fonts: cut names read from a font's name table to 128 characters

A name record can hold 64K characters, and every face copies its family into each localized name and every variable-font named instance (with their localized names again), so a small font could claim hundreds of megabytes of name strings before any later check. name() now keeps at most MAX_NAME_CHARS characters, the limit localized names already used, so every copy stays small.
Release 0.15.0
2026-11-03 00:40:33 -06:00 · 2026-11-02 20:36:38 -05:00 · 2026-11-02 19:25:56 -05:00 · 2026-11-02 17:32:04 -05:00 · 2026-11-02 16:53:31 -05:00 · 2026-11-02 16:52:47 -05:00
217 changed files with 8739 additions and 60501 deletions
@@ -6,6 +6,7 @@ on:
  pull_request:
    paths:
      - 'Cargo*'
+      - 'craft-i18n.rev'
      - 'apps/wordcraft/**'
      - 'apps/wordcraft-web/**'
      - 'crates/ui-egui/**'
@@ -14,6 +15,7 @@ on:
    branches: [main]
    paths:
      - 'Cargo*'
+      - 'craft-i18n.rev'
      - 'apps/wordcraft/**'
      - 'apps/wordcraft-web/**'
      - 'crates/ui-egui/**'
@@ -41,6 +43,26 @@ jobs:
      CARGO_INCREMENTAL: '0'
    steps:
      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          persist-credentials: false
+
+      # The interface catalogs and craft-proof's Language data come from craft-i18n at the commit in
+      # craft-i18n.rev, so these tests build with the same data as the gate (ci.yml).
+      - name: Check out craft-i18n
+        uses: storytold/craft-common-actions/actions/checkout-craft-i18n@main
+
+      # The catalogs include CJK interface languages, so the UI tests need the suite's fonts, as in
+      # the gate (same commit as ci.yml and release.yml).
+      - name: Check out craft-fonts
+        uses: storytold/craft-common-actions/actions/checkout-craft-fonts@main
+        with:
+          ref: b076417b5880bd4154bd22f00695b1d28c6bea83
+
+      - name: Build with craft-fonts
+        shell: bash
+        run: |
+          echo "CRAFT_FONTS_DIR=$GITHUB_WORKSPACE/craft-fonts" >> "$GITHUB_ENV"
+          echo "CRAFT_FONTS_REQUIRED=1" >> "$GITHUB_ENV"

      - name: Install Linux build and software-rendering dependencies
        if: runner.os == 'Linux'
@@ -20,7 +20,7 @@ People trust WordCraft with their writing; a crash loses their work. **This outr
 - Never copy, trace, screenshot-crop or recreate icons, images, templates, themes, clip art, fonts, cursors, sounds or any other asset from Microsoft Word/Office or any Microsoft, Adobe, Avid or Autodesk product. Not even "temporarily".
 - Every asset must be **original** (drawn in code — prefer this: `crates/ui-egui/src/icons.rs`), **public domain / CC0**, **OSI open source**, or **Creative Commons that allows redistribution**, or licensed open source by a contributor who made it.
 - **Every asset file has a row in [`ATTRIBUTION.md`](ATTRIBUTION.md)** (author, source, licence). `cargo xtask assets` (part of `cargo xtask ci`) fails otherwise. Add the row in the same commit as the file.
- Fonts: new font files go to [`storytold/craft-fonts`](https://github.com/storytold/craft-fonts), never this repo; it's an optional build input (`CRAFT_FONTS_DIR`). Documents asking for proprietary fonts (Calibri, Aptos, Cambria…) are rendered with installed system fonts or metric-compatible open substitutes (`crates/fonts/src/word.rs`) — never bundle proprietary fonts.
+- Fonts: new font files go to [`storytold/craft-fonts`](https://github.com/storytold/craft-fonts), never this repo; it's an optional build input (`CRAFT_FONTS_DIR`). Spelling word lists and hyphenation patterns live in [`storytold/craft-i18n`](https://github.com/storytold/craft-i18n) `language-data/`, never this repo; they are embedded from `CRAFT_I18N_DIR` (pinned by `craft-i18n.rev`, which CI checks out). Without it the app builds, but English spelling falls back to an installed Hunspell English or the built-in Moby word list (with no American/British variant table) and pattern hyphenation is off (Bulgarian, Turkish and Russian still hyphenate by rule), so proofing tests need it. Documents asking for proprietary fonts (Calibri, Aptos, Cambria…) are rendered with installed system fonts or metric-compatible open substitutes (`crates/fonts/src/word.rs`) — never bundle proprietary fonts.
 - Templates, sample documents and themes are our own text and colours (`crates/engine/src/sample.rs`, `cmd/design.rs`). Demo images are public domain, with sources recorded.

 ## Clean room
@@ -34,11 +34,12 @@ People trust WordCraft with their writing; a crash loses their work. **This outr
 | L0 | `geom` | units, rects, measurement parsing |
 | L1 | `doc`, `fonts` | document model + edit primitives; font DB, shaping, Word font substitution |
 | L2 | `layout`, `docx`, `formats` | line breaking, pagination, tables, headers/footers, hit testing; OOXML I/O; TXT/MD/HTML/RTF/ODT |
-| L3 | `render`, `pdf`, `proof` | vello_cpu rasteriser; krilla PDF; proofing |
+| L3 | `render`, `pdf` | vello_cpu rasteriser; krilla PDF |
 | L4 | `engine` | Session, undo, **command registry**, Word feature catalog, samples, I/O dispatch |
 | L5 | `mcp` | MCP server (headless or bridged to the app) |
 | L6 | `ui-egui` | Word-style egui front end (swappable) |
 | apps | `apps/wordcraft`, `apps/wordcraft-cli`, `apps/wordcraft-web` | binaries |
+| external | `craft-proof` (craft-libs, used as `wordcraft-proof`, sits at L1) | proofing: spelling, grammar, hyphenation; its word lists and patterns are craft-i18n's `language-data/` (`CRAFT_I18N_DIR`) |

 - **Everything is a command.** User-visible behaviour = a `CommandSpec` in `crates/engine/src/cmd/*.rs` (id, label, ribbon location, shortcut, params doc, `enabled`, `run`) + tests. UI-only commands (`ui.*`) live in `crates/ui-egui/src/lib.rs`. The ribbon, shortcuts, command search, CLI, control channel and MCP all dispatch by id. Programmatic calls never open dialogs.
 - **Parity is measured:** `crates/engine/src/catalog.rs` lists Word's ribbon/menu features with command ids. `cargo xtask parity` writes `docs/parity-checklist.md`; a test enforces a floor that only rises. Implement missing ids to raise it.
@@ -10,15 +10,6 @@ Generated-in-code art is original and has no file to list: the UI icon set (`cra
 |---|---|---|---|---|
 | `egui` dependency: `epaint_default_fonts/fonts/egui-icons.ttf` | John Slegers (emoji-icon-font) | https://github.com/storytold/egui/blob/173cc7375d35dc5fbc5b5d0994fdf787f547c5d1/crates/epaint_default_fonts/fonts/egui-icons.txt; original https://github.com/jslegers/emoji-icon-font | MIT (full copyright and permission text in NOTICE) | Small egui special-icon subset; embedded by default_fonts, retained in FontDefinitions::default() |
 | `egui` dependency: `epaint_default_fonts/fonts/emoji-icon-font.ttf` | John Slegers | https://github.com/storytold/egui/blob/173cc7375d35dc5fbc5b5d0994fdf787f547c5d1/crates/epaint_default_fonts/fonts/emoji-icon-font-mit-license.txt; original https://github.com/jslegers/emoji-icon-font | MIT (full copyright and permission text in NOTICE) | Parent monochrome icon/emoji face embedded by monochrome_emoji_fonts; separate from the egui-icons subset |
-| `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 | ~240 lowercase English base forms supplementing spelling (#520), including British spellings (#40) |
-| `assets/spelling/en-variants.txt` | WordCraft contributors | original list written for this repository from general knowledge of English spelling; no other dictionary or variant list copied | MIT OR Apache-2.0 | ~280 American/British spelling pairs (-or/-our, -er/-re, -ize/-ise, -yze/-yse, -ense/-ence, doubled l, ae/oe…) with a policy per English: en-US, en-CA and the British-spelling Englishes (#1281, #1045) |
-| `assets/spelling/ru-extra.txt` | WordCraft contributors | original word list written for this repository (common modern Russian words older Hunspell dictionaries lack, with their declension classes); no other dictionary copied | MIT OR Apache-2.0 | ~200 Russian base forms supplementing spelling; their forms are generated by `crates/proof/src/ru/extra.rs` |
-| `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/hyphenation/it.pat` | Claudio Beccari (Italian TeX hyphenation patterns, 2008–2014); hyph-utf8 package | https://github.com/hyphenation/tex-hyphen (`hyph-utf8/tex/generic/hyph-utf8/patterns/tex/hyph-it.tex`, SHA-256 `6ce56ed6ed2ca7c0688838e366e6595e0429f984f4fc6e96ec045cebb1e76a47`), bundled unmodified | MIT (the file is offered under LPPL 1.3+ or MIT; WordCraft uses MIT, whose notice is in the file's header) | 355 Liang patterns, no exception words; Italian hyphenation |
-| `assets/hyphenation/de.pat` | Deutschsprachige Trennmustermannschaft: Stephan Hennig, Werner Lemberg, Günter Milde, Sander van Geloven, Georg Pfeiffer, Gisbert W. Selke, Tobias Wendorf, Keno Wehr (German TeX hyphenation patterns, reformed spelling, 2013–2024); hyph-utf8 package | https://github.com/hyphenation/tex-hyphen (`hyph-utf8/tex/generic/hyph-utf8/patterns/tex/hyph-de-1996.tex`, SHA-256 `374ad1ce3263f8a2791ec070ef9b516c532dae64456f3fb656e617e8ae53a9d3`), bundled unmodified | MIT (the notice is in the file's header) | 36,709 Liang patterns, no exception words; German hyphenation (de-DE, de-AT, de-CH) |
 | `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 `<image>`, 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") |
@@ -70,3 +61,4 @@ Generated-in-code art is original and has no file to list: the UI icon set (`cra
 | `docs/brand/artcraft-mark.svg` | ArtCraft Team | ArtCraft brand kit (project owner) | ArtCraft trademark, `docs/brand/LICENSE-brand.txt` | Not open source; usable unmodified only |
 | 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 |
 | craft-i18n (optional build input, not files in this repo): the interface translations | see https://github.com/storytold/craft-i18n/blob/main/ATTRIBUTION.md, which credits every translator of WordCraft's former in-repo catalogs (`crates/ui-egui/src/i18n/*.tsv`, migrated at `wordcraft@806a945`) | embedded only when built with `CRAFT_I18N_DIR`, at the commit in `craft-i18n.rev` (all official releases) | MIT OR Apache-2.0 | Interface languages other than English; clean-room translations from the English interface labels |
+| craft-i18n Language data (optional build input, not files in this repo): the spelling word lists and hyphenation patterns (`language-data/`: the Moby Hyphenator II list and our patterns for English, `en-extra.txt`, `en-variants.txt`, `ru-extra.txt`, and the hyph-utf8 German, Polish, Portuguese and Italian patterns) | see the Language data section of https://github.com/storytold/craft-i18n/blob/main/ATTRIBUTION.md (authors, upstream sources with SHA-256, licences); moved there from `assets/hyphenation/` and `assets/spelling/` byte for byte | embedded by craft-proof (storytold/craft-libs) only when built with `CRAFT_I18N_DIR`, at the commit in `craft-i18n.rev` (all official releases) | Public Domain / CC0-1.0, MIT, BSD-3-Clause, MIT OR Apache-2.0 (per file; see NOTICE) | Spelling and hyphenation; without it English spelling falls back to an installed Hunspell English or craft-proof's built-in Moby list, and pattern hyphenation is off |
@@ -20,7 +20,7 @@ People trust WordCraft with their writing; a crash loses their work. **This outr
 - Never copy, trace, screenshot-crop or recreate icons, images, templates, themes, clip art, fonts, cursors, sounds or any other asset from Microsoft Word/Office or any Microsoft, Adobe, Avid or Autodesk product. Not even "temporarily".
 - Every asset must be **original** (drawn in code — prefer this: `crates/ui-egui/src/icons.rs`), **public domain / CC0**, **OSI open source**, or **Creative Commons that allows redistribution**, or licensed open source by a contributor who made it.
 - **Every asset file has a row in [`ATTRIBUTION.md`](ATTRIBUTION.md)** (author, source, licence). `cargo xtask assets` (part of `cargo xtask ci`) fails otherwise. Add the row in the same commit as the file.
- Fonts: new font files go to [`storytold/craft-fonts`](https://github.com/storytold/craft-fonts), never this repo; it's an optional build input (`CRAFT_FONTS_DIR`). Documents asking for proprietary fonts (Calibri, Aptos, Cambria…) are rendered with installed system fonts or metric-compatible open substitutes (`crates/fonts/src/word.rs`) — never bundle proprietary fonts.
+- Fonts: new font files go to [`storytold/craft-fonts`](https://github.com/storytold/craft-fonts), never this repo; it's an optional build input (`CRAFT_FONTS_DIR`). Spelling word lists and hyphenation patterns live in [`storytold/craft-i18n`](https://github.com/storytold/craft-i18n) `language-data/`, never this repo; they are embedded from `CRAFT_I18N_DIR` (pinned by `craft-i18n.rev`, which CI checks out). Without it the app builds, but English spelling falls back to an installed Hunspell English or the built-in Moby word list (with no American/British variant table) and pattern hyphenation is off (Bulgarian, Turkish and Russian still hyphenate by rule), so proofing tests need it. Documents asking for proprietary fonts (Calibri, Aptos, Cambria…) are rendered with installed system fonts or metric-compatible open substitutes (`crates/fonts/src/word.rs`) — never bundle proprietary fonts.
 - Templates, sample documents and themes are our own text and colours (`crates/engine/src/sample.rs`, `cmd/design.rs`). Demo images are public domain, with sources recorded.

 ## Clean room
@@ -34,11 +34,12 @@ People trust WordCraft with their writing; a crash loses their work. **This outr
 | L0 | `geom` | units, rects, measurement parsing |
 | L1 | `doc`, `fonts` | document model + edit primitives; font DB, shaping, Word font substitution |
 | L2 | `layout`, `docx`, `formats` | line breaking, pagination, tables, headers/footers, hit testing; OOXML I/O; TXT/MD/HTML/RTF/ODT |
-| L3 | `render`, `pdf`, `proof` | vello_cpu rasteriser; krilla PDF; proofing |
+| L3 | `render`, `pdf` | vello_cpu rasteriser; krilla PDF |
 | L4 | `engine` | Session, undo, **command registry**, Word feature catalog, samples, I/O dispatch |
 | L5 | `mcp` | MCP server (headless or bridged to the app) |
 | L6 | `ui-egui` | Word-style egui front end (swappable) |
 | apps | `apps/wordcraft`, `apps/wordcraft-cli`, `apps/wordcraft-web` | binaries |
+| external | `craft-proof` (craft-libs, used as `wordcraft-proof`, sits at L1) | proofing: spelling, grammar, hyphenation; its word lists and patterns are craft-i18n's `language-data/` (`CRAFT_I18N_DIR`) |

 - **Everything is a command.** User-visible behaviour = a `CommandSpec` in `crates/engine/src/cmd/*.rs` (id, label, ribbon location, shortcut, params doc, `enabled`, `run`) + tests. UI-only commands (`ui.*`) live in `crates/ui-egui/src/lib.rs`. The ribbon, shortcuts, command search, CLI, control channel and MCP all dispatch by id. Programmatic calls never open dialogs.
 - **Parity is measured:** `crates/engine/src/catalog.rs` lists Word's ribbon/menu features with command ids. `cargo xtask parity` writes `docs/parity-checklist.md`; a test enforces a floor that only rises. Implement missing ids to raise it.
@@ -1011,6 +1011,45 @@ dependencies = [
 "libc",
 ]

+[[package]]
+name = "craft-atomic"
+version = "0.1.0"
+source = "git+https://github.com/storytold/craft-libs?rev=762a06a4c6b48b522b7d4de84a87917ef90265e6#762a06a4c6b48b522b7d4de84a87917ef90265e6"
+dependencies = [
+ "rustix 1.1.5",
+]
+
+[[package]]
+name = "craft-font-catalog"
+version = "0.1.0"
+source = "git+https://github.com/storytold/craft-libs?rev=762a06a4c6b48b522b7d4de84a87917ef90265e6#762a06a4c6b48b522b7d4de84a87917ef90265e6"
+dependencies = [
+ "fontique",
+ "skrifa 0.47.0",
+ "thiserror 2.0.21",
+ "windows-registry",
+]
+
+[[package]]
+name = "craft-font-store"
+version = "0.1.0"
+source = "git+https://github.com/storytold/craft-libs?rev=762a06a4c6b48b522b7d4de84a87917ef90265e6#762a06a4c6b48b522b7d4de84a87917ef90265e6"
+dependencies = [
+ "craft-atomic",
+ "craft-font-catalog",
+ "flate2",
+ "log",
+ "ruzstd",
+ "serde",
+ "serde_json",
+ "sha2 0.10.9",
+ "skrifa 0.47.0",
+ "tar",
+ "thiserror 2.0.21",
+ "toml",
+ "ureq",
+]
+
 [[package]]
 name = "craft-fonts-build"
 version = "0.1.0"
@@ -1061,6 +1100,17 @@ dependencies = [
 "winsafe",
 ]

+[[package]]
+name = "craft-proof"
+version = "0.1.0"
+source = "git+https://github.com/storytold/craft-libs?rev=89fa1559e94378ecef85cd7e4681400f4df053e0#89fa1559e94378ecef85cd7e4681400f4df053e0"
+dependencies = [
+ "log",
+ "miniz_oxide 0.9.1",
+ "toml",
+ "unicode-normalization",
+]
+
 [[package]]
 name = "craft-ui"
 version = "0.1.0"
@@ -1801,6 +1851,28 @@ dependencies = [
 "bytemuck",
 ]

+[[package]]
+name = "fontique"
+version = "0.11.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6688bc1294fe7117d788937b6c53480169b29c566954af490830d4c09da9516a"
+dependencies = [
+ "hashbrown 0.17.1",
+ "linebender_resource_handle",
+ "memmap2",
+ "objc2 0.6.5",
+ "objc2-core-foundation",
+ "objc2-core-text",
+ "objc2-foundation 0.3.2",
+ "parlance",
+ "read-fonts 0.41.0",
+ "roxmltree",
+ "smallvec",
+ "windows",
+ "windows-core",
+ "yeslogic-fontconfig-sys",
+]
+
 [[package]]
 name = "foreign-types"
 version = "0.5.0"
@@ -3334,6 +3406,16 @@ dependencies = [
 "objc2-foundation 0.2.2",
 ]

+[[package]]
+name = "objc2-core-text"
+version = "0.3.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0cde0dfb48d25d2b4862161a4d5fcc0e3c24367869ad306b0c9ec0073bfed92d"
+dependencies = [
+ "bitflags 2.13.2",
+ "objc2-core-foundation",
+]
+
 [[package]]
 name = "objc2-encode"
 version = "4.1.0"
@@ -3605,6 +3687,12 @@ dependencies = [
 "windows-link",
 ]

+[[package]]
+name = "parlance"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "73072a9470554c8e794778acece2312661218aaabb635600957d952cf01c716e"
+
 [[package]]
 name = "pdf-writer"
 version = "0.15.0"
@@ -4516,6 +4604,12 @@ dependencies = [
 "wait-timeout",
 ]

+[[package]]
+name = "ruzstd"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "69e4ce4a039f42bf25920b95667e4d789e98749ab11bbb8d2c2936f788421bb5"
+
 [[package]]
 name = "ryu"
 version = "1.0.23"
@@ -5723,6 +5817,7 @@ dependencies = [
 "rustls-platform-verifier",
 "ureq-proto",
 "utf8-zero",
+ "webpki-roots",
 ]

 [[package]]
@@ -6106,6 +6201,15 @@ dependencies = [
 "rustls-pki-types",
 ]

+[[package]]
+name = "webpki-roots"
+version = "1.0.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f5f280bbb2b287bc8e7de7e3852945a0ff50d1bbcb5ed42b6c7a568c546df171"
+dependencies = [
+ "rustls-pki-types",
+]
+
 [[package]]
 name = "weezl"
 version = "0.1.12"
@@ -6648,12 +6752,13 @@ dependencies = [

 [[package]]
 name = "wordcraft"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "arboard",
 "chrono",
 "clipboard-win",
 "craft-print",
+ "craft-proof",
 "craft-update",
 "dispatch2",
 "eframe",
@@ -6676,14 +6781,13 @@ dependencies = [
 "wordcraft-doc",
 "wordcraft-engine",
 "wordcraft-fonts",
- "wordcraft-proof",
 "wordcraft-ui-egui",
 "zbus",
 ]

 [[package]]
 name = "wordcraft-cli"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "serde_json",
 "sys-locale",
@@ -6696,11 +6800,11 @@ dependencies = [

 [[package]]
 name = "wordcraft-control-key"
-version = "0.14.0"
+version = "0.15.0"

 [[package]]
 name = "wordcraft-doc"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "base64 0.23.1",
 "log",
@@ -6714,7 +6818,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-docbin"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "cfb",
 "chrono",
@@ -6731,7 +6835,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-docx"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "aes",
 "base64 0.23.1",
@@ -6755,10 +6859,13 @@ dependencies = [

