4z3be +.ube2z3 +.ubezw2 +.ubr2 +.uch2 +.ucz2 +.udź2 +.udż2 +.udz2 +.ukr2 +.umk2 +.upch2 +.upo2d2 +.upo3ć2 +.upo3ł2 +.upo3ś2 +.upo3ź2 +.upo3ż2 +.upo3b2 +.upo3c2 +.upo3da +.upo3f2 +.upo3g2 +.upo3h2 +.upo3k2 +.upo3l2 +.upo3m2 +.upo3p2 +.upo3r2 +.upo3s2 +.upo3t2 +.upo3w2 +.upo3z2 +.upoch2 +.upocz2 +.upodź2 +.upodż2 +.upod3d +.uporz2 +.uposz2 +.urż2 +.uro2z3 +.urz2 +.usch2 +.usz2 +.utk2 +.utr2 +.uze3w2 +.vć8 +.vł8 +.vń8 +.vś8 +.vź8 +.vż8 +.v8 +.vb8 +.vc8 +.vd8 +.vf8 +.vg8 +.vh8 +.vj8 +.vk8 +.vl8 +.vm8 +.vn8 +.vp8 +.vr8 +.vs8 +.vt8 +.vv8 +.vw8 +.vx8 +.vz8 +.wć8 +.wł8 +.wń8 +.wś8 +.wź8 +.wż8 +.w8 +.wb8 +.wc8 +.wd8 +.we3ć2 +.we3ł2 +.we3ś2 +.we3ż2 +.we3b2 +.we3c2 +.we3d2 +.we3f2 +.we3g2 +.we3h2 +.we3k2 +.we3l2 +.we3m2 +.we3n2 +.we3p2 +.we3r2 +.we3s2 +.we3t2 +.we3w2 +.we3z2 +.we4ł3n +.we4k3t +.we4l3w +.we4n3d +.we4n3t +.we4r3b +.we4r3d +.we4r3n +.we4r3s +.we4r3t +.we4s3prz +.we4s3tch2 +.we4z3br +.we4z3gł +.wech2 +.wecz2 +.wedź2 +.wedż2 +.wedz2 +.wemk2 +.wepch2 +.werz2 +.wesz2 +.wetk2 +.wewną2trz3 +.wf8 +.wg8 +.wh8 +.wielo3ś2 +.wielo3d2 +.wielo3k2 +.wieluse2t3 +.wilczo3m2 +.wj8 +.wk8 +.wl8 +.wm8 +.wn8 +.wniebo3w2 +.wodo3w2 +.wp8 +.wr8 +.ws8 +.współi2 +.współo2b3w +.współu2 +.współw2 +.wspó2ł3 +.wsze2ch3 +.wszecho2 +.wszechw2 +.wt8 +.wv8 +.ww8 +.wx8 +.wy3ć2 +.wy3ł2 +.wy3ś2 +.wy3ź2 +.wy3ż2 +.wy3b2 +.wy3c2 +.wy3d2 +.wy3f2 +.wy3g2 +.wy3h2 +.wy3k2 +.wy3l2 +.wy3m2 +.wy3o2d3r +.wy3p2 +.wy3r2 +.wy3s2 +.wy3t2 +.wy3w2 +.wy3z2 +.wy4ż3sz +.wy4cz3ha +.wybr2 +.wybrz2 +.wych2 +.wycz2 +.wydź2 +.wydż2 +.wydr2 +.wydz2 +.wye2k2s3 +.wygrz2 +.wyi2zo +.wykl2 +.wykr2 +.wykrz2 +.wymk2 +.wypch2 +.wyprz2 +.wyrż2 +.wyrz2 +.wysch2 +.wysm2 +.wysz2 +.wytch2 +.wytk2 +.wytr2 +.wz8 +.xć8 +.xł8 +.xń8 +.xś8 +.xź8 +.xż8 +.x8 +.xb8 +.xc8 +.xd8 +.xf8 +.xg8 +.xh8 +.xj8 +.xk8 +.xl8 +.xm8 +.xn8 +.xp8 +.xr8 +.xs8 +.xt8 +.xv8 +.xw8 +.xx8 +.xz8 +.zć8 +.zł8 +.zło3w2 +.zń8 +.zś8 +.zź8 +.zż8 +.z8 +.za3ć2 +.za3ł2 +.za3ś2 +.za3ź2 +.za3ż2 +.za3b2 +.za3c2 +.za3d2 +.za3f2 +.za3g2 +.za3h2 +.za3k2 +.za3l2 +.za3m2 +.za3o2b3r +.za3o2b3s +.za3p2 +.za3r2 +.za3s2 +.za3t2 +.za3u2 +.za3w2 +.za3z2 +.za4k3t +.za4l3g +.za4l3k +.za4l3t +.za4m3k +.za4r3ch +.za4uto +.za5m4k2n +.zabr2 +.zabrz2 +.zach2 +.zacz2 +.zadź2 +.zadż2 +.zadośću4 +.zado2ść3 +.zadr2 +.zady2s3po +.zadz2 +.zagrz2 +.zai2n3 +.zai2zo +.zain4ic +.zakl2 +.zakr2 +.zakrz2 +.zanie3d2 +.zarż2 +.zarz2 +.zasch2 +.zasm2 +.zasz2 +.zatk2 +.zatr2 +.zb8 +.zc8 +.zd8 +.zde2z3 +.zde3z4awu +.zde3z4el +.zde3z4er +.zde3z4y +.zdy2s3kont +.zdy2s3kred +.zdy2s3kwal +.ze3ć2 +.ze3ł2 +.ze3ś2 +.ze3ź2 +.ze3ż2 +.ze3b2 +.ze3c2 +.ze3d2 +.ze3f2 +.ze3g2 +.ze3h2 +.ze3k2 +.ze3l2 +.ze3m2 +.ze3p2 +.ze3r2 +.ze3s2 +.ze3t2 +.ze3tk2 +.ze3w2 +.ze3z2 +.ze4r3k +.ze4t3e2m1e2s +.ze4t3e2s1e2l +.ze4t3emp +.ze4t3hap +.zech2 +.zecz2 +.zedź2 +.zedż2 +.zedz2 +.zekl2 +.zepch2 +.zerż2 +.zerz2 +.zesch2 +.zesm4 +.zesz2 +.zf8 +.zg8 +.zh8 +.zimno3kr2 +.zj8 +.zk8 +.zl8 +.zm8 +.zmartwy2ch3 +.zmartwychw2 +.zn8 +.znie3ć2 +.znie3ł2 +.znie3ń2 +.znie3ś2 +.znie3ź2 +.znie3ż2 +.znie3b2 +.znie3c2 +.znie3d2 +.znie3f2 +.znie3g2 +.znie3h2 +.znie3k2 +.znie3l2 +.znie3m2 +.znie3n2 +.znie3p2 +.znie3r2 +.znie3s2 +.znie3t2 +.znie3w2 +.znie3z2 +.znie4dź3 +.znie4m3c +.zniech2 +.zniecz2 +.zniedż2 +.zniedz2 +.znierz2 +.zniesz2 +.zo2o3 +.zp8 +.zr8 +.zro2z3 +.zro3z4u +.zs8 +.zt8 +.zv8 +.zw8 +.zx8 +.zz8 +ą1 +ę1 +ó1 +ó4w3cz +ś1c +ź2dź +1ś2ci +2ć1ń +2ć1ś +2ć1ź +2ć1ż +2ć1b +2ć1c +2ć1d +2ć1f +2ć1g +2ć1k +2ć1m +2ć1n +2ć1p +2ć1s +2ć1t +2ć1z +2ł1ć +2ł1ń +2ł1ś +2ł1ź +2ł1ż +2ł1b +2ł1c +2ł1d +2ł1f +2ł1g +2ł1h +2ł1j +2ł1k +2ł1l +2ł1m +2ł1n +2ł1p +2ł1r +2ł1s +2ł1t +2ł1w +2ł1z +2ń1ć +2ń1ł +2ń1ń +2ń1ś +2ń1ź +2ń1ż +2ń1b +2ń1c +2ń1d +2ń1f +2ń1g +2ń1h +2ń1j +2ń1k +2ń1l +2ń1m +2ń1n +2ń1p +2ń1r +2ń1s +2ń1t +2ń1w +2ń1z +2śćc +2ś1ś +2ś1ź +2ś1ż +2ś1b +2ś1d +2ś1f +2ś1g +2ś1k +2ś1p +2ś1s +2ś1t +2ś1z +2ślm +2śln +2ź1ć +2ź1ś +2ź1ż +2ź1b +2ź1c +2ź1d +2ź1f +2ź1g +2ź1k +2ź1l +2ź1m +2ź1n +2ź1p +2ź1s +2ź1t +2ź1w +2ź1z +2ż1ć +2ż1ł +2ż1ń +2ż1ś +2ż1ź +2ż1b +2ż1c +2ż1d +2ż1f +2ż1g +2ż1j +2ż1k +2ż1l +2ż1m +2ż1n +2ż1p +2ż1r +2ż1s +2ż1t +2ż1w +2ż1z +2błk +2b1ć +2b1ń +2b1ś +2b1ź +2b1ż +2b1c +2b1d +2b1f +2b1g +2b1k +2b1m +2b1n +2b1p +2b1s +2b1t +2b1z +2brn +2c1ć +2c1ń +2c1ś +2c1ź +2c1ż +2c1b +2c1d +2c1f +2c1g +2c1k +2c1l +2c1m +2c1n +2c1p +2c1s +2c1t +2ch1ć +2ch1ń +2ch1ś +2ch1ź +2ch1ż +2ch1b +2ch1c +2ch1d +2ch1f +2ch1g +2ch1k +2ch1m +2ch1n +2ch1p +2ch1s +2ch1t +2ch1z +2cz1ć +2cz1ń +2cz1ś +2cz1ź +2cz1ż +2cz1b +2cz1c +2cz1d +2cz1f +2cz1g +2cz1k +2cz1l +2cz1m +2cz1n +2cz1p +2cz1s +2cz1t +2cz1z +2dłb +2dłsz +2dź1ć +2dź1ń +2dź1ś +2dź1ź +2dź1ż +2dź1b +2dź1c +2dź1d +2dź1f +2dź1g +2dź1k +2dź1m +2dź1n +2dź1p +2dź1s +2dź1t +2dź1z +2dż1ć +2dż1ń +2dż1ś +2dż1ź +2dż1ż +2dż1b +2dż1c +2dż1d +2dż1f +2dż1g +2dż1k +2dż1m +2dż1n +2dż1p +2dż1s +2dż1t +2dż1z +2d1ć +2d1ń +2d1ś +2d1b +2d1c +2d1f +2d1g +2d1k +2d1m +2d1n +2d1p +2d1s +2d1t +2drn +2dz1ć +2dz1ń +2dz1ś +2dz1ź +2dz1ż +2dz1b +2dz1c +2dz1d +2dz1f +2dz1g +2dz1k +2dz1l +2dz1m +2dz1n +2dz1p +2dz1s +2dz1t +2dz1z +2f1c +2f1k +2f1m +2f1n +2głb +2g1ć +2g1ń +2g1ś +2g1ź +2g1ż +2g1b +2g1c +2g1d +2g1f +2g1k +2g1m +2g1p +2g1s +2g1t +2g1z +2h1ć +2h1ł +2h1ń +2h1ś +2h1ź +2h1ż +2h1b +2h1c +2h1d +2h1f +2h1g +2h1j +2h1k +2h1l +2h1m +2h1n +2h1p +2h1r +2h1s +2h1t +2h1w +2h1z +2j1ć +2j1ł +2j1ń +2j1ś +2j1ź +2j1ż +2j1b +2j1c +2j1d +2j1f +2j1g +2j1h +2j1k +2j1l +2j1m +2j1n +2j1p +2j1r +2j1s +2j1t +2j1w +2j1z +2kłb +2k1ć +2k1ń +2k1ś +2k1ź +2k1ż +2k1b +2k1c +2k1d +2k1f +2k1g +2k1m +2k1n +2k1p +2k1s +2k1sz +2k1t +2k1z +2l1ć +2l1ł +2l1ń +2l1ś +2l1ź +2l1ż +2l1b +2l1c +2l1d +2l1f +2l1g +2l1h +2l1j +2l1k +2l1m +2l1n +2l1p +2l1r +2l1s +2l1t +2l1w +2l1z +2m1ć +2m1ł +2m1ń +2m1ś +2m1ź +2m1ż +2m1b +2m1c +2m1d +2m1f +2m1g +2m1h +2m1j +2m1k +2m1l +2m1n +2m1p +2m1r +2m1s +2m1t +2m1w +2m1z +2n1ć +2n1ł +2n1ń +2n1ś +2n1ź +2n1ż +2n1b +2n1c +2n1d +2n1f +2n1g +2n1h +2n1j +2n1k +2n1l +2n1m +2n1p +2n1r +2n1s +2n1t +2n1w +2n1z +2ntn +2p1ć +2p1ń +2p1ś +2p1ź +2p1ż +2p1b +2p1c +2p1d +2p1f +2p1g +2p1k +2p1m +2p1n +2p1s +2p1sz +2p1t +2p1z +2pln +2r1ć +2r1ł +2r1ń +2r1ś +2r1ź +2r1ż +2r1b +2r1c +2r1d +2r1f +2r1g +2r1h +2r1j +2r1k +2r1l +2r1m +2r1n +2r1p +2r1s +2r1t +2r1w +2rz1ć +2rz1ł +2rz1ń +2rz1ś +2rz1ź +2rz1ż +2rz1b +2rz1c +2rz1d +2rz1f +2rz1g +2rz1h +2rz1j +2rz1k +2rz1l +2rz1m +2rz1n +2rz1p +2rz1r +2rz1s +2rz1t +2rz1w +2słb +2s1ź +2s1ż +2s1b +2s1d +2s1f +2s1g +2s1s +2snk +2stk +2stn +2stsz +2sz1ć +2sz1ś +2sz1c +2sz1f +2sz1k +2sz1l +2sz1m +2sz1n +2sz1p +2sz1s +2sz1t +2sz1w +2sz1z +2szln +2t1ć +2t1ń +2t1ś +2t1ź +2t1ż +2t1b +2t1c +2t1d +2t1f +2t1g +2t1k +2t1m +2t1n +2t1p +2t1s +2t1z +2tln +2trk +2trzn +2w1ć +2w1ł +2w1ń +2w1ś +2w1ź +2w1ż +2w1b +2w1c +2w1d +2w1f +2w1g +2w1j +2w1k +2w1l +2w1m +2w1n +2w1p +2w1r +2w1s +2w1t +2w1z +2z1ć +2z1ś +2z1c +2z1d +2z1f +2z1k +2z1p +2z1s +2z1t +2zdk +2zdn +3d2niow +3k2sz2t +3m2k2n +3m2nest +3m2nezj +3m2sk2n +3p2neu +3w2ład +3w2łos +3w2czas +4ć3ć +4ł3ł +4ź3ź +4ż3ż +4b3b +4c3c +4d3d +4f3f +4g3g +4h3h +4j3j +4k3k +4l3l +4m3m +4n3n +4p3p +4r3r +4t3t +4w3w +4z3z +8ć. +8ćć. +8ćł. +8ćń. +8ćś. +8ćź. +8ćż. +8ćb. +8ćc. +8ćd. +8ćf. +8ćg. +8ćh. +8ćj. +8ćk. +8ćl. +8ćm. +8ćn. +8ćp. +8ćr. +8ćs. +8ćt. +8ćv. +8ćw. +8ćx. +8ćz. +8ł. +8łć. +8łł. +8łń. +8łś. +8łź. +8łż. +8łb. +8łc. +8łd. +8łf. +8łg. +8łh. +8łj. +8łk. +8łl. +8łm. +8łn. +8łp. +8łr. +8łs. +8łt. +8łv. +8łw. +8łx. +8łz. +8ń. +8ńć. +8ńł. +8ńń. +8ńś. +8ńź. +8ńż. +8ńb. +8ńc. +8ńd. +8ńf. +8ńg. +8ńh. +8ńj. +8ńk. +8ńl. +8ńm. +8ńn. +8ńp. +8ńr. +8ńs. +8ńt. +8ńv. +8ńw. +8ńx. +8ńz. +8ś. +8ść. +8śł. +8śń. +8śś. +8śź. +8śż. +8śb. +8śc. +8śd. +8śf. +8śg. +8śh. +8śj. +8śk. +8śl. +8śm. +8śn. +8śp. +8śr. +8śs. +8śt. +8śv. +8św. +8śx. +8śz. +8ź. +8źć. +8źł. +8źń. +8źś. +8źź. +8źż. +8źb. +8źc. +8źd. +8źf. +8źg. +8źh. +8źj. +8źk. +8źl. +8źm. +8źn. +8źp. +8źr. +8źs. +8źt. +8źv. +8źw. +8źx. +8źz. +8ż. +8żć. +8żł. +8żń. +8żś. +8żź. +8żż. +8żb. +8żc. +8żd. +8żf. +8żg. +8żh. +8żj. +8żk. +8żl. +8żm. +8żn. +8żp. +8żr. +8żs. +8żt. +8żv. +8żw. +8żx. +8żz. +8b. +8bć. +8bł. +8bń. +8bś. +8bź. +8bż. +8bb. +8bc. +8bd. +8bf. +8bg. +8bh. +8bj. +8bk. +8bl. +8bm. +8bn. +8bp. +8br. +8brz. +8bs. +8bt. +8bv. +8bw. +8bx. +8bz. +8c. +8cć. +8cł. +8cń. +8cś. +8cź. +8cż. +8cb. +8cc. +8cd. +8cf. +8cg. +8ch. +8chł. +8chrz. +8chw. +8cj. +8ck. +8cl. +8cm. +8cn. +8cp. +8cr. +8cs. +8ct. +8cv. +8cw. +8cx. +8cz. +8czt. +8d. +8dć. +8dł. +8dń. +8dś. +8dź. +8dż. +8db. +8dc. +8dd. +8df. +8dg. +8dh. +8dj. +8dk. +8dl. +8dm. +8dn. +8dp. +8dr. +8drz. +8ds. +8dt. +8dv. +8dw. +8dx. +8dz. +8f. +8fć. +8fł. +8fń. +8fś. +8fź. +8fż. +8fb. +8fc. +8fd. +8ff. +8fg. +8fh. +8fj. +8fk. +8fl. +8fm. +8fn. +8fp. +8fr. +8fs. +8ft. +8fv. +8fw. +8fx. +8fz. +8g. +8gć. +8gł. +8gń. +8gś. +8gź. +8gż. +8gb. +8gc. +8gd. +8gf. +8gg. +8gh. +8gj. +8gk. +8gl. +8gm. +8gn. +8gp. +8gr. +8gs. +8gt. +8gv. +8gw. +8gx. +8gz. +8h. +8hć. +8hł. +8hń. +8hś. +8hź. +8hż. +8hb. +8hc. +8hd. +8hf. +8hg. +8hh. +8hj. +8hk. +8hl. +8hm. +8hn. +8hp. +8hr. +8hs. +8ht. +8hv. +8hw. +8hx. +8hz. +8j. +8jć. +8jł. +8jń. +8jś. +8jź. +8jż. +8jb. +8jc. +8jd. +8jf. +8jg. +8jh. +8jj. +8jk. +8jl. +8jm. +8jn. +8jp. +8jr. +8js. +8jt. +8jv. +8jw. +8jx. +8jz. +8k. +8kć. +8kł. +8kń. +8kś. +8kź. +8kż. +8kb. +8kc. +8kd. +8kf. +8kg. +8kh. +8kj. +8kk. +8kl. +8km. +8kn. +8kp. +8kr. +8ks. +8kst. +8kt. +8kv. +8kw. +8kx. +8kz. +8l. +8lć. +8lł. +8lń. +8lś. +8lź. +8lż. +8lb. +8lc. +8ld. +8lf. +8lg. +8lh. +8lj. +8lk. +8ll. +8lm. +8ln. +8lp. +8lr. +8ls. +8lt. +8lv. +8lw. +8lx. +8lz. +8m. +8mć. +8mł. +8mń. +8mś. +8mź. +8mż. +8mb. +8mc. +8md. +8mf. +8mg. +8mh. +8mj. +8mk. +8ml. +8mm. +8mn. +8mp. +8mr. +8ms. +8mst. +8mt. +8mv. +8mw. +8mx. +8mz. +8n. +8nć. +8nł. +8nń. +8nś. +8nź. +8nż. +8nb. +8nc. +8nd. +8nf. +8ng. +8nh. +8nj. +8nk. +8nl. +8nm. +8nn. +8np. +8nr. +8ns. +8nt. +8nv. +8nw. +8nx. +8nz. +8p. +8pć. +8pł. +8pń. +8pś. +8pź. +8pż. +8pb. +8pc. +8pd. +8pf. +8pg. +8ph. +8pj. +8pk. +8pl. +8pm. +8pn. +8pp. +8pr. +8prz. +8ps. +8pt. +8pv. +8pw. +8px. +8pz. +8r. +8rć. +8rł. +8rń. +8rś. +8rź. +8rż. +8rb. +8rc. +8rd. +8rf. +8rg. +8rh. +8rj. +8rk. +8rl. +8rm. +8rn. +8rp. +8rr. +8rs. +8rsz. +8rt. +8rv. +8rw. +8rx. +8rz. +8rzł. +8s. +8sć. +8sł. +8sń. +8sś. +8sź. +8sż. +8sb. +8sc. +8sch. +8sd. +8sf. +8sg. +8sh. +8sj. +8sk. +8skrz. +8sl. +8sm. +8sn. +8sp. +8sr. +8ss. +8st. +8str. +8strz. +8stw. +8sv. +8sw. +8sx. +8sz. +8szcz. +8szczb. +8szk. +8szn. +8szt. +8sztr. +8t. +8tć. +8tł. +8tń. +8tś. +8tź. +8tż. +8tb. +8tc. +8td. +8tf. +8tg. +8th. +8tj. +8tk. +8tl. +8tm. +8tn. +8tp. +8tr. +8trz. +8ts. +8tt. +8tv. +8tw. +8tx. +8tz. +8v. +8vć. +8vł. +8vń. +8vś. +8vź. +8vż. +8vb. +8vc. +8vd. +8vf. +8vg. +8vh. +8vj. +8vk. +8vl. +8vm. +8vn. +8vp. +8vr. +8vs. +8vt. +8vv. +8vw. +8vx. +8vz. +8w. +8wć. +8wł. +8wń. +8wś. +8wź. +8wż. +8wb. +8wc. +8wd. +8wf. +8wg. +8wh. +8wj. +8wk. +8wl. +8wm. +8wn. +8wp. +8wr. +8ws. +8wt. +8wv. +8ww. +8wx. +8wz. +8x. +8xć. +8xł. +8xń. +8xś. +8xź. +8xż. +8xb. +8xc. +8xd. +8xf. +8xg. +8xh. +8xj. +8xk. +8xl. +8xm. +8xn. +8xp. +8xr. +8xs. +8xt. +8xv. +8xw. +8xx. +8xz. +8z. +8zć. +8zł. +8zń. +8zś. +8zź. +8zż. +8zb. +8zc. +8zd. +8zdr. +8zdrz. +8zf. +8zg. +8zh. +8zj. +8zk. +8zl. +8zm. +8zn. +8zp. +8zr. +8zs. +8zt. +8zv. +8zw. +8zx. +8zz. +a1 +a2u +a2y +aa2 +ae2 +ai2 +ao2 +be2eth +be2f3sz2 +be2k1hend +bi2n3o2ku +bi2sz3kop +bi2z3ne2s3m +bi2z3nes +birmin2g1ham +blo2k1hauz +bo2s3ma +br2d +bro2a2d3way +bu2sz3me +buk2sz3pan +busine2ss3m +busines2s +c4h +c4z +cal2d1well +ch2ł +ch2j +ch2l +ch2r +ch2w +chus1t +cu2r7zon +dż2ł +dż2j +dż2l +dż2r +dż2w +dże4z3b +dże4z3m +d4ź +d4ż +d4z +deut4sch3land +drz2w +du2sz3past +e1 +e2r5zac +e2u +e2y +e3u2sz +ea2 +ee2 +ei2 +eo2 +fi2s3harm +fi2sz3bin +fo2k2s3t +fo2r5zac +fol2k1lor +fos2f1a2zot +ga3d2get +gado3p2ta +gol2f3s +golfsz2 +gran2d1ilo +gro4t3r +hi2sz3p +hu2cz1w +hu2x3ley +i1 +i2ą +i2ę +i2ó +i2a +i2e +i2i +i2o +i2u +i2y +in4nsbruck +in4sbruc +j2t1ł +j2t1r +ja4z4z3b +ja4z4z3m +karl2s1kron +karl2s1ruhe +kir2chhoff +kongre2s3m +led1w +lu2ft3waffe +lu2ks1fer +ly2o +ma2r5zł +ma2r5zl +ma2r5zn +mi2sz1masz +mie2r5zł +mie2r5zi +mon2t3real +moza2i3k +mu2r7zasich3l +na4ł3kows +na4r3v +o1 +o2y +oa2 +och3mistrz +oe2 +of2f3set +oi2 +oo2 +ou2 +pa2n3a2mer +pa2s3cal +pa2s3ch +połu3d2ni +po3d4nieprz +po3m2ną +po3m2nę +po3m2ni +po4rt2s3mo2uth +po4rt3land +poli3e2t +poli3u2re +powsze3d2ni +pr2chal +pre2sz3pa +r4z +ro2e3nt2gen +ro2k3rocz +ro2s3to3c2k +s4z +se2t3le +sko2r5zoner +sm2r +sowi3z2 +sy2n3opt +sy2s1tem +sza2sz1ły +sze2z1long +sze4ść +szto2k1holm +szyn2k1was +to3y2o3t +turboo2d3rzut +tygo3d2ni +u1 +u2y +ua2 +ue2 +ui2 +uo2 +uu2 +vo2lk2s3 +we2e2k1end +we4st3f +we4st3m +y1 +ya2 +ye2 +yi2 +yo2 +yu2 +ze4p3p +} +\hyphenation{ +be-zach +be-zami +by-naj-mniej +gdzie-nie-gdzie +ina-czej +na-dal +ni-gdy +ni-gdzie +niech-że +niech-by +ow-szem +pó-łach +pó-łami +pó-łek +pod-ów-czas +przy-naj-mniej +skąd-inąd +tró-jach +tró-jami +tró-jek +} diff --git a/assets/hyphenation/pt.pat b/assets/hyphenation/pt.pat new file mode 100644 index 000000000..cb947c43b --- /dev/null +++ b/assets/hyphenation/pt.pat @@ -0,0 +1,503 @@ +% title: Hyphenation patterns for Portuguese +% copyright: Copyright (C) 1987, 1994, 1996, 2015 Pedro J. de Rezende, 1996, 2015 J. Joao Dias Almeida, 2024 Leonardo Araujo and Aline Benevides +% notice: This file is part of the hyph-utf8 package. +% See http://www.hyphenation.org/tex for more information. +% language: +% name: Portuguese +% tag: pt +% version: 1.4 2024-07-13 +% authors: +% - +% name: Pedro J. de Rezende +% contact: rezende (at) ic.unicamp.br +% - +% name: J. Joao Dias Almeida +% contact: jj (at) di.uminho.pt +% - +% name: Leonardo Araujo +% contact: leolca (at) gmail.com +% - +% name: Aline Benevides +% contact: benevides.aline12 (at) gmail.com +% licence: +% name: BSD 3-clause licence +% url: https://opensource.org/licenses/BSD-3-Clause +% text: > +% Redistribution and use in source and binary forms, with or without +% modification, are permitted provided that the following conditions +% are met: +% * Redistributions of source code must retain the above copyright +% notice, this list of conditions and the following disclaimer. +% * Redistributions in binary form must reproduce the above copyright +% notice, this list of conditions and the following disclaimer in the +% documentation and/or other materials provided with the +% distribution. +% * Neither the name of the University of Campinas, of the University +% of Minho nor the names of its contributors may be used to endorse +% or promote products derived from this software without specific +% prior written permission. +% +% THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +% "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +% LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +% A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL PEDRO J. DE +% REZENDE OR J.JOAO DIAS ALMEIDA BE LIABLE FOR ANY DIRECT, INDIRECT, +% INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, +% BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS +% OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED +% AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +% LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY +% WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +% POSSIBILITY OF SUCH DAMAGE. +% hyphenmins: +% typesetting: +% left: 2 +% right: 3 +% changes: +% - Version 1.4 Release date: 13/07/2024 Leonardo Araujo and Aline Benevides +% - Version 1.3 Release date: 12/08/2015 Pedro J. de Rezende and J. Joao Dias Almeida +% - Version 1.2 Release date: 07/21/1996 Pedro J. de Rezende and J. Joao Dias Almeida +% - Version 1.1 Release date: 04/12/1994 Pedro J. de Rezende +% - Version 1.0 Release date: 02/13/1987 Pedro J. de Rezende +% texlive: +% synonyms: +% - portuges +% encoding: ec +% babelname: portuguese +% legacy_patterns: pthyph.tex +% message: Portuguese hyphenation patterns +% description: Hyphenation patterns for Portuguese in T1/EC and UTF-8 encodings. +% +\patterns{ +1b2l +1b2r +1ba +1be +1bi +1bo +1bu +1bá +1bâ +1bã +1bé +1bí +1bó +1bú +1bê +1bõ +1c2h +1c2l +1c2r +1ca +1ce +1ci +1co +1cu +1cá +1câ +1cã +1cé +1cí +1có +1cú +1cê +1cõ +1ça +1çe +1çi +1ço +1çu +1çá +1çâ +1çã +1çé +1çí +1çó +1çú +1çê +1çõ +1d2l +1d2r +1da +1de +1di +1do +1du +1dá +1dâ +1dã +1dé +1dí +1dó +1dú +1dê +1dõ +1f2l +1f2r +1fa +1fe +1fi +1fo +1fu +1fá +1fâ +1fã +1fé +1fí +1fó +1fú +1fê +1fõ +1g2l +1g2r +1ga +1ge +1gi +1go +1gu +1gu4a +1gu4e +1gu4i +1gu4o +1gá +1gâ +1gã +1gé +1gí +1gó +1gú +1gê +1gõ +1ja +1je +1ji +1jo +1ju +1já +1jâ +1jã +1jé +1jí +1jó +1jú +1jê +1jõ +1k2l +1k2r +1ka +1ke +1ki +1ko +1ku +1ká +1kâ +1kã +1ké +1kí +1kó +1kú +1kê +1kõ +1l2h +1la +1le +1li +1lo +1lu +1lá +1lâ +1lã +1lé +1lí +1ló +1lú +1lê +1lõ +1ma +1me +1mi +1mo +1mu +1má +1mâ +1mã +1mé +1mí +1mó +1mú +1mê +1mõ +1n2h +1na +1ne +1ni +1no +1nu +1ná +1nâ +1nã +1né +1ní +1nó +1nú +1nê +1nõ +1p2l +1p2r +1pa +1pe +1pi +1po +1pu +1pá +1pâ +1pã +1pé +1pí +1pó +1pú +1pê +1põ +1qu4a +1qu4e +1qu4i +1qu4o +1ra +1re +1ri +1ro +1ru +1rá +1râ +1rã +1ré +1rí +1ró +1rú +1rê +1rõ +1sa +1se +1si +1so +1su +1sá +1sâ +1sã +1sé +1sí +1só +1sú +1sê +1sõ +1t2l +1t2r +1ta +1te +1ti +1to +1tu +1tá +1tâ +1tã +1té +1tí +1tó +1tú +1tê +1tõ +1v2l +1v2r +1va +1ve +1vi +1vo +1vu +1vá +1vâ +1vã +1vé +1ví +1vó +1vú +1vê +1võ +1w2l +1w2r +1xa +1xe +1xi +1xo +1xu +1xá +1xâ +1xã +1xé +1xí +1xó +1xú +1xê +1xõ +1za +1ze +1zi +1zo +1zu +1zá +1zâ +1zã +1zé +1zí +1zó +1zú +1zê +1zõ +a3a +a3e +a3o +c3c +e3a +e3e +e3o +i3a +i3e +i3i +i3o +i3â +i3ê +i3ô +o3a +o3e +o3o +r3r +s3s +u3a +u3e +u3o +u3u +.p2si +.p2sí +.g2no +.g2nó +.g2nô +t2c +1p2neu +.t2m +.p2t +su2b3r +su2b3l +.su3b4li +a1ir. +u1ir. +1qu +1vô +1lô +1cô +1gô +1bô +1tô +1rô +1pô +a1é +a1í +a1ó +a1ú +e1á +e1â +e1ã +e1é +e1ê +e1í +e1ó +é1o +e1ú +i1á +i1ã +í1a +i1é +i1í +i1ó +í1o +i1u +i1ú +o1á +o1é +o1í +o1ó +u1á +u1ã +u1â +u1í +ú1o +1qu2á +1qu2â +1qu2í +1gu2í +a1ind +a1i1nh +e1imp +e1inc +e1inf +e1ing +e1ins +e1int +e1inv +u1iz. +a1iz. +4a. +4e. +4o. +1gu2á +1gu2ã +1qu2ã +.m2n +c2za +.s2 +1p2seu1d +1dô +1fô +1mô +1nô +1sô +1zô +tu1i +tu2it +tu2id +bu1i +nu1i +o1in +u1in +su1i +í1e +ju1i +fu1i +du1i +do1im +au1i +u1i1ç +u1ê +1gu2ê +1qu2ê +1çô +u1é +1gu2é +1qu2é +1xô +a1â +a1ã +a1ô +e1ô +.ne4o +o1ã +o1ê +o1im +o2i1na +pro1i1b +co2ima +.t2 +1- +} +\hyphenation{% Do NOT make any alterations to this list! --- PdR +hard-ware +soft-ware +} diff --git a/contributors/people.toml b/contributors/people.toml index d6a5d204a..bd0aa39e8 100644 --- a/contributors/people.toml +++ b/contributors/people.toml @@ -38,6 +38,11 @@ real_name = "Dmitrii Khlapov" display_name = "Dima Khlapov" sync_github_name = false +[people.DanielKnourek] +real_name = "Daniel Kňourek" +display_name = "Daniel Kňourek" +sync_github_name = false + [people.arcanii] real_name = "Bryan Mark" display_name = "Arcanii" @@ -46,14 +51,22 @@ sync_github_name = false [people.henno] sync_github_name = true -[people.GhostMitra] +[people.yeniufuklar-com] +real_name = "Ufuk Efe" +display_name = "Ufuk Efe" sync_github_name = true + +[people.GhostMitra] [people.DevCoderMax] real_name = "Gleyson do Nascimento Carvalho" display_name = "Gleyson Carvalho" sync_github_name = false [people.alexanderalber] real_name = "Alexander Alber" - [people.edurbs] -sync_github_name = true +[people.manucarcor] +real_name = "Manuel Carmona" +display_name = "manucarcor" +sync_github_name = false + +[people.odedigo] diff --git a/crates/doc/src/bidi.rs b/crates/doc/src/bidi.rs index 5da0dab5c..e4c91835f 100644 --- a/crates/doc/src/bidi.rs +++ b/crates/doc/src/bidi.rs @@ -37,6 +37,26 @@ pub fn has_rtl(text: &str) -> bool { text.chars().any(is_rtl_script) } +/// Does the first strong directional character in `text` belong to a right-to-left script? +/// +/// Follows UAX #9 rules P2 and P3: characters with strong bidirectional class AL or R +/// (Arabic, Hebrew, Syriac, Thaana…) give `true`; strong L characters (Latin, Greek, +/// Cyrillic, CJK…) give `false`. Neutral, weak and whitespace characters are skipped. +pub fn first_strong_is_rtl(text: &str) -> bool { + for c in text.chars() { + if c == '\u{200F}' { + return true; + } + if c == '\u{200E}' { + return false; + } + if c.is_alphabetic() { + return is_rtl_script(c); + } + } + false +} + #[cfg(test)] mod tests { use super::*; @@ -56,5 +76,11 @@ mod tests { assert!("می‌خواهم".chars().all(is_word_char), "ZWNJ is inside the word"); assert!("کِتاب".chars().all(is_word_char), "so are harakat"); assert!(!is_word_char(' ') && !is_word_char('،')); + assert!(first_strong_is_rtl("سلام")); + assert!(first_strong_is_rtl("من ناوم هێڤاره")); + assert!(first_strong_is_rtl(" 123. سلام")); + assert!(!first_strong_is_rtl("Hello سلام")); + assert!(!first_strong_is_rtl("123 Hello")); + assert!(!first_strong_is_rtl("")); } } diff --git a/crates/doc/src/collate.rs b/crates/doc/src/collate.rs new file mode 100644 index 000000000..db034e8ac --- /dev/null +++ b/crates/doc/src/collate.rs @@ -0,0 +1,520 @@ +//! Language-aware sorting of text: a small, pure-Rust collation. +//! +//! Every letter gets a *primary* weight (its place in the alphabet) and a *secondary* one (its +//! diacritic). Strings compare by primary weights first, then by secondary ones, so `é` sorts +//! with `e` (`resume` < `résumé` < `rex`) unless the language makes it a letter of its own: +//! Polish puts `ą` right after `a` and `ż` after `ź` after `z`, Swedish puts `å ä ö` after `z`, +//! Czech puts `ch` after `h`, Spanish `ñ` after `n`, Hungarian `cs`, `gy`, `sz`… after their +//! first letter. The per-language tables follow the published alphabets of those languages. +//! Text without diacritics sorts exactly as a case-insensitive code-point compare. + +use std::cmp::Ordering; + +/// The diacritic of a precomposed lowercase Latin letter: `(base letters, mark)`. Marks: 1 acute, +/// 2 grave, 3 circumflex, 4 diaeresis, 5 tilde, 6 ring, 7 caron, 8 cedilla, 9 ogonek, 10 dot, +/// 11 stroke, 12 breve, 13 macron, 14 double acute, 15 comma below, 16 ligature, 17 dotless, +/// 18 eth. +const DECOMP: &[(char, &str, u16)] = &[ + ('à', "a", 2), + ('á', "a", 1), + ('â', "a", 3), + ('ã', "a", 5), + ('ä', "a", 4), + ('å', "a", 6), + ('ā', "a", 13), + ('ă', "a", 12), + ('ą', "a", 9), + ('æ', "ae", 16), + ('ç', "c", 8), + ('ć', "c", 1), + ('ĉ', "c", 3), + ('ċ', "c", 10), + ('č', "c", 7), + ('ď', "d", 7), + ('đ', "d", 11), + ('ð', "d", 18), + ('è', "e", 2), + ('é', "e", 1), + ('ê', "e", 3), + ('ë', "e", 4), + ('ē', "e", 13), + ('ĕ', "e", 12), + ('ė', "e", 10), + ('ę', "e", 9), + ('ě', "e", 7), + ('ĝ', "g", 3), + ('ğ', "g", 12), + ('ġ', "g", 10), + ('ģ', "g", 8), + ('ĥ', "h", 3), + ('ħ', "h", 11), + ('ì', "i", 2), + ('í', "i", 1), + ('î', "i", 3), + ('ï', "i", 4), + ('ĩ', "i", 5), + ('ī', "i", 13), + ('ĭ', "i", 12), + ('į', "i", 9), + ('ı', "i", 17), + ('ij', "ij", 16), + ('ĵ', "j", 3), + ('ķ', "k", 8), + ('ĺ', "l", 1), + ('ļ', "l", 8), + ('ľ', "l", 7), + ('ŀ', "l", 10), + ('ł', "l", 11), + ('ñ', "n", 5), + ('ń', "n", 1), + ('ņ', "n", 8), + ('ň', "n", 7), + ('ò', "o", 2), + ('ó', "o", 1), + ('ô', "o", 3), + ('õ', "o", 5), + ('ö', "o", 4), + ('ø', "o", 11), + ('ō', "o", 13), + ('ŏ', "o", 12), + ('ő', "o", 14), + ('œ', "oe", 16), + ('ŕ', "r", 1), + ('ŗ', "r", 8), + ('ř', "r", 7), + ('ś', "s", 1), + ('ŝ', "s", 3), + ('ş', "s", 8), + ('š', "s", 7), + ('ș', "s", 15), + ('ß', "ss", 16), + ('ţ', "t", 8), + ('ť', "t", 7), + ('ŧ', "t", 11), + ('ț', "t", 15), + ('ù', "u", 2), + ('ú', "u", 1), + ('û', "u", 3), + ('ü', "u", 4), + ('ũ', "u", 5), + ('ū', "u", 13), + ('ŭ', "u", 12), + ('ů', "u", 6), + ('ű', "u", 14), + ('ų', "u", 9), + ('ŵ', "w", 3), + ('ý', "y", 1), + ('ÿ', "y", 4), + ('ŷ', "y", 3), + ('ź', "z", 1), + ('ż', "z", 10), + ('ž', "z", 7), +]; + +/// A letter (or letter group) a language sorts as a letter of its own: right after `after` +/// (`rank` places after it), with secondary weight `sec` among letters of the same place. +struct Tailor { + seq: &'static str, + after: char, + rank: u32, + sec: u16, +} + +const fn t(seq: &'static str, after: char, rank: u32) -> Tailor { + Tailor { seq, after, rank, sec: 0 } +} + +const fn ts(seq: &'static str, after: char, rank: u32, sec: u16) -> Tailor { + Tailor { seq, after, rank, sec } +} + +const PL: &[Tailor] = + &[t("ą", 'a', 1), t("ć", 'c', 1), t("ę", 'e', 1), t("ł", 'l', 1), t("ń", 'n', 1), t("ó", 'o', 1), t("ś", 's', 1), t("ź", 'z', 1), t("ż", 'z', 2)]; +const CS: &[Tailor] = &[t("č", 'c', 1), t("ch", 'h', 1), t("ř", 'r', 1), t("š", 's', 1), t("ž", 'z', 1)]; +const SK: &[Tailor] = &[t("ä", 'a', 1), t("č", 'c', 1), t("ch", 'h', 1), t("ô", 'o', 1), t("š", 's', 1), t("ž", 'z', 1)]; +const HU: &[Tailor] = &[ + t("cs", 'c', 1), + t("dz", 'd', 1), + t("dzs", 'd', 2), + t("gy", 'g', 1), + t("ly", 'l', 1), + t("ny", 'n', 1), + t("ö", 'o', 1), + ts("ő", 'o', 1, 14), + t("sz", 's', 1), + t("ty", 't', 1), + t("ü", 'u', 1), + ts("ű", 'u', 1, 14), + t("zs", 'z', 1), +]; +const SV: &[Tailor] = &[t("å", 'z', 1), t("ä", 'z', 2), ts("æ", 'z', 2, 16), t("ö", 'z', 3), ts("ø", 'z', 3, 11)]; +const DA: &[Tailor] = &[t("æ", 'z', 1), ts("ä", 'z', 1, 4), t("ø", 'z', 2), ts("ö", 'z', 2, 4), t("å", 'z', 3)]; +const ES: &[Tailor] = &[t("ñ", 'n', 1)]; +const TR: &[Tailor] = &[t("ç", 'c', 1), t("ğ", 'g', 1), t("ı", 'h', 1), t("ö", 'o', 1), t("ş", 's', 1), t("ü", 'u', 1)]; +const HR: &[Tailor] = + &[t("č", 'c', 1), t("ć", 'c', 2), t("dž", 'd', 1), t("đ", 'd', 2), t("lj", 'l', 1), t("nj", 'n', 1), t("š", 's', 1), t("ž", 'z', 1)]; +const RO: &[Tailor] = &[t("ă", 'a', 1), t("â", 'a', 2), t("î", 'i', 1), t("ș", 's', 1), ts("ş", 's', 1, 8), t("ț", 't', 1), ts("ţ", 't', 1, 8)]; +const ET: &[Tailor] = &[t("š", 's', 1), t("z", 's', 2), t("ž", 's', 3), t("õ", 'w', 1), t("ä", 'w', 2), t("ö", 'w', 3), t("ü", 'w', 4)]; +const LT: &[Tailor] = &[t("č", 'c', 1), t("š", 's', 1), t("ž", 'z', 1)]; +const LV: &[Tailor] = &[t("č", 'c', 1), t("ģ", 'g', 1), t("ķ", 'k', 1), t("ļ", 'l', 1), t("ņ", 'n', 1), t("š", 's', 1), t("ž", 'z', 1)]; +const IS: &[Tailor] = &[ + t("á", 'a', 1), + t("ð", 'd', 1), + t("é", 'e', 1), + t("í", 'i', 1), + t("ó", 'o', 1), + t("ú", 'u', 1), + t("ý", 'y', 1), + t("þ", 'z', 1), + t("æ", 'z', 2), + t("ö", 'z', 3), +]; + +/// The primary language subtag of a BCP 47 tag (`pl-PL` → `pl`), lowercased. +pub fn language(lang: Option<&str>) -> String { + lang.and_then(|l| l.split(['-', '_']).next()).unwrap_or("").trim().to_ascii_lowercase() +} + +fn tailoring(lang: Option<&str>) -> &'static [Tailor] { + match language(lang).as_str() { + "pl" => PL, + "cs" => CS, + "sk" => SK, + "hu" => HU, + "sv" | "fi" => SV, + "da" | "nb" | "nn" | "no" => DA, + "es" | "gl" => ES, + "tr" | "az" => TR, + "hr" | "bs" | "sr" => HR, + "ro" => RO, + "et" => ET, + "lt" => LT, + "lv" => LV, + "is" => IS, + "sl" => LT, + _ => &[], + } +} + +/// A lowercase Latin letter followed by a combining mark, and the single letter Unicode +/// composes them to (NFC; generated from the Unicode Character Database). Text pasted from +/// decomposed sources (macOS file names: `z` + U+0307) collates like the composed letter. +const COMPOSE: &[(char, char, char)] = &[ + ('a', '\u{300}', 'à'), + ('a', '\u{301}', 'á'), + ('a', '\u{302}', 'â'), + ('a', '\u{303}', 'ã'), + ('a', '\u{304}', 'ā'), + ('a', '\u{306}', 'ă'), + ('a', '\u{307}', 'ȧ'), + ('a', '\u{308}', 'ä'), + ('a', '\u{30a}', 'å'), + ('a', '\u{30c}', 'ǎ'), + ('a', '\u{328}', 'ą'), + ('b', '\u{307}', 'ḃ'), + ('c', '\u{301}', 'ć'), + ('c', '\u{302}', 'ĉ'), + ('c', '\u{307}', 'ċ'), + ('c', '\u{30c}', 'č'), + ('c', '\u{327}', 'ç'), + ('d', '\u{307}', 'ḋ'), + ('d', '\u{30c}', 'ď'), + ('d', '\u{327}', 'ḑ'), + ('e', '\u{300}', 'è'), + ('e', '\u{301}', 'é'), + ('e', '\u{302}', 'ê'), + ('e', '\u{303}', 'ẽ'), + ('e', '\u{304}', 'ē'), + ('e', '\u{306}', 'ĕ'), + ('e', '\u{307}', 'ė'), + ('e', '\u{308}', 'ë'), + ('e', '\u{30c}', 'ě'), + ('e', '\u{327}', 'ȩ'), + ('e', '\u{328}', 'ę'), + ('f', '\u{307}', 'ḟ'), + ('g', '\u{301}', 'ǵ'), + ('g', '\u{302}', 'ĝ'), + ('g', '\u{304}', 'ḡ'), + ('g', '\u{306}', 'ğ'), + ('g', '\u{307}', 'ġ'), + ('g', '\u{30c}', 'ǧ'), + ('g', '\u{327}', 'ģ'), + ('h', '\u{302}', 'ĥ'), + ('h', '\u{307}', 'ḣ'), + ('h', '\u{308}', 'ḧ'), + ('h', '\u{30c}', 'ȟ'), + ('h', '\u{327}', 'ḩ'), + ('i', '\u{300}', 'ì'), + ('i', '\u{301}', 'í'), + ('i', '\u{302}', 'î'), + ('i', '\u{303}', 'ĩ'), + ('i', '\u{304}', 'ī'), + ('i', '\u{306}', 'ĭ'), + ('i', '\u{308}', 'ï'), + ('i', '\u{30c}', 'ǐ'), + ('i', '\u{328}', 'į'), + ('j', '\u{302}', 'ĵ'), + ('j', '\u{30c}', 'ǰ'), + ('k', '\u{301}', 'ḱ'), + ('k', '\u{30c}', 'ǩ'), + ('k', '\u{327}', 'ķ'), + ('l', '\u{301}', 'ĺ'), + ('l', '\u{30c}', 'ľ'), + ('l', '\u{327}', 'ļ'), + ('m', '\u{301}', 'ḿ'), + ('m', '\u{307}', 'ṁ'), + ('n', '\u{300}', 'ǹ'), + ('n', '\u{301}', 'ń'), + ('n', '\u{303}', 'ñ'), + ('n', '\u{307}', 'ṅ'), + ('n', '\u{30c}', 'ň'), + ('n', '\u{327}', 'ņ'), + ('o', '\u{300}', 'ò'), + ('o', '\u{301}', 'ó'), + ('o', '\u{302}', 'ô'), + ('o', '\u{303}', 'õ'), + ('o', '\u{304}', 'ō'), + ('o', '\u{306}', 'ŏ'), + ('o', '\u{307}', 'ȯ'), + ('o', '\u{308}', 'ö'), + ('o', '\u{30b}', 'ő'), + ('o', '\u{30c}', 'ǒ'), + ('o', '\u{328}', 'ǫ'), + ('p', '\u{301}', 'ṕ'), + ('p', '\u{307}', 'ṗ'), + ('r', '\u{301}', 'ŕ'), + ('r', '\u{307}', 'ṙ'), + ('r', '\u{30c}', 'ř'), + ('r', '\u{327}', 'ŗ'), + ('s', '\u{301}', 'ś'), + ('s', '\u{302}', 'ŝ'), + ('s', '\u{307}', 'ṡ'), + ('s', '\u{30c}', 'š'), + ('s', '\u{326}', 'ș'), + ('s', '\u{327}', 'ş'), + ('t', '\u{307}', 'ṫ'), + ('t', '\u{308}', 'ẗ'), + ('t', '\u{30c}', 'ť'), + ('t', '\u{326}', 'ț'), + ('t', '\u{327}', 'ţ'), + ('u', '\u{300}', 'ù'), + ('u', '\u{301}', 'ú'), + ('u', '\u{302}', 'û'), + ('u', '\u{303}', 'ũ'), + ('u', '\u{304}', 'ū'), + ('u', '\u{306}', 'ŭ'), + ('u', '\u{308}', 'ü'), + ('u', '\u{30a}', 'ů'), + ('u', '\u{30b}', 'ű'), + ('u', '\u{30c}', 'ǔ'), + ('u', '\u{328}', 'ų'), + ('v', '\u{303}', 'ṽ'), + ('w', '\u{300}', 'ẁ'), + ('w', '\u{301}', 'ẃ'), + ('w', '\u{302}', 'ŵ'), + ('w', '\u{307}', 'ẇ'), + ('w', '\u{308}', 'ẅ'), + ('w', '\u{30a}', 'ẘ'), + ('x', '\u{307}', 'ẋ'), + ('x', '\u{308}', 'ẍ'), + ('y', '\u{300}', 'ỳ'), + ('y', '\u{301}', 'ý'), + ('y', '\u{302}', 'ŷ'), + ('y', '\u{303}', 'ỹ'), + ('y', '\u{304}', 'ȳ'), + ('y', '\u{307}', 'ẏ'), + ('y', '\u{308}', 'ÿ'), + ('y', '\u{30a}', 'ẙ'), + ('z', '\u{301}', 'ź'), + ('z', '\u{302}', 'ẑ'), + ('z', '\u{307}', 'ż'), + ('z', '\u{30c}', 'ž'), +]; + +/// `s` lowercased, with each letter + combining mark that has a single-letter form composed. +fn lower_composed(s: &str) -> String { + let mut out = String::with_capacity(s.len()); + for c in s.to_lowercase().chars() { + if let Some(prev) = out.chars().next_back() + && let Some((_, _, composed)) = COMPOSE.iter().find(|(b, m, _)| *b == prev && *m == c) + { + out.pop(); + out.push(*composed); + } else { + out.push(c); + } + } + out +} + +fn weight(c: char, rank: u32) -> u32 { + ((c as u32) << 4).saturating_add(rank) +} + +/// One collation element: primary weight, secondary weight, the text it covers. +struct Elem<'a> { + primary: Vec, + sec: u16, + text: &'a str, + /// Set when the text is a letter of its own in the language (shown as is in headings). + own: bool, +} + +/// The next collation element of `rest` (lowercase), and its length in bytes. +fn next_elem<'a>(rest: &'a str, rules: &[Tailor]) -> Option<(Elem<'a>, usize)> { + if let Some(r) = rules.iter().filter(|r| rest.starts_with(r.seq)).max_by_key(|r| r.seq.len()) { + let text = rest.get(..r.seq.len())?; + return Some((Elem { primary: vec![weight(r.after, r.rank)], sec: r.sec, text, own: true }, r.seq.len())); + } + let c = rest.chars().next()?; + let n = c.len_utf8(); + let text = rest.get(..n)?; + if let Some((_, base, mark)) = DECOMP.iter().find(|(x, _, _)| *x == c) { + return Some((Elem { primary: base.chars().map(|b| weight(b, 0)).collect(), sec: *mark, text, own: false }, n)); + } + // A combining mark adds a secondary difference to the letter before it. + if ('\u{300}'..='\u{36f}').contains(&c) { + return Some((Elem { primary: Vec::new(), sec: 32 + (c as u32 - 0x300) as u16, text, own: false }, n)); + } + Some((Elem { primary: vec![weight(c, 0)], sec: 0, text, own: false }, n)) +} + +/// A sort key: compare keys to sort strings in a language's order. Equal keys mean the strings +/// differ at most in letter case. +#[derive(Clone, Debug, PartialEq, Eq, PartialOrd, Ord, Hash, Default)] +pub struct SortKey { + primary: Vec, + secondary: Vec, + text: String, +} + +/// The sort key of `s` for language `lang` (a BCP 47 tag such as `pl-PL`; `None` sorts like +/// English). +pub fn sort_key(s: &str, lang: Option<&str>) -> SortKey { + let text = lower_composed(s); + let rules = tailoring(lang); + let mut key = SortKey::default(); + let mut i = 0; + while let Some(rest) = text.get(i..).filter(|r| !r.is_empty()) { + let Some((e, n)) = next_elem(rest, rules) else { break }; + i += n.max(1); + if e.primary.is_empty() { + if let Some(last) = key.secondary.last_mut() { + *last = last.saturating_add(e.sec); + } + continue; + } + for (k, w) in e.primary.iter().enumerate() { + key.primary.push(*w); + key.secondary.push(if k == 0 { e.sec } else { 0 }); + } + } + key.text = text; + key +} + +/// Compare two strings in the order of language `lang`, ignoring case. +pub fn collate(a: &str, b: &str, lang: Option<&str>) -> Ordering { + sort_key(a, lang).cmp(&sort_key(b, lang)) +} + +/// The index heading letter `s` files under in language `lang`: `Ł` for `Łódź` in Polish, `L` +/// in English; `Ch` for `chata` in Czech; `A` for `Äpfel` in German, `Ä` in Swedish. +pub fn heading_letter(s: &str, lang: Option<&str>) -> Option { + let low = lower_composed(s.trim_start()); + let (e, _) = next_elem(&low, tailoring(lang))?; + let shown = if e.own || e.sec == 0 { + e.text.to_string() + } else { + DECOMP.iter().find(|(x, _, _)| e.text.starts_with(*x)).map(|(_, b, _)| b.chars().take(1).collect()).unwrap_or_else(|| e.text.to_string()) + }; + let mut cs = shown.chars(); + let first = cs.next()?; + Some(first.to_uppercase().chain(cs).collect()) +} + +#[cfg(test)] +mod tests { + use super::*; + + fn sorted(words: &[&str], lang: &str) -> Vec { + let mut v: Vec = words.iter().map(|w| w.to_string()).collect(); + v.sort_by(|a, b| collate(a, b, Some(lang))); + v + } + + #[test] + fn polish_letters_follow_their_base_letter() { + assert_eq!(sorted(&["Zakopane", "Łódź", "Kraków"], "pl-PL"), ["Kraków", "Łódź", "Zakopane"]); + assert_eq!(sorted(&["zebra", "ćma", "dom"], "pl-PL"), ["ćma", "dom", "zebra"]); + assert_eq!(sorted(&["żaba", "źle", "zebra", "ząb", "lato", "łapa", "mama"], "pl"), ["lato", "łapa", "mama", "ząb", "zebra", "źle", "żaba"]); + assert_eq!(sorted(&["b", "ą", "a"], "pl"), ["a", "ą", "b"]); + // Every Polish letter is a letter of its own, after its base letter (review of #844). + assert_eq!(sorted(&["Malbork", "Łódź", "Lublin"], "pl"), ["Lublin", "Łódź", "Malbork"]); + assert_eq!(sorted(&["ósmy", "owca", "pies"], "pl"), ["owca", "ósmy", "pies"]); + assert_eq!(sorted(&["ściana", "sum", "tak"], "pl"), ["sum", "ściana", "tak"]); + assert_eq!(sorted(&["ęsi", "ezop", "fala"], "pl"), ["ezop", "ęsi", "fala"]); + assert_eq!(sorted(&["ńa", "nz", "oko"], "pl"), ["nz", "ńa", "oko"]); + assert_eq!(heading_letter("ósmy", Some("pl")).as_deref(), Some("Ó")); + assert_eq!(heading_letter("ńa", Some("pl")).as_deref(), Some("Ń")); + // Decomposed letters (`z` + combining dot above, as macOS file names write `ż`) sort like + // the composed ones. + assert_eq!(collate("z\u{307}aba", "żaba", Some("pl-PL")), Ordering::Equal); + assert_eq!(collate("z\u{307}aba", "zebra", Some("pl-PL")), Ordering::Greater); + assert_eq!(heading_letter("Z\u{307}aba", Some("pl")).as_deref(), Some("Ż")); + assert_eq!(collate("a\u{328}", "ą", Some("pl")), Ordering::Equal); + } + + #[test] + fn swedish_puts_a_ring_and_umlauts_after_z() { + assert_eq!(sorted(&["ödla", "zebra", "äpple", "åka", "apa"], "sv-SE"), ["apa", "zebra", "åka", "äpple", "ödla"]); + } + + #[test] + fn german_sorts_umlauts_with_their_base_letter() { + assert_eq!(sorted(&["Zug", "Äpfel", "Apfel", "Bach"], "de-DE"), ["Apfel", "Äpfel", "Bach", "Zug"]); + assert_eq!(sorted(&["Straße", "Strasse", "Strand"], "de"), ["Strand", "Strasse", "Straße"]); + } + + #[test] + fn spanish_n_tilde_after_n() { + assert_eq!(sorted(&["oso", "ñu", "nube"], "es-ES"), ["nube", "ñu", "oso"]); + } + + #[test] + fn czech_slovak_hungarian_alphabets() { + assert_eq!(sorted(&["chata", "hrad", "ihla", "cesta", "čaj", "dům"], "cs-CZ"), ["cesta", "čaj", "dům", "hrad", "chata", "ihla"]); + assert_eq!(sorted(&["ôsmy", "oko", "pes", "ä", "b"], "sk"), ["ä", "b", "oko", "ôsmy", "pes"]); + assert_eq!(sorted(&["csak", "cukor", "dob", "szó", "sűrű", "tál"], "hu-HU"), ["cukor", "csak", "dob", "sűrű", "szó", "tál"]); + // ö and ő are one letter after o, told apart by the accent only. + assert_eq!(sorted(&["ötös", "őr", "pálya", "olló", "ör"], "hu"), ["olló", "ör", "őr", "ötös", "pálya"]); + } + + #[test] + fn english_keeps_case_insensitive_order_and_folds_accents() { + assert_eq!(sorted(&["rex", "résumé", "resume", "Apple", "banana"], "en-US"), ["Apple", "banana", "resume", "résumé", "rex"]); + assert_eq!(collate("Łódź", "Lodz", None), Ordering::Greater); + assert_eq!(collate("Łódź", "Lodzia", None), Ordering::Less); + assert_eq!(collate("ABC", "abc", None), Ordering::Equal); + } + + #[test] + fn heading_letters() { + assert_eq!(heading_letter("Łódź", Some("pl-PL")).as_deref(), Some("Ł")); + assert_eq!(heading_letter("Łódź", Some("en-US")).as_deref(), Some("L")); + assert_eq!(heading_letter("chata", Some("cs")).as_deref(), Some("Ch")); + assert_eq!(heading_letter("Äpfel", Some("de")).as_deref(), Some("A")); + assert_eq!(heading_letter("Äpple", Some("sv")).as_deref(), Some("Ä")); + assert_eq!(heading_letter("", Some("sv")), None); + } + + #[test] + fn hostile_input_never_panics() { + for s in ["", "\u{301}", "a\u{301}\u{301}\u{301}", "dzs", "\u{10FFFF}", "ẞ", "İstanbul"] { + for l in [None, Some("hu"), Some(""), Some("-"), Some("tr-TR")] { + let _ = sort_key(s, l); + let _ = heading_letter(s, l); + } + } + } +} diff --git a/crates/doc/src/graphic.rs b/crates/doc/src/graphic.rs index 091a0b426..c6e448f67 100644 --- a/crates/doc/src/graphic.rs +++ b/crates/doc/src/graphic.rs @@ -1,4 +1,4 @@ -//! Charts and SmartArt diagrams: drawings kept as neutral items ready to draw. Charts and SmartArt +//! Charts, SmartArt diagrams and native ink: drawings kept as neutral items ready to draw. Charts and SmartArt //! WordCraft can edit also keep their model ([`crate::chart::ChartSpec`], //! [`crate::smart_art::SmartArtSpec`]); the rest are shown, not edited. @@ -7,7 +7,7 @@ use serde::{Deserialize, Serialize}; use crate::para::ShapeKind; use crate::props::Rgb; -/// A chart or a SmartArt diagram, as items ready to draw (and, for an editable chart or SmartArt +/// A chart, SmartArt diagram or native ink preview, as items ready to draw (and, for an editable chart or SmartArt /// graphic, its model). #[derive(Clone, Debug, Default, PartialEq, Serialize, Deserialize)] #[serde(rename_all = "camelCase")] @@ -22,7 +22,7 @@ pub struct Graphic { pub w: f32, #[serde(default)] pub h: f32, - /// Where the graphic was read from, so saving writes the chart or diagram itself back (not + /// Where the graphic was read from, so saving writes the original drawing itself back (not /// only these items). `None` for graphics made some other way. #[serde(skip)] pub source: Option>, @@ -44,7 +44,8 @@ pub struct Graphic { #[derive(Clone, Debug, Default, PartialEq)] pub struct Embedded { /// The object's markup as a standalone XML element with its namespace declarations (DOCX: - /// the `a:graphic` of a chart or SmartArt diagram, the `w:object` of an OLE object). + /// the `a:graphic` of a chart, SmartArt diagram or native ink, an ink `mc:AlternateContent`, + /// or the `w:object` of an OLE object). pub xml: String, /// The relationships it uses, by the ids in `xml` (or in its parts, for a SmartArt drawing). pub refs: Vec, @@ -68,6 +69,8 @@ pub enum GraphicKind { #[default] Chart, Diagram, + /// Word's native InkML, kept opaque; displayed as its fallback picture when available. + Ink, } impl GraphicKind { @@ -76,6 +79,7 @@ impl GraphicKind { match self { GraphicKind::Chart => "chart", GraphicKind::Diagram => "diagram", + GraphicKind::Ink => "ink", } } } diff --git a/crates/doc/src/lib.rs b/crates/doc/src/lib.rs index a0f244006..07d2e3783 100644 --- a/crates/doc/src/lib.rs +++ b/crates/doc/src/lib.rs @@ -14,6 +14,7 @@ pub mod bidi; pub mod chart; mod cjk_numbers; pub mod cluster; +pub mod collate; pub mod connector; pub mod control; pub mod edit; @@ -24,6 +25,7 @@ pub mod fields; pub mod freeform; pub mod graphic; pub mod link; +pub mod locale_number; pub mod math; pub mod math_edit; pub mod math_latex; @@ -136,6 +138,11 @@ impl BoxBudget { self.open.pop(); } + /// Whether a text box is being expanded now. + pub fn inside(&self) -> bool { + !self.open.is_empty() + } + /// Boxes expanded so far. pub fn total(&self) -> usize { self.total @@ -274,12 +281,28 @@ pub struct TextBody { /// continues in when it doesn't fit this one. The linked box shows the rest of this box's /// story; its own blocks are unused. See [`Document::text_box_chains`]. pub next: Option, + /// Internal margins (left, top, right, bottom), EMU, when the file gives its own + /// (`wps:bodyPr` `lIns`… ECMA-376 Part 1 §21.1.2.1.1); `None` is Word's defaults. + pub inset: Option<[u32; 4]>, + /// Resize shape to fit text (`a:spAutoFit`): Word grows the box to its text, so none of + /// the text is cut. + pub auto_fit: bool, } +/// Word's default text box margins (left, top, right, bottom), EMU: 0.1" and 0.05". +pub const TEXT_BOX_INSET: [u32; 4] = [91_440, 45_720, 91_440, 45_720]; +/// Widest text box margin kept, EMU (22", the largest page side). +pub const MAX_TEXT_BOX_INSET: u32 = 20_116_800; + impl TextBody { pub fn is_default(&self) -> bool { *self == TextBody::default() } + + /// The internal margins (left, top, right, bottom) in points. + pub fn insets_pt(&self) -> [f32; 4] { + self.inset.unwrap_or(TEXT_BOX_INSET).map(|v| v.min(MAX_TEXT_BOX_INSET) as f32 / 12_700.0) + } } /// Most boxes one chain of linked text boxes runs through (hostile files). diff --git a/crates/doc/src/locale_number.rs b/crates/doc/src/locale_number.rs new file mode 100644 index 000000000..f9a505327 --- /dev/null +++ b/crates/doc/src/locale_number.rs @@ -0,0 +1,177 @@ +//! Numbers written the way a language writes them: `1 250,50` in Polish, `1.250,50` in German, +//! `1,250.50` in English. Used by table sort and table formulas. + +use crate::collate::language; + +/// The decimal separator of language `lang` (a BCP 47 tag; `None` means English). +pub fn decimal_separator(lang: Option<&str>) -> char { + let tag = lang.unwrap_or("").trim().to_ascii_lowercase().replace('_', "-"); + // Regions that write a decimal point where the language usually writes a comma. + const POINT_REGIONS: &[&str] = + &["de-ch", "de-li", "it-ch", "es-mx", "es-us", "es-419", "es-pr", "es-do", "es-gt", "es-hn", "es-ni", "es-pa", "es-sv"]; + if POINT_REGIONS.iter().any(|r| tag == *r || tag.starts_with(&format!("{r}-"))) { + return '.'; + } + const COMMA: &[&str] = &[ + "af", "az", "be", "bg", "bs", "ca", "cs", "da", "de", "el", "es", "et", "eu", "fi", "fo", "fr", "gl", "hr", "hu", "hy", "id", "is", "it", + "ka", "kk", "ky", "lb", "lt", "lv", "mk", "mn", "nb", "nl", "nn", "no", "pl", "pt", "ro", "ru", "sk", "sl", "sq", "sr", "sv", "tr", "uk", + "uz", "vi", + ]; + if COMMA.contains(&language(lang).as_str()) { ',' } else { '.' } +} + +const SPACES: &[char] = &[' ', '\u{a0}', '\u{202f}', '\u{2009}']; +const CURRENCY: &[char] = &['$', '€', '£', '¥', '¢', '₽', '₴', '₺', '₹', '₩', '₪', '₫', '฿', '₱', '₸', '₼', '₾']; + +/// Strip a currency symbol or a short currency word (`zł`, `kr`, `PLN`) from either end. +fn strip_currency(s: &str) -> &str { + let mut s = s.trim_matches(|c: char| c.is_whitespace() || CURRENCY.contains(&c)); + if let Some((head, word)) = s.rsplit_once(SPACES) + && (1..=3).contains(&word.chars().count()) + && word.chars().all(|c| c.is_alphabetic() || c == '.') + { + s = head; + } + if let Some((word, tail)) = s.split_once(SPACES) + && word.chars().count() == 3 + && word.chars().all(|c| c.is_ascii_uppercase()) + { + s = tail; + } + s.trim_matches(|c: char| c.is_whitespace() || CURRENCY.contains(&c)) +} + +/// Digits of an integer part with group separators `groups`: the first group has 1–3 digits, +/// every later one exactly 3. +fn grouped_digits(int: &str, groups: &[char]) -> Option { + if int.chars().all(|c| c.is_ascii_digit()) { + return Some(int.to_string()); + } + let parts: Vec<&str> = int.split(groups).collect(); + let (first, rest) = parts.split_first()?; + // A first group of `0` is never grouping (`0.125` is a fraction, not 125). + let ok = (1..=3).contains(&first.len()) + && !first.starts_with('0') + && first.chars().all(|c| c.is_ascii_digit()) + && rest.iter().all(|g| g.len() == 3 && g.chars().all(|c| c.is_ascii_digit())); + ok.then(|| parts.concat()) +} + +/// Parse `s` as a number written in language `lang`, or `None` when it isn't one. The decimal +/// separator is the language's; group separators are spaces (and `.` or `'` where the decimal +/// separator is a comma, `,` or `'` where it is a point) between groups of three digits. A +/// currency sign or short currency word is ignored, and so is a trailing `%` (`12 %` is 12, as +/// table sort reads it). In comma languages a lone `.` that cannot be a group separator (`1.5`) +/// is read as a decimal point. A trailing separator with no digits after it (`1.`, an ordinal +/// in a numbered column) is the whole number. A minus may be `-`, `−` or an en dash. +pub fn parse_number(s: &str, lang: Option<&str>) -> Option { + let dec = decimal_separator(lang); + let s = s.trim(); + let s = s.strip_suffix('%').map_or(s, str::trim_end); + let mut body = strip_currency(s); + let mut neg = false; + if let Some(rest) = body.strip_prefix(['-', '\u{2212}', '\u{2013}']) { + neg = true; + body = strip_currency(rest); + } else if let Some(rest) = body.strip_prefix('+') { + body = strip_currency(rest); + } + if body.is_empty() || body.len() > 400 { + return None; + } + let mut groups: Vec = SPACES.to_vec(); + groups.extend(['\'', '\u{2019}']); + groups.push(if dec == ',' { '.' } else { ',' }); + let (int, frac) = match body.split_once(dec) { + Some((i, f)) => (i, Some(f)), + None => (body, None), + }; + let int_digits = match grouped_digits(int, &groups) { + Some(d) => d, + None if dec == ',' && frac.is_none() && int.matches('.').count() == 1 => { + return parse_number(&format!("{}{}", if neg { "-" } else { "" }, int.replace('.', ",")), lang); + } + None => return None, + }; + let frac_digits = match frac { + Some(f) if f.chars().all(|c| c.is_ascii_digit()) => f, + Some(_) => return None, + None => "", + }; + if int_digits.is_empty() && frac_digits.is_empty() { + return None; + } + let v: f64 = format!("0{int_digits}.{frac_digits}0").parse().ok()?; + Some(if neg { -v } else { v }) +} + +/// Write `v` the way language `lang` writes a formula result: whole numbers without decimals, +/// others with two decimals and the language's decimal separator (`5,75` in Polish). +pub fn format_number(v: f64, lang: Option<&str>) -> String { + if !v.is_finite() { + return "0".into(); + } + if v.fract() == 0.0 && v.abs() < 1e15 { + return format!("{}", v as i64); + } + let s = format!("{v:.2}"); + if decimal_separator(lang) == ',' { s.replace('.', ",") } else { s } +} + +#[cfg(test)] +mod tests { + use super::*; + + #[test] + fn polish_numbers() { + let pl = Some("pl-PL"); + assert_eq!(parse_number("1,5", pl), Some(1.5)); + assert_eq!(parse_number("3,50", pl), Some(3.5)); + assert_eq!(parse_number("1 250,00", pl), Some(1250.0)); + assert_eq!(parse_number("1\u{a0}250,00 zł", pl), Some(1250.0)); + assert_eq!(parse_number("1.250,00", pl), Some(1250.0)); + assert_eq!(parse_number("1.5", pl), Some(1.5)); + assert_eq!(parse_number("-2,25", pl), Some(-2.25)); + assert_eq!(parse_number("PLN 12", pl), Some(12.0)); + assert_eq!(parse_number("abc", pl), None); + assert_eq!(parse_number("1,2,3", pl), None); + // From review: ordinals, a leading zero, percentages and an en dash minus. + assert_eq!(parse_number("1.", pl), Some(1.0), "a numbered column: 1."); + assert_eq!(parse_number("10.", pl), Some(10.0)); + assert_eq!(parse_number("0.125", pl), Some(0.125), "a leading 0 is never a group"); + assert_eq!(parse_number("0,125", pl), Some(0.125)); + assert_eq!(parse_number("12 %", pl), Some(12.0)); + assert_eq!(parse_number("12%", pl), Some(12.0)); + assert_eq!(parse_number("\u{2013}2,5", pl), Some(-2.5)); + assert_eq!(parse_number("%", pl), None); + assert_eq!(parse_number(".", pl), None); + assert_eq!(format_number(5.75, pl), "5,75"); + assert_eq!(format_number(1350.0, pl), "1350"); + } + + #[test] + fn english_numbers() { + let en = Some("en-US"); + assert_eq!(parse_number("1,250.50", en), Some(1250.5)); + assert_eq!(parse_number("$1,250", en), Some(1250.0)); + assert_eq!(parse_number("1.5", en), Some(1.5)); + assert_eq!(parse_number("12", None), Some(12.0)); + assert_eq!(parse_number(".5", None), Some(0.5)); + assert_eq!(parse_number("1,5", en), None); + assert_eq!(parse_number("1.", en), Some(1.0)); + assert_eq!(parse_number("0,125", en), None, "never 125"); + assert_eq!(format_number(5.75, en), "5.75"); + assert_eq!(decimal_separator(Some("de-CH")), '.'); + assert_eq!(decimal_separator(Some("de-DE")), ','); + } + + #[test] + fn hostile_numbers_never_panic() { + for s in ["", "-", "+", ",", ".", "1e400", "inf", "NaN", "--1", "₽", "1 ,", " zł", "\u{2212}", "9".repeat(500).as_str()] { + for l in [None, Some("pl"), Some("en"), Some("")] { + let _ = parse_number(s, l); + } + } + assert_eq!(format_number(f64::NAN, None), "0"); + } +} diff --git a/crates/doc/src/para.rs b/crates/doc/src/para.rs index 8fec42071..287f12ac3 100644 --- a/crates/doc/src/para.rs +++ b/crates/doc/src/para.rs @@ -447,8 +447,8 @@ pub enum InlineObject { code: bool, }, /// The start of a field whose result is ordinary content — formatted, possibly spanning - /// paragraphs — up to the matching [`InlineObject::FieldEnd`]. Citation managers' `ADDIN` - /// fields (Zotero, Mendeley, EndNote) are kept this way; `instr` is the field code. See + /// paragraphs — up to the matching [`InlineObject::FieldEnd`]. Imported TOCs and citation + /// managers' `ADDIN` fields are kept this way; `instr` is the field code. See /// [`crate::fields`]. FieldStart { instr: String, diff --git a/crates/doc/src/props.rs b/crates/doc/src/props.rs index 4405d4d9b..97d653572 100644 --- a/crates/doc/src/props.rs +++ b/crates/doc/src/props.rs @@ -303,10 +303,10 @@ pub struct CharProps { /// inherit, then `font`. #[serde(skip_serializing_if = "Option::is_none")] pub font_ea: Option, - /// Complex-script size, points (`w:szCs`). `None` = the same as `size` (see [`CharProps::overlay`]). + /// Complex-script size, points (`w:szCs`). `None` leaves the inherited value unchanged. #[serde(skip_serializing_if = "Option::is_none")] pub size_cs: Option, - /// Complex-script bold / italic (`w:bCs`, `w:iCs`). `None` = the same as `bold` / `italic`. + /// Complex-script bold / italic (`w:bCs`, `w:iCs`). `None` leaves the inherited value unchanged. #[serde(skip_serializing_if = "Option::is_none")] pub bold_cs: Option, #[serde(skip_serializing_if = "Option::is_none")] @@ -390,11 +390,8 @@ impl CharProps { /// Apply every `Some` field of `patch` on top of `self`. /// - /// Size, bold and italic come in pairs with their complex-script values (Word's own buttons - /// set both): a patch that sets the plain value and leaves the complex-script one `None` sets - /// both, so a run's size or bold isn't overridden for its Persian/Arabic characters by an - /// inherited `w:szCs` / `w:bCs`. The complex-script font inherits on its own (Word often - /// sets only the Latin font and lets complex-script text keep the style's). + /// Plain and complex-script properties inherit independently. Commands that intend to + /// format both scripts must set both values explicitly. pub fn overlay(&mut self, patch: &CharProps) { overlay_fields!(self, patch; style, font, size, bold, italic, underline, underline_color, strike, double_strike, color, highlight, shading, border, vert_align, caps, small_caps, hidden, spacing, scale, position, kern, outline, shadow, emboss, engrave, lang, no_proof, rtl, @@ -402,15 +399,6 @@ impl CharProps { if let Some(p) = &patch.text_effects { self.text_effects.get_or_insert_with(Default::default).overlay(p); } - if patch.size.is_some() && patch.size_cs.is_none() { - self.size_cs = None; - } - if patch.bold.is_some() && patch.bold_cs.is_none() { - self.bold_cs = None; - } - if patch.italic.is_some() && patch.italic_cs.is_none() { - self.italic_cs = None; - } } pub fn overlaid(mut self, patch: &CharProps) -> CharProps { self.overlay(patch); @@ -638,6 +626,33 @@ pub struct NumRef { pub level: u8, } +/// A text frame's placement (`w:framePr`, ECMA-376 §17.3.1.11), in points. Offsets apply when +/// the matching alignment is unset; anchors and the wrap keep OOXML's names. +#[derive(Clone, Debug, PartialEq, Default, Serialize, Deserialize)] +#[serde(default, rename_all = "camelCase")] +pub struct TextFrame { + /// Width; `None` fits the text. + pub w: Option, + /// Height, with `h_exact` exact rather than at least; `None` fits the text. + pub h: Option, + pub h_exact: bool, + pub x: f32, + pub y: f32, + /// `left`, `center`, `right`, `inside`, `outside`. + pub x_align: Option, + /// `top`, `center`, `bottom`, `inside`, `outside` (`inline` is no alignment). + pub y_align: Option, + /// `text` (the column), `margin` or `page`. + pub h_anchor: String, + /// `text` (the paragraph), `margin` or `page`. + pub v_anchor: String, + /// `around`, `auto`, `notBeside`, `none`, `tight` or `through`. + pub wrap: String, + /// Distance from the text left and right, and above and below. + pub h_space: f32, + pub v_space: f32, +} + /// Paragraph formatting. `None` = inherit. #[derive(Clone, Debug, PartialEq, Default, Serialize, Deserialize)] #[serde(default, rename_all = "camelCase")] @@ -671,6 +686,13 @@ pub struct ParaProps { pub space_before: Option, #[serde(skip_serializing_if = "Option::is_none")] pub space_after: Option, + /// HTML-style automatic space before (`w:beforeAutospacing`): 14 pt, none at the top of the + /// document or of a table cell, and collapsing with the space after the paragraph above. + #[serde(skip_serializing_if = "Option::is_none")] + pub space_before_auto: Option, + /// HTML-style automatic space after (`w:afterAutospacing`): 14 pt, none at the end of a cell. + #[serde(skip_serializing_if = "Option::is_none")] + pub space_after_auto: Option, #[serde(skip_serializing_if = "Option::is_none")] pub line_spacing: Option, /// Ignore space before/after between paragraphs of the same style. @@ -709,6 +731,10 @@ pub struct ParaProps { /// Drop cap: lines to drop (0 = none) — applies to the first character(s). #[serde(skip_serializing_if = "Option::is_none")] pub drop_cap: Option, + /// A text frame (`w:framePr` other than a drop cap): the paragraph is placed apart from the + /// text, which may wrap around it — a footer's page number in a frame, for instance. + #[serde(skip_serializing_if = "Option::is_none")] + pub frame: Option, /// Asian typography (`w:kinsoku`): East Asian line-breaking rules — no line starts with /// closing punctuation such as 、。」 or ends with opening punctuation such as 「（. /// Unset = on. @@ -743,9 +769,9 @@ pub struct ParaProps { impl ParaProps { pub fn overlay(&mut self, patch: &ParaProps) { - overlay_fields!