 [[package]]
 name = "wordcraft-engine"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "cfb",
 "chrono",
+ "craft-font-catalog",
+ "craft-font-store",
+ "craft-proof",
 "encoding_rs",
 "getrandom 0.3.4",
 "image",
@@ -6778,7 +6885,6 @@ dependencies = [
 "wordcraft-geom",
 "wordcraft-layout",
 "wordcraft-pdf",
- "wordcraft-proof",
 "wordcraft-render",
 "zeroize",
 "zip 3.0.0",
@@ -6786,7 +6892,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-fonts"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "brotli-decompressor",
 "harfrust 0.13.3",
@@ -6800,7 +6906,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-formats"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "cfb",
 "chardetng",
@@ -6819,7 +6925,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-geom"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "image",
 "serde",
@@ -6827,8 +6933,9 @@ dependencies = [

 [[package]]
 name = "wordcraft-layout"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
+ "craft-proof",
 "image",
 "js-sys",
 "kurbo",
@@ -6842,12 +6949,11 @@ dependencies = [
 "wordcraft-fonts",
 "wordcraft-geom",
 "wordcraft-metafile",
- "wordcraft-proof",
 ]

 [[package]]
 name = "wordcraft-mcp"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "log",
 "serde_json",
@@ -6859,7 +6965,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-metafile"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "kurbo",
 "thiserror 2.0.21",
@@ -6867,7 +6973,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-pdf"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "chrono",
 "craft-pdf",
@@ -6886,19 +6992,9 @@ dependencies = [
 "wordcraft-metafile",
 ]

-[[package]]
-name = "wordcraft-proof"
-version = "0.14.0"
-dependencies = [
- "log",
- "miniz_oxide 0.9.1",
- "proptest",
- "unicode-normalization",
-]
-
 [[package]]
 name = "wordcraft-render"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "image",
 "log",
@@ -6912,7 +7008,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-ui-egui"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "cfb",
 "chrono",
@@ -6943,7 +7039,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-web"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "eframe",
 "egui",
@@ -6962,7 +7058,7 @@ dependencies = [

 [[package]]
 name = "wordcraft-zotero"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "log",
 "proptest",
@@ -7125,11 +7221,22 @@ checksum = "9440ea3e5aeabb0ac63af70daf835274065238cdd0cec83418f417eae38bacee"

 [[package]]
 name = "xtask"
-version = "0.14.0"
+version = "0.15.0"
 dependencies = [
 "serde_json",
 ]

+[[package]]
+name = "yeslogic-fontconfig-sys"
+version = "6.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1d8b8abf912b9a29ff112e1671c97c33636903d13a69712037190e6805af4f76"
+dependencies = [
+ "dlib",
+ "once_cell",
+ "pkg-config",
+]
+
 [[package]]
 name = "yoke"
 version = "0.8.4"
@@ -4,7 +4,7 @@ members = ["crates/*", "apps/*", "xtask"]
 default-members = ["crates/*", "apps/wordcraft", "apps/wordcraft-cli", "xtask"]

 [workspace.package]
-version = "0.14.0"
+version = "0.15.0"
 edition = "2024"
 license = "MIT OR Apache-2.0"
 rust-version = "1.95"
@@ -21,7 +21,16 @@ wordcraft-docbin = { path = "crates/docbin" }
 wordcraft-formats = { path = "crates/formats" }
 wordcraft-render = { path = "crates/render" }
 wordcraft-pdf = { path = "crates/pdf" }
-wordcraft-proof = { path = "crates/proof" }
+# Proofing is craft-libs' craft-proof (it was crates/proof); its Language data is craft-i18n's
+# language-data/, embedded from CRAFT_I18N_DIR (craft-i18n.rev). The package keeps the
+# `wordcraft-proof` name here so `wordcraft_proof::` paths are unchanged.
+# craft-proof is the head of storytold/craft-libs#495 (review fixes), which contains #485 (CP4.2,
+# Language packs, on #483 SCOWL English and #482); the font store's `language-pack` kind is #486
+# (CP4.1). GitHub serves PR heads from storytold/craft-libs, and merged with merge commits they
+# stay in its main.
+wordcraft-proof = { package = "craft-proof", git = "https://github.com/storytold/craft-libs", rev = "89fa1559e94378ecef85cd7e4681400f4df053e0", version = "=0.1.0" }
+craft-font-store = { git = "https://github.com/storytold/craft-libs", rev = "762a06a4c6b48b522b7d4de84a87917ef90265e6", version = "=0.1.0" }
+craft-font-catalog = { git = "https://github.com/storytold/craft-libs", rev = "762a06a4c6b48b522b7d4de84a87917ef90265e6", version = "=0.1.0" }
 wordcraft-engine = { path = "crates/engine" }
 wordcraft-mcp = { path = "crates/mcp" }
 wordcraft-zotero = { path = "crates/zotero" }
@@ -20,13 +20,77 @@ list (author, source, license) is in ATTRIBUTION.md.
  https://github.com/storytold/craft-fonts/blob/main/ATTRIBUTION.md (BIZ UDPGothic, BIZ UDMincho,
  Shippori Mincho), under the SIL Open Font License 1.1. Their licence texts ship with the
  packages as OFL-<family>.txt. Builds without CRAFT_FONTS_DIR do not include them.
- English word list (assets/hyphenation/en-us.dic) — derived from the Moby Hyphenator II list by
+- Language data (craft-i18n): the word lists, dictionaries and hyphenation patterns below are embedded at build
+  time from https://github.com/storytold/craft-i18n (language-data/, CRAFT_I18N_DIR at the commit
+  in craft-i18n.rev; all official releases) by craft-proof (storytold/craft-libs). Builds without
+  CRAFT_I18N_DIR do not include them.
+- English word list (craft-i18n language-data/en-us/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
+  (craft-i18n language-data/en-us/hyphenation.pat) were trained from it and are released as CC0-1.0.
+- English spelling dictionaries (craft-i18n language-data/en-us, en-gb and en-ca: spelling.aff and
+  spelling.dic) — the SCOWL rel-2026.02.25 Hunspell dictionaries en_US, en_GB-ise and en_CA from
+  https://wordlist.aspell.net, bundled unmodified. Their upstream notice (embedded too, for the
+  About window) is language-data/en-us/LICENCE-SCOWL.txt in craft-i18n. The word lists:
+
+    Copyright 2000-2026 by Kevin Atkinson
+
+    Permission to use, copy, modify, distribute, and sell any part of SCOWLv2, or
+    word lists created from it, is hereby granted without fee, provided that the
+    above copyright notice appears in all copies and that both the above
+    copyright notice and this notice appear in supporting documentation.  Kevin
+    Atkinson makes no representations about the suitability of this database for
+    any purpose.  It is provided "as is" without express or implied warranty.
+
+  The Australian data:
+
+    Copyright 2016 by Benjamin Titze
+
+    Permission to use, copy, modify, distribute and sell this array, the
+    associated software, and its documentation for any purpose is hereby
+    granted without fee, provided that the above copyright notice appears
+    in all copies and that both that copyright notice and this permission
+    notice appear in supporting documentation. Benjamin Titze makes no
+    representations about the suitability of this array for any
+    purpose. It is provided "as is" without express or implied warranty.
+
+  The affix files, a modified version of Geoff Kuenning's Ispell english.aff:
+
+    Copyright 1993, Geoff Kuenning, Granada Hills, CA
+    All rights reserved.
+
+    Redistribution and use in source and binary forms, with or without
+    modification, are permitted provided that the following conditions
+    are met:
+
+    1. Redistributions of source code must retain the above copyright
+       notice, this list of conditions and the following disclaimer.
+    2. 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.
+    3. All modifications to the source code must be clearly marked as
+       such.  Binary redistributions based on modified source code
+       must be clearly marked as modified versions in the documentation
+       and/or other materials provided with the distribution.
+    (clause 4 removed with permission from Geoff Kuenning)
+    5. The name of Geoff Kuenning may not be used to endorse or promote
+       products derived from this software without specific prior
+       written permission.
+
+    THIS SOFTWARE IS PROVIDED BY GEOFF KUENNING 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 GEOFF KUENNING OR CONTRIBUTORS 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.
+- Polish hyphenation patterns (craft-i18n language-data/pl/hyphenation.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
+- Portuguese hyphenation patterns (craft-i18n language-data/pt/hyphenation.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,
@@ -52,7 +116,7 @@ list (author, source, license) is in ATTRIBUTION.md.
    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.
- German hyphenation patterns (assets/hyphenation/de.pat) — hyph-de-1996.tex (reformed spelling)
+- German hyphenation patterns (craft-i18n language-data/de/hyphenation.pat) — hyph-de-1996.tex (reformed spelling)
  from the hyph-utf8 package, bundled unmodified and used under the MIT License; the copyright
  and permission notice is in the file:

@@ -74,7 +138,7 @@ list (author, source, license) is in ATTRIBUTION.md.
    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.
- Italian hyphenation patterns (assets/hyphenation/it.pat) — hyph-it.tex from the hyph-utf8
+- Italian hyphenation patterns (craft-i18n language-data/it/hyphenation.pat) — hyph-it.tex from the hyph-utf8
  package, Copyright (C) 2008-2011 Claudio Beccari, bundled unmodified and used under the MIT
  License (offered under LPPL 1.3+ or MIT); the copyright and permission notice is in the file:

@@ -140,6 +140,19 @@ cargo run --release -p wordcraft -- --sample        # the desktop app with the s
 cargo run --release -p wordcraft -- report.docx     # open a document
 ```

+The interface translations and the spelling and hyphenation data come from
+[craft-i18n](https://github.com/storytold/craft-i18n), an optional build input. To build with them,
+check it out at the commit in `craft-i18n.rev` and point `CRAFT_I18N_DIR` at it (an absolute
+path); without it WordCraft builds with an English interface, English spelling falls back to an
+installed Hunspell English or craft-proof's built-in word list, and only Bulgarian, Turkish and
+Russian hyphenate (by rule):
+
+```sh
+git clone https://github.com/storytold/craft-i18n ../craft-i18n
+git -C ../craft-i18n checkout "$(cat craft-i18n.rev)"
+CRAFT_I18N_DIR="$PWD/../craft-i18n" cargo run --release -p wordcraft -- --sample
+```
+
 Command line:

 ```sh
@@ -194,7 +207,7 @@ covers every target starting with it (`wordcraft*=debug`). The logger is
 | Layer | Crate | Job |
 |---|---|---|
 | L0 | `wordcraft-geom` | units and measurements |
-| L1 | `wordcraft-doc`, `wordcraft-fonts`, `wordcraft-proof` | document model and editing; fonts and shaping; spelling, grammar, hyphenation |
+| L1 | `wordcraft-doc`, `wordcraft-fonts`, `wordcraft-proof` ([`craft-proof`](https://github.com/storytold/craft-libs/tree/main/crates/proof) from craft-libs) | document model and editing; fonts and shaping; spelling, grammar, hyphenation (data: craft-i18n `language-data/`, `CRAFT_I18N_DIR`) |
 | L2 | `wordcraft-layout`, `wordcraft-docx`, `wordcraft-formats` | line breaking, pagination, tables, notes, hit testing; OOXML; ODT/RTF/HTML/Markdown/TXT |
 | L3 | `wordcraft-render`, `wordcraft-pdf` | rasteriser (vello_cpu); PDF ([craft-pdf](docs/shared-pdf.md), Krilla) |
 | L4 | `wordcraft-engine` | session, undo, 422 commands, document feature catalog |
@@ -758,6 +758,12 @@ fn parity_markdown(p: &Value) -> String {

 fn main() -> ExitCode {
    let args: Vec<OsString> = std::env::args_os().skip(1).collect();
+    // Language packs installed in the suite's font store are used; `proof.installPack` installs
+    // only while the person has turned downloading on in the app.
+    wordcraft_engine::packs::use_font_store(false);
+    if let Some(source) = wordcraft_engine::packs::index_override() {
+        eprintln!("wordcraft-cli: Language packs come from a test source: {source}");
+    }
    match run(&args) {
        Ok(()) | Err(Failure::StdoutClosed) => ExitCode::SUCCESS,
        Err(Failure::Error(e)) => {
@@ -9,7 +9,8 @@ use wordcraft_engine::Session;
 use wordcraft_ui_egui::async_saves::{AsyncSaves, SaveOutcome};
 use wordcraft_ui_egui::{Inbox, Services, WordApp};

-const DOC_EXTS: &[&str] = &["docx", "docm", "dotx", "dotm", "doc", "dot", "odt", "rtf", "txt", "md", "html", "htm", "tex", "json", "xml"];
+const DOC_EXTS: &[&str] =
+    &["docx", "docm", "dotx", "dotm", "doc", "dot", "odt", "rtf", "txt", "md", "html", "htm", "mht", "mhtml", "tex", "json", "xml"];
 const IMAGE_EXTS: &[&str] = &["png", "jpg", "jpeg", "gif", "webp", "bmp", "wmf", "emf"];
 const RECIPIENT_EXTS: &[&str] = wordcraft_engine::cmd::mailings::RECIPIENT_EXTENSIONS;
 mod font_store;
@@ -490,6 +490,8 @@ fn main() -> eframe::Result {
            app.set_startup_update_check(true);
            load_prefs(&mut app);
            load_user_dictionary(&mut app);
+            // Language packs from the suite's font store; an Install click is the person's own.
+            wordcraft_engine::packs::use_font_store(true);
            // 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() {
@@ -1,431 +0,0 @@
-% title: Hyphenation patterns for Italian
-% copyright: Copyright (C) 2008-2011 Claudio Beccari
-% notice: This file is part of the hyph-utf8 package.
-%     See http://www.hyphenation.org/tex for more information.
-% language:
-%     name: Italian
-%     tag: it
-% version: 4.9 2014/04/22
-% authors:
-%   -
-%     name: Claudio Beccari
-%     contact: claudio.beccari (at) gmail.com
-% licence:
-%     - This file is available under any of the following licences:
-%     -
-%         name: LPPL
-%         version: 1.3
-%         or_later: true
-%         url: http://www.latex-project.org/lppl.txt
-%         status: maintained
-%         maintainer: Claudio Beccari, e-mail claudio dot beccari at gmail dot com
-%     -
-%         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.
-% hyphenmins:
-%     typesetting:
-%         left: 2
-%         right: 2
-% changes:
-%     - 2014-04-22 - Add few patterns involving `h'
-%     - 2011-08-16 - Change the licence from GNU LGPL into LPPL v1.3.
-%     - 2010-05-24 - Fix for Italian patterns for proper hyphenation of -ich and Ljubljana.
-%     - 2008-06-09 - Import of original ithyph.tex into hyph-utf8 package.
-%     - 2008-03-08 - (last change in ithyph.tex)
-% texlive:
-%     encoding: ascii
-%     babelname: italian
-%     legacy_patterns: ithyph.tex
-%     message: Italian hyphenation patterns
-%     description: |-
-%         Hyphenation patterns for Italian in ASCII encoding.
-%         Compliant with the Recommendation UNI 6461 on hyphenation
-%         issued by the Italian Standards Institution
-%         (Ente Nazionale di Unificazione UNI).
-% ==========================================
-%
-% These hyphenation patterns for the Italian language are supposed to comply
-% with the Recommendation UNI 6461 on hyphenation issued by the Italian
-% Standards Institution (Ente Nazionale di Unificazione UNI).  No guarantee
-% or declaration of fitness to any particular purpose is given and any
-% liability is disclaimed.
-%
-\patterns{
-.a3p2n               % After the Garzanti dictionary: a-pnea, a-pnoi-co,...
-.anti1
-.anti3m2n
-.bio1
-.ca4p3s
-.circu2m1
-.contro1
-.di2s3cine
-.e2x1eu
-.fran2k3
-.free3
-.li3p2sa
-.narco1
-.opto1
-.orto3p2
-.para1
-.ph2l
-.ph2r
-.poli3p2
-.pre1
-.p2s
-.re1i2scr
-.sha2re3
-.tran2s3c
-.tran2s3d
-.tran2s3l
-.tran2s3n
-.tran2s3p
-.tran2s3r
-.tran2s3t
-.su2b3lu
-.su2b3r
-.wa2g3n
-.wel2t1
-2'2
-a1ia
-a1ie
-a1io
-a1iu
-a1uo
-a1ya
-2at.
-e1iu
-e2w
-o1ia
-o1ie
-o1io
-o1iu
-1b
-2bb
-2bc
-2bd
-2bf
-2bm
-2bn
-2bp
-2bs
-2bt
-2bv
-b2l
-b2r
-2b.
-2b'
-1c
-2cb
-2cc
-2cd
-2cf
-2ck
-2cm
-2cn
-2cq
-2cs
-2ct
-2cz
-2chh
-c2h
-2ch.
-2ch'.
-2ch''.
-2chb
-ch2r
-2chn
-c2l
-c2r
-2c.
-2c'
-.c2
-1d
-2db
-2dd
-2dg
-2dl
-2dm
-2dn
-2dp
-d2r
-2ds
-2dt
-2dv
-2dw
-2d.
-2d'
-.d2
-1f
-2fb
-2fg
-2ff
-2fn
-f2l
-f2r
-2fs
-2ft
-2f.
-2f'
-1g
-2gb
-2gd
-2gf
-2gg
-g2h
-g2l
-2gm
-g2n
-2gp
-g2r
-2gs
-2gt
-2gv
-2gw
-2gz
-2gh2t
-2g.
-2g'
-.h2
-1h
-2hb
-2hd
-2hh
-hi3p2n
-h2l
-2hm
-2hn
-2hr
-2hv
-2h.
-2h'
-.j2
-1j
-2j.
-2j'
-.k2
-1k
-2kg
-2kf
-k2h
-2kk
-k2l
-2km
-k2r
-2ks
-2kt
-2k.
-2k'
-1l
-2lb
-2lc
-2ld
-2l3f2
-2lg
-l2h
-l2j
-2lk
-2ll
-2lm
-2ln
-2lp
-2lq
-2lr
-2ls
-2lt
-2lv
-2lw
-2lz
-2l.
-2l'.
-2l''
-1m
-2mb
-2mc
-2mf
-2ml
-2mm
-2mn
-2mp
-2mq
-2mr
-2ms
-2mt
-2mv
-2mw
-2m.
-2m'
-1n
-2nb
-2nc
-2nd
-2nf
-2ng
-2nk
-2nl
-2nm
-2nn
-2np
-2nq
-2nr
-2ns
-n2s3fer
-2nt
-2nv
-2nz
-n2g3n
-2nheit
-2n.
-2n'
-1p
-2pd
-p2h
-p2l
-2pn
-3p2ne
-2pp
-p2r
-2ps
-3p2sic
-2pt
-2pz
-2p.
-2p'
-1q
-2qq
-2q.
-2q'
-1r
-2rb
-2rc
-2rd
-2rf
-r2h
-2rg
-2rk
-2rl
-2rm
-2rn
-2rp
-2rq
-2rr
-2rs
-2rt
-r2t2s3
-2rv
-2rx
-2rw
-2rz
-2r.
-2r'
-1s2
-2shm
-2sh.
-2sh'
-2s3s
-s4s3m
-2s3p2n
-2stb
-2stc
-2std
-2stf
-2stg
-2stm
-2stn
-2stp
-2sts
-2stt
-2stv
-2sz
-4s.
-4s'.
-4s''
-.t2
-1t
-2tb
-2tc
-2td
-2tf
-2tg
-t2h
-2th.
-t2l
-2tm
-2tn
-2tp
-t2r
-t2s
-3t2sch
-2tt
-t2t3s
-2tv
-2tw
-t2z
-2tzk
-tz2s
-2t.
-2t'.
-2t''
-1v
-2vc
-v2l
-v2r
-2vv
-2v.
-2v'.
-2v''
-1w
-w2h
-wa2r
-2w1y
-2w.
-2w'
-1x
-2xb
-2xc
-2xf
-2xh
-2xm
-2xp
-2xt
-2xw
-2x.
-2x'
-y1ou
-y1i
-1z
-2zb
-2zd
-2zl
-2zn
-2zp
-2zt
-2zs
-2zv
-2zz
-2z.
-2z'.
-2z''
-.z2
-}                          % Pattern end
-
@@ -1,503 +0,0 @@
-% 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
-}
@@ -1,241 +0,0 @@
-# English spelling supplement: common words the public-domain Moby list (1990s) lacks.
-# Written by WordCraft contributors (MIT OR Apache-2.0); see ATTRIBUTION.md.
-# One lowercase base form per line, sorted. Plurals, -ed, -ing, -er, -ly… come from the
-# inflection rules in crates/proof/src/lib.rs, so list "desktop", not "desktops".
-airbag
-airfare
-artcraft
-autocomplete
-autocorrect
-autofill
-autoformat
-autosave
-backend
-bitcoin
-bitmap
-blockchain
-blog
-blogger
-bluetooth
-captcha
-caregiver
-carpool
-cellphone
-chatbot
-chatroom
-checkbox
-checklist
-childcare
-clamour
-clickable
-colourful
-colourisation
-colourise
-colourless
-configurable
-copyedit
-coronavirus
-covid
-crowdfunding
-crowdsource
-crypto
-cryptocurrency
-customizable
-customization
-cyber
-cybersecurity
-cyberspace
-database
-dataset
-daycare
-desktop
-download
-downloadable
-downsize
-dropdown
-earbud
-ebook
-ecommerce
-email
-emoji
-emoticon
-endeavour
-esports
-facebook
-favicon
-fintech
-firewall
-flowchart
-foodie
-freeware
-frontend
-fundraiser
-gameplay
-gamification
-geeky
-geolocation
-gigabit
-gigabyte
-github
-glamour
-globalization
-google
-groupthink
-hackathon
-hardcode
-hashtag
-healthcare
-homepage
-homeschool
-honour
-honourable
-honourably
-hotspot
-hyperlink
-hypertext
-inbox
-incentivize
-infographic
-instagram
-internet
-interoperability
-intranet
-ipad
-iphone
-javascript
-kickstart
-kilobyte
-laptop
-lifestyle
-linkedin
-linux
-livestream
-lockdown
-login
-logout
-lowercase
-makeover
-malware
-manoeuvrable
-megabyte
-megapixel
-meme
-metadata
-microchip
-microsoft
-middleware
-mindset
-multicultural
-multiplayer
-multitask
-nerdy
-netbook
-netflix
-newsfeed
-offboard
-offline
-onboard
-online
-onscreen
-outsource
-paediatric
-paralegal
-passcode
-petabyte
-phishing
-photobomb
-playdate
-playlist
-plugin
-podcast
-popup
-prequel
-printout
-prioritise
-prioritization
-prioritize
-privatise
-ransomware
-recyclable
-rigour
-ringtone
-roadmap
-rollout
-runtime
-scalability
-screencast
-screensaver
-screenshot
-scrollbar
-selfie
-signup
-sitemap
-skillset
-smartphone
-smartwatch
-smiley
-snowboard
-spellcheck
-spellchecker
-spreadsheet
-spyware
-stakeholder
-standup
-startup
-subfolder
-sunscreen
-superfood
-superuser
-synthesise
-tagline
-takeout
-techie
-telecommute
-teleconference
-telehealth
-telemedicine
-terabyte
-textbox
-timeline
-timestamp
-toolbar
-toolkit
-touchpad
-touchscreen
-trackball
-trackpad
-travelogue
-upcycle
-upload
-uppercase
-upsell
-upskill
-usability
-username
-videoconference
-videogame
-vlog
-vlogger
-voicemail
-walkthrough
-webcam
-webcast
-webinar
-weblog
-webmaster
-webpage
-website
-whatsapp
-whiteboard
-wifi
-wiki
-wikipedia
-wireframe
-wordcraft
-workaround
-workflow
-workforce
-workplace
-workspace
-workstation
-youtube
@@ -1,305 +0,0 @@
-# English spelling variants: American and British forms of the same word, for the English
-# dialects (#1281, #1045). Written by WordCraft contributors from general knowledge of English
-# spelling (MIT OR Apache-2.0); no other word list copied. See ATTRIBUTION.md.
-#
-# Each line: POLICY AMERICAN BRITISH
-# POLICY is three letters, for American English (en-US), Canadian English (en-CA) and the
-# British-spelling Englishes (en-GB, en-AU, en-NZ, en-IE, en-ZA, en-IN…), in that order:
-#   u  only the American form is right there (the British one is flagged, American suggested)
-#   b  only the British form is right there
-#   *  both are right
-# Inflections (colours, centred, organising, organisation…) come from the rules in
-# crates/proof/src/en_variants.rs for -or/-our, -er/-re and -ize/-ise style pairs; other pairs
-# get only a plural, so list their other forms on lines of their own. When in doubt a form is
-# accepted: a false alarm is worse than a missed variant.
-
-# -or / -our
-ubb arbor arbour
-ubb ardor ardour
-ubb armor armour
-ubb armory armoury
-ubb behavior behaviour
-ubb behavioral behavioural
-ubb candor candour
-ubb clamor clamour
-ubb color colour
-ubb demeanor demeanour
-ubb discolor discolour
-ubb dishonor dishonour
-ubb endeavor endeavour
-ubb favor favour
-ubb favorite favourite
-ubb favoritism favouritism
-ubb fervor fervour
-ubb flavor flavour
-ubb harbor harbour
-ubb honor honour
-ubb humor humour
-ubb labor labour
-ubb misbehavior misbehaviour
-ubb neighbor neighbour
-ubb neighborhood neighbourhood
-ubb neighborly neighbourly
-ubb odor odour
-ubb parlor parlour
-ubb rancor rancour
-ubb rigor rigour
-ubb rumor rumour
-ubb savior saviour
-ubb savor savour
-ubb savory savoury
-ubb splendor splendour
-ubb succor succour
-ubb tumor tumour
-ubb unfavorable unfavourable
-ubb valor valour
-ubb vapor vapour
-ubb vigor vigour
-
-# -er / -re
-ubb caliber calibre
-ubb center centre
-ubb centimeter centimetre
-ubb epicenter epicentre
-ubb fiber fibre
-ubb kilometer kilometre
-ubb liter litre
-ubb luster lustre
-ubb maneuver manoeuvre
-ubb meager meagre
-ubb millimeter millimetre
-ubb miter mitre
-ubb ocher ochre
-ubb saber sabre
-ubb sepulcher sepulchre
-ubb somber sombre
-ubb specter spectre
-*bb theater theatre
-# A meter is also a measuring device in British English.
-u** meter metre
-
-# -ize / -ise (British English accepts both: Oxford spelling uses -ize)
-uu* agonize agonise
-uu* apologize apologise
-uu* authorize authorise
-uu* capitalize capitalise
-uu* categorize categorise
-uu* centralize centralise
-uu* characterize characterise
-uu* civilize civilise
-uu* colonize colonise
-uu* commercialize commercialise
-uu* computerize computerise
-uu* criticize criticise
-uu* crystallize crystallise
-uu* customize customise
-uu* decentralize decentralise
-uu* demonize demonise
-uu* deodorize deodorise
-uu* destabilize destabilise
-uu* digitize digitise
-uu* dramatize dramatise
-uu* economize economise
-uu* emphasize emphasise
-uu* energize energise
-uu* equalize equalise
-uu* familiarize familiarise
-uu* fantasize fantasise
-uu* fertilize fertilise
-uu* finalize finalise
-uu* formalize formalise
-uu* fossilize fossilise
-uu* galvanize galvanise
-uu* generalize generalise
-uu* globalize globalise
-uu* harmonize harmonise
-uu* hospitalize hospitalise
-uu* humanize humanise
-uu* hypnotize hypnotise
-uu* hypothesize hypothesise
-uu* idealize idealise
-uu* idolize idolise
-uu* immobilize immobilise
-uu* immortalize immortalise
-uu* immunize immunise
-uu* industrialize industrialise
-uu* initialize initialise
-uu* italicize italicise
-uu* itemize itemise
-uu* jeopardize jeopardise
-uu* legalize legalise
-uu* legitimize legitimise
-uu* liberalize liberalise
-uu* localize localise
-uu* magnetize magnetise
-uu* marginalize marginalise
-uu* maximize maximise
-uu* mechanize mechanise
-uu* memorize memorise
-uu* mesmerize mesmerise
-uu* minimize minimise
-uu* mobilize mobilise
-uu* modernize modernise
-uu* monopolize monopolise
-uu* moralize moralise
-uu* nationalize nationalise
-uu* naturalize naturalise
-uu* neutralize neutralise
-uu* normalize normalise
-uu* optimize optimise
-uu* organize organise
-uu* ostracize ostracise
-uu* oxidize oxidise
-uu* pasteurize pasteurise
-uu* patronize patronise
-uu* penalize penalise
-uu* personalize personalise
-uu* plagiarize plagiarise
-uu* polarize polarise
-uu* popularize popularise
-uu* pressurize pressurise
-uu* prioritize prioritise
-uu* privatize privatise
-uu* publicize publicise
-uu* pulverize pulverise
-uu* randomize randomise
-uu* rationalize rationalise
-uu* realize realise
-uu* recognize recognise
-uu* reorganize reorganise
-uu* revolutionize revolutionise
-uu* satirize satirise
-uu* scrutinize scrutinise
-uu* sensitize sensitise
-uu* socialize socialise
-uu* specialize specialise
-uu* stabilize stabilise
-uu* standardize standardise
-uu* sterilize sterilise
-uu* stigmatize stigmatise
-uu* subsidize subsidise
-uu* summarize summarise
-uu* symbolize symbolise
-uu* sympathize sympathise
-uu* synchronize synchronise
-uu* synthesize synthesise
-uu* systematize systematise
-uu* tantalize tantalise
-uu* terrorize terrorise
-uu* theorize theorise
-uu* trivialize trivialise
-uu* urbanize urbanise
-uu* utilize utilise
-uu* vandalize vandalise
-uu* verbalize verbalise
-uu* victimize victimise
-uu* visualize visualise
-uu* vocalize vocalise
-uu* westernize westernise
-
-# -yze / -yse
-u*b analyze analyse
-u*b catalyze catalyse
-u*b dialyze dialyse
-u*b electrolyze electrolyse
-u*b hydrolyze hydrolyse
-u*b paralyze paralyse
-u*b psychoanalyze psychoanalyse
-
-# -ense / -ence
-ubb defense defence
-ubb offense offence
-ubb pretense pretence
-ubb defenseless defenceless
-# In British English "license" and "practise" are the verbs, so only American English flags.
-u** license licence
-u** practice practise
-
-# Doubled l (American English accepts both)
-*bb canceled cancelled
-*bb canceling cancelling
-*bb channeled channelled
-*bb channeling channelling
-*bb counseled counselled
-*bb counseling counselling
-*bb counselor counsellor
-*bb counselors counsellors
-*bb dueling duelling
-*bb equaled equalled
-*bb equaling equalling
-*bb fueled fuelled
-*bb fueling fuelling
-*bb jeweler jeweller
-*bb jewelers jewellers
-*bb jewelry jewellery
-*bb labeled labelled
-*bb labeling labelling
-*bb leveled levelled
-*bb leveling levelling
-*bb marveled marvelled
-*bb marveling marvelling
-*bb marvelous marvellous
-*bb modeled modelled
-*bb modeling modelling
-*bb quarreled quarrelled
-*bb quarreling quarrelling
-*bb signaled signalled
-*bb signaling signalling
-*bb totaled totalled
-*bb totaling totalling
-*bb traveled travelled
-*bb traveler traveller
-*bb travelers travellers
-*bb traveling travelling
-*bb tunneled tunnelled
-*bb tunneling tunnelling
-*bb woolen woollen
-
-# -og / -ogue
-*bb catalog catalogue
-*bb cataloged catalogued
-*bb cataloging cataloguing
-
-# ae / oe
-u*b anemia anaemia
-u*b anemic anaemic
-u*b anesthesia anaesthesia
-u*b anesthetic anaesthetic
-**b archeology archaeology
-**b archeologist archaeologist
-u*b diarrhea diarrhoea
-u*b edema oedema
-u*b esophagus oesophagus
-u*b estrogen oestrogen
-u*b gynecologist gynaecologist
-u*b gynecology gynaecology
-u*b hematology haematology
-u*b hemoglobin haemoglobin
-u*b hemorrhage haemorrhage
-u*b leukemia leukaemia
-u*b orthopedic orthopaedic
-u*b pediatric paediatric
-u*b pediatrician paediatrician
-u*b pediatrics paediatrics
-
-# Other words
-u*b aluminum aluminium
-u*b cozy cosy
-ubb mold mould
-ubb molded moulded
-ubb molding moulding
-ubb moldy mouldy
-ubb molt moult
-ubb pajamas pyjamas
-ubb plow plough
-ubb plowed ploughed
-ubb plowing ploughing
-u** program programme
-u*b skeptic sceptic
-u*b skeptical sceptical
-u*b skepticism scepticism
-ubb smolder smoulder
-ubb smoldering smouldering
-# "Tire" and "curb" are also verbs everywhere, so only American English flags.
-uu* tire tyre
-uu* curb kerb
@@ -1,214 +0,0 @@
-# Russian spelling supplement: common words older Russian Hunspell dictionaries lack.
-# Written by WordCraft contributors (MIT OR Apache-2.0); see ATTRIBUTION.md.
-# One word per line: the base form, a TAB, then its declension, from which
-# crates/proof/src/ru/extra.rs makes every form:
-#   m     masculine noun ending in a consonant (маркетплейс, -а, -у, -ом, -е, -ы/-и, -ов/-ей…)
-#   f     feminine noun in -а (инфографика, -и, -е, -у, -ой…; genitive plural without ending)
-#   fia   feminine noun in -ия (интеграция, -ии, -ию, -ией, -ий…)
-#   fs    feminine noun in -ь (соцсеть, -и, -ью, -ей, -ям…)
-#   nie   neuter noun in -ие (согласие, -ия, -ию, -ием, -ий…)
-#   no    neuter noun in -о (юрлицо, -а, -у, -ом, -е; genitive plural without ending)
-#   adj   adjective or participle in -ый, -ий, -ой (бонусный, газоиспользующий)
-#   ovat  verb in -овать with its conjugation and participles (колеровать, администрировать)
-#   =     the forms themselves, space-separated (verbs, nouns with a fleeting vowel)
-# A word alone on its line doesn't decline (онлайн, медиа). Capitalized words are names and match
-# only when written with a capital.
-
-# Commerce, marketing and the web
-маркетплейс	m
-селлер	m
-фулфилмент	m
-ритейл	m
-ритейлер	m
-дистрибьютор	m
-дропшиппинг	m
-мерчандайзер	m
-эквайринг	m
-промокод	m
-кешбэк	m
-кэшбэк	m
-лендинг	m
-лид	m
-тренд	m
-бренд	m
-брендбук	m
-ребрендинг	m
-таргетинг	m
-геотаргетинг	m
-ретаргетинг	m
-таргетолог	m
-маркетолог	m
-копирайтер	m
-рерайтер	m
-контент	m
-блогер	m
-инфлюенсер	m
-хештег	m
-хэштег	m
-репост	m
-лайк	m
-подкаст	m
-стриминг	m
-вебинар	m
-чат	m
-бот	m
-чат-бот	m
-стартап	m
-франчайзинг	m
-аутсорсинг	m
-аутстаффинг	m
-коворкинг	m
-дедлайн	m
-фидбэк	m
-фидбек	m
-кейс	m
-скриншот	m
-скринкаст	m
-баннер	m
-виджет	m
-дашборд	m
-сервис	m
-сайт	m
-скан	m
-токен	m
-аккаунт	m
-логин	m
-пароль	m
-смартфон	m
-гаджет	m
-планшет	m
-ноутбук	m
-апгрейд	m
-апдейт	m
-бэкап	m
-бэкенд	m
-фронтенд	m
-датасет	m
-промпт	m
-интерфейс	m
-хостинг	m
-домен	m
-сниппет	m
-трафик	m
-юзер	m
-стикер	m
-чекбокс	m
-инпут	m
-плейсхолдер	m
-тултип	m
-гипсокартон	m
-пеноплекс	m
-праймер	m
-колер	m
-колорант	m
-Яндекс	m
-Битрикс	m
-Вебвизор	m
-Дзен	m
-инфографика	f
-аналитика	f
-оплата	f
-предоплата	f
-постоплата	f
-доплата	f
-переплата	f
-недоплата	f
-маркировка	= маркировка маркировки маркировке маркировку маркировкой маркировкою маркировок маркировкам маркировками маркировках
-отгрузка	= отгрузка отгрузки отгрузке отгрузку отгрузкой отгрузкою отгрузок отгрузкам отгрузками отгрузках
-колеровка	= колеровка колеровки колеровке колеровку колеровкой колеровкою колеровок колеровкам колеровками колеровках
-рассылка	= рассылка рассылки рассылке рассылку рассылкой рассылкою рассылок рассылкам рассылками рассылках
-выгрузка	= выгрузка выгрузки выгрузке выгрузку выгрузкой выгрузкою выгрузок выгрузкам выгрузками выгрузках
-интеграция	fia
-пагинация	fia
-госрегистрация	fia
-геолокация	fia
-монетизация	fia
-кастомизация	fia
-персонализация	fia
-сегментация	fia
-верификация	fia
-валидация	fia
-авторизация	fia
-аутентификация	fia
-оптимизация	fia
-визуализация	fia
-соцсеть	fs
-согласие	nie
-домовладение	nie
-рождение	nie
-паллетирование	nie
-неисполнение	nie
-непредоставление	nie
-бонусный	adj
-маркетинговый	adj
-мотивационный	adj
-гипоаллергенный	adj
-тиксотропный	adj
-сополимерный	adj
-покупочный	adj
-таргетированный	adj
-брендированный	adj
-кастомный	adj
-онлайновый	adj
-офлайновый	adj
-колеруемый	adj
-колерованный	adj
-заколерованный	adj
-газоиспользующий	adj
-колеровать	ovat
-заколеровать	ovat
-администрировать	ovat
-модерировать	ovat
-отмодерировать	ovat
-таргетировать	ovat
-монетизировать	ovat
-брендировать	ovat
-кастомизировать	ovat
-детектировать	ovat
-верифицировать	ovat
-валидировать	ovat
-генерировать	ovat
-сгенерировать	ovat
-мониторить	= мониторить мониторю мониторишь мониторит мониторим мониторите мониторят мониторил мониторила мониторили мониторя мониторинг
-гуглить	= гуглить гуглю гуглишь гуглит гуглим гуглите гуглят гуглил гуглила гуглили
-скачивать	= скачивать скачиваю скачиваешь скачивает скачиваем скачиваете скачивают скачивал скачивала скачивали скачивается скачиваются
-оценивание	nie
-медучреждение	nie
-юрлицо	no
-учётка	= учётка учётки учётке учётку учёткой учёток учёткам учётками учётках учетка учетки учетке учетку учеткой учеток учеткам учетками учетках
-перелинковка	= перелинковка перелинковки перелинковке перелинковку перелинковкой перелинковок
-поклейка	= поклейка поклейки поклейке поклейку поклейкой поклеек
-детекция	fia
-досудебный	adj
-категорийный	adj
-оркестратор	m
-оффер	m
-трекер	m
-бриф	m
-гайд	m
-монтажёр	m
-тенант	m
-мерчант	m
-амбассадор	m
-Роскомнадзор	m
-Директ	m
-Маркет	m
-вовлечённость	fs
-вовлеченность	fs
-укрывистость	fs
-# Words that don't decline
-онлайн
-офлайн
-оффлайн
-медиа
-демо
-франчайзи
-киноа
-смс
-пн
-вт
-ср
-чт
-пт
-сб
-вс
@@ -1 +1 @@
-9205108cfdced5177706264038848b68b2422d2b
+059cbf95c2a3eb95d637ca8584faca1129d17f48
@@ -9,7 +9,8 @@ description = "WordCraft document model: stories, paragraphs, runs, tables, sect
 workspace = true

 [features]
-# Test inputs other crates' tests share (`cfb_guard::repeating_difat_sample`); turned on only as a
+# Test inputs other crates' tests share (`cfb_guard::repeating_difat_sample`,
+# `zip_guard::zip64_entry_count_sample`); turned on only as a
 # dev-dependency, so never built into the app.
 test-support = []