(self, patch; style, align, indent_left, indent_right, indent_first, indent_left_chars, indent_right_chars, indent_first_chars, space_before, space_after, line_spacing, + overlay_fields!(self, patch; style, align, indent_left, indent_right, indent_first, indent_left_chars, indent_right_chars, indent_first_chars, space_before, space_after, space_before_auto, space_after_auto, line_spacing, contextual_spacing, mirror_indents, keep_next, keep_lines, page_break_before, widow_control, outline_level, numbering, tabs, shading, borders, - suppress_hyphens, suppress_line_numbers, bidi, drop_cap, kinsoku, word_wrap, overflow_punct, top_line_punct, auto_space_de, auto_space_dn, + suppress_hyphens, suppress_line_numbers, bidi, drop_cap, frame, kinsoku, word_wrap, overflow_punct, top_line_punct, auto_space_de, auto_space_dn, snap_to_grid); } pub fn overlaid(mut self, patch: &ParaProps) -> ParaProps { @@ -1025,6 +1051,10 @@ pub struct CellProps { /// Which way the cell's text runs. pub text_direction: TextDirection, pub no_wrap: bool, + /// The cell's last paragraph takes no room when it is empty (`w:hideMark`): a row of such + /// empty cells shrinks to its borders. + #[serde(skip_serializing_if = "std::ops::Not::not")] + pub hide_mark: bool, /// Tracked formatting change (`w:tcPrChange`). #[serde(skip_serializing_if = "Option::is_none")] pub fmt_change: Option>>, diff --git a/crates/doc/src/resolve.rs b/crates/doc/src/resolve.rs index a96f35280..4ef0137fd 100644 --- a/crates/doc/src/resolve.rs +++ b/crates/doc/src/resolve.rs @@ -8,6 +8,10 @@ use crate::props::{ }; use crate::styles::{StyleKind, StyleSheet}; +/// Automatic (HTML-style) paragraph spacing, points: what Word for Windows gives `w:beforeAutospacing` +/// and `w:afterAutospacing` at any font size (measured). +pub const AUTO_SPACING: f32 = 14.0; + /// Concrete character formatting. #[derive(Clone, Debug, PartialEq, Serialize, Deserialize)] #[serde(rename_all = "camelCase")] @@ -32,6 +36,9 @@ pub struct ResolvedChar { pub spacing: f32, pub scale: f32, pub position: f32, + /// Kerning: the smallest font size, in points, that is kerned (`w:kern`); 0 = never kerned. + #[serde(default)] + pub kern: f32, pub outline: bool, pub shadow: bool, pub emboss: bool, @@ -82,6 +89,7 @@ impl ResolvedChar { spacing: c.spacing.unwrap_or(0.0).clamp(-100.0, 100.0), scale: c.scale.unwrap_or(100.0).clamp(1.0, 600.0), position: c.position.unwrap_or(0.0).clamp(-800.0, 800.0), + kern: c.kern.filter(|k| k.is_finite()).unwrap_or(0.0).clamp(0.0, 1638.0), outline: c.outline.unwrap_or(false), shadow: c.shadow.unwrap_or(false), emboss: c.emboss.unwrap_or(false), @@ -108,6 +116,11 @@ impl ResolvedChar { pub fn complex(&self) -> ResolvedChar { ResolvedChar { font: self.font_cs.clone(), size: self.size_cs, bold: self.bold_cs, italic: self.italic_cs, ..self.clone() } } + /// Whether the text is kerned: Word kerns only a run with Kerning for fonts set (`w:kern`) at + /// a size of at least that many points, never by default (measured in Word for Windows). + pub fn kerned(&self) -> bool { + self.kern > 0.0 && self.size >= self.kern + } /// Does a character of this run use the complex-script properties? pub fn uses_complex(&self, c: char) -> bool { self.rtl || self.cs || crate::bidi::is_complex_script(c) @@ -138,6 +151,12 @@ pub struct ResolvedPara { pub indent_first: f32, pub space_before: f32, pub space_after: f32, + /// The space before or after is automatic (HTML-style, [`AUTO_SPACING`]): see + /// `ParaProps::space_before_auto`. + #[serde(default)] + pub before_auto: bool, + #[serde(default)] + pub after_auto: bool, pub line_spacing: LineSpacing, pub contextual_spacing: bool, /// Left/right indents are inside/outside: they swap on even pages. @@ -153,6 +172,9 @@ pub struct ResolvedPara { pub borders: Option, pub bidi: bool, pub drop_cap: u8, + /// A text frame the paragraph sits in (see `ParaProps::frame`). + #[serde(default, skip_serializing_if = "Option::is_none")] + pub frame: Option, pub suppress_hyphens: bool, pub suppress_line_numbers: bool, /// Asian typography (see `ParaProps`): East Asian line-breaking rules. @@ -206,8 +228,10 @@ impl ResolvedPara { indent_left: clampi(p.indent_left), indent_right: clampi(p.indent_right), indent_first: clampi(p.indent_first), - space_before: p.space_before.unwrap_or(0.0).clamp(0.0, 1584.0), - space_after: p.space_after.unwrap_or(0.0).clamp(0.0, 1584.0), + space_before: if p.space_before_auto == Some(true) { AUTO_SPACING } else { p.space_before.unwrap_or(0.0).clamp(0.0, 1584.0) }, + space_after: if p.space_after_auto == Some(true) { AUTO_SPACING } else { p.space_after.unwrap_or(0.0).clamp(0.0, 1584.0) }, + before_auto: p.space_before_auto == Some(true), + after_auto: p.space_after_auto == Some(true), line_spacing: match p.line_spacing.unwrap_or_default() { LineSpacing::Multiple(m) => LineSpacing::Multiple(if m.is_finite() { m.clamp(0.06, 132.0) } else { 1.0 }), LineSpacing::AtLeast(v) => LineSpacing::AtLeast(if v.is_finite() { v.clamp(0.0, 1584.0) } else { 0.0 }), @@ -226,6 +250,7 @@ impl ResolvedPara { borders: p.borders.filter(Borders::any_visible), bidi: p.bidi.unwrap_or(false), drop_cap: p.drop_cap.unwrap_or(0).min(10), + frame: p.frame.clone().filter(|_| p.drop_cap.unwrap_or(0) == 0), suppress_hyphens: p.suppress_hyphens.unwrap_or(false), suppress_line_numbers: p.suppress_line_numbers.unwrap_or(false), kinsoku: p.kinsoku.unwrap_or(true), diff --git a/crates/doc/src/styles.rs b/crates/doc/src/styles.rs index 0053c03e5..4c9a398c4 100644 --- a/crates/doc/src/styles.rs +++ b/crates/doc/src/styles.rs @@ -174,7 +174,7 @@ impl StyleSheet { /// The built-in style set used for new documents. pub fn builtin() -> Self { let mut s = StyleSheet { - default_chr: CharProps { font: Some(BODY_FONT.into()), size: Some(12.0), kern: Some(0.0), lang: Some("en-US".into()), ..chr() }, + default_chr: CharProps { font: Some(BODY_FONT.into()), size: Some(12.0), kern: Some(1.0), lang: Some("en-US".into()), ..chr() }, default_para: ParaProps { space_after: Some(8.0), space_before: Some(0.0), @@ -444,6 +444,19 @@ impl StyleSheet { out.reverse(); out } + /// Which indents (left, first line) of a paragraph that gets its list from style `id` the style + /// chain sets over the list level's own. Word applies the level's indents on top of the most + /// derived style that gives a list, so that style and the ones based on it win over the level, + /// and the styles it is based on don't (Heading 3 with its own left indent on Heading 2's + /// list). `(false, false)` when no style in the chain gives a list. + pub fn indents_over_list(&self, id: &str) -> (bool, bool) { + let chain = self.chain(id); + let Some(k) = chain.iter().rposition(|s| s.para.numbering.is_some()) else { return (false, false) }; + let set = |v: Option, chars: Option| v.is_some() || chars.is_some_and(|c| c != 0.0); + chain.get(k..).unwrap_or_default().iter().fold((false, false), |(l, f), s| { + (l || set(s.para.indent_left, s.para.indent_left_chars), f || set(s.para.indent_first, s.para.indent_first_chars)) + }) + } /// Table style `id` with its based-on chain merged, base styles first (`None` when there is no /// such table style). Styles of other kinds in the chain are skipped. pub fn table_style(&self, id: &str) -> Option { diff --git a/crates/docbin/Cargo.toml b/crates/docbin/Cargo.toml index 253a0b298..3f9d43662 100644 --- a/crates/docbin/Cargo.toml +++ b/crates/docbin/Cargo.toml @@ -12,6 +12,8 @@ workspace = true wordcraft-doc = { workspace = true } wordcraft-geom = { workspace = true } cfb = { workspace = true } +image = { workspace = true } +miniz_oxide = { workspace = true } thiserror = { workspace = true } log = { workspace = true } diff --git a/crates/docbin/src/drawing.rs b/crates/docbin/src/drawing.rs new file mode 100644 index 000000000..ef403d158 --- /dev/null +++ b/crates/docbin/src/drawing.rs @@ -0,0 +1,262 @@ +//! Floating shapes and pictures: the `PlcfSpa` anchors of the main document and the headers +//! ([MS-DOC] §2.8.27 PlcfSpa, §2.9.253 Spa) joined with the OfficeArt shapes they name, and the +//! text box stories (`PlcftxbxTxt`, §2.8.30) that fill text boxes. + +use std::collections::HashMap; + +use wordcraft_doc::para::{Anchor, Float, FloatAlign, GroupChild, InlineObject, ShapeKind, Wrap}; +use wordcraft_doc::props::Rgb; + +use crate::media::Pictures; +use crate::officeart::{Node, Sp}; + +/// Most anchors (and text box stories) read per PLC. +const MAX_ANCHORS: usize = 10_000; +/// `Spa` and `FTXBXS` record sizes. +const CB_SPA: usize = 26; +const CB_FTXBXS: usize = 22; +/// Largest frame or offset kept (points; 22 inches, Word's own limit for sizes). +const MAX_PT: f32 = 1584.0; +/// Deepest group nesting flattened. +const MAX_GROUP_DEPTH: usize = 8; + +/// The (CP, data) entries of a PLC with `cb`-byte data elements, the trailing CP dropped. +fn plc(table: &[u8], (fc, lcb): (u32, u32), cb: usize) -> Vec<(u32, &[u8])> { + let Some(blob) = table.get(fc as usize..(fc as usize).saturating_add(lcb as usize)) else { return Vec::new() }; + let n = blob.len().saturating_sub(4) / (4 + cb); + let n = n.min(MAX_ANCHORS); + let data_at = (n + 1) * 4; + (0..n) + .filter_map(|i| { + let cp = u32::from_le_bytes(blob.get(i * 4..i * 4 + 4)?.try_into().ok()?); + let at = data_at + i * cb; + Some((cp, blob.get(at..at + cb)?)) + }) + .collect() +} + +/// The text box stories of a `PlcftxbxTxt` (or `PlcfHdrtxbxTxt`): (shape spid, story CP range +/// relative to the text box subdocument). +pub(crate) fn textbox_stories(table: &[u8], pair: Option<(u32, u32)>) -> Vec<(u32, (u32, u32))> { + let Some(pair) = pair else { return Vec::new() }; + let Some(blob) = table.get(pair.0 as usize..(pair.0 as usize).saturating_add(pair.1 as usize)) else { return Vec::new() }; + let mut out = Vec::new(); + for (i, (cp, ftxbxs)) in plc(table, pair, CB_FTXBXS).iter().enumerate() { + // The story runs to the next entry's CP (the PLC's last CP for the last entry). + let Some(end) = blob.get((i + 1) * 4..(i + 1) * 4 + 4).and_then(|b| b.try_into().ok()).map(u32::from_le_bytes) else { continue }; + let Some(lid) = ftxbxs.get(14..18).and_then(|b| b.try_into().ok()).map(u32::from_le_bytes) else { continue }; + if end > *cp && lid != 0 { + out.push((lid, (*cp, end))); + } + } + out +} + +/// What building the floating objects needs. +pub(crate) struct Build<'a> { + /// The BLIP store and the stream holding BLIPs not embedded in it (WordDocument). + pub(crate) bstore: &'a [&'a [u8]], + pub(crate) delay: &'a [u8], + /// Text box part ids by shape spid. + pub(crate) stories: &'a HashMap, +} + +/// The floating objects anchored by `PlcfSpa` (at `pair`), as (absolute anchor CP, object), +/// sorted by CP. `cp_base` turns the PLC's CPs into absolute ones (the header subdocument's +/// start for `PlcSpaHdr`). +pub(crate) fn anchors( + table: &[u8], + pair: Option<(u32, u32)>, + cp_base: u32, + shapes: &HashMap, + b: &Build, + pics: &mut Pictures, +) -> Vec<(u32, InlineObject)> { + let Some(pair) = pair else { return Vec::new() }; + let mut out = Vec::new(); + for (cp, spa) in plc(table, pair, CB_SPA) { + let i32_at = |o: usize| spa.get(o..o + 4).and_then(|b| b.try_into().ok()).map(i32::from_le_bytes).unwrap_or(0); + let spid = i32_at(0) as u32; + let [xl, yt, xr, yb] = [4, 8, 12, 16].map(i32_at); + let flags = spa.get(20..22).map(|b| u16::from_le_bytes([b[0], b[1]])).unwrap_or(0); + let Some(node) = shapes.get(&spid) else { continue }; + let tw = |v: i64| (v as f32 / 20.0).clamp(-2.0 * MAX_PT, 2.0 * MAX_PT); + let (mut x, mut y) = (tw(xl as i64), tw(yt as i64)); + let (mut w, mut h) = (tw(xr as i64 - xl as i64).clamp(0.0, MAX_PT), tw(yb as i64 - yt as i64).clamp(0.0, MAX_PT)); + let sp = node.sp(); + let rot = wordcraft_geom::normalize_degrees(if sp.props.rot.is_finite() { sp.props.rot } else { 0.0 }); + // A shape turned by 45°–135° (or 225°–315°) is anchored by its turned bounds: the + // unrotated frame has the other aspect, about the same centre. + if (45.0..135.0).contains(&rot) || (225.0..315.0).contains(&rot) { + x += (w - h) / 2.0; + y += (h - w) / 2.0; + std::mem::swap(&mut w, &mut h); + } + let mut float = spa_float(flags, &sp.props); + float.x = x; + float.y = y; + float.rot = rot; + float.flip_h = sp.flip_h(); + float.flip_v = sp.flip_v(); + let obj = match node { + Node::Shape(sp) => leaf(sp, w, h, float, b, pics), + Node::Group(sp, members) => group(sp, members, w, h, float, b, pics), + }; + match obj { + Some(obj) => out.push((cp_base.saturating_add(cp), obj)), + None => pics.dropped(cp_base.saturating_add(cp)), + } + } + out.sort_by_key(|(cp, _)| *cp); + out +} + +/// Placement from a `Spa`'s flags (fHdr, bx, by, wr, wrk, fRcaSimple, fBelowText, fAnchorLock) +/// and the shape's position properties, which take precedence when present. +fn spa_float(flags: u16, p: &crate::officeart::Props) -> Float { + let bx = (flags >> 1) & 3; + let by = (flags >> 3) & 3; + let wr = (flags >> 5) & 0xF; + let below = flags & (1 << 14) != 0; + let wrap = match wr { + 1 => Wrap::TopAndBottom, + 3 => { + if below { + Wrap::BehindText + } else { + Wrap::InFrontOfText + } + } + 4 => Wrap::Tight, + 5 => Wrap::Through, + _ => Wrap::Square, + }; + let h_rel = match p.posrelh.unwrap_or(bx as u32) { + 0 => Anchor::Margin, + 1 => Anchor::Page, + 3 => Anchor::Character, + _ => Anchor::Column, + }; + let v_rel = match p.posrelv.unwrap_or(by as u32) { + 0 => Anchor::Margin, + 1 => Anchor::Page, + 3 => Anchor::Line, + _ => Anchor::Paragraph, + }; + let align = |v: Option| match v { + Some(1) => Some(FloatAlign::Start), + Some(2) => Some(FloatAlign::Center), + Some(3) => Some(FloatAlign::End), + Some(4) => Some(FloatAlign::Inside), + Some(5) => Some(FloatAlign::Outside), + _ => None, + }; + // Distances from text, EMUs; Word's defaults are 0.125" at the sides, none above and below. + let emu = |v: Option, def: f32| v.map_or(def, |e| (e as f32 / 12_700.0).clamp(0.0, MAX_PT)); + let [dl, dt, dr, db] = p.wrap_dist; + Float { + wrap, + h_rel, + v_rel, + h_align: align(p.posh), + v_align: align(p.posv), + dist: emu(dl, 9.0).max(emu(dr, 9.0)), + dist_top: emu(dt, 0.0), + dist_bottom: emu(db, 0.0), + ..Float::default() + } +} + +/// One shape as a picture or a drawn shape, `w` × `h` points. +fn leaf(sp: &Sp, w: f32, h: f32, float: Float, b: &Build, pics: &mut Pictures) -> Option { + let p = &sp.props; + if let Some(pib) = p.pib { + let media = pics.bse(pib, b.bstore, b.delay)?; + return Some(InlineObject::Image { media, w, h, alt: p.alt.clone(), float, crop: p.crop, ole: None }); + } + let story = b.stories.get(&sp.spid).copied().filter(|_| sp.textbox || sp.spt == 202); + let mut kind = match sp.spt { + 1 => ShapeKind::Rectangle, + 2 => ShapeKind::RoundedRectangle, + 3 => ShapeKind::Ellipse, + 4 => ShapeKind::Diamond, + 5 | 6 => ShapeKind::Triangle, + 12 | 187 => ShapeKind::Star, + 13 | 66 | 67 | 68 | 69 => ShapeKind::Arrow, + 20 | 32 => ShapeKind::Line, + 74 => ShapeKind::Heart, + 202 => ShapeKind::TextBox, + // A freeform's path is not read: better left out than drawn as a box. + 0 => return None, + _ => ShapeKind::Rectangle, + }; + if story.is_some() && kind == ShapeKind::Rectangle { + kind = ShapeKind::TextBox; + } + let line = kind == ShapeKind::Line; + let fill = if line || p.filled == Some(false) { None } else { Some(p.fill_color.unwrap_or(Rgb::WHITE)) }; + let stroke = if p.line == Some(false) { None } else { Some(p.line_color.unwrap_or(Rgb::BLACK)) }; + // Line widths are EMUs; Word's default is 0.75 pt. + let stroke_width = p.line_width.map_or(0.75, |e| (e as f32 / 12_700.0).clamp(0.0, 100.0)); + Some(InlineObject::Shape { + kind, + w, + h, + fill, + stroke, + stroke_width, + float, + story, + freeform: None, + effects: Default::default(), + extra: Default::default(), + }) +} + +/// A group: its members flattened into the group's own coordinate space (nested groups mapped +/// through their child anchors). +fn group(sp: &Sp, members: &[Node], w: f32, h: f32, float: Float, b: &Build, pics: &mut Pictures) -> Option { + let [l, t, r, btm] = sp.group_space.unwrap_or([0, 0, 0, 0]); + let (cw, ch) = ((r as f64 - l as f64) as f32, (btm as f64 - t as f64) as f32); + let (ch_w, ch_h) = if cw > 0.0 && ch > 0.0 { (cw, ch) } else { (w.max(1.0), h.max(1.0)) }; + let mut children = Vec::new(); + // Map: outer = x0 + inner × s; the top group's space starts at its FSPGR origin. + members_into(members, (-(l as f32), -(t as f32), 1.0, 1.0), 0, b, pics, &mut children); + if children.is_empty() { + return None; + } + Some(InlineObject::Group { w, h, float, ch_w, ch_h, children, canvas: None }) +} + +fn members_into(nodes: &[Node], t: (f32, f32, f32, f32), depth: usize, b: &Build, pics: &mut Pictures, out: &mut Vec) { + let (x0, y0, sx, sy) = t; + let fit = |v: f32| if v.is_finite() { v.clamp(-1e7, 1e7) } else { 0.0 }; + for n in nodes { + if out.len() >= wordcraft_doc::para::MAX_GROUP_CHILDREN { + return; + } + let sp = n.sp(); + let Some([l, t, r, btm]) = sp.child else { continue }; + let (ox, oy) = (l as f32, t as f32); + let (ew, eh) = ((r as f64 - l as f64) as f32, (btm as f64 - t as f64) as f32); + match n { + Node::Group(_, members) => { + if depth >= MAX_GROUP_DEPTH { + continue; + } + let [gl, gt, gr, gb] = sp.group_space.unwrap_or([l, t, r, btm]); + let (gw, gh) = ((gr as f64 - gl as f64) as f32, (gb as f64 - gt as f64) as f32); + let (kx, ky) = if gw > 0.0 && gh > 0.0 { (ew / gw, eh / gh) } else { (1.0, 1.0) }; + let nt = (fit(x0 + (ox - gl as f32 * kx) * sx), fit(y0 + (oy - gt as f32 * ky) * sy), fit(sx * kx), fit(sy * ky)); + members_into(members, nt, depth + 1, b, pics, out); + } + Node::Shape(sp) => { + let float = Float { rot: sp.props.rot, flip_h: sp.flip_h(), flip_v: sp.flip_v(), ..Float::default() }; + let (w, h) = (fit(ew * sx).max(0.0), fit(eh * sy).max(0.0)); + if let Some(obj) = leaf(sp, w, h, float, b, pics) { + out.push(GroupChild { x: fit(x0 + ox * sx), y: fit(y0 + oy * sy), obj }); + } + } + } + } +} diff --git a/crates/docbin/src/fib.rs b/crates/docbin/src/fib.rs index 959240fa3..983e45bca 100644 --- a/crates/docbin/src/fib.rs +++ b/crates/docbin/src/fib.rs @@ -33,8 +33,16 @@ pub(crate) mod pair { pub(crate) const PLCF_BKF: usize = 22; pub(crate) const PLCF_BKL: usize = 23; pub(crate) const CLX: usize = 33; + /// `fcPlcSpaMom` / `fcPlcSpaHdr`: floating shape anchors of the main document / headers. + pub(crate) const PLC_SPA_MOM: usize = 40; + pub(crate) const PLC_SPA_HDR: usize = 41; pub(crate) const PLCF_END_REF: usize = 46; pub(crate) const PLCF_END_TXT: usize = 47; + /// `fcDggInfo`: the OfficeArt drawing content (BLIP store, shapes). + pub(crate) const DGG_INFO: usize = 50; + /// `fcPlcftxbxTxt` / `fcPlcfHdrtxbxTxt`: text box stories of the main document / headers. + pub(crate) const PLCF_TXBX_TXT: usize = 56; + pub(crate) const PLCF_HDR_TXBX_TXT: usize = 58; /// `fcPlfLst`: the list definitions (LSTF array, with the LVLs appended after the blob). pub(crate) const PLF_LST: usize = 73; /// `fcPlfLfo`: the list format overrides (LFO array with LFOData appended). diff --git a/crates/docbin/src/fmt.rs b/crates/docbin/src/fmt.rs index 30b830115..6c4232f2c 100644 --- a/crates/docbin/src/fmt.rs +++ b/crates/docbin/src/fmt.rs @@ -1,7 +1,7 @@ //! Sprms → model properties, and style-chain resolution. use wordcraft_doc::Rgb; -use wordcraft_doc::props::{Align, CharProps, Highlight, LineSpacing, ParaProps, TextColor, Underline, VertAlign}; +use wordcraft_doc::props::{Align, CharProps, Highlight, LineSpacing, ParaProps, TabAlign, TabLeader, TabStop, TextColor, Underline, VertAlign}; use wordcraft_doc::styles::{Style, StyleKind, StyleSheet}; use crate::sprm::{self, Prl, Toggle}; @@ -187,11 +187,63 @@ pub(crate) fn apply_char(p: &mut CharProps, prl: &Prl, base: &CharProps, fonts: } } +/// Custom tab stops from `sprmPChgTabsPapx` / `sprmPChgTabs` ([MS-DOC] PChgTabsPapxOperand, +/// PChgTabsOperand): the stops removed become `Clear` stops (they hide inherited ones) and the +/// stops added carry their alignment and leader (TBD: jc in bits 0–2, tlc in bits 3–5). +fn change_tabs(p: &mut ParaProps, prl: &Prl) { + let op = prl.operand; + let i16_at = |i: usize| op.get(i..i + 2).map(|b| f32::from(i16::from_le_bytes([b[0], b[1]])) / 20.0); + // Both operands start with cb; sprmPChgTabs's removals also carry a "close" distance each. + let close = prl.op == sprm::P_CHG_TABS; + let mut at = 1; + let Some(&del) = op.get(at) else { return }; + let del = usize::from(del.min(64)); + at += 1; + let mut removed = Vec::with_capacity(del); + for k in 0..del { + if let Some(x) = i16_at(at + 2 * k) { + removed.push(x); + } + } + at += 2 * del * if close { 2 } else { 1 }; + let Some(&add) = op.get(at) else { return }; + let add = usize::from(add.min(64)); + at += 1; + let mut tabs = p.tabs.take().unwrap_or_default(); + // Removed within 25 twips (1.25 pt), as the format says. + tabs.retain(|t| !removed.iter().any(|x| (t.pos - x).abs() <= 1.25)); + for x in removed { + tabs.push(TabStop { pos: x, align: TabAlign::Clear, leader: TabLeader::None }); + } + for k in 0..add { + let (Some(pos), Some(&tbd)) = (i16_at(at + 2 * k), op.get(at + 2 * add + k)) else { break }; + let align = match tbd & 7 { + 1 => TabAlign::Center, + 2 => TabAlign::Right, + 3 => TabAlign::Decimal, + 4 => TabAlign::Bar, + _ => TabAlign::Left, + }; + let leader = match (tbd >> 3) & 7 { + 1 => TabLeader::Dot, + 2 => TabLeader::Hyphen, + 3 | 4 => TabLeader::Underscore, + 5 => TabLeader::MiddleDot, + _ => TabLeader::None, + }; + tabs.retain(|t| (t.pos - pos).abs() > 0.05); + tabs.push(TabStop { pos, align, leader }); + } + tabs.sort_by(|a, b| a.pos.total_cmp(&b.pos)); + p.tabs = Some(tabs); +} + /// Apply a paragraph-property sprm. pub(crate) fn apply_para(p: &mut ParaProps, prl: &Prl) { let tw = |v: i32| (v as f32 / 20.0).clamp(-1000.0, 1000.0); let s16 = |prl: &Prl| -> Option { u16_of(prl).map(|v| v as i16) }; match prl.op { + sprm::P_CHG_TABS | sprm::P_CHG_TABS_PAPX => change_tabs(p, prl), P_JC | P_JC80 => { p.align = match prl.operand.first().copied().unwrap_or(0) { 1 => Some(Align::Center), diff --git a/crates/docbin/src/lib.rs b/crates/docbin/src/lib.rs index dde7ebb22..9ff438620 100644 --- a/crates/docbin/src/lib.rs +++ b/crates/docbin/src/lib.rs @@ -8,12 +8,14 @@ //! capped), so hostile input yields an error or a best-effort document, never a panic. #![deny(clippy::unwrap_used, clippy::expect_used, clippy::panic, clippy::unimplemented, clippy::todo, clippy::unreachable)] +mod drawing; mod fib; mod fkp; mod fmt; mod list; mod media; mod notes; +mod officeart; mod piece; mod sections; mod sprm; @@ -139,8 +141,36 @@ pub fn read(bytes: &[u8]) -> Result { sheet: &sheet, notes: &[], marks: &[], + drawn: &[], + sect_ends: &[], }; + // Drawings: text box stories become parts first, then the floating shapes and pictures + // anchored in the main document and the headers, found by their anchor CP while walking. + let txbx_base = edn_base.saturating_add(fib.ccp.edn); + let hdr_txbx_base = txbx_base.saturating_add(fib.ccp.txbx); + let art = fib + .pair(fib::pair::DGG_INFO) + .and_then(|(f, l)| table.get(f as usize..(f as usize).saturating_add(l as usize))) + .map(officeart::content) + .unwrap_or_default(); + let mut tb_stories = std::collections::HashMap::new(); + for (pair, base, len) in [(fib::pair::PLCF_TXBX_TXT, txbx_base, fib.ccp.txbx), (fib::pair::PLCF_HDR_TXBX_TXT, hdr_txbx_base, fib.ccp.hdr_txbx)] { + for (spid, ab) in drawing::textbox_stories(&table, fib.pair(pair)) { + let Some((a, b)) = story(base, len, ab) else { continue }; + if doc.parts.len() >= MAX_PARTS || tb_stories.contains_key(&spid) { + continue; + } + let blocks = table::assemble(walk(&ctx, a, b, &mut pics)); + tb_stories.insert(spid, doc.add_part(PartKind::TextBox, blocks)); + } + } + let build = drawing::Build { bstore: &art.bstore, delay: &word, stories: &tb_stories }; + let mut drawn = drawing::anchors(&table, fib.pair(fib::pair::PLC_SPA_MOM), 0, &art.main, &build, &mut pics); + drawn.extend(drawing::anchors(&table, fib.pair(fib::pair::PLC_SPA_HDR), hdd_base, &art.header, &build, &mut pics)); + drawn.sort_by_key(|(cp, _)| *cp); + let ctx = WalkCtx { drawn: &drawn, ..ctx }; + // Header and footer parts: skip the six separator stories, then six stories per section // in the order even header, odd (default) header, even footer, odd (default) footer, // first header, first footer. Empty stories inherit from the previous section. @@ -186,7 +216,9 @@ pub fn read(bytes: &[u8]) -> Result { // Sorted so the main walk finds each reference by binary search. note_links.sort_by_key(|(cp, _, _)| *cp); let marks = notes::parse_bookmarks(&table, &fib); - let ctx = WalkCtx { notes: ¬e_links, marks: &marks, ..ctx }; + let mut sect_ends: Vec = sects.iter().map(|s| s.end_cp).collect(); + sect_ends.sort_unstable(); + let ctx = WalkCtx { notes: ¬e_links, marks: &marks, sect_ends: §_ends, ..ctx }; // Sections: each section's properties attach to the paragraph that ends it (its story of // header/footer parts included); the section reaching the end of the text is the final one. @@ -246,6 +278,11 @@ struct WalkCtx<'a> { notes: &'a [(u32, NoteKind, u32)], /// Bookmark boundaries (main-document CPs), sorted for a sequential walk. marks: &'a [(u32, notes::Mark)], + /// Floating shapes and pictures by anchor CP (absolute), sorted. + drawn: &'a [(u32, InlineObject)], + /// Where sections end (main-document CPs, sorted): the section mark (0x0C) just before + /// each ends the section's last paragraph, it isn't a page break inside it. + sect_ends: &'a [u32], } /// State while assembling one paragraph from per-CP formatting. @@ -319,8 +356,10 @@ fn walk(ctx: &WalkCtx, cp_start: u32, cp_end: u32, pics: &mut media::Pictures) - continue; } }; + // A section mark (0x0C) at a section's end is that section's last paragraph mark. + let section_mark = c == '\u{C}' && ctx.sect_ends.binary_search(&cp.saturating_add(1)).is_ok(); match c { - '\r' | '\u{7}' => { + '\r' | '\u{7}' | '\u{C}' if c != '\u{C}' || section_mark => { // Paragraph mark: its PAPX covers the FC range of the paragraph it ends, so // querying at the mark's own FC finds that paragraph's properties. let fc_mark = ctx.pieces.fc_of_cp(cp).unwrap_or(0); @@ -405,21 +444,29 @@ fn walk(ctx: &WalkCtx, cp_start: u32, cp_end: u32, pics: &mut media::Pictures) - let fc = ctx.pieces.fc_of_cp(cp).unwrap_or(0); let fc_pic = pic_location(ctx.chpx_bins.chpx(ctx.word, fc)); match fc_pic.and_then(|at| pics.get(ctx.data, at)) { - Some((key, w, h)) => { + Some(p) => { pb.push('\u{FFFC}', CharProps::default()); pb.objects.push(InlineObject::Image { - media: key, - w, - h, - alt: String::new(), + media: p.key, + w: p.w, + h: p.h, + alt: p.alt, float: Default::default(), - crop: [0.0; 4], + crop: p.crop, ole: None, }); } None => pics.dropped(cp), } } + '\u{8}' => { + // Drawn object anchor: the floating shape or picture anchored at this CP. + let at = ctx.drawn.partition_point(|(c, _)| *c < cp); + for (_, obj) in ctx.drawn.get(at..).unwrap_or_default().iter().take_while(|(c, _)| *c == cp) { + pb.push('\u{FFFC}', CharProps::default()); + pb.objects.push(obj.clone()); + } + } '\u{2}' => { // Footnote or endnote reference; custom symbols share the same character. let at = ctx.notes.partition_point(|(c, _, _)| *c < cp); diff --git a/crates/docbin/src/media.rs b/crates/docbin/src/media.rs index 82ca60f1b..1cb46e9e2 100644 --- a/crates/docbin/src/media.rs +++ b/crates/docbin/src/media.rs @@ -1,10 +1,13 @@ //! Inline pictures: `sprmCPicLocation` → a `PICF` in the Data stream → embedded image bytes -//! ([MS-DOC] §2.9.190 PICF, §1.3.5 Pictures). +//! ([MS-DOC] §2.9.190 PICF, §1.3.5 Pictures), and the shared store of every picture the +//! document reads (inline and floating). use std::collections::{BTreeMap, HashMap}; use std::ops::Range; use std::sync::Arc; +use crate::officeart; + /// `PICF.cbHeader` — the fixed size of the PICF structure. const CB_HEADER: usize = 0x44; /// Offsets inside PICF: `PICMID.dxaGoal/dyaGoal/mx/my` start at 28. @@ -17,10 +20,13 @@ pub(crate) const MAX_MEDIA_BYTES: usize = 256 << 20; /// The document's pictures, keyed by their location in the Data stream: anchors that share a /// `fc_pic` share one media entry, and the total bytes copied are capped, so a file with -/// thousands of anchors on one huge picture costs one copy, not thousands. +/// thousands of anchors on one huge picture costs one copy, not thousands. Pictures of the +/// drawing layer's BLIP store are shared the same way, by store index. pub(crate) struct Pictures { - /// fc_pic → (media key, width, height) of pictures already materialised. - by_fc: HashMap, + /// fc_pic → the inline picture already materialised. + by_fc: HashMap, + /// BLIP store index → media key (`None`: unreadable, not retried). + by_bse: HashMap>, /// The media entries, merged into the document at the end. pub(crate) media: BTreeMap>>, total: usize, @@ -28,33 +34,89 @@ pub(crate) struct Pictures { pub(crate) limit: usize, /// Anchors whose picture was dropped (only the first few are logged). dropped: usize, + /// Pictures kept as placeholders (formats we can't draw). + placeholders: usize, } impl Default for Pictures { fn default() -> Self { - Pictures { by_fc: HashMap::new(), media: BTreeMap::new(), total: 0, limit: MAX_MEDIA_BYTES, dropped: 0 } + Pictures { + by_fc: HashMap::new(), + by_bse: HashMap::new(), + media: BTreeMap::new(), + total: 0, + limit: MAX_MEDIA_BYTES, + dropped: 0, + placeholders: 0, + } } } +/// An inline picture: its media key, display size (points), crop fractions and alt text. +#[derive(Clone, Debug)] +pub(crate) struct Inline { + pub(crate) key: String, + pub(crate) w: f32, + pub(crate) h: f32, + pub(crate) crop: [f32; 4], + pub(crate) alt: String, +} + impl Pictures { - /// The media key and display size (points) of the picture at `fc_pic`, copying its - /// bytes the first time. `None` when the header is malformed, the payload is a format - /// we do not embed (metafiles), or a picture count/byte cap is reached. - pub(crate) fn get(&mut self, data: &[u8], fc_pic: u32) -> Option<(String, f32, f32)> { + /// The inline picture at `fc_pic` in the Data stream, copying its bytes the first time. + /// `None` when the header is malformed, the picture can't be read, or a picture + /// count/byte cap is reached. + pub(crate) fn get(&mut self, data: &[u8], fc_pic: u32) -> Option { if let Some(hit) = self.by_fc.get(&fc_pic) { return Some(hit.clone()); } - if self.by_fc.len() >= MAX_PICTURES { + if self.by_fc.len().saturating_add(self.by_bse.len()) >= MAX_PICTURES { return None; } - let pic = locate(data, fc_pic)?; - let bytes = data.get(pic.range)?; + let pic = match locate(data, fc_pic)? { + Found::Range(pic) => { + let bytes = data.get(pic.range.clone())?.to_vec(); + let key = self.insert(bytes, pic.ext, false)?; + Inline { key, w: pic.w, h: pic.h, crop: [0.0; 4], alt: String::new() } + } + Found::Shape { w, h, blip, props } => { + let key = self.insert(blip.bytes, blip.ext, blip.placeholder)?; + Inline { key, w, h, crop: props.crop, alt: props.alt } + } + }; + self.by_fc.insert(fc_pic, pic.clone()); + Some(pic) + } + + /// The media key of BLIP store entry `pib` (1-based), decoding it the first time; `delay` + /// is the stream that holds BLIPs not embedded in the store (the WordDocument stream). + pub(crate) fn bse(&mut self, pib: u32, bstore: &[&[u8]], delay: &[u8]) -> Option { + if let Some(hit) = self.by_bse.get(&pib) { + return hit.clone(); + } + if self.by_fc.len().saturating_add(self.by_bse.len()) >= MAX_PICTURES { + return None; + } + let fbse = bstore.get((pib as usize).checked_sub(1)?)?; + let room = self.limit.saturating_sub(self.total); + let key = officeart::fbse_blip(fbse, delay, room).and_then(|b| self.insert(b.bytes, b.ext, b.placeholder)); + self.by_bse.insert(pib, key.clone()); + key + } + + /// Add picture bytes as a new media entry, within the byte cap. + fn insert(&mut self, bytes: Vec, ext: &str, placeholder: bool) -> Option { let total = self.total.checked_add(bytes.len()).filter(|&t| t <= self.limit)?; - let key = format!("image{}.{}", self.by_fc.len() + 1, pic.ext); - self.media.insert(key.clone(), Arc::new(bytes.to_vec())); + let key = format!("image{}.{ext}", self.media.len() + 1); + if placeholder { + if self.placeholders < 10 { + log::warn!("docbin: picture `{key}` is in a format WordCraft can't draw ({ext}); it shows as a placeholder"); + } + self.placeholders = self.placeholders.saturating_add(1); + } + self.media.insert(key.clone(), Arc::new(bytes)); self.total = total; - self.by_fc.insert(fc_pic, (key.clone(), pic.w, pic.h)); - Some((key, pic.w, pic.h)) + Some(key) } /// Note a picture anchor that could not be materialised. @@ -74,27 +136,64 @@ struct Located { ext: &'static str, } -/// Locate the picture at `fc_pic` in the Data stream without copying it. -fn locate(data: &[u8], fc_pic: u32) -> Option { +/// What a `PICFAndOfficeArtData` holds: raw image bytes (older writers), or an +/// `OfficeArtInlineSpContainer` (Word 97 and later) whose BLIP is decoded. +enum Found { + Range(Located), + Shape { w: f32, h: f32, blip: officeart::Blip, props: officeart::Props }, +} + +/// `PICF.mfpf.mm` values of pictures stored as a shape (`MM_SHAPE`) or a shape with a file +/// name (`MM_SHAPEFILE`). +const MM_SHAPE: u16 = 0x0064; +const MM_SHAPEFILE: u16 = 0x0066; + +/// Locate the picture at `fc_pic` in the Data stream ([MS-DOC] §2.9.192 PICFAndOfficeArtData). +fn locate(data: &[u8], fc_pic: u32) -> Option { let base = fc_pic as usize; let pic = data.get(base..)?; let lcb = u32::from_le_bytes(pic.get(0..4)?.try_into().ok()?) as usize; let cb_header = u16::from_le_bytes(pic.get(4..6)?.try_into().ok()?) as usize; let cb_header = if (0x20..=0x100).contains(&cb_header) { cb_header } else { CB_HEADER }; + let mm = u16::from_le_bytes(pic.get(6..8)?.try_into().ok()?); let m = pic.get(OFF_PICMID..OFF_PICMID + 8)?; let dxa_goal = i16::from_le_bytes([m[0], m[1]]) as f32; let dya_goal = i16::from_le_bytes([m[2], m[3]]) as f32; let mx = u16::from_le_bytes([m[4], m[5]]) as f32; let my = u16::from_le_bytes([m[6], m[7]]) as f32; let scale = |r: f32| if r > 0.0 { (r / 1000.0).clamp(0.01, 100.0) } else { 1.0 }; + let end = lcb.min(pic.len()); + if mm == MM_SHAPE || mm == MM_SHAPEFILE { + let mut at = cb_header; + if mm == MM_SHAPEFILE { + at = at.checked_add(1 + *pic.get(at)? as usize)?; + } + let (blip, props) = inline_shape(pic.get(at..end)?, data)?; + // The goal size is before cropping; the crop fractions take their share off it. + let [cl, ct, cr, cb] = props.crop; + let w = (dxa_goal / 20.0 * (1.0 - cl - cr).max(0.01) * scale(mx)).clamp(1.0, 1584.0); + let h = (dya_goal / 20.0 * (1.0 - ct - cb).max(0.01) * scale(my)).clamp(1.0, 1584.0); + return Some(Found::Shape { w, h, blip, props }); + } let w = (dxa_goal / 20.0 * scale(mx)).clamp(1.0, 1584.0); let h = (dya_goal / 20.0 * scale(my)).clamp(1.0, 1584.0); - let end = lcb.min(pic.len()); let body = pic.get(cb_header..end)?; let at = sniff(body)?; let image = body.get(at..)?; let start = base.checked_add(cb_header)?.checked_add(at)?; - Some(Located { w, h, range: start..start.checked_add(image.len())?, ext: ext_of(image) }) + Some(Found::Range(Located { w, h, range: start..start.checked_add(image.len())?, ext: ext_of(image) })) +} + +/// An `OfficeArtInlineSpContainer`: the picture shape, then the `OfficeArtFBSE` records its +/// `pib` indexes (BLIPs embedded, or at `foDelay` in the Data stream). +fn inline_shape(buf: &[u8], data: &[u8]) -> Option<(officeart::Blip, officeart::Props)> { + let mut budget = 10_000usize; + let recs = officeart::children(buf, &mut budget); + let sp = recs.iter().find(|r| r.typ == officeart::rt::SP_CONTAINER).map(|r| officeart::shape(r.body, &mut budget))?; + let fbses: Vec<&[u8]> = recs.iter().filter(|r| r.typ == officeart::rt::FBSE).map(|r| r.body).collect(); + let fbse = sp.props.pib.and_then(|p| fbses.get((p as usize).checked_sub(1)?)).or(fbses.first())?; + let blip = officeart::fbse_blip(fbse, data, MAX_MEDIA_BYTES)?; + Some((blip, sp.props)) } /// Offset of the first embedded raster image inside the payload, skipping the small diff --git a/crates/docbin/src/officeart.rs b/crates/docbin/src/officeart.rs new file mode 100644 index 000000000..188dbd1b0 --- /dev/null +++ b/crates/docbin/src/officeart.rs @@ -0,0 +1,463 @@ +//! OfficeArt (Escher) drawing records, as Word 97-2003 stores them ([MS-ODRAW], public Microsoft +//! Open Specifications): the record header, the BLIP store and its pictures, shape containers with +//! their property tables, and the `OfficeArtContent` of the Table stream ([MS-DOC] §2.9.172). +//! +//! Everything here walks hostile bytes: record lengths are clamped to their parent, nesting depth +//! and record counts are capped, and decompressed metafiles are size-limited. + +use std::collections::HashMap; + +use wordcraft_doc::props::Rgb; + +/// Deepest container nesting walked (groups inside groups…). +const MAX_DEPTH: usize = 12; +/// Most records walked in one drawing tree or BLIP store. +const MAX_RECORDS: usize = 200_000; +/// Most BLIP store entries kept. +const MAX_BSE: usize = 10_000; +/// Most drawings (main document, headers) read from `OfficeArtContent`. +const MAX_DRAWINGS: usize = 8; +/// Largest decompressed metafile (bytes). +const MAX_METAFILE: usize = 64 << 20; +/// Largest DIB converted to PNG, per side (pixels). +const MAX_DIB_SIDE: u32 = 16_384; + +/// Record types ([MS-ODRAW] §2.2, §2.3). +pub(crate) mod rt { + pub(crate) const DGG_CONTAINER: u16 = 0xF000; + pub(crate) const BSTORE_CONTAINER: u16 = 0xF001; + pub(crate) const DG_CONTAINER: u16 = 0xF002; + pub(crate) const SPGR_CONTAINER: u16 = 0xF003; + pub(crate) const SP_CONTAINER: u16 = 0xF004; + pub(crate) const FBSE: u16 = 0xF007; + pub(crate) const FSPGR: u16 = 0xF009; + pub(crate) const FSP: u16 = 0xF00A; + pub(crate) const FOPT: u16 = 0xF00B; + pub(crate) const CLIENT_TEXTBOX: u16 = 0xF00D; + pub(crate) const CHILD_ANCHOR: u16 = 0xF00F; + pub(crate) const SECONDARY_FOPT: u16 = 0xF121; + pub(crate) const TERTIARY_FOPT: u16 = 0xF122; + pub(crate) const BLIP_EMF: u16 = 0xF01A; + pub(crate) const BLIP_WMF: u16 = 0xF01B; + pub(crate) const BLIP_PICT: u16 = 0xF01C; + pub(crate) const BLIP_JPEG: u16 = 0xF01D; + pub(crate) const BLIP_PNG: u16 = 0xF01E; + pub(crate) const BLIP_DIB: u16 = 0xF01F; + pub(crate) const BLIP_TIFF: u16 = 0xF029; + pub(crate) const BLIP_JPEG_CMYK: u16 = 0xF02A; +} + +/// One record: header fields and its body (clamped to what the parent holds). +#[derive(Clone, Copy, Debug)] +pub(crate) struct Rec<'a> { + pub(crate) inst: u16, + pub(crate) typ: u16, + pub(crate) body: &'a [u8], +} + +/// The record at the start of `buf` and the bytes it spans (header included). A body longer +/// than `buf` is cut to what is there. +pub(crate) fn record(buf: &[u8]) -> Option<(Rec<'_>, usize)> { + let h = buf.get(0..8)?; + let ver_inst = u16::from_le_bytes([h[0], h[1]]); + let typ = u16::from_le_bytes([h[2], h[3]]); + let len = u32::from_le_bytes([h[4], h[5], h[6], h[7]]) as usize; + let rest = buf.get(8..)?; + let len = len.min(rest.len()); + let body = rest.get(..len)?; + Some((Rec { inst: ver_inst >> 4, typ, body }, 8 + len)) +} + +/// The sibling records of a container body, at most `budget` of them (the budget is shared +/// across a whole walk, so a deep or wide tree cannot multiply the work). +pub(crate) fn children<'a>(body: &'a [u8], budget: &mut usize) -> Vec> { + let mut out = Vec::new(); + let mut at = 0usize; + while *budget > 0 { + let Some((rec, used)) = body.get(at..).and_then(record) else { break }; + *budget -= 1; + out.push(rec); + at = at.saturating_add(used); + } + out +} + +fn u16_at(b: &[u8], at: usize) -> Option { + Some(u16::from_le_bytes(b.get(at..at.checked_add(2)?)?.try_into().ok()?)) +} + +fn u32_at(b: &[u8], at: usize) -> Option { + Some(u32::from_le_bytes(b.get(at..at.checked_add(4)?)?.try_into().ok()?)) +} + +fn i32_at(b: &[u8], at: usize) -> Option { + u32_at(b, at).map(|v| v as i32) +} + +// --------------------------------------------------------------------------------------------- +// Pictures (BLIPs). + +/// A decoded picture: bytes ready for `Document::media` and their file extension. +pub(crate) struct Blip { + pub(crate) bytes: Vec, + pub(crate) ext: &'static str, + /// The format cannot be drawn (Macintosh PICT): the picture is kept as a placeholder of + /// the right size. + pub(crate) placeholder: bool, +} + +/// The picture of an `OfficeArtFBSE` body: its embedded BLIP, or the BLIP stored at `foDelay` +/// in `delay` (the WordDocument stream for the document's BLIP store, the Data stream for +/// inline pictures). `limit` caps the bytes produced. +pub(crate) fn fbse_blip(fbse: &[u8], delay: &[u8], limit: usize) -> Option { + // btWin32, btMacOS, rgbUid[16], tag, size, cRef, foDelay, unused, cbName, unused×2. + let size = u32_at(fbse, 20)? as usize; + let fo_delay = u32_at(fbse, 28)? as usize; + let cb_name = *fbse.get(33)? as usize; + let at = 36usize.checked_add(cb_name)?; + if let Some((rec, _)) = fbse.get(at..).and_then(record) { + return blip(&rec, limit); + } + // Not embedded: the record sits in the delay stream. + let end = fo_delay.checked_add(size.max(8))?.min(delay.len()); + let (rec, _) = record(delay.get(fo_delay..end)?)?; + blip(&rec, limit) +} + +/// Decode one `OfficeArtBlip*` record ([MS-ODRAW] §2.2.23-§2.2.31). +pub(crate) fn blip(rec: &Rec, limit: usize) -> Option { + // Each type has two instance values; the odd one carries a second 16-byte UID. + let uids = if rec.inst & 1 == 1 { 32 } else { 16 }; + match rec.typ { + rt::BLIP_PNG | rt::BLIP_JPEG | rt::BLIP_JPEG_CMYK | rt::BLIP_DIB | rt::BLIP_TIFF => { + let data = rec.body.get(uids + 1..)?; // + tag byte + if data.is_empty() || data.len() > limit { + return None; + } + match rec.typ { + rt::BLIP_PNG => Some(Blip { bytes: data.to_vec(), ext: "png", placeholder: false }), + rt::BLIP_JPEG | rt::BLIP_JPEG_CMYK => Some(Blip { bytes: data.to_vec(), ext: "jpeg", placeholder: false }), + rt::BLIP_TIFF => Some(Blip { bytes: data.to_vec(), ext: "tiff", placeholder: false }), + _ => dib_to_png(data), + } + } + rt::BLIP_EMF | rt::BLIP_WMF | rt::BLIP_PICT => { + // OfficeArtMetafileHeader: cbSize, rcBounds, ptSize, cbSave, compression, filter. + let hdr = rec.body.get(uids..uids + 34)?; + let cb_size = u32_at(hdr, 0)? as usize; + let cb_save = u32_at(hdr, 28)? as usize; + let compression = *hdr.get(32)?; + let raw = rec.body.get(uids + 34..)?; + let raw = raw.get(..cb_save.min(raw.len()))?; + let cap = limit.min(MAX_METAFILE); + let bytes = if compression == 0 { + let want = if cb_size > 0 { cb_size.min(cap) } else { cap }; + miniz_oxide::inflate::decompress_to_vec_zlib_with_limit(raw, want).ok()? + } else { + if raw.len() > cap { + return None; + } + raw.to_vec() + }; + if bytes.is_empty() { + return None; + } + match rec.typ { + rt::BLIP_EMF => Some(Blip { bytes, ext: "emf", placeholder: false }), + rt::BLIP_WMF => Some(Blip { bytes, ext: "wmf", placeholder: false }), + _ => Some(Blip { bytes, ext: "pict", placeholder: true }), + } + } + _ => None, + } +} + +/// A device-independent bitmap (BITMAPINFOHEADER + palette + pixels) as a PNG: the DIB gets +/// the 14-byte BMP file header it lacks, is decoded with size limits and re-encoded. When the +/// bitmap can't be decoded the BMP bytes are kept (and render as a grey placeholder). +fn dib_to_png(dib: &[u8]) -> Option { + let header = u32_at(dib, 0)? as usize; + if !(12..=124).contains(&header) { + return None; + } + let (bits, used, compression) = + if header == 12 { (u16_at(dib, 10)? as usize, 0, 0) } else { (u16_at(dib, 14)? as usize, u32_at(dib, 32)? as usize, u32_at(dib, 16)?) }; + let entry = if header == 12 { 3 } else { 4 }; + let colors = if used > 0 { + used.min(256) + } else if bits <= 8 { + 1usize << bits + } else { + 0 + }; + // BI_BITFIELDS masks follow a 40-byte header. + let masks = if header == 40 && (compression == 3 || compression == 6) { if compression == 3 { 12 } else { 16 } } else { 0 }; + let off_bits = 14usize.checked_add(header)?.checked_add(colors * entry)?.checked_add(masks)?; + let file_size = 14usize.checked_add(dib.len())?; + let mut bmp = Vec::with_capacity(file_size); + bmp.extend_from_slice(b"BM"); + bmp.extend_from_slice(&u32::try_from(file_size).ok()?.to_le_bytes()); + bmp.extend_from_slice(&[0, 0, 0, 0]); + bmp.extend_from_slice(&u32::try_from(off_bits).ok()?.to_le_bytes()); + bmp.extend_from_slice(dib); + let mut reader = image::ImageReader::new(std::io::Cursor::new(&bmp)); + reader.set_format(image::ImageFormat::Bmp); + let mut limits = image::Limits::default(); + limits.max_image_width = Some(MAX_DIB_SIDE); + limits.max_image_height = Some(MAX_DIB_SIDE); + limits.max_alloc = Some(256 << 20); + reader.limits(limits); + let mut png = Vec::new(); + match reader.decode() { + Ok(img) if img.write_to(&mut std::io::Cursor::new(&mut png), image::ImageFormat::Png).is_ok() => { + Some(Blip { bytes: png, ext: "png", placeholder: false }) + } + _ => Some(Blip { bytes: bmp, ext: "bmp", placeholder: true }), + } +} + +// --------------------------------------------------------------------------------------------- +// Shapes. + +/// The shape properties we use, from `OfficeArtFOPT` and its secondary/tertiary tables +/// ([MS-ODRAW] §2.3). +#[derive(Clone, Debug, Default, PartialEq)] +pub(crate) struct Props { + /// `pib`: 1-based index of the picture in the BLIP store. + pub(crate) pib: Option, + /// Rotation, degrees clockwise. + pub(crate) rot: f32, + /// Crop fractions (left, top, right, bottom). + pub(crate) crop: [f32; 4], + pub(crate) fill_color: Option, + pub(crate) filled: Option, + pub(crate) line_color: Option, + /// Line width in EMUs. + pub(crate) line_width: Option, + pub(crate) line: Option, + /// An arrowhead at either end of a line. + pub(crate) arrowhead: bool, + /// `posh`/`posrelh`/`posv`/`posrelv`: alignment and reference frame (Word 2000 and later). + pub(crate) posh: Option, + pub(crate) posrelh: Option, + pub(crate) posv: Option, + pub(crate) posrelv: Option, + /// Distance from wrapped text (left, top, right, bottom), EMUs. + pub(crate) wrap_dist: [Option; 4], + /// `wzDescription`: the alternative text. + pub(crate) alt: String, +} + +impl Props { + /// Apply one property table (`OfficeArtFOPT` or a secondary/tertiary one): `count` 6-byte + /// entries, then the complex data of the entries flagged complex, in order. + pub(crate) fn apply(&mut self, body: &[u8], count: u16) { + let n = (count as usize).min(body.len() / 6); + let mut complex_at = n * 6; + for i in 0..n { + let Some(opid) = u16_at(body, i * 6) else { break }; + let Some(op) = u32_at(body, i * 6 + 2) else { break }; + let pid = opid & 0x3FFF; + if opid & 0x8000 != 0 { + let start = complex_at; + complex_at = complex_at.saturating_add(op as usize); + if pid == 0x0381 { + let end = complex_at.min(body.len()); + if let Some(data) = body.get(start..end) { + let units: Vec = + data.as_chunks::<2>().0.iter().map(|c| u16::from_le_bytes(*c)).take_while(|u| *u != 0).take(4096).collect(); + self.alt = String::from_utf16_lossy(&units); + } + } + continue; + } + let fixed = |v: u32| (v as i32) as f32 / 65_536.0; + match pid { + 0x0004 => self.rot = fixed(op), + 0x0100..=0x0103 => { + // cropFromTop, cropFromBottom, cropFromLeft, cropFromRight. + let slot = [1, 3, 0, 2][(pid - 0x0100) as usize]; + let v = fixed(op); + self.crop[slot] = if v.is_finite() { v.clamp(0.0, 0.99) } else { 0.0 }; + } + 0x0104 => self.pib = Some(op), + 0x0181 => self.fill_color = color(op), + 0x01BF => { + // fFilled (bit 4) is valid when fUsefFilled (bit 20) is set. + if op & (1 << 20) != 0 { + self.filled = Some(op & (1 << 4) != 0); + } + } + 0x01C0 => self.line_color = color(op), + 0x01CB => self.line_width = Some(op), + 0x01D0 | 0x01D1 => self.arrowhead |= op != 0, + 0x01FF => { + // fLine (bit 3) is valid when fUsefLine (bit 19) is set. + if op & (1 << 19) != 0 { + self.line = Some(op & (1 << 3) != 0); + } + } + 0x0384..=0x0387 => self.wrap_dist[(pid - 0x0384) as usize] = Some(op), + 0x038F => self.posh = Some(op), + 0x0390 => self.posrelh = Some(op), + 0x0391 => self.posv = Some(op), + 0x0392 => self.posrelv = Some(op), + _ => {} + } + } + } +} + +/// An `OfficeArtCOLORREF` holding a literal RGB value (palette, scheme and system colour +/// references are not resolved). +fn color(v: u32) -> Option { + let [r, g, b, flags] = v.to_le_bytes(); + // fPaletteIndex, fSchemeIndex, fSysIndex. + if flags & 0x19 != 0 { + return None; + } + Some(Rgb(r, g, b)) +} + +/// One shape (`OfficeArtSpContainer`). +#[derive(Clone, Debug, Default)] +pub(crate) struct Sp { + pub(crate) spid: u32, + /// Shape type (`MSOSPT`), from the FSP's instance. + pub(crate) spt: u16, + /// FSP flags: fGroup, fChild, fPatriarch, fDeleted, fOleShape, fHaveMaster, fFlipH, fFlipV… + pub(crate) flags: u32, + pub(crate) props: Props, + /// `OfficeArtChildAnchor` (left, top, right, bottom) in the parent group's space. + pub(crate) child: Option<[i32; 4]>, + /// `OfficeArtFSPGR`: a group's own coordinate space. + pub(crate) group_space: Option<[i32; 4]>, + /// Has an `OfficeArtClientTextbox` (its text is in the text box story). + pub(crate) textbox: bool, +} + +impl Sp { + pub(crate) fn flip_h(&self) -> bool { + self.flags & (1 << 6) != 0 + } + pub(crate) fn flip_v(&self) -> bool { + self.flags & (1 << 7) != 0 + } + fn deleted(&self) -> bool { + self.flags & (1 << 3) != 0 + } +} + +/// A shape or a group of shapes. +#[derive(Clone, Debug)] +pub(crate) enum Node { + Shape(Sp), + /// A group: its own shape record (FSPGR, anchor) and members. + Group(Sp, Vec), +} + +impl Node { + pub(crate) fn sp(&self) -> &Sp { + match self { + Node::Shape(sp) | Node::Group(sp, _) => sp, + } + } +} + +/// Parse one `OfficeArtSpContainer`. +pub(crate) fn shape(body: &[u8], budget: &mut usize) -> Sp { + let mut sp = Sp::default(); + for rec in children(body, budget) { + match rec.typ { + rt::FSP => { + sp.spt = rec.inst; + sp.spid = u32_at(rec.body, 0).unwrap_or(0); + sp.flags = u32_at(rec.body, 4).unwrap_or(0); + } + rt::FOPT | rt::SECONDARY_FOPT | rt::TERTIARY_FOPT => sp.props.apply(rec.body, rec.inst), + rt::FSPGR | rt::CHILD_ANCHOR => { + let r = [0, 4, 8, 12].map(|o| i32_at(rec.body, o)); + if let [Some(l), Some(t), Some(r), Some(b)] = r { + if rec.typ == rt::FSPGR { + sp.group_space = Some([l, t, r, b]); + } else { + sp.child = Some([l, t, r, b]); + } + } + } + rt::CLIENT_TEXTBOX => sp.textbox = true, + _ => {} + } + } + sp +} + +/// Parse an `OfficeArtSpgrContainer`: the first shape is the group itself, the rest its members. +fn group(body: &[u8], depth: usize, budget: &mut usize) -> Option { + let recs = children(body, budget); + let mut it = recs.iter(); + let head = it.find(|r| r.typ == rt::SP_CONTAINER).map(|r| shape(r.body, budget))?; + let mut members = Vec::new(); + for rec in it { + if let Some(n) = node(rec, depth + 1, budget) { + members.push(n); + } + } + Some(Node::Group(head, members)) +} + +/// A shape or group record as a node (`None` for other records, deleted shapes or past the +/// depth cap). +fn node(rec: &Rec, depth: usize, budget: &mut usize) -> Option { + match rec.typ { + rt::SP_CONTAINER => Some(shape(rec.body, budget)).filter(|s| !s.deleted()).map(Node::Shape), + rt::SPGR_CONTAINER if depth < MAX_DEPTH => group(rec.body, depth, budget), + _ => None, + } +} + +/// The document's drawings: the BLIP store and the top-level shapes of each drawing by spid. +#[derive(Default)] +pub(crate) struct Content<'a> { + /// `OfficeArtFBSE` bodies, in BLIP store order (`pib` 1 = the first). + pub(crate) bstore: Vec<&'a [u8]>, + /// The main document's shapes (`dgglbl` 0) and the headers' (`dgglbl` 1), by spid. + pub(crate) main: HashMap, + pub(crate) header: HashMap, +} + +/// Parse `OfficeArtContent` (the `fcDggInfo` blob of the Table stream): an +/// `OfficeArtDggContainer` followed by (`dgglbl`, `OfficeArtDgContainer`) drawings. +pub(crate) fn content(blob: &[u8]) -> Content<'_> { + let mut out = Content::default(); + let mut budget = MAX_RECORDS; + let Some((dgg, mut at)) = record(blob) else { return out }; + if dgg.typ == rt::DGG_CONTAINER { + for rec in children(dgg.body, &mut budget) { + if rec.typ == rt::BSTORE_CONTAINER { + out.bstore = children(rec.body, &mut budget).into_iter().filter(|r| r.typ == rt::FBSE).map(|r| r.body).take(MAX_BSE).collect(); + } + } + } + for _ in 0..MAX_DRAWINGS { + let Some(&label) = blob.get(at) else { break }; + let Some((dg, used)) = blob.get(at + 1..).and_then(record) else { break }; + at = at.saturating_add(1 + used); + if dg.typ != rt::DG_CONTAINER { + continue; + } + let into = if label == 0 { &mut out.main } else { &mut out.header }; + for rec in children(dg.body, &mut budget) { + if rec.typ != rt::SPGR_CONTAINER { + continue; + } + // The top group: its first shape is the patriarch; the members are the drawing's shapes. + if let Some(Node::Group(_, members)) = group(rec.body, 0, &mut budget) { + for n in members { + into.entry(n.sp().spid).or_insert(n); + } + } + } + } + out +} diff --git a/crates/docbin/src/sprm.rs b/crates/docbin/src/sprm.rs index aec6218fe..7db632caf 100644 --- a/crates/docbin/src/sprm.rs +++ b/crates/docbin/src/sprm.rs @@ -1,8 +1,10 @@ //! Sprm decoding ([MS-DOC] §2.6). A Prl is a 2-byte opcode followed by an operand whose //! size comes from the opcode's `spra` field (with two variable-length exceptions). -/// `sprmPChgTabs`: custom tab stops ([MS-DOC] §2.6.29 — handled in a later pass). -const P_CHG_TABS: u16 = 0xC615; +/// `sprmPChgTabs`: custom tab stops added and removed ([MS-DOC] §2.6.3, PChgTabsOperand). +pub(crate) const P_CHG_TABS: u16 = 0xC615; +/// `sprmPChgTabsPapx`: the same for paragraph and style formatting (PChgTabsPapxOperand). +pub(crate) const P_CHG_TABS_PAPX: u16 = 0xC60D; /// `sprmTDefTable`: table cell borders and widths. const T_DEF_TABLE: u16 = 0xD608; @@ -32,7 +34,16 @@ pub(crate) fn iter(grpprl: &[u8]) -> impl Iterator> { let after = &rest[2..]; // Operand size per spra, with the exceptions whose size byte lies. let len = match (op >> 13, op) { - (_, P_CHG_TABS) => after.get(..2).map(|b| 2 + u16::from_le_bytes([b[0], b[1]]) as usize), + // PChgTabsOperand.cb is one byte, the size after it; 255 means "work it out" from the + // two counts: cb, cTabs, 4 bytes per removed stop, cTabs, 3 bytes per added stop. + (_, P_CHG_TABS) => match after.first().copied() { + Some(255) => after.get(1).and_then(|&del| { + let add = after.get(2 + 4 * del as usize).copied()?; + Some(1 + 1 + 4 * del as usize + 1 + 3 * add as usize) + }), + Some(cb) => Some(1 + cb as usize), + None => None, + }, // TDefTableOperand.cb counts the bytes after cb, incremented by 1, so the // operand is one byte longer than cb. (_, T_DEF_TABLE) => after.get(..2).map(|b| 1 + u16::from_le_bytes([b[0], b[1]]) as usize), diff --git a/crates/docbin/src/tests.rs b/crates/docbin/src/tests.rs index 52582193d..acad98605 100644 --- a/crates/docbin/src/tests.rs +++ b/crates/docbin/src/tests.rs @@ -7,6 +7,8 @@ use proptest::prelude::*; use super::read; +mod drawing; + /// Offset in the WordDocument stream where test text pieces are placed (past the FIB). const TEXT_FC: u32 = 0x600; /// Offset in the Table stream where the test Clx is placed. @@ -1628,7 +1630,7 @@ fn picture_bytes_are_capped_per_document() { pics.limit = 1000; let first = pics.get(&data, 0).expect("first fits"); // Same location again: no new bytes. - assert_eq!