@@ -48,6 +48,131 @@ impl LineCap {
    }
 }

+/// An arrowhead's width across its line or length along it (ST_LineEndWidth and
+/// ST_LineEndLength, ECMA-376 §20.1.10.39–40): 2, 3 or 5 times the line's width, as Word draws it.
+#[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub enum ArrowSize {
+    Small,
+    #[default]
+    Medium,
+    Large,
+}
+
+impl ArrowSize {
+    /// Read a DrawingML size (`sm`, `med`, `lg`); absent or unknown values are medium.
+    pub fn from_ooxml(value: &str) -> Self {
+        match value {
+            "sm" => Self::Small,
+            "lg" => Self::Large,
+            _ => Self::Medium,
+        }
+    }
+
+    /// The DrawingML spelling.
+    pub fn ooxml(self) -> &'static str {
+        match self {
+            Self::Small => "sm",
+            Self::Medium => "med",
+            Self::Large => "lg",
+        }
+    }
+
+    /// Read a VML width (`narrow`, `medium`, `wide`) or length (`short`, `medium`, `long`).
+    pub fn from_vml(value: &str) -> Self {
+        match value {
+            "narrow" | "short" => Self::Small,
+            "wide" | "long" => Self::Large,
+            _ => Self::Medium,
+        }
+    }
+
+    /// How many [`arrow_unit`]s the arrowhead measures.
+    pub fn factor(self) -> f32 {
+        match self {
+            Self::Small => 2.0,
+            Self::Medium => 3.0,
+            Self::Large => 5.0,
+        }
+    }
+}
+
+/// The kind of an arrowhead (ST_LineEndType, ECMA-376 §20.1.10.33), kept for saving; every kind
+/// is drawn as a triangle.
+#[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(rename_all = "camelCase")]
+pub enum ArrowKind {
+    #[default]
+    Triangle,
+    Stealth,
+    Diamond,
+    Oval,
+    Arrow,
+}
+
+impl ArrowKind {
+    /// Read a DrawingML end type other than `none`; unknown values are triangles.
+    pub fn from_ooxml(value: &str) -> Self {
+        match value {
+            "stealth" => Self::Stealth,
+            "diamond" => Self::Diamond,
+            "oval" => Self::Oval,
+            "arrow" => Self::Arrow,
+            _ => Self::Triangle,
+        }
+    }
+
+    /// The DrawingML spelling.
+    pub fn ooxml(self) -> &'static str {
+        match self {
+            Self::Triangle => "triangle",
+            Self::Stealth => "stealth",
+            Self::Diamond => "diamond",
+            Self::Oval => "oval",
+            Self::Arrow => "arrow",
+        }
+    }
+
+    /// Read a VML arrow (`block`, `classic`, `diamond`, `oval`, `open`) other than `none`.
+    pub fn from_vml(value: &str) -> Self {
+        match value {
+            "classic" => Self::Stealth,
+            "diamond" => Self::Diamond,
+            "oval" => Self::Oval,
+            "open" => Self::Arrow,
+            _ => Self::Triangle,
+        }
+    }
+}
+
+/// An arrowhead's kind and size (`a:headEnd` / `a:tailEnd`). The default is a medium triangle.
+#[derive(Clone, Copy, Debug, Default, PartialEq, Eq, Serialize, Deserialize)]
+#[serde(default, rename_all = "camelCase")]
+pub struct ArrowStyle {
+    pub kind: ArrowKind,
+    pub width: ArrowSize,
+    pub length: ArrowSize,
+}
+
+impl ArrowStyle {
+    pub fn is_default(&self) -> bool {
+        *self == Self::default()
+    }
+
+    /// The arrowhead's length along its line and width across it, for a line `line_width` wide.
+    pub fn dims(self, line_width: f32) -> (f32, f32) {
+        let unit = arrow_unit(line_width);
+        (self.length.factor() * unit, self.width.factor() * unit)
+    }
+}
+
+/// What an arrowhead's size is a multiple of: the line's width, but at least 2 points (Word draws
+/// a thinner line's arrowheads as for a 2-point line).
+pub fn arrow_unit(line_width: f32) -> f32 {
+    let w = if line_width.is_finite() { line_width.clamp(0.25, 200.0) } else { 0.75 };
+    w.max(2.0)
+}
+
 /// The rest of a shape's properties. Empty (the default) for most shapes.
 #[derive(Clone, Debug, PartialEq, Default, Serialize, Deserialize)]
 #[serde(default, rename_all = "camelCase")]
@@ -69,6 +194,11 @@ pub struct ShapeExtra {
    pub arrow_start: bool,
    #[serde(skip_serializing_if = "is_false")]
    pub arrow_end: bool,
+    /// The kind and size of the start and end arrowheads, when there are any.
+    #[serde(skip_serializing_if = "ArrowStyle::is_default")]
+    pub start_style: ArrowStyle,
+    #[serde(skip_serializing_if = "ArrowStyle::is_default")]
+    pub end_style: ArrowStyle,
    /// A rounded rectangle's corner radius (`roundRect`'s `adj`, ECMA-376 §20.1.10.56): in
    /// 100000ths of the shorter side, at most [`MAX_CORNER`]; `None` draws [`DEFAULT_CORNER`].
    #[serde(skip_serializing_if = "Option::is_none")]
@@ -83,6 +213,39 @@ pub struct ShapeExtra {
    /// main colour (the shape's `stroke`).
    #[serde(skip_serializing_if = "Option::is_none")]
    pub outline_fill: Option<crate::fill::Fill>,
+    /// A horizontal rule (VML `o:hr`: a web page's `<hr>` pasted into Word): it takes its whole
+    /// line, and is drawn across [`HorizontalRule::pct`] of it or at the shape's own width.
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub hr: Option<HorizontalRule>,
+}
+
+/// How a horizontal rule sits on its line (VML `o:hrpct`, `o:hralign`), as Word draws one:
+/// across `pct` percent of the line's width (the paragraph's, inside its indents), or at the
+/// shape's own width when `pct` is 0; `align` places it when it is narrower than the line.
+#[derive(Clone, Copy, Debug, PartialEq, Default, Serialize, Deserialize)]
+#[serde(default, rename_all = "camelCase")]
+pub struct HorizontalRule {
+    pub pct: f32,
+    pub align: crate::props::Align,
+    /// Word's standard look (`o:hrstd`), kept for saving.
+    pub standard: bool,
+}
+
+impl HorizontalRule {
+    /// The rule's width on a line `line` wide, for a shape `own` wide.
+    pub fn width(&self, line: f32, own: f32) -> f32 {
+        let line = line.max(0.0);
+        if self.pct.is_finite() && self.pct > 0.0 { line * (self.pct.min(100.0) / 100.0) } else { own.clamp(0.0, line) }
+    }
+
+    /// Where a rule `w` wide starts on a line from `x`, `line` wide.
+    pub fn left(&self, x: f32, line: f32, w: f32) -> f32 {
+        match self.align {
+            crate::props::Align::Center => x + (line - w) / 2.0,
+            crate::props::Align::Right => x + line - w,
+            _ => x,
+        }
+    }
 }

 fn is_zero(v: &u32) -> bool {
@@ -205,11 +205,15 @@ struct BoxBudget {
    costs: BTreeMap<u32, usize>,
    /// Boxes left without a story because the paste's budget couldn't hold theirs.
    skipped: usize,
+    /// A box the fragment has no story for may show this document's story of that id (a
+    /// fragment copied from this document without its parts). Off for another document's
+    /// fragment, whose ids mean nothing here.
+    local: bool,
 }

 impl BoxBudget {
    fn new() -> BoxBudget {
-        BoxBudget { stories: MAX_PASTED_BOXES, bytes: MAX_PASTED_BOX_BYTES, active: Vec::new(), costs: BTreeMap::new(), skipped: 0 }
+        BoxBudget { stories: MAX_PASTED_BOXES, bytes: MAX_PASTED_BOX_BYTES, active: Vec::new(), costs: BTreeMap::new(), skipped: 0, local: true }
    }
 }

@@ -385,6 +389,167 @@ fn restyle_block(b: &mut Block, from: &str, to: &Option<String>, depth: usize) -
    }
 }

+/// What [`Document::adopt_with`] renumbers in another document's content: media keys, lists,
+/// note and comment stories, comment ids, text boxes (one it doesn't bring is dropped), revisions
+/// (after ours, at `rev`; an id past the `revs` it brings is dropped) and styles (those kept under new ids; unstyled paragraphs take
+/// `default_style` when it's given).
+struct Relink<'a> {
+    media: &'a BTreeMap<String, String>,
+    nums: &'a BTreeMap<u32, u32>,
+    notes: &'a BTreeMap<u32, u32>,
+    comments: &'a BTreeMap<u32, u32>,
+    boxes: &'a BTreeMap<u32, Part>,
+    /// Added to permission range ids: ours come first.
+    perms: u32,
+    rev: u32,
+    revs: u32,
+    styles: &'a BTreeMap<String, String>,
+    default_style: Option<&'a str>,
+}
+
+impl Relink<'_> {
+    fn block(&self, b: &mut Block) {
+        // Its own stack, like `each_para_mut`: tables however deeply nested are relinked (#1339).
+        let mut todo = vec![b];
+        while let Some(b) = todo.pop() {
+            match b {
+                Block::Para(p) => self.para(p),
+                Block::Table(t) => {
+                    self.style(&mut t.props.style);
+                    self.shift_change(&mut t.props.fmt_change);
+                    if let Some(c) = t.props.fmt_change.as_mut() {
+                        self.style(&mut c.old.style);
+                    }
+                    for row in t.rows.iter_mut() {
+                        self.shift_change(&mut row.props.fmt_change);
+                        for cell in row.cells.iter_mut() {
+                            self.shift_change(&mut cell.props.fmt_change);
+                            todo.extend(cell.blocks.iter_mut().map(Arc::make_mut));
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    fn style(&self, id: &mut Option<String>) {
+        if let Some(new) = id.as_ref().and_then(|i| self.styles.get(i)) {
+            *id = Some(new.clone());
+        }
+    }
+
+    /// A paragraph's style: unstyled ones take `default_style` when it's given.
+    fn para_style(&self, id: &mut Option<String>) {
+        match (&*id, self.default_style) {
+            (None, Some(d)) => *id = Some(d.to_string()),
+            _ => self.style(id),
+        }
+    }
+
+    fn num(&self, n: &mut Option<crate::props::NumRef>) {
+        if let Some(n) = n.as_mut()
+            && n.num != 0
+        {
+            n.num = self.nums.get(&n.num).copied().unwrap_or(0);
+        }
+    }
+
+    /// Our id for the other document's revision `r`, if it has one.
+    fn rev_id(&self, r: u32) -> Option<u32> {
+        if r < self.revs { self.rev.checked_add(r) } else { None }
+    }
+
+    fn shift_change<T>(&self, c: &mut Option<Box<crate::props::PropChange<T>>>) {
+        match (c.as_ref().and_then(|c| self.rev_id(c.rev)), c.as_mut()) {
+            (Some(v), Some(x)) => x.rev = v,
+            _ => *c = None,
+        }
+    }
+
+    fn para(&self, p: &mut Paragraph) {
+        self.num(&mut p.props.numbering);
+        self.para_style(&mut p.props.style);
+        for r in &mut p.runs {
+            self.style(&mut r.props.style);
+        }
+        self.style(&mut p.mark.style);
+        for c in p.runs.iter_mut().map(|r| &mut r.props).chain([&mut p.mark]) {
+            c.ins = c.ins.and_then(|r| self.rev_id(r));
+            c.del = c.del.and_then(|r| self.rev_id(r));
+            self.shift_change(&mut c.fmt_change);
+            // A tracked change's old formatting names styles and lists as the live one does.
+            if let Some(f) = c.fmt_change.as_mut() {
+                self.style(&mut f.old.style);
+            }
+        }
+        self.shift_change(&mut p.props.fmt_change);
+        if let Some(f) = p.props.fmt_change.as_mut() {
+            self.para_style(&mut f.old.style);
+            self.num(&mut f.old.numbering);
+        }
+        if let Some(s) = p.section.as_deref_mut() {
+            self.shift_change(&mut s.fmt_change);
+        }
+        match (p.props.num_change.as_ref().and_then(|c| self.rev_id(c.rev)), p.props.num_change.as_mut()) {
+            (Some(v), Some(c)) => c.rev = v,
+            _ => p.props.num_change = None,
+        }
+        for o in &mut p.objects {
+            self.object(o, 0);
+        }
+        p.touch();
+    }
+
+    fn object(&self, o: &mut InlineObject, depth: usize) {
+        // An id that names nothing here becomes one that names nothing at all.
+        let map = |m: &BTreeMap<u32, u32>, id: &mut u32| *id = m.get(id).copied().unwrap_or(u32::MAX);
+        match o {
+            InlineObject::Image { media, .. } => {
+                if let Some(k) = self.media.get(media.as_str()) {
+                    *media = k.clone();
+                }
+            }
+            InlineObject::Graphic { graphic, .. } => {
+                let mapped = |m: &str| self.media.get(m).filter(|k| k.as_str() != m);
+                if graphic.items.iter().any(|i| matches!(i, crate::graphic::GraphicItem::Image { media, .. } if mapped(media).is_some())) {
+                    for i in &mut Arc::make_mut(graphic).items {
+                        if let crate::graphic::GraphicItem::Image { media, .. } = i
+                            && let Some(k) = self.media.get(media.as_str())
+                        {
+                            *media = k.clone();
+                        }
+                    }
+                }
+            }
+            InlineObject::Group { children, .. } if depth < MAX_BOX_DEPTH => children.iter_mut().for_each(|c| self.object(&mut c.obj, depth + 1)),
+            InlineObject::NoteRef { id, .. } => map(self.notes, id),
+            InlineObject::CommentStart { id } | InlineObject::CommentEnd { id } => map(self.comments, id),
+            InlineObject::PermStart { id, .. } | InlineObject::PermEnd { id } => *id = id.saturating_add(self.perms),
+            _ => {}
+        }
+        for story in o.text_box_slots() {
+            if story.is_some_and(|id| !self.boxes.contains_key(&id)) {
+                *story = None;
+            }
+        }
+    }
+}
+
+/// One past the highest permission range id in `blocks`.
+fn perm_end<'a>(blocks: impl Iterator<Item = &'a Arc<Block>>) -> u32 {
+    let mut end = 0u32;
+    for b in blocks {
+        each_para(b, 0, &mut |p| {
+            for o in &p.objects {
+                if let InlineObject::PermStart { id, .. } | InlineObject::PermEnd { id } = o {
+                    end = end.max(id.saturating_add(1));
+                }
+            }
+        });
+    }
+    end
+}
+
 /// Text box story ids of a paragraph's shapes.
 fn box_ids(p: &Paragraph) -> impl Iterator<Item = u32> + '_ {
    p.objects.iter().flat_map(InlineObject::text_boxes)
@@ -862,6 +1027,186 @@ impl Document {
        Fragment { blocks, parts }
    }

+    /// Another document's body as a fragment of this one, to insert: the media, lists, styles,
+    /// notes, comments and tracked changes it refers to come along (lists, notes, comments and
+    /// revisions renumbered so they can't collide with ours; a style we already have by id stays
+    /// ours), and its text boxes ride in [`Fragment::parts`].
+    pub fn adopt(&mut self, src: Document) -> Fragment {
+        self.adopt_with(src, false)
+    }
+
+    /// [`Self::adopt`]; with `keep_styles` its content keeps its own formatting (Word's
+    /// `w:matchSrc`): a style whose id we already have comes along under a new id, and its
+    /// document defaults go into its base styles, which its unstyled paragraphs then use.
+    pub fn adopt_with(&mut self, src: Document, keep_styles: bool) -> Fragment {
+        self.adopt_in(src, keep_styles, &mut 0)
+    }
+
+    /// [`Self::adopt_with`], its permission range ids from at least `perms` on (and past ours);
+    /// `perms` moves past them.
+    fn adopt_in(&mut self, mut src: Document, keep_styles: bool, perms: &mut u32) -> Fragment {
+        let mut renamed = BTreeMap::new();
+        let mut default_style = None;
+        if keep_styles {
+            if src.styles.get("Normal").is_none() {
+                src.styles.upsert(crate::styles::Style { id: "Normal".into(), name: "Normal".into(), ..Default::default() });
+            }
+            // Ids given out here are taken too: two styles never get the same new one.
+            let mut given = std::collections::HashSet::new();
+            let room = src.styles.styles.len().saturating_mul(2).saturating_add(self.styles.styles.len()).saturating_add(1);
+            for st in &src.styles.styles {
+                if self.styles.get(&st.id).is_some() {
+                    let fresh = (1..=room)
+                        .map(|n| format!("{}{n}", st.id))
+                        .find(|id| self.styles.get(id).is_none() && src.styles.get(id).is_none() && !given.contains(id));
+                    if let Some(fresh) = fresh {
+                        given.insert(fresh.clone());
+                        renamed.insert(st.id.clone(), fresh);
+                    }
+                }
+            }
+            default_style = Some(renamed.get("Normal").cloned().unwrap_or_else(|| "Normal".into()));
+        }
+        let restyle = |id: &mut Option<String>| {
+            if let Some(new) = id.as_ref().and_then(|i| renamed.get(i)) {
+                *id = Some(new.clone());
+            }
+        };
+        let mut media = BTreeMap::new();
+        for (key, data) in &src.media {
+            let ext = key.rsplit_once('.').map(|(_, e)| e).unwrap_or("png");
+            media.insert(key.clone(), self.add_media(data.to_vec(), ext));
+        }
+        let next_abs = self.numbering.abstracts.iter().map(|a| a.id.saturating_add(1)).max().unwrap_or(0);
+        let mut abs_map = BTreeMap::new();
+        for (i, a) in src.numbering.abstracts.iter().enumerate() {
+            let id = next_abs.saturating_add(i as u32);
+            abs_map.insert(a.id, id);
+            let mut a = crate::numbering::AbstractNum { id, ..a.clone() };
+            a.levels.iter_mut().for_each(|l| restyle(&mut l.style));
+            self.numbering.abstracts.push(a);
+        }
+        let next_num = self.numbering.nums.iter().map(|n| n.id.saturating_add(1)).max().unwrap_or(1).max(1);
+        let mut nums = BTreeMap::new();
+        for (i, n) in src.numbering.nums.iter().enumerate() {
+            let id = next_num.saturating_add(i as u32);
+            nums.insert(n.id, id);
+            let abstract_id = abs_map.get(&n.abstract_id).copied().unwrap_or(n.abstract_id);
+            let mut n = crate::numbering::Num { id, abstract_id, ..n.clone() };
+            n.level_overrides.iter_mut().for_each(|(_, l)| restyle(&mut l.style));
+            self.numbering.nums.push(n);
+        }
+        let renumber = |n: &mut Option<crate::props::NumRef>| {
+            if let Some(n) = n.as_mut()
+                && n.num != 0
+            {
+                n.num = nums.get(&n.num).copied().unwrap_or(0);
+            }
+        };
+        // What a style in the other document doesn't set comes from its defaults there, and
+        // Word's own (no space, single lines) where those don't say.
+        let mut para_defaults = src.styles.default_para.clone();
+        para_defaults.space_before.get_or_insert(0.0);
+        para_defaults.space_after.get_or_insert(0.0);
+        para_defaults.line_spacing.get_or_insert(crate::props::LineSpacing::Multiple(1.0));
+        for st in &src.styles.styles {
+            let mut st = st.clone();
+            // Our defaults stay ours.
+            st.default = false;
+            if keep_styles {
+                if let Some(new) = renamed.get(&st.id) {
+                    st.name = format!("{} ({new})", st.name);
+                    st.id = new.clone();
+                }
+                restyle(&mut st.based_on);
+                restyle(&mut st.next);
+                restyle(&mut st.linked);
+                // A character style's runs get the defaults through their paragraph's style.
+                if st.based_on.is_none() && st.kind != crate::styles::StyleKind::Character {
+                    let mut chr = src.styles.default_chr.clone();
+                    chr.overlay(&st.chr);
+                    st.chr = chr;
+                    let mut para = para_defaults.clone();
+                    para.overlay(&st.para);
+                    st.para = para;
+                }
+            }
+            if self.styles.get(&st.id).is_none() {
+                renumber(&mut st.para.numbering);
+                self.styles.upsert(st);
+            }
+        }
+        let rev = u32::try_from(self.revisions.len()).unwrap_or(u32::MAX);
+        self.revisions.extend(src.revisions.iter().cloned());
+        // Notes and comments get stories (and comments ids) of their own here; their text is
+        // relinked below, once every new id is known.
+        let mut notes = BTreeMap::new();
+        for (id, part) in src.parts.iter().filter(|(_, p)| matches!(p.kind, PartKind::Footnote | PartKind::Endnote)) {
+            notes.insert(*id, self.add_part(part.kind, Vec::new()));
+        }
+        let mut comments = BTreeMap::new();
+        let mut next_comment = self.comments.keys().next_back().map_or(0, |k| k.saturating_add(1));
+        for (id, c) in &src.comments {
+            if src.parts.get(&c.part).is_some_and(|p| p.kind == PartKind::Comment) {
+                notes.insert(c.part, self.add_part(PartKind::Comment, Vec::new()));
+                comments.insert(*id, next_comment);
+                next_comment = next_comment.saturating_add(1);
+            }
+        }
+        let boxes: BTreeMap<u32, Part> = src.parts.iter().filter(|(_, p)| p.kind == PartKind::TextBox).map(|(k, p)| (*k, p.clone())).collect();
+        // Its permission ranges are numbered after ours, so no end of theirs closes one of ours.
+        let first_perm = perm_end(self.body.iter().chain(self.parts.values().flat_map(|p| p.blocks.iter()))).max(*perms);
+        *perms = first_perm.saturating_add(perm_end(src.body.iter().chain(src.parts.values().flat_map(|p| p.blocks.iter()))));
+        let link = Relink {
+            media: &media,
+            nums: &nums,
+            notes: &notes,
+            comments: &comments,
+            boxes: &boxes,
+            perms: first_perm,
+            rev,
+            revs: u32::try_from(src.revisions.len()).unwrap_or(u32::MAX),
+            styles: &renamed,
+            default_style: default_style.as_deref(),
+        };
+        let mut parts = boxes.clone();
+        for part in parts.values_mut() {
+            part.blocks.iter_mut().for_each(|b| link.block(Arc::make_mut(b)));
+        }
+        // A note's or comment's text boxes get stories here now; the body's go with the fragment.
+        let mut budget = BoxBudget::new();
+        for (old, new) in &notes {
+            let Some(part) = src.parts.get(old) else { continue };
+            let mut blocks: Vec<Block> = part.blocks.iter().map(|b| (**b).clone()).collect();
+            blocks.iter_mut().for_each(|b| link.block(b));
+            self.adopt_text_boxes(&mut blocks, &parts, 0, &mut budget);
+            if let Some(p) = self.parts.get_mut(new) {
+                p.blocks = blocks.into_iter().map(Arc::new).collect();
+            }
+        }
+        for (id, c) in &src.comments {
+            if let (Some(new), Some(part)) = (comments.get(id), notes.get(&c.part)) {
+                let parent = c.parent.and_then(|p| comments.get(&p).copied());
+                // Its durable id is unique in its own document only: the writer gives the copy one.
+                self.comments.insert(*new, crate::Comment { parent, part: *part, durable_id: None, ..c.clone() });
+            }
+        }
+        let mut blocks: Vec<Block> = src.body.iter().map(|b| (**b).clone()).collect();
+        blocks.iter_mut().for_each(|b| link.block(b));
+        Fragment { blocks, parts }
+    }
+
+    /// [`Self::adopt_with`] as blocks to place in a story directly (a reader building one): its
+    /// text boxes already have stories of their own here. `perms` is the first permission range
+    /// id the reader hasn't given out (ranges it holds aren't in this document yet); it moves past
+    /// the adopted ones.
+    pub fn adopt_blocks(&mut self, src: Document, keep_styles: bool, perms: &mut u32) -> Vec<Block> {
+        let Fragment { mut blocks, parts } = self.adopt_in(src, keep_styles, perms);
+        let mut budget = BoxBudget::new();
+        self.adopt_text_boxes(&mut blocks, &parts, 0, &mut budget);
+        blocks
+    }
+
    /// Copies of the text box stories shapes in `blocks` show, nested boxes included.
    fn text_box_parts(&self, blocks: &[Block]) -> BTreeMap<u32, Part> {
        let mut out = BTreeMap::new();
@@ -971,7 +1316,7 @@ impl Document {
            budget.skipped = budget.skipped.saturating_add(1);
            return None;
        }
-        let src = parts.get(&old).or_else(|| self.parts.get(&old).filter(|p| p.kind == PartKind::TextBox))?;
+        let src = parts.get(&old).or_else(|| self.parts.get(&old).filter(|p| budget.local && p.kind == PartKind::TextBox))?;
        // Charge what the copy really allocates before making any of it: the block slots, and the
        // blocks holding text boxes (deep-copied below). The rest are shared `Arc`s, however many
        // boxes show them, so they cost nothing more. Worked out once per paste.
@@ -1012,6 +1357,22 @@ impl Document {
        Some(id)
    }

+    /// The blocks of `frag`, a fragment of another document, with every text box given a new
+    /// story of its own here copied from the fragment's parts (as a paste does), to place them
+    /// in this document; a box the fragment has no story for keeps none.
+    pub fn adopt_fragment(&mut self, frag: &Fragment) -> Vec<Block> {
+        let mut blocks = frag.blocks.clone();
+        let mut budget = BoxBudget { local: false, ..BoxBudget::new() };
+        self.adopt_text_boxes(&mut blocks, &frag.parts, 0, &mut budget);
+        if budget.skipped > 0 {
+            log::warn!(
+                "{} text boxes are left empty; their stories are past what one copy brings ({MAX_PASTED_BOXES} stories, {MAX_PASTED_BOX_BYTES} bytes)",
+                budget.skipped
+            );
+        }
+        blocks
+    }
+
    /// Insert a fragment at `pos`; returns the position after the pasted content.
    pub fn insert_fragment(&mut self, pos: &Pos, frag: &Fragment) -> Result<Pos> {
        // Tables that would nest deeper than the limit here become their paragraphs (#1596).
@@ -664,3 +664,230 @@ fn a_table_nested_deeper_than_the_paragraph_walk_is_never_removed_whole() {
    d.delete_range(&Pos { story: StoryRef::Body, path: first, off: 0 }, &Pos::body(2, 0)).unwrap();
    assert!(d.body.get(1).is_some_and(|b| b.as_table().is_some()), "a table with paragraphs the walk never saw was removed");
 }
+#[test]
+fn adopting_a_document_renumbers_what_it_refers_to() {
+    use crate::para::NoteKind;
+    use crate::{Comment, Revision, RevisionKind};
+    // Ours already has a footnote, a comment, a list and a revision with the ids the other's use.
+    let mut ours = doc(&["ours"]);
+    let our_note = ours.add_part(PartKind::Footnote, vec![para_block(Paragraph::with_text("our note", CharProps::default()))]);
+    ours.comments.insert(0, Comment { part: our_note, ..Comment::default() });
+    let our_list = ours.numbering.add_list(crate::numbering::ListKind::Bullet);
+    ours.revisions.push(Revision { kind: RevisionKind::Insert, ..Revision::default() });
+
+    let mut theirs = doc(&["theirs"]);
+    let note = theirs.add_part(PartKind::Footnote, vec![para_block(Paragraph::with_text("their note", CharProps::default()))]);
+    let said = theirs.add_part(PartKind::Comment, vec![para_block(Paragraph::with_text("their comment", CharProps::default()))]);
+    theirs.comments.insert(0, Comment { author: "Them".into(), part: said, durable_id: Some("0A0B0C0D".into()), ..Comment::default() });
+    let list = theirs.numbering.add_list(crate::numbering::ListKind::Bullet);
+    assert_eq!(list, our_list);
+    theirs.revisions.push(Revision { kind: RevisionKind::Delete, ..Revision::default() });
+    let mut style = crate::styles::Style { id: "TheirList".into(), name: "Their List".into(), ..Default::default() };
+    style.para.numbering = Some(crate::props::NumRef { num: list, level: 0 });
+    theirs.styles.upsert(style);
+    let p = theirs.para_mut(StoryRef::Body, &Path::top(0)).unwrap();
+    p.props.numbering = Some(crate::props::NumRef { num: list, level: 0 });
+    p.runs[0].props.del = Some(0);
+    // An insertion naming a revision it doesn't have.
+    p.runs[0].props.ins = Some(7);
+    let props = p.props_of_char(0).clone();
+    p.insert_object(0, InlineObject::NoteRef { kind: NoteKind::Footnote, id: note, custom: String::new() }, &props).unwrap();
+    p.insert_object(0, InlineObject::CommentStart { id: 0 }, &props).unwrap();
+
+    let frag = ours.adopt(theirs);
+    let Some(Block::Para(p)) = frag.blocks.first() else { panic!() };
+    let new_list = p.props.numbering.unwrap().num;
+    assert_ne!(new_list, our_list);
+    assert!(ours.numbering.num(new_list).is_some());
+    assert_eq!(ours.styles.get("TheirList").and_then(|s| s.para.numbering).map(|n| n.num), Some(new_list));
+    // Their deletion is their revision, after ours.
+    let del = p.runs.iter().find_map(|r| r.props.del).unwrap();
+    assert_eq!(ours.revisions.get(del as usize).map(|r| r.kind), Some(RevisionKind::Delete));
+    assert!(p.runs.iter().all(|r| r.props.ins.is_none()));
+    // Their note and comment come along under ids of their own.
+    let (note_id, comment_id) = p.objects.iter().fold((None, None), |(n, c), o| match o {
+        InlineObject::NoteRef { id, .. } => (Some(*id), c),
+        InlineObject::CommentStart { id } => (n, Some(*id)),
+        _ => (n, c),
+    });
+    let text = |id: u32| {
+        ours.parts
+            .get(&id)
+            .map(|p| p.blocks.iter().map(|b| if let Block::Para(p) = &**b { p.text.clone() } else { String::new() }).collect::<String>())
+    };
+    assert_eq!(text(note_id.unwrap()).as_deref(), Some("their note"));
+    let comment = ours.comments.get(&comment_id.unwrap()).unwrap();
+    assert_eq!((comment.author.as_str(), text(comment.part).as_deref()), ("Them", Some("their comment")));
+    // Its durable id is the other document's: the writer gives the copy one of its own.
+    assert_eq!(comment.durable_id, None);
+    assert_eq!(text(our_note).as_deref(), Some("our note"));
+}
+
+#[test]
+fn a_kept_character_style_leaves_what_it_doesnt_set_to_the_paragraph_style() {
+    use crate::styles::{Style, StyleKind, StyleSheet};
+    let mut ours = doc(&["ours"]);
+    let mut theirs = doc(&["theirs"]);
+    theirs.styles = StyleSheet::empty();
+    theirs.styles.default_chr = CharProps { size: Some(9.0), ..Default::default() };
+    let big = CharProps { size: Some(20.0), ..Default::default() };
+    theirs.styles.upsert(Style { id: "Big".into(), name: "Big".into(), chr: big, ..Default::default() });
+    theirs.styles.upsert(Style { id: "Link".into(), name: "Link".into(), kind: StyleKind::Character, ..Default::default() });
+    let p = theirs.para_mut(StoryRef::Body, &Path::top(0)).unwrap();
+    p.props.style = Some("Big".into());
+    p.runs[0].props.style = Some("Link".into());
+
+    let frag = ours.adopt_with(theirs, true);
+    let Some(Block::Para(p)) = frag.blocks.first() else { panic!() };
+    assert_eq!(ours.styles.resolve_char(p.props.style.as_deref(), &p.runs[0].props).size, 20.0);
+}
+
+#[test]
+fn kept_styles_renamed_on_adoption_are_renamed_in_list_levels_too() {
+    use crate::styles::Style;
+    let mut ours = doc(&["ours"]);
+    ours.styles.upsert(Style { id: "Heading1".into(), name: "heading 1".into(), ..Default::default() });
+    let mut theirs = doc(&["theirs"]);
+    theirs.styles.upsert(Style { id: "Heading1".into(), name: "heading 1".into(), ..Default::default() });
+    let list = theirs.numbering.add_list(crate::numbering::ListKind::Numbered);
+    let abs = theirs.numbering.nums.iter().find(|n| n.id == list).map(|n| n.abstract_id).unwrap();
+    let a = theirs.numbering.abstracts.iter_mut().find(|a| a.id == abs).unwrap();
+    a.levels[0].style = Some("Heading1".into());
+    let level = a.levels[0].clone();
+    theirs.numbering.nums.iter_mut().find(|n| n.id == list).unwrap().level_overrides.push((1, level));
+
+    let before = ours.numbering.abstracts.len();
+    ours.adopt_with(theirs, true);
+    let copied = &ours.numbering.abstracts[before..];
+    let styles: Vec<_> = copied
+        .iter()
+        .flat_map(|a| a.levels.iter())
+        .chain(ours.numbering.nums.iter().flat_map(|n| n.level_overrides.iter().map(|(_, l)| l)))
+        .filter_map(|l| l.style.clone())
+        .collect();
+    assert_eq!(styles.len(), 2, "{styles:?}");
+    for st in &styles {
+        assert_ne!(st, "Heading1");
+        assert!(ours.styles.get(st).is_some(), "{st}");
+    }
+}
+
+#[test]
+fn an_adopted_section_breaks_tracked_change_names_its_revision() {
+    use crate::{Revision, RevisionKind, SectionProps};
+    let mut ours = doc(&["ours"]);
+    ours.revisions.push(Revision { kind: RevisionKind::Insert, ..Revision::default() });
+    let mut theirs = doc(&["theirs"]);
+    theirs.revisions.push(Revision { kind: RevisionKind::Format, ..Revision::default() });
+    let p = theirs.para_mut(StoryRef::Body, &Path::top(0)).unwrap();
+    p.section = Some(Box::new(SectionProps { fmt_change: crate::props::PropChange::boxed(0, SectionProps::default()), ..SectionProps::default() }));
+
+    let frag = ours.adopt(theirs);
+    let Some(Block::Para(p)) = frag.blocks.first() else { panic!() };
+    let rev = p.section.as_ref().and_then(|s| s.fmt_change.as_ref()).map(|c| c.rev);
+    assert_eq!(rev.and_then(|r| ours.revisions.get(r as usize)).map(|r| r.kind), Some(RevisionKind::Format));
+}
+
+#[test]
+fn an_adopted_tracked_format_change_keeps_its_old_styles_and_list() {
+    use crate::props::{NumRef, ParaProps, PropChange};
+    use crate::{Revision, RevisionKind};
+    for keep in [false, true] {
+        let mut ours = doc(&["ours"]);
+        let our_list = ours.numbering.add_list(crate::numbering::ListKind::Bullet);
+        let mut theirs = doc(&["theirs"]);
+        theirs.revisions.push(Revision { kind: RevisionKind::Format, ..Revision::default() });
+        let list = theirs.numbering.add_list(crate::numbering::ListKind::Bullet);
+        assert_eq!(list, our_list);
+        let p = theirs.para_mut(StoryRef::Body, &Path::top(0)).unwrap();
+        // Live and old props name the same styles (ours has them too) and their list.
+        p.props.style = Some("Quote".into());
+        p.props.numbering = Some(NumRef { num: list, level: 0 });
+        let old = ParaProps { style: Some("Quote".into()), numbering: Some(NumRef { num: list, level: 0 }), ..ParaProps::default() };
+        p.props.fmt_change = PropChange::boxed(0, old);
+        p.runs[0].props.style = Some("Strong".into());
+        p.runs[0].props.fmt_change = PropChange::boxed(0, CharProps { style: Some("Strong".into()), ..CharProps::default() });
+
+        let frag = ours.adopt_with(theirs, keep);
+        let Some(Block::Para(p)) = frag.blocks.first() else { panic!() };
+        let old = &p.props.fmt_change.as_ref().unwrap().old;
+        assert_eq!((&old.style, old.numbering), (&p.props.style, p.props.numbering), "keep_styles {keep}");
+        assert_ne!(old.numbering.map(|n| n.num), Some(our_list));
+        let run = &p.runs[0].props;
+        assert_eq!(run.fmt_change.as_ref().unwrap().old.style, run.style, "keep_styles {keep}");
+    }
+}
+
+#[test]
+fn adopted_styles_never_become_our_defaults() {
+    use crate::styles::{Style, StyleKind};
+    for keep in [false, true] {
+        let mut ours = doc(&["ours"]);
+        let mut theirs = doc(&["theirs"]);
+        // A localized default character style, and one whose id we have too.
+        theirs.styles.upsert(Style { id: "Absatz".into(), kind: StyleKind::Character, default: true, ..Default::default() });
+        theirs.styles.upsert(Style { id: "Strong".into(), kind: StyleKind::Character, default: true, ..Default::default() });
+        ours.adopt_with(theirs, keep);
+        assert!(ours.styles.styles.iter().all(|s| !s.default), "keep_styles {keep}");
+    }
+}
+
+#[test]
+fn adopted_permission_ranges_keep_ids_of_their_own() {
+    let perm = |id, editor: &str| InlineObject::PermStart { id, group: String::new(), editor: editor.into() };
+    let mut ours = doc(&["ours"]);
+    ours.insert_object(&Pos::body(0, 4), InlineObject::PermEnd { id: 0 }, &CharProps::default()).unwrap();
+    ours.insert_object(&Pos::body(0, 0), perm(0, "us"), &CharProps::default()).unwrap();
+    // Theirs: a closed range and one left open, numbered from 0 like ours.
+    let mut theirs = doc(&["theirs"]);
+    theirs.insert_object(&Pos::body(0, 6), perm(1, "them"), &CharProps::default()).unwrap();
+    theirs.insert_object(&Pos::body(0, 6), InlineObject::PermEnd { id: 0 }, &CharProps::default()).unwrap();
+    theirs.insert_object(&Pos::body(0, 0), perm(0, "them"), &CharProps::default()).unwrap();
+
+    let frag = ours.adopt(theirs);
+    let Some(Block::Para(p)) = frag.blocks.first() else { panic!() };
+    let ids: Vec<u32> = p
+        .objects
+        .iter()
+        .filter_map(|o| match o {
+            InlineObject::PermStart { id, .. } | InlineObject::PermEnd { id } => Some(*id),
+            _ => None,
+        })
+        .collect();
+    assert_eq!(ids.len(), 3);
+    assert!(!ids.contains(&0), "{ids:?}");
+    assert_eq!(ids[0], ids[1], "a range's start and end still pair: {ids:?}");
+    assert_ne!(ids[1], ids[2]);
+}
+
+#[test]
+fn an_adopted_notes_text_box_gets_a_story_of_its_own() {
+    use crate::para::NoteKind;
+    // Ours has a text box under the id their note's text box has.
+    let mut ours = doc(&["ours"]);
+    ours.add_part(PartKind::Footnote, Vec::new());
+    let our_box = text_box(&mut ours, &Pos::body(0, 0), "our box");
+    let mut theirs = doc(&["theirs"]);
+    let note = theirs.add_part(PartKind::Footnote, vec![para_block(Paragraph::with_text("note", CharProps::default()))]);
+    let their_box = theirs.add_part(PartKind::TextBox, vec![para_block(Paragraph::with_text("their box", CharProps::default()))]);
+    assert_eq!(their_box, our_box);
+    theirs
+        .insert_object(&Pos { story: StoryRef::Part(note), path: Path::top(0), off: 0 }, box_shape(Some(their_box)), &CharProps::default())
+        .unwrap();
+    theirs
+        .insert_object(&Pos::body(0, 0), InlineObject::NoteRef { kind: NoteKind::Footnote, id: note, custom: String::new() }, &CharProps::default())
+        .unwrap();
+
+    let blocks = ours.adopt(theirs).blocks;
+    let Some(Block::Para(p)) = blocks.first() else { panic!() };
+    let new_note = p.objects.iter().find_map(|o| if let InlineObject::NoteRef { id, .. } = o { Some(*id) } else { None }).unwrap();
+    let story = ours.parts[&new_note]
+        .blocks
+        .iter()
+        .find_map(|b| if let Block::Para(p) = &**b { p.objects.iter().find_map(|o| o.text_boxes().first().copied()) } else { None })
+        .unwrap();
+    assert_ne!(story, our_box);
+    let part = &ours.parts[&story];
+    assert_eq!(part.kind, PartKind::TextBox);
+    assert!(matches!(part.blocks.first().map(|b| &**b), Some(Block::Para(p)) if p.text == "their box"));
+}
@@ -48,6 +48,7 @@ pub mod symbol_fonts;
 pub mod table;
 pub mod wordart;
 pub mod wrap;
+pub mod zip_guard;

 use std::collections::BTreeMap;
 use std::sync::Arc;
@@ -338,6 +338,23 @@ impl Counters {
                *f = LevelFlags { started: false, first: Some(*start) };
            }
        }
+        // A level above that hasn't counted yet counts its first number here, as in Word: after
+        // 1.1 and 1.2 with no level-1 item before them, the next level-1 item is 2, whether or not
+        // the deeper level's text shows it.
+        for (k, (v, f)) in st.iter_mut().zip(fl.iter_mut()).enumerate().take(lv) {
+            if f.started {
+                continue;
+            }
+            let start = match self.seen.entry((num, k as u8)) {
+                std::collections::hash_map::Entry::Vacant(e) => {
+                    e.insert(true);
+                    n.start_overrides.iter().find(|(l, _)| *l as usize == k).map(|(_, s)| *s)
+                }
+                std::collections::hash_map::Entry::Occupied(_) => None,
+            };
+            *v = start.or(f.first).unwrap_or_else(|| level_def(k).map_or(1, |l| l.start));
+            *f = LevelFlags { started: true, first: None };
+        }
        if let (Some(v), Some(f)) = (st.get_mut(lv), fl.get_mut(lv)) {
            *v = if f.started { v.saturating_add(1) } else { f.first.unwrap_or(def.start) };
            *f = LevelFlags { started: true, first: None };
@@ -426,8 +443,9 @@ mod tests {
        assert_eq!(c.next_label(&n, id, 2).unwrap().0, "1.02");
        assert_eq!(c.next_label(&n, id, 1).unwrap().0, "");
        assert_eq!(c.next_label(&n, id, 2).unwrap().0, "2.01");
-        // Levels it doesn't define still come from the abstract list.
-        assert_eq!(c.next_label(&n, id, 0).unwrap().0, "1.");
+        // Levels it doesn't define still come from the abstract list; level 0 took its first number
+        // with the first level-1 item, so this is its second.
+        assert_eq!(c.next_label(&n, id, 0).unwrap().0, "2.");
        assert_eq!(n.level(id, 2).map(|l| l.format), Some(NumFormat::DecimalZero));
    }

@@ -476,6 +494,40 @@ mod tests {
        assert_eq!(l(2), "(1)", "level 1 restarts level 3");
    }

+    #[test]
+    fn a_deeper_item_first_counts_the_levels_above_it() {
+        // Measured in Word: a level above that hasn't counted yet takes its first number at the
+        // first deeper item, whether or not that item's text shows it.
+        let labels_with = |starts: [u32; 3], texts: [&str; 3], overrides: &[(u8, u32)], seq: &[u8]| {
+            let mut n = Numbering::default();
+            let id = n.add_list(ListKind::Legal);
+            if let Some(a) = n.abstracts.first_mut() {
+                for (l, (s, t)) in a.levels.iter_mut().zip(starts.into_iter().zip(texts)) {
+                    l.start = s;
+                    l.text = t.into();
+                }
+            }
+            if let Some(num) = n.nums.iter_mut().find(|x| x.id == id) {
+                num.start_overrides = overrides.to_vec();
+            }
+            let mut c = Counters::default();
+            seq.iter().map(|lv| c.next_label(&n, id, *lv).unwrap().0).collect::<Vec<_>>()
+        };
+        let labels = |starts: [u32; 3], texts: [&str; 3], seq: &[u8]| labels_with(starts, texts, &[], seq);
+        let std = ["%1.", "%1.%2.", "%1.%2.%3."];
+        assert_eq!(labels([1, 1, 1], std, &[1, 1, 0, 0]), ["1.1.", "1.2.", "2.", "3."]);
+        assert_eq!(labels([10, 1, 1], std, &[1, 1, 1, 1, 0, 0]), ["10.1.", "10.2.", "10.3.", "10.4.", "11.", "12."]);
+        assert_eq!(labels([1, 1, 1], ["%1.", "(%2)", "%1.%2.%3."], &[1, 1, 0, 0]), ["(1)", "(2)", "2.", "3."]);
+        assert_eq!(labels([1, 1, 1], std, &[2, 1, 0, 1]), ["1.1.1.", "1.2.", "2.", "2.1."]);
+        assert_eq!(labels([1, 1, 1], ["%1.", "%1.%2.", "(%3)"], &[2, 1, 0]), ["(1)", "1.2.", "2."]);
+        assert_eq!(labels([1, 1, 1], std, &[0, 2, 1, 0]), ["1.", "1.1.1.", "1.2.", "2."]);
+        // The num's own start for the level above counts in its place, as the level's start does.
+        assert_eq!(labels_with([1, 1, 1], std, &[(0, 10)], &[1, 1, 0]), ["10.1.", "10.2.", "11."]);
+        assert_eq!(labels_with([1, 1, 1], std, &[(0, 5)], &[1, 0, 1]), ["5.1.", "6.", "6.1."]);
+        // A level-1 item after the first deeper ones restarts level 2 below it, as usual.
+        assert_eq!(labels([1, 1, 1], std, &[1, 1, 0, 1, 1, 0, 2]), ["1.1.", "1.2.", "2.", "2.1.", "2.2.", "3.", "3.1.1."]);
+    }
+
    #[test]
    fn a_level_starting_at_zero_counts_up() {
        let mut n = Numbering::default();
@@ -448,6 +448,25 @@ fn box_para(story: u32, id: u32) -> Blocks {
    vec![para_block(p)]
 }