(pics.get(&data, 0).map(|p| p.0), Some(first.0.clone())); + assert_eq!(pics.get(&data, 0).map(|p| p.key), Some(first.key.clone())); // A second distinct picture would exceed the cap. assert!(pics.get(&data, a.len() as u32).is_none()); assert_eq!(pics.media.len(), 1); @@ -1801,3 +1803,106 @@ fn alternate_font_name_keeps_later_entries() { assert_eq!(fonts, [Some("Arial".to_string()), Some("Courier New".to_string())], "alt {alt:?}"); } } + +/// PChgTabsPapxOperand bytes: cb, removed stops, added stops with their TBDs. +fn tabs_papx(del: &[i16], add: &[(i16, u8)]) -> Vec { + let mut v = vec![0u8, del.len() as u8]; + for d in del { + v.extend_from_slice(&d.to_le_bytes()); + } + v.push(add.len() as u8); + for (x, _) in add { + v.extend_from_slice(&x.to_le_bytes()); + } + v.extend(add.iter().map(|(_, tbd)| *tbd)); + v[0] = (v.len() - 1) as u8; + v +} + +/// Custom tab stops ([MS-DOC] sprmPChgTabsPapx, sprmPChgTabs) were skipped, so tab-aligned text +/// (centred titles, right-aligned dates, dot leaders in forms) lost its alignment. Stops added by +/// a style and by the paragraph, stops it removes, and the TBD's alignment and leader are read. +#[test] +fn tab_stops_read_from_styles_and_paragraphs() { + use wordcraft_doc::props::{TabAlign, TabLeader}; + // Normal: centre at 3.25" (4680 tw), right at 6.5" (9360 tw) with a dot leader (tlc 1). + let normal = grpprl(&[(0xC60D, &tabs_papx(&[], &[(4680, 1), (9360, 2 | 1 << 3)]))]); + // The paragraph drops the centre stop and adds a left stop at 0.5" and a bar at 1". + let direct = grpprl(&[(0xC60D, &tabs_papx(&[4680], &[(720, 0), (1440, 4)]))]); + // The second paragraph uses the older sprmPChgTabs: removals carry a "close" distance. + let mut old = vec![0u8, 1]; + old.extend_from_slice(&9360i16.to_le_bytes()); + old.extend_from_slice(&25i16.to_le_bytes()); + old.push(1); + old.extend_from_slice(&2880i16.to_le_bytes()); + old.push(3 | 5 << 3); // decimal, middle dot + old[0] = (old.len() - 1) as u8; + let spec = RichSpec { + text: "\tA\tB\r\tC\r", + styles: vec![(1, 0x0FFF, "Normal", normal, Vec::new())], + fonts: vec!["Calibri"], + chpx_runs: vec![(0, Vec::new())], + papx_entries: vec![(0, 0, direct), (5, 0, grpprl(&[(0xC615, &old)]))], + }; + let doc = read(&rich_doc(&spec)).expect("opens"); + let tabs = |i: usize| { + let p = doc.body.get(i).and_then(|b| b.as_para()).expect("para"); + doc.styles.resolve_para(&p.props).tabs.iter().map(|t| (t.pos, t.align, t.leader)).collect::>() + }; + assert_eq!( + tabs(0), + [(36.0, TabAlign::Left, TabLeader::None), (72.0, TabAlign::Bar, TabLeader::None), (468.0, TabAlign::Right, TabLeader::Dot)], + "the style's stops, less the removed one, plus the paragraph's" + ); + assert_eq!(tabs(1), [(144.0, TabAlign::Decimal, TabLeader::MiddleDot), (234.0, TabAlign::Center, TabLeader::None)], "sprmPChgTabs"); + assert_eq!(doc.styles.styles.iter().find(|s| s.name == "Normal").and_then(|s| s.para.tabs.as_ref()).map(Vec::len), Some(2)); +} + +/// Hostile tab operands (counts past the end, more than 64 stops, cb 255) never panic and +/// never swallow the sprms after them. +#[test] +fn hostile_tab_operands_are_safe() { + for operand in [vec![2u8, 200, 9], vec![0], vec![255, 1, 0x10, 0x00], vec![255, 3], vec![5, 0, 70, 0, 1, 2]] { + let papx = grpprl(&[(0xC615, &operand), (0x2403, &[1])]); + let spec = RichSpec { + text: "x\r", + styles: vec![(1, 0x0FFF, "Normal", Vec::new(), Vec::new())], + fonts: vec!["Calibri"], + chpx_runs: vec![(0, Vec::new())], + papx_entries: vec![(0, 0, papx)], + }; + let _ = read(&rich_doc(&spec)); + } + let mut iter = crate::sprm::iter(&[0x15, 0xC6, 3, 0, 0, 0, 0x03, 0x24, 1]); + assert_eq!(iter.next().map(|p| (p.op, p.operand.len())), Some((0xC615, 4)), "cb is one byte: the operand is cb + 1 bytes"); + assert_eq!(iter.next().map(|p| p.op), Some(0x2403), "the next sprm is still read"); +} + +/// A section mark (0x0C) at a section's end is the section's last paragraph mark, not a page +/// break inside a paragraph: the section's properties (here continuous) land on that paragraph. +/// Before, every section of such a file was lost and each section mark became a page break. +#[test] +fn section_marks_end_paragraphs() { + use wordcraft_doc::section::SectionStart; + let spec = StructSpec { + text: "first\u{C}second\r", + header_stories: vec![], + sections: vec![(6, grpprl(&[(0x3009, &[0x00])])), (13, Vec::new())], + chpx_runs: vec![(0, Vec::new())], + papx_entries: vec![(0, 0, Vec::new()), (6, 0, Vec::new())], + plf_lst: Vec::new(), + plf_lfo: Vec::new(), + note_stories: vec![], + endnote_stories: vec![], + extra_pairs: Vec::new(), + data_stream: Vec::new(), + }; + let doc = read(&struct_doc(&spec)).expect("opens"); + assert_eq!(body_texts(&doc), ["first", "second"]); + let p0 = doc.body.first().and_then(|b| b.as_para()).expect("para"); + assert_eq!(p0.section.as_ref().map(|s| s.start), Some(SectionStart::Continuous), "the section ends at its mark"); + // A page break inside a section stays a page break. + let spec = StructSpec { text: "a\u{C}b\r", sections: vec![(4, Vec::new())], papx_entries: vec![(0, 0, Vec::new())], ..spec }; + let doc = read(&struct_doc(&spec)).expect("opens"); + assert_eq!(body_texts(&doc), ["a\u{C}b"]); +} diff --git a/crates/docbin/src/tests/drawing.rs b/crates/docbin/src/tests/drawing.rs new file mode 100644 index 000000000..ef3f269ac --- /dev/null +++ b/crates/docbin/src/tests/drawing.rs @@ -0,0 +1,276 @@ +//! Pictures and drawings: OfficeArt records ([MS-ODRAW]) hand-built for these tests from the +//! specification, with tiny pictures generated in code. No file produced by Word is used. + +use proptest::prelude::*; +use wordcraft_doc::para::{Anchor, InlineObject, ShapeKind, Wrap}; +use wordcraft_doc::props::Rgb; + +use super::{StructSpec, grpprl, plc, read, struct_doc, structured_spec}; +use crate::officeart; + +/// One OfficeArt record: header (recVer, recInstance, recType, recLen) + body. +fn rec(ver: u16, inst: u16, typ: u16, body: &[u8]) -> Vec { + let vi = (inst << 4) | (ver & 0xF); + [vi.to_le_bytes().as_slice(), &typ.to_le_bytes(), &(body.len() as u32).to_le_bytes(), body].concat() +} + +fn container(typ: u16, kids: &[Vec]) -> Vec { + rec(0xF, 0, typ, &kids.concat()) +} + +/// `OfficeArtFSP`: shape type, spid, flags. +fn fsp(spt: u16, spid: u32, flags: u32) -> Vec { + rec(2, spt, 0xF00A, &[spid.to_le_bytes(), flags.to_le_bytes()].concat()) +} + +/// `OfficeArtFOPT` with simple properties and complex ones (data appended in order). +fn fopt(simple: &[(u16, u32)], complex: &[(u16, Vec)]) -> Vec { + let mut body = Vec::new(); + for (pid, op) in simple { + body.extend_from_slice(&pid.to_le_bytes()); + body.extend_from_slice(&op.to_le_bytes()); + } + for (pid, data) in complex { + body.extend_from_slice(&(pid | 0x8000).to_le_bytes()); + body.extend_from_slice(&(data.len() as u32).to_le_bytes()); + } + for (_, data) in complex { + body.extend_from_slice(data); + } + rec(3, (simple.len() + complex.len()) as u16, 0xF00B, &body) +} + +/// A shape container with the given records. +fn sp(kids: &[Vec]) -> Vec { + container(0xF004, kids) +} + +/// A real 2×2 PNG. +fn tiny_png() -> Vec { + let img = image::RgbaImage::from_pixel(2, 2, image::Rgba([10, 20, 30, 255])); + let mut out = Vec::new(); + img.write_to(&mut std::io::Cursor::new(&mut out), image::ImageFormat::Png).unwrap(); + out +} + +/// `OfficeArtBlipPNG` (one UID, tag) and the `OfficeArtFBSE` embedding it. +fn fbse_png(png: &[u8]) -> Vec { + let blip = rec(0, 0x6E0, 0xF01E, &[[0u8; 16].as_slice(), &[0xFF], png].concat()); + let mut body = vec![6u8, 6]; + body.extend_from_slice(&[0u8; 16]); // rgbUid + body.extend_from_slice(&0u16.to_le_bytes()); // tag + body.extend_from_slice(&(blip.len() as u32).to_le_bytes()); // size + body.extend_from_slice(&1u32.to_le_bytes()); // cRef + body.extend_from_slice(&0u32.to_le_bytes()); // foDelay + body.extend_from_slice(&[0, 0, 0, 0]); // unused1, cbName, unused2, unused3 + body.extend_from_slice(&blip); + rec(2, 6, 0xF007, &body) +} + +/// UTF-16LE, NUL-terminated (a complex string property). +fn wz(s: &str) -> Vec { + s.encode_utf16().chain([0]).flat_map(|u| u.to_le_bytes()).collect() +} + +/// A `Spa` (26 bytes): spid, bounds in twips, flags, cTxbx. +fn spa(spid: u32, [l, t, r, b]: [i32; 4], flags: u16) -> Vec { + [spid.to_le_bytes().as_slice(), &l.to_le_bytes(), &t.to_le_bytes(), &r.to_le_bytes(), &b.to_le_bytes(), &flags.to_le_bytes(), &0u32.to_le_bytes()] + .concat() +} + +/// `Spa` flags from bx, by, wr and fBelowText. +fn spa_flags(bx: u16, by: u16, wr: u16, below: bool) -> u16 { + (bx << 1) | (by << 3) | (wr << 5) | (u16::from(below) << 14) +} + +/// `OfficeArtContent`: a DggContainer holding a BLIP store, then one main-document drawing. +fn art_content(fbses: &[Vec], shapes: &[Vec]) -> Vec { + let bstore = rec(0xF, fbses.len() as u16, 0xF001, &fbses.concat()); + let dgg = container(0xF000, &[bstore]); + let patriarch = sp(&[fsp(0, 1024, 0b101)]); + let spgr = container(0xF003, &[[patriarch].as_slice(), shapes].concat()); + let dg = container(0xF002, &[rec(0, 1, 0xF008, &[0u8; 8]), spgr]); + [dgg, vec![0u8], dg].concat() +} + +/// A PICF for an inline OfficeArt picture: `MM_SHAPE`, 1 × 0.5 inch goal size at 100 %. +fn picf_shape(rest: &[u8]) -> Vec { + let mut picf = vec![0u8; 0x44]; + picf[0..4].copy_from_slice(&((0x44 + rest.len()) as u32).to_le_bytes()); + picf[4..6].copy_from_slice(&0x44u16.to_le_bytes()); + picf[6..8].copy_from_slice(&0x0064u16.to_le_bytes()); // mfpf.mm = MM_SHAPE + picf[28..30].copy_from_slice(&1440i16.to_le_bytes()); + picf[30..32].copy_from_slice(&720i16.to_le_bytes()); + picf[32..34].copy_from_slice(&1000u16.to_le_bytes()); + picf[34..36].copy_from_slice(&1000u16.to_le_bytes()); + [picf.as_slice(), rest].concat() +} + +#[test] +fn inline_officeart_picture_with_crop_and_alt_text() { + let png = tiny_png(); + // A picture frame whose pib names the first FBSE that follows it, a quarter cropped + // off the left, with alt text. + let shape = sp(&[fsp(75, 1025, 0xA00), fopt(&[(0x0102, 0x4000), (0x0104, 1)], &[(0x0381, wz("Logo"))])]); + let data = picf_shape(&[shape, fbse_png(&png)].concat()); + let text = "Pic\u{1}!\r"; + let ccp = text.chars().count() as u32; + let spec = StructSpec { text, data_stream: data, chpx_runs: vec![(0, grpprl(&[(0x6A03, &0u32.to_le_bytes())]))], ..structured_spec(text, ccp) }; + let doc = read(&struct_doc(&spec)).expect("opens"); + let p0 = doc.body.first().and_then(|b| b.as_para()).expect("para"); + assert_eq!(p0.text, "Pic\u{FFFC}!"); + let Some(InlineObject::Image { media, w, h, alt, crop, float, .. }) = p0.objects.first() else { panic!("no image: {:?}", p0.objects) }; + assert_eq!((*w, *h), (54.0, 36.0)); // 72 pt less a quarter, × 36 pt + assert_eq!(*crop, [0.25, 0.0, 0.0, 0.0]); + assert_eq!(alt, "Logo"); + assert_eq!(float.wrap, Wrap::Inline); + assert_eq!(doc.media.get(media).map(|b| b.as_slice()), Some(png.as_slice())); +} + +#[test] +fn floating_picture_from_the_blip_store_keeps_position_and_wrap() { + let png = tiny_png(); + let pic = sp(&[fsp(75, 1025, 0xA00), fopt(&[(0x0104, 1)], &[])]); + let art = art_content(&[fbse_png(&png)], &[pic]); + // Two inches by one, 1" from the page's left edge and 0.5" below the top margin; square wrap. + let spas = plc(&[1, 3], &spa(1025, [1440, 720, 4320, 2160], spa_flags(1, 0, 2, false))); + let text = "A\u{8}B\r"; + let ccp = text.chars().count() as u32; + let spec = StructSpec { text, extra_pairs: vec![(40, spas), (50, art)], ..structured_spec(text, ccp) }; + let doc = read(&struct_doc(&spec)).expect("opens"); + let p0 = doc.body.first().and_then(|b| b.as_para()).expect("para"); + assert_eq!(p0.text, "A\u{FFFC}B"); + let Some(InlineObject::Image { media, w, h, float, .. }) = p0.objects.first() else { panic!("no image: {:?}", p0.objects) }; + assert_eq!((*w, *h), (144.0, 72.0)); + assert_eq!((float.wrap, float.h_rel, float.v_rel, float.x, float.y), (Wrap::Square, Anchor::Page, Anchor::Margin, 72.0, 36.0)); + assert_eq!(doc.media.get(media).map(|b| b.as_slice()), Some(png.as_slice())); +} + +#[test] +fn shapes_keep_kind_colours_wrap_and_text_box_text() { + // A red ellipse with a 2 pt blue outline behind the text, and an unfilled text box. + let ellipse = sp(&[fsp(3, 1026, 0xA00), fopt(&[(0x0181, 0x0000_00FF), (0x01C0, 0x00FF_0000), (0x01CB, 25_400)], &[])]); + let tb = sp(&[fsp(202, 1027, 0xA00), fopt(&[(0x01BF, 1 << 20)], &[]), rec(0, 0, 0xF00D, &0x0001_0000u32.to_le_bytes())]); + let art = art_content(&[], &[ellipse, tb]); + let spas = + plc(&[1, 2, 4], &[spa(1026, [0, 0, 1440, 1440], spa_flags(2, 2, 3, true)), spa(1027, [0, 0, 2880, 720], spa_flags(0, 0, 1, false))].concat()); + // The text box story: an FTXBXS whose lid is the shape's spid (cTxbx, cReusable, fReusable, + // reserved, lid, txidUndo). + let ftxbxs = + [1i32.to_le_bytes().as_slice(), &0i32.to_le_bytes(), &0i16.to_le_bytes(), &0i32.to_le_bytes(), &1027u32.to_le_bytes(), &0i32.to_le_bytes()] + .concat(); + let txbx = plc(&[0, 7], &ftxbxs); + let text = "A\u{8}\u{8}B\r"; + let ccp = text.chars().count() as u32; + // The text box story is stored as the subdocument right after the main text; the endnote + // slot of the fixture writes it there and the FIB is then patched to call it the text box + // subdocument (ccpEdn = 0, ccpTxbx = 7). + let spec = + StructSpec { text, endnote_stories: vec!["Inside\r"], extra_pairs: vec![(40, spas), (50, art), (56, txbx)], ..structured_spec(text, ccp) }; + let file = patch_word(&struct_doc(&spec), &[(0x60, 0), (0x64, 7)]); + let doc = read(&file).expect("opens"); + let p0 = doc.body.first().and_then(|b| b.as_para()).expect("para"); + assert_eq!(p0.objects.len(), 2, "{:?}", p0.objects); + let Some(InlineObject::Shape { kind, w, h, fill, stroke, stroke_width, float, .. }) = p0.objects.first() else { panic!("{:?}", p0.objects) }; + assert_eq!((*kind, *w, *h), (ShapeKind::Ellipse, 72.0, 72.0)); + assert_eq!((*fill, *stroke, *stroke_width), (Some(Rgb(255, 0, 0)), Some(Rgb(0, 0, 255)), 2.0)); + assert_eq!((float.wrap, float.h_rel, float.v_rel), (Wrap::BehindText, Anchor::Column, Anchor::Paragraph)); + let Some(InlineObject::Shape { kind, fill, float, story, .. }) = p0.objects.get(1) else { panic!("{:?}", p0.objects) }; + assert_eq!((*kind, *fill, float.wrap), (ShapeKind::TextBox, None, Wrap::TopAndBottom)); + let part = story.and_then(|id| doc.parts.get(&id)).expect("text box story part"); + assert_eq!(part.kind, wordcraft_doc::PartKind::TextBox); + assert_eq!(part.blocks.first().and_then(|b| b.as_para()).map(|p| p.text.as_str()), Some("Inside")); +} + +/// Overwrite little-endian u32s in the WordDocument stream of a compound file. +fn patch_word(file: &[u8], at: &[(usize, u32)]) -> Vec { + use std::io::Read; + let mut comp = cfb::CompoundFile::open(std::io::Cursor::new(file.to_vec())).unwrap(); + let mut streams = Vec::new(); + for name in ["WordDocument", "0Table", "Data"] { + if let Ok(mut s) = comp.open_stream(format!("/{name}")) { + let mut v = Vec::new(); + s.read_to_end(&mut v).unwrap(); + if name == "WordDocument" { + for (o, x) in at { + v[*o..*o + 4].copy_from_slice(&x.to_le_bytes()); + } + } + streams.push((name, v)); + } + } + let refs: Vec<(&str, &[u8])> = streams.iter().map(|(n, v)| (*n, v.as_slice())).collect(); + super::cfb_file(&refs) +} + +#[test] +fn dib_becomes_png_and_compressed_metafiles_inflate() { + // A 2×1 24-bit DIB (BITMAPINFOHEADER, rows padded to 4 bytes). + let mut dib = Vec::new(); + for v in [40u32, 2, 1] { + dib.extend_from_slice(&v.to_le_bytes()); + } + dib.extend_from_slice(&1u16.to_le_bytes()); + dib.extend_from_slice(&24u16.to_le_bytes()); + dib.extend_from_slice(&[0u8; 24]); + dib.extend_from_slice(&[0, 0, 255, 0, 255, 0, 0, 0]); + let r = rec(0, 0x7A8, 0xF01F, &[[0u8; 16].as_slice(), &[0xFF], &dib].concat()); + let (dib_rec, _) = officeart::record(&r).unwrap(); + let blip = officeart::blip(&dib_rec, 1 << 20).expect("DIB decodes"); + assert_eq!(blip.ext, "png"); + let img = image::load_from_memory(&blip.bytes).unwrap().to_rgb8(); + assert_eq!((img.width(), img.height(), img.get_pixel(0, 0).0), (2, 1, [255, 0, 0])); + + // A deflated WMF stays a WMF (the metafile renderer draws it). + let wmf: Vec = [1u16, 9, 0x300].iter().flat_map(|v| v.to_le_bytes()).chain([0u8; 12]).collect(); + let packed = miniz_oxide::deflate::compress_to_vec_zlib(&wmf, 6); + let mut hdr = Vec::new(); + hdr.extend_from_slice(&(wmf.len() as u32).to_le_bytes()); + hdr.extend_from_slice(&[0u8; 24]); // rcBounds, ptSize + hdr.extend_from_slice(&(packed.len() as u32).to_le_bytes()); + hdr.extend_from_slice(&[0x00, 0xFE]); // compression = deflate, filter + let r = rec(0, 0x216, 0xF01B, &[[0u8; 16].as_slice(), &hdr, &packed].concat()); + let (wmf_rec, _) = officeart::record(&r).unwrap(); + let blip = officeart::blip(&wmf_rec, 1 << 20).expect("WMF inflates"); + assert_eq!((blip.ext, blip.bytes), ("wmf", wmf)); +} + +#[test] +fn deeply_nested_groups_and_lying_lengths_are_bounded() { + // 2000 nested group containers: the walk stops at its depth cap (no stack overflow). + let mut g = sp(&[fsp(1, 2000, 0)]); + for _ in 0..2_000 { + g = container(0xF003, &[sp(&[fsp(0, 2001, 1)]), g]); + } + let _ = officeart::content(&art_content(&[], &[g])); + // A metafile claiming 4 GB uncompressed with garbage data, and an FBSE pointing far out. + let mut hdr = vec![0xFFu8; 4]; + hdr.extend_from_slice(&[0u8; 24]); + hdr.extend_from_slice(&u32::MAX.to_le_bytes()); + hdr.extend_from_slice(&[0, 0]); + let r = rec(0, 0x3D4, 0xF01A, &[[0u8; 16].as_slice(), &hdr, &[0x78, 0x9C, 1, 2, 3]].concat()); + let (emf, _) = officeart::record(&r).unwrap(); + assert!(officeart::blip(&emf, 1 << 20).is_none()); + let mut fbse = vec![0u8; 36]; + fbse[28..32].copy_from_slice(&u32::MAX.to_le_bytes()); + assert!(officeart::fbse_blip(&fbse, &[0u8; 64], 1 << 20).is_none()); +} + +proptest! { + #![proptest_config(ProptestConfig::with_cases(128))] + + #[test] + fn junk_drawings_never_panic(art in proptest::collection::vec(any::(), 0..2048), spas in proptest::collection::vec(any::(), 0..512), data in proptest::collection::vec(any::(), 0..1024)) { + let text = "A\u{8}\u{1}\u{8}B\r"; + let ccp = text.chars().count() as u32; + let picf = picf_shape(&data); + let spec = StructSpec { + text, + data_stream: picf, + chpx_runs: vec![(0, grpprl(&[(0x6A03, &0u32.to_le_bytes())]))], + extra_pairs: vec![(40, spas), (50, art.clone()), (56, art)], + ..structured_spec(text, ccp) + }; + let _ = read(&struct_doc(&spec)); + } +} diff --git a/crates/docx/src/package.rs b/crates/docx/src/package.rs index 1e20db246..dcb956a87 100644 --- a/crates/docx/src/package.rs +++ b/crates/docx/src/package.rs @@ -24,6 +24,7 @@ pub mod rt { pub const NUMBERING: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/numbering"; pub const SETTINGS: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/settings"; pub const ATTACHED_TEMPLATE: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/attachedTemplate"; + pub const FONT_TABLE: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/fontTable"; pub const THEME: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/theme"; pub const FOOTNOTES: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/footnotes"; pub const ENDNOTES: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/endnotes"; diff --git a/crates/docx/src/read/embed.rs b/crates/docx/src/read/embed.rs index e684dde0f..0fadb2b52 100644 --- a/crates/docx/src/read/embed.rs +++ b/crates/docx/src/read/embed.rs @@ -6,7 +6,8 @@ //! reaches through relationships (chart, embedded workbook, colours and style; diagram data, layout, //! quick style, colours and drawing; OLE embedding and its picture…) as opaque bytes in //! `Document::passthrough`, with their content types and relationships in an -//! [`EmbeddedManifest`](crate::package::EmbeddedManifest). +//! [`EmbeddedManifest`](crate::package::EmbeddedManifest). The font table and the fonts it embeds +//! are kept the same way. use std::sync::Arc; @@ -73,7 +74,7 @@ impl Reader<'_> { /// Keep part `root` and every part it reaches through its relationships. False when `root` /// itself can't be kept. - fn keep_part_tree(&mut self, root: &str) -> bool { + pub(super) fn keep_part_tree(&mut self, root: &str) -> bool { if self.embedded.parts.contains_key(root) { return true; } diff --git a/crates/docx/src/read/math.rs b/crates/docx/src/read/math.rs index d5d12c776..c6a8adc36 100644 --- a/crates/docx/src/read/math.rs +++ b/crates/docx/src/read/math.rs @@ -219,11 +219,9 @@ fn node(k: &El, depth: usize, out: &mut Arg) { }); } // Tracked insertions show; deletions don't. - "w:ins" | "w:moveTo" | "w:smartTag" | "w:customXml" | "m:oMath" => { - if depth <= MAX_DEPTH { - for c in k.els() { - node(c, d, out); - } + "w:ins" | "w:moveTo" | "w:smartTag" | "w:customXml" | "m:oMath" if depth <= MAX_DEPTH => { + for c in k.els() { + node(c, d, out); } } "w:sdt" => { diff --git a/crates/docx/src/read/mod.rs b/crates/docx/src/read/mod.rs index c86bb4fc7..5f16f8236 100644 --- a/crates/docx/src/read/mod.rs +++ b/crates/docx/src/read/mod.rs @@ -161,6 +161,10 @@ pub(crate) fn read_opened(pkg: &Package) -> Result { r.pc.minor_font = r.doc.settings.minor_font.clone(); let part = |t: &str| rels.by_type(t).filter(|x| !x.external).map(|x| x.target.clone()); + // The font table and the fonts embedded through it ride along as opaque parts. + if let Some(p) = part(rt::FONT_TABLE) { + r.keep_part_tree(&p); + } if let Some(p) = part(rt::THEME) { r.read_theme(&p); } @@ -229,8 +233,9 @@ pub(crate) fn read_opened(pkg: &Package) -> Result { let mut blocks = Blocks::new(); r.read_blocks(&mut sc, body, &rels, &mut blocks, 0); r.flush_pending(&mut sc, &mut blocks); - toc_field_to_heading(&mut blocks); r.doc.body = blocks; + unstyle_toc_links(&mut r.doc); + toc_field_to_heading(&mut r.doc.body); if let Some(s) = body.child("w:sectPr") { r.doc.last_section = r.read_section(s, &rels); } @@ -385,8 +390,25 @@ fn keep_glued_ids(doc: &mut Document, glued: &HashSet) { }); } -/// Word starts a TOC field in the first entry; the engine keeps it at the end of the TOC heading -/// (the paragraph holding the field marks where the TOC is). Move it there when one precedes it. +/// Word draws a table of contents' entry links without the Hyperlink character style (no blue, +/// no underline) though the entries carry it (measured in Word for Windows: a Hyperlink-styled run +/// in a hyperlink inside a TOC field's result is plain; one outside a hyperlink, another style, +/// direct formatting, or a hyperlink in another field keep their look). The style is dropped from +/// those runs so they look the same here; the links stay. +fn unstyle_toc_links(doc: &mut Document) { + for f in doc.field_ranges(wordcraft_doc::StoryRef::Body) { + if is_toc(&f.instr) { + let _ = doc.format_range(&f.start, &f.end, &|c| { + if c.link.is_some() && c.style.as_deref() == Some("Hyperlink") { + c.style = None; + } + }); + } + } +} + +/// Normalize only a complete TOC range that the engine's heading/entry representation can +/// reproduce exactly. Other imported TOCs keep their explicit boundaries, independent of styles. fn toc_field_to_heading(body: &mut Blocks) { for i in 1..body.len() { let heading = body[i - 1].as_para().filter(|p| p.props.style.as_deref() == Some("TOCHeading")); @@ -394,12 +416,23 @@ fn toc_field_to_heading(body: &mut Blocks) { if heading.is_none_or(|h| h.objects.iter().any(has_toc)) { continue; } - let Some((field, props)) = body[i].as_para().and_then(|p| match p.object_at(0) { - Some(f @ InlineObject::Field { result, .. }) if result.is_empty() && has_toc(f) => Some((f.clone(), p.props_of_char(0).clone())), - _ => None, + let Some((field, props, end)) = body[i].as_para().and_then(|p| { + let (field, end) = match p.object_at(0) { + Some(InlineObject::FieldStart { instr, locked }) if is_toc(instr) => { + (InlineObject::Field { instr: instr.clone(), result: String::new(), locked: *locked, code: false }, Some(toc_range_end(body, i)?)) + } + Some(f @ InlineObject::Field { result, .. }) if result.is_empty() && has_toc(f) => (f.clone(), None), + _ => return None, + }; + Some((field, p.props_of_char(0).clone(), end)) }) else { continue; }; + // Delete the end first: the two markers may occupy the same paragraph. + if let Some(Block::Para(p)) = end.and_then(|end| body.get_mut(end)).map(Arc::make_mut) { + let off = p.len().saturating_sub(OBJ.len_utf8()); + let _ = p.delete(off, p.len()); + } if let Some(Block::Para(p)) = body.get_mut(i).map(Arc::make_mut) { let _ = p.delete(0, OBJ.len_utf8()); } @@ -411,7 +444,32 @@ fn toc_field_to_heading(body: &mut Blocks) { } fn is_toc(instr: &str) -> bool { - instr.trim_start().get(..3).is_some_and(|k| k.eq_ignore_ascii_case("TOC")) + instr.split_whitespace().next().is_some_and(|k| k.eq_ignore_ascii_case("TOC")) +} + +/// The actual matching end must also be the end the styled writer would choose: at the end +/// of the last contiguous TOC entry, with no section break or trailing content outside it. +fn toc_range_end(body: &Blocks, start: usize) -> Option { + let entry = + |p: &wordcraft_doc::Paragraph| p.section.is_none() && p.props.style.as_deref().is_some_and(|s| s.starts_with("TOC") && s != "TOCHeading"); + let mut depth = 0_usize; + for (i, b) in body.iter().enumerate().skip(start) { + let p = b.as_para().filter(|p| entry(p))?; + for (off, o) in p.object_offsets().into_iter().zip(&p.objects) { + match o { + InlineObject::FieldStart { .. } => depth = depth.checked_add(1)?, + InlineObject::FieldEnd => { + depth = depth.checked_sub(1)?; + if depth == 0 { + let following_entry = body.get(i + 1).and_then(|b| b.as_para()).is_some_and(entry); + return (off.checked_add(OBJ.len_utf8()) == Some(p.len()) && !following_entry).then_some(i); + } + } + _ => {} + } + } + } + None } /// Built-in style names Word stores in lower case, and how the UI shows them. @@ -734,6 +792,25 @@ impl Reader<'_> { } sheet.styles.push(st); } + // Styles that give only a list level take the list from the style they are based on + // (Heading 3: level 3 of Heading 2's outline list). + let fixed: Vec> = sheet + .styles + .iter() + .map(|s| match s.para.numbering { + Some(n) if n.num == props::LEVEL_ONLY => { + let base = s.based_on.as_deref().and_then(|b| props::chain_list(&sheet, b)); + base.map(|num| wordcraft_doc::props::NumRef { num, level: n.level }) + } + other => other, + }) + .collect(); + for (s, n) in sheet.styles.iter_mut().zip(fixed) { + s.para.numbering = n; + } + if sheet.default_para.numbering.is_some_and(|n| n.num == props::LEVEL_ONLY) { + sheet.default_para.numbering = None; + } self.doc.styles = sheet; Ok(()) } diff --git a/crates/docx/src/read/props.rs b/crates/docx/src/read/props.rs index 7a307db3a..ce1668351 100644 --- a/crates/docx/src/read/props.rs +++ b/crates/docx/src/read/props.rs @@ -14,6 +14,28 @@ use wordcraft_doc::section::{ use super::drawing_color::adjust_lum; use crate::units::{int, measure, on_off, tw, u32_of}; + +/// A text frame (`w:framePr` without a drop cap, ECMA-376 §17.3.1.11), sizes and offsets in +/// points (clamped), names kept as OOXML writes them. +fn text_frame(k: &El) -> wordcraft_doc::props::TextFrame { + let len = |name: &str| tw(k, name).filter(|v| v.is_finite()).map(|v| v.clamp(-1584.0, 1584.0)); + let word = |name: &str, allowed: &[&str], default: &str| k.attr(name).filter(|v| allowed.contains(v)).unwrap_or(default).to_string(); + let align = |name: &str, allowed: &[&str]| k.attr(name).filter(|v| allowed.contains(v)).map(str::to_string); + wordcraft_doc::props::TextFrame { + w: len("w:w").filter(|v| *v > 0.0), + h: len("w:h").map(f32::abs).filter(|v| *v > 0.0), + h_exact: k.attr("w:hRule") == Some("exact"), + x: len("w:x").unwrap_or(0.0), + y: len("w:y").unwrap_or(0.0), + x_align: align("w:xAlign", &["left", "center", "right", "inside", "outside"]), + y_align: align("w:yAlign", &["top", "center", "bottom", "inside", "outside"]), + h_anchor: word("w:hAnchor", &["text", "margin", "page"], "page"), + v_anchor: word("w:vAnchor", &["text", "margin", "page"], "page"), + wrap: word("w:wrap", &["around", "auto", "notBeside", "none", "tight", "through"], "auto"), + h_space: len("w:hSpace").unwrap_or(0.0).max(0.0), + v_space: len("w:vSpace").unwrap_or(0.0).max(0.0), + } +} use crate::xml::El; /// Context for property reading: theme fonts and colours, and style id aliases. @@ -33,6 +55,27 @@ pub struct PropCtx { pub theme: Vec, } +/// A `w:numPr` that gives a level (`w:ilvl`) and no list (`w:numId`): the list is the one the +/// paragraph's style chain uses (ECMA-376 §17.3.1.19). A placeholder while reading, never left in +/// a document (`resolve_list_levels`). +pub(crate) const LEVEL_ONLY: u32 = u32::MAX; + +/// The list a level-only `numPr` refers to: the nearest numbering in style `style`'s based-on chain +/// (its own included) that names a list. `None` when no style in the chain has one. +pub(crate) fn chain_list(sheet: &wordcraft_doc::styles::StyleSheet, style: &str) -> Option { + sheet.chain(style).iter().rev().filter_map(|s| s.para.numbering).map(|n| n.num).find(|n| *n != LEVEL_ONLY) +} + +/// Resolve level-only numbering in `p` against its style chain (`style` when it has none). +pub(crate) fn resolve_level_only(sheet: &wordcraft_doc::styles::StyleSheet, p: &mut ParaProps, style: &str) { + if let Some(n) = p.numbering.as_mut().filter(|n| n.num == LEVEL_ONLY) { + match chain_list(sheet, p.style.as_deref().unwrap_or(style)) { + Some(num) => n.num = num, + None => p.numbering = None, + } + } +} + impl PropCtx { pub fn style_id(&self, id: &str) -> String { self.alias.get(id).cloned().unwrap_or_else(|| id.to_string()) @@ -189,22 +232,12 @@ impl PropCtx { _ => {} } } - // A complex-script value equal to its plain one is the same as none (see - // `CharProps::overlay`), which keeps documents round-tripping exactly. - if c.size_cs.is_some() && c.size_cs == c.size { - c.size_cs = None; - } - if c.bold_cs.is_some() && c.bold_cs == c.bold { - c.bold_cs = None; - } - if c.italic_cs.is_some() && c.italic_cs == c.italic { - c.italic_cs = None; - } c } /// `w:pPr` (without `w:sectPr`; `w:rPr` is the paragraph mark, returned separately). pub fn ppr(&self, e: &El) -> (ParaProps, CharProps) { + // (`w:numPr` with a level and no list sets `LEVEL_ONLY`, resolved by the caller.) let mut p = ParaProps::default(); let mut mark = CharProps::default(); for k in e.els() { @@ -229,13 +262,21 @@ impl PropCtx { "w:framePr" => { if matches!(k.attr("w:dropCap"), Some("drop") | Some("margin")) { p.drop_cap = Some(k.attr("w:lines").and_then(u32_of).unwrap_or(3).clamp(1, 10) as u8); + } else { + p.frame = Some(text_frame(k)); } } "w:numPr" => { let num = k.child_val("w:numId").and_then(u32_of); - let level = k.child_val("w:ilvl").and_then(u32_of).unwrap_or(0).min(8) as u8; - if let Some(num) = num { - p.numbering = Some(NumRef { num, level }); + let ilvl = k.child_val("w:ilvl").and_then(u32_of); + let level = ilvl.unwrap_or(0).min(8) as u8; + match (num, ilvl) { + (Some(num), _) => p.numbering = Some(NumRef { num, level }), + // A level without a list: the list comes from the based-on style, as in + // a Heading 3 that only says "level 3" of Heading 2's list. Resolved once + // the styles are read (`resolve_list_levels`). + (None, Some(_)) => p.numbering = Some(NumRef { num: LEVEL_ONLY, level }), + (None, None) => {} } } "w:pBdr" => p.borders = Some(borders(k)), @@ -251,6 +292,14 @@ impl PropCtx { if let Some(v) = tw(k, "w:after") { p.space_after = Some(v.max(0.0)); } + // HTML-style automatic spacing; "0" switches an inherited one off. + let auto = |name| k.attr(name).map(|v| matches!(v.trim(), "1" | "true" | "on")); + if let Some(a) = auto("w:beforeAutospacing") { + p.space_before_auto = Some(a); + } + if let Some(a) = auto("w:afterAutospacing") { + p.space_after_auto = Some(a); + } if let Some(line) = k.attr("w:line") { // Word reads the rule case-insensitively (`atleast` is seen in the wild). let rule = k.attr("w:lineRule").unwrap_or("auto").to_ascii_lowercase(); @@ -489,6 +538,7 @@ pub fn tcpr(pc: &PropCtx, e: &El) -> (CellProps, bool) { "w:tcBorders" => c.borders = Some(borders(k)), "w:shd" => c.shading = pc.shd_fill(k), "w:noWrap" => c.no_wrap = on_off(k), + "w:hideMark" => c.hide_mark = on_off(k), "w:tcMar" => c.margins = Some(margins(k)), "w:textDirection" => c.text_direction = TextDirection::from_ooxml(k.attr("w:val").unwrap_or("")), "w:vAlign" => { diff --git a/crates/docx/src/read/story.rs b/crates/docx/src/read/story.rs index c6f724ca2..fc6302e4b 100644 --- a/crates/docx/src/read/story.rs +++ b/crates/docx/src/read/story.rs @@ -14,11 +14,14 @@ use wordcraft_doc::{Block, Blocks, InlineObject, Paragraph, PartKind, RevisionKi use super::Reader; use super::props::{sectpr, tcpr, trpr}; use super::shape_style::{preset_kind, shape_paint, vml_color}; -use crate::package::{Rels, rt}; +use crate::package::{ContentTypes, Rels, rt}; use crate::units::{int, measure, u32_of}; -use crate::xml::El; +use crate::xml::{El, Node}; use wordcraft_doc::wordart::TextWarp; +const INK_URI: &str = "http://schemas.microsoft.com/office/word/2010/wordprocessingInk"; +const INK_REL: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/customXml"; + /// Deepest table nesting we build (deeper tables become plain paragraphs). const MAX_TABLE_DEPTH: usize = 24; /// Deepest field nesting we track. @@ -83,9 +86,9 @@ fn enclosure(f: &FieldState) -> Option { } /// Fields kept as ranges: those whose result another program writes, with formatting and often -/// several paragraphs (citation managers' `ADDIN` fields: Zotero, Mendeley, EndNote). +/// several paragraphs (TOCs and citation managers' `ADDIN` fields). fn is_range_instr(instr: &str) -> bool { - instr.trim_start().get(..5).is_some_and(|k| k.eq_ignore_ascii_case("ADDIN")) + super::is_toc(instr) || instr.trim_start().get(..5).is_some_and(|k| k.eq_ignore_ascii_case("ADDIN")) } /// The last paragraph of a block list: its last block, or inside a final table the last @@ -163,7 +166,7 @@ fn clean_text(s: &str) -> String { fn children_of_choice(ac: &El) -> Option<&El> { // Prefer a Choice whose requirements we understand; else the Fallback; else any Choice. - const KNOWN: &[&str] = &["wps", "wpg", "wpc", "w14", "w15", "wp14", "a14", "w16se", "w16cid", "w16", "w16cex", "w16sdtdh", "v"]; + const KNOWN: &[&str] = &["wps", "wpg", "wpc", "wpi", "w14", "w15", "wp14", "a14", "w16se", "w16cid", "w16", "w16cex", "w16sdtdh", "v"]; let ok = |c: &El| c.attr("Requires").unwrap_or("").split_whitespace().all(|r| KNOWN.contains(&r)); ac.children("mc:Choice").find(|c| ok(c)).or_else(|| ac.child("mc:Fallback")).or_else(|| ac.child("mc:Choice")) } @@ -331,7 +334,8 @@ impl Reader<'_> { pb.push_obj(o, &CharProps::default()); } if let Some(ppr) = e.child("w:pPr") { - let (props, mark) = self.pc.ppr(ppr); + let (mut props, mark) = self.pc.ppr(ppr); + super::props::resolve_level_only(&self.doc.styles, &mut props, "Normal"); para.props = props; para.mark = mark; // A tracked paragraph mark (ECMA-376 §17.13.5.16 / §17.13.5.15): `w:ins` / `w:del` @@ -353,7 +357,14 @@ impl Reader<'_> { // Tracked formatting changes (§17.13.5.29, §17.13.5.19): the paragraph properties // before the change, and a change of the list numbering. if let Some(ch) = ppr.child("w:pPrChange") { - let old = ch.child("w:pPr").map(|o| self.pc.ppr(o).0.formatting()).unwrap_or_default(); + let old = ch + .child("w:pPr") + .map(|o| { + let mut old = self.pc.ppr(o).0; + super::props::resolve_level_only(&self.doc.styles, &mut old, "Normal"); + old.formatting() + }) + .unwrap_or_default(); para.props.fmt_change = PropChange::boxed(self.revision(RevisionKind::Format, ch), old); } if para.props.numbering.is_some() @@ -500,7 +511,10 @@ impl Reader<'_> { } } "mc:AlternateContent" => { - if let Some(c) = children_of_choice(k) { + if let Some(o) = self.read_ink_alternate(sc, k, rels) { + let props = self.run_props_none(ctx); + self.emit_obj(sc, pb, o, &props); + } else if let Some(c) = children_of_choice(k) { self.read_inline_children(sc, pb, c, rels, ctx, depth + 1); } } @@ -661,7 +675,9 @@ impl Reader<'_> { } "w:ruby" => self.read_ruby(sc, pb, k, props), "mc:AlternateContent" => { - if let Some(c) = children_of_choice(k) { + if let Some(o) = self.read_ink_alternate(sc, k, rels) { + self.emit_obj(sc, pb, o, props); + } else if let Some(c) = children_of_choice(k) { for x in c.els() { self.read_run_child(sc, pb, x, props, rels, note, depth + 1); } @@ -881,6 +897,25 @@ impl Reader<'_> { let gd = c.child("a:graphic").and_then(|g| g.child("a:graphicData"))?; // Checked by URI before the blip search: a diagram's data can hold a blip further down. let uri = gd.attr("uri").unwrap_or(""); + if uri == INK_URI { + let content = gd.child("w14:contentPart")?; + let part = super::part_of(rels, content.attr("r:id")?, INK_REL)?; + let pkg = self.pkg; + if self.content_types.get_or_insert_with(|| ContentTypes::read(pkg)).of(&part) != Some("application/inkml+xml") { + return None; + } + let source = self.keep_embedded(c.child("a:graphic")?, rels, &[])?; + float.set_spin(xfrm_spin(content.child("w14:xfrm"))); + let (px, py) = float.spin_pad(w, h); + for (e, p) in float.effect.iter_mut().zip([px, py, px, py]) { + *e = (*e - p).clamp(0.0, 1584.0); + } + // The renderer's existing missing-picture box marks ink without a usable preview. + // InkML stays opaque: no approximation of its strokes or conversion to freeforms. + let items = vec![GraphicItem::Image { rect: [0.0, 0.0, w, h], media: String::new() }]; + let graphic = Arc::new(Graphic { kind: GraphicKind::Ink, items, w, h, source: Some(source), ..Default::default() }); + return Some(InlineObject::Graphic { w, h, alt, float, graphic }); + } let graphic_kind = if uri.ends_with("/chart") { Some(GraphicKind::Chart) } else if uri.ends_with("/diagram") { @@ -892,7 +927,7 @@ impl Reader<'_> { // Kept so saving writes the chart or diagram back: the frame around it is written // from the object's (possibly moved or resized) size and position. let extra: Vec = match kind { - GraphicKind::Chart => Vec::new(), + GraphicKind::Chart | GraphicKind::Ink => Vec::new(), GraphicKind::Diagram => self.diagram_drawing_rel(gd, rels).into_iter().collect(), }; let source = c.child("a:graphic").and_then(|g| self.keep_embedded(g, rels, &extra)); @@ -909,17 +944,14 @@ impl Reader<'_> { } else { let wsp = gd.find("wps:wsp")?; let mut obj = self.read_wsp(sc, wsp, rels, w, h, float); - // A freeform: its custom geometry, and whether it's ink WordCraft wrote. - if let InlineObject::Shape { kind, freeform, .. } = &mut obj - && let Some(sppr) = wsp.child("wps:spPr") - && let Some(mut f) = super::freeform::cust_geom(sppr, w, h) + // A freeform: whether it's ink WordCraft wrote. + if let InlineObject::Shape { freeform: Some(f), .. } = &mut obj + && let Some(name) = c.child("wp:docPr").and_then(|p| p.attr("name")) + && let Some(tool) = super::freeform::ink_tool(name) { - f.alpha = super::freeform::line_alpha(sppr.child("a:ln")); - let name = c.child("wp:docPr").and_then(|p| p.attr("name")); - f.ink = name.and_then(super::freeform::ink_tool); - f.order = name.filter(|_| f.ink.is_some()).and_then(super::freeform::ink_order); - *kind = ShapeKind::Freeform; - *freeform = Some(Arc::new(f)); + let f = Arc::make_mut(f); + f.ink = Some(tool); + f.order = super::freeform::ink_order(name); } obj }; @@ -938,6 +970,34 @@ impl Reader<'_> { Some(obj) } + /// Keep the selected native-ink branch and its picture fallback as one object. Reading only + /// the fallback flattens the ink; reading only the choice loses its preview on the next save. + fn read_ink_alternate(&mut self, sc: &mut StoryCtx, ac: &El, rels: &Rels) -> Option { + let choice = children_of_choice(ac).filter(|e| e.name == "mc:Choice")?; + let drawing = choice.child("w:drawing")?; + if drawing.find("a:graphicData")?.attr("uri") != Some(INK_URI) { + return None; + } + let fallback = ac.child("mc:Fallback"); + let preview = fallback.and_then(|f| f.child("w:pict")).and_then(|p| self.read_vml(sc, p, rels)); + let Some(mut object) = self.read_drawing(sc, drawing, rels) else { return preview }; + if let InlineObject::Graphic { w, h, graphic, .. } = &mut object { + let graphic = Arc::make_mut(graphic); + if let Some(InlineObject::Image { media, .. }) = preview { + graphic.items = vec![GraphicItem::Image { rect: [0.0, 0.0, *w, *h], media }]; + } + if let Some(fallback) = fallback { + let kept = + El { name: "mc:AlternateContent".into(), attrs: Vec::new(), kids: vec![Node::El(choice.clone()), Node::El(fallback.clone())] }; + if let Some(source) = self.keep_embedded(&kept, rels, &[]) { + graphic.source = Some(source); + } + // A malformed or over-budget fallback must not discard a valid native part. + } + } + Some(object) + } + /// The chart or SmartArt diagram in an `a:graphicData`, as items in points from its top-left /// corner. Charts are drawn from their cached data, SmartArt diagrams from the drawing Word /// stores beside them. Built once per part and size; once the file's graphic budget is spent, @@ -946,6 +1006,7 @@ impl Reader<'_> { let part = match kind { GraphicKind::Chart => gd.child("c:chart").and_then(|c| c.attr("r:id")).and_then(|id| super::part_of(rels, id, rt::CHART)), GraphicKind::Diagram => self.diagram_part(gd, rels), + GraphicKind::Ink => None, }; let Some(path) = part else { return Arc::new(Graphic { kind, items: Vec::new(), w, h, source, chart: None, smart_art: None }) }; let key = (kind, path, w.to_bits(), h.to_bits()); @@ -978,6 +1039,7 @@ impl Reader<'_> { None => Vec::new(), }, GraphicKind::Diagram => self.diagram_drawing(path, w, h), + GraphicKind::Ink => Vec::new(), } } @@ -1160,6 +1222,15 @@ impl Reader<'_> { if let Some(a) = bp.attr("anchor") { part.body.anchor = wordcraft_doc::props::valign_from_anchor(a); } + // Internal margins, each defaulting to Word's (#743). + let mut inset = wordcraft_doc::TEXT_BOX_INSET; + for (slot, n) in inset.iter_mut().zip(["lIns", "tIns", "rIns", "bIns"]) { + if let Some(v) = bp.attr(n).and_then(int) { + *slot = u32::try_from(v.clamp(0, i64::from(wordcraft_doc::MAX_TEXT_BOX_INSET))).unwrap_or(0); + } + } + part.body.inset = (inset != wordcraft_doc::TEXT_BOX_INSET).then_some(inset); + part.body.auto_fit = bp.child("a:spAutoFit").is_some(); } let mut extra = ShapeExtra::default(); // Arrowheads: any end type but `none` (ECMA-376 §20.1.10.33) is drawn as a triangle. @@ -1179,7 +1250,13 @@ impl Reader<'_> { self.glued.extend(extra.start.iter().chain(extra.end.iter()).map(|e| e.id)); } extra.warp = wsp.child("wps:bodyPr").and_then(|b| b.child("a:prstTxWarp")).and_then(text_warp); - InlineObject::Shape { kind, w, h, fill, stroke, stroke_width, float, story, freeform: None, effects, extra } + // A freeform: its custom geometry (#743: in a group too). + let freeform = sppr.and_then(|s| super::freeform::cust_geom(s, w, h)).map(|mut f| { + f.alpha = super::freeform::line_alpha(ln); + kind = ShapeKind::Freeform; + Arc::new(f) + }); + InlineObject::Shape { kind, w, h, fill, stroke, stroke_width, float, story, freeform, effects, extra } } /// Keep drawing id `id` on a shape read from a file (see [`Reader::glued`]). @@ -1649,7 +1726,9 @@ struct Xfrm { /// The rotation and flips of the `a:xfrm` in shape properties `sppr` (ECMA-376 §20.1.7.6: /// `rot` in 60000ths of a degree, clockwise; `flipH`, `flipV`). None when missing or junk. fn xfrm_spin(sppr: Option<&El>) -> wordcraft_geom::Spin { - let Some(x) = sppr.and_then(|s| s.child("a:xfrm")) else { return wordcraft_geom::Spin::default() }; + let Some(x) = sppr.and_then(|s| if matches!(s.name.as_str(), "a:xfrm" | "w14:xfrm") { Some(s) } else { s.child("a:xfrm") }) else { + return wordcraft_geom::Spin::default(); + }; let rot = x.attr("rot").and_then(int).map(|v| v.rem_euclid(21_600_000) as f32 / 60_000.0).unwrap_or(0.0); wordcraft_geom::Spin::new(rot, on_off_attr(x, "flipH"), on_off_attr(x, "flipV")) } diff --git a/crates/docx/src/write/embed.rs b/crates/docx/src/write/embed.rs index 5f66f0b4e..8fd0c7881 100644 --- a/crates/docx/src/write/embed.rs +++ b/crates/docx/src/write/embed.rs @@ -1,4 +1,4 @@ -//! Writing back charts, SmartArt diagrams and OLE objects read from a file (see `read::embed`). +//! Writing back charts, SmartArt diagrams, native ink and OLE objects read from a file (see `read::embed`). //! //! Each written object gets its kept markup with fresh relationship ids, pointing at copies of the //! parts it needs. The first object to use a part writes it under its original name (unless that @@ -16,7 +16,7 @@ use super::PartRels; use crate::package::{ CUSTOM_XML_ROOTS, EMBEDDED_PARTS, EmbeddedManifest, MAX_EMBEDDED_PARTS, MAX_EMBEDDED_RELS, VBA_PROJECT_PART, VBA_RELATED, rel_is, relative, rt, }; -use crate::units::emu; +use crate::units::{MAX_LEN_PT, emu, int}; use crate::xml::{self, El, MAX_DEPTH}; /// Most bytes written for copies of parts (objects pasted more than once). @@ -29,6 +29,9 @@ const PART_COST: u64 = 4096; /// Where the stories' parts live: every story this writer makes is a part in `word/`. const STORY_PART: &str = "word/document.xml"; +/// The font table kept from a file (read by `Reader::keep_part_tree`). +const FONT_TABLE_PART: &str = "word/fontTable.xml"; + /// The embedded-object parts of a package being written. pub(crate) struct EmbedWriter<'d> { manifest: EmbeddedManifest, @@ -90,6 +93,14 @@ impl<'d> EmbedWriter<'d> { } } + /// Write the font table kept from a file, with the fonts it embeds, relating it from `rels`. + pub fn font_table(&mut self, rels: &mut PartRels) { + let mut written = Vec::new(); + if let Some(name) = self.copy_tree(FONT_TABLE_PART, &mut HashMap::new(), &mut written) { + rels.add(rt::FONT_TABLE, &relative(STORY_PART, &name), false); + } + } + /// Reserve the names of the media files this writer writes. pub fn reserve_media<'a>(&mut self, files: impl Iterator) { for f in files { @@ -251,6 +262,12 @@ impl super::Writer<'_> { /// the object's current size and position into its own markup (a chart or diagram's frame is /// written by the caller). `None` when it can't be written back (its parts are gone). pub(super) fn embedded_xml(&mut self, src: &Embedded, rels: &mut PartRels, resize: Option<(f32, f32, &Float, &str)>) -> Option { + self.embedded_tree(src, rels, resize).map(|root| root.to_xml()) + } + + /// The same bounded preservation operation, retaining the parsed tree for a caller that + /// must rebuild an ink AlternateContent drawing frame without reparsing unbound prefixes. + pub(super) fn embedded_tree(&mut self, src: &Embedded, rels: &mut PartRels, resize: Option<(f32, f32, &Float, &str)>) -> Option { let Some(left) = self.embeds.markup_bytes.checked_sub(src.xml.len() as u64) else { log::warn!("docx: too much embedded-object markup; the rest is left out"); return None; @@ -305,7 +322,7 @@ impl super::Writer<'_> { if let Some((w, h, float, alt)) = resize { self.resize_object(&mut root, w, h, float, alt, 0); } - Some(root.to_xml()) + Some(root) } /// The `a:graphic` of a SmartArt graphic written from its model, laid out `w` × `h` points: @@ -437,6 +454,51 @@ impl super::Writer<'_> { } } } + + /// Ink's own transform need not fill its outer drawing frame. Scale its original offsets + /// and extents proportionally; an unrelated edit must not stretch the preserved strokes. + pub(super) fn ink_transform(e: &mut El, from: (f32, f32), to: (f32, f32), float: &Float, depth: usize) { + if depth > MAX_DEPTH { + return; + } + if e.name == "w14:xfrm" { + e.attrs.retain(|(k, _)| !matches!(k.as_str(), "rot" | "flipH" | "flipV")); + for (k, v) in super::story::spin_attrs(float.spin()) { + e.attrs.push((k.to_string(), v)); + } + let scale = |old: f32, new: f32| if old.is_finite() && old > 0.0 && new.is_finite() { (new.max(0.0) / old) as f64 } else { 1.0 }; + let (sx, sy) = (scale(from.0, to.0), scale(from.1, to.1)); + let max = MAX_LEN_PT as f64 * 12_700.0; + for c in e.kids.iter_mut().filter_map(|k| match k { + xml::Node::El(c) => Some(c), + _ => None, + }) { + if !matches!(c.name.as_str(), "a:off" | "a:ext") { + continue; + } + for (key, value) in &mut c.attrs { + let factor = match key.as_str() { + "x" | "cx" => sx, + "y" | "cy" => sy, + _ => continue, + }; + if factor != 1.0 + && let Some(n) = int(value) + { + let min = if c.name == "a:ext" { 0.0 } else { -max }; + *value = ((n as f64 * factor).clamp(min, max).round() as i64).to_string(); + } + } + } + return; + } + for c in e.kids.iter_mut().filter_map(|k| match k { + xml::Node::El(c) => Some(c), + _ => None, + }) { + Self::ink_transform(c, from, to, float, depth + 1); + } + } } /// Set an extent's `cx` and `cy` (those it has) to `w` × `h` points. @@ -527,9 +589,9 @@ fn vml_style_with(style: &str, w: f32, h: f32, float: &Float) -> String { .split(';') .filter(|d| { let key = d.split_once(':').map_or("", |(k, _)| k).trim().to_ascii_lowercase(); - !d.trim().is_empty() - && !matches!(key.as_str(), "width" | "height" | "rotation" | "flip") - && !(absolute && matches!(key.as_str(), "left" | "top" | "margin-left" | "margin-top")) + !(d.trim().is_empty() + || matches!(key.as_str(), "width" | "height" | "rotation" | "flip") + || (absolute && matches!(key.as_str(), "left" | "top" | "margin-left" | "margin-top"))) }) .map(|d| d.trim().to_string()) .collect(); diff --git a/crates/docx/src/write/mod.rs b/crates/docx/src/write/mod.rs index 4562ca689..bb6fb98bb 100644 --- a/crates/docx/src/write/mod.rs +++ b/crates/docx/src/write/mod.rs @@ -368,6 +368,8 @@ pub(crate) fn write_parts(doc: &Document, flavor: Flavor) -> Result = BTreeMap::new(); for (key, file) in &wr.media_files { diff --git a/crates/docx/src/write/props.rs b/crates/docx/src/write/props.rs index b9e083793..16ed267f6 100644 --- a/crates/docx/src/write/props.rs +++ b/crates/docx/src/write/props.rs @@ -26,9 +26,7 @@ pub fn rpr_inner(w: &mut W, c: &CharProps) { if let Some(s) = &c.style { w.val("w:rStyle", s); } - // The complex-script font is written only when set: without it, Persian/Arabic text keeps - // the style's. Complex-script bold, italic and size default to the plain ones (they come in - // pairs, see `CharProps::overlay`), so text formatted here looks the same in Word. + // Emit only explicit properties so absent values keep inheriting from the style chain. let mut fonts: Vec<(&str, &str)> = Vec::new(); if let Some(f) = &c.font { fonts.extend([("w:ascii", f.as_str()), ("w:hAnsi", f.as_str())]); @@ -43,9 +41,9 @@ pub fn rpr_inner(w: &mut W, c: &CharProps) { w.empty("w:rFonts", &fonts); } toggle(w, "w:b", c.bold); - toggle(w, "w:bCs", c.bold_cs.or(c.bold)); + toggle(w, "w:bCs", c.bold_cs); toggle(w, "w:i", c.italic); - toggle(w, "w:iCs", c.italic_cs.or(c.italic)); + toggle(w, "w:iCs", c.italic_cs); toggle(w, "w:caps", c.caps); toggle(w, "w:smallCaps", c.small_caps); toggle(w, "w:strike", c.strike); @@ -77,7 +75,7 @@ pub fn rpr_inner(w: &mut W, c: &CharProps) { if let Some(s) = c.size { w.val("w:sz", &half(s)); } - if let Some(s) = c.size_cs.or(c.size) { + if let Some(s) = c.size_cs { w.val("w:szCs", &half(s)); } if let Some(h) = c.highlight { @@ -211,6 +209,35 @@ pub fn ppr_inner(w: &mut W, p: &ParaProps, framed: bool, num_change: Option<&Cha "w:framePr", &[("w:dropCap", "drop"), ("w:lines", &n(lines.min(10) as i64)), ("w:wrap", "around"), ("w:vAnchor", "text"), ("w:hAnchor", "text")], ); + } else if let Some(f) = &p.frame { + let (fw, fh, fx, fy) = (f.w.map(twips), f.h.map(twips), twips(f.x), twips(f.y)); + let (hs, vs) = (twips(f.h_space), twips(f.v_space)); + let mut a: Vec<(&str, &str)> = Vec::new(); + if let Some(v) = &fw { + a.push(("w:w", v)); + } + if let Some(v) = &fh { + a.push(("w:h", v)); + a.push(("w:hRule", if f.h_exact { "exact" } else { "atLeast" })); + } + if f.h_space > 0.0 { + a.push(("w:hSpace", &hs)); + } + if f.v_space > 0.0 { + a.push(("w:vSpace", &vs)); + } + a.push(("w:wrap", &f.wrap)); + a.push(("w:vAnchor", &f.v_anchor)); + a.push(("w:hAnchor", &f.h_anchor)); + match &f.x_align { + Some(x) => a.push(("w:xAlign", x)), + None => a.push(("w:x", &fx)), + } + match &f.y_align { + Some(y) => a.push(("w:yAlign", y)), + None => a.push(("w:y", &fy)), + } + w.empty("w:framePr", &a); } toggle(w, "w:widowControl", p.widow_control); if let Some(nr) = p.numbering { @@ -268,7 +295,12 @@ pub fn ppr_inner(w: &mut W, p: &ParaProps, framed: bool, num_change: Option<&Cha toggle(w, "w:autoSpaceDN", p.auto_space_dn); toggle(w, "w:bidi", p.bidi); toggle(w, "w:snapToGrid", p.snap_to_grid); - if p.space_before.is_some() || p.space_after.is_some() || p.line_spacing.is_some() { + if p.space_before.is_some() + || p.space_after.is_some() + || p.line_spacing.is_some() + || p.space_before_auto.is_some() + || p.space_after_auto.is_some() + { let before = p.space_before.map(|v| twips(v.max(0.0))); let after = p.space_after.map(|v| twips(v.max(0.0))); let line = p.line_spacing.map(|l| match l { @@ -280,9 +312,15 @@ pub fn ppr_inner(w: &mut W, p: &ParaProps, framed: bool, num_change: Option<&Cha if let Some(b) = &before { a.push(("w:before", b)); } + if let Some(on) = p.space_before_auto { + a.push(("w:beforeAutospacing", if on { "1" } else { "0" })); + } if let Some(x) = &after { a.push(("w:after", x)); } + if let Some(on) = p.space_after_auto { + a.push(("w:afterAutospacing", if on { "1" } else { "0" })); + } if let Some((l, r)) = &line { a.push(("w:line", l)); a.push(("w:lineRule", r)); @@ -499,6 +537,9 @@ fn tcpr_inner(w: &mut W, c: &CellProps) { VAlign::Center => w.val("w:vAlign", "center"), VAlign::Bottom => w.val("w:vAlign", "bottom"), } + if c.hide_mark { + w.empty("w:hideMark", &[]); + } } /// `w:tblpPr` and `w:tblOverlap` for a floating table. diff --git a/crates/docx/src/write/story.rs b/crates/docx/src/write/story.rs index 508f4fdbd..8d87d6166 100644 --- a/crates/docx/src/write/story.rs +++ b/crates/docx/src/write/story.rs @@ -524,38 +524,50 @@ impl Writer<'_> { w.close("w:r"); self.rev_close(w, props); } - // A chart or diagram read from a file: its frame from the object's size and position, - // the graphic inside as read. Ones made some other way have nothing to write, except - // charts and SmartArt with a model (made or edited in WordCraft): their parts are - // written from it. + // Editable charts and SmartArt keep main's model writer; opaque graphics and native + // ink keep their original source and related parts. InlineObject::Graphic { w: gw, h: gh, alt, float, graphic } => { - let inner = match (graphic.chart.as_deref(), graphic.smart_art.as_deref(), graphic.source.as_deref()) { - // A chart WordCraft can edit: a chart part written from its model. - (Some(spec), _, _) => self.chart_graphic(spec, rels), - (None, Some(spec), _) => self.smart_art_graphic(spec, *gw, *gh, rels), - (None, None, Some(src)) => { - let Some(inner) = self.embedded_xml(src, rels, None) else { return }; - inner - } - (None, None, None) => return, + let ink = graphic.kind == wordcraft_doc::graphic::GraphicKind::Ink; + let root = match (graphic.chart.as_deref(), graphic.smart_art.as_deref(), graphic.source.as_deref()) { + (Some(spec), _, _) => crate::xml::parse(self.chart_graphic(spec, rels).as_bytes()).ok(), + (None, Some(spec), _) => crate::xml::parse(self.smart_art_graphic(spec, *gw, *gh, rels).as_bytes()).ok(), + (None, None, Some(src)) => self.embedded_tree(src, rels, ink.then_some((*gw, *gh, float, alt.as_str()))), + (None, None, None) => None, }; - if inner.is_empty() { + let Some(mut root) = root else { return }; + if ink { + Self::ink_transform(&mut root, (graphic.w, graphic.h), (*gw, *gh), float, 0); + } + let alternate = (ink && root.name == "mc:AlternateContent").then_some(&root); + let fallback = alternate.and_then(|e| e.child("mc:Fallback")); + let native = alternate.and_then(|e| e.child("mc:Choice")).and_then(|e| e.find("a:graphic")); + if alternate.is_some() && native.is_none() { return; } self.rev_open(w, props); w.open("w:r", &[]); self.rpr(w, props); + if fallback.is_some() { + w.open("mc:AlternateContent", &[]); + w.open("mc:Choice", &[("Requires", "wpi")]); + } w.open("w:drawing", &[]); let docpr = self.next_docpr(); let name = match graphic.kind { wordcraft_doc::graphic::GraphicKind::Chart => format!