+#[test]
+fn adopting_relinks_lists_in_deeply_nested_tables() {
+    // As #1339: a list paragraph 20 tables down takes the list it is brought in with.
+    let mut ours = Document::from_text("ours");
+    let our_list = ours.numbering.add_list(numbering::ListKind::Bullet);
+    let mut theirs = Document::from_text("theirs");
+    let list = theirs.numbering.add_list(numbering::ListKind::Bullet);
+    assert_eq!(list, our_list);
+    let mut p = Paragraph::with_text("item", CharProps::default());
+    p.props.numbering = Some(props::NumRef { num: list, level: 0 });
+    theirs.body = nested_tables(20, vec![para_block(p)]);
+    let mut nums = Vec::new();
+    for b in &ours.adopt(theirs).blocks {
+        edit::each_para(b, 0, &mut |p| nums.extend(p.props.numbering.map(|n| n.num)));
+    }
+    assert_eq!(nums.len(), 1);
+    assert_ne!(nums[0], our_list);
+}
+
 #[test]
 fn text_boxes_shown_from_deeply_nested_tables_are_kept() {
    // #1339: the paragraph walk stopped 16 tables down, so a box shown only from deeper was taken
@@ -0,0 +1,191 @@
+//! A bytes-only check of a ZIP package's declared entry count before the `zip` crate opens it
+//! (#3660). Every reader of `.docx`, `.odt` and spreadsheet packages calls [`check`] first.
+//!
+//! `zip::ZipArchive::new` reserves room for, and parses, as many central directory headers as the
+//! end record declares before its caller can look at `len()`. A ZIP64 end record can declare
+//! millions of entries for a file of a few hundred megabytes (only 46 bytes each have to fit), and
+//! each parsed entry costs several times that in memory. Here the end records are found the way
+//! `zip` finds them (APPNOTE.TXT §4.3.14 to §4.3.16), and a package declaring more than the
+//! reader's limit is refused before `zip` parses anything. Anything else that is wrong is left for
+//! `zip` to report.
+
+const EOCD: &[u8; 4] = b"PK\x05\x06";
+const EOCD64_LOCATOR: &[u8; 4] = b"PK\x06\x07";
+const EOCD64: &[u8; 4] = b"PK\x06\x06";
+const CENTRAL_HEADER: &[u8; 4] = b"PK\x01\x02";
+
+/// End records looked at, at most. A package has one; a few more can turn up inside stored data.
+const MAX_END_RECORDS: usize = 16;
+
+/// Refuse `bytes` when the end record `zip` would use declares more than `max_entries` entries.
+///
+/// `zip` searches the file backwards for an end of central directory record. A ZIP64 one points
+/// through its locator at the ZIP64 end record, which `zip` looks for from the locator's offset
+/// on: every such record there is checked, and as `zip` may fall back to an earlier end record
+/// when one doesn't parse, the search goes on. A plain one is used when it declares no entries, or
+/// a central directory header lies between its directory offset and itself: it is the last one
+/// looked at.
+pub fn check(bytes: &[u8], max_entries: u64) -> Result<(), String> {
+    let u16_at = |at: usize| bytes.get(at..at.checked_add(2)?).map(|b| u16::from_le_bytes([b[0], b[1]]));
+    let u32_at = |at: usize| bytes.get(at..at.checked_add(4)?).map(|b| u32::from_le_bytes([b[0], b[1], b[2], b[3]]));
+    let u64_at = |at: usize| bytes.get(at..at.checked_add(8)?).and_then(|b| <[u8; 8]>::try_from(b).ok()).map(u64::from_le_bytes);
+    let refuse = |n: u64| -> Result<(), String> {
+        if n > max_entries { Err(format!("the package declares {n} entries, more than {max_entries}")) } else { Ok(()) }
+    };
+    let mut end = bytes.len();
+    let mut seen = 0;
+    while let Some(at) = bytes.get(..end).and_then(|b| b.windows(4).rposition(|w| w == EOCD)) {
+        end = at;
+        seen += 1;
+        if seen > MAX_END_RECORDS {
+            return Err(format!("more than {MAX_END_RECORDS} end of central directory records"));
+        }
+        // `zip` skips a record whose comment runs past the end of the file.
+        let Some(comment) = u16_at(at + 20) else { continue };
+        if at + 22 + usize::from(comment) > bytes.len() {
+            continue;
+        }
+        let (Some(on_disk), Some(total), Some(cd_offset)) = (u16_at(at + 8), u16_at(at + 10), u32_at(at + 16)) else { continue };
+        let zip64 = total == u16::MAX || cd_offset == u32::MAX;
+        if zip64
+            && let Some(locator) = at.checked_sub(20)
+            && bytes.get(locator..locator + 4) == Some(EOCD64_LOCATOR.as_slice())
+        {
+            let (Some(eocd64), Some(disks)) = (u64_at(locator + 8), u32_at(locator + 16)) else { continue };
+            let Some(from) = usize::try_from(eocd64).ok().filter(|&f| f < locator) else { continue };
+            if disks > 1 {
+                continue;
+            }
+            let records = bytes.get(from..locator).unwrap_or_default().windows(4).enumerate().filter(|(_, w)| *w == EOCD64);
+            for (i, _) in records {
+                let p = from + i;
+                refuse(u64_at(p + 24).unwrap_or(0).max(u64_at(p + 32).unwrap_or(0)))?;
+            }
+            continue;
+        }
+        if total == 0 {
+            return Ok(());
+        }
+        let cd = cd_offset as usize;
+        if cd < at && bytes.get(cd..at).is_some_and(|b| b.windows(4).any(|w| w == CENTRAL_HEADER)) {
+            return refuse(u64::from(on_disk.max(total)));
+        }
+    }
+    Ok(())
+}
+
+/// A package of one empty stored file, `a`, whose central directory lists it `entries` times (at
+/// most as many as fit in 64 MiB) under a ZIP64 end record that declares them all: `zip` 3.0.0
+/// parses and keeps every one of them before its caller can count them. Regression input for the
+/// readers' tests: built only for tests, this crate's and (through the `test-support` feature,
+/// which they turn on as a dev-dependency) the readers', never into the app.
+#[cfg(any(test, feature = "test-support"))]
+pub fn zip64_entry_count_sample(entries: u64) -> Vec<u8> {
+    // Local file header (APPNOTE.TXT §4.3.7): version 10, stored, no data, name "a".
+    let mut f = b"PK\x03\x04".to_vec();
+    f.extend_from_slice(&[10, 0, 0, 0, 0, 0, 0, 0, 0x21, 0]); // version, flags, method, time, date
+    f.extend_from_slice(&[0; 12]); // crc, sizes
+    f.extend_from_slice(&[1, 0, 0, 0]); // name and extra lengths
+    f.push(b'a');
+    let cd_at = f.len() as u64;
+    // Its central header (§4.3.12), 47 bytes.
+    let mut header = CENTRAL_HEADER.to_vec();
+    header.extend_from_slice(&[10, 0, 10, 0, 0, 0, 0, 0, 0, 0, 0x21, 0]); // versions, flags, method, time, date
+    header.extend_from_slice(&[0; 12]); // crc, sizes
+    header.extend_from_slice(&[1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]); // name, extra, comment, disk, attributes
+    header.extend_from_slice(&[0; 4]); // local header offset
+    header.push(b'a');
+    let count = entries.min((64 << 20) / header.len() as u64);
+    for _ in 0..count {
+        f.extend_from_slice(&header);
+    }
+    let cd_len = f.len() as u64 - cd_at;
+    let eocd64_at = f.len() as u64;
+    // ZIP64 end of central directory record (§4.3.14), 56 bytes.
+    f.extend_from_slice(EOCD64);
+    f.extend_from_slice(&44u64.to_le_bytes()); // size of the rest of the record
+    f.extend_from_slice(&[45, 0, 45, 0]); // versions
+    f.extend_from_slice(&[0; 8]); // this disk, disk with the central directory
+    f.extend_from_slice(&entries.to_le_bytes()); // entries on this disk
+    f.extend_from_slice(&entries.to_le_bytes()); // entries
+    f.extend_from_slice(&cd_len.to_le_bytes());
+    f.extend_from_slice(&cd_at.to_le_bytes());
+    // Its locator (§4.3.15), 20 bytes.
+    f.extend_from_slice(EOCD64_LOCATOR);
+    f.extend_from_slice(&0u32.to_le_bytes());
+    f.extend_from_slice(&eocd64_at.to_le_bytes());
+    f.extend_from_slice(&1u32.to_le_bytes());
+    // End of central directory record (§4.3.16), 22 bytes, each count and offset "see ZIP64".
+    f.extend_from_slice(EOCD);
+    f.extend_from_slice(&[0; 4]);
+    f.extend_from_slice(&[0xFF; 12]);
+    f.extend_from_slice(&[0; 2]);
+    f
+}
+
+#[cfg(test)]
+mod tests {
+    use super::{CENTRAL_HEADER, EOCD, check, zip64_entry_count_sample};
+
+    /// A plain package's tail: one central header at 0, then an end record declaring `entries`.
+    fn plain(entries: u16, comment: &[u8]) -> Vec<u8> {
+        let mut f = CENTRAL_HEADER.to_vec();
+        f.resize(46, 0);
+        f.extend_from_slice(EOCD);
+        f.extend_from_slice(&[0, 0, 0, 0]);
+        f.extend_from_slice(&entries.to_le_bytes());
+        f.extend_from_slice(&entries.to_le_bytes());
+        f.extend_from_slice(&46u32.to_le_bytes());
+        f.extend_from_slice(&0u32.to_le_bytes());
+        f.extend_from_slice(&(comment.len() as u16).to_le_bytes());
+        f.extend_from_slice(comment);
+        f
+    }
+
+    #[test]
+    fn a_zip64_end_record_declaring_too_many_entries_is_refused() {
+        // #3660: `zip` would reserve and parse them all before the reader's cap.
+        let sample = zip64_entry_count_sample(20_001);
+        assert!(check(&sample, 20_000).is_err_and(|e| e.contains("20001 entries")));
+        assert!(check(&zip64_entry_count_sample(5), 20_000).is_ok());
+        // Junk before the package (a self-extracting stub) doesn't hide it.
+        let mut prefixed = vec![0x55u8; 4096];
+        prefixed.extend_from_slice(&sample);
+        assert!(check(&prefixed, 20_000).is_err());
+    }
+
+    #[test]
+    fn a_plain_end_record_is_checked_as_zip_reads_it() {
+        assert!(check(&plain(3, b""), 20_000).is_ok());
+        assert!(check(&plain(30_000, b"a comment"), 20_000).is_err_and(|e| e.contains("30000 entries")));
+        // A record whose comment runs past the end is not the one `zip` uses.
+        let mut cut = plain(30_000, b"");
+        cut.truncate(cut.len() - 2);
+        cut.extend_from_slice(&500u16.to_le_bytes());
+        assert!(check(&cut, 20_000).is_ok());
+        assert!(check(b"not a package", 20_000).is_ok());
+        assert!(check(b"", 20_000).is_ok());
+    }
+
+    #[test]
+    fn a_fake_end_record_after_the_real_one_does_not_hide_it() {
+        // `zip` skips a last record with no central header before it and uses the one before.
+        let mut f = zip64_entry_count_sample(1_000_000);
+        f.extend_from_slice(EOCD);
+        f.extend_from_slice(&[0, 0, 0, 0]);
+        f.extend_from_slice(&7u16.to_le_bytes());
+        f.extend_from_slice(&7u16.to_le_bytes());
+        f.extend_from_slice(&0u32.to_le_bytes());
+        f.extend_from_slice(&((f.len() - 12) as u32).to_le_bytes());
+        f.extend_from_slice(&0u16.to_le_bytes());
+        assert!(check(&f, 20_000).is_err());
+    }
+
+    #[test]
+    fn many_end_records_are_refused() {
+        // Each declares one entry and has no central header before it, so `zip` would try them all.
+        let record = [0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0];
+        let f: Vec<u8> = (0..20).flat_map(|_| EOCD.iter().copied().chain(record)).collect();
+        assert!(check(&f, 20_000).is_err_and(|e| e.contains("end of central directory records")));
+    }
+}
@@ -20,7 +20,9 @@ const C_ICO: u16 = 0x2A42; // palette colour
 const C_CR: u16 = 0x6870; // explicit COLORREF
 const C_HPS: u16 = 0x4A43; // half-points
 const C_HPS_POS: u16 = 0x4845; // super/subscript by half-points
-const C_FTC0: u16 = 0x4A4F; // font index
+const C_FTC0: u16 = 0x4A4F; // font index (ASCII)
+const C_FTC1: u16 = 0x4A50; // East Asian font index (sprmCRgFtc1)
+const C_FTC_BI: u16 = 0x4A5E; // complex-script font index (sprmCFtcBi)
 const C_HIGHLIGHT: u16 = 0x2A0C;
 const C_ISTD: u16 = 0x4A30; // character style

@@ -254,6 +256,16 @@ pub(crate) fn apply_char(p: &mut CharProps, prl: &Prl, base: &CharProps, fonts:
                p.font = Some(name.clone());
            }
        }
+        C_FTC1 => {
+            if let Some(name) = u16_of(prl).and_then(|i| fonts.get(i as usize)) {
+                p.font_ea = Some(name.clone());
+            }
+        }
+        C_FTC_BI => {
+            if let Some(name) = u16_of(prl).and_then(|i| fonts.get(i as usize)) {
+                p.font_cs = Some(name.clone());
+            }
+        }
        C_HIGHLIGHT => {
            if let Some(h) = HL.get(prl.operand.first().copied().unwrap_or(0) as usize) {
                p.highlight = Some(*h);
@@ -417,10 +429,15 @@ pub(crate) fn apply_para(p: &mut ParaProps, prl: &Prl) {

 /// Build the model stylesheet: resolve each raw style's chain (base → this) and the ids we
 /// reference from paragraphs.
-pub(crate) fn stylesheet(raw: &[crate::stsh::RawStyle], fonts: &[String]) -> StyleSheet {
+pub(crate) fn stylesheet(raw: &[crate::stsh::RawStyle], fonts: &[String], standard: crate::stsh::StandardFonts) -> StyleSheet {
    let mut sheet = StyleSheet::empty();
-    // Word's base character properties: font index 0 of the table, 10 pt ([MS-DOC] sprmCHps).
-    sheet.default_chr.font = fonts.first().cloned();
+    // Word's base character properties: the stylesheet's standard fonts (Aptos in a Word 365
+    // file, whose font table still starts with Times New Roman, #5066), else font index 0 of the
+    // table; 10 pt ([MS-DOC] sprmCHps).
+    let font = |i: Option<u16>| i.and_then(|i| fonts.get(usize::from(i))).cloned();
+    sheet.default_chr.font = font(standard.ascii).or_else(|| fonts.first().cloned());
+    sheet.default_chr.font_ea = font(standard.east_asian);
+    sheet.default_chr.font_cs = font(standard.bidi);
    if sheet.default_chr.size.is_none() {
        sheet.default_chr.size = Some(10.0);
    }
@@ -233,6 +233,7 @@ pub fn read_with_password(bytes: &[u8], password: Option<&str>) -> Result<Docume
        Some((f, l)) => stsh::parse(&table, f, l)?,
        None => Vec::new(),
    };
+    let standard_fonts = fib.pair(fib::pair::STSHF).map(|(f, l)| stsh::standard_fonts(&table, f, l)).unwrap_or_default();
    let chpx_bins = match fib.pair(fib::pair::PLCF_BTE_CHPX) {
        Some((f, l)) => Bins::parse(&table, f, l)?,
        None => Bins::parse(&[], 0, 0)?,
@@ -247,7 +248,7 @@ pub fn read_with_password(bytes: &[u8], password: Option<&str>) -> Result<Docume
        doc.core = core;
    }
    doc.body.clear();
-    doc.styles = fmt::stylesheet(&raw_styles, &fonts);
+    doc.styles = fmt::stylesheet(&raw_styles, &fonts, standard_fonts);
    doc.numbering = list::parse(&table, &fib, &fonts);
    let sheet: StyleSheet = doc.styles.clone();
    let sects = sections::parse(&word, &table, &fib);
@@ -20,6 +20,41 @@ pub(crate) struct RawStyle {
    pub(crate) chpx: Vec<u8>,
 }

+/// The stylesheet's standard fonts, the ones text takes when neither its style nor its own
+/// formatting names a font: font-table indexes of the ASCII, East Asian and other
+/// (non-East-Asian, non-ASCII) fonts (`Stshif.rgftcStandardChpStsh`, [MS-DOC] §2.9.273) and the
+/// complex-script font (`Stshi.ftcBi`, §2.9.272). `None` where the header is too short for one.
+#[derive(Clone, Copy, Debug, Default, PartialEq, Eq)]
+pub(crate) struct StandardFonts {
+    pub(crate) ascii: Option<u16>,
+    pub(crate) east_asian: Option<u16>,
+    pub(crate) other: Option<u16>,
+    pub(crate) bidi: Option<u16>,
+}
+
+/// `StshiLsd.cbLSD`: the size of a latent style's `LSD` ([MS-DOC] `StshiLsd`).
+const LSD_SIZE: u16 = 4;
+
+/// The [`StandardFonts`] in the header (`Stshi`) of the `STSH` at `table[fc..fc+lcb]`: Word 365
+/// writes Aptos there while the font table's first entry is still Times New Roman (#5066).
+pub(crate) fn standard_fonts(table: &[u8], fc: u32, lcb: u32) -> StandardFonts {
+    let Some(stsh) = (fc as usize).checked_add(lcb as usize).and_then(|end| table.get(fc as usize..end)) else { return StandardFonts::default() };
+    let cb_stshi = stsh.get(0..2).map_or(0, |b| u16::from_le_bytes([b[0], b[1]]) as usize);
+    let stshi = stsh.get(2..).map_or(&[][..], |s| s.get(..cb_stshi.min(s.len())).unwrap_or(&[]));
+    // Stshif: cstd, cbSTDBaseInFile, flags, stiMaxWhenSaved, istdMaxFixedWhenSaved,
+    // nVerBuiltInNamesWhenSaved (2 bytes each), then rgftcStandardChpStsh (3 × 2); ftcBi follows.
+    let at = |o: usize| stshi.get(o..o + 2).map(|b| u16::from_le_bytes([b[0], b[1]]));
+    // ftcBi is optional, and the latent style data (`StshiLsd`, whose `cbLSD` MUST be 4) follows
+    // it: a 4 right after it says it is there, a 4 in its place with more data after says it was
+    // left out. A header that ends right after the field (as Word 97's) has it, a 4 included.
+    let bidi = match (at(18), at(20)) {
+        (b, Some(LSD_SIZE) | None) => b,
+        (Some(LSD_SIZE), Some(_)) => None,
+        (b, _) => b,
+    };
+    StandardFonts { ascii: at(12), east_asian: at(14), other: at(16), bidi }
+}
+
 /// Parse the `STSH` at `table[fc..fc+lcb]`.
 pub(crate) fn parse(table: &[u8], fc: u32, lcb: u32) -> Result<Vec<RawStyle>, DocbinError> {
    let fc = fc as usize;
@@ -421,7 +421,17 @@ fn std_bytes_next(stk: u8, istd_base: u16, istd_next: u16, name: &str, papx: &[u

 /// An STSH with the given styles.
 fn stsh_bytes(styles: &[Vec<u8>]) -> Vec<u8> {
-    let stshi = [(styles.len() as u16).to_le_bytes(), 10u16.to_le_bytes()].concat(); // cstd, cbSTDBaseInFile
+    stsh_bytes_with(styles, None)
+}
+
+/// An STSH with the given styles; with `standard`, a full Stshif and ftcBi giving the standard
+/// fonts' indexes (ASCII, East Asian, other, complex script), as Word writes them.
+fn stsh_bytes_with(styles: &[Vec<u8>], standard: Option<[u16; 4]>) -> Vec<u8> {
+    let mut stshi = [(styles.len() as u16).to_le_bytes(), 10u16.to_le_bytes()].concat(); // cstd, cbSTDBaseInFile
+    if let Some(fonts) = standard {
+        stshi.extend_from_slice(&[0u8; 8]); // flags, stiMaxWhenSaved, istdMaxFixedWhenSaved, nVerBuiltInNamesWhenSaved
+        stshi.extend(fonts.iter().flat_map(|f| f.to_le_bytes()));
+    }
    let mut v = (stshi.len() as u16).to_le_bytes().to_vec();
    v.extend_from_slice(&stshi);
    for s in styles {
@@ -532,6 +542,8 @@ struct RichExtra {
    prm: u16,
    alt_fonts: Vec<Option<&'static str>>,
    style_next: Vec<u16>,
+    /// The stylesheet's standard fonts (ASCII, East Asian, other, complex script).
+    standard_fonts: Option<[u16; 4]>,
 }

 fn rich_doc(spec: &RichSpec) -> Vec<u8> {
@@ -571,7 +583,7 @@ fn rich_doc_with(spec: &RichSpec, extra: &RichExtra) -> Vec<u8> {
        .enumerate()
        .map(|(i, s)| std_bytes_next(s.0, s.1, extra.style_next.get(i).copied().unwrap_or(0x0FFF), s.2, &s.3, &s.4))
        .collect();
-    let stsh = stsh_bytes(&styles);
+    let stsh = stsh_bytes_with(&styles, extra.standard_fonts);
    let fonts: Vec<(&str, Option<&str>)> = spec.fonts.iter().enumerate().map(|(i, n)| (*n, extra.alt_fonts.get(i).copied().flatten())).collect();
    let ffn = ffn_bytes_alt(&fonts);
    let clx_at = 0x40;
@@ -654,6 +666,67 @@ fn font_and_size_from_style_and_direct() {
    assert!(p.runs.first().is_some_and(|r| r.props.font.as_deref() == Some("Arial")), "run font: {:?}", p.runs);
 }

+/// #5066: the default font is the stylesheet's standard font (Aptos in a Word 365 file), not
+/// the font table's first entry (still Times New Roman there); the East Asian and complex-script
+/// standard fonts are the defaults for those scripts, and runs name theirs with sprmCRgFtc1 and
+/// sprmCFtcBi.
+#[test]
+fn the_default_font_is_the_stylesheets_standard_font() {
+    let spec = |runs: Vec<(u32, Vec<u8>)>| RichSpec {
+        text: "Hello\r",
+        styles: vec![(1, 0x0FFF, "Normal", Vec::new(), grpprl(&[(0x4A43, &24u16.to_le_bytes())]))],
+        fonts: vec!["Times New Roman", "Symbol", "Arial", "Aptos", "MS Mincho", "Arial Unicode"],
+        chpx_runs: runs,
+        papx_entries: vec![(0, 0, Vec::new())],
+    };
+    let with = |standard: Option<[u16; 4]>, runs| {
+        read(&rich_doc_with(&spec(runs), &RichExtra { standard_fonts: standard, ..Default::default() })).expect("opens")
+    };
+    let doc = with(Some([3, 4, 3, 2]), vec![(0, Vec::new())]);
+    let d = &doc.styles.default_chr;
+    assert_eq!((d.font.as_deref(), d.font_ea.as_deref(), d.font_cs.as_deref()), (Some("Aptos"), Some("MS Mincho"), Some("Arial")));
+    let p = para0(&doc);
+    let resolved = doc.styles.resolve_char(p.props.style.as_deref(), &p.runs[0].props);
+    assert_eq!(resolved.font, "Aptos", "Normal names no font, so the text takes the standard one");
+    // A run's own East Asian and complex-script fonts.
+    let doc = with(Some([3, 4, 3, 2]), vec![(0, grpprl(&[(0x4A50, &5u16.to_le_bytes()), (0x4A5E, &3u16.to_le_bytes())]))]);
+    let run = &para0(&doc).runs[0].props;
+    assert_eq!((run.font_ea.as_deref(), run.font_cs.as_deref()), (Some("Arial Unicode"), Some("Aptos")));
+    // Indexes past the font table, or a header too short to hold them, keep entry 0.
+    for standard in [Some([99, 99, 99, 99]), None] {
+        let doc = with(standard, vec![(0, Vec::new())]);
+        let d = &doc.styles.default_chr;
+        assert_eq!((d.font.as_deref(), d.font_ea.as_deref(), d.font_cs.as_deref()), (Some("Times New Roman"), None, None), "{standard:?}");
+    }
+}
+
+/// #5066: `ftcBi` is optional in the stylesheet header, and the latent style data (`cbLSD`, 4,
+/// then the LSDs) follows it: a header that leaves it out doesn't make `cbLSD` the complex-script
+/// font.
+#[test]
+fn the_complex_script_standard_font_is_read_only_when_present() {
+    let header = |tail: &[u16]| -> Vec<u8> {
+        // cstd, cbSTDBaseInFile, flags, stiMaxWhenSaved, istdMaxFixedWhenSaved,
+        // nVerBuiltInNamesWhenSaved; ASCII, East Asian, other.
+        let words = [&[0u16, 10, 0, 1, 0, 0, 3, 4, 3][..], tail].concat();
+        let stshi: Vec<u8> = words.iter().flat_map(|w| w.to_le_bytes()).collect();
+        [(stshi.len() as u16).to_le_bytes().to_vec(), stshi].concat()
+    };
+    let bidi = |tail: &[u16]| {
+        let b = header(tail);
+        crate::stsh::standard_fonts(&b, 0, b.len() as u32).bidi
+    };
+    assert_eq!(bidi(&[2]), Some(2), "a header that ends with it (Word 97)");
+    assert_eq!(bidi(&[4]), Some(4), "font 4, at the end of the header");
+    assert_eq!(bidi(&[2, 4, 0, 0]), Some(2), "followed by the latent style data");
+    assert_eq!(bidi(&[4, 4, 0, 0]), Some(4), "font 4, followed by the latent style data");
+    assert_eq!(bidi(&[4, 0, 0]), None, "left out: cbLSD is in its place");
+    assert_eq!(bidi(&[]), None, "a header too short for it");
+    let b = header(&[4, 0, 0]);
+    let fonts = crate::stsh::standard_fonts(&b, 0, b.len() as u32);
+    assert_eq!((fonts.ascii, fonts.east_asian, fonts.other), (Some(3), Some(4), Some(3)), "the other fonts are read all the same");
+}
+
 #[test]
 fn paragraph_props_and_heading_style() {
    let spec = RichSpec {
@@ -76,7 +76,7 @@ pub fn sniff_xml(bytes: &[u8]) -> XmlKind {

 /// Read a Flat OPC file.
 pub fn read(bytes: &[u8]) -> Result<Document, DocxError> {
-    crate::read::read_opened(&unpack(bytes)?)
+    crate::read::read_opened(&unpack(bytes)?, None)
 }

 /// The data element of the part being read.
@@ -32,7 +32,7 @@ pub use crypt::{
 };
 pub use flat::{XmlKind, sniff_xml, write as write_flat};
 pub use fonts::{EmbeddedFont, embedded_fonts};
-pub use read::read;
+pub use read::{AltChunk, ChunkImporter, ReadOptions, read};
 pub use smart_art::smart_art_items;
 pub use wordml2003_write::write as write_wordml2003;
 pub use write::{write, write_as};
@@ -56,11 +56,19 @@ pub fn read_xml(bytes: &[u8]) -> Result<wordcraft_doc::Document, DocxError> {

 /// The title (`dc:title`) in a package's core properties part (`docProps/core.xml`, ECMA-376
 /// Part 2 §11), trimmed; `None` when the part can't be read or has no title. Only the package's
-/// directory and at most `max_part` bytes of that part are read: for listing templates by name
-/// without reading the whole package.
-pub fn package_title<R: std::io::Read + std::io::Seek>(package: R, max_part: u64) -> Option<String> {
+/// directory and at most `max_part` bytes of that part are inflated: for listing templates by
+/// name without unpacking the whole package. A package over 64 MiB, or one declaring more entries
+/// than `MAX_ENTRIES` (20,000), has none (#3660).
+pub fn package_title<R: std::io::Read>(package: R, max_part: u64) -> Option<String> {
    use std::io::Read;
-    let mut zip = zip::ZipArchive::new(package).ok()?;
+    const MAX_PACKAGE: u64 = 64 << 20;
+    let mut bytes = Vec::new();
+    package.take(MAX_PACKAGE + 1).read_to_end(&mut bytes).ok()?;
+    if bytes.len() as u64 > MAX_PACKAGE {
+        return None;
+    }
+    wordcraft_doc::zip_guard::check(&bytes, package::MAX_ENTRIES as u64).ok()?;
+    let mut zip = zip::ZipArchive::new(std::io::Cursor::new(bytes)).ok()?;
    let part = zip.by_name("docProps/core.xml").ok()?;
    let mut xml = Vec::new();
    part.take(max_part).read_to_end(&mut xml).ok()?;
@@ -84,11 +92,18 @@ pub(crate) const CANVAS_SPIN_EXT: &str = "urn:wordcraft:canvas-xfrm";
 /// `password` first. An encrypted package without a password gives
 /// [`DocxError::PasswordRequired`]; a wrong one, [`DocxError::WrongPassword`].
 pub fn read_with_password(bytes: &[u8], password: Option<&str>) -> Result<wordcraft_doc::Document, DocxError> {
+    read_with(bytes, &ReadOptions { password, ..Default::default() })
+}
+
+/// Read a package with options: the password of an encrypted one ([`read_with_password`]) and
+/// an importer for `w:altChunk` content in formats other than WordprocessingML (see
+/// [`ChunkImporter`]).
+pub fn read_with(bytes: &[u8], opts: &ReadOptions) -> Result<wordcraft_doc::Document, DocxError> {
    if is_encrypted(bytes) {
-        let package = zeroize::Zeroizing::new(decrypt(bytes, password)?);
-        return read::read_package(&package);
+        let package = zeroize::Zeroizing::new(decrypt(bytes, opts.password)?);
+        return read::read_package(&package, opts.chunks);
    }
-    read(bytes)
+    read::read_package(bytes, opts.chunks)
 }

 /// Which kind of WordprocessingML package to write. The main part's content type differs per
@@ -47,6 +47,7 @@ pub mod rt {
    pub const DIAGRAM_COLORS: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/diagramColors";
    pub const CUSTOM_XML: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/customXml";
    pub const GLOSSARY: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/glossaryDocument";
+    pub const AF_CHUNK: &str = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/aFChunk";
 }

 /// Content type of a glossary document part (ECMA-376 Part 1 §11.3.8).
@@ -212,6 +213,8 @@ impl Package {
    }

    pub fn open_limited(bytes: &[u8], max_part: u64, max_total: u64) -> Result<Package, DocxError> {
+        // Before `zip` parses every entry the end record declares (#3660).
+        wordcraft_doc::zip_guard::check(bytes, MAX_ENTRIES as u64).map_err(DocxError::Limit)?;
        let mut zip = zip::ZipArchive::new(Cursor::new(bytes)).map_err(|e| DocxError::Zip(e.to_string()))?;
        if zip.len() > MAX_ENTRIES {
            return Err(DocxError::Limit(format!("{} zip entries", zip.len())));
@@ -252,6 +255,11 @@ impl Package {
        Ok(Package::new(files))
    }

+    /// The bytes of all its parts.
+    pub fn size(&self) -> u64 {
+        self.files.values().map(|f| f.len() as u64).sum()
+    }
+
    /// A package from parts already in memory (a Flat OPC file, a converted document), by part
    /// name without the leading `/`.
    pub fn from_parts(files: BTreeMap<String, Vec<u8>>) -> Package {
@@ -524,6 +532,21 @@ mod tests {
        assert!(folded.iter().all(|w| w.contains("2 zip entries")), "{folded:?}");
    }

+    #[test]
+    fn a_package_declaring_too_many_entries_is_refused_before_zip_reads_them() {
+        // #3660: a ZIP64 end record declaring more than MAX_ENTRIES.
+        let bomb = wordcraft_doc::zip_guard::zip64_entry_count_sample(MAX_ENTRIES as u64 + 1);
+        assert!(matches!(Package::open(&bomb), Err(DocxError::Limit(e)) if e.contains("declares")));
+        assert_eq!(crate::package_title(std::io::Cursor::new(bomb), 1 << 20), None);
+        // A real package still opens, and its title still reads.
+        let z = zip_of(&[(
+            "docProps/core.xml",
+            br#"<cp:coreProperties xmlns:cp="http://schemas.openxmlformats.org/package/2006/metadata/core-properties" xmlns:dc="http://purl.org/dc/elements/1.1/"><dc:title>Plan</dc:title></cp:coreProperties>"#.to_vec(),
+        )]);
+        assert!(Package::open(&z).is_ok());
+        assert_eq!(crate::package_title(std::io::Cursor::new(z), 1 << 20).as_deref(), Some("Plan"));
+    }
+
    #[test]
    fn decompression_bomb_is_rejected() {
        let z = zip_of(&[("word/document.xml", vec![b' '; 2_000_000])]);
@@ -587,7 +610,7 @@ mod tests {
            ("word/media/pic.png", b"\x89PNG picture bytes".to_vec()),
        ]);
        let pkg = Package::open(&z).unwrap();
-        let doc = crate::read::read_opened(&pkg).unwrap();
+        let doc = crate::read::read_opened(&pkg, None).unwrap();
        let held = doc.media.values().next().expect("the picture is read");
        assert!(Arc::ptr_eq(held, pkg.get_shared("word/media/pic.png").unwrap()), "the same bytes, not a copy");
        assert_eq!(pkg.get("/WORD/media/PIC.png"), Some(&b"\x89PNG picture bytes"[..]), "lookups are unchanged");
@@ -1,6 +1,7 @@
 //! DOCX → [`Document`].

 pub(crate) mod chart;
+mod chunk;
 pub(crate) mod diagram;
 mod drawing_color;
 mod embed;
@@ -13,7 +14,7 @@ mod shape_style;
 mod story;
 mod wordart;

-use std::collections::{BTreeMap, HashMap, HashSet};
+use std::collections::{HashMap, HashSet};
 use std::sync::Arc;

 use wordcraft_doc::control::{ControlEvent, ControlKind, control_events, pair_events};
@@ -46,6 +47,9 @@ pub(crate) use meta::PROTECTION_HASH_ATTRS;
 pub(crate) use styles::LOWER_NAMES;
 use toc::{is_toc, toc_control_to_toc, toc_field_to_heading, unstyle_toc_links};

+use chunk::ChunkState;
+pub use chunk::{AltChunk, ChunkImporter, ReadOptions};
+
 /// Most fonts read from a font table, and the longest name (hostile files).
 const MAX_TABLE_FONTS: usize = 2000;
 const MAX_FONT_NAME: usize = 256;
@@ -87,12 +91,12 @@ pub(crate) struct Reader<'p> {
    pub bookmarks: HashMap<String, String>,
    /// Restrict Editing exception ranges (`w:permStart`): the file's id → ours.
    pub perms: HashMap<String, u32>,
+    /// The next permission range id to give out (chunks' ranges take theirs from it too).
+    pub next_perm: u32,
    /// Chart and diagram parts by path, parsed once.
    graphic_parts: HashMap<String, Option<Arc<El>>>,
    /// Charts and diagrams by (kind, part path, width bits, height bits), built once.
    graphics: HashMap<(GraphicKind, String, u32, u32), Arc<Graphic>>,
-    /// Graphic work left before charts and diagrams are left empty (see [`MAX_GRAPHIC_WORK`]).
-    graphic_budget: usize,
    /// Drawing ids connectors are glued to (`a:stCxn`, `a:endCxn`), and whether any shape's id
    /// was kept while reading (those not glued to are dropped at the end).
    pub glued: HashSet<u32>,
@@ -103,6 +107,8 @@ pub(crate) struct Reader<'p> {
    content_types: Option<ContentTypes>,
    /// Parts kept for charts, diagrams and OLE objects (see `embed`).
    embedded: EmbeddedManifest,
+    /// What alternative format chunks may still bring in, and how (see [`chunk`]).
+    chunks: ChunkState<'p>,
    /// Linked text boxes as read: (`wps:txbx/@id` or `wps:linkedTxbx/@id`, `seq` (0 for the box
    /// holding the text), text box part). See [`link_text_boxes`].
    pub txbx_links: Vec<(u32, u32, u32)>,
@@ -125,17 +131,24 @@ pub fn read(bytes: &[u8]) -> Result<Document, DocxError> {
    if crate::is_encrypted(bytes) {
        return crate::read_with_password(bytes, None);
    }
-    read_package(bytes)
+    read_package(bytes, None)
 }

-/// [`read`] for bytes known not to be an encrypted package.
-pub(crate) fn read_package(bytes: &[u8]) -> Result<Document, DocxError> {
-    read_opened(&Package::open(bytes)?)
+/// [`read`] for bytes known not to be an encrypted package; `chunks` imports alternative format
+/// chunks in formats other than WordprocessingML.
+pub(crate) fn read_package(bytes: &[u8], chunks: Option<ChunkImporter>) -> Result<Document, DocxError> {
+    read_opened(&Package::open(bytes)?, chunks)
 }

 /// Read a package whose parts are already unpacked (from a zip, a Flat OPC file or a converted
 /// Word 2003 XML document).
-pub(crate) fn read_opened(pkg: &Package) -> Result<Document, DocxError> {
+pub(crate) fn read_opened(pkg: &Package, chunks: Option<ChunkImporter>) -> Result<Document, DocxError> {
+    read_opened_in(pkg, ChunkState::new(chunks)).map(|(doc, ..)| doc)
+}
+
+/// [`read_opened`] with what chunks may still bring in; also gives the chunks, chunk bytes and
+/// graphic work left of that.
+fn read_opened_in(pkg: &Package, chunks: ChunkState) -> Result<(Document, usize, u64, usize), DocxError> {
    let root_rels = pkg.rels("");
    let main = root_rels.by_type(rt::OFFICE_DOC).filter(|r| !r.external).map(|r| r.target.clone()).unwrap_or_else(|| "word/document.xml".to_string());
    let Some(root) = pkg.xml_strict(&main)? else {
@@ -164,9 +177,9 @@ pub(crate) fn read_opened(pkg: &Package) -> Result<Document, DocxError> {
        comments_ended: HashSet::new(),
        bookmarks: HashMap::new(),
        perms: HashMap::new(),
+        next_perm: 0,
        graphic_parts: HashMap::new(),
        graphics: HashMap::new(),
-        graphic_budget: MAX_GRAPHIC_WORK,
        glued: HashSet::new(),
        shape_ids: false,
        main: main.clone(),
@@ -175,6 +188,7 @@ pub(crate) fn read_opened(pkg: &Package) -> Result<Document, DocxError> {
        txbx_links: Vec::new(),
        theme_fills: Default::default(),
        group_fill: None,
+        chunks,
    };
    r.pc.major_font = r.doc.settings.major_font.clone();
    r.pc.minor_font = r.doc.settings.minor_font.clone();
@@ -267,7 +281,7 @@ pub(crate) fn read_opened(pkg: &Package) -> Result<Document, DocxError> {
    if let Some(body) = root.child("w:body") {
        let mut sc = StoryCtx::default();
        let mut blocks = Blocks::new();
-        r.read_blocks(&mut sc, body, &rels, &mut blocks, 0);
+        r.read_blocks(&mut sc, body, &rels, &mut blocks, r.chunks.depth);
        r.flush_pending(&mut sc, &mut blocks);
        r.doc.body = blocks;
        unstyle_toc_links(&mut r.doc);
@@ -321,9 +335,10 @@ pub(crate) fn read_opened(pkg: &Package) -> Result<Document, DocxError> {
    if r.shape_ids {
        keep_glued_ids(&mut r.doc, &r.glued);
    }
+    let (left, bytes, graphic) = (r.chunks.left, r.chunks.bytes, r.chunks.graphic);
    let mut doc = r.doc;
    doc.ensure_nonempty();
-    Ok(doc)
+    Ok((doc, left, bytes, graphic))
 }