("Chart {docpr}"), wordcraft_doc::graphic::GraphicKind::Diagram => format!("Diagram {docpr}"), + wordcraft_doc::graphic::GraphicKind::Ink => format!("Ink {docpr}"), }; self.drawing_open(w, float, *gw, *gh, &docpr, &name, alt); w.empty("wp:cNvGraphicFramePr", &[]); - w.raw(&inner); + w.raw(&native.unwrap_or(&root).to_xml()); w.close(if float.wrap == Wrap::Inline { "wp:inline" } else { "wp:anchor" }); w.close("w:drawing"); + if let Some(fallback) = fallback { + w.close("mc:Choice"); + w.raw(&fallback.to_xml()); + w.close("mc:AlternateContent"); + } w.close("w:r"); self.rev_close(w, props); } @@ -1019,13 +1031,25 @@ impl Writer<'_> { if body.anchor != wordcraft_doc::props::VAlign::Top { attrs.push(("anchor", wordcraft_doc::props::anchor_ooxml(body.anchor))); } - match extra.warp.as_ref().and_then(|t| t.sanitized()) { - Some(warp) => { - w.open("wps:bodyPr", &attrs); - text_warp(w, &warp); - w.close("wps:bodyPr"); + // Internal margins, when the box has its own. + let inset = body.inset.map(|i| i.map(|v| v.min(wordcraft_doc::MAX_TEXT_BOX_INSET).to_string())); + if let Some(inset) = &inset { + for (n, v) in ["lIns", "tIns", "rIns", "bIns"].into_iter().zip(inset) { + attrs.push((n, v)); } - None => w.empty("wps:bodyPr", &attrs), + } + let warp = extra.warp.as_ref().and_then(|t| t.sanitized()); + if warp.is_none() && !body.auto_fit { + w.empty("wps:bodyPr", &attrs); + } else { + w.open("wps:bodyPr", &attrs); + if let Some(warp) = &warp { + text_warp(w, warp); + } + if body.auto_fit { + w.empty("a:spAutoFit", &[]); + } + w.close("wps:bodyPr"); } w.close("wps:wsp"); } @@ -1488,7 +1512,7 @@ const TOC_SDT_OPEN: &str = r#""#; fn is_toc(instr: &str) -> bool { - instr.trim_start().get(..3).is_some_and(|k| k.eq_ignore_ascii_case("TOC")) + instr.split_whitespace().next().is_some_and(|k| k.eq_ignore_ascii_case("TOC")) } /// The engine's TOC: a paragraph holding an empty `TOC` field, then its TOC-styled entries. diff --git a/crates/docx/src/xml.rs b/crates/docx/src/xml.rs index 925122c70..2d4747841 100644 --- a/crates/docx/src/xml.rs +++ b/crates/docx/src/xml.rs @@ -6,7 +6,7 @@ use quick_xml::XmlVersion; use quick_xml::events::Event; -use quick_xml::name::ResolveResult; +use quick_xml::name::{QName, ResolveResult}; use quick_xml::reader::NsReader; use crate::DocxError; @@ -26,6 +26,7 @@ pub const NAMESPACES: &[(&str, &str)] = &[ ("pic", "http://schemas.openxmlformats.org/drawingml/2006/picture"), ("mc", "http://schemas.openxmlformats.org/markup-compatibility/2006"), ("w14", "http://schemas.microsoft.com/office/word/2010/wordml"), + ("wpi", "http://schemas.microsoft.com/office/word/2010/wordprocessingInk"), ("w15", "http://schemas.microsoft.com/office/word/2012/wordml"), ("wp14", "http://schemas.microsoft.com/office/word/2010/wordprocessingDrawing"), ("wps", "http://schemas.microsoft.com/office/word/2010/wordprocessingShape"), @@ -81,6 +82,24 @@ fn prefix_for(uri: &[u8], extra: &[(&'static str, &str)]) -> &'static str { "?" } +/// Compatibility feature namespaces whose elements are not otherwise handled by this tree. +/// Keep the reader's existing feature set when canonicalizing `mc:Choice/@Requires` aliases. +fn requirement_prefix(uri: &[u8]) -> &'static str { + let prefix = prefix_for(uri, &[]); + if prefix != "?" { + return prefix; + } + match uri { + b"http://schemas.microsoft.com/office/drawing/2010/main" => "a14", + b"http://schemas.microsoft.com/office/word/2015/wordml/symex" => "w16se", + b"http://schemas.microsoft.com/office/word/2016/wordml/cid" => "w16cid", + b"http://schemas.microsoft.com/office/word/2018/wordml" => "w16", + b"http://schemas.microsoft.com/office/word/2018/wordml/cex" => "w16cex", + b"http://schemas.microsoft.com/office/word/2020/wordml/sdtdatahash" => "w16sdtdh", + _ => "?", + } +} + #[derive(Clone, Debug, PartialEq)] pub enum Node { El(El), @@ -288,6 +307,28 @@ pub fn parse_ns(bytes: &[u8], extra: &[(&'static str, &str)]) -> Result v.into_owned(), Err(_) => String::from_utf8_lossy(&a.value).into_owned(), }; + // Requires contains namespace prefixes, not feature names. Resolve them while + // the namespace scope still exists; declarations are omitted from this tree. + let val = if el.name == "mc:Choice" && aname == "Requires" { + let requirements: Vec<_> = val.split_whitespace().take(65).collect(); + if requirements.len() > 64 || requirements.iter().any(|prefix| prefix.len() > 128) { + "?".to_string() + } else { + requirements + .into_iter() + .map(|prefix| { + let qualified = format!("{prefix}:requirement"); + match r.resolver_mut().resolve_attribute(QName(qualified.as_bytes())).0 { + ResolveResult::Bound(ns) => requirement_prefix(ns.as_ref()), + _ => "?", + } + }) + .collect::>() + .join(" ") + } + } else { + val + }; el.attrs.push((aname, val)); } if matches!(ev, Event::Start(_)) { @@ -489,6 +530,14 @@ mod tests { assert_eq!(t.attr("xml:space"), Some("preserve")); } + #[test] + fn choice_requirements_resolve_namespace_aliases_and_rebindings() { + let x = br#""#; + let e = parse(x).unwrap(); + let values: Vec<_> = e.children("mc:Choice").map(|c| c.attr("Requires")).collect(); + assert_eq!(values, [Some("wpi"), Some("?"), Some("?")]); + } + #[test] fn deep_nesting_is_error() { let mut s = String::new(); diff --git a/crates/docx/tests/bidi.rs b/crates/docx/tests/bidi.rs index ede391b8d..68b008b49 100644 --- a/crates/docx/tests/bidi.rs +++ b/crates/docx/tests/bidi.rs @@ -114,8 +114,8 @@ fn reads_word_style_bidi_markup() { assert_eq!(fa.font.as_deref(), Some("Times New Roman")); assert_eq!(fa.font_cs.as_deref(), Some("B Lotus")); assert_eq!((fa.size, fa.size_cs), (Some(12.0), Some(16.0))); - // b + bCs alike is stored once: an unset complex-script bold is the same as `bold`. - assert_eq!((fa.bold, fa.bold_cs, fa.rtl), (Some(true), None, Some(true))); + // Both explicit values survive, even when they happen to be equal. + assert_eq!((fa.bold, fa.bold_cs, fa.rtl), (Some(true), Some(true), Some(true))); assert_eq!((fa.lang.as_deref(), fa.lang_bidi.as_deref()), (Some("en-US"), Some("fa-IR"))); assert_eq!(ps[1].props.bidi, Some(false)); let x = &ps[1].runs[0].props; diff --git a/crates/docx/tests/char_units.rs b/crates/docx/tests/char_units.rs index 3aa09e362..03d99c5a8 100644 --- a/crates/docx/tests/char_units.rs +++ b/crates/docx/tests/char_units.rs @@ -53,3 +53,30 @@ fn character_indents_read_and_round_trip() { let again = props(&wordcraft_docx::read(&docx(&xml[xml.find("").unwrap() + 8..xml.find("").unwrap()])).unwrap()); assert_eq!(again[..3], p[..3]); } + +/// HTML-style automatic paragraph spacing (`w:beforeAutospacing`, `w:afterAutospacing`, ECMA-376 +/// §17.3.1.33), common in text pasted from web pages: read, "0" switching an inherited one off, +/// and written back. +#[test] +fn automatic_spacing_reads_and_round_trips() { + let body = r#"a + b + c"#; + let d = wordcraft_docx::read(&docx(body)).unwrap(); + let p = props(&d); + assert_eq!((p[0].space_before_auto, p[0].space_after_auto, p[0].space_before), (Some(true), Some(true), Some(5.0))); + assert_eq!((p[1].space_before_auto, p[1].space_after_auto), (Some(false), Some(true))); + assert_eq!((p[2].space_before_auto, p[2].space_after_auto), (None, None)); + let rp = d.styles.resolve_para(&p[0]); + assert_eq!((rp.space_before, rp.space_after, rp.before_auto), (14.0, 14.0, true), "automatic spacing is 14 pt"); + assert_eq!(d.styles.resolve_para(&p[2]).space_after, 5.0); + + let mut z = zip::ZipArchive::new(std::io::Cursor::new(wordcraft_docx::write(&d).unwrap())).unwrap(); + let mut xml = String::new(); + z.by_name("word/document.xml").unwrap().read_to_string(&mut xml).unwrap(); + for attr in [r#"w:beforeAutospacing="1""#, r#"w:afterAutospacing="1""#, r#"w:beforeAutospacing="0""#] { + assert!(xml.contains(attr), "{attr} missing"); + } + let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap(); + assert_eq!(props(&again)[..3], p[..3]); +} diff --git a/crates/docx/tests/complex_script_inheritance.rs b/crates/docx/tests/complex_script_inheritance.rs new file mode 100644 index 000000000..92e11ea3c --- /dev/null +++ b/crates/docx/tests/complex_script_inheritance.rs @@ -0,0 +1,135 @@ +//! Independent complex-script formatting through style inheritance and DOCX round trips. +//! Synthetic packages, written here; no external documents or installed fonts are needed. + +use std::io::Write; +use wordcraft_doc::{Document, Paragraph}; + +fn package(run: &str, defaults: &str, styles: &str, paragraph_style: &str) -> Vec { + let ns = r#"xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main""#; + let root = r#""#; + let rels = r#""#; + let document = format!( + r#"{paragraph_style}{run}שָׁלוֹם ABC 123"# + ); + let styles = format!("{defaults}{styles}"); + let mut z = zip::ZipWriter::new(std::io::Cursor::new(Vec::new())); + for (name, data) in [ + ("_rels/.rels", root), + ("word/_rels/document.xml.rels", rels), + ("word/document.xml", document.as_str()), + ("word/styles.xml", styles.as_str()), + ] { + z.start_file(name, zip::write::SimpleFileOptions::default()).unwrap(); + z.write_all(data.as_bytes()).unwrap(); + } + z.finish().unwrap().into_inner() +} + +fn paragraph(d: &Document) -> &Paragraph { + d.body.iter().find_map(|b| b.as_para()).unwrap() +} + +fn check_generations(bytes: &[u8], check: impl Fn(&Document)) { + let mut d = wordcraft_docx::read(bytes).unwrap(); + for _ in 0..3 { + check(&d); + let next = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap(); + assert_eq!(paragraph(&next).text, paragraph(&d).text); + assert_eq!(paragraph(&next).runs, paragraph(&d).runs); + assert_eq!(paragraph(&next).props.bidi, Some(true)); + d = next; + } +} + +#[test] +fn separate_hebrew_font_size_language_and_flags_survive() { + let run = r#""#; + check_generations(&package(run, "", "", ""), |d| { + let c = ¶graph(d).runs[0].props; + assert_eq!((c.font.as_deref(), c.font_cs.as_deref()), (Some("Source Serif 4"), Some("Arial Hebrew"))); + assert_eq!((c.size, c.size_cs), (Some(12.0), Some(24.0))); + assert_eq!((c.bold, c.bold_cs, c.italic, c.italic_cs), (Some(false), Some(true), Some(false), Some(true))); + assert_eq!((c.rtl, c.cs), (Some(true), Some(true))); + assert_eq!((c.lang.as_deref(), c.lang_bidi.as_deref()), (Some("en-US"), Some("he-IL"))); + }); +} + +#[test] +fn latin_direct_properties_leave_complex_script_defaults_inherited() { + let defaults = r#""#; + let run = r#""#; + check_generations(&package(run, defaults, "", ""), |d| { + let p = paragraph(d); + let c = &p.runs[0].props; + assert_eq!((c.size_cs, c.bold_cs, c.italic_cs), (None, None, None), "Absent direct values must remain absent when saved"); + let resolved = d.styles.resolve_char(p.props.style.as_deref(), c); + assert_eq!((resolved.size, resolved.bold, resolved.italic), (10.0, false, false)); + let cs = resolved.complex(); + assert_eq!((cs.size, cs.bold, cs.italic), (18.0, true, true)); + }); +} + +#[test] +fn latin_style_properties_leave_complex_script_style_chain_inherited() { + let styles = r#" + + +"#; + let run = r#""#; + check_generations(&package(run, "", styles, r#""#), |d| { + let p = paragraph(d); + let c = d.styles.resolve_char(p.props.style.as_deref(), &p.runs[0].props); + assert_eq!((c.font.as_str(), c.size, c.bold, c.italic), ("Source Serif 4", 10.0, false, false)); + let cs = c.complex(); + assert_eq!((cs.font.as_str(), cs.size, cs.bold, cs.italic), ("Arial Hebrew", 18.0, true, true)); + assert_eq!(d.styles.get("HebrewBase").unwrap().chr.lang_bidi.as_deref(), Some("he-IL")); + }); +} + +#[test] +fn equal_explicit_complex_script_values_are_not_discarded() { + let run = r#""#; + let defaults = r#""#; + check_generations(&package(run, defaults, "", ""), |d| { + let p = paragraph(d); + let c = &p.runs[0].props; + assert_eq!((c.size_cs, c.bold_cs, c.italic_cs), (Some(10.0), Some(false), Some(false))); + let cs = d.styles.resolve_char(p.props.style.as_deref(), c).complex(); + assert_eq!((cs.size, cs.bold, cs.italic), (10.0, false, false)); + }); +} + +#[test] +fn complex_script_direct_properties_leave_latin_defaults_inherited() { + let run = r#""#; + let defaults = r#""#; + check_generations(&package(run, defaults, "", ""), |d| { + let p = paragraph(d); + let c = &p.runs[0].props; + assert_eq!((c.size, c.bold, c.italic), (None, None, None)); + assert_eq!((c.rtl, c.cs), (Some(false), Some(false))); + let resolved = d.styles.resolve_char(p.props.style.as_deref(), c); + assert_eq!((resolved.size, resolved.bold, resolved.italic), (18.0, true, true)); + let cs = resolved.complex(); + assert_eq!((cs.size, cs.bold, cs.italic), (10.0, false, false)); + }); +} + +#[test] +fn tracked_formatting_keeps_independent_complex_script_values() { + let defaults = r#""#; + let run = r#""#; + check_generations(&package(run, defaults, "", ""), |d| { + let p = paragraph(d); + let c = &p.runs[0].props; + let current = d.styles.resolve_char(None, c).complex(); + assert_eq!((current.size, current.bold, current.italic), (18.0, true, true)); + let change = c.fmt_change.as_ref().unwrap(); + assert_eq!((change.old.size_cs, change.old.bold_cs, change.old.italic_cs), (Some(10.0), Some(false), Some(false))); + let old = d.styles.resolve_char(None, &change.old).complex(); + assert_eq!((old.size, old.bold, old.italic), (10.0, false, false)); + let mut overlaid = wordcraft_doc::props::CharProps::default(); + overlaid.overlay(c); + assert_eq!(overlaid.fmt_change, c.fmt_change); + }); +} diff --git a/crates/docx/tests/fixtures.rs b/crates/docx/tests/fixtures.rs index 912d21397..3b616b83f 100644 --- a/crates/docx/tests/fixtures.rs +++ b/crates/docx/tests/fixtures.rs @@ -97,6 +97,29 @@ fn paras(d: &Document) -> Vec<&Paragraph> { d.body.iter().filter_map(|b| b.as_para()).collect() } +#[test] +fn single_entry_toc_keeps_formatted_result_and_nested_page_field() { + let d = read_body( + r#" +Contents + + + TOC \o "1-1" \h + + Intro + 1 + +"#, + ); + let p = paras(&d); + assert_eq!(p[0].plain_text(), "Contents"); + assert_eq!(p[0].objects, vec![InlineObject::Field { instr: r#"TOC \o "1-1" \h"#.into(), result: String::new(), locked: true, code: false }]); + assert_eq!(p[1].plain_text(), "Intro\t1"); + assert_eq!(p[1].props_of_char(0).bold, Some(true)); + assert_eq!(p[1].props_of_char(0).link.as_deref(), Some("#_Toc1")); + assert_eq!(p[1].objects, vec![InlineObject::Field { instr: r#"PAGEREF _Toc1 \h"#.into(), result: "1".into(), locked: false, code: false }]); +} + #[test] fn toc_complex_field_spanning_paragraphs() { let body = r#" @@ -112,13 +135,13 @@ fn toc_complex_field_spanning_paragraphs() { let p = paras(&d); assert_eq!(p.len(), 5); assert_eq!(p[0].text, "\u{FFFC}Intro\t\u{FFFC}"); - assert_eq!(p[0].objects[0], InlineObject::Field { instr: r#"TOC \o "1-3" \h \z \u"#.into(), result: String::new(), locked: false, code: false }); + assert_eq!(p[0].objects[0], InlineObject::FieldStart { instr: r#"TOC \o "1-3" \h \z \u"#.into(), locked: false }); assert_eq!(p[0].objects[1], InlineObject::Field { instr: r#"PAGEREF _Toc1 \h"#.into(), result: "1".into(), locked: false, code: false }); assert_eq!(p[0].props_of_char(3).link.as_deref(), Some("#_Toc1")); assert_eq!(p[0].props.style.as_deref(), Some("TOC1")); assert_eq!(p[1].text, "Body\t2"); assert_eq!(p[1].props_of_char(0).link.as_deref(), Some("#_Toc2")); - assert_eq!(p[2].text, ""); + assert_eq!(p[2].objects, vec![InlineObject::FieldEnd]); assert_eq!(p[3].objects, vec![InlineObject::BookmarkStart { name: "_Toc1".into() }, InlineObject::BookmarkEnd { name: "_Toc1".into() }]); assert_eq!(p[4].plain_text(), "Page 7"); assert_eq!(p[4].objects[0], InlineObject::Field { instr: "PAGE".into(), result: "7".into(), locked: false, code: false }); @@ -1123,7 +1146,8 @@ fn formatting_revisions_round_trip() { assert!(xml.contains(parent), "{tag} last: {xml}"); } assert!(xml.contains(r#""#), "{xml}"); - assert!(xml.contains(r#""#; + let member = |id: u32, txbox: bool, off: (u32, u32), ext: (u32, u32), geom: &str, paint: &str, inner: &str| { + let nv = if txbox { r#""# } else { "" }; + format!( + r#"{nv}{geom}{paint}{style}{inner}"#, + off.0, off.1, ext.0, ext.1 + ) + }; + let rect = r#""#; + let band = r#""#; + let title = r#"Cover Title"#; + let members = [ + member(11, false, (0, 0), (5486400, 5486400), rect, r#""#, ""), + member( + 12, + false, + (914400, 4572000), + (4572000, 914400), + band, + r#""#, + "", + ), + member(13, true, (457200, 457200), (4572000, 1828800), rect, r#""#, title), + ] + .concat(); + let group = format!( + r##"centertop{members}"## + ); + let subtitle = anchored_wsp( + "Subtitle", + 5753100, + 420000, + r#""#, + "", + r#"PlanAuthor"#, + ) + // `anchored_wsp` writes an empty `wps:bodyPr`: give this one margins and Resize to fit. + .replace("", r#""#); + let check = |d: &Document| { + let theme = &d.settings.theme_colors; + let objs = ¶s(d)[0].objects; + let InlineObject::Group { children, .. } = &objs[0] else { panic!("{:?}", objs[0]) }; + assert_eq!(children.len(), 3); + let shape = |i: usize| match &children[i].obj { + InlineObject::Shape { kind, fill, freeform, story, .. } => (*kind, *fill, freeform.is_some(), *story), + o => panic!("{o:?}"), + }; + let (kind, fill, free, _) = shape(0); + assert_eq!((kind, fill, free), (ShapeKind::Rectangle, Some(theme[4]), false), "the style's accent fill"); + let (kind, fill, free, _) = shape(1); + assert_eq!((kind, fill, free), (ShapeKind::Freeform, Some(theme[1]), true), "a freeform in a group keeps its outline"); + let (kind, _, _, story) = shape(2); + assert_eq!(kind, ShapeKind::TextBox); + let body = d.parts[&story.unwrap()].body; + assert_eq!((body.inset, body.auto_fit), (Some([0; 4]), false), "the title box has no margins"); + let InlineObject::Shape { story: Some(sub), .. } = &objs[1] else { panic!("{:?}", objs[1]) }; + let body = d.parts[sub].body; + assert_eq!((body.inset, body.auto_fit), (Some([0, 12700, 0, 0]), true), "the subtitle resizes to fit its text"); + }; + let d = read_body(&format!("{group}{subtitle}")); + check(&d); + check(&wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap()); +} + +#[test] +fn embedded_fonts_survive_a_round_trip() { + let table = r#""#; + let table_rels = rels(&[("rId1", "font", "fonts/Roboto-regular.ttf"), ("rId2", "font", "fonts/Roboto-bold.ttf")]); + let bytes = docx( + "Hi", + &[("rId9", "fontTable", "fontTable.xml")], + &[ + ("word/fontTable.xml", table), + ("word/_rels/fontTable.xml.rels", &table_rels), + ("word/fonts/Roboto-regular.ttf", "regular font bytes"), + ("word/fonts/Roboto-bold.ttf", "bold font bytes"), + ], + ); + let out = wordcraft_docx::write(&wordcraft_docx::read(&bytes).unwrap()).unwrap(); + let files = unzip(&out); + assert_eq!(text(&files, "word/fontTable.xml"), table); + assert_eq!(text(&files, "word/fonts/Roboto-regular.ttf"), "regular font bytes"); + assert_eq!(text(&files, "word/fonts/Roboto-bold.ttf"), "bold font bytes"); + let table_rels = text(&files, "word/_rels/fontTable.xml.rels"); + assert!(table_rels.contains(r#"Id="rId1""#) && table_rels.contains("fonts/Roboto-regular.ttf"), "{table_rels}"); + assert!(text(&files, "word/_rels/document.xml.rels").contains("fontTable.xml")); + // A second save keeps them too. + let again = unzip(&wordcraft_docx::write(&wordcraft_docx::read(&out).unwrap()).unwrap()); + assert_eq!(text(&again, "word/fonts/Roboto-bold.ttf"), "bold font bytes"); +} + +// ---- native Word ink: opaque preservation, never re-encoded as WordCraft freeforms ---- + +const INK_NS: &str = "http://schemas.microsoft.com/office/word/2010/wordprocessingInk"; +const INK_XML: &str = r##"0 0, 1 1, 2 0"##; +const INK_TYPES: &str = r#""#; + +fn ink_docx(fallback: bool, alias: bool) -> Vec { + let content = r#""#; + let drawing = format!( + r#"{content}"# + ); + let prefix = if alias { "pen" } else { "wpi" }; + let markup = if fallback { + format!( + r##"{drawing}"## + ) + } else { + drawing + }; + let body = format!("Original{markup}"); + let mut relationships = vec![("rInk", "customXml", "ink/ink1.xml")]; + let mut extra = vec![("[Content_Types].xml", INK_TYPES), ("word/ink/ink1.xml", INK_XML)]; + if fallback { + relationships.push(("rPreview", "image", "media/preview.png")); + extra.push(("word/media/preview.png", "original preview bytes")); + } + docx(&body, &relationships, &extra) +} + +fn check_ink(files: &std::collections::BTreeMap>) -> String { + let body = text(files, "word/document.xml"); + let [id] = attr_values(&body, "w14:contentPart", "r:id").try_into().unwrap_or_else(|_| panic!("ink reference lost: {body}")); + let (kind, target) = rel(&text(files, "word/_rels/document.xml.rels"), &id); + assert!(kind.ends_with("/customXml")); + let part = part_at("word", &target); + assert_eq!(text(files, &part), INK_XML); + assert!(text(files, "[Content_Types].xml").contains(&format!(r#"PartName="/{part}" ContentType="application/inkml+xml""#))); + part +} + +#[test] +fn native_ink_survives_unrelated_edit_and_two_saves() { + let mut d = wordcraft_docx::read(&ink_docx(false, false)).unwrap(); + let g = graphics(&d); + assert_eq!(g.len(), 1, "native ink must remain a model object"); + let InlineObject::Graphic { graphic, .. } = &g[0] else { panic!() }; + assert_eq!(graphic.kind.noun(), "ink"); + assert!(graphic.source.is_some()); + assert!(!graphic.items.is_empty(), "a missing preview must not silently leave a blank frame"); + let p = std::sync::Arc::make_mut(&mut d.body[0]); + let Block::Para(p) = p else { panic!() }; + p.delete(0, p.len()).unwrap(); + p.insert_text(0, "Edited", &Default::default()).unwrap(); + for _ in 0..2 { + let out = wordcraft_docx::write(&d).unwrap(); + let files = unzip(&out); + check_ink(&files); + assert!(text(&files, "word/document.xml").contains("Edited")); + d = wordcraft_docx::read(&out).unwrap(); + assert_eq!(graphics(&d).len(), 1); + } +} + +#[test] +fn native_ink_keeps_alternate_content_and_preview_with_renamed_prefixes() { + for alias in [false, true] { + let mut d = wordcraft_docx::read(&ink_docx(true, alias)).unwrap(); + let g = graphics(&d); + assert_eq!(g.len(), 1, "fallback must not replace native ink"); + let InlineObject::Graphic { graphic, .. } = &g[0] else { panic!() }; + assert!( + graphic.items.iter().any(|item| matches!(item, wordcraft_doc::graphic::GraphicItem::Image { media, .. } if d.media.contains_key(media))) + ); + for _ in 0..2 { + let out = wordcraft_docx::write(&d).unwrap(); + let files = unzip(&out); + check_ink(&files); + let body = text(&files, "word/document.xml"); + assert!(body.contains("") && body.contains("")); + let [id] = attr_values(&body, "v:imagedata", "r:id").try_into().unwrap(); + let (kind, target) = rel(&text(&files, "word/_rels/document.xml.rels"), &id); + assert!(kind.ends_with("/image")); + assert_eq!(text(&files, &part_at("word", &target)), "original preview bytes"); + d = wordcraft_docx::read(&out).unwrap(); + assert_eq!(graphics(&d).len(), 1); + } + } +} + +#[test] +fn native_ink_missing_part_uses_available_fallback() { + let bytes = ink_docx(true, true); + let mut files = unzip(&bytes); + files.remove("word/ink/ink1.xml"); + let entries: Vec<_> = files.iter().map(|(name, bytes)| (name.as_str(), bytes.as_slice())).collect(); + let d = wordcraft_docx::read(&zip(&entries)).unwrap(); + assert!(graphics(&d).is_empty()); + assert!(paras(&d).iter().flat_map(|p| &p.objects).any(|o| matches!(o, InlineObject::Image { .. }))); + let files = unzip(&wordcraft_docx::write(&d).unwrap()); + assert!(!text(&files, "word/document.xml").contains("w14:contentPart")); + assert!(files.values().any(|bytes| bytes == b"original preview bytes")); +} + +#[test] +fn native_ink_missing_optional_preview_keeps_native_part() { + let mut files = unzip(&ink_docx(true, false)); + files.remove("word/media/preview.png"); + let entries: Vec<_> = files.iter().map(|(name, bytes)| (name.as_str(), bytes.as_slice())).collect(); + let d = wordcraft_docx::read(&zip(&entries)).unwrap(); + assert_eq!(graphics(&d).len(), 1); + let files = unzip(&wordcraft_docx::write(&d).unwrap()); + check_ink(&files); + assert!(!text(&files, "word/document.xml").contains("")); +} + +#[test] +fn native_ink_requires_the_bound_namespace_and_ink_content_type() { + for wrong_namespace in [true, false] { + let mut files = unzip(&ink_docx(true, false)); + let name = if wrong_namespace { "word/document.xml" } else { "[Content_Types].xml" }; + let old = text(&files, name); + let changed = if wrong_namespace { + old.replace(&format!(r#"xmlns:wpi="{INK_NS}""#), r#"xmlns:wpi="urn:future:unsupported""#) + } else { + old.replace("application/inkml+xml", "application/xml") + }; + files.insert(name.into(), changed.into_bytes()); + let entries: Vec<_> = files.iter().map(|(name, bytes)| (name.as_str(), bytes.as_slice())).collect(); + let d = wordcraft_docx::read(&zip(&entries)).unwrap(); + assert!(graphics(&d).is_empty()); + assert!(paras(&d).iter().flat_map(|p| &p.objects).any(|o| matches!(o, InlineObject::Image { .. }))); + } +} + +#[test] +fn native_ink_copies_have_independent_parts_and_current_frames() { + let mut d = wordcraft_docx::read(&ink_docx(true, false)).unwrap(); + let [mut object] = graphics(&d).try_into().unwrap(); + let InlineObject::Graphic { w, h, float, .. } = &mut object else { panic!() }; + (*w, *h) = (288.0, 144.0); + *float = wordcraft_doc::para::Float { wrap: Wrap::Square, h_rel: Anchor::Page, v_rel: Anchor::Page, x: 72.0, y: 36.0, ..Default::default() }; + let mut p = Paragraph::new(); + p.insert_object(0, object.clone(), &Default::default()).unwrap(); + p.insert_object(p.len(), object, &Default::default()).unwrap(); + d.body = vec![wordcraft_doc::para_block(p)]; + let out = wordcraft_docx::write(&d).unwrap(); + let files = unzip(&out); + let body = text(&files, "word/document.xml"); + let ids = attr_values(&body, "w14:contentPart", "r:id"); + assert_eq!(ids.len(), 2); + let parts: Vec<_> = ids.iter().map(|id| part_at("word", &rel(&text(&files, "word/_rels/document.xml.rels"), id).1)).collect(); + assert_ne!(parts[0], parts[1]); + for part in parts { + assert_eq!(text(&files, &part), INK_XML); + } + assert_eq!(attr_values(&body, "wp:extent", "cx"), ["3657600", "3657600"]); + assert_eq!(attr_values(&body, "a:ext", "cx"), ["3657600", "3657600"]); + assert_eq!(attr_values(&body, "v:shape", "style"), ["width:288pt;height:144pt", "width:288pt;height:144pt"]); + assert_eq!(body.matches("914400").count(), 2); + let back = wordcraft_docx::read(&out).unwrap(); + let g = graphics(&back); + assert_eq!(g.len(), 2); + for object in g { + let InlineObject::Graphic { w, h, float, .. } = object else { panic!() }; + assert_eq!((w, h, float.x, float.y), (288.0, 144.0, 72.0, 36.0)); + } +} + +#[test] +fn native_ink_markup_limit_does_not_write_a_dangling_reference() { + let mut files = unzip(&ink_docx(false, false)); + let body = text(&files, "word/document.xml").replace(" = files.iter().map(|(name, bytes)| (name.as_str(), bytes.as_slice())).collect(); + let d = wordcraft_docx::read(&zip(&entries)).unwrap(); + let files = unzip(&wordcraft_docx::write(&d).unwrap()); + assert!(!text(&files, "word/document.xml").contains("w14:contentPart")); +} + +#[test] +fn native_ink_preserves_inner_transform_and_rotation_padding_on_unrelated_saves() { + let mut files = unzip(&ink_docx(false, false)); + let body = text(&files, "word/document.xml") + .replace("", r#""#) + .replace(r#""#, r#""#); + files.insert("word/document.xml".into(), body.into_bytes()); + let entries: Vec<_> = files.iter().map(|(name, bytes)| (name.as_str(), bytes.as_slice())).collect(); + let mut d = wordcraft_docx::read(&zip(&entries)).unwrap(); + for _ in 0..2 { + let [object] = graphics(&d).try_into().unwrap(); + let InlineObject::Graphic { float, .. } = object else { panic!() }; + assert_eq!(float.spin().deg, 45.0); + assert!(float.effect.iter().all(|v| v.abs() < 0.001), "rotation padding is not counted twice: {:?}", float.effect); + let out = wordcraft_docx::write(&d).unwrap(); + let files = unzip(&out); + check_ink(&files); + let body = text(&files, "word/document.xml"); + assert_eq!(attr_values(&body, "a:off", "x"), ["127000"]); + assert_eq!(attr_values(&body, "a:ext", "cx"), ["914400"]); + assert_eq!(attr_values(&body, "w14:xfrm", "rot"), ["2700000"]); + d = wordcraft_docx::read(&out).unwrap(); + } +} + +/// Word draws a table of contents' entry links without the Hyperlink character style (measured in +/// Word for Windows); the import drops the style from them, keeping the link. A Hyperlink-styled run +/// outside a hyperlink, and a hyperlink in another field, keep it. +#[test] +fn toc_entry_links_lose_the_hyperlink_style() { + let styles = format!( + r#" + "# + ); + let field = |instr: &str, inner: &str| { + format!( + r#"{instr}{inner}"# + ) + }; + let link = + |t: &str| format!(r#"{t}"#); + let body = [ + field(r#" TOC \o "1-3" \h \z \u "#, &link("entry")), + field(" TOC ", r#"styled"#), + field(" QUOTE x ", &link("quoted")), + format!("{}", link("plain")), + r#"Target"#.to_string(), + ] + .concat(); + let d = wordcraft_docx::read(&docx(&body, &[("rId1", "styles", "styles.xml")], &[("word/styles.xml", &styles)])).unwrap(); + let run = |word: &str| { + paras(&d).into_iter().find_map(|p| p.text.find(word).map(|at| p.props_of_char(at).clone())).unwrap_or_else(|| panic!("no run with {word}")) + }; + let entry = run("entry"); + assert_eq!((entry.style.as_deref(), entry.link.as_deref()), (None, Some("#_Toc1")), "a TOC entry: plain, still a link"); + assert_eq!(run("styled").style.as_deref(), Some("Hyperlink"), "not in a hyperlink: kept"); + assert_eq!(run("quoted").style.as_deref(), Some("Hyperlink"), "another field: kept"); + assert_eq!(run("plain").style.as_deref(), Some("Hyperlink"), "outside fields: kept"); +} + +/// A style or paragraph whose `w:numPr` gives only a level (`w:ilvl`) takes the list from its +/// style chain (ECMA-376 §17.3.1.19): a Heading 3 based on a numbered Heading 2 is level 3 of the +/// same list. WordCraft dropped the level, so Heading 3s numbered as Heading 2s (6.15 where Word +/// shows 6.4.5 in NIH's monitoring-plan template). +#[test] +fn a_list_level_without_a_list_takes_the_style_s_list() { + let styles = format!( + r#" + + + + + + "# + ); + let body = r#"three + direct level + no list anywhere"#; + let d = wordcraft_docx::read(&docx(body, &[("rId1", "styles", "styles.xml")], &[("word/styles.xml", &styles)])).unwrap(); + let num = |id: &str| d.styles.get(id).and_then(|s| s.para.numbering).map(|n| (n.num, n.level)); + assert_eq!(num("Heading3"), Some((15, 2)), "Heading 3 is level 3 of Heading 2's list"); + assert_eq!(num("Heading4"), Some((15, 3)), "through two level-only styles"); + assert_eq!(num("Loose"), None, "no list in the chain: no numbering"); + let p = paras(&d); + assert_eq!(d.styles.resolve_para(&p[0].props).numbering.map(|n| (n.num, n.level)), Some((15, 2))); + assert_eq!(p[1].props.numbering.map(|n| (n.num, n.level)), Some((15, 2)), "a paragraph's own level, its style's list"); + assert_eq!(p[2].props.numbering, None); + // Saved and read again it is an ordinary list reference. + let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap(); + assert_eq!(again.styles.get("Heading3").and_then(|s| s.para.numbering).map(|n| (n.num, n.level)), Some((15, 2))); +} + +/// `w:hideMark` on a cell is read and written back (it was dropped, and the empty rows NIH's +/// tables collapse with it showed as blank lines). +#[test] +fn cell_hide_mark_reads_and_round_trips() { + let body = r#""#; + let hides = |d: &Document| -> Vec { + d.body + .iter() + .find_map(|b| { + if let wordcraft_doc::Block::Table(t) = &**b { Some(t.rows[0].cells.iter().map(|c| c.props.hide_mark).collect()) } else { None } + }) + .unwrap() + }; + let d = read_body(body); + assert_eq!(hides(&d), [true, false]); + let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap(); + assert_eq!(hides(&again), [true, false]); +} + +/// Text frames (`w:framePr` without a drop cap), as NIH's templates put a footer's page number in +/// one: read with their size, anchors, alignment and wrap, and written back (they were dropped on +/// save). +#[test] +fn text_frames_read_and_round_trip() { + let body = r#"3 + boxed + junk + D + rop cap text"#; + let d = read_body(body); + let p = paras(&d); + let f = p[0].props.frame.clone().expect("a frame"); + assert_eq!( + (f.wrap.as_str(), f.v_anchor.as_str(), f.h_anchor.as_str(), f.x_align.as_deref(), f.y), + ("around", "text", "margin", Some("center"), 0.05) + ); + let g = p[1].props.frame.clone().expect("a frame"); + assert_eq!((g.w, g.h, g.h_exact, g.h_space, g.wrap.as_str(), g.x, g.y), (Some(144.0), Some(36.0), true, 9.0, "notBeside", 72.0, 144.0)); + let cap = p.iter().find(|q| q.props.drop_cap.is_some()).expect("the drop cap"); + assert!(cap.props.frame.is_none(), "a drop cap stays a drop cap"); + let junk = p.iter().find(|q| q.text.contains("junk")).and_then(|q| q.props.frame.clone()).expect("a frame"); + assert_eq!((junk.wrap.as_str(), junk.h_anchor.as_str(), junk.w, junk.x), ("auto", "page", None, 0.0), "unknown values fall back"); + let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap(); + let q = paras(&again); + assert_eq!(q[0].props.frame, p[0].props.frame); + assert_eq!(q[1].props.frame, p[1].props.frame); + assert!(q.iter().any(|x| x.props.drop_cap == Some(2) && x.props.frame.is_none())); +} diff --git a/crates/docx/tests/roundtrip.rs b/crates/docx/tests/roundtrip.rs index 4d9759616..791cf203f 100644 --- a/crates/docx/tests/roundtrip.rs +++ b/crates/docx/tests/roundtrip.rs @@ -141,6 +141,8 @@ fn every_para_prop_round_trips() { indent_first_chars: Some(-1.0), space_before: Some(6.0), space_after: Some(12.0), + space_before_auto: Some(true), + space_after_auto: Some(false), line_spacing: Some(LineSpacing::Multiple(1.5)), contextual_spacing: Some(true), mirror_indents: Some(true), @@ -163,6 +165,20 @@ fn every_para_prop_round_trips() { suppress_line_numbers: Some(true), bidi: Some(false), drop_cap: None, + frame: Some(wordcraft_doc::props::TextFrame { + w: Some(144.0), + h: Some(36.0), + h_exact: true, + x: 0.0, + y: 12.0, + x_align: Some("right".into()), + y_align: None, + h_anchor: "margin".into(), + v_anchor: "text".into(), + wrap: "around".into(), + h_space: 9.0, + v_space: 0.0, + }), kinsoku: Some(false), word_wrap: Some(false), overflow_punct: Some(false), @@ -685,41 +701,46 @@ fn word_note_mark_round_trips_once() { /// or Word shows them as plain text with no Update Table. #[test] fn toc_field_wraps_its_entries() { - let mut head = Paragraph::with_text("Contents", CharProps::default()).styled("TOCHeading"); - let end = head.len(); - head.insert_object( - end, - InlineObject::Field { instr: "TOC \\o \"1-2\" \\h \\z \\u".into(), result: String::new(), locked: false, code: false }, - &CharProps::default(), - ) - .unwrap(); - let entries = [("Intro\t1", "TOC1"), ("Detail\t2", "TOC2")]; - let mut blocks = vec![head]; - blocks.extend(entries.iter().map(|(t, st)| Paragraph::with_text(t, CharProps::default()).styled(st))); - blocks.push(Paragraph::with_text("Body", CharProps::default())); - let d = doc_with(blocks); + for entries in [&[("Intro\t1", "TOC1")][..], &[("Intro\t1", "TOC1"), ("Detail\t2", "TOC2")][..]] { + let mut head = Paragraph::with_text("Contents", CharProps::default()).styled("TOCHeading"); + let end = head.len(); + head.insert_object( + end, + InlineObject::Field { instr: "TOC \\o \"1-2\" \\h \\z \\u".into(), result: String::new(), locked: false, code: false }, + &CharProps::default(), + ) + .unwrap(); + let mut blocks = vec![head]; + blocks.extend(entries.iter().map(|(t, st)| Paragraph::with_text(t, CharProps::default()).styled(st))); + blocks.push(Paragraph::with_text("Body", CharProps::default())); + let d = doc_with(blocks); - let bytes = wordcraft_docx::write(&d).expect("write"); - let mut zip = zip::ZipArchive::new(std::io::Cursor::new(&bytes)).unwrap(); - let mut xml = String::new(); - std::io::Read::read_to_string(&mut zip.by_name("word/document.xml").unwrap(), &mut xml).unwrap(); - let sdt = xml.find("").expect("TOC content control"); - assert!(xml.contains(r#""#), "{xml}"); - let (begin, fend) = (xml.find(r#"fldCharType="begin""#).unwrap(), xml.find(r#"fldCharType="end""#).unwrap()); - let last_entry = xml.find("Detail").unwrap(); - assert!(sdt < xml.find("Contents").unwrap() && xml.find("").unwrap() < xml.find(">Body<").unwrap(), "{xml}"); - assert!(begin < xml.find("Intro").unwrap() && fend > last_entry && fend < xml.find("").unwrap(), "field must span the entries: {xml}"); - // Word starts the field in the first entry; started in the heading, LibreOffice splits the heading. - let heading_end = xml[..xml.find("Intro").unwrap()].rfind("").unwrap(); - assert!(begin > heading_end, "field must start in the first entry, not the heading: {xml}"); + let bytes = wordcraft_docx::write(&d).expect("write"); + let mut zip = zip::ZipArchive::new(std::io::Cursor::new(&bytes)).unwrap(); + let mut xml = String::new(); + std::io::Read::read_to_string(&mut zip.by_name("word/document.xml").unwrap(), &mut xml).unwrap(); + let sdt = xml.find("").expect("TOC content control"); + assert!(xml.contains(r#""#), "{xml}"); + let (begin, fend) = (xml.find(r#"fldCharType="begin""#).unwrap(), xml.find(r#"fldCharType="end""#).unwrap()); + let last_entry = xml.find(entries.last().unwrap().0.split('\t').next().unwrap()).unwrap(); + assert!(sdt < xml.find("Contents").unwrap() && xml.find("").unwrap() < xml.find(">Body<").unwrap(), "{xml}"); + assert!( + begin < xml.find("Intro").unwrap() && fend > last_entry && fend < xml.find("").unwrap(), + "field must span the entries: {xml}" + ); + // Word starts the field in the first entry; started in the heading, LibreOffice splits the heading. + let heading_end = xml[..xml.find("Intro").unwrap()].rfind("").unwrap(); + assert!(begin > heading_end, "field must start in the first entry, not the heading: {xml}"); - // WordCraft reads its own TOC back unchanged. - let r = wordcraft_docx::read(&bytes).expect("read"); - let ps = paras(&r); - assert_eq!(ps.iter().map(|p| p.plain_text()).collect::>(), ["Contents", "Intro\t1", "Detail\t2", "Body"]); - assert!(matches!(ps[0].objects.first(), Some(InlineObject::Field { instr, .. }) if instr.starts_with("TOC")), "{:?}", ps[0].objects); - assert!(ps[1].objects.is_empty(), "the field goes back on the heading: {:?}", ps[1].objects); - assert_eq!(ps[1].props.style.as_deref(), Some("TOC1")); + // WordCraft reads its own TOC back unchanged. + let r = wordcraft_docx::read(&bytes).expect("read"); + let ps = paras(&r); + let expected: Vec<_> = std::iter::once("Contents").chain(entries.iter().map(|(t, _)| *t)).chain(std::iter::once("Body")).collect(); + assert_eq!(ps.iter().map(|p| p.plain_text()).collect::>(), expected); + assert!(matches!(ps[0].objects.first(), Some(InlineObject::Field { instr, .. }) if instr.starts_with("TOC")), "{:?}", ps[0].objects); + assert!(ps[1].objects.is_empty(), "the field goes back on the heading: {:?}", ps[1].objects); + assert_eq!(ps[1].props.style.as_deref(), Some("TOC1")); + } } /// Table Layout › Text Direction (#226): each direction is written as Word's `w:textDirection` diff --git a/crates/docx/tests/text_boxes.rs b/crates/docx/tests/text_boxes.rs index df18ff7ed..2831c9ca4 100644 --- a/crates/docx/tests/text_boxes.rs +++ b/crates/docx/tests/text_boxes.rs @@ -53,9 +53,9 @@ fn text_box(d: &mut Document, text: &str, body: TextBody) -> InlineObject { #[test] fn direction_alignment_and_links_round_trip() { let mut d = Document::from_text("a\nb\nc"); - let a = text_box(&mut d, "Flows on", TextBody { vert: TextVert::Vert, anchor: VAlign::Center, next: None }); + let a = text_box(&mut d, "Flows on", TextBody { vert: TextVert::Vert, anchor: VAlign::Center, ..Default::default() }); let b = text_box(&mut d, "", TextBody::default()); - let c = text_box(&mut d, "", TextBody { vert: TextVert::Stacked, anchor: VAlign::Bottom, next: None }); + let c = text_box(&mut d, "", TextBody { vert: TextVert::Stacked, anchor: VAlign::Bottom, ..Default::default() }); for (i, o) in [a, b, c].into_iter().enumerate() { d.insert_object(&wordcraft_doc::Pos::body(i, 1), o, &Default::default()).unwrap(); } diff --git a/crates/docx/tests/toc_ranges.rs b/crates/docx/tests/toc_ranges.rs new file mode 100644 index 000000000..d8010bd49 --- /dev/null +++ b/crates/docx/tests/toc_ranges.rs @@ -0,0 +1,177 @@ +//! Original field-boundary fixtures from the public OOXML complex-field contract. +use std::io::Write; + +use wordcraft_doc::{InlineObject, Path, StoryRef}; + +fn package(body: &str) -> Vec { + let xml = format!(r#"{body}"#); + let mut zip = zip::ZipWriter::new(std::io::Cursor::new(Vec::new())); + for (name, value) in [ + ("word/document.xml", xml.as_str()), + ( + "_rels/.rels", + r#""#, + ), + ] { + zip.start_file(name, zip::write::SimpleFileOptions::default()).unwrap(); + zip.write_all(value.as_bytes()).unwrap(); + } + zip.finish().unwrap().into_inner() +} + +fn start(code: &str) -> String { + format!( + r#"{code}"# + ) +} + +const END: &str = r#""#; +const PAGE: &str = r#"7"#; +const ENTRY: &str = r#"Scope"#; + +#[test] +fn unstyled_toc_keeps_range_nested_field_and_formatting_through_two_edited_saves() { + for split in [false, true] { + let content = if split { + format!("{ENTRY}{PAGE}Delivery") + } else { + format!("{ENTRY}{PAGE}") + }; + let body = format!( + r#"Before.{}{content}{END}After."#, + start(r#"TOC \o "1-2" \h"#) + ); + let mut doc = wordcraft_docx::read(&package(&body)).unwrap(); + for cycle in 0..=2 { + let ranges = doc.field_ranges(StoryRef::Body); + assert_eq!(ranges.len(), 1); + assert_eq!(ranges[0].instr, r#"TOC \o "1-2" \h"#); + assert!(ranges[0].locked); + assert_eq!(ranges[0].start.path, Path::top(1)); + assert_eq!(ranges[0].start.off, 0); + assert_eq!(ranges[0].end.path, Path::top(if split { 2 } else { 1 })); + let entry = doc.body[1].as_para().unwrap(); + assert_eq!(entry.plain_text(), "Scope\t7"); + let props = entry.props_of_char(wordcraft_doc::para::OBJ.len_utf8()); + assert_eq!(props.bold, Some(true)); + assert_eq!(props.italic, Some(true)); + assert_eq!(props.link.as_deref(), Some("#Original")); + assert!( + entry.objects.iter().any(|o| matches!(o, InlineObject::Field { instr, result, .. } if instr == "PAGEREF Original" && result == "7")) + ); + assert_eq!(doc.body.last().unwrap().as_para().unwrap().plain_text(), "After."); + if cycle == 2 { + break; + } + doc.para_mut(StoryRef::Body, &Path::top(0)).unwrap().insert_text(0, "Independent edit. ", &Default::default()).unwrap(); + doc = wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap(); + } + } +} + +#[test] +fn styled_heading_does_not_expand_partial_custom_or_short_ranges() { + let heading = r#"Contents"#; + let extra = r#"Outside entry."#; + for (style, prefix, suffix, following) in + [("TOC1", "Prefix.", "", ""), ("TOC1", "", "Suffix.", ""), ("TOC1", "", "", extra), ("CompanyContents", "", "", "")] + { + let body = format!( + r#"{heading}{prefix}{}Scope{END}{suffix}{following}Body."#, + start("TOC") + ); + let mut doc = wordcraft_docx::read(&package(&body)).unwrap(); + let text = doc.plain_text(StoryRef::Body); + for cycle in 0..=2 { + let ranges = doc.field_ranges(StoryRef::Body); + assert_eq!(ranges.len(), 1, "{style} {prefix} {suffix} {following}"); + assert_eq!(ranges[0].start.path, Path::top(1)); + assert_eq!(ranges[0].end.path, Path::top(1)); + assert_eq!(ranges[0].start.off, prefix.len()); + assert_eq!(ranges[0].end.off, prefix.len() + wordcraft_doc::para::OBJ.len_utf8() + "Scope".len()); + assert_eq!(doc.plain_text(StoryRef::Body), text); + if cycle == 2 { + break; + } + doc = wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap(); + } + } +} + +#[test] +fn simple_lowercase_toc_and_multiple_tocs_keep_their_own_ranges() { + let body = format!( + r#"{ENTRY}{PAGE}Between.{}Other result.{END}After."#, + start(r#"TOC \o "1-1""#) + ); + let mut doc = wordcraft_docx::read(&package(&body)).unwrap(); + for cycle in 0..=2 { + let r = doc.field_ranges(StoryRef::Body); + assert_eq!(r.len(), 2); + assert_eq!(r[0].instr, r#"toc \h"#); + assert_eq!(r[1].instr, r#"TOC \o "1-1""#); + assert!(r.iter().all(|r| r.locked)); + assert!(r[0].end.off < r[1].start.off); + assert_eq!(doc.plain_text(StoryRef::Body), "Scope\t7Between.Other result.After."); + if cycle == 2 { + break; + } + doc = wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap(); + } +} + +#[test] +fn toc_in_table_cell_does_not_capture_the_next_cell() { + let body = format!( + r#"{}{ENTRY}{PAGE}{END}Other cell."#, + start("TOC") + ); + let mut doc = wordcraft_docx::read(&package(&body)).unwrap(); + for cycle in 0..=2 { + let ranges = doc.field_ranges(StoryRef::Body); + assert_eq!(ranges.len(), 1); + assert_eq!(ranges[0].start.path, wordcraft_doc::Path(vec![0, 0, 0, 0])); + assert_eq!(ranges[0].end.path, ranges[0].start.path); + assert_eq!(doc.body[0].as_table().unwrap().rows[0].cells[1].blocks[0].as_para().unwrap().plain_text(), "Other cell."); + if cycle == 2 { + break; + } + doc = wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap(); + } +} + +#[test] +fn missing_end_and_excessive_nesting_stay_bounded_and_save_balanced() { + for (starts, ends) in [(1, 0), (40, 40)] { + let body = format!("{}Visible result.{}", start("TOC").repeat(starts), END.repeat(ends)); + let doc = wordcraft_docx::read(&package(&body)).unwrap(); + let saved = wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap(); + assert!(!saved.has_unbalanced_field_ranges()); + assert_eq!(saved.plain_text(StoryRef::Body), "Visible result."); + assert!(saved.field_ranges(StoryRef::Body).len() <= 32); + } +} + +#[test] +fn section_break_keeps_an_explicit_range_and_toc_prefix_is_not_a_toc() { + let body = format!( + r#"Contents{}Scope{END}{}Other.{END}"#, + start("TOC"), + start("TOC_OTHER") + ); + let doc = wordcraft_docx::read(&package(&body)).unwrap(); + for d in [&doc, &wordcraft_docx::read(&wordcraft_docx::write(&doc).unwrap()).unwrap()] { + let ranges = d.field_ranges(StoryRef::Body); + assert_eq!(ranges.len(), 1); + assert_eq!(ranges[0].start.path, Path::top(1)); + assert!(d.body[1].as_para().unwrap().section.is_some()); + assert!( + d.body[2] + .as_para() + .unwrap() + .objects + .iter() + .any(|o| matches!(o, InlineObject::Field { instr, result, .. } if instr == "TOC_OTHER" && result == "Other.")) + ); + } +} diff --git a/crates/engine/src/catalog.rs b/crates/engine/src/catalog.rs index 92d2cdf98..dffb839b7 100644 --- a/crates/engine/src/catalog.rs +++ b/crates/engine/src/catalog.rs @@ -20,6 +20,7 @@ File|Backstage|Properties|file.properties File|Backstage|Export PDF|file.exportPdf File|Backstage|Export Image|file.exportPng File|Backstage|Options|file.options +File|Options|Numerals|view.numeral File|Backstage|Share|file.share File|Backstage|Protect Document|file.protect File|Backstage|Inspect Document|file.inspect @@ -30,6 +31,7 @@ File|Backstage|Recover Unsaved|file.recover File|Backstage|New from Template|file.newFromTemplate File|Backstage|Save as Template|file.saveTemplate Quick Access|Toolbar|Undo|edit.undo +Quick Access|Toolbar|Undo List|edit.undoTo Quick Access|Toolbar|Redo|edit.redo Quick Access|Toolbar|Repeat|edit.repeat Quick Access|Toolbar|AutoSave|file.autosave @@ -355,6 +357,7 @@ Table Design|Borders|Borders|table.borders Table Design|Borders|Border Painter|table.borderPainter Table Layout|Table|Select Table|table.selectTable Table Layout|Table|Select Row|table.selectRow +Table Layout|Table|Select Column|table.selectColumn Table Layout|Table|Select Cell|table.selectCell Table Layout|Table|View Gridlines|table.viewGridlines Table Layout|Table|Properties|table.properties diff --git a/crates/engine/src/cmd/citations.rs b/crates/engine/src/cmd/citations.rs index 2793b1363..54bebe535 100644 --- a/crates/engine/src/cmd/citations.rs +++ b/crates/engine/src/cmd/citations.rs @@ -3,6 +3,7 @@ //! of authorities, footnote options. use serde_json::{Value, json}; +use wordcraft_doc::collate::{SortKey, heading_letter, sort_key}; use wordcraft_doc::para::InlineObject; use wordcraft_doc::props::{CharProps, TabAlign, TabLeader, TabStop}; use wordcraft_doc::{Block, Document, Paragraph, Path, Pos, Source, StoryRef}; @@ -563,7 +564,7 @@ fn citation(s: &mut Session, v: &Value) -> CmdResult { let src = s.doc.sources.iter().find(|x| x.tag == tag).cloned().ok_or_else(|| CmdError::Params(format!("no source `{tag}`")))?; let pages = p::str(v, "pages").unwrap_or("").to_string(); let st = style(&s.bib_style); - let n = number_of(&s.doc, &bib_order(&s.doc, st), &tag); + let n = number_of(&s.doc, &bib_order(&s.doc, st, super::default_lang(s)), &tag); let props = s.typing_props(); let at = delete_selection(s)?; let instr = if pages.is_empty() { format!("CITATION {tag}") } else { format!("CITATION {tag} \\p {pages}") }; @@ -611,7 +612,7 @@ fn author_key(src: &Source) -> String { /// Indices into `doc.sources` in bibliography order: the cited sources (every source while /// nothing is cited), sorted or in order of first citation as the style asks. -fn bib_order(doc: &Document, st: &CiteStyle) -> Vec { +fn bib_order(doc: &Document, st: &CiteStyle, lang: Option) -> Vec { let cited = cited_tags(doc); let mut idx: Vec = if cited.is_empty() { (0..doc.sources.len()).collect() @@ -621,8 +622,10 @@ fn bib_order(doc: &Document, st: &CiteStyle) -> Vec { let cited: std::collections::HashSet<&str> = cited.iter().map(String::as_str).collect(); doc.sources.iter().enumerate().filter(|(_, x)| cited.contains(x.tag.as_str())).map(|(i, _)| i).collect() }; - let key = - |i: &usize, by_title: bool| doc.sources.get(*i).map(|x| if by_title { x.title.to_lowercase() } else { author_key(x) }).unwrap_or_default(); + // Sorted in the document's language (Polish authors `Łukasiewicz` before `Mazur`). + let key = |i: &usize, by_title: bool| { + sort_key(&doc.sources.get(*i).map(|x| if by_title { x.title.clone() } else { author_key(x) }).unwrap_or_default(), lang.as_deref()) + }; match st.order { Order::Author => idx.sort_by_cached_key(|i| key(i, false)), Order::Title => idx.sort_by_cached_key(|i| key(i, true)), @@ -645,7 +648,7 @@ pub fn update_citations(s: &mut Session) -> Result<(), CmdError> { /// Refresh the results of the CITATION fields in the body. fn refresh_citations(s: &mut Session) -> Result<(), CmdError> { let st = style(&s.bib_style); - let order = bib_order(&s.doc, st); + let order = bib_order(&s.doc, st, super::default_lang(s)); for path in s.doc.para_paths(StoryRef::Body) { let Some(p) = s.doc.para(StoryRef::Body, &path) else { continue }; let updates: Vec<(usize, String)> = p @@ -763,7 +766,7 @@ fn page_of(s: &mut Session, pos: &Pos) -> u32 { fn entries_for(s: &mut Session, kind: &str) -> Vec<(String, bool)> { if kind.starts_with("BIBLIOGRAPHY") { let st = style(&s.bib_style); - return bib_order(&s.doc, st) + return bib_order(&s.doc, st, super::default_lang(s)) .into_iter() .enumerate() .filter_map(|(k, i)| s.doc.sources.get(i).map(|x| (entry(x, st.name, k.saturating_add(1)), !st.numbered()))) @@ -824,21 +827,22 @@ fn entries_for(s: &mut Session, kind: &str) -> Vec<(String, bool)> { return with_pages.into_iter().map(|(t, p)| (format!("{t}\t{p}"), false)).collect(); } // Index / TOA: group by entry, sort, list pages. - let mut grouped: std::collections::BTreeMap> = Default::default(); + let lang = super::default_lang(s); + let mut grouped: std::collections::BTreeMap<(SortKey, String), Vec> = Default::default(); for (e, p) in with_pages { - let pages = grouped.entry(e).or_default(); + let pages = grouped.entry((sort_key(&e, lang.as_deref()), e)).or_default(); if !pages.contains(&p) { pages.push(p); } } let mut out = Vec::new(); let mut letter = None; - for (e, pages) in grouped { - let first = e.chars().next().map(|c| c.to_ascii_uppercase()); + for ((_, e), pages) in grouped { + let first = heading_letter(&e, lang.as_deref()); if kind.starts_with("INDEX") && first != letter { - letter = first; + letter = first.clone(); if let Some(l) = first { - out.push((l.to_string(), false)); + out.push((l, false)); } } let list = pages.iter().map(|p| p.to_string()).collect::>().join(", "); diff --git a/crates/engine/src/cmd/column.rs b/crates/engine/src/cmd/column.rs index f1552eee0..1db3cd1b4 100644 --- a/crates/engine/src/cmd/column.rs +++ b/crates/engine/src/cmd/column.rs @@ -3,6 +3,10 @@ //! Backspace remove just the block, character formatting applies to each row's piece, paragraph //! formatting to the paragraphs it touches. Typing or pasting first deletes the block; any other //! edit collapses it to its top-left corner. Any ordinary caret move or selection drops it. +//! Whole table cells that aren't one run of the text (part of a column, `CellRange`) are a block +//! of their paragraphs: edits and formatting go cell by cell; table commands take the cells. +//! Extending the selection from one cell into another (Shift+arrows, Shift+click, a drag) makes +//! such cells, as in Word (`table::extend_selection`). use serde_json::{Value, json}; use wordcraft_doc::edit::Fragment; @@ -94,7 +98,7 @@ fn make(s: &mut Session, a: Pos, b: Pos, xa: f32, xb: f32) { let l = s.layout(); let segments = l.column_segments(&s.doc, &a, &b, left, right, s.page_hint, MAX_LINES); s.sel = Selection { anchor: a, focus: b }; - s.column = Some(ColumnBlock { sel: s.sel.clone(), anchor_x: xa, focus_x: xb, segments }); + s.column = Some(ColumnBlock { sel: s.sel.clone(), anchor_x: xa, focus_x: xb, segments, cells: None }); s.pending = None; } @@ -116,6 +120,11 @@ pub(crate) fn dispatch(s: &mut Session, id: &str, mutates: bool, run: fn(&mut Se s.column = None; s.column_mode = false; } + // Placing the caret ends a block even where its corners are (a selected empty cell's are + // the same place). + if matches!(id, "caret.set" | "select.collapse") { + s.column = None; + } if s.column_mode && id.starts_with("caret.") && id != "caret.set" { let mut params = v.clone(); match params.as_object_mut() { @@ -132,8 +141,25 @@ pub(crate) fn dispatch(s: &mut Session, id: &str, mutates: bool, run: fn(&mut Se if matches!(id, "caret.set" | "select.collapse") || (mutates && !id.starts_with("format.") && !id.starts_with("para.")) { s.column_mode = false; } + // Extending the selection (Shift+arrows, Shift+click, a drag) goes on from where it really + // runs, then takes whole cells or rows the way Word does around tables. + if id.starts_with("caret.") && p::bool(v, "extend") == Some(true) { + let shown = s.sel.clone(); + s.sel = s.sel_ends().clone(); + if let Err(e) = run(s, v) { + s.sel = shown; + return Err(e); + } + let to = s.sel.focus.clone(); + super::table::extend_selection(s, to); + return super::sel_result(s); + } let Some(segs) = s.column_segments().map(<[(Pos, Pos)]>::to_vec) else { return run(s, v) }; + let cells = s.cell_range().is_some(); match id { + // Whole cells: table commands take the rectangle between the corners, and Tab moves on + // from it without clearing it, as in Word. + _ if cells && (id.starts_with("table.") || matches!(id, "text.tab" | "text.backTab")) => run(s, v), "edit.copy" => copy(s, &segs), "edit.cut" => { let r = copy(s, &segs)?; @@ -146,7 +172,8 @@ pub(crate) fn dispatch(s: &mut Session, id: &str, mutates: bool, run: fn(&mut Se } "format.changeCase" | "format.clear" | "format.resetChar" => per_row(s, run, v, &segs), _ if id.starts_with("format.") || id.starts_with("para.") || id == "select.column" => run(s, v), - _ if mutates && (id.starts_with("text.") || id.starts_with("edit.paste")) => { + // Pasting formatting (the format painter) leaves the text alone. + _ if mutates && (id.starts_with("text.") || (id.starts_with("edit.paste") && id != "edit.pasteFormat")) => { delete(s, &segs)?; run(s, v) } diff --git a/crates/engine/src/cmd/edit.rs b/crates/engine/src/cmd/edit.rs index e7243a33f..d8b78a572 100644 --- a/crates/engine/src/cmd/edit.rs +++ b/crates/engine/src/cmd/edit.rs @@ -7,6 +7,26 @@ use wordcraft_doc::{Pos, StoryRef}; use super::{delete_selection, find_pattern, pos_json, sel_result}; use crate::{CmdError, CmdResult, CommandSpec, Selection, Session, p}; +/// Undo several steps at once: the Undo button's list (#701). `count` is validated here and +/// clamped to the undo depth, so hostile numbers never reach the stack. +fn undo_to(s: &mut Session, v: &Value) -> CmdResult { + let n = p::u64(v, "count").ok_or_else(|| CmdError::Params("`count` (whole number, 1 or more) is required".into()))?; + if n == 0 { + return Err(CmdError::Params("`count` must be 1 or more".into())); + } + let n = usize::try_from(n).unwrap_or(usize::MAX).min(s.undo_depth()); + if s.step_lifts_protection(false) { + return Err(CmdError::Disabled("edit.undoTo: use Stop Protection to lift the document's protection".into())); + } + let undone = s.undo_steps(n); + let mut r = sel_result(s)?; + if let Some(m) = r.as_object_mut() { + m.insert("undone".into(), json!(undone)); + m.insert("depth".into(), json!(s.undo_depth())); + } + Ok(r) +} + fn has_sel(s: &Session) -> Option<&'static str> { if s.sel.is_collapsed() { Some("nothing is selected") } else { None } } @@ -22,6 +42,14 @@ pub fn specs() -> Vec { }) .key("Mod+Z") .pure(), + CommandSpec::new("edit.undoTo", "Undo Actions", "Quick Access Toolbar", undo_to) + .params(r#"{"count": n (1 = the newest step, like edit.undo; n undoes the newest n steps at once; at most the undo depth)}"#) + .when(|s| if s.can_undo() { None } else { Some("there is nothing to undo") }) + .pure(), + CommandSpec::new("edit.undoHistory", "Undo History", "Quick Access Toolbar", |s, _| { + Ok(json!({"undo": s.undo_history(), "depth": s.undo_depth(), "redo": s.redo_label()})) + }) + .pure(), CommandSpec::new("edit.redo", "Redo", "Quick Access Toolbar", |s, _| { if s.step_lifts_protection(true) { return Err(CmdError::Disabled("edit.redo: use Stop Protection to lift the document's protection".into())); @@ -33,11 +61,12 @@ pub fn specs() -> Vec { .pure(), CommandSpec::new("edit.cut", "Cut", "Home › Clipboard", cut).key("Mod+X").when(has_sel), CommandSpec::new("edit.copy", "Copy", "Home › Clipboard", copy).key("Mod+C").pure(), - CommandSpec::new("edit.paste", "Paste", "Home › Clipboard", paste).key("Mod+V").params(r#"{"text"?: string}"#), + CommandSpec::new("edit.paste", "Paste", "Home › Clipboard", paste).key("Mod+V").params(r#"{"text"?: string, "html"?: string, "rtf"?: string}"#), CommandSpec::new("edit.pasteText", "Paste: Keep Text Only", "Home › Clipboard › Paste", paste_text) .params(r#"{"text"?: string}"#) .key("Mod+Shift+Alt+V"), - CommandSpec::new("edit.pasteMerge", "Paste: Merge Formatting", "Home › Clipboard › Paste", paste_text).params(r#"{"text"?: string}"#), + CommandSpec::new("edit.pasteMerge", "Paste: Merge Formatting", "Home › Clipboard › Paste", paste_merge) + .params(r#"{"text"?: string, "html"?: string, "rtf"?: string}"#), CommandSpec::new("edit.clipboardPane", "Clipboard Pane", "Home › Clipboard", |s, v| { let on = p::bool(v, "value").unwrap_or(!s.view.clipboard_pane); s.view.clipboard_pane = on; @@ -133,14 +162,18 @@ fn cut(s: &mut Session, v: &Value) -> CmdResult { Ok(r) } +/// Paste keeps the source formatting: our own copy while the system clipboard still holds it, +/// else another program's HTML or RTF (#687), else the system clipboard's text. fn paste(s: &mut Session, v: &Value) -> CmdResult { - // An explicit text param (from the system clipboard) wins unless it matches our own copy. - let frag = match (p::str(v, "text"), &s.clipboard) { - (Some(t), Some(f)) if f.plain_text() == t.replace("\r\n", "\n") => f.clone(), - (Some(t), _) => Fragment::from_text(&t.replace("\r\n", "\n")), - (None, Some(f)) => f.clone(), - (None, None) => return Err(CmdError::Failed("the clipboard is empty".into())), - }; + let frag = super::paste::paste_fragment(s, v)?; + super::paste::insert(s, frag)?; + sel_result(s) +} + +/// Merge Formatting: the pasted text takes the destination's formatting, keeping its emphasis. +fn paste_merge(s: &mut Session, v: &Value) -> CmdResult { + let mut frag = super::paste::paste_fragment(s, v)?; + super::paste::merge_formatting(s, &mut frag); super::paste::insert(s, frag)?; sel_result(s) } @@ -284,6 +317,9 @@ fn paste_text(s: &mut Session, v: &Value) -> CmdResult { Some(t) => t.to_string(), None => s.clipboard.as_ref().map(Fragment::plain_text).unwrap_or_default(), }; + if text.is_empty() { + return Err(CmdError::Failed("the clipboard is empty".into())); + } super::paste_text(s, &text.replace("\r\n", "\r").replace('\n', "\r"))?; sel_result(s) } diff --git a/crates/engine/src/cmd/file.rs b/crates/engine/src/cmd/file.rs index e1376555b..1c451fbf7 100644 --- a/crates/engine/src/cmd/file.rs +++ b/crates/engine/src/cmd/file.rs @@ -119,7 +119,15 @@ fn new(s: &mut Session, v: &Value) -> CmdResult { "letter" => crate::sample::letter(), "resume" => crate::sample::resume(), "report" => crate::sample::report(), - _ => Document::new(), + // The Normal template: the region's paper (#838), written in the editing language (the + // templates' text is English). + _ => { + let mut d = s.blank_document(); + if let Some(lang) = &s.editing_lang { + d.styles.default_chr.lang = Some(lang.clone()); + } + d + } }; s.set_document(doc); s.path = None; @@ -132,12 +140,19 @@ fn open(s: &mut Session, v: &Value) -> CmdResult { return sel_result(s); }; let password = p::str(v, "password"); - let (doc, encrypted) = if let Some(data) = p::str(v, "data") { + let (mut doc, encrypted) = if let Some(data) = p::str(v, "data") { let bytes = super::insert::base64_decode(data).ok_or_else(|| CmdError::Params("bad base64".into()))?; crate::io::open_bytes_with(path, &bytes, password).map_err(CmdError::Failed)? } else { crate::io::open_path_with(std::path::Path::new(path), password).map_err(CmdError::Failed)? }; + // Plain text and Markdown carry no language: they are in the editing language. + let ext = std::path::Path::new(path).extension().and_then(|e| e.to_str()).unwrap_or("").to_ascii_lowercase(); + if let Some(lang) = &s.editing_lang + && matches!(ext.as_str(), "txt" | "text" | "md" | "markdown") + { + doc.styles.default_chr.lang = Some(lang.clone()); + } s.set_document(doc); s.path = Some(path.into()); // Like Word, a document opened with a password is saved with it again. @@ -219,8 +234,7 @@ fn export_png(s: &mut Session, v: &Value) -> CmdResult { let pg = l.pages.get(page).ok_or_else(|| CmdError::Params(format!("no page {}", page + 1)))?; let img = wordcraft_render::render_page(&s.doc, pg, scale, &Default::default()); let png = img.to_png(); - #[cfg(not(target_arch = "wasm32"))] - std::fs::write(path, &png).map_err(|e| CmdError::Failed(format!("{path}: {e}")))?; + crate::io::write_file(path, &png).map_err(CmdError::Failed)?; Ok(json!({"path": path, "width": img.width, "height": img.height, "bytes": png.len()})) } diff --git a/crates/engine/src/cmd/font.rs b/crates/engine/src/cmd/font.rs index 703a86858..bcdee205d 100644 --- a/crates/engine/src/cmd/font.rs +++ b/crates/engine/src/cmd/font.rs @@ -33,14 +33,8 @@ fn font_json(f: &UserFont) -> Value { /// Read a font file named by the user, refusing one larger than the cap before reading it. #[cfg(not(target_arch = "wasm32"))] fn read_font_file(path: &str) -> Result, CmdError> { - let meta = std::fs::metadata(path).map_err(|e| CmdError::Failed(format!("{path}: {e}")))?; - if !meta.is_file() { - return Err(CmdError::Failed(format!("{path}: not a file"))); - } - if meta.len() > wordcraft_fonts::MAX_FONT_FILE_BYTES as u64 { - return Err(CmdError::Failed(format!("{path}: the font file is larger than {} MB", wordcraft_fonts::MAX_FONT_FILE_BYTES >> 20))); - } - std::fs::read(path).map_err(|e| CmdError::Failed(format!("{path}: {e}"))) + // Through `io`, which keeps automation inside its read root (#769). + crate::io::read_file_capped(path, wordcraft_fonts::MAX_FONT_FILE_BYTES as u64, "font file").map_err(CmdError::Failed) } fn add(s: &mut Session, v: &Value) -> CmdResult { diff --git a/crates/engine/src/cmd/format.rs b/crates/engine/src/cmd/format.rs index ef9a353e5..85ca49e13 100644 --- a/crates/engine/src/cmd/format.rs +++ b/crates/engine/src/cmd/format.rs @@ -15,15 +15,14 @@ pub const SIZES: [f32; 17] = [8.0, 9.0, 10.0, 10.5, 11.0, 12.0, 14.0, 16.0, 18.0 pub fn specs() -> Vec { vec![ CommandSpec::new("format.bold", "Bold", "Home › Font", |s, v| { - // Like Word's button, for complex-script (Persian, Arabic) text too: an unset `bold_cs` - // follows `bold`. + // The button formats both plain and complex-script text explicitly. toggle( s, v, |r| r.bold, |c, on| { c.bold = Some(on); - c.bold_cs = None; + c.bold_cs = Some(on); }, ) }) @@ -36,7 +35,7 @@ pub fn specs() -> Vec { |r| r.italic, |c, on| { c.italic = Some(on); - c.italic_cs = None; + c.italic_cs = Some(on); }, ) }) @@ -168,13 +167,15 @@ fn target(s: &Session) -> Option<(Pos, Pos)> { None } -/// The non-empty rows of a column selection, if one is active. +/// The non-empty rows of a column selection, if one is active. Of whole cells, every paragraph: +/// an empty one's mark takes the formatting. fn column_rows(s: &Session) -> Option> { let segs = s.column_segments()?; + let cells = s.cell_range().is_some(); let rows: Vec<(Pos, Pos)> = segs .iter() .map(|(a, b)| (s.doc.clamp(a), s.doc.clamp(b))) - .filter(|(a, b)| a != b && a.path == b.path && a.story == b.story) + .filter(|(a, b)| (a != b || cells) && a.path == b.path && a.story == b.story) .map(|(a, b)| if a <= b { (a, b) } else { (b, a) }) .collect(); // An empty block (column mode just started) formats like a caret. @@ -187,6 +188,10 @@ pub fn apply(s: &mut Session, f: &dyn Fn(&mut CharProps)) -> CmdResult { if let Some(rows) = column_rows(s) { for (a, b) in rows { s.doc.format_range(&a, &b, f)?; + // An empty cell's paragraph only had its mark formatted: lay it out again. + if a == b { + s.doc.para_mut(a.story, &a.path)?.touch(); + } } return sel_result(s); } @@ -217,6 +222,10 @@ fn resolved(s: &Session) -> Vec { if let Some(rows) = column_rows(s) { for (a, b) in rows { let Some(p) = s.doc.para_at(&a) else { continue }; + // An empty cell: its mark's formatting. + if a == b { + out.push(s.doc.styles.resolve_char(p.props.style.as_deref(), &p.mark)); + } for (r, c) in p.run_ranges() { if r.end > a.off && r.start < b.off { out.push(s.doc.styles.resolve_char(p.props.style.as_deref(), c)); @@ -319,7 +328,7 @@ fn size(s: &mut Session, v: &Value) -> CmdResult { let sz = (sz * 2.0).round() / 2.0; apply(s, &|c| { c.size = Some(sz); - c.size_cs = None; + c.size_cs = Some(sz); }) } @@ -338,7 +347,7 @@ fn step_size(s: &mut Session, dir: i32) -> CmdResult { // Each run steps from its own size when the selection mixes sizes. apply(s, &|c| { c.size = Some(next); - c.size_cs = None; + c.size_cs = Some(next); }) } @@ -346,7 +355,7 @@ fn nudge_size(s: &mut Session, d: f32) -> CmdResult { let next = (current_size(s) + d).clamp(1.0, 1638.0); apply(s, &|c| { c.size = Some(next); - c.size_cs = None; + c.size_cs = Some(next); }) } @@ -568,7 +577,12 @@ fn char_style(s: &mut Session, v: &Value) -> CmdResult { } fn set(s: &mut Session, v: &Value) -> CmdResult { - let props: CharProps = serde_json::from_value(v.get("props").cloned().unwrap_or(Value::Null)).map_err(|e| CmdError::Params(e.to_string()))?; + let mut props: CharProps = serde_json::from_value(v.get("props").cloned().unwrap_or(Value::Null)).map_err(|e| CmdError::Params(e.to_string()))?; + // The font dialog and callers supplying one value format both scripts. Explicit + // complex-script values still take precedence; importing a DOCX never uses this command. + props.size_cs = props.size_cs.or(props.size); + props.bold_cs = props.bold_cs.or(props.bold); + props.italic_cs = props.italic_cs.or(props.italic); apply(s, &|c| c.overlay(&props)) } @@ -602,6 +616,9 @@ pub fn state(s: &Session) -> Value { "highlight": same(&|r| json!(r.highlight.map(|c| c.hex()))), "style": style, "styleName": style_name, + // The proofing language (a BCP 47 tag; null when mixed or unset) and "Do not check". + "lang": same(&|r| json!(r.lang)), + "noProof": all.iter().all(|r| r.no_proof), }) } diff --git a/crates/engine/src/cmd/index.rs b/crates/engine/src/cmd/index.rs index 3b1cc4c25..a94b300b5 100644 --- a/crates/engine/src/cmd/index.rs +++ b/crates/engine/src/cmd/index.rs @@ -4,6 +4,7 @@ use std::collections::BTreeMap; use serde_json::{Value, json}; +use wordcraft_doc::collate::{SortKey, heading_letter, sort_key}; use wordcraft_doc::para::InlineObject; use wordcraft_doc::props::{CharProps, TabAlign, TabLeader, TabStop}; use wordcraft_doc::section::SectionStart; @@ -167,14 +168,15 @@ struct PageRef { #[derive(Default)] struct Node { - /// The entry's levels as written (the key is their lower-case form). + /// The entry's levels as written (the key is their sort keys). levels: Vec, pages: Vec, see: Vec, } -/// Every XE entry in the body with its page (or page range), grouped by entry. -fn collect(s: &mut Session) -> BTreeMap, Node> { +/// Every XE entry in the body with its page (or page range), grouped by entry and sorted in the +/// order of language `lang` (Polish files `Łódź` between `L` and `M`). +fn collect(s: &mut Session, lang: Option<&str>) -> BTreeMap, Node> { let mut marks: Vec<(FieldCode, Pos)> = Vec::new(); for path in s.doc.para_paths(StoryRef::Body) { let Some(p) = s.doc.para(StoryRef::Body, &path) else { continue }; @@ -197,7 +199,7 @@ fn collect(s: &mut Session) -> BTreeMap, Node> { }) }; let starts = s.doc.bookmarks(); - let mut tree: BTreeMap, Node> = BTreeMap::new(); + let mut tree: BTreeMap, Node> = BTreeMap::new(); for (f, pos) in marks { let text = f.args.first().cloned().unwrap_or_default(); let levels: Vec = text.split(':').map(|x| x.trim().to_string()).filter(|x| !x.is_empty()).take(9).collect(); @@ -207,10 +209,10 @@ fn collect(s: &mut Session) -> BTreeMap, Node> { // Every level above gets a line of its own, even without pages. for k in 1..levels.len() { let head: Vec = levels.iter().take(k).cloned().collect(); - tree.entry(head.iter().map(|x| x.to_lowercase()).collect()).or_insert_with(|| Node { levels: head, ..Default::default() }); + tree.entry(head.iter().map(|x| sort_key(x, lang)).collect()).or_insert_with(|| Node { levels: head, ..Default::default() }); } let node = - tree.entry(levels.iter().map(|x| x.to_lowercase()).collect()).or_insert_with(|| Node { levels: levels.clone(), ..Default::default() }); + tree.entry(levels.iter().map(|x| sort_key(x, lang)).collect()).or_insert_with(|| Node { levels: levels.clone(), ..Default::default() }); if let Some(see) = f.arg('t') { if !node.see.iter().any(|x| x == see) { node.see.push(see.to_string()); @@ -278,7 +280,9 @@ pub fn rebuild(s: &mut Session, i: usize, leader: Option) -> Result<( let sep = f.arg('e').map(str::to_string).unwrap_or_else(|| ", ".into()); let run_in = f.has('r'); let headings = f.switches.iter().find(|(c, _)| *c == 'h').map(|(_, a)| a.clone().unwrap_or_default()); - let tree = collect(s); + // Entries sort in the language of the index's paragraph (else the document's). + let lang = s.doc.body.get(i).and_then(|b| b.as_para()).and_then(|p| super::para_lang(&s.doc, p)).or_else(|| super::default_lang(s)); + let tree = collect(s, lang.as_deref()); let mut lines: Vec<(usize, Line)> = Vec::new(); let mut letter: Option = None; let mut keys = tree.keys().peekable(); @@ -286,7 +290,7 @@ pub fn rebuild(s: &mut Session, i: usize, leader: Option) -> Result<( let Some(node) = tree.get(key) else { continue }; let Some(main) = node.levels.first() else { continue }; if let Some(h) = &headings { - let first = main.chars().next().map(|c| c.to_uppercase().to_string()); + let first = heading_letter(main, lang.as_deref()); if first != letter { if letter.is_some() || !h.is_empty() { lines.push((0, vec![(if h.is_empty() { String::new() } else { first.clone().unwrap_or_default() }, CharProps::default())])); diff --git a/crates/engine/src/cmd/insert.rs b/crates/engine/src/cmd/insert.rs index 9d84e5f81..c541c3661 100644 --- a/crates/engine/src/cmd/insert.rs +++ b/crates/engine/src/cmd/insert.rs @@ -183,19 +183,11 @@ pub fn base64_encode(b: &[u8]) -> String { out } -fn picture(s: &mut Session, v: &Value) -> CmdResult { +pub(crate) fn picture(s: &mut Session, v: &Value) -> CmdResult { let bytes = if let Some(d) = p::str(v, "data") { base64_decode(d).ok_or_else(|| CmdError::Params("bad base64 data".into()))? } else if let Some(path) = p::str(v, "path") { - #[cfg(not(target_arch = "wasm32"))] - { - std::fs::read(path).map_err(|e| CmdError::Failed(format!("{path}: {e}")))? - } - #[cfg(target_arch = "wasm32")] - { - let _ = path; - return Err(CmdError::Failed("reading files by path isn't available on the web".into())); - } + crate::io::read_file(path).map_err(CmdError::Failed)? } else { s.ui_requests.push(json!({"open": "insertPicture"})); return sel_result(s); diff --git a/crates/engine/src/cmd/mailings.rs b/crates/engine/src/cmd/mailings.rs index ba6611619..d8f7891c6 100644 --- a/crates/engine/src/cmd/mailings.rs +++ b/crates/engine/src/cmd/mailings.rs @@ -320,14 +320,8 @@ fn extension(path: &str) -> Option { std::path::Path::new(path).extension().map(|e| e.to_string_lossy().to_ascii_lowercase()) } -#[cfg(not(target_arch = "wasm32"))] fn read_file(path: &str) -> Result, CmdError> { - std::fs::read(path).map_err(|e| CmdError::Failed(format!("{path}: {e}"))) -} - -#[cfg(target_arch = "wasm32")] -fn read_file(_path: &str) -> Result, CmdError> { - Err(CmdError::Failed("paths aren't available on the web; pass `data`".into())) + crate::io::read_file(path).map_err(CmdError::Failed) } /// Field names and records from a table's rows: with `headers`, the first row names the fields diff --git a/crates/engine/src/cmd/mod.rs b/crates/engine/src/cmd/mod.rs index 5e8161a83..33dffa579 100644 --- a/crates/engine/src/cmd/mod.rs +++ b/crates/engine/src/cmd/mod.rs @@ -39,6 +39,8 @@ pub mod smart_art; pub mod speech; pub mod table; pub mod table_draw; +pub mod table_mouse; +pub mod table_resize; pub mod table_style; pub mod templates; pub mod text; @@ -74,6 +76,8 @@ pub fn registry() -> Registry { v.extend(table::specs()); v.extend(table_style::specs()); v.extend(table_draw::specs()); + v.extend(table_resize::specs()); + v.extend(table_mouse::specs()); v.extend(review::specs()); v.extend(file::specs()); v.extend(options::specs()); @@ -128,6 +132,19 @@ pub fn sel_result(s: &Session) -> CmdResult { Ok(json!({"anchor": pos_json(&s.sel.anchor), "focus": pos_json(&s.sel.focus)})) } +/// The paragraphs the selection covers: those between its ends (not the next one when it ends at +/// that one's start, #418) or, for whole table cells that aren't one run of the text (part of a +/// column), just the cells' own. +pub(crate) fn selected_paths(s: &Session) -> Vec { + if s.cell_range().is_some() + && let Some(segs) = s.column_segments() + { + return segs.iter().map(|(a, _)| a.path.clone()).collect(); + } + let (a, b) = s.sel.ordered(); + s.doc.para_paths_in(&a, &b) +} + /// Delete the selection (respecting track changes); returns the collapsed position. pub fn delete_selection(s: &mut Session) -> Result { let (a, b) = s.sel.ordered(); @@ -273,7 +290,18 @@ fn insert_text(s: &mut Session, text: &str, enter: bool) -> Result<(), CmdError> first = false; let part = part.replace('\u{000B}', "\n"); if !part.is_empty() { + let starts_para = direction_unset(s, &at); at = s.doc.insert_text(&at, &part, &props)?; + // Typing Arabic, Hebrew… as the first letters of a paragraph whose direction nothing + // sets makes it a right-to-left paragraph, so the caret and text start at the right + // margin. Paragraphs that already have letters, or a direction, keep theirs. + if starts_para + && let Ok(p) = s.doc.para_mut(at.story, &at.path) + && wordcraft_doc::bidi::first_strong_is_rtl(&p.text) + { + p.props.bidi = Some(true); + p.touch(); + } } } s.sel = crate::Selection::caret(at); @@ -281,6 +309,17 @@ fn insert_text(s: &mut Session, text: &str, enter: bool) -> Result<(), CmdError> Ok(()) } +/// Is the paragraph at `at` still without letters (empty, or only spaces, digits and +/// punctuation), with no direction set on it, its style chain or the document defaults? +fn direction_unset(s: &Session, at: &Pos) -> bool { + let Some(p) = s.doc.para_at(at) else { return false }; + if p.props.bidi.is_some() || p.text.chars().any(char::is_alphabetic) || s.doc.styles.default_para.bidi.is_some() { + return false; + } + let style = p.props.style.as_deref().unwrap_or("Normal"); + !s.doc.styles.chain(style).iter().any(|st| st.para.bidi.is_some()) +} + /// Take a paragraph out of its list like Word does when Enter or Backspace ends a list: a /// "List Paragraph" goes back to Normal; any other style keeps itself with numbering switched off. pub fn leave_list(para: &mut wordcraft_doc::Paragraph) { @@ -396,6 +435,24 @@ pub fn story_param(s: &Session, v: &Value) -> StoryRef { v.get("story").and_then(|x| serde_json::from_value(x.clone()).ok()).unwrap_or(s.sel.focus.story) } +/// The language (`w:lang`) of a paragraph's text: that of its first run with text (else of the +/// paragraph mark), through styles and the document defaults. Sorting and number parsing follow it. +pub fn para_lang(doc: &wordcraft_doc::Document, p: &wordcraft_doc::Paragraph) -> Option { + let run = p.runs.iter().find(|r| r.len > 0).map(|r| &r.props).unwrap_or(&p.mark); + doc.styles.resolve_char(p.props.style.as_deref(), run).lang +} + +/// The document's default text language (`w:docDefaults`, the Normal style). +pub fn doc_lang(doc: &wordcraft_doc::Document) -> Option { + doc.styles.resolve_char(None, &wordcraft_doc::CharProps::default()).lang +} + +/// The language of text that names none: the document's default, else the editing language (a +/// Polish interface writes Polish), else none (English). As AutoCorrect decides it. +pub fn default_lang(s: &Session) -> Option { + doc_lang(&s.doc).or_else(|| s.editing_lang.clone()) +} + #[cfg(test)] mod iso_tests { use super::iso_from_unix_secs; diff --git a/crates/engine/src/cmd/objects.rs b/crates/engine/src/cmd/objects.rs index fc658de12..11493302c 100644 --- a/crates/engine/src/cmd/objects.rs +++ b/crates/engine/src/cmd/objects.rs @@ -203,15 +203,7 @@ pub fn specs() -> Vec { let bytes = if let Some(d) = p::str(v, "data") { super::insert::base64_decode(d).ok_or_else(|| CmdError::Params("bad base64".into()))? } else if let Some(path) = p::str(v, "path") { - #[cfg(not(target_arch = "wasm32"))] - { - std::fs::read(path).map_err(|e| CmdError::Failed(format!("{path}: {e}")))? - } - #[cfg(target_arch = "wasm32")] - { - let _ = path; - return Err(CmdError::Failed("paths aren't available on the web".into())); - } + crate::io::read_file(path).map_err(CmdError::Failed)? } else { return Err(CmdError::Params("`path` or `data` required".into())); }; diff --git a/crates/engine/src/cmd/options.rs b/crates/engine/src/cmd/options.rs index c7a8a49c5..d9a96252e 100644 --- a/crates/engine/src/cmd/options.rs +++ b/crates/engine/src/cmd/options.rs @@ -20,7 +20,7 @@ pub fn specs() -> Vec { vec![ CommandSpec::new("file.options", "Options", "File", options) .params( - r#"{"pane"?: "general|display|proofing|save|language|accessibility|advanced|agents", "author"?: string, "initials"?: string, "units"?: "inches|centimeters|millimeters|points|picas", "typingReplacesSelection"?: bool, "insertKeyOvertype"?: bool, "overtype"?: bool, "marks"?: {"tabs"?: bool, "spaces"?: bool, "paragraphs"?: bool, "hidden"?: bool}, "checkSpelling"?: bool, "markGrammar"?: bool, "ignoreUppercase"?: bool, "ignoreNumbers"?: bool, "ignoreInternet"?: bool, "addWords"?: [string], "removeWords"?: [string]} — with any setting it changes them and opens nothing; without, it opens File › Options (at `pane`). Returns the settings."#, + r#"{"pane"?: "general|display|proofing|save|language|accessibility|advanced|agents", "author"?: string, "initials"?: string, "units"?: "inches|centimeters|millimeters|points|picas|auto" (auto: the region's), "region"?: string (a locale tag like pl-PL, saved: the paper File › New starts on, the default units and the decimal separator; auto: the system's), "typingReplacesSelection"?: bool, "insertKeyOvertype"?: bool, "overtype"?: bool, "marks"?: {"tabs"?: bool, "spaces"?: bool, "paragraphs"?: bool, "hidden"?: bool}, "checkSpelling"?: bool, "markGrammar"?: bool, "ignoreUppercase"?: bool, "ignoreNumbers"?: bool, "ignoreInternet"?: bool, "addWords"?: [string], "removeWords"?: [string]} — with any setting it changes them and opens nothing; without, it opens File › Options (at `pane`). Returns the settings."#, ) .pure(), CommandSpec::new("text.overtype", "Overtype", "File › Options › Advanced", |s, v| { @@ -95,10 +95,23 @@ fn options(s: &mut Session, v: &Value) -> CmdResult { }; let units = match v.get("units") { None | Some(Value::Null) => None, - Some(u) => Some( + // `auto` forgets the choice: lengths follow the region again (#838). + Some(Value::String(a)) if a == "auto" => Some(None), + Some(u) => Some(Some( serde_json::from_value::(u.clone()) - .map_err(|_| CmdError::Params("`units` is one of inches, centimeters, millimeters, points, picas".into()))?, - ), + .map_err(|_| CmdError::Params("`units` is one of inches, centimeters, millimeters, points, picas, auto".into()))?, + )), + }; + let region = match v.get("region") { + None | Some(Value::Null) => None, + // `auto` (or an empty tag) forgets the choice: the system's region again. + Some(Value::String(t)) if t.trim().is_empty() || t.trim().eq_ignore_ascii_case("auto") => Some(None), + // Kept as the tag (normalised, short), so it is saved and read back as chosen. + Some(Value::String(t)) => { + let r = wordcraft_geom::Region::from_tag(t); + Some(Some(if r.territory.is_empty() { r.language } else { format!("{}-{}", r.language, r.territory) })) + } + Some(_) => return Err(CmdError::Params("`region` must be a locale tag like pl-PL, or auto".into())), }; let mut flag_values = Vec::new(); for (k, f) in flags() { @@ -125,6 +138,7 @@ fn options(s: &mut Session, v: &Value) -> CmdResult { let setting = author.is_some() || initials.is_some() || units.is_some() + || region.is_some() || !flag_values.is_empty() || !mark_values.is_empty() || check_spelling.is_some() @@ -139,6 +153,9 @@ fn options(s: &mut Session, v: &Value) -> CmdResult { if let Some(u) = units { s.prefs.units = u; } + if let Some(r) = region { + s.prefs.region = r; + } for (f, b) in flag_values { *f(&mut s.prefs) = b; } @@ -154,6 +171,9 @@ fn options(s: &mut Session, v: &Value) -> CmdResult { for w in &remove { wordcraft_proof::remove_word(w); } + if !add.is_empty() || !remove.is_empty() { + s.persist_user_dictionary(&remove); + } if setting { // Marks, hidden text and proofing change what is laid out. s.relayout(); @@ -170,10 +190,14 @@ fn options(s: &mut Session, v: &Value) -> CmdResult { /// The settings `file.options` reads and sets. pub fn current(s: &Session) -> Value { let pr = &s.prefs; + let region = s.region(); json!({ "author": s.author, "initials": s.user_initials(), - "units": pr.units, + "units": s.units(), + "unitsFromRegion": pr.units.is_none(), + "region": {"language": region.language, "territory": region.territory, "paper": region.paper().0, "decimal": region.decimal().to_string()}, + "regionFromSystem": pr.region.is_none(), "typingReplacesSelection": pr.typing_replaces_selection, "insertKeyOvertype": pr.insert_key_overtype, "overtype": pr.overtype, diff --git a/crates/engine/src/cmd/para.rs b/crates/engine/src/cmd/para.rs index 1bb48312f..cdbe79027 100644 --- a/crates/engine/src/cmd/para.rs +++ b/crates/engine/src/cmd/para.rs @@ -1,6 +1,7 @@ //! Paragraph formatting, lists and styles (Home › Paragraph, Home › Styles). use serde_json::{Value, json}; +use wordcraft_doc::collate::{SortKey, sort_key}; use wordcraft_doc::numbering::ListKind; use wordcraft_doc::props::{Align, Border, BorderStyle, Borders, LineSpacing, NumRef, ParaProps, Rgb, TabAlign, TabLeader, TabStop}; use wordcraft_doc::styles::{Style, StyleKind}; @@ -78,6 +79,11 @@ pub fn specs() -> Vec { CommandSpec::new("para.oneAndHalf", "1.5 Line Spacing", "Home › Paragraph › Line Spacing", |s, _| ls(s, LineSpacing::Multiple(1.5))).key("Mod+5"), CommandSpec::new("para.spacing", "Paragraph Spacing", "Layout › Paragraph", |s, v| { let (b, a) = (p::f32(v, "before"), p::f32(v, "after")); + let (b_auto, a_auto) = (p::bool(v, "beforeAuto"), p::bool(v, "afterAuto")); + // A number replaces automatic spacing, as typing one in Word's Paragraph dialog does. + let rp = cur(s); + let b_auto = b_auto.or((b.is_some() && rp.before_auto).then_some(false)); + let a_auto = a_auto.or((a.is_some() && rp.after_auto).then_some(false)); fmt(s, &|p| { if let Some(x) = b { p.space_before = Some(x.clamp(0.0, 1584.0)); @@ -85,17 +91,35 @@ pub fn specs() -> Vec { if let Some(x) = a { p.space_after = Some(x.clamp(0.0, 1584.0)); } + if let Some(on) = b_auto { + p.space_before_auto = Some(on); + } + if let Some(on) = a_auto { + p.space_after_auto = Some(on); + } }) }) - .params(r#"{"before"?: pt, "after"?: pt}"#), + .params(r#"{"before"?: pt, "after"?: pt, "beforeAuto"?: bool, "afterAuto"?: bool (automatic, HTML-style spacing: 14 pt)}"#), CommandSpec::new("para.addSpaceBefore", "Add Space Before Paragraph", "Home › Paragraph › Line Spacing", |s, _| { - let has = cur(s).space_before > 0.0; - fmt(s, &|p| p.space_before = Some(if has { 0.0 } else { 12.0 })) + let rp = cur(s); + let has = rp.space_before > 0.0; + fmt(s, &|p| { + p.space_before = Some(if has { 0.0 } else { 12.0 }); + if rp.before_auto { + p.space_before_auto = Some(false); + } + }) }) .key("Mod+0"), CommandSpec::new("para.removeSpaceAfter", "Remove Space After Paragraph", "Home › Paragraph › Line Spacing", |s, _| { - let has = cur(s).space_after > 0.0; - fmt(s, &|p| p.space_after = Some(if has { 0.0 } else { 8.0 })) + let rp = cur(s); + let has = rp.space_after > 0.0; + fmt(s, &|p| { + p.space_after = Some(if has { 0.0 } else { 8.0 }); + if rp.after_auto { + p.space_after_auto = Some(false); + } + }) }), // Back to the paragraph style: direct indents, spacing, alignment… go; the style, list // membership and drop cap stay. ⌘Q quits on macOS, so Word for Mac uses ⌃Q. @@ -230,8 +254,12 @@ fn cur(s: &Session) -> wordcraft_doc::resolve::ResolvedPara { } pub fn fmt(s: &mut Session, f: &dyn Fn(&mut ParaProps)) -> CmdResult { - let (a, b) = s.sel.ordered(); - s.doc.format_paragraphs(&a, &b, f)?; + let story = s.sel.focus.story; + for path in super::selected_paths(s) { + let para = s.doc.para_mut(story, &path)?; + f(&mut para.props); + para.touch(); + } sel_result(s) } @@ -311,8 +339,8 @@ fn asian_typography(s: &mut Session, v: &Value) -> CmdResult { } pub fn indent(s: &mut Session, _: &Value) -> CmdResult { - let (a, b) = s.sel.ordered(); - let paths = s.doc.para_paths_in(&a, &b); + let a = s.sel.ordered().0; + let paths = super::selected_paths(s); let tab = s.doc.settings.default_tab.max(1.0); for path in paths { let (num, left) = { @@ -333,8 +361,8 @@ pub fn indent(s: &mut Session, _: &Value) -> CmdResult { } pub fn outdent(s: &mut Session, _: &Value) -> CmdResult { - let (a, b) = s.sel.ordered(); - let paths = s.doc.para_paths_in(&a, &b); + let a = s.sel.ordered().0; + let paths = super::selected_paths(s); let tab = s.doc.settings.default_tab.max(1.0); for path in paths { let (num, left) = { @@ -396,7 +424,7 @@ fn apply_style(s: &mut Session, v: &Value) -> CmdResult { return super::format::apply(s, &|c| c.style = Some(id.clone())); } let (a, b) = s.sel.ordered(); - s.doc.format_paragraphs(&a, &b, &|p| { + fmt(s, &|p| { p.style = Some(id.clone()); // Applying a heading or Normal removes direct list numbering (except list styles). if (id.starts_with("Heading") || id == "Title" || id == "Normal") && p.numbering.is_some_and(|n| n.num != 0) { @@ -433,10 +461,10 @@ fn apply_style(s: &mut Session, v: &Value) -> CmdResult { fn list(s: &mut Session, v: &Value, default: ListKind) -> CmdResult { let kind = p::str(v, "kind").and_then(ListKind::parse).unwrap_or(default); let off = p::bool(v, "off").unwrap_or(false); - let (a, b) = s.sel.ordered(); + let a = s.sel.ordered().0; // Toggle: if every selected paragraph already has this kind of list, remove it. let current: Vec> = - s.doc.para_paths_in(&a, &b).iter().map(|p| s.doc.para(a.story, p).and_then(|x| x.props.numbering).filter(|n| n.num != 0)).collect(); + super::selected_paths(s).iter().map(|p| s.doc.para(a.story, p).and_then(|x| x.props.numbering).filter(|n| n.num != 0)).collect(); let same_kind = |n: &NumRef| { s.doc.numbering.level(n.num, 0).is_some_and(|l| { let want = wordcraft_doc::numbering::levels_for(kind); @@ -802,28 +830,42 @@ fn sort(s: &mut Session, v: &Value) -> CmdResult { return Err(CmdError::Failed("select paragraphs in one place to sort".into())); } let (i0, i1) = (a.path.last(), b.path.last()); + let lang = s.doc.para(a.story, &a.path).and_then(|p| super::para_lang(&s.doc, p)).or_else(|| super::default_lang(s)); let bl = s.doc.container_mut(a.story, &a.path)?; let Some(slice) = bl.get_mut(i0..=i1) else { return Err(CmdError::Failed("bad range".into())) }; if slice.iter().any(|b| !matches!(**b, Block::Para(_))) { return Err(CmdError::Failed("sorting tables isn't supported here; use Table › Sort".into())); } - let key = |b: &std::sync::Arc| b.as_para().map(|p| p.plain_text().to_lowercase()).unwrap_or_default(); - slice.sort_by(|x, y| { - let o = natural_cmp(&key(x), &key(y)); + let mut keyed: Vec<_> = + slice.iter().map(|b| (natural_key(&b.as_para().map(|p| p.plain_text()).unwrap_or_default(), lang.as_deref()), b.clone())).collect(); + keyed.sort_by(|(x, _), (y, _)| { + let o = natural_cmp(x, y); if desc { o.reverse() } else { o } }); + for (dst, (_, blk)) in slice.iter_mut().zip(keyed) { + *dst = blk; + } // Keep a section break on the last paragraph of the range where it was. s.sel = crate::Selection { anchor: Pos { off: 0, ..a }, focus: b }; sel_result(s) } -/// Compare strings with embedded numbers numerically ("item 2" < "item 10"). -fn natural_cmp(a: &str, b: &str) -> std::cmp::Ordering { - let num = |s: &str| s.trim().split(|c: char| !c.is_ascii_digit()).next().and_then(|n| n.parse::().ok()); - match (num(a), num(b)) { - (Some(x), Some(y)) if x != y => x.total_cmp(&y), - _ => a.cmp(b), - } +/// A paragraph's sort key: the number it starts with, then its text in its language's order. +fn natural_key(text: &str, lang: Option<&str>) -> (Option, SortKey) { + let num = text.trim().split(|c: char| !c.is_ascii_digit()).next().and_then(|n| n.parse::().ok()); + (num, sort_key(text, lang)) +} + +/// Paragraphs starting with a number come first, in numeric order ("item 2" < "item 10" within +/// the same number, by text); the rest follow in collation order. A total order. +fn natural_cmp(a: &(Option, SortKey), b: &(Option, SortKey)) -> std::cmp::Ordering { + let num = match (a.0, b.0) { + (Some(x), Some(y)) => x.total_cmp(&y), + (Some(_), None) => std::cmp::Ordering::Less, + (None, Some(_)) => std::cmp::Ordering::Greater, + (None, None) => std::cmp::Ordering::Equal, + }; + num.then_with(|| a.1.cmp(&b.1)) } fn create_style(s: &mut Session, v: &Value) -> CmdResult { @@ -1146,3 +1188,29 @@ fn own_format(st: &Style) -> Value { pub fn touch_all(s: &mut Session) { s.dirty = true; } + +/// Which Home › Paragraph list buttons show as on. +#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)] +pub struct ListButtons { + pub bullets: bool, + pub numbering: bool, + pub multilevel: bool, +} + +/// The list buttons for a paragraph with resolved `numbering` (the ribbon passes the paragraph +/// props it already resolved). The paragraph's own level decides Bullets (a bulleted level) or +/// Numbering (a numbered one); Multilevel is on too when the list numbers its levels together: +/// some level's label shows another level's number (`1.1.`) or uses legal numbering. +pub fn list_buttons(doc: &wordcraft_doc::Document, numbering: Option) -> ListButtons { + let Some(n) = numbering.filter(|n| n.num != 0) else { return ListButtons::default() }; + let Some(level) = doc.numbering.level(n.num, n.level) else { return ListButtons::default() }; + let bullets = level.format == wordcraft_doc::section::NumFormat::Bullet; + let joined = |i: u8| { + doc.numbering.level(n.num, i).is_some_and(|l| { + let own = u32::from(i) + 1; + l.legal || l.text.split('%').skip(1).filter_map(|t| t.chars().next()?.to_digit(10)).any(|k| k != own) + }) + }; + let multilevel = (0..9u8).any(joined); + ListButtons { bullets, numbering: !bullets, multilevel } +} diff --git a/crates/engine/src/cmd/paste.rs b/crates/engine/src/cmd/paste.rs index dcb3d3699..4e23ae4fb 100644 --- a/crates/engine/src/cmd/paste.rs +++ b/crates/engine/src/cmd/paste.rs @@ -1,9 +1,11 @@ -//! Paste Special: paste the clipboard in a chosen format (formatted, RTF, HTML, unformatted). +//! Paste Special: paste the clipboard in a chosen format (formatted, RTF, HTML, picture, +//! unformatted), and the formatted content Paste and Merge Formatting take from other programs. //! //! The clipboard can hold several representations of the same content. WordCraft's own copy keeps //! a rich fragment; the system clipboard gives text, and callers (agents, the CLI, a front end -//! with a richer clipboard) may pass `html` or `rtf` too. Text that is itself RTF or HTML source -//! can also be pasted as that format. +//! with a richer clipboard) may pass `html`, `rtf` or a `picture` too. Text that is itself RTF or +//! HTML source can also be pasted as that format. Paste keeps the source formatting of another +//! program's HTML or RTF; Merge Formatting keeps only its emphasis (#687). use std::collections::BTreeMap; use std::sync::Arc; @@ -17,18 +19,25 @@ use crate::{CmdError, CmdResult, CommandSpec, Selection, Session, p}; /// Largest clipboard payload Paste Special accepts, in bytes. pub const MAX_PASTE_BYTES: usize = 16 << 20; +/// Largest clipboard picture Paste Special accepts (base64), in bytes. +pub const MAX_PICTURE_BYTES: usize = 96 << 20; /// Nesting depth we follow when adjusting pasted tables. const MAX_DEPTH: usize = 64; /// The formats, in the order the dialog lists them: (id, label). -pub const FORMATS: [(&str, &str); 4] = - [("formatted", "Formatted Text"), ("rtf", "Rich Text Format (RTF)"), ("html", "HTML Format"), ("text", "Unformatted Text")]; +pub const FORMATS: [(&str, &str); 5] = [ + ("formatted", "Formatted Text"), + ("rtf", "Rich Text Format (RTF)"), + ("html", "HTML Format"), + ("picture", "Picture"), + ("text", "Unformatted Text"), +]; pub fn specs() -> Vec { vec![ CommandSpec::new("edit.pasteSpecial", "Paste Special", "Home › Clipboard › Paste", paste_special) .key("Mod+Alt+V") - .params(r#"{"as"?: "formatted"|"rtf"|"html"|"text", "text"?: string, "html"?: string, "rtf"?: string}"#), + .params(r#"{"as"?: "formatted"|"rtf"|"html"|"picture"|"text", "text"?: string, "html"?: string, "rtf"?: string, "picture"?: base64}"#), ] } @@ -63,6 +72,7 @@ fn source(s: &Session, v: &Value, fmt: &str) -> Option { } "rtf" => p::str(v, "rtf").map(str::to_string).or_else(|| clip_text(s, v).filter(|t| looks_like_rtf(t))).and_then(non_empty), "html" => p::str(v, "html").map(str::to_string).or_else(|| clip_text(s, v).filter(|t| looks_like_html(t))).and_then(non_empty), + "picture" => p::str(v, "picture").map(str::to_string).and_then(non_empty), "text" => clip_text(s, v).and_then(non_empty), _ => None, } @@ -83,11 +93,14 @@ fn paste_special(s: &mut Session, v: &Value) -> CmdResult { return Err(CmdError::Params(format!("`{k}` is larger than {} MB", MAX_PASTE_BYTES >> 20))); } } + if p::str(v, "picture").is_some_and(|t| t.len() > MAX_PICTURE_BYTES) { + return Err(CmdError::Params(format!("`picture` is larger than {} MB", MAX_PICTURE_BYTES >> 20))); + } let ids = available(s, v); let Some(fmt) = p::str(v, "as") else { // No format chosen: list them and ask a front end (if any) to show the dialog. let mut req = json!({"open": "pasteSpecial", "formats": ids}); - for k in ["text", "html", "rtf"] { + for k in ["text", "html", "rtf", "picture"] { if let Some(t) = p::str(v, k) { req[k] = json!(t); } @@ -99,8 +112,9 @@ fn paste_special(s: &mut Session, v: &Value) -> CmdResult { "formatted" | "keep" | "wordcraft" => "formatted", "rtf" => "rtf", "html" => "html", + "picture" | "png" | "image" => "picture", "text" | "unformatted" | "plain" => "text", - other => return Err(CmdError::Params(format!("unknown format `{other}` (formatted, rtf, html or text)"))), + other => return Err(CmdError::Params(format!("unknown format `{other}` (formatted, rtf, html, picture or text)"))), }; let Some(src) = source(s, v, fmt) else { return Err(CmdError::Failed(if ids.is_empty() { @@ -116,8 +130,13 @@ fn paste_special(s: &mut Session, v: &Value) -> CmdResult { r["pastedAs"] = json!(fmt); return Ok(r); } + "picture" => { + let mut r = super::insert::picture(s, &json!({"data": src}))?; + r["pastedAs"] = json!(fmt); + return Ok(r); + } "formatted" => s.clipboard.clone().ok_or_else(|| CmdError::Failed("the clipboard is empty".into()))?, - "html" => adopt(s, wordcraft_formats::html::import(&src)), + "html" => adopt(s, wordcraft_formats::html::import(html_fragment(&src))), _ => adopt(s, wordcraft_formats::rtf::import(src.as_bytes()).map_err(|e| CmdError::Failed(format!("bad RTF: {e}")))?), }; insert(s, frag)?; @@ -126,6 +145,102 @@ fn paste_special(s: &mut Session, v: &Value) -> CmdResult { Ok(r) } +/// The HTML itself from a Windows "HTML Format" (CF_HTML) clipboard payload — a `Version:` header +/// giving byte offsets — or the text unchanged. Hostile offsets fall back to the text after the +/// header. +pub fn html_fragment(html: &str) -> &str { + if !html.starts_with("Version:") { + return html; + } + let offset = |key: &str| -> Option { + let at = html.find(key)?.checked_add(key.len())?; + let digits: &str = html.get(at..)?.split(['\r', '\n']).next()?; + digits.trim().parse().ok() + }; + if let (Some(a), Some(b)) = (offset("StartFragment:"), offset("EndFragment:")) + && let Some(frag) = html.get(a..b) + { + return frag; + } + html.find('<').and_then(|i| html.get(i..)).unwrap_or("") +} + +/// Whether an imported document has anything to paste: text, a table, or a picture it carries. +fn has_content(d: &Document) -> bool { + d.body.iter().any(|b| match &**b { + Block::Para(p) => !p.text.trim().is_empty() || (!p.objects.is_empty() && !d.media.is_empty()), + Block::Table(_) => true, + }) +} + +/// Another program's formatted copy, from the system clipboard's HTML, else its RTF (#687), as a +/// fragment of this document. `None` when neither is given, either is larger than +/// [`MAX_PASTE_BYTES`], or it holds nothing to paste (the caller pastes the text then). +pub fn rich_fragment(s: &mut Session, v: &Value) -> Option { + let fits = |k: &str| p::str(v, k).filter(|t| !t.trim().is_empty() && t.len() <= MAX_PASTE_BYTES); + let doc = fits("html").map(|html| wordcraft_formats::html::import(html_fragment(html))); + let doc = match doc.filter(has_content) { + Some(d) => d, + None => { + let rtf = fits("rtf").filter(|t| looks_like_rtf(t))?; + wordcraft_formats::rtf::import(rtf.as_bytes()).ok().filter(has_content)? + } + }; + Some(adopt(s, doc)) +} + +/// What Paste inserts: WordCraft's own copy while the system clipboard still holds it, else +/// another program's formatted copy (HTML, RTF), else the system clipboard's text. +pub fn paste_fragment(s: &mut Session, v: &Value) -> Result { + let text = p::str(v, "text").map(|t| t.replace("\r\n", "\n")); + if let Some(f) = &s.clipboard + && text.as_ref().is_none_or(|t| f.plain_text() == *t) + && (text.is_some() || (p::str(v, "html").is_none() && p::str(v, "rtf").is_none())) + { + return Ok(f.clone()); + } + if let Some(f) = rich_fragment(s, v) { + return Ok(f); + } + match text { + Some(t) if !t.is_empty() => Ok(Fragment::from_text(&t)), + _ => Err(CmdError::Failed("the clipboard is empty".into())), + } +} + +/// Merge Formatting (#687): the pasted text takes the formatting where it lands — the paragraph's +/// and the caret's — keeping only its emphasis (bold, italic, underline, strikethrough). +pub fn merge_formatting(s: &Session, frag: &mut Fragment) { + let mut base = s.typing_props(); + base.ins = None; + base.del = None; + base.fmt_change = None; + let (start, _) = s.sel.ordered(); + let para = s.doc.para_at(&start).map(|p| p.props.clone()).unwrap_or_default(); + for b in &mut frag.blocks { + each_para(b, 0, &mut |p| { + p.props = para.clone(); + for run in &mut p.runs { + let src = std::mem::replace(&mut run.props, base.clone()); + if src.bold == Some(true) { + run.props.bold = Some(true); + } + if src.italic == Some(true) { + run.props.italic = Some(true); + } + if src.underline.is_some_and(|u| u != wordcraft_doc::props::Underline::None) { + run.props.underline = src.underline; + } + if src.strike == Some(true) { + run.props.strike = Some(true); + } + run.props.link = src.link; + } + p.touch(); + }); + } +} + /// Replace the selection with `frag`, marking it as an insertion under Track Changes. pub fn insert(s: &mut Session, mut frag: Fragment) -> Result<(), CmdError> { super::controls::prepare_edit(s)?; @@ -171,7 +286,7 @@ pub fn insert(s: &mut Session, mut frag: Fragment) -> Result<(), CmdError> { } /// Every paragraph in a block, tables included (bounded depth). -fn each_para(b: &mut Block, depth: usize, f: &mut dyn FnMut(&mut Paragraph)) { +pub(crate) fn each_para(b: &mut Block, depth: usize, f: &mut dyn FnMut(&mut Paragraph)) { if depth > MAX_DEPTH { return; } @@ -218,6 +333,11 @@ fn adopt(s: &mut Session, src: Document) -> Fragment { } } let mut blocks: Vec = src.body.iter().map(|b| (**b).clone()).collect(); + // Pasted pictures fit our text width, like Insert › Pictures (layout keeps their size). + let max_w = crate::io::text_width(&s.doc); + for b in &mut blocks { + crate::io::fit_block_pictures(b, max_w); + } for b in &mut blocks { each_para(b, 0, &mut |p| { if let Some(n) = p.props.numbering.as_mut() diff --git a/crates/engine/src/cmd/references.rs b/crates/engine/src/cmd/references.rs index 4c28a5091..c16ded371 100644 --- a/crates/engine/src/cmd/references.rs +++ b/crates/engine/src/cmd/references.rs @@ -1,7 +1,7 @@ //! References tab: table of contents, footnotes/endnotes, captions; field updates. use serde_json::{Value, json}; -use wordcraft_doc::para::{InlineObject, NoteKind}; +use wordcraft_doc::para::{InlineObject, NoteKind, OBJ}; use wordcraft_doc::props::{CharProps, TabAlign, TabLeader, TabStop}; use wordcraft_doc::{Block, Paragraph, PartKind, Path, Pos, StoryRef, para_block}; @@ -309,6 +309,41 @@ fn toc_range(s: &Session) -> Option<(usize, usize)> { Some((start, end)) } +/// A TOC imported as a range field (`FieldStart`…`FieldEnd`: one without a TOC heading, or with +/// content around it the engine's TOC can't reproduce) becomes the engine's TOC before an update, +/// so Update Table refreshes it rather than adding a second one: an empty `TOC` field where the +/// range started, its old result gone, and the regenerated entries after that paragraph (text +/// that followed the range moves below them). Only an unlocked, top-level TOC that isn't inside +/// another range field, and only when the body has no engine TOC already. +fn range_toc_to_field(s: &mut Session) -> Result<(), CmdError> { + if toc_range(s).is_some() { + return Ok(()); + } + let is_toc = |instr: &str| { + let field = FieldCode::parse(instr); + field.name == "TOC" && !field.has('c') + }; + let Some(r) = s + .doc + .field_ranges(StoryRef::Body) + .into_iter() + .find(|r| r.depth == 0 && !r.locked && is_toc(&r.instr) && r.start.path.depth() == 0 && r.end.path.depth() == 0) + else { + return Ok(()); + }; + let Some(after_end) = r.end.off.checked_add(OBJ.len_utf8()) else { return Ok(()) }; + let end = Pos { off: after_end, ..r.end.clone() }; + let props = s.doc.para_at(&r.start).map(|p| p.props_of_char(r.start.off).clone()).unwrap_or_default(); + let at = s.doc.delete_range(&r.start, &end)?; + let field = InlineObject::Field { instr: r.instr.trim().to_string(), result: String::new(), locked: false, code: false }; + let after = s.doc.insert_object(&at, field, &props)?; + if s.doc.para_at(&after).is_some_and(|p| after.off < p.len()) { + s.doc.split_paragraph(&after)?; + } + s.clamp_selection(); + Ok(()) +} + /// The TOC field instruction in `p` (not a table of figures). fn toc_field(p: &Paragraph) -> Option { p.objects.iter().find_map(|o| match o { @@ -321,6 +356,7 @@ fn toc_field(p: &Paragraph) -> Option { } fn insert_toc(s: &mut Session, v: &Value) -> CmdResult { + range_toc_to_field(s)?; let replace = v.as_object().is_some_and(|m| !m.is_empty()); // Replacing a table puts the new one where the old one was, in place of an empty paragraph. let mut placeholder = false; @@ -408,6 +444,7 @@ pub fn update_toc(s: &mut Session) -> Result<(), CmdError> { /// Rebuild the entries the TOC field's switches ask for. The tab leader is `leader`, or the one /// the entries had (Word keeps it in the entries' tab stops, not in the field). fn regenerate_toc(s: &mut Session, leader: Option) -> Result<(), CmdError> { + range_toc_to_field(s)?; let Some((start, end)) = toc_range(s) else { return Ok(()) }; let Some(instr) = s.doc.body.get(start).and_then(|b| b.as_para()).and_then(toc_field) else { return Ok(()) }; let f = FieldCode::parse(&instr); diff --git a/crates/engine/src/cmd/review.rs b/crates/engine/src/cmd/review.rs index b89f0fab6..752939459 100644 --- a/crates/engine/src/cmd/review.rs +++ b/crates/engine/src/cmd/review.rs @@ -65,7 +65,13 @@ pub fn specs() -> Vec { if w.is_empty() { return Err(CmdError::Params("no word".into())); } - wordcraft_proof::add_word(&w); + // The same rule as loading the dictionary file, so a word added is never dropped later. + if wordcraft_proof::custom_word(&w).is_none() { + let max = wordcraft_proof::MAX_CUSTOM_WORD_CHARS; + return Err(CmdError::Params(format!("{w:?} isn't a single word of at most {max} characters"))); + } + wordcraft_proof::add_word(w.trim()); + s.persist_user_dictionary(&[]); s.relayout(); Ok(json!({"added": w})) }) @@ -78,7 +84,10 @@ pub fn specs() -> Vec { .map(|(a, b, _)| s.doc.para_at(&a).and_then(|p| p.text.get(a.off..b.off).map(str::to_string)).unwrap_or_default()) .unwrap_or_default(), }; - wordcraft_proof::add_word(&w); + // For this session only, as in Word: not added to (or saved with) the dictionary. + if !w.trim().is_empty() { + wordcraft_proof::ignore_word(w.trim()); + } s.relayout(); Ok(json!({"ignored": w})) }) @@ -442,9 +451,42 @@ fn resolve_sel(s: &mut Session, accept: bool) -> CmdResult { // The change at the caret: the run around it, else the paragraph mark right after it, // else the paragraph's formatting, else the table cell's. let found = s.doc.para_at(&a).and_then(|p| { - p.run_ranges() - .find(|(r, c)| r.start <= a.off && a.off <= r.end && (c.ins.is_some() || c.del.is_some() || c.fmt_change.is_some())) - .map(|(r, _)| r) + let runs: Vec<_> = p.run_ranges().collect(); + let i = runs + .iter() + .position(|(r, c)| r.start <= a.off && a.off <= r.end && (c.ins.is_some() || c.del.is_some() || c.fmt_change.is_some()))?; + let (mut start, mut end) = (runs[i].0.start, runs[i].0.end); + let key = (runs[i].1.ins, runs[i].1.del); + if key == (None, None) { + return Some(start..end); + } + // Comment anchors split a revision's text into several runs; the revision goes on past them. + let is_anchor = + |k: usize| matches!(p.object_at(runs[k].0.start), Some(InlineObject::CommentStart { .. } | InlineObject::CommentEnd { .. })); + let same = |k: usize| (runs[k].1.ins, runs[k].1.del) == key; + let mut k = i; + while k > 0 { + if same(k - 1) { + k -= 1; + } else if is_anchor(k - 1) && k > 1 && same(k - 2) { + k -= 2; + } else { + break; + } + start = runs[k].0.start; + } + let mut k = i; + while k + 1 < runs.len() { + if same(k + 1) { + k += 1; + } else if is_anchor(k + 1) && k + 2 < runs.len() && same(k + 2) { + k += 2; + } else { + break; + } + end = runs[k].0.end; + } + Some(start..end) }); let para_fmt = s.doc.para_at(&a).is_some_and(|p| p.props.fmt_change.is_some() || p.props.num_change.is_some()); let mark_fmt = s.doc.para_at(&a).is_some_and(|p| a.off == p.len() && p.mark.fmt_change.is_some()); @@ -738,15 +780,13 @@ fn word_count(s: &mut Session, v: &Value) -> CmdResult { })) } -/// Issues (spelling + grammar) in one paragraph as positions. +/// Issues (spelling + grammar) in one paragraph as positions, each in its run's language and +/// skipping text left unchecked — the same issues the layout underlines. fn para_issues(s: &Session, story: StoryRef, path: &wordcraft_doc::Path) -> Vec<(Pos, Pos, wordcraft_proof::Issue)> { let Some(p) = s.doc.para(story, path) else { return Vec::new() }; - let text = wordcraft_layout::para::proof_text(p); - let mut v: Vec = wordcraft_proof::check_spelling_with(&text, &s.prefs.spelling()); - v.extend(wordcraft_proof::check_grammar(&text)); - v.sort_by_key(|i| i.start); - v.into_iter() - .filter(|i| !p.run_ranges().any(|(r, c)| r.start < i.end && i.start < r.end && (c.no_proof == Some(true) || c.link.is_some()))) + let runs = wordcraft_layout::para::proof_runs(&s.doc, p, None); + wordcraft_layout::para::paragraph_issues(p, &runs, &s.prefs.spelling(), true) + .into_iter() .map(|i| (Pos { story, path: path.clone(), off: i.start }, Pos { story, path: path.clone(), off: i.end }, i)) .collect() } @@ -758,8 +798,12 @@ fn issue_at(s: &Session, at: &Pos) -> Option<(Pos, Pos, wordcraft_proof::Issue)> fn issue_json(s: &Session, a: &Pos, b: &Pos, i: &wordcraft_proof::Issue) -> Value { let word = s.doc.para_at(a).and_then(|p| p.text.get(a.off..b.off)).unwrap_or("").to_string(); - let sugg = if i.kind == wordcraft_proof::IssueKind::Spelling { wordcraft_proof::suggest(&word, 6) } else { i.suggestions.clone() }; - json!({"start": pos_json(a), "end": pos_json(b), "text": word, "kind": if i.kind == wordcraft_proof::IssueKind::Spelling { "spelling" } else { "grammar" }, "message": i.message, "suggestions": sugg}) + let sugg = if i.kind == wordcraft_proof::IssueKind::Spelling { wordcraft_proof::suggest_in(&word, 6, i.lang) } else { i.suggestions.clone() }; + json!({ + "start": pos_json(a), "end": pos_json(b), "text": word, + "kind": if i.kind == wordcraft_proof::IssueKind::Spelling { "spelling" } else { "grammar" }, + "message": i.message, "suggestions": sugg, "lang": i.lang.code(), + }) } /// F7: select the next issue after the caret and return it with suggestions. @@ -1164,6 +1208,29 @@ mod tests { s.doc.para(wordcraft_doc::StoryRef::Body, &wordcraft_doc::Path::top(0)).unwrap() } + /// A comment inside a tracked insertion splits it into several runs; Accept and Reject at the + /// caret resolve the whole insertion, not just the run around the caret. + #[test] + fn accept_and_reject_cover_a_change_split_by_a_comment() { + for accept in [true, false] { + let mut s = Session::new(Document::new()); + s.view.proofing = false; + s.author = "Ana".into(); + s.run("review.trackChanges", &json!({"value": true})).unwrap(); + s.run("text.insert", &json!({"text": "hello world"})).unwrap(); + s.run("select.range", &json!({"anchor": {"block": 0, "off": 3}, "focus": {"block": 0, "off": 8}})).unwrap(); + s.run("review.newComment", &json!({"text": "note"})).unwrap(); + assert!(body_para(&s).runs.len() > 3, "the comment anchors split the insertion"); + s.run("select.range", &json!({"anchor": {"block": 0, "off": 1}, "focus": {"block": 0, "off": 1}})).unwrap(); + s.run(if accept { "review.accept" } else { "review.reject" }, &json!({})).unwrap(); + let p = body_para(&s); + assert!(p.runs.iter().all(|r| r.props.ins.is_none()), "no tracked insertion is left"); + // The comment's anchors are objects in the text; read past them. + let shown: String = p.text.chars().filter(|c| *c != wordcraft_doc::para::OBJ).collect(); + assert_eq!(shown.contains("hello"), accept, "{:?}", p.text); + } + } + /// "plain words here" with "words" made bold, then 14 pt, and the paragraph centred, under /// Track Changes by Ana. fn formatted_while_tracking() -> Session { @@ -1198,7 +1265,7 @@ mod tests { let list = s.run("review.changes", &json!({})).unwrap(); let desc: Vec<&str> = list.as_array().unwrap().iter().filter(|c| c["kind"] == "format").filter_map(|c| c["description"].as_str()).collect(); - assert_eq!(desc, ["Formatted: Align: center", "Formatted: Bold, Size: 14"], "{list:#}"); + assert_eq!(desc, ["Formatted: Align: center", "Formatted: Bold, Bold cs, Size: 14, Size cs: 14"], "{list:#}"); let (d, body_x) = draws(&mut s); let bars = lines(&d).into_iter().filter(|(x0, _, x1, _)| x0 == x1 && *x0 < body_x).count(); diff --git a/crates/engine/src/cmd/table.rs b/crates/engine/src/cmd/table.rs index fed691fcb..e1e09f7fb 100644 --- a/crates/engine/src/cmd/table.rs +++ b/crates/engine/src/cmd/table.rs @@ -1,8 +1,10 @@ //! Table Design and Table Layout tabs. use serde_json::{Value, json}; +use wordcraft_doc::collate::{SortKey, sort_key}; +use wordcraft_doc::locale_number::{format_number, parse_number}; use wordcraft_doc::props::{Align, Border, BorderStyle, Borders, HeightRule, Rgb, TextDirection, VAlign}; -use wordcraft_doc::{Block, Paragraph, Path, Pos, StoryRef, Table, para_block}; +use wordcraft_doc::{Block, Document, Paragraph, Path, Pos, StoryRef, Table, para_block}; use super::sel_result; use crate::{CmdError, CmdResult, CommandSpec, Selection, Session, p}; @@ -175,6 +177,7 @@ pub fn specs() -> Vec { select_cells(s, &tp, Some((r, None))) }) .pure()), + t(CommandSpec::new("table.selectColumn", "Select Column", "Table Layout › Table › Select", select_column).pure()), t(CommandSpec::new("table.selectCell", "Select Cell", "Table Layout › Table › Select", |s, _| { let (tp, r, c) = cell(s)?; select_cells(s, &tp, Some((r, Some(c)))) @@ -384,6 +387,13 @@ fn with_table(s: &mut Session, f: impl Fn(&mut Table)) -> CmdResult { /// Selected cells: (row, cell) pairs covered by the selection (at least the caret's cell). fn selected_cells(s: &Session) -> Vec<(usize, usize)> { + // Whole cells selected (a column, say): just those, by grid column, so merged cells count. + if let Some(range) = s.cell_range() + && s.sel.focus.path.cell().is_some_and(|(tp, _, _)| tp == range.table) + && let Some(t) = s.doc.table(range.story, &range.table) + { + return range.cells(t); + } let (a, b) = s.sel.ordered(); let (Some((ta, ra, ca)), Some((tb, rb, cb))) = (a.path.cell(), b.path.cell()) else { return Vec::new() }; if ta != tb { @@ -454,7 +464,8 @@ fn fix_caret(s: &mut Session, tp: &Path, r: usize, c: usize) -> CmdResult { sel_result(s) } -fn row(s: &mut Session, below: bool) -> CmdResult { +/// Insert Above / Below: rows like the caret's, as many as the selection spans; the caret goes into the first. +pub(super) fn row(s: &mut Session, below: bool) -> CmdResult { let (tp, r, c) = cell(s)?; let cells = selected_cells(s); let n = cells.iter().map(|x| x.0).collect::>().len().max(1); @@ -465,7 +476,8 @@ fn row(s: &mut Session, below: bool) -> CmdResult { fix_caret(s, &tp, if below { r + 1 } else { r }, c) } -fn col(s: &mut Session, right: bool) -> CmdResult { +/// Insert Left / Right: a column beside the caret's cell; the caret goes into it. +pub(super) fn col(s: &mut Session, right: bool) -> CmdResult { let (tp, r, c) = cell(s)?; let t = s.doc.table_mut(s.sel.focus.story, &tp)?; let g = t.grid_col(r, c); @@ -618,22 +630,47 @@ fn cell_text(c: &wordcraft_doc::Cell) -> String { c.blocks.iter().filter_map(|b| b.as_para().map(|p| p.plain_text())).collect::>().join("\n") } +/// The language of a cell's text: that of its first paragraph with text. +fn cell_lang(doc: &wordcraft_doc::Document, c: &wordcraft_doc::Cell) -> Option { + c.blocks.iter().filter_map(|b| b.as_para()).find(|p| !p.plain_text().trim().is_empty()).and_then(|p| super::para_lang(doc, p)) +} + +/// Sort the table's rows by a column. Numbers are read in each cell's language (`1,5` is one and +/// a half in Polish) and sort first, numerically; other text sorts in the column's language. fn sort_table(s: &mut Session, v: &Value) -> CmdResult { let col = p::u64(v, "column").unwrap_or(0) as usize; let desc = p::bool(v, "descending").unwrap_or(false); let header = p::bool(v, "header"); - with_table(s, |t| { - let skip = usize::from(header.unwrap_or(t.rows.first().is_some_and(|r| r.props.header))); - let Some(body) = t.rows.get_mut(skip..) else { return }; - body.sort_by(|a, b| { - let ka = a.cells.get(col).map(cell_text).unwrap_or_default(); - let kb = b.cells.get(col).map(cell_text).unwrap_or_default(); - let o = match (ka.trim().replace(',', "").parse::(), kb.trim().replace(',', "").parse::()) { - (Ok(x), Ok(y)) => x.total_cmp(&y), - _ => ka.to_lowercase().cmp(&kb.to_lowercase()), + let (tp, _, _) = cell(s)?; + let t = s.doc.table(s.sel.focus.story, &tp).ok_or_else(|| CmdError::Failed("no table".into()))?; + let skip = usize::from(header.unwrap_or(t.rows.first().is_some_and(|r| r.props.header))).min(t.rows.len()); + let column = || t.rows.iter().skip(skip).map(|r| r.cells.get(col)); + let lang = column().flatten().find_map(|c| cell_lang(&s.doc, c)).or_else(|| super::default_lang(s)); + // A total order (numbers, then text), so sorting never sees inconsistent comparisons. + let keys: Vec<(Option, SortKey)> = column() + .map(|c| { + let text = c.map(cell_text).unwrap_or_default(); + let num_lang = c.and_then(|c| cell_lang(&s.doc, c)).or_else(|| lang.clone()); + let num = parse_number(text.trim(), num_lang.as_deref()).filter(|x| x.is_finite()); + (num, if num.is_some() { SortKey::default() } else { sort_key(text.trim(), lang.as_deref()) }) + }) + .collect(); + with_table(s, move |t| { + if skip > t.rows.len() { + return; + } + let body = t.rows.split_off(skip); + let mut rows: Vec<_> = keys.iter().cloned().zip(body).collect(); + rows.sort_by(|((a, ka), _), ((b, kb), _)| { + let o = match (a, b) { + (Some(x), Some(y)) => x.total_cmp(y), + (Some(_), None) => std::cmp::Ordering::Less, + (None, Some(_)) => std::cmp::Ordering::Greater, + (None, None) => ka.cmp(kb), }; if desc { o.reverse() } else { o } }); + t.rows.extend(rows.into_iter().map(|(_, r)| r)); }) } @@ -664,14 +701,27 @@ fn formula(s: &mut Session, v: &Value) -> CmdResult { let (tp, r, c) = cell(s)?; let f = p::str(v, "formula").unwrap_or("=SUM(ABOVE)").to_ascii_uppercase(); let t = s.doc.table(s.sel.focus.story, &tp).cloned().ok_or_else(|| CmdError::Failed("no table".into()))?; - let nums: Vec = if f.contains("LEFT") { - t.rows - .get(r) - .map(|row| row.cells.iter().take(c).filter_map(|x| cell_text(x).trim().replace([',', '$'], "").parse().ok()).collect()) - .unwrap_or_default() + let story = s.sel.focus.story; + // Each cell's number is read in that cell's language (`3,50` in Polish is three and a half). + let sources: Vec<&wordcraft_doc::Cell> = if f.contains("LEFT") { + t.rows.get(r).map(|row| row.cells.iter().take(c).collect()).unwrap_or_default() } else { - t.rows.iter().take(r).filter_map(|row| row.cells.get(c)).filter_map(|x| cell_text(x).trim().replace([',', '$'], "").parse().ok()).collect() + t.rows.iter().take(r).filter_map(|row| row.cells.get(c)).collect() }; + let default_lang = super::default_lang(s); + let nums: Vec = sources + .iter() + .filter_map(|x| parse_number(cell_text(x).trim(), cell_lang(&s.doc, x).or_else(|| default_lang.clone()).as_deref())) + .filter(|x| x.is_finite()) + .collect(); + // The result is written in the language of the formula's cell, else of the numbers it adds. + let out_lang = s + .doc + .para(story, &s.sel.focus.path) + .filter(|p| !p.plain_text().trim().is_empty()) + .and_then(|p| super::para_lang(&s.doc, p)) + .or_else(|| sources.iter().find_map(|x| cell_lang(&s.doc, x))) + .or(default_lang); let val = if f.starts_with("=AVERAGE") { if nums.is_empty() { 0.0 } else { nums.iter().sum::() / nums.len() as f64 } } else if f.starts_with("=COUNT") { @@ -685,7 +735,7 @@ fn formula(s: &mut Session, v: &Value) -> CmdResult { } else { nums.iter().sum() }; - let text = if val.fract() == 0.0 && val.abs() < 1e15 { format!("{}", val as i64) } else { format!("{val:.2}") }; + let text = format_number(val, out_lang.as_deref()); let props = s.typing_props(); let at = s.sel.focus.clone(); let end = s.doc.insert_object( @@ -700,26 +750,164 @@ fn formula(s: &mut Session, v: &Value) -> CmdResult { fn select_cells(s: &mut Session, tp: &Path, which: Option<(usize, Option)>) -> CmdResult { let story = s.sel.focus.story; let Some(t) = s.doc.table(story, tp) else { return sel_result(s) }; - let (r0, r1, c0, c1) = match which { - None => (0, t.rows.len().saturating_sub(1), 0, usize::MAX), - Some((r, None)) => (r, r, 0, usize::MAX), - Some((r, Some(c))) => (r, r, c, c), + let (last_row, last_col) = (t.rows.len().saturating_sub(1), t.cols().saturating_sub(1)); + let (rows, cols) = match which { + None => ((0, last_row), (0, last_col)), + Some((r, None)) => ((r, r), (0, last_col)), + Some((r, Some(c))) => ((r, r), grid_cols(t, r, c)), }; - let last_c = t.rows.get(r1).map(|x| x.cells.len().saturating_sub(1)).unwrap_or(0).min(c1); - let mut a = tp.0.clone(); - a.extend([r0 as u32, c0 as u32, 0]); - let mut b = tp.0.clone(); - b.extend([r1 as u32, last_c as u32]); - let end_path = s.doc.para_paths(story).into_iter().rfind(|q| q.0.starts_with(&b)).unwrap_or_else(|| { - let mut x = b.clone(); - x.push(0); - Path(x) - }); - let off = s.doc.para(story, &end_path).map(|x| x.len()).unwrap_or(0); - s.sel = Selection { anchor: Pos { story, path: Path(a), off: 0 }, focus: Pos { story, path: end_path, off } }; + select_range(s, story, tp, rows, cols) +} + +/// The grid columns cell `c` of row `r` covers (first, last). +pub(super) fn grid_cols(t: &Table, r: usize, c: usize) -> (usize, usize) { + let g = t.grid_col(r, c); + let span = t.rows.get(r).and_then(|row| row.cells.get(c)).map_or(1, wordcraft_doc::Cell::span); + (g, g + span - 1) +} + +/// Table Layout › Table › Select › Select Column: the grid columns of the selected cells, every row. +fn select_column(s: &mut Session, _: &Value) -> CmdResult { + let (tp, r, c) = cell(s)?; + let story = s.sel.focus.story; + let t = s.doc.table(story, &tp).ok_or_else(|| CmdError::Disabled("the cursor isn't in a table".into()))?; + let (mut g0, mut g1) = grid_cols(t, r, c); + for (r, c) in selected_cells(s) { + let (a, b) = grid_cols(t, r, c); + (g0, g1) = (g0.min(a), g1.max(b)); + } + let rows = (0, t.rows.len().saturating_sub(1)); + select_range(s, story, &tp, rows, (g0, g1)) +} + +/// Select whole cells of table `tp`: rows `rows`, and of them every cell reaching into grid +/// columns `cols` (both inclusive, either order). The selection runs from the first cell's start +/// to the last cell's end; a block of the cells' paragraphs (`CellRange`) says which cells they +/// are, so the app shades them and, when they aren't one run of the text (part of a column), +/// edits and formatting leave the cells between the corners alone. +pub(crate) fn select_range(s: &mut Session, story: StoryRef, tp: &Path, rows: (usize, usize), cols: (usize, usize)) -> CmdResult { + let t = s.doc.table(story, tp).ok_or_else(|| CmdError::Failed("there's no table there".into()))?; + let (last_row, last_col) = (t.rows.len().saturating_sub(1), t.cols().saturating_sub(1)); + let rows = (rows.0.min(rows.1).min(last_row), rows.0.max(rows.1).min(last_row)); + let cols = (cols.0.min(cols.1).min(last_col), cols.0.max(cols.1).min(last_col)); + let stream = rows.0 == rows.1 || (cols.0 == 0 && cols.1 == last_col); + let range = crate::CellRange { story, table: tp.clone(), rows, cols, stream }; + let picked: std::collections::BTreeSet<(usize, usize)> = range.cells(t).into_iter().collect(); + // Every paragraph of the picked cells (nested tables' too), in document order. + let depth = tp.0.len(); + let in_picked = |q: &Path| match (q.0.get(depth), q.0.get(depth + 1)) { + (Some(r), Some(c)) => q.0.len() > depth + 2 && q.0.starts_with(&tp.0) && picked.contains(&(*r as usize, *c as usize)), + _ => false, + }; + let segments: Vec<(Pos, Pos)> = s + .doc + .para_paths(story) + .into_iter() + .filter(|q| in_picked(q)) + .map(|q| { + let len = s.doc.para(story, &q).map_or(0, Paragraph::len); + (Pos { story, path: q.clone(), off: 0 }, Pos { story, path: q, off: len }) + }) + .collect(); + let (Some(first), Some(last)) = (segments.first(), segments.last()) else { + return Err(CmdError::Failed("there are no cells to select there".into())); + }; + // The end corner is in the last cell's own last paragraph (not a nested table's), so it + // names the cell. + let focus = segments.iter().rev().find(|(_, b)| b.path.0.len() == depth + 3).map_or(&last.1, |(_, b)| b).clone(); + let sel = Selection { anchor: first.0.clone(), focus }; + s.sel = sel.clone(); + s.column = Some(crate::ColumnBlock { sel, anchor_x: 0.0, focus_x: 0.0, segments, cells: Some(range) }); + s.column_mode = false; + s.widened = None; + s.goal_x = None; + s.pending = None; sel_result(s) } +/// Extend the selection from its anchor to `to` the way Word does around tables: from one cell +/// into another of the same table it takes whole cells, the rectangle between them; between a +/// table's cell and text outside the table, whole rows, from the cell's row out to the table's +/// edge facing the other end. Where the selection really runs is kept (`Session::sel_ends`), so +/// extending it again goes on from there. +pub(crate) fn extend_selection(s: &mut Session, to: Pos) { + let ends = Selection { anchor: s.sel.anchor.clone(), focus: to }; + // Shift+Up and Shift+Down keep their goal column. + let goal_x = s.goal_x; + s.sel = ends.clone(); + s.column = None; + widen(s, &ends); + s.widened = if s.sel == ends { None } else { Some(crate::session::Widened { sel: s.sel.clone(), ends }) }; + s.goal_x = goal_x; +} + +/// Widen the selection between `ends` (already `s.sel`) by Word's table rule (see +/// `extend_selection`). +fn widen(s: &mut Session, ends: &Selection) { + let (a, f) = (&ends.anchor, &ends.focus); + if a.story != f.story { + return; + } + // Down through the tables both ends are in, one nesting level (three indexes) at a time. + let mut i = 0; + while let (Some(ab), Some(fb)) = (a.path.0.get(i), f.path.0.get(i)) { + match (cell_at(&a.path.0, i), cell_at(&f.path.0, i)) { + // The same cell: look inside it. + (Some(x), Some(y)) if ab == fb && x == y => i += 3, + // Two cells of one table: the rectangle between them. + (Some((tp, ra, ca)), Some((_, rf, cf))) if ab == fb => { + let cols = s.doc.table(a.story, &tp).map(|t| { + let ((g0, g1), (h0, h1)) = (grid_cols(t, ra, ca), grid_cols(t, rf, cf)); + (g0.min(h0), g1.max(h1)) + }); + if let Some(cols) = cols { + // A table without paragraphs keeps the plain selection. + let _ = select_range(s, a.story, &tp, (ra, rf), cols); + } + return; + } + // Different blocks: an end in a table takes its whole row, out to the table's edge + // facing the other end. + (x, y) if ab != fb => { + let forward = ab < fb; + if let Some((tp, r, _)) = x + && let Some((start, end)) = row_ends(&s.doc, a.story, &tp, r) + { + s.sel.anchor = if forward { start } else { end }; + } + if let Some((tp, r, _)) = y + && let Some((start, end)) = row_ends(&s.doc, f.story, &tp, r) + { + s.sel.focus = if forward { end } else { start }; + } + return; + } + // One paragraph. + _ => return, + } + } +} + +/// The cell of the table at index `i` of paragraph path `p`, if the paragraph is in one there: +/// (the table's path, row, cell). +fn cell_at(p: &[u32], i: usize) -> Option<(Path, usize, usize)> { + match (p.get(..=i), p.get(i + 1), p.get(i + 2)) { + (Some(t), Some(r), Some(c)) if p.len() > i + 3 => Some((Path(t.to_vec()), *r as usize, *c as usize)), + _ => None, + } +} + +/// Where row `r` of table `tp` starts and ends: its first cell's start and its last cell's end. +fn row_ends(doc: &Document, story: StoryRef, tp: &Path, r: usize) -> Option<(Pos, Pos)> { + let mut row = tp.0.clone(); + row.push(u32::try_from(r).ok()?); + let mut paths = doc.para_paths(story).into_iter().filter(|q| q.0.starts_with(&row)); + let first = paths.next()?; + let last = paths.next_back().unwrap_or_else(|| first.clone()); + let len = doc.para(story, &last).map_or(0, Paragraph::len); + Some((Pos { story, path: first, off: 0 }, Pos { story, path: last, off: len })) +} + fn from_text(s: &mut Session, v: &Value) -> CmdResult { let sep = if p::str(v, "separator") == Some("comma") { ',' } else { '\t' }; let (a, b) = s.sel.ordered(); diff --git a/crates/engine/src/cmd/table_mouse.rs b/crates/engine/src/cmd/table_mouse.rs new file mode 100644 index 000000000..0bb1eb50b --- /dev/null +++ b/crates/engine/src/cmd/table_mouse.rs @@ -0,0 +1,864 @@ +//! The rest of Word's table mouse (resizing is `table_resize`): selection areas just above a +//! column, left of a row and along a cell's left edge inside it select the column, row or cell, +//! and a drag from one selects every one up to the pointer; insert buttons left of a row border +//! and above a column border add a row or column there; the move handle just past the table's +//! top-left corner selects the table or, dragged, moves it (the table then floats where it's +//! dropped, the text wrapping around it, as in Word); a double-click on a column border fits the +//! column left of it to its contents. The app sends each click or finished drag here in page +//! points, so each is one command (one undo step when it changes the document). + +use serde_json::Value; +use wordcraft_doc::para::Anchor; +use wordcraft_doc::props::TableFloat; +use wordcraft_doc::{Document, Path, Pos, StoryRef, Table}; +use wordcraft_geom::Rect; +use wordcraft_layout::{DocLayout, Placed}; + +use super::table_resize::{self, CellBox, Edge, REACH, ResizeTarget}; +use super::{sel_result, table}; +use crate::{CellRange, CmdError, CmdResult, CommandSpec, Selection, Session, p}; + +/// The column selection area: from this far above a table's top border (points) to this far +/// below it. +const COLUMN_ABOVE: f32 = 6.0; +const COLUMN_BELOW: f32 = 1.5; +/// How far left of a table (points) the row selection area and the row insert buttons reach. +const ROW_LEFT: f32 = 24.0; +/// The cell selection area's width inside a cell's left edge (points), at the least; it starts +/// past the column border's resize reach. +const CELL_EDGE: f32 = 5.5; +/// The insert buttons answer this near a border (points), and above the table this far out. +const INSERT_REACH: f32 = 3.5; +const INSERT_OUT: f32 = 14.0; +/// An insert button's centre sits this far out from the table's edge (points). +const BUTTON_OUT: f32 = 8.0; +/// The move handle's side and its gap from the table's top-left corner (points). +const HANDLE: f32 = 8.0; +const HANDLE_GAP: f32 = 2.0; +/// The farthest a drag moves a table (points; Word's 22 inches). +const MAX_MOVE: f32 = 1584.0; + +pub fn specs() -> Vec { + vec![ + CommandSpec::new("table.selectAt", "Select Table Cells by Clicking", "Table Layout › Table › Select", select_at) + .params( + r#"{"page": n, "x": pt, "y": pt, "toX"?: pt, "toY"?: pt, "reach"?: pt} (a click in page points, 0-based page, on a selection area of a table in the caret's story: just above a column selects the column, left of a row the row, along a cell's left edge inside it the cell, the move handle past the top-left corner the table; with toX/toY (a drag) every column, row or cell from there to that point. `reach` (default 3) is the column borders' resize reach, which the cell areas keep clear of.)"#, + ) + .pure(), + CommandSpec::new("table.insertAt", "Insert Row or Column by Clicking", "Table Layout › Rows & Columns", insert_at).params( + r#"{"page": n, "x": pt, "y": pt, "reach"?: pt} (a click on a table's insert button: left of a row border adds a row there, above a column border a column; the new row or column is selected, as in Word)"#, + ), + CommandSpec::new("table.moveDrag", "Move Table by Dragging", "Table Layout › Table", move_drag).params( + r#"{"page": n, "x": pt, "y": pt, "toX": pt, "toY": pt, "reach"?: pt} (a drag from the move handle past a table's top-left corner: the table floats where it's dropped, the text wrapping around it; tables in the main text only)"#, + ), + CommandSpec::new("table.fitColumn", "AutoFit Column by Double-Clicking", "Table Layout › Cell Size", fit_column).params( + r#"{"page": n, "x": pt, "y": pt, "reach"?: pt} (a double-click on a column border: the column left of it fits its contents and the columns right of it keep their widths, moving with the border, so the table narrows or widens (up to the right margin), as in Word)"#, + ), + ] +} + +/// What the pointer is over. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +pub enum Spot { + /// The move handle. + Handle, + /// The insert button of horizontal border `n` (0 = the table's top, `n` = the bottom of row + /// `n - 1`). + InsertRow(usize), + /// The insert button of grid line `n` (0 = the table's left edge). + InsertColumn(usize), + /// The selection area above grid column `n`. + Column(usize), + /// The selection area left of row `n`. + Row(usize), + /// The selection area along cell (`row`, `cell`)'s left edge. + Cell(usize, usize), +} + +/// A spot and the table's drawn geometry on the page (`geom.edge` means nothing here). +#[derive(Clone, Debug, PartialEq)] +pub struct Target { + pub spot: Spot, + pub geom: ResizeTarget, +} + +/// An insert button: its centre and the border it inserts at, end to end (page points). +#[derive(Clone, Copy, Debug, PartialEq)] +pub struct InsertMark { + pub center: (f32, f32), + pub from: (f32, f32), + pub to: (f32, f32), +} + +/// The move handle's square (page points), just past the table's top-left corner. +pub fn handle(geom: &ResizeTarget) -> Rect { + Rect::new(geom.left() - HANDLE_GAP - HANDLE, geom.top() - HANDLE_GAP - HANDLE, HANDLE, HANDLE) +} + +/// Whether the table's first row is on this page (its move handle is there). +fn starts_here(geom: &ResizeTarget) -> bool { + geom.rows.first().is_some_and(|r| r.row == 0) +} + +fn on_handle(geom: &ResizeTarget, x: f32, y: f32, reach: f32) -> bool { + let h = handle(geom); + starts_here(geom) && x >= h.x - reach / 2.0 && x <= h.right() + reach / 2.0 && y >= h.y - reach / 2.0 && y <= h.bottom() + reach / 2.0 +} + +fn clean_reach(reach: f32) -> f32 { + if reach.is_finite() { reach.clamp(0.5, 24.0) } else { REACH } +} + +/// Of `lines` (index, position), the one nearest `at` within `reach`. +fn nearest(lines: impl Iterator, at: f32, reach: f32) -> Option { + lines.map(|(i, v)| (i, (at - v).abs())).filter(|(_, d)| *d <= reach).min_by(|a, b| a.1.total_cmp(&b.1)).map(|(i, _)| i) +} + +/// The grid column at page x `x` (the first or the last beyond the table). +fn column_at(geom: &ResizeTarget, x: f32) -> usize { + let n = geom.cols.len().saturating_sub(1); + geom.cols.iter().skip(1).take_while(|cx| **cx <= x).count().min(n.saturating_sub(1)) +} + +/// The drawn row at page y `y` (the first or the last beyond the table). +fn row_at(geom: &ResizeTarget, y: f32) -> usize { + geom.rows.iter().find(|r| y < r.bottom).or(geom.rows.last()).map_or(0, |r| r.row) +} + +/// The spot of one table at (`x`, `y`): the handle, then the insert buttons, then the selection +/// areas. `cells` are its cell boxes on the page. +fn spot_in(geom: &ResizeTarget, cells: &[&CellBox], x: f32, y: f32, reach: f32) -> Option { + if on_handle(geom, x, y, reach) { + return Some(Spot::Handle); + } + let (left, right, top, bottom) = (geom.left(), geom.right(), geom.top(), geom.bottom()); + // Above the table: the insert buttons over its column borders, and the column areas. + if y >= top - INSERT_OUT && y <= top + COLUMN_BELOW && x >= left - INSERT_REACH && x <= right + INSERT_REACH { + let lines = geom.cols.iter().copied().enumerate().filter(|(i, _)| geom.rows.iter().any(|r| r.edges.contains(i))); + if y < top + && let Some(line) = nearest(lines, x, INSERT_REACH) + { + return Some(Spot::InsertColumn(line)); + } + if y >= top - COLUMN_ABOVE && x >= left && x <= right { + return Some(Spot::Column(column_at(geom, x))); + } + } + // Left of the table: the insert buttons beside its row borders, and the row areas. + if x >= left - ROW_LEFT && x < left && y >= top - INSERT_REACH && y <= bottom + INSERT_REACH { + let borders = geom.rows.iter().flat_map(|r| [(r.row, r.top), (r.row + 1, r.bottom)]); + if let Some(line) = nearest(borders, y, INSERT_REACH) { + return Some(Spot::InsertRow(line)); + } + if let Some(r) = geom.rows.iter().find(|r| y >= r.top && y <= r.bottom) { + return Some(Spot::Row(r.row)); + } + } + // Inside a cell, along its left edge. + let edge = CELL_EDGE.max(reach + 2.5); + cells.iter().find(|b| x >= b.rect.x && x <= b.rect.x + edge && y >= b.rect.y && y <= b.rect.bottom()).map(|b| Spot::Cell(b.row, b.cell)) +} + +/// The spot under page point (`x`, `y`) of a table in `story`, of nested tables the innermost's. +/// None on a column or row border or the resize handle: those resize (`table_resize`). `reach` +/// is their resize reach (points). +pub fn target_at(l: &DocLayout, doc: &Document, story: StoryRef, page: usize, x: f32, y: f32, reach: f32) -> Option { + if !x.is_finite() || !y.is_finite() { + return None; + } + let reach = clean_reach(reach); + if table_resize::resize_target(l, doc, page, x, y, reach).is_some() { + return None; + } + let boxes = table_resize::cell_boxes(l, page); + table_resize::tables_on(l, doc, page) + .into_iter() + .filter(|g| g.story == story) + .filter_map(|geom| { + let cells: Vec<&CellBox> = boxes.iter().filter(|b| b.story == geom.story && b.table == geom.table).collect(); + spot_in(&geom, &cells, x, y, reach).map(|spot| Target { spot, geom }) + }) + .max_by_key(|t| t.geom.table.0.len()) +} + +/// The table whose move handle shows with the pointer at (`x`, `y`): the innermost table of +/// `story` starting on this page that the pointer is over, or on its way to the handle. +pub fn handle_table(l: &DocLayout, doc: &Document, story: StoryRef, page: usize, x: f32, y: f32, reach: f32) -> Option { + if !x.is_finite() || !y.is_finite() { + return None; + } + let pad = clean_reach(reach) / 2.0; + table_resize::tables_on(l, doc, page) + .into_iter() + .filter(|g| { + let h = handle(g); + g.story == story && starts_here(g) && x >= h.x - pad && x <= g.right() && y >= h.y - pad && y <= g.bottom() + }) + .max_by_key(|g| g.table.0.len()) +} + +/// The insert button a target is, if it's one. +pub fn insert_mark(t: &Target) -> Option { + let g = &t.geom; + match t.spot { + Spot::InsertRow(line) => { + // The bottom of the row above the border, or the top of the first row. + let y = g.rows.iter().find(|r| r.row + 1 == line).map(|r| r.bottom).or_else(|| g.rows.iter().find(|r| r.row == line).map(|r| r.top))?; + Some(InsertMark { center: (g.left() - BUTTON_OUT, y), from: (g.left(), y), to: (g.right(), y) }) + } + Spot::InsertColumn(line) => { + let x = g.cols.get(line).copied()?; + Some(InsertMark { center: (x, g.top() - BUTTON_OUT), from: (x, g.top()), to: (x, g.bottom()) }) + } + _ => None, + } +} + +/// Whether dragging the move handle moves the table: tables in the main text, not nested ones. +pub fn movable(geom: &ResizeTarget) -> bool { + geom.story == StoryRef::Body && geom.table.0.len() == 1 +} + +/// Where a move-handle drag by (`dx`, `dy`) puts the table's top-left corner on `page`: on the +/// page, its part there above the bottom margin. +pub fn moved_to(l: &DocLayout, page: usize, geom: &ResizeTarget, dx: f32, dy: f32) -> (f32, f32) { + let clean = |d: f32| if d.is_finite() { d.clamp(-MAX_MOVE, MAX_MOVE) } else { 0.0 }; + let (w, h) = (geom.right() - geom.left(), geom.bottom() - geom.top()); + let (page_w, body_bottom) = l.pages.get(page).map_or((0.0, 0.0), |pg| (pg.w, pg.body.bottom())); + let x = (geom.left() + clean(dx)).clamp(0.0, (page_w - w).max(0.0)); + let y = (geom.top() + clean(dy)).clamp(0.0, (body_bottom - h).max(0.0)); + (x, y) +} + +/// Where the selected cells are drawn (page, rect), for shading them. +pub fn cell_rects(l: &DocLayout, doc: &Document, range: &CellRange) -> Vec<(usize, Rect)> { + let Some(t) = doc.table(range.story, &range.table) else { return Vec::new() }; + let picked: std::collections::BTreeSet<(usize, usize)> = range.cells(t).into_iter().collect(); + let mut out = Vec::new(); + for (pi, pg) in l.pages.iter().enumerate() { + for it in &pg.items { + // A vertically merged cell's box spans its rows (#640); its continuation rows' + // `covered` boxes would shade it twice. + if let Placed::Cell { rect, table, row, cell, story, covered: false } = it + && *story == range.story + && *table == range.table + && picked.contains(&(*row, *cell)) + { + out.push((pi, *rect)); + } + } + } + out +} + +/// The click: `page`, `x`, `y` and the resize `reach`. +fn point(v: &Value) -> Result<(usize, f32, f32, f32), CmdError> { + let page = + p::u64(v, "page").map(|n| usize::try_from(n).unwrap_or(usize::MAX)).ok_or_else(|| CmdError::Params("`page` (number) is required".into()))?; + Ok((page, p::req_f32(v, "x")?, p::req_f32(v, "y")?, p::f32(v, "reach").unwrap_or(REACH))) +} + +/// The spot under the click, in the caret's story. +fn target(s: &mut Session, page: usize, x: f32, y: f32, reach: f32) -> Result { + let l = s.layout(); + if l.pages.get(page).is_none() { + return Err(CmdError::Params(format!("no page {page}"))); + } + target_at(&l, &s.doc, s.sel.focus.story, page, x, y, reach) + .ok_or_else(|| CmdError::Failed("there's no table selection area or button there".into())) +} + +fn select_at(s: &mut Session, v: &Value) -> CmdResult { + let (page, x, y, reach) = point(v)?; + let Target { spot, geom } = target(s, page, x, y, reach)?; + let to = p::f32(v, "toX").zip(p::f32(v, "toY")); + let story = geom.story; + let t = s.doc.table(story, &geom.table).ok_or_else(|| CmdError::Failed("there's no table there".into()))?; + let (last_row, last_col) = (t.rows.len().saturating_sub(1), t.cols().saturating_sub(1)); + let (rows, cols) = match spot { + Spot::Column(g) => ((0, last_row), (g, to.map_or(g, |(tx, _)| column_at(&geom, tx)))), + Spot::Row(r) => ((r, to.map_or(r, |(_, ty)| row_at(&geom, ty))), (0, last_col)), + Spot::Cell(r, c) => { + let (g0, g1) = table::grid_cols(t, r, c); + match to { + Some((tx, ty)) => { + let g = column_at(&geom, tx); + ((r, row_at(&geom, ty)), (g0.min(g), g1.max(g))) + } + None => ((r, r), (g0, g1)), + } + } + Spot::Handle => ((0, last_row), (0, last_col)), + Spot::InsertRow(_) | Spot::InsertColumn(_) => return Err(CmdError::Failed("there's an insert button there, not a selection area".into())), + }; + table::select_range(s, story, &geom.table, rows, cols) +} + +/// A cell to insert a column beside at grid line `line`: one starting there (the column goes +/// before it) or, at the table's right edge, a row's last cell (after it). (row, cell, after) +fn beside_line(t: &Table, line: usize) -> Option<(usize, usize, bool)> { + t.rows.iter().enumerate().find_map(|(r, row)| { + let mut g = 0usize; + for (c, cell) in row.cells.iter().enumerate() { + if g == line { + return Some((r, c, false)); + } + g = g.saturating_add(cell.span()); + if g == line && c + 1 == row.cells.len() { + return Some((r, c, true)); + } + } + None + }) +} + +fn insert_at(s: &mut Session, v: &Value) -> CmdResult { + let (page, x, y, reach) = point(v)?; + let Target { spot, geom } = target(s, page, x, y, reach)?; + let story = geom.story; + // The row or column goes beside a cell at the border, as Insert Above / Below / Left / Right + // would put it there, so it takes the border's index (`line`); then, as in Word, it's selected. + let (r, c, after, rows, line) = match spot { + Spot::InsertRow(line) => match line.checked_sub(1) { + Some(above) => (above, 0, true, true, line), + None => (0, 0, false, true, line), + }, + Spot::InsertColumn(line) => { + let t = s.doc.table(story, &geom.table).ok_or_else(|| CmdError::Failed("there's no table there".into()))?; + let (r, c, after) = beside_line(t, line).ok_or_else(|| CmdError::Failed("there's no column border there".into()))?; + (r, c, after, false, line) + } + _ => return Err(CmdError::Failed("there's a selection area there, not an insert button".into())), + }; + let size = |s: &Session| s.doc.table(story, &geom.table).map(|t| (t.rows.len(), t.cols())); + let before = size(s); + let mut path = geom.table.0.clone(); + path.extend([r as u32, c as u32, 0]); + s.column = None; + s.sel = Selection::caret(Pos { story, path: Path(path), off: 0 }); + if rows { + table::row(s, after)?; + } else { + table::col(s, after)?; + } + if size(s) == before { + return Err(CmdError::Failed("the table can't take another row or column".into())); + } + let (sel_rows, sel_cols) = if rows { ((line, line), (0, usize::MAX)) } else { ((0, usize::MAX), (line, line)) }; + table::select_range(s, story, &geom.table, sel_rows, sel_cols) +} + +/// Where table `table` of `story` is first drawn: its page and top-left corner. +fn corner(l: &DocLayout, story: StoryRef, table: &Path) -> Option<(usize, f32, f32)> { + l.pages.iter().enumerate().find_map(|(pi, pg)| { + let cells = pg.items.iter().filter_map(|it| match it { + Placed::Cell { rect, table: t, story: st, .. } if *st == story && t == table => Some(*rect), + _ => None, + }); + cells.fold(None, |acc: Option<(f32, f32)>, r| Some(acc.map_or((r.x, r.y), |(x, y)| (x.min(r.x), y.min(r.y))))).map(|(x, y)| (pi, x, y)) + }) +} + +fn move_drag(s: &mut Session, v: &Value) -> CmdResult { + let (page, x, y, reach) = point(v)?; + let (to_x, to_y) = (p::req_f32(v, "toX")?, p::req_f32(v, "toY")?); + let Target { spot, geom } = target(s, page, x, y, reach)?; + if spot != Spot::Handle { + return Err(CmdError::Failed("there's no table move handle there".into())); + } + if !movable(&geom) { + return Err(CmdError::Failed("only a table in the main text can be moved".into())); + } + let (gx, gy) = moved_to(&s.layout(), page, &geom, to_x - x, to_y - y); + if (gx - geom.left()).abs() < 0.01 && (gy - geom.top()).abs() < 0.01 { + return sel_result(s); + } + // Dragged, a table floats where it's dropped, the text wrapping around it (Word): placed from + // its column and the text where it stands, 0.125" from the text at its sides. + let t = s.doc.table_mut(geom.story, &geom.table)?; + let mut f = t.props.float.unwrap_or(TableFloat { + h_rel: Anchor::Column, + v_rel: Anchor::Paragraph, + dist: [9.0, 0.0, 9.0, 0.0], + overlap: true, + ..Default::default() + }); + f.h_align = None; + f.v_align = None; + t.props.float = Some(f); + table_resize::touch(s, &geom)?; + // Placed by offsets, the table moves as much as they do: one correction from where it landed + // puts it where it was dropped. + if let Some((p0, x0, y0)) = corner(&s.layout(), geom.story, &geom.table) + && p0 == page + { + let fix = |v: f32| if v.is_finite() { v.clamp(-MAX_MOVE, MAX_MOVE) } else { 0.0 }; + let t = s.doc.table_mut(geom.story, &geom.table)?; + if let Some(f) = t.props.float.as_mut() { + f.x = fix(f.x + gx - x0); + f.y = fix(f.y + gy - y0); + } + table_resize::touch(s, &geom)?; + } + sel_result(s) +} + +fn fit_column(s: &mut Session, v: &Value) -> CmdResult { + let (page, x, y, reach) = point(v)?; + let l = s.layout(); + let Some((target, line)) = table_resize::resize_target(&l, &s.doc, page, x, y, reach).and_then(|t| match t.edge { + Edge::Column(line) => Some((t, line)), + _ => None, + }) else { + return Err(CmdError::Failed("there's no column border there".into())); + }; + let t = s.doc.table(target.story, &target.table).ok_or_else(|| CmdError::Failed("there's no table there".into()))?; + let g = line.checked_sub(1).ok_or_else(|| CmdError::Failed("there's no column left of that border".into()))?; + let (narrowest, widest) = + wordcraft_layout::measure_table_columns(&s.doc, t).get(g).copied().ok_or_else(|| CmdError::Failed("there's no such column".into()))?; + let current = target.widths().get(g).copied().unwrap_or(0.0); + // As in Word, the columns right of it keep their widths and move with the border, so the + // table narrows or widens: up to the right margin, never below the column's narrowest content. + let room = current + l.pages.get(page).map_or(0.0, |pg| (pg.body.right() - target.right()).max(0.0)); + let want = widest.min(room.max(narrowest)); + if let Some(w) = target.plan(want - current, 0.0, true).widths { + let t = s.doc.table_mut(target.story, &target.table)?; + table_resize::set_widths(t, w); + table_resize::touch(s, &target)?; + } + sel_result(s) +} + +#[cfg(test)] +mod tests { + use super::*; + use serde_json::json; + use wordcraft_doc::props::Align; + + fn table_session(rows: usize, cols: usize) -> Session { + let mut s = Session::new(Document::new()); + s.run("insert.table", &json!({"rows": rows, "cols": cols})).unwrap(); + s + } + + fn table_path(s: &Session) -> Path { + Path(vec![s.doc.body.iter().position(|b| b.as_table().is_some()).unwrap() as u32]) + } + + fn first_table(s: &Session) -> Table { + s.doc.body.iter().find_map(|b| b.as_table()).cloned().unwrap() + } + + fn cell_pos(s: &Session, r: usize, c: usize) -> Pos { + let mut path = table_path(s).0; + path.extend([r as u32, c as u32, 0]); + Pos { story: StoryRef::Body, path: Path(path), off: 0 } + } + + /// Type "r{row}c{cell}" into every cell. + fn fill(s: &mut Session) { + let t = first_table(s); + for (r, row) in t.rows.iter().enumerate() { + for c in 0..row.cells.len() { + s.sel = Selection::caret(cell_pos(s, r, c)); + s.run("text.insert", &json!({"text": format!("r{r}c{c}")})).unwrap(); + } + } + } + + fn para(s: &Session, r: usize, c: usize) -> wordcraft_doc::Paragraph { + s.doc.para(StoryRef::Body, &cell_pos(s, r, c).path).cloned().unwrap() + } + + fn bold(s: &Session, r: usize, c: usize) -> bool { + let p = para(s, r, c); + if p.is_empty() { p.mark.bold == Some(true) } else { p.run_ranges().all(|(_, ch)| ch.bold == Some(true)) } + } + + fn cell_rect(s: &mut Session, row: usize, cell: usize) -> Rect { + s.layout().pages[0] + .items + .iter() + .find_map(|it| match it { + Placed::Cell { rect, row: r, cell: c, .. } if *r == row && *c == cell => Some(*rect), + _ => None, + }) + .unwrap() + } + + fn spot_at(s: &mut Session, x: f32, y: f32) -> Option { + let l = s.layout(); + target_at(&l, &s.doc, StoryRef::Body, 0, x, y, REACH).map(|t| t.spot) + } + + fn mid_y(r: Rect) -> f32 { + r.y + r.h / 2.0 + } + + /// The areas around and in a table: column, row and cell selection areas, the insert buttons + /// and the move handle; borders stay the resize module's, the text the caret's. + #[test] + fn the_areas_of_a_table_are_found() { + let mut s = table_session(3, 3); + let (a, b) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1)); + let (left, top) = (a.x, a.y); + assert_eq!(spot_at(&mut s, b.x + b.w / 2.0, top - 3.0), Some(Spot::Column(1))); + assert_eq!(spot_at(&mut s, b.x + b.w / 2.0, top + 1.0), Some(Spot::Column(1))); + let r1 = cell_rect(&mut s, 1, 0); + assert_eq!(spot_at(&mut s, left - 10.0, mid_y(r1)), Some(Spot::Row(1))); + let c = cell_rect(&mut s, 1, 2); + assert_eq!(spot_at(&mut s, c.x + 4.5, mid_y(c)), Some(Spot::Cell(1, 2))); + assert_eq!(spot_at(&mut s, r1.x + 2.0, mid_y(r1)), Some(Spot::Cell(1, 0)), "the first column's edge has no resize border"); + assert_eq!(spot_at(&mut s, c.x + 20.0, mid_y(c)), None, "the text is the caret's"); + assert_eq!(spot_at(&mut s, b.x, mid_y(a)), None, "a column border resizes"); + assert_eq!(spot_at(&mut s, b.x, top - 8.0), Some(Spot::InsertColumn(1))); + assert_eq!(spot_at(&mut s, left - 10.0, a.bottom()), Some(Spot::InsertRow(1))); + assert_eq!(spot_at(&mut s, left - 10.0, top), Some(Spot::InsertRow(0))); + let h = handle(&table_resize::tables_on(&s.layout(), &s.doc, 0)[0]); + assert_eq!(spot_at(&mut s, h.x + h.w / 2.0, h.y + h.h / 2.0), Some(Spot::Handle)); + // Another story's tables aren't in reach. + let l = s.layout(); + assert!(target_at(&l, &s.doc, StoryRef::Part(7), 0, b.x + b.w / 2.0, top - 3.0, REACH).is_none()); + } + + /// A selected column's text is its own cells', so the status bar and Word Count count only + /// them, not the cells between its ends in the text. + #[test] + fn a_selected_columns_text_is_its_cells() { + let mut s = table_session(3, 3); + fill(&mut s); + let b = cell_rect(&mut s, 0, 1); + s.run("table.selectAt", &json!({"page": 0, "x": b.x + b.w / 2.0, "y": b.y - 3.0})).unwrap(); + assert_eq!(s.selected_text(), "r0c1\nr1c1\nr2c1"); + assert_eq!(s.run("review.wordCount", &json!({})).unwrap()["words"], 3); + } + + /// A click above a column selects just its cells: formatting, alignment and Delete leave the + /// cells between its ends alone, each one undo step; table commands take the column. + #[test] + fn a_selected_column_is_edited_cell_by_cell() { + let mut s = table_session(3, 3); + fill(&mut s); + let b = cell_rect(&mut s, 0, 1); + s.run("table.selectAt", &json!({"page": 0, "x": b.x + b.w / 2.0, "y": b.y - 3.0})).unwrap(); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.rows, range.cols, range.stream), ((0, 2), (1, 1), false)); + assert_eq!(s.column_segments().unwrap().len(), 3); + s.run("format.bold", &json!({})).unwrap(); + for r in 0..3 { + assert!(bold(&s, r, 1), "row {r} of the column is bold"); + assert!(!bold(&s, r, 0) && !bold(&s, r, 2), "row {r}'s other cells aren't"); + } + s.run("para.alignCenter", &json!({})).unwrap(); + assert_eq!(para(&s, 1, 1).props.align, Some(Align::Center)); + assert_eq!(para(&s, 0, 2).props.align, None, "the cells between the corners keep theirs"); + assert!(s.cell_range().is_some(), "formatting keeps the selection"); + s.run("table.shading", &json!({"color": "FF0000"})).unwrap(); + let t = first_table(&s); + assert!(t.rows.iter().all(|row| row.cells[1].props.shading.is_some() && row.cells[0].props.shading.is_none())); + let depth = s.undo_depth(); + s.run("text.delete", &json!({})).unwrap(); + assert_eq!(s.undo_depth(), depth + 1); + for r in 0..3 { + assert!(para(&s, r, 1).is_empty()); + assert_eq!(para(&s, r, 0).plain_text(), format!("r{r}c0")); + assert_eq!(para(&s, r, 2).plain_text(), format!("r{r}c2")); + } + s.run("edit.undo", &json!({})).unwrap(); + assert_eq!(para(&s, 2, 1).plain_text(), "r2c1"); + assert_eq!(s.cell_range().map(|r| r.cols), Some((1, 1)), "undo brings the column selection back"); + } + + /// Numbering a column numbers its cells as one list, like Word. + #[test] + fn numbering_a_column_makes_one_list() { + let mut s = table_session(3, 2); + fill(&mut s); + s.sel = Selection::caret(cell_pos(&s, 1, 0)); + s.run("table.selectColumn", &json!({})).unwrap(); + s.run("para.numbering", &json!({})).unwrap(); + let nums: Vec<_> = (0..3).map(|r| para(&s, r, 0).props.numbering.map(|n| n.num)).collect(); + assert!(nums[0].is_some_and(|n| n != 0) && nums.iter().all(|n| *n == nums[0]), "{nums:?}"); + assert!((0..3).all(|r| para(&s, r, 1).props.numbering.is_none())); + } + + /// An empty column's cells take character formatting in their marks; Tab moves on without + /// clearing a selected column; a click in place ends a selected empty cell. + #[test] + fn empty_cells_tab_and_clicks() { + let mut s = table_session(2, 2); + s.sel = Selection::caret(cell_pos(&s, 0, 0)); + s.run("table.selectColumn", &json!({})).unwrap(); + s.run("format.bold", &json!({})).unwrap(); + assert!(bold(&s, 0, 0) && bold(&s, 1, 0) && !bold(&s, 0, 1) && !bold(&s, 1, 1)); + fill(&mut s); + s.sel = Selection::caret(cell_pos(&s, 0, 1)); + s.run("table.selectColumn", &json!({})).unwrap(); + s.run("text.tab", &json!({})).unwrap(); + assert_eq!(para(&s, 0, 1).plain_text(), "r0c1"); + assert_eq!(para(&s, 1, 1).plain_text(), "r1c1"); + // A selected empty cell's corners are one place; clicking there ends the selection. + let mut s = table_session(2, 2); + let c = cell_rect(&mut s, 1, 1); + s.run("table.selectAt", &json!({"page": 0, "x": c.x + 4.5, "y": mid_y(c)})).unwrap(); + assert!(s.cell_range().is_some() && s.sel.is_collapsed()); + let at = serde_json::to_value(s.sel.focus.clone()).unwrap(); + s.run("caret.set", &json!({"pos": at})).unwrap(); + assert!(s.cell_range().is_none()); + } + + /// Rows, cells and drags: whole rows and cells of one row are an ordinary stream selection; + /// a drag over cells of several rows selects their rectangle only. + #[test] + fn rows_cells_and_drags_select_rectangles() { + let mut s = table_session(3, 3); + fill(&mut s); + let r1 = cell_rect(&mut s, 1, 0); + s.run("table.selectAt", &json!({"page": 0, "x": r1.x - 10.0, "y": mid_y(r1)})).unwrap(); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.rows, range.cols, range.stream), ((1, 1), (0, 2), true)); + assert!(s.column_segments().is_none(), "a row is one run of the text"); + let r2 = cell_rect(&mut s, 2, 0); + s.run("table.selectAt", &json!({"page": 0, "x": r1.x - 10.0, "y": mid_y(r1), "toX": r1.x - 10.0, "toY": mid_y(r2) + 50.0})).unwrap(); + assert_eq!(s.cell_range().map(|r| r.rows), Some((1, 2))); + // From cell (0, 0) to cell (1, 1): four cells, not the row between. + let (a, d) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 1, 1)); + s.run("table.selectAt", &json!({"page": 0, "x": a.x + 2.0, "y": mid_y(a), "toX": d.x + d.w / 2.0, "toY": mid_y(d)})).unwrap(); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.rows, range.cols, range.stream), ((0, 1), (0, 1), false)); + s.run("format.italic", &json!({})).unwrap(); + let italic = |s: &Session, r, c| para(s, r, c).run_ranges().all(|(_, ch)| ch.italic == Some(true)); + assert!(italic(&s, 0, 0) && italic(&s, 1, 1) && !italic(&s, 0, 2) && !italic(&s, 1, 2)); + // Dragging across columns: from the first to the last is every column, all rows. + let (c0, c2) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 2)); + s.run("table.selectAt", &json!({"page": 0, "x": c0.x + 20.0, "y": c0.y - 3.0, "toX": c2.right() + 30.0, "toY": c0.y})).unwrap(); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.rows, range.cols, range.stream), ((0, 2), (0, 2), true)); + // The move handle selects the table. + let h = handle(&table_resize::tables_on(&s.layout(), &s.doc, 0)[0]); + s.run("table.selectAt", &json!({"page": 0, "x": h.x + 4.0, "y": h.y + 4.0})).unwrap(); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((0, 2), (0, 2)))); + // The ribbon's Select Row is the same whole-cell selection. + s.sel = Selection::caret(cell_pos(&s, 2, 1)); + s.run("table.selectRow", &json!({})).unwrap(); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols, r.stream)), Some(((2, 2), (0, 2), true))); + } + + /// The insert buttons add a row or column at their border, one undo step, and select it (as + /// Word does); typing then goes into its first cell. + #[test] + fn the_insert_buttons_add_rows_and_columns() { + let mut s = table_session(2, 3); + fill(&mut s); + let a = cell_rect(&mut s, 0, 0); + s.run("table.insertAt", &json!({"page": 0, "x": a.x - 10.0, "y": a.bottom()})).unwrap(); + let t = first_table(&s); + assert_eq!(t.rows.len(), 3); + assert!(para(&s, 1, 0).is_empty() && para(&s, 2, 0).plain_text() == "r1c0", "the new row is the second"); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((1, 1), (0, 2))), "it's selected"); + s.run("text.insert", &json!({"text": "x"})).unwrap(); + assert_eq!(para(&s, 1, 0).plain_text(), "x"); + assert_eq!((first_table(&s).rows.len(), para(&s, 2, 0).plain_text()), (3, "r1c0".to_string())); + s.run("edit.undo", &json!({})).unwrap(); + s.run("edit.undo", &json!({})).unwrap(); + assert_eq!(first_table(&s).rows.len(), 2); + // Above the first row. + s.run("table.insertAt", &json!({"page": 0, "x": a.x - 10.0, "y": a.y})).unwrap(); + assert!(para(&s, 0, 0).is_empty() && para(&s, 1, 0).plain_text() == "r0c0"); + assert_eq!(s.cell_range().map(|r| r.rows), Some((0, 0))); + s.run("edit.undo", &json!({})).unwrap(); + // A column at the border between the first two, the table keeping its width. + let width: f32 = first_table(&s).grid.iter().sum(); + let b = cell_rect(&mut s, 0, 1); + s.run("table.insertAt", &json!({"page": 0, "x": b.x, "y": a.y - 8.0})).unwrap(); + let t = first_table(&s); + assert_eq!(t.cols(), 4); + assert!(para(&s, 0, 1).is_empty() && para(&s, 0, 2).plain_text() == "r0c1"); + assert!((t.grid.iter().sum::() - width).abs() < 0.5); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((0, 1), (1, 1))), "it's selected"); + // And at the right edge. + let e = cell_rect(&mut s, 0, 3); + s.run("table.insertAt", &json!({"page": 0, "x": e.right(), "y": a.y - 8.0})).unwrap(); + assert_eq!(first_table(&s).cols(), 5); + assert!(para(&s, 0, 4).is_empty() && para(&s, 0, 3).plain_text() == "r0c2"); + assert_eq!(s.cell_range().map(|r| r.cols), Some((4, 4))); + } + + /// Dragging the move handle floats the table where it's dropped; one undo puts it back; a + /// click on the handle moves nothing. + #[test] + fn dragging_the_handle_moves_the_table() { + let mut s = table_session(2, 2); + let a = cell_rect(&mut s, 0, 0); + let h = handle(&table_resize::tables_on(&s.layout(), &s.doc, 0)[0]); + let (hx, hy) = (h.x + h.w / 2.0, h.y + h.h / 2.0); + s.run("table.moveDrag", &json!({"page": 0, "x": hx, "y": hy, "toX": hx, "toY": hy})).unwrap(); + assert!(first_table(&s).props.float.is_none(), "a click doesn't move it"); + s.run("table.moveDrag", &json!({"page": 0, "x": hx, "y": hy, "toX": hx + 50.0, "toY": hy + 30.0})).unwrap(); + let f = first_table(&s).props.float.unwrap(); + assert_eq!((f.h_rel, f.v_rel, f.overlap), (Anchor::Column, Anchor::Paragraph, true)); + let moved = cell_rect(&mut s, 0, 0); + assert!((moved.x - (a.x + 50.0)).abs() < 0.5 && (moved.y - (a.y + 30.0)).abs() < 0.5, "{a:?} → {moved:?}"); + // Moving it again starts from where it floats. + let h = handle(&table_resize::tables_on(&s.layout(), &s.doc, 0)[0]); + let (hx, hy) = (h.x + h.w / 2.0, h.y + h.h / 2.0); + s.run("table.moveDrag", &json!({"page": 0, "x": hx, "y": hy, "toX": hx - 20.0, "toY": hy + 10.0})).unwrap(); + let again = cell_rect(&mut s, 0, 0); + assert!((again.x - (moved.x - 20.0)).abs() < 0.5 && (again.y - (moved.y + 10.0)).abs() < 0.5, "{moved:?} → {again:?}"); + s.run("edit.undo", &json!({})).unwrap(); + s.run("edit.undo", &json!({})).unwrap(); + assert!(first_table(&s).props.float.is_none()); + assert_eq!(cell_rect(&mut s, 0, 0), a); + // Nested tables and tables outside the main text stay where they are. + let mut nested = table_resize::tables_on(&s.layout(), &s.doc, 0)[0].clone(); + assert!(movable(&nested)); + nested.table = Path(vec![0, 0, 0, 1]); + assert!(!movable(&nested)); + } + + /// A double-click on a column border fits the column left of it to its text; as in Word, the + /// columns right of it keep their widths and move along, so the table narrows. The right + /// edge's column fits the same way. + #[test] + fn a_double_click_on_a_border_fits_the_column() { + let mut s = table_session(1, 3); + s.sel = Selection::caret(cell_pos(&s, 0, 0)); + s.run("text.insert", &json!({"text": "ab"})).unwrap(); + let (a, b, c) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1), cell_rect(&mut s, 0, 2)); + let fit = wordcraft_layout::measure_table_columns(&s.doc, &first_table(&s)); + s.run("table.fitColumn", &json!({"page": 0, "x": a.right(), "y": mid_y(a)})).unwrap(); + let (a1, b1, c1) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1), cell_rect(&mut s, 0, 2)); + assert!((a1.w - fit[0].1).abs() < 0.5 && a1.w < a.w - 50.0, "{a:?} → {a1:?}, {fit:?}"); + assert!((b1.w - b.w).abs() < 0.5 && (c1.w - c.w).abs() < 0.5, "the columns right of it keep their widths"); + assert!((b1.x - b.x - (a1.w - a.w)).abs() < 0.5 && (c1.right() - c.right() - (a1.w - a.w)).abs() < 0.5, "and move along"); + s.run("table.fitColumn", &json!({"page": 0, "x": c1.right(), "y": mid_y(c1)})).unwrap(); + let c2 = cell_rect(&mut s, 0, 2); + assert!((c2.w - fit[2].1).abs() < 0.5 && c2.w < c.w, "{c:?} → {c2:?}"); + } + + /// Pasting formatting onto a selected column leaves its text (it used to clear it). + #[test] + fn the_format_painter_keeps_a_selected_columns_text() { + let mut s = table_session(2, 2); + fill(&mut s); + s.sel = Selection::caret(cell_pos(&s, 0, 0)); + s.run("edit.formatPainter", &json!({})).unwrap(); + s.sel = Selection::caret(cell_pos(&s, 0, 1)); + s.run("table.selectColumn", &json!({})).unwrap(); + s.run("edit.pasteFormat", &json!({})).unwrap(); + assert_eq!(para(&s, 0, 1).plain_text(), "r0c1"); + assert_eq!(para(&s, 1, 1).plain_text(), "r1c1"); + } + + /// Where the app extends the selection while dragging (`caret.set` with `extend`). + fn drag_to(s: &mut Session, to: Pos) { + s.run("caret.set", &json!({"pos": to, "extend": true})).unwrap(); + } + + /// An ordinary drag from one cell into another selects the rectangle of whole cells between + /// them, so bold takes just those, as in Word; back in the first cell it's text again, from + /// where the drag started. Out of a nested table, the cells are the table's around it. + #[test] + fn a_drag_across_cells_selects_their_rectangle() { + let mut s = table_session(3, 3); + fill(&mut s); + let start = Pos { off: 1, ..cell_pos(&s, 0, 1) }; + s.run("caret.set", &json!({"pos": start})).unwrap(); + let down = Pos { off: 2, ..cell_pos(&s, 2, 1) }; + drag_to(&mut s, down); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.rows, range.cols, range.stream), ((0, 2), (1, 1), false)); + s.run("format.bold", &json!({})).unwrap(); + for r in 0..3 { + assert!(bold(&s, r, 1) && !bold(&s, r, 0) && !bold(&s, r, 2), "row {r}: just the dragged column"); + } + // On to the left: the rectangle still runs from the cell the drag started in. + let left = cell_pos(&s, 1, 0); + drag_to(&mut s, left); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((0, 1), (0, 1)))); + // Back in that cell: its text, from where the drag started. + let back = Pos { off: 3, ..cell_pos(&s, 0, 1) }; + drag_to(&mut s, back.clone()); + assert!(s.cell_range().is_none()); + assert_eq!(s.sel, Selection { anchor: start, focus: back }); + // From a nested table's cell into another cell of the table around it. + let mut s = table_session(2, 2); + s.sel = Selection::caret(cell_pos(&s, 0, 0)); + s.run("insert.table", &json!({"rows": 2, "cols": 2})).unwrap(); + assert_eq!(s.sel.focus.path.0.len(), 7, "the caret is in the nested table"); + let to = cell_pos(&s, 1, 1); + drag_to(&mut s, to); + let range = s.cell_range().cloned().unwrap(); + assert_eq!((range.table, range.rows, range.cols), (table_path(&s), (0, 1), (0, 1))); + } + + /// Shift+Down from a cell takes it and the cell below, whole; on past the table, the anchor's + /// row and every row after it, then the text; Shift+Up comes back to the cells, then into the + /// first cell's text. Shift+click goes on from where the selection started, and from text + /// before the table a drag into a cell takes the rows down to it, whole. + #[test] + fn shift_arrows_and_clicks_select_cells_then_rows() { + let mut s = Session::new(Document::new()); + s.run("text.insert", &json!({"text": "Intro"})).unwrap(); + s.run("text.newParagraph", &json!({})).unwrap(); + s.run("insert.table", &json!({"rows": 3, "cols": 3})).unwrap(); + fill(&mut s); + let after = table_path(&s).last() + 1; + let shift = |s: &mut Session, id: &str| { + s.run(id, &json!({"extend": true})).unwrap(); + }; + s.sel = Selection::caret(cell_pos(&s, 1, 1)); + shift(&mut s, "caret.down"); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((1, 2), (1, 1)))); + shift(&mut s, "caret.down"); + assert!(s.cell_range().is_none()); + assert_eq!(s.sel.anchor, cell_pos(&s, 1, 0), "from the start of the anchor's row"); + assert_eq!(s.sel.focus.path, Path::top(after)); + s.run("format.italic", &json!({})).unwrap(); + let italic = |s: &Session, r, c| para(s, r, c).run_ranges().all(|(_, ch)| ch.italic == Some(true)); + assert!(italic(&s, 1, 0) && italic(&s, 2, 2) && !italic(&s, 0, 1), "whole rows from the anchor's"); + shift(&mut s, "caret.up"); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((1, 2), (1, 1)))); + shift(&mut s, "caret.up"); + assert!(s.cell_range().is_none()); + assert_eq!(s.sel.anchor, cell_pos(&s, 1, 1)); + assert_eq!(s.sel.focus.path, cell_pos(&s, 1, 1).path); + // Shift+click: the rectangle from the anchor's cell. + let corner = cell_pos(&s, 0, 0); + s.run("caret.set", &json!({"pos": corner, "extend": true})).unwrap(); + assert_eq!(s.cell_range().map(|r| (r.rows, r.cols)), Some(((0, 1), (0, 1)))); + // From the text before the table into a cell: the rows down to that cell's, whole. + s.sel = Selection::caret(Pos::body(0, 2)); + let into = cell_pos(&s, 1, 1); + drag_to(&mut s, into); + assert!(s.cell_range().is_none()); + assert_eq!(s.sel, Selection { anchor: Pos::body(0, 2), focus: Pos { off: 4, ..cell_pos(&s, 1, 2) } }); + } + + /// Clicks away from the areas, and hostile numbers: an error and no change, never a panic. + #[test] + fn misses_and_hostile_params_change_nothing() { + let mut s = table_session(2, 2); + let start = first_table(&s); + let c = cell_rect(&mut s, 1, 1); + let mid = json!({"page": 0, "x": c.x + c.w / 2.0, "y": mid_y(c), "toX": 1, "toY": 1}); + for id in ["table.selectAt", "table.insertAt", "table.moveDrag", "table.fitColumn"] { + assert!(s.run(id, &mid).is_err(), "{id}"); + assert!(s.run(id, &json!({"page": 99, "x": 1, "y": 1, "toX": 1, "toY": 1})).is_err(), "{id}"); + assert!(s.run(id, &json!({"page": 0, "x": 1e30, "y": -1e30, "toX": 1e30, "toY": 1e30, "reach": 1e30})).is_err(), "{id}"); + assert!(s.run(id, &json!({})).is_err(), "{id}"); + } + // An insert button isn't a selection area, nor a selection area a button. + let a = cell_rect(&mut s, 0, 0); + assert!(s.run("table.selectAt", &json!({"page": 0, "x": a.x - 10.0, "y": a.bottom()})).is_err()); + assert!(s.run("table.insertAt", &json!({"page": 0, "x": a.x - 10.0, "y": mid_y(a)})).is_err()); + assert_eq!(first_table(&s), start); + // A drag to the ends of the earth keeps the table on its page. + let h = handle(&table_resize::tables_on(&s.layout(), &s.doc, 0)[0]); + s.run("table.moveDrag", &json!({"page": 0, "x": h.x + 4.0, "y": h.y + 4.0, "toX": 3e38, "toY": -3e38})).unwrap(); + let f = first_table(&s).props.float.unwrap(); + assert!(f.x.is_finite() && f.y.is_finite()); + let moved = cell_rect(&mut s, 0, 0); + assert!(moved.x >= -0.5 && moved.right() <= 612.5 && moved.y >= -0.5, "{moved:?}"); + } +} diff --git a/crates/engine/src/cmd/table_resize.rs b/crates/engine/src/cmd/table_resize.rs new file mode 100644 index 000000000..5c4314a6e --- /dev/null +++ b/crates/engine/src/cmd/table_resize.rs @@ -0,0 +1,576 @@ +//! Resizing a table with the mouse, like Word: drag a border between columns (or the table's +//! right edge), a row's bottom border, or the small handle just outside the table's bottom-right +//! corner to scale the whole table. The geometry comes from the layout (what is on the page, in +//! page points), so the result is what the drag showed even when autofit or a percentage width +//! made the drawn columns differ from the stored grid. The app sends each finished drag here, +//! so every drag is one undo step. + +use serde_json::Value; +use wordcraft_doc::props::HeightRule; +use wordcraft_doc::{Document, Path, StoryRef, Table}; +use wordcraft_geom::Rect; +use wordcraft_layout::{DocLayout, Placed}; + +use super::sel_result; +use crate::{CmdError, CmdResult, CommandSpec, Session, p}; + +/// How near (points) a border has to be to the pointer, by default. +pub const REACH: f32 = 3.0; +/// The corner handle's side (points); it sits just outside the table's bottom-right corner. +pub const HANDLE: f32 = 6.0; +/// The narrowest column a drag leaves (points). +pub const MIN_COL: f32 = 6.0; +/// The lowest row a drag leaves (points). +pub const MIN_ROW: f32 = 4.0; +/// The largest column width or row height a drag sets (points; Word's 22 inches). +const MAX_LEN: f32 = 1584.0; + +pub fn specs() -> Vec { + vec![CommandSpec::new("table.resizeDrag", "Resize Table by Dragging", "Table Layout › Cell Size", resize_drag).params( + r#"{"page": n, "x": pt, "y": pt, "toX": pt, "toY": pt, "shift"?: bool, "reach"?: pt} (a mouse drag in page points, 0-based page, starting on a table border or the resize handle: a column border moves (the column to its right gives or takes the difference, so the table keeps its width; with shift, or on the table's right edge, only the column left of it changes); a row's bottom border sets that row's height; the handle just outside the bottom-right corner scales every column and the rows on that page. `reach` (default 3) is how near the start has to be.)"#, + )] +} + +/// What a press on the page would grab. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +pub enum Edge { + /// The vertical border at grid line `line` (1 = between the first two columns; the last is + /// the table's right edge). + Column(usize), + /// The bottom border of row `row`. + Row(usize), + /// The resize handle at the bottom-right corner. + Corner, +} + +/// A row of the table as drawn on one page. +#[derive(Clone, Debug, PartialEq)] +pub struct DrawnRow { + pub row: usize, + pub top: f32, + pub bottom: f32, + /// Grid lines that are cell borders in this row (a merged cell hides the ones it covers). + pub(crate) edges: Vec, +} + +/// A table border (or the handle) under the pointer, with the table's drawn geometry. +#[derive(Clone, Debug, PartialEq)] +pub struct ResizeTarget { + pub edge: Edge, + pub story: StoryRef, + pub table: Path, + /// Drawn grid lines, page x, left to right (columns + 1 of them). + pub cols: Vec, + /// Drawn rows on this page, top to bottom. + pub rows: Vec, + /// Whether the table's last row is on this page (the handle is shown only then). + pub ends_here: bool, +} + +/// What a drag to a point would make of the table. +#[derive(Clone, Debug, Default, PartialEq)] +pub struct Plan { + /// New column widths (points), when the columns change. + pub widths: Option>, + /// New heights (points) of these rows. + pub heights: Vec<(usize, f32)>, +} + +impl ResizeTarget { + pub fn left(&self) -> f32 { + self.cols.first().copied().unwrap_or(0.0) + } + pub fn right(&self) -> f32 { + self.cols.last().copied().unwrap_or(0.0) + } + pub fn top(&self) -> f32 { + self.rows.first().map_or(0.0, |r| r.top) + } + pub fn bottom(&self) -> f32 { + self.rows.last().map_or(0.0, |r| r.bottom) + } + /// The drawn column widths. + pub fn widths(&self) -> Vec { + self.cols.windows(2).map(|w| (w[1] - w[0]).max(0.0)).collect() + } + /// The resize handle's square (page points). + pub fn handle(&self) -> Rect { + Rect::new(self.right() + 1.0, self.bottom() + 1.0, HANDLE, HANDLE) + } + + /// The new widths and heights a drag that moved the pointer by (`dx`, `dy`) points gives. + pub fn plan(&self, dx: f32, dy: f32, shift: bool) -> Plan { + let (dx, dy) = + (if dx.is_finite() { dx.clamp(-MAX_LEN, MAX_LEN) } else { 0.0 }, if dy.is_finite() { dy.clamp(-MAX_LEN, MAX_LEN) } else { 0.0 }); + let mut w = self.widths(); + match self.edge { + Edge::Column(line) => { + let Some(left) = line.checked_sub(1).and_then(|i| w.get(i).copied()) else { return Plan::default() }; + let right = w.get(line).copied(); + let (new_left, new_right) = match right { + // An inner border: the neighbour on the right makes up the difference. + Some(right) if !shift => { + let d = dx.clamp(MIN_COL - left, (right - MIN_COL).max(MIN_COL - left)); + (left + d, Some(right - d)) + } + // The right edge (or Shift): only the column on the left changes. + _ => ((left + dx).clamp(MIN_COL, MAX_LEN), None), + }; + if let Some(x) = w.get_mut(line - 1) { + *x = new_left; + } + if let (Some(r), Some(x)) = (new_right, w.get_mut(line)) { + *x = r; + } + Plan { widths: Some(w), heights: Vec::new() } + } + Edge::Row(row) => { + let Some(r) = self.rows.iter().find(|r| r.row == row) else { return Plan::default() }; + Plan { widths: None, heights: vec![(row, (r.bottom - r.top + dy).clamp(MIN_ROW, MAX_LEN))] } + } + Edge::Corner => { + let (width, height) = (self.right() - self.left(), self.bottom() - self.top()); + let kx = if width > 1.0 { ((width + dx) / width).clamp(0.01, 100.0) } else { 1.0 }; + let ky = if height > 1.0 { ((height + dy) / height).clamp(0.01, 100.0) } else { 1.0 }; + let widths = ((kx - 1.0).abs() > 0.001).then(|| w.iter().map(|x| (x * kx).clamp(MIN_COL, MAX_LEN)).collect()); + let heights = if (ky - 1.0).abs() > 0.001 { + self.rows.iter().map(|r| (r.row, ((r.bottom - r.top) * ky).clamp(MIN_ROW, MAX_LEN))).collect() + } else { + Vec::new() + }; + Plan { widths, heights } + } + } + } +} + +/// A table cell's area on a page. +pub(crate) struct CellBox { + pub(crate) rect: Rect, + pub(crate) table: Path, + pub(crate) story: StoryRef, + pub(crate) row: usize, + pub(crate) cell: usize, +} + +pub(crate) fn cell_boxes(l: &DocLayout, page: usize) -> Vec { + let Some(pg) = l.pages.get(page) else { return Vec::new() }; + pg.items + .iter() + .filter_map(|it| match it { + Placed::Cell { rect, table, row, cell, story, .. } => { + Some(CellBox { rect: *rect, table: table.clone(), story: *story, row: *row, cell: *cell }) + } + _ => None, + }) + .collect() +} + +/// The drawn geometry of one table on a page, from its cells' boxes. +fn geometry(t: &Table, boxes: &[&CellBox]) -> Option<(Vec, Vec)> { + let n = t.cols(); + if n == 0 || n > wordcraft_doc::table::MAX_COLS || boxes.is_empty() { + return None; + } + let mut lines: Vec> = vec![None; n + 1]; + let mut rows: Vec = Vec::new(); + for b in boxes { + let g = t.grid_col(b.row, b.cell); + let span = t.rows.get(b.row).and_then(|r| r.cells.get(b.cell)).map_or(1, wordcraft_doc::Cell::span); + let end = g.saturating_add(span).min(n); + if let Some(x) = lines.get_mut(g) { + x.get_or_insert(b.rect.x); + } + if let Some(x) = lines.get_mut(end) { + x.get_or_insert(b.rect.right()); + } + // A vertically merged cell's box spans all its rows (#640), so a row ends at its + // shortest box. + match rows.iter_mut().find(|r| r.row == b.row) { + Some(r) => { + r.top = r.top.min(b.rect.y); + r.bottom = r.bottom.min(b.rect.bottom()); + r.edges.extend([g, end]); + } + None => rows.push(DrawnRow { row: b.row, top: b.rect.y, bottom: b.rect.bottom(), edges: vec![g, end] }), + } + } + // A line no cell on this page ends at: between its drawn neighbours, as the grid shares it. + let (first, last) = (lines.first().copied().flatten()?, lines.last().copied().flatten()?); + let weight = |i: usize| t.grid.get(i).copied().filter(|w| t.grid.len() == n && w.is_finite() && *w > 0.0).unwrap_or(1.0); + let mut cols = Vec::with_capacity(n + 1); + let mut prev = (0usize, first); + for i in 0..=n { + let x = match lines.get(i).copied().flatten() { + Some(x) => x, + None => { + let next = (i + 1..=n).find_map(|j| lines.get(j).copied().flatten().map(|x| (j, x))).unwrap_or((n, last)); + let span: f32 = (prev.0..next.0).map(weight).sum(); + let part: f32 = (prev.0..i).map(weight).sum(); + prev.1 + (next.1 - prev.1) * if span > 0.0 { part / span } else { 0.0 } + } + }; + let x = x.max(cols.last().copied().unwrap_or(x)); + if lines.get(i).copied().flatten().is_some() { + prev = (i, x); + } + cols.push(x); + } + rows.sort_by_key(|r| r.row); + // A row whose every cell starts a vertical merge has only boxes that span further: it ends + // where the next row starts. + let tops: Vec<(usize, f32)> = rows.iter().map(|r| (r.row, r.top)).collect(); + for (r, (next, top)) in rows.iter_mut().zip(tops.into_iter().skip(1)) { + if next == r.row.saturating_add(1) && top < r.bottom { + r.bottom = top.max(r.top); + } + } + Some((cols, rows)) +} + +/// Every table drawn on a page, with its geometry there (`edge` is a placeholder). +pub(crate) fn tables_on(l: &DocLayout, doc: &Document, page: usize) -> Vec { + let boxes = cell_boxes(l, page); + let mut tables: Vec<(StoryRef, &Path)> = Vec::new(); + for b in &boxes { + if !tables.iter().any(|(s, p)| *s == b.story && **p == b.table) { + tables.push((b.story, &b.table)); + } + } + tables + .into_iter() + .filter_map(|(story, path)| { + let t = doc.table(story, path)?; + let mine: Vec<&CellBox> = boxes.iter().filter(|b| b.story == story && b.table == *path).collect(); + let (cols, rows) = geometry(t, &mine)?; + let ends_here = rows.last().is_some_and(|r| r.row + 1 == t.rows.len()); + Some(ResizeTarget { edge: Edge::Corner, story, table: path.clone(), cols, rows, ends_here }) + }) + .collect() +} + +fn on_handle(t: &ResizeTarget, x: f32, y: f32, reach: f32) -> bool { + let h = t.handle(); + t.ends_here && x >= h.x - reach / 2.0 && x <= h.right() + reach / 2.0 && y >= h.y - reach / 2.0 && y <= h.bottom() + reach / 2.0 +} + +/// The innermost table under page point (`x`, `y`) whose resize handle is on this page, or whose +/// handle is under the point: the one whose handle the app shows. +pub fn table_under(l: &DocLayout, doc: &Document, page: usize, x: f32, y: f32, reach: f32) -> Option { + if !x.is_finite() || !y.is_finite() { + return None; + } + let reach = if reach.is_finite() { reach.clamp(0.5, 24.0) } else { REACH }; + tables_on(l, doc, page) + .into_iter() + .filter(|t| t.ends_here && ((x >= t.left() && x <= t.right() && y >= t.top() && y <= t.bottom()) || on_handle(t, x, y, reach))) + .max_by_key(|t| t.table.0.len()) +} + +/// The table border or resize handle within `reach` points of page point (`x`, `y`): the +/// nearest, and of nested tables the innermost. +pub fn resize_target(l: &DocLayout, doc: &Document, page: usize, x: f32, y: f32, reach: f32) -> Option { + if !x.is_finite() || !y.is_finite() { + return None; + } + let reach = if reach.is_finite() { reach.clamp(0.5, 24.0) } else { REACH }; + let mut best: Option<(f32, usize, ResizeTarget)> = None; + for mut target in tables_on(l, doc, page) { + let depth = target.table.0.len(); + let mut offer = |d: f32, edge: Edge, target: &mut ResizeTarget| { + if d <= reach && best.as_ref().is_none_or(|(bd, bdepth, _)| d < *bd - 0.01 || (d <= *bd + 0.01 && depth > *bdepth)) { + target.edge = edge; + best = Some((d, depth, target.clone())); + } + }; + if on_handle(&target, x, y, reach) { + offer(0.0, Edge::Corner, &mut target); + } + let (left, right, top, bottom) = (target.left(), target.right(), target.top(), target.bottom()); + if x < left - reach || x > right + reach || y < top - reach || y > bottom + reach { + continue; + } + // Column borders: in the row under the pointer, only lines that are cell borders there. + let row_here = target.rows.iter().find(|r| y >= r.top && y <= r.bottom).cloned(); + if let Some(r) = &row_here { + for line in 1..target.cols.len() { + if r.edges.contains(&line) + && let Some(cx) = target.cols.get(line).copied() + { + offer((x - cx).abs(), Edge::Column(line), &mut target); + } + } + } + // Row borders: the bottom of each drawn row. + if x >= left && x <= right { + for r in target.rows.clone() { + offer((y - r.bottom).abs(), Edge::Row(r.row), &mut target); + } + } + } + best.map(|b| b.2) +} + +fn resize_drag(s: &mut Session, v: &Value) -> CmdResult { + let page = + p::u64(v, "page").map(|n| usize::try_from(n).unwrap_or(usize::MAX)).ok_or_else(|| CmdError::Params("`page` (number) is required".into()))?; + let (x, y) = (p::req_f32(v, "x")?, p::req_f32(v, "y")?); + let (to_x, to_y) = (p::req_f32(v, "toX")?, p::req_f32(v, "toY")?); + let shift = p::bool(v, "shift").unwrap_or(false); + let reach = p::f32(v, "reach").unwrap_or(REACH); + let l = s.layout(); + if l.pages.get(page).is_none() { + return Err(CmdError::Params(format!("no page {page}"))); + } + let target = + resize_target(&l, &s.doc, page, x, y, reach).ok_or_else(|| CmdError::Failed("there's no table border or resize handle there".into()))?; + let plan = target.plan(to_x - x, to_y - y, shift); + if plan.widths.is_none() && plan.heights.is_empty() { + return sel_result(s); + } + let t = s.doc.table_mut(target.story, &target.table)?; + if let Some(widths) = plan.widths { + set_widths(t, widths); + } + for (r, h) in &plan.heights { + if let Some(row) = t.rows.get_mut(*r) { + row.props.height = Some(*h); + if row.props.height_rule != HeightRule::Exact { + row.props.height_rule = HeightRule::AtLeast; + } + } + } + touch(s, &target)?; + if plan.heights.is_empty() { + return sel_result(s); + } + // An "at least" height leaves out the cells' margins and border bands, which the layout adds + // (and which depend on the table style): measure what the rows came to and take the extra + // off once. A row its text holds taller keeps the dragged height. + let drawn = drawn_heights(&s.layout(), &s.doc, page, &target); + let mut fixed = Vec::new(); + let t = s.doc.table_mut(target.story, &target.table)?; + for (r, want) in &plan.heights { + let Some(got) = drawn.iter().find(|(i, _)| i == r).map(|(_, h)| *h) else { continue }; + if let Some(row) = t.rows.get_mut(*r) + && row.props.height_rule == HeightRule::AtLeast + && got > want + 0.05 + { + row.props.height = Some((want - (got - want)).max(1.0)); + fixed.push((*r, *want, got)); + } + } + if fixed.is_empty() { + return sel_result(s); + } + touch(s, &target)?; + let again = drawn_heights(&s.layout(), &s.doc, page, &target); + let t = s.doc.table_mut(target.story, &target.table)?; + for (r, want, got) in fixed { + let now = again.iter().find(|(i, _)| *i == r).map_or(got, |(_, h)| *h); + if now >= got - 0.05 + && let Some(row) = t.rows.get_mut(r) + { + row.props.height = Some(want); + } + } + touch(s, &target)?; + sel_result(s) +} + +/// Make `widths` (one per grid column) the table's column widths, as drawn: no preferred table +/// width scales them again. +pub(crate) fn set_widths(t: &mut Table, widths: Vec) { + if widths.len() == t.cols() { + t.grid = widths; + t.props.width = None; + t.props.width_pct = None; + super::table::sync_cell_widths(t); + } +} + +/// The drawn height of each of the target table's rows on `page`. +fn drawn_heights(l: &DocLayout, doc: &Document, page: usize, target: &ResizeTarget) -> Vec<(usize, f32)> { + tables_on(l, doc, page) + .into_iter() + .find(|t| t.story == target.story && t.table == target.table) + .map(|t| t.rows.iter().map(|r| (r.row, r.bottom - r.top)).collect()) + .unwrap_or_default() +} + +/// Mark the table's paragraphs changed, so the layout redoes them. +pub(crate) fn touch(s: &mut Session, target: &ResizeTarget) -> Result<(), CmdError> { + let story = target.story; + let paths: Vec = + s.doc.para_paths(story).into_iter().filter(|q| q.0.len() > target.table.0.len() && q.0.starts_with(&target.table.0)).collect(); + for q in paths { + s.doc.para_mut(story, &q)?.touch(); + } + // A new document revision: the next `layout()` lays the change out instead of reusing the + // cached pages. + s.touch(); + Ok(()) +} + +#[cfg(test)] +mod tests { + use super::*; + use serde_json::json; + + fn first_table(s: &Session) -> Table { + s.doc.body.iter().find_map(|b| b.as_table()).cloned().unwrap() + } + + fn cell_rect(s: &mut Session, row: usize, cell: usize) -> Rect { + s.layout().pages[0] + .items + .iter() + .find_map(|it| match it { + Placed::Cell { rect, row: r, cell: c, .. } if *r == row && *c == cell => Some(*rect), + _ => None, + }) + .unwrap() + } + + fn table_session(rows: usize, cols: usize) -> Session { + let mut s = Session::new(wordcraft_doc::Document::new()); + s.run("insert.table", &json!({"rows": rows, "cols": cols})).unwrap(); + s + } + + /// Dragging the border between two columns moves it; the right neighbour gives up the room, + /// so the table keeps its width; one undo puts it back. + #[test] + fn a_column_border_drag_trades_width_with_the_neighbour() { + let mut s = table_session(2, 3); + let start = first_table(&s); + let a = cell_rect(&mut s, 0, 0); + let (b0, total0) = (cell_rect(&mut s, 0, 1), cell_rect(&mut s, 0, 2).right() - a.x); + let y = a.y + a.h / 2.0; + s.run("table.resizeDrag", &json!({"page": 0, "x": a.right() + 1.0, "y": y, "toX": a.right() + 31.0, "toY": y})).unwrap(); + let (a1, b1) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1)); + assert!((a1.w - (a.w + 30.0)).abs() < 1.0, "{a:?} → {a1:?}"); + assert!((b1.w - (b0.w - 30.0)).abs() < 1.0, "{b0:?} → {b1:?}"); + let total1 = cell_rect(&mut s, 0, 2).right() - a1.x; + assert!((total1 - total0).abs() < 1.0, "the table keeps its width"); + // The second row follows: one border for the whole column. + assert!((cell_rect(&mut s, 1, 0).w - a1.w).abs() < 0.5); + s.run("edit.undo", &json!({})).unwrap(); + assert_eq!(first_table(&s), start, "one undo step"); + } + + /// With Shift only the left column changes and the table grows; the right edge does the same. + #[test] + fn shift_and_the_right_edge_change_the_table_width() { + let mut s = table_session(1, 2); + let (a, b) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1)); + let y = a.y + a.h / 2.0; + s.run("table.resizeDrag", &json!({"page": 0, "x": a.right(), "y": y, "toX": a.right() - 40.0, "toY": y, "shift": true})).unwrap(); + let (a1, b1) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1)); + assert!((a1.w - (a.w - 40.0)).abs() < 1.0 && (b1.w - b.w).abs() < 1.0, "{a1:?} {b1:?}"); + s.run("table.resizeDrag", &json!({"page": 0, "x": b1.right() - 1.0, "y": y, "toX": b1.right() - 51.0, "toY": y})).unwrap(); + let b2 = cell_rect(&mut s, 0, 1); + assert!((b2.w - (b1.w - 50.0)).abs() < 1.0, "{b1:?} → {b2:?}"); + } + + /// No column gets narrower than the minimum, however far the drag goes. + #[test] + fn column_drags_stop_at_the_minimum_width() { + let mut s = table_session(1, 2); + let (a, b) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1)); + let y = a.y + a.h / 2.0; + s.run("table.resizeDrag", &json!({"page": 0, "x": a.right(), "y": y, "toX": a.right() + 5000.0, "toY": y})).unwrap(); + let t = first_table(&s); + assert!((t.grid[1] - MIN_COL).abs() < 0.01, "{:?}", t.grid); + assert!((t.grid[0] + t.grid[1] - (a.w + b.w)).abs() < 1.0); + } + + /// Dragging a row's bottom border sets its height ("at least"), and the row grows. + #[test] + fn a_row_border_drag_sets_the_row_height() { + let mut s = table_session(2, 2); + let a = cell_rect(&mut s, 0, 0); + let x = a.x + a.w / 2.0; + s.run("table.resizeDrag", &json!({"page": 0, "x": x, "y": a.bottom() + 1.0, "toX": x, "toY": a.bottom() + 41.0})).unwrap(); + let t = first_table(&s); + assert_eq!(t.rows[0].props.height_rule, HeightRule::AtLeast); + // The stored height leaves out the margins and borders; the drawn row is what was dragged. + assert!(t.rows[0].props.height.unwrap() < a.h + 40.0, "{:?}", t.rows[0].props); + let h = cell_rect(&mut s, 0, 0).h; + assert!((h - (a.h + 40.0)).abs() < 0.1, "{} → {h}", a.h); + assert_eq!(t.rows[1].props.height, None, "the other row is untouched"); + } + + /// The handle past the bottom-right corner scales every column and row. + #[test] + fn the_corner_handle_scales_the_table() { + let mut s = table_session(2, 2); + let (a, d) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 1, 1)); + let (w, h) = (d.right() - a.x, d.bottom() - a.y); + let (hx, hy) = (d.right() + 1.0 + HANDLE / 2.0, d.bottom() + 1.0 + HANDLE / 2.0); + let l = s.layout(); + let tgt = resize_target(&l, &s.doc, 0, hx, hy, REACH).unwrap(); + assert_eq!(tgt.edge, Edge::Corner); + // Half as wide, twice as tall: the handle moves by that much. + s.run("table.resizeDrag", &json!({"page": 0, "x": hx, "y": hy, "toX": hx - w * 0.5, "toY": hy + h})).unwrap(); + let (a1, d1) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 1, 1)); + assert!(((d1.right() - a1.x) - w * 0.5).abs() < 1.5, "{w} → {}", d1.right() - a1.x); + assert!(((d1.bottom() - a1.y) - h * 2.0).abs() < 0.5, "{h} → {}", d1.bottom() - a1.y); + } + + /// A merged cell hides the border it covers: no column border to grab in its middle. + #[test] + fn merged_cells_have_no_border_inside() { + let mut s = table_session(2, 2); + let a = cell_rect(&mut s, 0, 0); + s.run("table.eraser", &json!({"page": 0, "x": a.right(), "y": a.y + a.h / 2.0})).unwrap(); + let l = s.layout(); + let m = cell_rect(&mut s, 0, 0); + let mid = resize_target(&l, &s.doc, 0, a.right(), m.y + m.h / 2.0, REACH); + assert!(mid.as_ref().is_none_or(|t| t.edge != Edge::Column(1)), "{mid:?}"); + // The second row still has it. + let b = cell_rect(&mut s, 1, 0); + let t = resize_target(&l, &s.doc, 0, b.right(), b.y + b.h / 2.0, REACH).unwrap(); + assert_eq!(t.edge, Edge::Column(1)); + } + + /// A vertical merge lays out as one box as tall as its rows (#640); each row keeps its own + /// bottom border, and dragging one changes that row alone. + #[test] + fn a_vertical_merge_keeps_each_rows_border() { + let mut s = table_session(3, 2); + let (m, a, b) = (cell_rect(&mut s, 0, 0), cell_rect(&mut s, 0, 1), cell_rect(&mut s, 1, 1)); + s.run("table.eraser", &json!({"page": 0, "x": m.x + m.w / 2.0, "y": m.bottom() + 0.5})).unwrap(); + assert_eq!(first_table(&s).rows[1].cells[0].props.vmerge, wordcraft_doc::props::VMerge::Continue); + let l = s.layout(); + let x = b.x + b.w / 2.0; + let t = resize_target(&l, &s.doc, 0, x, a.bottom() + 1.0, REACH).unwrap(); + assert_eq!(t.edge, Edge::Row(0), "the border between the merged rows is still there"); + let t = resize_target(&l, &s.doc, 0, x, b.bottom() + 1.0, REACH).unwrap(); + assert_eq!(t.edge, Edge::Row(1), "the merge's bottom is the second row's border"); + s.run("table.resizeDrag", &json!({"page": 0, "x": x, "y": a.bottom() + 1.0, "toX": x, "toY": a.bottom() + 21.0})).unwrap(); + let t = first_table(&s); + let h = cell_rect(&mut s, 0, 1).h; + assert!((h - (a.h + 20.0)).abs() < 0.1, "{} → {h}", a.h); + assert_eq!(t.rows[1].props.height, None, "the other rows are untouched"); + assert!((cell_rect(&mut s, 1, 1).h - b.h).abs() < 0.1); + } + + /// Away from any border, or with hostile numbers: an error and no change, never a panic. + #[test] + fn misses_and_hostile_params_change_nothing() { + let mut s = table_session(2, 2); + let start = first_table(&s); + let c = cell_rect(&mut s, 0, 0); + let mid = json!({"page": 0, "x": c.x + c.w / 2.0, "y": c.y + c.h / 2.0, "toX": 10, "toY": 10}); + assert!(s.run("table.resizeDrag", &mid).is_err()); + assert!(s.run("table.resizeDrag", &json!({"page": 0, "x": 1e30, "y": -1e30, "toX": 1e30, "toY": 1e30})).is_err()); + assert!(s.run("table.resizeDrag", &json!({"page": 99, "x": 1, "y": 1, "toX": 1, "toY": 1})).is_err()); + assert!(s.run("table.resizeDrag", &json!({})).is_err()); + let y = c.y + c.h / 2.0; + s.run("table.resizeDrag", &json!({"page": 0, "x": c.right(), "y": y, "toX": f32::MAX, "toY": y, "reach": 1e9})).unwrap(); + assert!(first_table(&s).grid.iter().all(|w| w.is_finite() && *w >= MIN_COL)); + s.run("edit.undo", &json!({})).unwrap(); + assert_eq!(first_table(&s), start); + } +} diff --git a/crates/engine/src/cmd/text.rs b/crates/engine/src/cmd/text.rs index 3eb42a7bc..c962cc9f1 100644 --- a/crates/engine/src/cmd/text.rs +++ b/crates/engine/src/cmd/text.rs @@ -509,8 +509,18 @@ fn backspace_inner(s: &mut Session, _: &Value) -> CmdResult { } // At the start of a paragraph. if para.props.numbering.is_some_and(|n| n.num != 0) { - // First Backspace removes the number but keeps the item's indent (Word). + // First Backspace removes the number but keeps the item's indent (Word): the text stays + // where it was, so the indent becomes the paragraph's own (`num: 0` alone drops it). + let text_at = para.props.numbering.and_then(|n| s.doc.numbering.level(n.num, n.level)).map(|l| l.indent); let p = s.doc.para_mut(f.story, &f.path)?; + if p.props.style.as_deref() == Some("ListParagraph") { + if p.props.indent_left.is_none() && p.props.indent_left_chars.is_none_or(|c| c == 0.0) { + p.props.indent_left = text_at; + } + if p.props.indent_first.is_none() && p.props.indent_first_chars.is_none_or(|c| c == 0.0) { + p.props.indent_first = Some(0.0); + } + } p.props.numbering = Some(NumRef { num: 0, level: 0 }); p.touch(); return sel_result(s); diff --git a/crates/engine/src/cmd/tools.rs b/crates/engine/src/cmd/tools.rs index 6ad3fc46c..31d8a18cf 100644 --- a/crates/engine/src/cmd/tools.rs +++ b/crates/engine/src/cmd/tools.rs @@ -60,6 +60,11 @@ pub const AUTOCORRECT: &[(&str, &str)] = &[ ("3/4", "¾"), ]; +/// A built-in entry that is an English word (letters and apostrophes only), not a symbol. +fn is_english_entry(from: &str) -> bool { + from.chars().any(char::is_alphabetic) && from.chars().all(|c| c.is_alphabetic() || c == '\'') +} + /// Compare and Combine take the same settings. const COMPARE_PARAMS: &str = r#"{"path"?: string (revised document), "text"?: string (revised text), "original"?: string (original document; default: this one), "author"?: string (label changes with), "caseChanges"?: bool, "whiteSpace"?: bool, "formatting"?: bool (default true: differences count), "level"?: "word|character", "showIn"?: "original|revised|new" (revised/new: an untitled document), "moves"|"comments"|"tables"|"headersFooters"|"footnotes"|"textBoxes"|"fields"?: bool (accepted; not honoured yet except tables, always compared)}"#; @@ -71,12 +76,16 @@ pub struct AutoCorrectPrefs { pub enabled: bool, /// Replace text as you type (the replacement list). pub replace_text: bool, + /// Correct TWo INitial CApitals: "WOrd" → "Word" (not the exceptions, e.g. "IDs"). + pub two_initial_caps: bool, /// Capitalize the first letter of sentences. pub cap_sentences: bool, /// Capitalize the first letter of table cells. pub cap_cells: bool, /// Capitalize names of days. pub cap_days: bool, + /// Correct accidental usage of cAPS LOCK: "wORD" → "Word". + pub caps_lock: bool, /// AutoFormat As You Type: "straight quotes" with “smart quotes”. pub smart_quotes: bool, /// Fractions (1/2) with fraction characters (½). @@ -101,6 +110,8 @@ pub struct AutoCorrectPrefs { pub exceptions: Vec, /// Built-in abbreviations the user deleted from the exceptions. pub exceptions_removed: Vec, + /// The user's words that keep two initial capitals (besides the built-in ones). + pub initial_caps_exceptions: Vec, } impl Default for AutoCorrectPrefs { @@ -108,9 +119,11 @@ impl Default for AutoCorrectPrefs { AutoCorrectPrefs { enabled: true, replace_text: true, + two_initial_caps: true, cap_sentences: true, cap_cells: true, cap_days: true, + caps_lock: true, smart_quotes: true, fractions: true, ordinals: true, @@ -123,6 +136,7 @@ impl Default for AutoCorrectPrefs { removed: Vec::new(), exceptions: Vec::new(), exceptions_removed: Vec::new(), + initial_caps_exceptions: Vec::new(), } } } @@ -145,7 +159,7 @@ impl AutoCorrectPrefs { let mut seen = std::collections::HashSet::new(); self.entries.retain(|(a, b)| ok(a, MAX_FROM) && b.chars().count() <= MAX_TO && !b.contains(['\n', '\r']) && seen.insert(a.clone())); self.entries.truncate(MAX_ENTRIES); - for list in [&mut self.removed, &mut self.exceptions, &mut self.exceptions_removed] { + for list in [&mut self.removed, &mut self.exceptions, &mut self.exceptions_removed, &mut self.initial_caps_exceptions] { let mut seen = std::collections::HashSet::new(); list.retain(|a| ok(a, MAX_FROM) && seen.insert(a.clone())); list.truncate(MAX_ENTRIES); @@ -153,14 +167,23 @@ impl AutoCorrectPrefs { self } - /// The replacement for `word`, if any: the user's entry, else a built-in one not deleted. - /// Built-in lowercase entries also match a capitalised word ("Teh" → "The"). + /// The replacement for `word` in English text, if any: the user's entry, else a built-in + /// one not deleted. Built-in lowercase entries also match a capitalised word ("Teh" → "The"). pub fn replacement(&self, word: &str) -> Option { + self.replacement_in(word, true) + } + + /// The replacement for `word` in text that is (`english`) or isn't English: the built-in + /// English words ("teh", "i", "dont") apply only to English text, as Word keeps a list per + /// language (#835: Polish "i" means "and"); symbols, fractions and the user's entries apply + /// whatever the language. + pub fn replacement_in(&self, word: &str, english: bool) -> Option { if let Some((_, b)) = self.entries.iter().find(|(a, _)| a == word) { return Some(b.clone()); } AUTOCORRECT .iter() + .filter(|(a, _)| english || !is_english_entry(a)) .filter(|(a, _)| !self.removed.iter().any(|r| r == a) && !self.entries.iter().any(|(e, _)| e == a)) .filter(|(a, _)| self.fractions || !FRACTIONS.contains(a)) .find(|(a, _)| { @@ -216,9 +239,11 @@ fn autocorrect_options(s: &mut Session, v: &Value) -> CmdResult { for (k, f) in [ ("enabled", &mut ac.enabled as &mut bool), ("replaceText", &mut ac.replace_text), + ("twoInitialCaps", &mut ac.two_initial_caps), ("capSentences", &mut ac.cap_sentences), ("capCells", &mut ac.cap_cells), ("capDays", &mut ac.cap_days), + ("capsLock", &mut ac.caps_lock), ("smartQuotes", &mut ac.smart_quotes), ("fractions", &mut ac.fractions), ("ordinals", &mut ac.ordinals), @@ -273,13 +298,31 @@ fn autocorrect_options(s: &mut Session, v: &Value) -> CmdResult { ac.exceptions_removed.push(e); } } + if let Some(w) = p::str(v, "addInitialCapsException") { + let w = w.trim(); + if bad(w, MAX_FROM) || w.contains(char::is_whitespace) { + return Err(CmdError::Params("an INitial CAps exception is one word, like \"IDs\"".into())); + } + if !INITIAL_CAPS_EXCEPTIONS.contains(&w) + && !ac.initial_caps_exceptions.iter().any(|e| e == w) + && ac.initial_caps_exceptions.len() < MAX_ENTRIES + { + ac.initial_caps_exceptions.push(w.to_string()); + } + } + if let Some(w) = p::str(v, "deleteInitialCapsException") { + let w = w.trim(); + ac.initial_caps_exceptions.retain(|e| e != w); + } let ac = &s.prefs.autocorrect; Ok(json!({ "enabled": ac.enabled, "replaceText": ac.replace_text, + "twoInitialCaps": ac.two_initial_caps, "capSentences": ac.cap_sentences, "capCells": ac.cap_cells, "capDays": ac.cap_days, + "capsLock": ac.caps_lock, "smartQuotes": ac.smart_quotes, "fractions": ac.fractions, "ordinals": ac.ordinals, @@ -291,6 +334,7 @@ fn autocorrect_options(s: &mut Session, v: &Value) -> CmdResult { "entries": ac.list().into_iter().map(|(a, b, builtin)| json!({"from": a, "to": b, "builtIn": builtin})).collect::>(), "user": ac.entries, "exceptions": ac.exception_list(), + "initialCapsExceptions": INITIAL_CAPS_EXCEPTIONS.iter().map(|e| e.to_string()).chain(ac.initial_caps_exceptions.iter().cloned()).collect::>(), })) } @@ -321,7 +365,7 @@ pub fn specs() -> Vec { .pure(), CommandSpec::new("tools.autocorrect", "AutoCorrect Options", "File › Options › Proofing", autocorrect_options) .params( - r#"{"enabled"?: bool (all of AutoCorrect), "replaceText"|"capSentences"|"capCells"|"capDays"|"smartQuotes"|"fractions"|"ordinals"|"dashes"|"links"|"bullets"|"numbering"|"borderLines"?: bool, "add"?: {"from": string, "to": string} (adds or replaces an entry), "delete"?: string (an entry's "from", built in or added), "addException"?: string, "deleteException"?: string} → options, entries and exceptions"#, + r#"{"enabled"?: bool (all of AutoCorrect), "replaceText"|"twoInitialCaps"|"capSentences"|"capCells"|"capDays"|"capsLock"|"smartQuotes"|"fractions"|"ordinals"|"dashes"|"links"|"bullets"|"numbering"|"borderLines"?: bool, "add"?: {"from": string, "to": string} (adds or replaces an entry), "delete"?: string (an entry's "from", built in or added), "addException"?: string, "deleteException"?: string, "addInitialCapsException"?: string (keeps TWo INitial CApitals, like "IDs"), "deleteInitialCapsException"?: string} → options, entries and exceptions"#, ) .pure(), CommandSpec::new("tools.customizeKeyboard", "Customize Keyboard", "File › Options", customize_keyboard) @@ -577,7 +621,7 @@ pub fn specs() -> Vec { }) .params(r#"{"format": "decimal|upperRoman|lowerLetter…", "text"?: "%1.", "start"?: n}"#), CommandSpec::new("review.language", "Language", "Review › Language", language).params( - r#"{"lang"?: "en-US|en-GB|fr-FR|…", "noProof"?: bool, "detect"?: bool (detect language automatically, a preference), "default"?: bool (make `lang` the document's default language instead)}"#, + r#"{"lang"?: "en-US|en-GB|pl-PL|fr-FR|…" (BCP 47; default en-US), "noProof"?: bool, "detect"?: bool (detect language automatically, a preference), "default"?: bool (also the document's default language)}"#, ), CommandSpec::new("review.showMarkup", "Show Markup", "Review › Tracking", |s, v| { s.view.show_markup = p::bool(v, "value").unwrap_or(!s.view.show_markup); @@ -621,6 +665,7 @@ pub fn specs() -> Vec { .pure(), CommandSpec::new("view.sideToSide", "Side to Side", "View › Page Movement", |s, _| { s.view.multi_page = true; + s.view.single_column = false; s.view.fit = "multiplePages".into(); Ok(json!({"pageMovement": "sideToSide"})) }) @@ -633,8 +678,10 @@ pub fn specs() -> Vec { /// Nothing at all sets US English, as before. fn language(s: &mut Session, v: &Value) -> CmdResult { let lang = match p::str(v, "lang") { - Some(l) if !l.is_empty() && l.len() <= 35 && l.chars().all(|c| c.is_ascii_alphanumeric() || c == '-') => Some(l.to_string()), - Some(l) => return Err(CmdError::Params(format!("`{l}` is not a language tag"))), + Some(t) => Some( + wordcraft_proof::lang::normalize_tag(t) + .ok_or_else(|| CmdError::Params(format!("`lang` must be a language tag such as pl-PL or en-US, not {t:?}")))?, + ), None => None, }; let no_proof = p::bool(v, "noProof"); @@ -643,15 +690,14 @@ fn language(s: &mut Session, v: &Value) -> CmdResult { s.prefs.detect_language = d; } if p::bool(v, "default") == Some(true) { - let l = lang.ok_or_else(|| CmdError::Params("`lang` is required with `default`".into()))?; + let l = lang.clone().ok_or_else(|| CmdError::Params("`default` needs `lang`".into()))?; s.doc.styles.default_chr.lang = Some(l.clone()); // A Normal style that names its own language follows the new default. if let Some(n) = s.doc.styles.styles.iter_mut().find(|x| x.id == "Normal") && n.chr.lang.is_some() { - n.chr.lang = Some(l.clone()); + n.chr.lang = Some(l); } - return Ok(json!({"default": l, "detect": s.prefs.detect_language})); } let lang = match (lang, no_proof, detect) { (None, None, None) => Some("en-US".to_string()), @@ -713,7 +759,7 @@ pub fn autocorrect_word(s: &mut Session, on_enter: bool) -> Result<(), CmdError> return Ok(()); } let ac = s.prefs.autocorrect.clone(); - let f = s.sel.focus.clone(); + let mut f = s.sel.focus.clone(); let Some(text) = s.doc.para_at(&f).map(|p| p.text.clone()) else { return Ok(()) }; let Some(before) = text.get(..f.off) else { return Ok(()) }; let Some(last) = before.chars().next_back() else { return Ok(()) }; @@ -723,12 +769,39 @@ pub fn autocorrect_word(s: &mut Session, on_enter: bool) -> Result<(), CmdError> let trigger = if on_enter { 0 } else { last.len_utf8() }; let Some(body) = before.get(..before.len() - trigger) else { return Ok(()) }; let start = body.rfind(char::is_whitespace).map(|i| i + 1).unwrap_or(0); - let Some(word) = body.get(start..).map(str::to_string) else { return Ok(()) }; + let Some(mut word) = body.get(start..).map(str::to_string) else { return Ok(()) }; if word.is_empty() { return Ok(()); } + // TWo INitial CApitals and cAPS LOCK come first, so "TEh" is still a replacement's typo. + if let Some((at, reps)) = case_fix(&ac, &word) { + let para = s.doc.para_mut(f.story, &f.path)?; + let letters = word.get(at..).unwrap_or("").chars().take(reps.len()).collect::(); + let mut end = start + at + letters.len(); + for (c, rep) in letters.chars().rev().zip(reps.iter().rev()) { + let from = end.saturating_sub(c.len_utf8()); + if rep.chars().ne(std::iter::once(c)) { + let props = para.props_of_char(from).clone(); + para.delete(from, end)?; + para.insert_text(from, rep, &props)?; + } + end = from; + } + let fixed: String = reps.concat(); + f.off = (f.off + fixed.len()).saturating_sub(letters.len()); + s.sel = Selection::caret(f.clone()); + word = format!("{}{fixed}{}", word.get(..at).unwrap_or(""), word.get(at + letters.len()..).unwrap_or("")); + } // A whole-word replacement, else "wait..." → "wait…". - let found = if ac.replace_text { ac.replacement(&word) } else { None }; + // The word's language (run, paragraph style, document default) picks the list (#835). + // Text with no language at all is in the editing language (a Polish interface writes Polish), + // else English. Proofing decides what is English the same way. + let english = s.doc.para_at(&f).is_none_or(|p| { + let rc = s.doc.styles.resolve_char(p.props.style.as_deref(), p.props_of_char(start)); + let lang = rc.lang.as_deref().or(s.editing_lang.as_deref()); + wordcraft_proof::ProofLang::from_tag(lang) == wordcraft_proof::ProofLang::En + }); + let found = if ac.replace_text { ac.replacement_in(&word, english) } else { None }; let (from, rep) = match found { Some(r) => (start, Some(r)), None if ac.replace_text && word.len() > 3 && word.ends_with("...") => (start + word.len() - 3, Some("…".to_string())), @@ -803,6 +876,49 @@ pub fn autocorrect_word(s: &mut Session, on_enter: bool) -> Result<(), CmdError> Ok(()) } +/// Built-in words that keep their two initial capitals (plurals of two-letter abbreviations). +pub const INITIAL_CAPS_EXCEPTIONS: &[&str] = &["CDs", "IDs", "PCs", "TVs", "DJs", "GBs", "KBs", "MBs", "TBs", "VMs"]; + +/// Words with a lowercase first letter and the rest in capitals that aren't cAPS LOCK slips. +const CAPS_LOCK_EXCEPTIONS: &[&str] = &["iOS", "mRNA", "tRNA", "rRNA", "cDNA", "eSIM", "eSATA", "iSCSI", "vCPU"]; + +/// The capitalisation fix for a finished word: TWo INitial CApitals ("WOrd" → "Word") or cAPS +/// LOCK ("wORD" → "Word"). Returns the byte offset of the word's letters and each letter's +/// replacement. Only words of three or more letters, optionally in quotes or brackets. +fn case_fix(caps: &AutoCorrectPrefs, word: &str) -> Option<(usize, Vec)> { + if !caps.two_initial_caps && !caps.caps_lock { + return None; + } + let at = word.find(char::is_alphabetic)?; + let rest = word.get(at..)?; + let letters = rest.get(..rest.find(|c: char| !c.is_alphabetic()).unwrap_or(rest.len()))?; + let tail = rest.get(letters.len()..)?; + if !word.get(..at)?.chars().all(|c| "\"'“‘([".contains(c)) || !tail.chars().all(|c| "\"'”’)]".contains(c)) { + return None; + } + let cs: Vec = letters.chars().collect(); + let (&a, &b, more) = match cs.as_slice() { + [a, b, more @ ..] if !more.is_empty() => (a, b, more), + _ => return None, + }; + let one = |c: char| c.to_string(); + let lower = |c: char| c.to_lowercase().collect::(); + if caps.two_initial_caps + && a.is_uppercase() + && b.is_uppercase() + && more.iter().all(|c| c.is_lowercase()) + && !INITIAL_CAPS_EXCEPTIONS.contains(&letters) + && !caps.initial_caps_exceptions.iter().any(|e| e == letters) + { + return Some((at, std::iter::once(one(a)).chain(std::iter::once(lower(b))).chain(more.iter().map(|&c| one(c))).collect())); + } + if caps.caps_lock && a.is_lowercase() && b.is_uppercase() && more.iter().all(|c| c.is_uppercase()) && !CAPS_LOCK_EXCEPTIONS.contains(&letters) { + let up = a.to_uppercase().collect::(); + return Some((at, std::iter::once(up).chain(std::iter::once(lower(b))).chain(more.iter().map(|&c| lower(c))).collect())); + } + None +} + /// Byte length of an ordinal suffix to superscript ("1st" → 2), if `word` is a correct ordinal. fn ordinal_suffix(word: &str) -> Option { let digits = word.find(|c: char| !c.is_ascii_digit())?; diff --git a/crates/engine/src/cmd/view.rs b/crates/engine/src/cmd/view.rs index 1e2e0cfce..da156c596 100644 --- a/crates/engine/src/cmd/view.rs +++ b/crates/engine/src/cmd/view.rs @@ -1,7 +1,7 @@ //! View tab and status bar: views, zoom, show/hide, panes. use serde_json::{Value, json}; -use wordcraft_layout::ViewMode; +use wordcraft_layout::{NumeralMode, ViewMode}; use super::sel_result; use crate::{CmdError, CmdResult, CommandSpec, Session, p}; @@ -27,6 +27,15 @@ pub fn specs() -> Vec { // The body as an outline of its headings, with the Outlining tab (`outline.*`). CommandSpec::new("view.outline", "Outline", "View › Views", |s, _| mode(s, ViewMode::Outline, false)).key("Mod+Alt+O").pure(), CommandSpec::new("view.readMode", "Read Mode", "View › Views", |s, _| mode(s, ViewMode::Print, true)).pure(), + CommandSpec::new("view.readColumnWidth", "Reading Column Width", "Reading › Layout", read_column_width) + .params(r#"{"value": "auto" | "narrow" | "medium" | "wide" | "page"}"#) + .pure(), + CommandSpec::new("view.readTextSize", "Reading Text Size", "Reading › Layout", read_text_size) + .params(r#"{"value": percent (75–150)}"#) + .pure(), + CommandSpec::new("view.readLineSpacing", "Reading Line Spacing", "Reading › Layout", read_line_spacing) + .params(r#"{"value": "document" | 1.0 | 1.25 | 1.5 | 2.0}"#) + .pure(), // Print Layout: hide the white space (margins, headers, footers) between pages. Screen only. CommandSpec::new("view.hideWhiteSpace", "Hide White Space", "View › Show", |s, v| toggle(s, v, |x| &mut x.hide_white_space)) .params(r#"{"value"?: bool}"#) @@ -83,21 +92,79 @@ pub fn specs() -> Vec { .pure(), CommandSpec::new("view.state", "View State", "View", |s, _| serde_json::to_value(&s.view).map_err(|e| CmdError::Failed(e.to_string()))) .pure(), + // Word › File › Options › Advanced › Numeral: how digits look on screen (display + // only; the document text keeps the typed codepoints). + CommandSpec::new("view.numeral", "Numerals", "File › Options › Advanced", |s, v| match v.get("value") { + None => Ok(json!({"value": s.view.numeral})), + Some(Value::String(name)) => { + let Some(mode) = NumeralMode::parse(name) else { + return Err(CmdError::Params("numeral mode: arabic|hindi|context|system".into())); + }; + s.view.numeral = mode; + s.relayout(); + Ok(json!({"value": s.view.numeral})) + } + Some(_) => Err(CmdError::Params("numeral mode: arabic|hindi|context|system".into())), + }) + .params(r#"{"value"?: "arabic|hindi|context|system"}"#) + .pure(), ] } fn mode(s: &mut Session, m: ViewMode, read: bool) -> CmdResult { + if read && !s.view.read_mode { + // Reading controls are session-only, and reset on each entry to the view. + s.view.read_column_width = None; + s.view.read_zoom = 1.0; + s.view.read_line_spacing = None; + } s.view.mode = m; s.view.read_mode = read; s.relayout(); Ok(json!({"mode": m, "readMode": read})) } +fn read_column_width(s: &mut Session, v: &Value) -> CmdResult { + let width = match v.get("value").and_then(Value::as_str) { + Some("auto") => None, + Some("narrow") => Some(180.0), + Some("medium") => Some(240.0), + Some("wide") => Some(350.0), + Some("page") => Some(612.0), + _ => return Err(CmdError::Params("column width must be auto, narrow, medium, wide, or page".into())), + }; + s.view.read_column_width = width; + Ok(json!({"width": width})) +} + +fn read_text_size(s: &mut Session, v: &Value) -> CmdResult { + let Some(pct) = v.get("value").and_then(Value::as_f64) else { + return Err(CmdError::Params("reading text size must be a percentage".into())); + }; + if !pct.is_finite() || !(75.0..=150.0).contains(&pct) { + return Err(CmdError::Params("reading text size must be 75–150%".into())); + } + s.view.read_zoom = pct as f32 / 100.0; + Ok(json!({"percent": pct})) +} + +fn read_line_spacing(s: &mut Session, v: &Value) -> CmdResult { + let value = v.get("value"); + let spacing = match value { + Some(Value::String(s)) if s == "document" => None, + Some(Value::Number(n)) if matches!(n.as_f64(), Some(1.0 | 1.25 | 1.5 | 2.0)) => n.as_f64().map(|n| n as f32), + _ => return Err(CmdError::Params("reading line spacing must be document, 1, 1.25, 1.5, or 2".into())), + }; + s.view.read_line_spacing = spacing; + Ok(json!({"spacing": spacing})) +} + fn zoom(s: &mut Session, v: &Value) -> CmdResult { match v.get("value") { Some(Value::String(f)) if ["pageWidth", "onePage", "multiplePages", "textWidth"].contains(&f.as_str()) => { s.view.fit = f.clone(); s.view.multi_page = f == "multiplePages"; + s.view.single_column = f == "onePage" || f == "pageWidth"; } Some(x) => { let pct = x.as_f64().ok_or_else(|| CmdError::Params("zoom percent or fit mode".into()))? as f32; diff --git a/crates/engine/src/file_access.rs b/crates/engine/src/file_access.rs new file mode 100644 index 000000000..6744e7b46 --- /dev/null +++ b/crates/engine/src/file_access.rs @@ -0,0 +1,480 @@ +//! The folders automation may read and write (#769). +//! +//! `wordcraft-cli mcp --automation-read-root --automation-write-root ` confines the +//! files an agent's commands reach. [`AutomationRoots`] holds the two folders. While they are in force, every path a command +//! reads goes through [`check_read`] and every path it writes through [`check_write`]: +//! +//! - The path is made absolute against the working directory, as the file system reads it. Its +//! links are followed: for a file that doesn't exist yet, the links of its nearest existing +//! folder, the rest of the path being plain names (`..` below a missing folder is refused, and so +//! is a link that leads nowhere). +//! - The result must lie inside the root, compared by whole components (`/work2` isn't inside +//! `/work`), without regard to case on Windows. +//! - Read and write access are separate: a root given alone grants only its own kind of access. +//! +//! Roots are in force on a thread inside [`confine`], and in the whole process after +//! [`confine_process`] (`wordcraft-cli mcp`), except inside [`unconfined`]. Without roots +//! nothing is checked. +//! +//! A path is checked, then opened: a process that can change the folders in between (swapping in +//! a link) can still win that race. The roots keep an agent to the files it was given; they don't +//! guard against other programs on the computer. + +use std::cell::RefCell; +use std::ffi::OsString; +use std::path::{Component, Path, PathBuf}; +use std::sync::{Arc, OnceLock}; + +/// Read access or write access. +#[derive(Clone, Copy, Debug, PartialEq, Eq)] +pub enum Access { + Read, + Write, +} + +/// The folders automation may read and write; a root left out grants no access of its kind. +#[derive(Debug)] +pub struct AutomationRoots { + read: Option, + write: Option, +} + +#[derive(Debug)] +struct Root { + /// The folder made absolute, links followed: what paths are compared with. + resolved: PathBuf, + /// The same without Windows' `\\?\` prefix, for people. + shown: PathBuf, +} + +impl Root { + /// The folder at `path`, which must exist. + fn open(path: &Path, access: Access) -> Result { + let kind = match access { + Access::Read => "read", + Access::Write => "write", + }; + let resolved = std::fs::canonicalize(path) + .and_then(|p| if p.is_dir() { Ok(p) } else { Err(std::io::ErrorKind::NotADirectory.into()) }) + .map_err(|e| format!("cannot open the automation {kind} root `{}`: {e}", path.display()))?; + Ok(Self { shown: plain(resolved.clone()), resolved }) + } +} + +impl AutomationRoots { + /// The roots `read` and `write`, which must be existing folders; `None` when neither is given + /// (nothing is confined then). + pub fn new(read: Option<&Path>, write: Option<&Path>) -> Result>, String> { + if read.is_none() && write.is_none() { + return Ok(None); + } + let read = read.map(|p| Root::open(p, Access::Read)).transpose()?; + let write = write.map(|p| Root::open(p, Access::Write)).transpose()?; + Ok(Some(Arc::new(Self { read, write }))) + } + + /// The read root, absolute with its links followed. + pub fn read_root(&self) -> Option<&Path> { + self.read.as_ref().map(|r| r.shown.as_path()) + } + + /// The write root, absolute with its links followed. + pub fn write_root(&self) -> Option<&Path> { + self.write.as_ref().map(|r| r.shown.as_path()) + } + + /// May the file or folder at `path` be read (or written)? `Err` says why not. + pub fn check(&self, path: &str, access: Access) -> Result<(), String> { + let root = match access { + Access::Read => self.read.as_ref(), + Access::Write => self.write.as_ref(), + }; + let Some(Root { resolved: folder, shown }) = root else { + return match access { + Access::Read => Err(format!("automation filesystem access is not granted: read authority is absent: {path}")), + Access::Write => Err(format!("automation filesystem access is not granted: write authority is absent: {path}")), + }; + }; + if inside(&resolve(path)?, folder) { + return Ok(()); + } + let root = shown.display(); + match access { + Access::Read => Err(format!("automation path rejected: outside the read root {root}: {path}")), + Access::Write => Err(format!("automation path rejected: outside the write root {root}: {path}")), + } + } +} + +/// Where the roots are in force on a thread ([`confine`], [`unconfined`]). +#[derive(Clone, Debug)] +enum Scope { + Confined(Arc), + Unconfined, +} + +thread_local! { + /// This thread's scope; `None`: the process's roots ([`confine_process`]). + static SCOPE: RefCell> = const { RefCell::new(None) }; +} + +/// The roots in force in every thread outside [`confine`] and [`unconfined`]. +static PROCESS: OnceLock> = OnceLock::new(); + +/// Put `roots` in force in the whole process, for good. `false` when roots were in force already +/// (they stay as they were). +pub fn confine_process(roots: Arc) -> bool { + PROCESS.set(roots).is_ok() +} + +/// Puts the scope it replaced back when dropped (also when `f` panics). +struct Restore(Option); + +impl Drop for Restore { + fn drop(&mut self) { + let previous = self.0.take(); + SCOPE.with(|s| *s.borrow_mut() = previous); + } +} + +fn with_scope(scope: Option, f: impl FnOnce() -> T) -> T { + let _restore = Restore(SCOPE.with(|s| s.replace(scope))); + f() +} + +/// Run `f` with `roots` in force on this thread (`None`: as things stand). +pub fn confine(roots: Option<&Arc>, f: impl FnOnce() -> T) -> T { + match roots { + Some(r) => with_scope(Some(Scope::Confined(r.clone())), f), + None => f(), + } +} + +/// Run `f` with no roots in force on this thread: what the person at the keyboard does, and the +/// app's own folders. +pub fn unconfined(f: impl FnOnce() -> T) -> T { + with_scope(Some(Scope::Unconfined), f) +} + +/// The roots in force on this thread, if any. +pub fn current() -> Option> { + match SCOPE.with(|s| s.borrow().clone()) { + Some(Scope::Confined(r)) => Some(r), + Some(Scope::Unconfined) => None, + None => PROCESS.get().cloned(), + } +} + +/// `path` without Windows' `\\?\` prefix (which `canonicalize` adds), for people to read. +fn plain(path: PathBuf) -> PathBuf { + if !cfg!(windows) { + return path; + } + let text = path.to_string_lossy(); + if let Some(rest) = text.strip_prefix(r"\\?\UNC\") { + return PathBuf::from(format!(r"\\{rest}")); + } + match text.strip_prefix(r"\\?\") { + Some(rest) => PathBuf::from(rest), + None => path, + } +} + +/// May `path` be read here? Always, without roots in force. +pub fn check_read(path: &str) -> Result<(), String> { + current().map_or(Ok(()), |r| r.check(path, Access::Read)) +} + +/// May `path` be written here? Always, without roots in force. +pub fn check_write(path: &str) -> Result<(), String> { + current().map_or(Ok(()), |r| r.check(path, Access::Write)) +} + +/// `path` made absolute with its links followed; for a path that doesn't exist (yet), its nearest +/// existing folder's links followed and the rest appended. +fn resolve(path: &str) -> Result { + if path.is_empty() { + return Err("automation path rejected: path is empty".into()); + } + // Checked as given: making `…\NUL` absolute gives the device path `\\.\NUL` on Windows. + if cfg!(windows) && Path::new(path).components().any(|c| matches!(c, Component::Normal(n) if is_device_name(&n.to_string_lossy()))) { + return Err(format!("automation path rejected: reserved device names are not allowed: {path}")); + } + let absolute = std::path::absolute(path).map_err(|e| format!("automation path rejected: cannot resolve {path}: {e}"))?; + // The names below the nearest existing folder, last first. + let mut missing: Vec = vec![]; + let mut at = absolute.as_path(); + loop { + match std::fs::canonicalize(at) { + Ok(mut found) => { + found.extend(missing.iter().rev()); + return Ok(found); + } + Err(e) => match std::fs::symlink_metadata(at) { + Ok(m) if m.file_type().is_symlink() => return Err(format!("automation path rejected: a link that leads nowhere: {path}")), + // There, yet its real path can't be had (no permission): nothing to compare. + Ok(_) => return Err(format!("automation path rejected: cannot resolve {path}: {e}")), + Err(e) if e.kind() == std::io::ErrorKind::NotFound => {} + Err(e) => return Err(format!("automation path rejected: cannot resolve {path}: {e}")), + }, + } + let parent = at.parent(); + match (at.components().next_back(), parent) { + (Some(Component::Normal(name)), Some(parent)) => { + missing.push(name.to_os_string()); + at = parent; + } + (Some(Component::ParentDir), _) => { + return Err(format!("automation path rejected: parent traversal below a folder that doesn't exist: {path}")); + } + // Down to the root (or drive) and it doesn't exist either. + _ => { + let e = std::io::Error::from(std::io::ErrorKind::NotFound); + return Err(format!("automation path rejected: cannot resolve {path}: {e}")); + } + } + } +} + +/// Is `path` (resolved) `root` or inside it, component by component (without regard to case on +/// Windows, whose file systems ignore it)? +fn inside(path: &Path, root: &Path) -> bool { + let mut path = path.components(); + root.components().all(|r| path.next().is_some_and(|p| same_component(p, r))) +} + +fn same_component(a: Component, b: Component) -> bool { + if a == b { + return true; + } + if !cfg!(windows) { + return false; + } + match (a.as_os_str().to_str(), b.as_os_str().to_str()) { + (Some(a), Some(b)) => a.to_lowercase() == b.to_lowercase(), + _ => false, + } +} + +/// A Windows device name (`CON`, `NUL`, `COM1`… also with an extension), which names no file. +fn is_device_name(component: &str) -> bool { + let stem = component.split('.').next().unwrap_or(component).trim_end_matches(' ').to_ascii_uppercase(); + let numbered = |prefix: &str| stem.strip_prefix(prefix).is_some_and(|n| matches!(n, "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9")); + matches!(stem.as_str(), "CON" | "PRN" | "AUX" | "NUL" | "CLOCK$" | "CONIN$" | "CONOUT$") || numbered("COM") || numbered("LPT") +} + +#[cfg(test)] +mod tests { + use super::*; + + /// A fresh folder holding `in` (read root), `out` (write root) and `outside`. + fn folders(name: &str) -> (PathBuf, Arc) { + let base = std::env::temp_dir().join(format!("wc-file-access-{name}-{}", std::process::id())); + let _ = std::fs::remove_dir_all(&base); + for d in ["in/sub", "out", "outside", "in2"] { + std::fs::create_dir_all(base.join(d)).unwrap(); + } + std::fs::write(base.join("in/a.docx"), "x").unwrap(); + std::fs::write(base.join("outside/secret.docx"), "x").unwrap(); + let roots = AutomationRoots::new(Some(&base.join("in")), Some(&base.join("out"))).unwrap().unwrap(); + (base, roots) + } + + fn s(p: PathBuf) -> String { + p.to_string_lossy().into_owned() + } + + #[test] + fn paths_inside_the_roots_pass_and_others_dont() { + let (base, roots) = folders("inside"); + assert_eq!(roots.check(&s(base.join("in/a.docx")), Access::Read), Ok(())); + assert_eq!(roots.check(&s(base.join("in")), Access::Read), Ok(()), "the root itself"); + assert_eq!(roots.check(&s(base.join("in/sub/new.docx")), Access::Read), Ok(()), "missing files are checked by their folder"); + assert_eq!(roots.check(&s(base.join("out/new/deeper/x.png")), Access::Write), Ok(()), "and by their nearest folder"); + let e = roots.check(&s(base.join("outside/secret.docx")), Access::Read).unwrap_err(); + assert!(e.starts_with("automation path rejected: outside the read root"), "{e}"); + // Read and write access are separate. + let e = roots.check(&s(base.join("in/a.docx")), Access::Write).unwrap_err(); + assert!(e.contains("outside the write root"), "{e}"); + assert!(roots.check(&s(base.join("out/x.png")), Access::Read).unwrap_err().contains("outside the read root")); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn a_sibling_with_the_same_prefix_is_outside() { + let (base, roots) = folders("prefix"); + std::fs::write(base.join("in2/b.docx"), "x").unwrap(); + assert!(roots.check(&s(base.join("in2/b.docx")), Access::Read).is_err(), "in2 is not inside in"); + assert!(roots.check(&format!("{}2/c.docx", s(base.join("in"))), Access::Read).is_err()); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn parent_traversal_is_resolved_or_refused() { + let (base, roots) = folders("dotdot"); + let sep = std::path::MAIN_SEPARATOR; + let up = format!("{}{sep}..{sep}outside{sep}secret.docx", s(base.join("in"))); + assert!(roots.check(&up, Access::Read).unwrap_err().contains("outside the read root"), "`..` out of the root"); + let back = format!("{}{sep}..{sep}a.docx", s(base.join("in/sub"))); + assert_eq!(roots.check(&back, Access::Read), Ok(()), "`..` that stays inside"); + let missing = format!("{}{sep}nope{sep}..{sep}..{sep}outside{sep}x.docx", s(base.join("out"))); + let e = roots.check(&missing, Access::Write).unwrap_err(); + // Windows resolves `..` by name first (so does opening the file there): then it's outside. + assert!(e.contains("parent traversal below a folder that doesn't exist") || e.contains("outside the write root"), "{e}"); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn relative_paths_resolve_against_the_working_directory() { + let (base, roots) = folders("relative"); + // The tests' working directory is the crate's folder: not inside the roots. + assert!(roots.check("a.docx", Access::Read).unwrap_err().contains("outside the read root")); + assert!(roots.check("", Access::Read).unwrap_err().contains("path is empty")); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn a_missing_root_is_an_error_and_none_is_no_confinement() { + let (base, _) = folders("missing"); + let e = AutomationRoots::new(Some(&base.join("nope")), None).unwrap_err(); + assert!(e.starts_with("cannot open the automation read root"), "{e}"); + let e = AutomationRoots::new(None, Some(&base.join("in/a.docx"))).unwrap_err(); + assert!( + e.starts_with("cannot open the automation write root") + && e.ends_with(&std::io::Error::from(std::io::ErrorKind::NotADirectory).to_string()), + "{e}" + ); + assert!(AutomationRoots::new(Some(Path::new("")), None).unwrap_err().starts_with("cannot open the automation read root ``")); + assert!(AutomationRoots::new(None, None).unwrap().is_none()); + // One root alone grants only its own access. + let read_only = AutomationRoots::new(Some(&base.join("in")), None).unwrap().unwrap(); + let e = read_only.check(&s(base.join("in/x.docx")), Access::Write).unwrap_err(); + assert!(e.contains("write authority is absent"), "{e}"); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn links_are_followed_before_comparing() { + let (base, roots) = folders("links"); + #[cfg(unix)] + let made = |target: &Path, link: &Path| std::os::unix::fs::symlink(target, link).is_ok(); + #[cfg(windows)] + let made = |target: &Path, link: &Path| { + let made = + if target.is_dir() { std::os::windows::fs::symlink_dir(target, link) } else { std::os::windows::fs::symlink_file(target, link) }; + made.is_ok() + }; + #[cfg(not(any(unix, windows)))] + let made = |_: &Path, _: &Path| false; + // Windows needs Developer Mode or elevation to make links: nothing to test without them. + if !made(&base.join("outside/secret.docx"), &base.join("in/leak.docx")) { + eprintln!("skipped: symbolic links can't be made here"); + let _ = std::fs::remove_dir_all(base); + return; + } + assert!(roots.check(&s(base.join("in/leak.docx")), Access::Read).unwrap_err().contains("outside the read root"), "a file link out"); + assert!(made(&base.join("outside"), &base.join("out/away"))); + let e = roots.check(&s(base.join("out/away/new.png")), Access::Write).unwrap_err(); + assert!(e.contains("outside the write root"), "a folder link out: {e}"); + assert!(made(&base.join("outside/gone.docx"), &base.join("out/dangling.png"))); + let e = roots.check(&s(base.join("out/dangling.png")), Access::Write).unwrap_err(); + assert!(e.contains("a link that leads nowhere"), "writing would create the file outside: {e}"); + // A link to a folder inside the root stays inside. + assert!(made(&base.join("in/sub"), &base.join("in/alias"))); + assert_eq!(roots.check(&s(base.join("in/alias/x.docx")), Access::Read), Ok(())); + // A root given through a link is the folder it leads to. + assert!(made(&base.join("in"), &base.join("in-link"))); + let linked = AutomationRoots::new(Some(&base.join("in-link")), None).unwrap().unwrap(); + assert_eq!(linked.check(&s(base.join("in/a.docx")), Access::Read), Ok(())); + let _ = std::fs::remove_dir_all(base); + } + + #[cfg(windows)] + #[test] + fn windows_ignores_case_and_refuses_devices() { + let (base, roots) = folders("case"); + let upper = s(base.join("IN/A.SVG")); + assert_eq!(roots.check(&upper, Access::Read), Ok(()), "{upper}"); + assert_eq!(roots.check(&s(base.join("In/Sub/New.docx")), Access::Read), Ok(())); + let e = roots.check(&s(base.join("out/NUL")), Access::Write).unwrap_err(); + assert!(e.contains("reserved device names"), "{e}"); + assert!(roots.check(&s(base.join("out/com1.txt")), Access::Write).is_err()); + assert_eq!(roots.check(&s(base.join("out/console.txt")), Access::Write), Ok(())); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn scopes_nest_and_stay_on_their_thread() { + let (base, roots) = folders("scopes"); + let outside = s(base.join("outside/secret.docx")); + assert_eq!(check_read(&outside), Ok(()), "nothing is confined by default"); + confine(Some(&roots), || { + assert!(check_read(&outside).is_err()); + assert_eq!(unconfined(|| check_read(&outside)), Ok(())); + assert!(check_read(&outside).is_err(), "restored after unconfined"); + let o = outside.clone(); + let there = std::thread::spawn(move || check_read(&o).is_ok()).join().unwrap(); + assert!(there, "another thread is not confined by this thread's scope"); + }); + assert_eq!(check_read(&outside), Ok(()), "restored after confine"); + let r = std::panic::catch_unwind(|| confine(Some(&roots), || panic!("boom"))); + assert!(r.is_err()); + assert_eq!(check_read(&outside), Ok(()), "restored after a panic"); + assert_eq!(confine(None, || check_read(&outside)), Ok(())); + let _ = std::fs::remove_dir_all(base); + } + + #[test] + fn device_names() { + for d in ["CON", "nul", "Com1", "LPT9.txt", "aux.docx", "NUL "] { + assert!(is_device_name(d), "{d}"); + } + for f in ["console", "com10", "COM0", "lpt", "a.docx", "nul-x"] { + assert!(!is_device_name(f), "{f}"); + } + } + + /// Every file a command reads or writes by path goes through `io` (`open_path_with`, + /// `save_path_with`, `read_file`, `write_file`), which check the roots. Engine code that + /// reaches for `std::fs` itself would bypass them, so none may: every file under `src/` + /// (commands, `print.rs`, `session.rs`…) except `io.rs` with its `io/` helpers (`safe_save`, + /// reached only after `io` checked the roots), this module, and the test files. + #[test] + fn commands_do_not_touch_the_file_system_themselves() { + fn walk(dir: &Path, files: &mut Vec) { + for entry in std::fs::read_dir(dir).unwrap() { + let path = entry.unwrap().path(); + if path.is_dir() { + walk(&path, files); + } else if path.extension().is_some_and(|e| e == "rs") { + files.push(path); + } + } + } + let src = Path::new(env!("CARGO_MANIFEST_DIR")).join("src"); + let mut files = vec![]; + walk(&src, &mut files); + assert!(files.iter().any(|f| f.ends_with("print.rs")) && files.iter().any(|f| f.ends_with("cmd/file.rs")), "{files:?}"); + let mut offenders = vec![]; + for path in files { + let name = path.file_name().and_then(|n| n.to_str()).unwrap_or(""); + // `io` and this module are the checked gateway; `tests*.rs` are test-only modules. + if path.parent() == Some(src.as_path()) && (name == "io.rs" || name == "file_access.rs" || name.starts_with("tests")) + || path.starts_with(src.join("io")) + { + continue; + } + let text = std::fs::read_to_string(&path).unwrap(); + // Test modules (which build their own fixtures) come last in each file. + let code = text.split("#[cfg(test)]").next().unwrap_or(""); + for (n, line) in code.lines().enumerate() { + let line = line.trim_start(); + let touches = ["std::fs", "File::open", "File::create", "OpenOptions"].iter().any(|w| line.contains(w)); + if !line.starts_with("//") && touches { + offenders.push(format!("{}:{}: {line}", path.display(), n + 1)); + } + } + } + assert!(offenders.is_empty(), "use crate::io::{{read_file, write_file, open_path_with, save_path_with}}: {offenders:#?}"); + } +} diff --git a/crates/engine/src/io.rs b/crates/engine/src/io.rs index f3fc25819..cdc1d63b5 100644 --- a/crates/engine/src/io.rs +++ b/crates/engine/src/io.rs @@ -126,6 +126,43 @@ pub fn open_bytes_with(name: &str, bytes: &[u8], password: Option<&str>) -> Resu Ok((doc, false)) } +/// The narrowest text width of the document's sections, points: what a picture must fit. +pub(crate) fn text_width(doc: &Document) -> f32 { + doc.sections().iter().map(|(_, sp)| sp.text_width()).fold(f32::INFINITY, f32::min).clamp(36.0, 4000.0) +} + +/// Shrink inline pictures wider than `max_w` to it, keeping their proportions. Layout keeps a +/// picture's size (as Word does), so formats whose pictures come at their pixel size (HTML, +/// Markdown, ODT, RTF) are fitted when they are read, as Insert › Pictures does. +pub(crate) fn fit_block_pictures(b: &mut wordcraft_doc::Block, max_w: f32) { + crate::cmd::paste::each_para(b, 0, &mut |p| { + let mut changed = false; + for o in &mut p.objects { + if let wordcraft_doc::InlineObject::Image { w, h, float, .. } = o + && float.wrap == wordcraft_doc::para::Wrap::Inline + && w.is_finite() + && *w > max_w + { + let k = max_w / *w; + *w = max_w; + *h = (*h * k).max(1.0); + changed = true; + } + } + if changed { + p.touch(); + } + }); +} + +/// [`fit_block_pictures`] for a whole document read from a file, against its own text width. +pub(crate) fn fit_pictures(doc: &mut Document) { + let max_w = text_width(doc); + for b in &mut doc.body { + fit_block_pictures(std::sync::Arc::make_mut(b), max_w); + } +} + /// A file whose extension names no format WordCraft knows (`.m3u`, `.log`, `.ini`, none…), /// opened by what it holds, as Word does: a Word document or RTF file by its signature, any /// other text through the plain-text import with encoding detection (#569). Anything else is @@ -191,6 +228,7 @@ pub fn open_path(path: &std::path::Path) -> Result { /// [`open_path`] for a file that may be password-protected (see [`open_bytes_with`]). #[cfg(not(target_arch = "wasm32"))] pub fn open_path_with(path: &std::path::Path, password: Option<&str>) -> Result<(Document, bool), String> { + crate::file_access::check_read(&path.to_string_lossy())?; let meta = std::fs::metadata(path).map_err(|e| format!("{}: {e}", path.display()))?; if meta.len() > 2 << 30 { return Err(format!("{}: file is larger than 2 GB", path.display())); @@ -201,6 +239,7 @@ pub fn open_path_with(path: &std::path::Path, password: Option<&str>) -> Result< // Pictures referenced by a relative path load from beside the HTML file. let images = LocalImages::new(path.parent().unwrap_or(std::path::Path::new(""))); let mut doc = wordcraft_formats::html::import_with(&wordcraft_formats::html::decode(&bytes), &|src| images.load(src)); + fit_pictures(&mut doc); doc.ensure_nonempty(); return Ok((doc, false)); } @@ -209,7 +248,8 @@ pub fn open_path_with(path: &std::path::Path, password: Option<&str>) -> Result< /// The pictures an HTML file names by a path relative to its folder. /// -/// Only regular files inside the folder (after resolving `..` and symlinks) are read. URLs +/// Only regular files inside the folder (after resolving `..` and symlinks) are read. The folder is +/// the HTML file's own, which `open_path_with` has already checked against the automation read root. URLs /// (`http:`, `file:`…) and absolute paths are not followed. Each file is read at most once and /// at most `max_file` bytes of it; every reference, cached or not, counts against a budget for /// the whole document, so one large picture can't be multiplied by referencing it many times. @@ -310,6 +350,46 @@ fn read_regular_file(path: &std::path::Path, limit: u64) -> Option> { Some(out) } +/// The bytes of the file at `path`, for commands that take a path (pictures, mail-merge data). +/// Every such read goes through here so the automation read root applies ([`crate::file_access`]). +#[cfg(not(target_arch = "wasm32"))] +pub fn read_file(path: &str) -> Result, String> { + crate::file_access::check_read(path)?; + std::fs::read(path).map_err(|e| format!("{path}: {e}")) +} + +#[cfg(target_arch = "wasm32")] +pub fn read_file(path: &str) -> Result, String> { + Err(format!("{path}: files are opened through the browser on the web")) +} + +/// [`read_file`] for a file that must be a regular file of at most `max` bytes, checked before +/// anything is read (a font named by the user, which could be any size). +#[cfg(not(target_arch = "wasm32"))] +pub fn read_file_capped(path: &str, max: u64, what: &str) -> Result, String> { + crate::file_access::check_read(path)?; + let meta = std::fs::metadata(path).map_err(|e| format!("{path}: {e}"))?; + if !meta.is_file() { + return Err(format!("{path}: not a file")); + } + if meta.len() > max { + return Err(format!("{path}: the {what} is larger than {} MB", max >> 20)); + } + std::fs::read(path).map_err(|e| format!("{path}: {e}")) +} + +/// Write `bytes` to the file at `path`, inside the automation write root when there is one. +#[cfg(not(target_arch = "wasm32"))] +pub fn write_file(path: &str, bytes: &[u8]) -> Result<(), String> { + crate::file_access::check_write(path)?; + std::fs::write(path, bytes).map_err(|e| format!("{path}: {e}")) +} + +#[cfg(target_arch = "wasm32")] +pub fn write_file(path: &str, _bytes: &[u8]) -> Result<(), String> { + Err(format!("{path}: files are saved through the browser on the web")) +} + #[cfg(target_arch = "wasm32")] pub fn open_path_with(path: &std::path::Path, _password: Option<&str>) -> Result<(Document, bool), String> { Err(format!("{}: files are opened through the browser on the web", path.display())) @@ -322,6 +402,7 @@ pub fn save_path(path: &std::path::Path, doc: &Document) -> Result<(), String> { /// [`save_path`], encrypted with `password` as [`save_bytes_with`] does. #[cfg(not(target_arch = "wasm32"))] pub fn save_path_with(path: &std::path::Path, doc: &Document, password: Option<&str>, opts: &ExportOptions) -> Result<(), String> { + crate::file_access::check_write(&path.to_string_lossy())?; let bytes = save_bytes_with(&path.to_string_lossy(), doc, password, opts)?; safe_save::write(path, &bytes).map_err(|e| format!("{}: {e}", path.display())) } diff --git a/crates/engine/src/io_ext.rs b/crates/engine/src/io_ext.rs index a0693d447..04bd7e484 100644 --- a/crates/engine/src/io_ext.rs +++ b/crates/engine/src/io_ext.rs @@ -9,7 +9,13 @@ pub fn open(ext: &str, bytes: &[u8]) -> Option> { "doc" | "dot" => Some(wordcraft_docbin::read(bytes).map_err(|e| e.to_string())), // Word XML Document (Flat OPC) or Word 2003 XML, told apart by the root element. "xml" => Some(wordcraft_docx::read_xml(bytes).map_err(|e| e.to_string())), - other => wordcraft_formats::import(other, bytes), + // HTML, Markdown, ODT, RTF…: pictures fit the text width (see `io::fit_pictures`). + other => wordcraft_formats::import(other, bytes).map(|r| { + r.map(|mut doc| { + crate::io::fit_pictures(&mut doc); + doc + }) + }), } } diff --git a/crates/engine/src/lib.rs b/crates/engine/src/lib.rs index 2ec789301..99edb750b 100644 --- a/crates/engine/src/lib.rs +++ b/crates/engine/src/lib.rs @@ -8,6 +8,7 @@ pub mod catalog; pub mod cmd; +pub mod file_access; pub mod io; mod io_ext; pub mod keymap; @@ -25,8 +26,11 @@ use serde_json::Value; pub use io::Password; pub use keymap::KeyMap; pub use ribbon::RibbonLayout; -pub use session::{ColumnBlock, EditSnapshot, FindState, MathEdit, Prefs, Selection, Session, ViewState}; +pub use session::{ + CellRange, ColumnBlock, EditSnapshot, FindState, MathEdit, Prefs, Selection, Session, UNDO_LIST_MAX, ViewState, blank_document, set_system_locale, +}; pub use wordcraft_doc as doc; +pub use wordcraft_geom::Region; pub use wordcraft_layout as layout; pub use wordcraft_proof as proof; pub use wordcraft_render as render; diff --git a/crates/engine/src/print.rs b/crates/engine/src/print.rs index 290dfdb94..9d05060d9 100644 --- a/crates/engine/src/print.rs +++ b/crates/engine/src/print.rs @@ -161,8 +161,9 @@ pub(crate) fn print_pdf_to_file(s: &mut Session, v: &Value) -> Result = OnceLock::new(); + +/// Tell the engine the system locale (the shell calls this once at startup; layout never reads +/// the environment). Headless sessions without it show Western digits for Numeral › System. +pub fn set_system_locale(tag: &str) { + let _ = SYSTEM_LOCALE.set(tag.to_string()); +} use serde::{Deserialize, Serialize}; use serde_json::Value; use wordcraft_doc::edit::Fragment; use wordcraft_doc::props::CharProps; -use wordcraft_doc::{Document, Path, Pos, StoryRef}; -use wordcraft_layout::{DocLayout, LayoutCache, LayoutOptions, ViewMode}; +use wordcraft_doc::{Document, Path, Pos, StoryRef, Table}; +use wordcraft_layout::{DocLayout, LayoutCache, LayoutOptions, NumeralMode, ViewMode}; use crate::{CmdError, Registry}; @@ -42,6 +51,52 @@ pub struct ColumnBlock { pub focus_x: f32, /// One `(start, end)` per line, in document order, each inside one paragraph. pub segments: Vec<(Pos, Pos)>, + /// Set when the block is whole table cells (Select Cell, Row, Column or Table, or the mouse in + /// a table's selection areas): one segment per paragraph of the cells. + pub cells: Option, +} + +/// Whole table cells selected: rows `rows.0..=rows.1` and every cell of them that reaches into +/// grid columns `cols.0..=cols.1`. +#[derive(Clone, Debug, PartialEq)] +pub struct CellRange { + pub story: StoryRef, + pub table: Path, + pub rows: (usize, usize), + pub cols: (usize, usize), + /// The cells are one run of the text (cells of one row, or whole rows), so the stream + /// selection between the corners holds just them and commands act on that as usual. + /// Otherwise (part of a column) text edits and formatting go cell by cell. + pub stream: bool, +} + +impl CellRange { + /// The selected cells of table `t` (row, cell index), in document order. + pub fn cells(&self, t: &Table) -> Vec<(usize, usize)> { + let ((r0, r1), (g0, g1)) = (self.rows, self.cols); + let mut out = Vec::new(); + for (r, row) in t.rows.iter().enumerate().take(r1.saturating_add(1)).skip(r0) { + let mut g = 0usize; + for (c, cell) in row.cells.iter().enumerate() { + let end = g.saturating_add(cell.span()); + if g <= g1 && end > g0 { + out.push((r, c)); + } + g = end; + } + } + out + } +} + +/// A selection Word's table rule widened to whole cells or rows (`cmd::table::extend_selection`), +/// and where it really runs. +#[derive(Clone, Debug, PartialEq)] +pub(crate) struct Widened { + /// The selection it became; this is current while `Session::sel` still equals it. + pub sel: Selection, + /// Where the selection was really extended from and to: extending it again goes on from here. + pub ends: Selection, } /// View state that commands can change (ribbon View tab, status bar). @@ -78,14 +133,33 @@ pub struct ViewState { #[serde(default)] pub alt_text_pane: bool, pub multi_page: bool, + /// One Page or Page Width was chosen last: keep one page per row when a typed or wheel zoom replaces the fit mode. + #[serde(default)] + pub single_column: bool, /// Zoom to fit: "pageWidth", "onePage", "multiplePages", or empty. pub fit: String, pub web_width: f32, + /// Transient viewport metrics for reading pagination; never saved in a document or preferences. + #[serde(skip, default = "default_read_width")] + pub read_width: f32, + #[serde(skip, default = "default_read_height")] + pub read_height: f32, + /// Reader-only layout controls; they are neither document formatting nor saved UI prefs. + /// `None` chooses responsive column width and keeps the document's line spacing. + #[serde(skip)] + pub read_column_width: Option, + #[serde(skip, default = "default_read_zoom")] + pub read_zoom: f32, + #[serde(skip)] + pub read_line_spacing: Option, pub dark_mode: bool, pub show_markup: bool, pub track_changes_pane: bool, /// Check spelling and grammar as you type. pub proofing: bool, + /// How digits look on screen (Word › File › Options › Advanced › Numeral). + #[serde(default)] + pub numeral: NumeralMode, /// Review › Hide Ink: ink strokes aren't shown on screen (they stay in the document). #[serde(default)] pub hide_ink: bool, @@ -112,6 +186,18 @@ fn on() -> bool { true } +fn default_read_width() -> f32 { + 300.0 +} + +fn default_read_height() -> f32 { + 500.0 +} + +fn default_read_zoom() -> f32 { + 1.0 +} + impl Default for ViewState { fn default() -> Self { ViewState { @@ -132,12 +218,19 @@ impl Default for ViewState { accessibility_pane: false, alt_text_pane: false, multi_page: false, + single_column: false, fit: String::new(), web_width: 800.0, + read_width: default_read_width(), + read_height: default_read_height(), + read_column_width: None, + read_zoom: default_read_zoom(), + read_line_spacing: None, dark_mode: false, show_markup: true, track_changes_pane: false, proofing: true, + numeral: NumeralMode::default(), hide_ink: false, draw: Default::default(), field_codes: false, @@ -183,6 +276,9 @@ struct Undo { sel: Selection, /// Formatting waiting at the caret (Bold with nothing selected), so Undo takes it back too. pending: Option, + /// The column block or whole cells the selection was (undoing brings the selection back as + /// it was). + column: Option, } /// What [`Session::restore`] needs to take commands back exactly: the document, the selection, @@ -232,6 +328,11 @@ pub struct Session { pub pending: Option, pub path: Option, pub dirty: bool, + /// Where the custom dictionary is kept on disk, if the host wired one up. Only the desktop + /// app does: headless sessions (MCP, `wordcraft-cli run`), tests and the web build leave + /// `None`, so an agent's words stay in memory and never change the person's dictionary. + /// Add to Dictionary and File › Options › Proofing save through it (#40). + pub user_dict_path: Option, pub clipboard: Option, /// A tracked cut the next paste can turn into a move (see `cmd::moves`). pub pending_move: Option, @@ -268,9 +369,21 @@ pub struct Session { /// mail merge, recover), never on an edit. document_id: u64, cache: LayoutCache, - /// (revision, web width, view, layout, proofing, outline, (deletions left out, body - /// deletions in balloons)). - layout: Option<(u64, f32, ViewMode, Arc, bool, Option, (bool, bool))>, + /// (revision, width, read height, view, read line spacing, layout, proofing, outline, + /// (deletions left out, body deletions in balloons), numeral display). + #[allow(clippy::type_complexity)] + layout: Option<( + u64, + f32, + f32, + ViewMode, + Option, + Arc, + bool, + Option, + (bool, bool), + NumeralMode, + )>, /// Picture edits: edited media key → original media key (Reset Picture). pub originals: std::collections::HashMap, /// Last mutating command (Repeat). @@ -293,10 +406,19 @@ pub struct Session { pub read_aloud: crate::speech::ReadAloud, /// Preferences the front end saves between runs. pub prefs: Prefs, + /// The system region (#838): the paper a new blank document starts on, the default + /// measurement unit and the decimal separator. The front end sets it from the platform (the + /// browser's locale on the web); headless sessions keep `en-US`. A region the user chose + /// ([`Prefs::region`]) wins: read it with [`Session::region`]. + pub system_region: wordcraft_geom::Region, /// Custom keyboard shortcuts (`tools.customizeKeyboard`); the front end saves them too. pub keymap: crate::KeyMap, /// Custom ribbon and Quick Access Toolbar (`tools.customizeRibbon`); the front end saves it too. pub ribbon: crate::RibbonLayout, + /// The language new blank documents (and opened plain text) are written in, a BCP 47 tag; + /// `None` is WordCraft's default, US English. The front end sets it from the interface + /// language with [`Session::set_editing_language`] (a Polish interface writes Polish). + pub editing_lang: Option, /// Editing inside an equation: which one and the caret in it. pub math: Option, /// Equations are typed in LaTeX rather than the linear format. @@ -314,6 +436,9 @@ pub struct Session { pub column: Option, /// Column selection mode (Ctrl+Shift+F8): caret movement extends the block. pub column_mode: bool, + /// A selection widened to whole table cells or rows, and where it really runs (see + /// [`Session::sel_ends`]). + pub(crate) widened: Option, /// Where the last edits happened, oldest first (at most [`EDIT_MARKS`]), for Go Back /// (Shift+F5). Positions aren't mapped through later edits; they're clamped when used. edit_marks: Vec, @@ -355,8 +480,14 @@ pub struct Prefs { pub insert_key_overtype: bool, /// Overtype: typed characters replace the ones after the caret. pub overtype: bool, - /// The unit rulers and dialogs show lengths in. - pub units: wordcraft_geom::Unit, + /// The unit rulers and dialogs show lengths in, when the user chose one (File › Options › + /// Advanced › Measurement units); `None` follows the region ([`Session::units`]). Saved under + /// a new key; an older `units` value other than inches (the old default) carries over on load. + #[serde(rename = "measurementUnits")] + pub units: Option, + /// The region the user chose for new documents, units and the decimal separator (a locale + /// tag like `pl-PL`, File › Options › Advanced › Region); `None` follows the system. + pub region: Option, /// Formatting marks shown even while Show/Hide ¶ is off. pub marks: wordcraft_layout::display::FormattingMarks, /// Mark grammar errors as you type (with Check spelling as you type on). @@ -380,6 +511,7 @@ impl Default for Prefs { insert_key_overtype: false, overtype: false, units: Default::default(), + region: None, marks: Default::default(), mark_grammar: true, ignore_uppercase: true, @@ -410,8 +542,24 @@ fn doc_bib_style(doc: &Document) -> String { /// Maximum undo depth. const MAX_UNDO: usize = 500; +/// How many steps the Undo button's list shows (#701); `edit.undoTo` reaches no further. +pub const UNDO_LIST_MAX: usize = 100; impl Session { + /// File name of the persisted custom dictionary inside the settings folder. + pub const USER_DICT_FILE: &str = "custom-dictionary.txt"; + + /// Save the custom dictionary through [`Session::user_dict_path`], if the host wired one up, + /// with `removed` (words just taken out) left out of the merge with the file. A failed save + /// keeps the words in memory and never fails the command (#40). + pub(crate) fn persist_user_dictionary(&self, removed: &[String]) { + if let Some(path) = self.user_dict_path.as_deref() + && let Err(e) = wordcraft_proof::save_user_dictionary(path, removed) + { + log::warn!("custom dictionary: save failed: {e}"); + } + } + pub fn new(doc: Document) -> Self { let start = doc.start_of(StoryRef::Body); let bib_style = doc_bib_style(&doc); @@ -421,6 +569,7 @@ impl Session { view: ViewState::default(), pending: None, path: None, + user_dict_path: None, dirty: false, clipboard: None, pending_move: None, @@ -458,12 +607,15 @@ impl Session { math_normal_text: false, also_selected: Vec::new(), prefs: Prefs::default(), + system_region: Default::default(), keymap: crate::KeyMap::default(), ribbon: crate::RibbonLayout::default(), + editing_lang: None, read_aloud: Default::default(), password: None, column: None, column_mode: false, + widened: None, edit_marks: Vec::new(), go_back_step: 0, } @@ -504,7 +656,20 @@ impl Session { /// The line pieces of the current column selection, if there is one (the selection hasn't /// moved since the block was made). pub fn column_segments(&self) -> Option<&[(Pos, Pos)]> { - self.column.as_ref().filter(|c| c.sel == self.sel).map(|c| c.segments.as_slice()) + // Cells that are one run of the text are an ordinary stream selection. + self.column.as_ref().filter(|c| c.sel == self.sel && !c.cells.as_ref().is_some_and(|r| r.stream)).map(|c| c.segments.as_slice()) + } + + /// The whole table cells selected, if the selection is that (see [`CellRange`]). + pub fn cell_range(&self) -> Option<&CellRange> { + self.column.as_ref().filter(|c| c.sel == self.sel).and_then(|c| c.cells.as_ref()) + } + + /// The selection as it was made: where a drag, Shift+click or Shift+arrows started and where + /// the caret went, even when Word's table rule widened it to whole cells or rows (`sel` holds + /// those). + pub fn sel_ends(&self) -> &Selection { + self.widened.as_ref().filter(|w| w.sel == self.sel).map_or(&self.sel, |w| &w.ends) } /// Document revision (bumped by every change). @@ -561,36 +726,45 @@ impl Session { /// The current layout (recomputed when the document or view changed). pub fn layout(&mut self) -> Arc { - let ww = self.view.web_width; + let mode = if self.view.read_mode { ViewMode::Read } else { self.view.mode }; + let width = if self.view.read_mode { self.view.read_width } else { self.view.web_width }; + let height = if self.view.read_mode { self.view.read_height } else { 0.0 }; + let spacing = if self.view.read_mode { self.view.read_line_spacing.map(f32::to_bits) } else { None }; // Deletions in balloons depend on the view (the Comments pane, Read Mode, side by side), so // a view change that moves them in or out of the text lays it out again. let hide_deleted = !self.view.show_markup || self.prefs.markup.hides_deletions(); let deleted = (hide_deleted, self.revisions_in_balloons()); - let outline = (self.view.mode == ViewMode::Outline).then(|| self.view.outline.view()); - if let Some((r, w, m, l, pf, o, hd)) = &self.layout + let outline = (mode == ViewMode::Outline).then(|| self.view.outline.view()); + if let Some((r, w, h, m, sp, l, pf, o, hd, n)) = &self.layout && *r == self.rev - && (*w == ww || self.view.mode == ViewMode::Print) - && *m == self.view.mode + && (*w == width || mode == ViewMode::Print) + && *h == height + && *m == mode + && *sp == spacing && *pf == self.view.proofing + && *n == self.view.numeral && *hd == deleted && *o == outline { return l.clone(); } let opts = LayoutOptions { - view: self.view.mode, - web_width: ww, + view: mode, + web_width: width, + read_height: height, + read_line_spacing: spacing.map(f32::from_bits), show_hidden: self.view.marks || self.prefs.marks.hidden, hide_deleted, deletions_in_balloons: deleted.1, proofing: self.view.proofing, + numeral: self.view.numeral.resolve(SYSTEM_LOCALE.get().map(String::as_str)), comment_balloons: false, field_codes: self.view.field_codes, outline: outline.clone(), proof: wordcraft_layout::ProofOptions { grammar: self.prefs.mark_grammar, spelling: self.prefs.spelling() }, }; let l = Arc::new(wordcraft_layout::layout(&self.doc, &mut self.cache, &opts)); - self.layout = Some((self.rev, ww, self.view.mode, l.clone(), self.view.proofing, outline, deleted)); + self.layout = Some((self.rev, width, height, mode, spacing, l.clone(), self.view.proofing, outline, deleted, self.view.numeral)); l } /// A layout for output (PDF, images, print): no proofing marks, print view. @@ -611,13 +785,37 @@ impl Session { self.cache.clear(); } + /// Set the language new documents are written in ([`Session::editing_lang`]; `None` or a + /// malformed tag: US English). A document nobody has touched yet — untitled, unchanged, no + /// undo steps, no text — switches too, so the blank document a session starts with follows + /// the interface language. Nothing else changes and no undo step is added. + pub fn set_editing_language(&mut self, tag: Option<&str>) { + let tag = tag.and_then(wordcraft_proof::lang::normalize_tag); + if self.editing_lang == tag { + return; + } + self.editing_lang = tag; + let untouched = self.path.is_none() && !self.dirty && self.history.is_empty() && self.redo.is_empty(); + if untouched && self.doc.plain_text(StoryRef::Body).trim().is_empty() && self.doc.paragraph_count() <= 1 { + self.doc.styles.default_chr.lang = Some(self.editing_lang.clone().unwrap_or_else(|| "en-US".into())); + self.rev = self.rev.wrapping_add(1); + self.relayout(); + } + } + /// Snapshot for undo before a change. pub fn checkpoint(&mut self, label: &str) { if label == "Typing" && self.typing_open { return; } self.typing_open = label == "Typing"; - self.history.push(Arc::new(Undo { label: label.to_string(), doc: self.doc.clone(), sel: self.sel.clone(), pending: self.pending.clone() })); + self.history.push(Arc::new(Undo { + label: label.to_string(), + doc: self.doc.clone(), + sel: self.sel.clone(), + pending: self.pending.clone(), + column: self.column.clone(), + })); if self.history.len() > MAX_UNDO { self.history.remove(0); self.undo_evicted += 1; @@ -627,7 +825,7 @@ impl Session { /// Record `doc`/`sel` as their own undo step after the fact and close the typing group, so /// Undo right after an automatic change (AutoFormat, AutoCorrect) reverts only that change. pub fn push_undo(&mut self, label: &str, doc: Document, sel: Selection) { - self.history.push(Arc::new(Undo { label: label.to_string(), doc, sel, pending: self.pending.clone() })); + self.history.push(Arc::new(Undo { label: label.to_string(), doc, sel, pending: self.pending.clone(), column: None })); if self.history.len() > MAX_UNDO { self.history.remove(0); self.undo_evicted += 1; @@ -675,6 +873,21 @@ impl Session { pub fn undo_labels(&self) -> Vec { self.history.iter().rev().map(|u| u.label.clone()).collect() } + /// The Undo button's list (#701): the undo steps' labels (English, as commands name them), + /// newest first, at most [`UNDO_LIST_MAX`]. + pub fn undo_history(&self) -> Vec<&str> { + self.history.iter().rev().take(UNDO_LIST_MAX).map(|u| u.label.as_str()).collect() + } + /// Undo the newest `count` steps in one go, as picking an entry in the Undo list does (#701). + /// More than the undo depth undoes everything; a step that would take the document's + /// protection away stops it (only Stop Protection lifts it). Returns the steps undone. + pub fn undo_steps(&mut self, count: usize) -> usize { + let mut done = 0; + while done < count && !self.step_lifts_protection(false) && self.undo() { + done += 1; + } + done + } pub fn undo(&mut self) -> bool { self.typing_open = false; let Some(u) = self.history.pop().map(Arc::unwrap_or_clone) else { return false }; @@ -683,6 +896,7 @@ impl Session { doc: std::mem::replace(&mut self.doc, u.doc), sel: std::mem::replace(&mut self.sel, u.sel), pending: std::mem::replace(&mut self.pending, u.pending), + column: std::mem::replace(&mut self.column, u.column), }; self.redo.push(Arc::new(cur)); self.touch(); @@ -701,6 +915,7 @@ impl Session { doc: std::mem::replace(&mut self.doc, u.doc), sel: std::mem::replace(&mut self.sel, u.sel), pending: std::mem::replace(&mut self.pending, u.pending), + column: std::mem::replace(&mut self.column, u.column), }; self.history.push(Arc::new(cur)); self.touch(); @@ -713,6 +928,31 @@ impl Session { self.typing_open = false; } + /// The region in effect (#838): the one the user chose ([`Prefs::region`]), else the + /// system's. + pub fn region(&self) -> wordcraft_geom::Region { + match self.prefs.region.as_deref() { + Some(tag) => wordcraft_geom::Region::from_tag(tag), + None => self.system_region.clone(), + } + } + + /// The unit rulers and dialogs show lengths in: the user's choice, else the region's. + pub fn units(&self) -> wordcraft_geom::Unit { + self.prefs.units.unwrap_or_else(|| self.region().units()) + } + + /// `pt` in [`Session::units`], written the region's way (`2,54 cm` in Poland). + pub fn format_length(&self, pt: f32) -> String { + self.region().format(self.units(), pt) + } + + /// A new blank document (the Normal template) on the region's paper: Letter in the US and + /// other Letter regions, A4 elsewhere (#838). Margins stay 2.54 cm (1") all round. + pub fn blank_document(&self) -> Document { + blank_document(&self.region()) + } + /// Replace the document (open/new). pub fn set_document(&mut self, doc: Document) { self.document_id = self.document_id.wrapping_add(1); @@ -725,6 +965,7 @@ impl Session { self.also_selected.clear(); self.column = None; self.column_mode = false; + self.widened = None; self.edit_marks.clear(); self.go_back_step = 0; self.view.outline.collapsed.clear(); @@ -757,6 +998,8 @@ impl Session { view: _, pending: _, path: _, + // The custom dictionary file, like the file path: host wiring, not an edit. + user_dict_path: _, dirty, clipboard: _, // Like the clipboard: what the next paste can make a move of. @@ -792,8 +1035,11 @@ impl Session { document_id: _, read_aloud: _, prefs: _, + system_region: _, keymap: _, ribbon: _, + // A preference for new documents, like `prefs`: not part of this document. + editing_lang: _, // Equation editing mode, like `view`: not part of the document or its history. math: _, math_latex: _, @@ -801,9 +1047,11 @@ impl Session { // The save password, like the file path: not an edit to the document. password: _, also_selected: _, - // Column selection, like `sel`'s shape: valid only while `sel` matches it. + // Column selection and a widened selection's ends, like `sel`'s shape: valid only while + // `sel` matches them. column: _, column_mode: _, + widened: _, // Go Back's places, like the caret history: not part of the document. edit_marks: _, go_back_step: _, @@ -935,7 +1183,8 @@ impl Session { self.typing_open = false; } // Only the paste right after a tracked cut makes it a move. - if (spec.mutates || matches!(id, "edit.undo" | "edit.redo")) && !matches!(id, "edit.cut" | "edit.paste" | "edit.pasteSpecial") { + if (spec.mutates || matches!(id, "edit.undo" | "edit.undoTo" | "edit.redo")) && !matches!(id, "edit.cut" | "edit.paste" | "edit.pasteSpecial") + { self.pending_move = None; } let sel_before = self.sel.clone(); @@ -1034,7 +1283,21 @@ impl Session { /// Selected plain text. pub fn selected_text(&self) -> String { + // A column block or whole cells that aren't one run of the text: each piece's own text, + // a line each, so the status bar and Word Count count those cells only. + if let Some(segs) = self.column_segments() { + return segs.iter().map(|(a, b)| self.doc.copy_range(a.min(b), a.max(b)).plain_text()).collect::>().join("\n"); + } let (a, b) = self.sel.ordered(); self.doc.copy_range(&a, &b).plain_text() } } + +/// A new blank document on `region`'s paper (see [`Session::blank_document`]). +pub fn blank_document(region: &wordcraft_geom::Region) -> Document { + let mut d = Document::new(); + let (_, w, h) = region.paper(); + d.last_section.page_w = w; + d.last_section.page_h = h; + d +} diff --git a/crates/engine/src/tests.rs b/crates/engine/src/tests.rs index 79af9d056..cd1e41f6a 100644 --- a/crates/engine/src/tests.rs +++ b/crates/engine/src/tests.rs @@ -125,6 +125,66 @@ fn reset_shortcuts_remove_direct_formatting_but_keep_styles() { assert_eq!(text(&s), "Hello world—"); } +#[test] +fn custom_dictionary_words_persist_through_their_file() { + // #40: adding a word through a command writes it to the session's file, and loading + // that file makes the word known (only the desktop app wires this up). + let dir = std::env::temp_dir().join(format!("wc-engdict-{}", std::process::id())); + let _ = std::fs::create_dir_all(&dir); + let path = dir.join("custom-dictionary.txt"); + let _ = std::fs::remove_file(&path); + let mut s = s(); + s.user_dict_path = Some(path.clone()); + run(&mut s, "review.addToDictionary", json!({"word": "zxqengpersist"})); + let saved = std::fs::read_to_string(&path).unwrap_or_default(); + assert!(saved.lines().any(|l| l == "zxqengpersist"), "the word reaches the file: {saved:?}"); + run(&mut s, "file.options", json!({"removeWords": ["zxqengpersist"]})); + let saved = std::fs::read_to_string(&path).unwrap_or_default(); + assert!(!saved.lines().any(|l| l == "zxqengpersist"), "removal reaches the file too: {saved:?}"); + + // Ignore All is for this session only, as in Word: the word is accepted, never saved. + assert!(!wordcraft_proof::is_correct("zxqengignored")); + run(&mut s, "review.ignoreAll", json!({"word": "zxqengignored"})); + assert!(wordcraft_proof::is_correct("zxqengignored")); + assert!(!wordcraft_proof::user_dictionary().contains(&"zxqengignored".to_string())); + let saved = std::fs::read_to_string(&path).unwrap_or_default(); + assert!(!saved.contains("zxqengignored"), "Ignore All doesn't reach the file: {saved:?}"); + let _ = std::fs::remove_dir_all(&dir); +} + +#[test] +fn add_to_dictionary_takes_only_words_the_file_keeps() { + // #40: the same rule as loading the dictionary file, so nothing added is dropped on the next load. + let mut s = s(); + for bad in ["two words", "tab\tword", "bell\u{7}", &"x".repeat(65)] { + assert!(s.run("review.addToDictionary", &json!({"word": bad})).is_err(), "{bad:?}"); + assert!(!wordcraft_proof::user_dictionary().iter().any(|w| w == &bad.to_lowercase())); + } + run(&mut s, "review.addToDictionary", json!({"word": " zxqengtrimmed "})); + assert!(wordcraft_proof::user_dictionary().contains(&"zxqengtrimmed".to_string())); +} + +#[test] +fn non_english_runs_are_not_proofed() { + // #100: runs tagged with a proofing language that has no dictionary are skipped + // by both proofing paths; English still is proofed. + let mut s = s(); + run(&mut s, "text.insert", json!({"text": "ąžuolas labas"})); + let before = run(&mut s, "review.issues", json!({})); + assert!(!before.as_array().map(|a| a.is_empty()).unwrap_or(true), "default language flags the words: {before}"); + + run(&mut s, "select.all", json!({})); + run(&mut s, "review.language", json!({"lang": "lt-LT"})); + let after = run(&mut s, "review.issues", json!({})); + assert_eq!(after, json!([]), "Lithuanian runs are skipped: {after}"); + + // English typos are still caught. + let mut e = Session::new(wordcraft_doc::Document::new()); + run(&mut e, "text.insert", json!({"text": "Hello helo world"})); + let en = run(&mut e, "review.issues", json!({})); + assert_eq!(en.as_array().map(|a| a.len()).unwrap_or(0), 1, "English typo still flagged: {en}"); +} + #[test] fn typing_enter_undo() { let mut s = s(); @@ -143,6 +203,68 @@ fn typing_enter_undo() { assert_eq!(text(&s), "Hello world"); } +/// The Undo button's list (#701): several steps undone at once, newest first. +#[test] +fn undo_to_takes_several_steps_back_at_once() { + let mut s = s(); + run(&mut s, "text.insert", json!({"text": "Hello"})); + run(&mut s, "select.all", json!({})); + run(&mut s, "format.bold", json!({})); + run(&mut s, "caret.docEnd", json!({})); + run(&mut s, "text.newParagraph", json!({})); + run(&mut s, "text.insert", json!({"text": "World"})); + let h = run(&mut s, "edit.undoHistory", json!({})); + assert_eq!(h["undo"], json!(["Typing", "New Paragraph", "Bold", "Typing"])); + assert_eq!(h["depth"], 4); + assert_eq!(s.undo_history(), ["Typing", "New Paragraph", "Bold", "Typing"]); + // Three steps in one command: World, the paragraph and the bold go; Hello stays. + let r = run(&mut s, "edit.undoTo", json!({"count": 3})); + assert_eq!(r["undone"], 3); + assert_eq!(r["depth"], 1); + assert_eq!(text(&s), "Hello"); + assert_eq!(s.doc.para_at(&s.sel.focus).map(|p| p.props_of_char(1).bold), Some(None)); + assert_eq!(s.undo_history(), ["Typing"]); + // Each undone step is one redo step. + run(&mut s, "edit.redo", json!({})); + run(&mut s, "edit.redo", json!({})); + run(&mut s, "edit.redo", json!({})); + assert_eq!(text(&s), "Hello\nWorld"); + // More than the depth undoes everything. + let r = run(&mut s, "edit.undoTo", json!({"count": 1e300})); + assert_eq!(r["undone"], 4); + assert_eq!(text(&s), ""); + assert!(!s.can_undo()); + assert!(s.run("edit.undoTo", &json!({"count": 1})).is_err(), "nothing to undo"); +} + +#[test] +fn undo_to_rejects_bad_counts_and_leaves_the_document_alone() { + let mut s = s(); + run(&mut s, "text.insert", json!({"text": "Hello"})); + run(&mut s, "text.newParagraph", json!({})); + let before = (s.doc.clone(), s.undo_depth()); + for junk in + [json!({}), json!(null), json!({"count": 0}), json!({"count": -3}), json!({"count": "2"}), json!({"count": [2]}), json!({"count": -1e308})] + { + assert!(s.run("edit.undoTo", &junk).is_err(), "{junk}"); + assert_eq!((s.doc.clone(), s.undo_depth()), before, "{junk}"); + } + // A fraction counts whole steps. + let r = run(&mut s, "edit.undoTo", json!({"count": 1.9})); + assert_eq!(r["undone"], 1); + assert_eq!(text(&s), "Hello"); +} + +#[test] +fn undo_history_is_capped() { + let mut s = s(); + for i in 0..(crate::UNDO_LIST_MAX + 20) { + run(&mut s, "para.indents", json!({"left": i as f64})); + } + assert_eq!(s.undo_history().len(), crate::UNDO_LIST_MAX); + assert_eq!(run(&mut s, "edit.undoHistory", json!({}))["undo"].as_array().map(Vec::len), Some(crate::UNDO_LIST_MAX)); +} + #[test] fn joined_commands_are_one_undo_step() { let mut s = s(); @@ -769,6 +891,92 @@ fn paste_special_respects_track_changes_and_brings_lists() { assert!(text(&s).starts_with("one\ntwo")); } +/// Character formatting at a body offset of the first paragraph. +fn props_at(s: &Session, off: usize) -> wordcraft_doc::CharProps { + let pos = Pos { story: StoryRef::Body, path: wordcraft_doc::Path::top(0), off }; + s.doc.para_at(&pos).map(|p| p.props_of_char(off).clone()).unwrap_or_default() +} + +/// Paste of another program's formatted copy keeps its formatting (#687); Merge Formatting keeps +/// only its emphasis, in the destination's font; Keep Text Only drops it all. +#[test] +fn paste_keeps_source_formatting_from_other_programs() { + let mut s = s(); + run(&mut s, "document.setText", json!({"text": "x"})); + run(&mut s, "caret.docEnd", json!({})); + let html = "

red bold

"; + let clip = json!({"text": "red bold", "html": html}); + run(&mut s, "edit.paste", clip.clone()); + assert_eq!(text(&s), "xred bold"); + assert_eq!(props_at(&s, 1).font.as_deref(), Some("Courier New")); + assert!(props_at(&s, 1).color.is_some()); + assert!(!bold_at(&s, 0, 1) && bold_at(&s, 0, 5)); + run(&mut s, "edit.undo", json!({})); + assert_eq!(text(&s), "x"); + run(&mut s, "edit.pasteMerge", clip.clone()); + assert_eq!(text(&s), "xred bold"); + assert_eq!(props_at(&s, 1).font, None, "Merge Formatting takes the destination's font"); + assert_eq!(props_at(&s, 1).color, None); + assert!(bold_at(&s, 0, 5), "and keeps the emphasis"); + run(&mut s, "edit.undo", json!({})); + run(&mut s, "edit.pasteText", clip); + assert_eq!(text(&s), "xred bold"); + assert!(!bold_at(&s, 0, 5)); + + // Windows' "HTML Format" carries a header with the fragment's byte offsets. + run(&mut s, "document.setText", json!({"text": ""})); + let body = "win"; + let head = "Version:0.9\r\nStartHTML:0000000000\r\nEndHTML:0000000000\r\nStartFragment:SSSSSSSSSS\r\nEndFragment:EEEEEEEEEE\r\n"; + let start = head.len() + body.find("").unwrap(); + let end = head.len() + body.find(" wordcraft + application/vnd.openxmlformats-officedocument.wordprocessingml.document + application/vnd.ms-word.document.macroEnabled.12 + application/vnd.openxmlformats-officedocument.wordprocessingml.template + application/vnd.ms-word.template.macroEnabled.12 + application/msword + application/msword-template + application/vnd.oasis.opendocument.text + application/rtf diff --git a/packaging/macos/Info.plist.in b/packaging/macos/Info.plist.in index 13024e5d8..1dce256c4 100644 --- a/packaging/macos/Info.plist.in +++ b/packaging/macos/Info.plist.in @@ -59,6 +59,7 @@ org.openxmlformats.wordprocessingml.document org.openxmlformats.wordprocessingml.template + com.microsoft.word.doc org.oasis-open.opendocument.text public.rtf net.daringfireball.markdown diff --git a/packaging/windows/wordcraft.wxs b/packaging/windows/wordcraft.wxs index 620f28943..c0fa1f0aa 100644 --- a/packaging/windows/wordcraft.wxs +++ b/packaging/windows/wordcraft.wxs @@ -76,6 +76,80 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/xtask/src/layers.rs b/xtask/src/layers.rs index 9341e2cdf..a1f43d86f 100644 --- a/xtask/src/layers.rs +++ b/xtask/src/layers.rs @@ -169,13 +169,11 @@ pub fn check(crates: &[Crate]) -> Vec { continue; } match classify(&d.name) { - // Unregistered deps are reported on their own entry. - None => {} - Some(Class::Testkit) => { - if d.kind != DepKind::Dev { - out.push(Violation::TestkitAsNormalDep { krate: c.name.clone() }); - } + Some(Class::Testkit) if d.kind != DepKind::Dev => { + out.push(Violation::TestkitAsNormalDep { krate: c.name.clone() }); } + // Unregistered deps are reported on their own entry; testkit dev deps are allowed. + None | Some(Class::Testkit) => {} Some(dc) => { let to = dc.layer().unwrap_or(u8::MAX); if to >= layer && !(to == layer && intra_layer_allowed(&c.name, &d.name)) {