 /// Most move ranges open at once, and most move names tracked, per file.
@@ -0,0 +1,187 @@
+//! `w:altChunk` (ECMA-376 §17.17.2.1): content kept in another format (HTML, MHT, RTF, plain
+//! text, another WordprocessingML document) that Word brings into the document where the chunk
+//! stands when it opens the file. A nested WordprocessingML package is read here; other formats
+//! go to the caller's [`ChunkImporter`] (the format crates sit beside this one), and without one
+//! the chunk is left out.
+
+use std::sync::Arc;
+
+use wordcraft_doc::{Block, Blocks, Document};
+
+use super::story::StoryCtx;
+use super::{Reader, part_of};
+use crate::package::{ContentTypes, MAX_PART, MAX_TOTAL, Package, Rels, rt};
+use crate::xml::El;
+
+/// Most chunks brought in from one file (nested documents' chunks included).
+pub(crate) const MAX_CHUNKS: usize = 256;
+/// Deepest a WordprocessingML chunk may nest in another.
+pub(crate) const MAX_CHUNK_NESTING: u8 = 4;
+
+/// An alternative format chunk's content, as a [`ChunkImporter`] gets it.
+pub struct AltChunk<'a> {
+    /// The part's content type (`application/xhtml+xml`, `message/rfc822`, `application/rtf`,
+    /// `text/plain`…), empty when the package doesn't give one.
+    pub content_type: &'a str,
+    /// The part's name (`word/afchunk.mht`): its extension tells the format when the content type
+    /// doesn't.
+    pub part: &'a str,
+    pub bytes: &'a [u8],
+}
+
+/// Turns a chunk the docx reader can't read itself into a document whose body takes the chunk's
+/// place; `None` leaves the chunk out.
+pub type ChunkImporter<'a> = &'a dyn Fn(&AltChunk) -> Option<Document>;
+
+/// How [`crate::read_with`] reads a package.
+#[derive(Clone, Copy, Default)]
+pub struct ReadOptions<'a> {
+    /// The password of an encrypted package.
+    pub password: Option<&'a str>,
+    /// Imports `w:altChunk` content in formats other than WordprocessingML.
+    pub chunks: Option<ChunkImporter<'a>>,
+}
+
+impl Reader<'_> {
+    /// Bring the chunk `e` names into `out` where it stands (`depth`: of the XML around it).
+    pub(crate) fn read_alt_chunk(&mut self, sc: &mut StoryCtx, e: &El, rels: &Rels, out: &mut Blocks, depth: usize) {
+        let Some(part) = e.attr("r:id").and_then(|id| part_of(rels, id, rt::AF_CHUNK)) else { return };
+        let pkg = self.pkg;
+        let Some(bytes) = pkg.get(&part) else {
+            log::warn!("docx: alternative format chunk {part} is missing");
+            return;
+        };
+        // Every chunk's bytes count against one budget for the whole file, however often the same
+        // part is named and however deep chunks nest.
+        if self.chunks.left == 0 || bytes.len() as u64 > self.chunks.bytes {
+            log::warn!("docx: more than {MAX_CHUNKS} alternative format chunks or {MAX_TOTAL} bytes of them; the rest are left out");
+            return;
+        }
+        self.chunks.left -= 1;
+        self.chunks.bytes -= bytes.len() as u64;
+        let ct = self.content_types.get_or_insert_with(|| ContentTypes::read(pkg)).of(&part).unwrap_or("").to_string();
+        let chunk = AltChunk { content_type: &ct, part: &part, bytes };
+        // A WordprocessingML chunk is a package; one that isn't Word's fails to read (logged).
+        let is_docx = bytes.starts_with(b"PK\x03\x04");
+        let doc = if is_docx { self.read_nested(bytes, depth) } else { self.chunks.import.and_then(|f| f(&chunk)) };
+        let Some(mut doc) = doc else {
+            log::warn!("docx: alternative format chunk {part} ({ct}) is in a format WordCraft can't bring in");
+            return;
+        };
+        // The chunk flows in the section it stands in; its own section breaks don't come along.
+        for b in &mut doc.body {
+            if let Block::Para(p) = &**b
+                && p.section.is_some()
+                && let Block::Para(p) = Arc::make_mut(b)
+            {
+                p.section = None;
+            }
+        }
+        // `w:matchSrc`: a Word document's content keeps its own styles (other formats' formatting
+        // is their own already).
+        let keep = is_docx && e.child("w:altChunkPr").and_then(|p| p.child("w:matchSrc")).is_some_and(crate::units::on_off);
+        let mut blocks = self.doc.adopt_blocks(doc, keep, &mut self.next_perm);
+        // Bookmarks and comment marks read before the chunk stay before it.
+        self.flush_pending_before(sc, out, &mut blocks);
+        out.extend(blocks.into_iter().map(Arc::new));
+    }
+
+    /// A WordprocessingML chunk: its document, read with the same importer and what is left of
+    /// the budgets, its XML depth counting on from `depth`.
+    fn read_nested(&mut self, bytes: &[u8], depth: usize) -> Option<Document> {
+        if self.chunks.nesting >= MAX_CHUNK_NESTING {
+            log::warn!("docx: alternative format chunks nest deeper than {MAX_CHUNK_NESTING}; the inner ones are left out");
+            return None;
+        }
+        let pkg =
+            Package::open_limited(bytes, MAX_PART, self.chunks.bytes).map_err(|e| log::warn!("docx: unreadable nested document chunk: {e}")).ok()?;
+        self.chunks.bytes = self.chunks.bytes.saturating_sub(pkg.size());
+        let state = ChunkState { nesting: self.chunks.nesting + 1, depth: depth + 1, ..self.chunks };
+        let (doc, left, bytes, graphic) =
+            super::read_opened_in(&pkg, state).map_err(|e| log::warn!("docx: unreadable nested document chunk: {e}")).ok()?;
+        self.chunks.left = left;
+        self.chunks.bytes = bytes;
+        self.chunks.graphic = graphic;
+        Some(doc)
+    }
+}
+
+/// What a reader may still bring in from chunks.
+#[derive(Clone, Copy)]
+pub(crate) struct ChunkState<'a> {
+    /// Chunks left in the file's budget ([`MAX_CHUNKS`]).
+    pub left: usize,
+    /// Chunk bytes (nested packages unpacked) left in the file's budget.
+    pub bytes: u64,
+    /// Graphic work left in the file's budget ([`super::MAX_GRAPHIC_WORK`]): nested documents'
+    /// charts and diagrams count against it too.
+    pub graphic: usize,
+    /// How deep in nested WordprocessingML chunks this reader is.
+    pub nesting: u8,
+    /// The XML depth its stories start at (a nested document's: that of the chunk naming it).
+    pub depth: usize,
+    pub import: Option<ChunkImporter<'a>>,
+}
+
+impl<'a> ChunkState<'a> {
+    pub fn new(import: Option<ChunkImporter<'a>>) -> Self {
+        ChunkState { left: MAX_CHUNKS, bytes: MAX_TOTAL, graphic: super::MAX_GRAPHIC_WORK, nesting: 0, depth: 0, import }
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use std::io::Write;
+
+    use wordcraft_doc::{Block, Document, InlineObject};
+
+    use super::ChunkState;
+    use crate::package::Package;
+
+    /// A package whose body is `body`, with the document relationship `rel` (id, type, target) to
+    /// the part `(name, bytes)`.
+    fn package(body: &str, (id, ty, target): (&str, &str, &str), (name, bytes): (&str, &[u8])) -> Vec<u8> {
+        let ns = r#"xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing" xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main""#;
+        let rels =
+            |inner: &str| format!(r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">{inner}</Relationships>"#);
+        let rel = |id: &str, ty: &str, target: &str| {
+            format!(r#"<Relationship Id="{id}" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/{ty}" Target="{target}"/>"#)
+        };
+        let files: [(&str, Vec<u8>); 4] = [
+            ("_rels/.rels", rels(&rel("rId1", "officeDocument", "word/document.xml")).into_bytes()),
+            ("word/document.xml", format!(r#"<w:document {ns}><w:body>{body}</w:body></w:document>"#).into_bytes()),
+            ("word/_rels/document.xml.rels", rels(&rel(id, ty, target)).into_bytes()),
+            (name, bytes.to_vec()),
+        ];
+        let mut zw = zip::ZipWriter::new(std::io::Cursor::new(Vec::new()));
+        for (name, bytes) in files {
+            zw.start_file(name, zip::write::SimpleFileOptions::default()).unwrap();
+            zw.write_all(&bytes).unwrap();
+        }
+        zw.finish().unwrap().into_inner()
+    }
+
+    /// Whether the body's first paragraph's chart is drawn.
+    fn drawn(d: &Document) -> bool {
+        let Some(Block::Para(p)) = d.body.first().map(|b| &**b) else { panic!("no paragraph") };
+        p.objects.iter().any(|o| matches!(o, InlineObject::Graphic { graphic, .. } if !graphic.items.is_empty()))
+    }
+
+    #[test]
+    fn a_nested_documents_charts_count_against_the_files_graphic_budget() {
+        let chart = r#"<c:chartSpace xmlns:c="http://schemas.openxmlformats.org/drawingml/2006/chart"><c:chart><c:plotArea><c:barChart><c:ser><c:val><c:numLit><c:ptCount val="2"/><c:pt idx="0"><c:v>1</c:v></c:pt><c:pt idx="1"><c:v>2</c:v></c:pt></c:numLit></c:val></c:ser></c:barChart></c:plotArea></c:chart></c:chartSpace>"#;
+        let drawing = r#"<w:p><w:r><w:drawing><wp:inline><wp:extent cx="2743200" cy="1828800"/><wp:docPr id="5" name="G"/><a:graphic><a:graphicData uri="http://schemas.openxmlformats.org/drawingml/2006/chart"><c:chart xmlns:c="http://schemas.openxmlformats.org/drawingml/2006/chart" r:id="rIdChart"/></a:graphicData></a:graphic></wp:inline></w:drawing></w:r></w:p>"#;
+        let inner = package(drawing, ("rIdChart", "chart", "charts/chart1.xml"), ("word/charts/chart1.xml", chart.as_bytes()));
+        let outer = package(r#"<w:altChunk r:id="rIdC"/>"#, ("rIdC", "aFChunk", "afchunk.docx"), ("word/afchunk.docx", &inner));
+        let pkg = Package::open(&outer).unwrap();
+        let read = |graphic| {
+            let (d, .., left) = super::super::read_opened_in(&pkg, ChunkState { graphic, ..ChunkState::new(None) }).unwrap();
+            (drawn(&d), left)
+        };
+        let (shown, left) = read(super::super::MAX_GRAPHIC_WORK);
+        assert!(shown);
+        assert!(left < super::super::MAX_GRAPHIC_WORK, "the nested chart's work is the file's");
+        // With the file's budget spent, the nested document doesn't get one of its own.
+        assert_eq!(read(0), (false, 0));
+    }
+}
@@ -1,6 +1,6 @@
 use super::{
-    Arc, BTreeMap, Blocks, Comment, DocxError, El, HashMap, HashSet, MAX_PARTS, MAX_SOURCES, MAX_TEMPLATE_PATH, NoteKind, NumFormat, Paragraph,
-    PartKind, Reader, Rels, Seen, StoryCtx, Zoom, ZoomFit, int, math, on_off, para_block, props, tw, u32_of,
+    Arc, Blocks, Comment, DocxError, El, HashMap, MAX_PARTS, MAX_SOURCES, MAX_TEMPLATE_PATH, NoteKind, NumFormat, Paragraph, PartKind, Reader, Rels,
+    Seen, StoryCtx, Zoom, ZoomFit, int, math, on_off, para_block, props, tw, u32_of,
 };

 impl Reader<'_> {
@@ -120,7 +120,7 @@ impl Reader<'_> {
            let mut sc = StoryCtx::default();
            let mut blocks = Blocks::new();
            self.current_note = Some((kind, id));
-            self.read_blocks(&mut sc, n, &rels, &mut blocks, 0);
+            self.read_blocks(&mut sc, n, &rels, &mut blocks, self.chunks.depth);
            self.flush_pending(&mut sc, &mut blocks);
            self.current_note = None;
            if let Some(p) = self.doc.parts.get_mut(&id)
@@ -152,7 +152,7 @@ impl Reader<'_> {
        let mut sc = StoryCtx::default();
        let mut blocks = Blocks::new();
        self.in_note_sep = true;
-        self.read_blocks(&mut sc, n, rels, &mut blocks, 0);
+        self.read_blocks(&mut sc, n, rels, &mut blocks, self.chunks.depth);
        self.flush_pending(&mut sc, &mut blocks);
        self.in_note_sep = false;
        if blocks.is_empty() {
@@ -188,11 +188,7 @@ impl Reader<'_> {
        let rels = self.pkg.rels(path);
        let mut by_para: HashMap<String, u32> = HashMap::new();
        let mut next_id = 0u32;
-        let mut entries: BTreeMap<u32, Comment> = BTreeMap::new();
-        // Ids given before this part, looked up rather than scanned for every generated id: a
-        // scan per comment made many non-numeric ids quadratic (#3692). Ids given here are in
-        // `entries`.
-        let given: HashSet<u32> = self.comment_map.values().copied().collect();
+        // Straight into the document, beside the comments chunks in notes brought in.
        for c in root.children("w:comment") {
            if self.doc.parts.len() >= MAX_PARTS {
                break;
@@ -202,23 +198,26 @@ impl Reader<'_> {
                continue;
            }
            let id = match file_id.parse::<u32>() {
-                Ok(v) if !entries.contains_key(&v) => v,
+                Ok(v) if !self.doc.comments.contains_key(&v) => v,
                _ => {
-                    while entries.contains_key(&next_id) || given.contains(&next_id) {
+                    // Every id handed out so far is a key of `doc.comments` (inserted below).
+                    while self.doc.comments.contains_key(&next_id) {
                        next_id += 1;
                    }
                    next_id
                }
            };
+            // Taken before its text is read: a chunk there brings comments of its own.
+            self.doc.comments.insert(id, Comment::default());
            let mut sc = StoryCtx::default();
            let mut blocks = Blocks::new();
-            self.read_blocks(&mut sc, c, &rels, &mut blocks, 0);
+            self.read_blocks(&mut sc, c, &rels, &mut blocks, self.chunks.depth);
            self.flush_pending(&mut sc, &mut blocks);
            let part = self.doc.add_part(PartKind::Comment, blocks);
            if let Some(pid) = c.children("w:p").last().and_then(|p| p.attr("w14:paraId")) {
                by_para.insert(pid.to_string(), id);
            }
-            entries.insert(
+            self.doc.comments.insert(
                id,
                Comment {
                    author: c.attr("w:author").unwrap_or("").to_string(),
@@ -239,7 +238,7 @@ impl Reader<'_> {
                let Some(cid) = e.attr("w15:paraId").and_then(|p| by_para.get(p)).copied() else { continue };
                let parent = e.attr("w15:paraIdParent").and_then(|p| by_para.get(p)).copied().filter(|p| *p != cid);
                let done = e.attr("w15:done").is_some_and(|v| v == "1" || v == "true");
-                if let Some(c) = entries.get_mut(&cid) {
+                if let Some(c) = self.doc.comments.get_mut(&cid) {
                    c.parent = parent;
                    c.resolved = done;
                }
@@ -258,7 +257,7 @@ impl Reader<'_> {
            for e in x.els().filter(|e| e.local() == "commentId") {
                let Some(cid) = local(e, "paraId").and_then(|p| by_para.get(&p).copied()) else { continue };
                let Some(durable) = local(e, "durableId").filter(|d| crate::write::durable_id_ok(d)) else { continue };
-                if let Some(c) = entries.get_mut(&cid) {
+                if let Some(c) = self.doc.comments.get_mut(&cid) {
                    c.durable_id = Some(durable.to_ascii_uppercase());
                    by_durable.insert(durable.to_ascii_uppercase(), cid);
                }
@@ -270,12 +269,11 @@ impl Reader<'_> {
            for e in x.els().filter(|e| e.local() == "commentExtensible") {
                let Some(cid) = local(e, "durableId").and_then(|d| by_durable.get(&d.to_ascii_uppercase()).copied()) else { continue };
                let date = local(e, "dateUtc").filter(|d| d.len() <= 40 && d.chars().all(|c| c.is_ascii_alphanumeric() || ":-.+".contains(c)));
-                if let (Some(c), Some(date)) = (entries.get_mut(&cid), date) {
+                if let (Some(c), Some(date)) = (self.doc.comments.get_mut(&cid), date) {
                    c.date_utc = Some(date);
                }
            }
        }
-        self.doc.comments = entries;
        Ok(())
    }

@@ -144,7 +144,7 @@ impl Reader<'_> {
        sc.in_hf = true;
        let mut blocks = Blocks::new();
        self.hf_depth += 1;
-        self.read_blocks(&mut sc, &root, &rels, &mut blocks, 0);
+        self.read_blocks(&mut sc, &root, &rels, &mut blocks, self.chunks.depth);
        self.hf_depth = self.hf_depth.saturating_sub(1);
        self.flush_pending(&mut sc, &mut blocks);
        if let Some(p) = self.doc.parts.get_mut(&id)
@@ -166,9 +166,9 @@ fn children_of_choice(ac: &El) -> Option<&El> {
    ac.children("mc:Choice").find(|c| ok(c)).or_else(|| ac.child("mc:Fallback")).or_else(|| ac.child("mc:Choice"))
 }

-/// Gather `name` children (rows, cells), looking through `w:sdt`/`w:customXml` wrappers; each
-/// with the content controls that open before it and how many close after it. `pending` holds
-/// controls opened but not yet given to an element.
+/// Gather `name` children (rows, cells), looking through content controls, custom XML and the
+/// selected alternate-content branch; each with the content controls that open before it and how
+/// many close after it. `pending` holds controls opened but not yet given to an element.
 fn collect<'a>(e: &'a El, name: &str, out: &mut Vec<(&'a El, ControlWrap)>, pending: &mut Vec<ContentControl>, depth: usize) {
    if depth > 8 {
        return;
@@ -194,6 +194,10 @@ fn collect<'a>(e: &'a El, name: &str, out: &mut Vec<(&'a El, ControlWrap)>, pend
            }
        } else if k.name == "w:customXml" {
            collect(k, name, out, pending, depth + 1);
+        } else if k.name == "mc:AlternateContent"
+            && let Some(branch) = children_of_choice(k)
+        {
+            collect(branch, name, out, pending, depth + 1);
        }
    }
 }
@@ -51,6 +51,7 @@ impl Reader<'_> {
                    }
                }
                "w:sdt" => self.read_block_sdt(sc, e, rels, out, depth),
+                "w:altChunk" => self.read_alt_chunk(sc, e, rels, out, depth),
                "w:customXml" | "w:smartTag" | "w:ins" | "w:moveTo" => self.read_blocks(sc, e, rels, out, depth + 1),
                // Math is allowed straight in a story (EG_BlockLevelElts takes EG_RunLevelElements,
                // which takes EG_MathContent): it gets a paragraph of its own (#1273).
@@ -135,6 +136,21 @@ impl Reader<'_> {
        sc.pending.clear();
    }

+    /// Markers read before content placed whole (a chunk's `next`) stay before it: at the end of
+    /// the paragraph before, or after a table or with none, at the start of the content's first
+    /// paragraph, as a paragraph read there would take them (with a table first, the next
+    /// paragraph read takes them).
+    pub fn flush_pending_before(&mut self, sc: &mut StoryCtx, out: &mut Blocks, next: &mut [Block]) {
+        if out.last().is_some_and(|b| matches!(**b, Block::Para(_))) {
+            return self.flush_pending(sc, out);
+        }
+        if let Some(Block::Para(p)) = next.first_mut() {
+            for o in sc.pending.drain(..).rev() {
+                let _ = p.insert_object(0, o, &CharProps::default());
+            }
+        }
+    }
+
    pub(super) fn marker(&mut self, e: &El) -> Option<InlineObject> {
        let id = e.attr("w:id").map(str::trim).unwrap_or("").to_string();
        match e.name.as_str() {
@@ -161,8 +177,15 @@ impl Reader<'_> {
                if self.perms.len() >= 100_000 {
                    return None;
                }
-                let n = self.perms.len() as u32;
-                let id = *self.perms.entry(id).or_insert(n);
+                let id = match self.perms.get(&id) {
+                    Some(n) => *n,
+                    None => {
+                        let n = self.next_perm;
+                        self.next_perm = n.saturating_add(1);
+                        self.perms.insert(id, n);
+                        n
+                    }
+                };
                let text = |a: &str| e.attr(a).filter(|v| v.len() <= 256).unwrap_or("").to_string();
                let (group, editor) = (text("w:edGrp"), text("w:ed"));
                if group.is_empty() && editor.is_empty() {
@@ -218,8 +218,8 @@ impl Reader<'_> {
        if let Some(g) = self.graphics.get(&key) {
            return g.clone();
        }
-        let items = if self.graphic_budget == 0 { Vec::new() } else { self.graphic_items(kind, &key.1, w, h) };
-        self.graphic_budget = self.graphic_budget.saturating_sub(graphic_work(&items));
+        let items = if self.chunks.graphic == 0 { Vec::new() } else { self.graphic_items(kind, &key.1, w, h) };
+        self.chunks.graphic = self.chunks.graphic.saturating_sub(graphic_work(&items));
        // A chart or SmartArt that is exactly what WordCraft writes for its model can be edited
        // (and is written from the model); any other stays as it was read.
        let chart = match kind {
@@ -499,10 +499,26 @@ impl Reader<'_> {
                .and_then(|g| preset_adj(g, "adj"))
                .and_then(|v| u32::try_from(v.clamp(0, i64::from(wordcraft_doc::connector::MAX_CORNER))).ok());
        }
-        // Arrowheads: any end type but `none` (ECMA-376 §20.1.10.33) is drawn as a triangle.
-        let arrow = |tag: &str| ln.and_then(|l| l.child(tag)).and_then(|e| e.attr("type")).is_some_and(|t| t != "none");
-        extra.arrow_start = arrow("a:headEnd");
-        extra.arrow_end = arrow("a:tailEnd");
+        // Arrowheads: any end type but `none` (ECMA-376 §20.1.10.33) is drawn as a triangle of its
+        // width and length.
+        let arrow = |tag: &str| {
+            let e = ln.and_then(|l| l.child(tag))?;
+            let kind = e.attr("type").filter(|t| *t != "none")?;
+            let size = |name: &str| wordcraft_doc::connector::ArrowSize::from_ooxml(e.attr(name).unwrap_or("med"));
+            Some(wordcraft_doc::connector::ArrowStyle {
+                kind: wordcraft_doc::connector::ArrowKind::from_ooxml(kind),
+                width: size("w"),
+                length: size("len"),
+            })
+        };
+        if let Some(style) = arrow("a:headEnd") {
+            extra.arrow_start = true;
+            extra.start_style = style;
+        }
+        if let Some(style) = arrow("a:tailEnd") {
+            extra.arrow_end = true;
+            extra.end_style = style;
+        }
        // A connector's glued ends (ECMA-376 §20.1.2.2.36, §20.1.2.2.13).
        if let Some(cn) = wsp.child("wps:cNvCnPr") {
            let end = |tag: &str| {
@@ -540,7 +556,7 @@ impl Reader<'_> {
        let mut blocks = Blocks::new();
        // Shapes in the text aren't members of a group the box may be in.
        let group = self.group_fill.take();
-        self.read_blocks(&mut inner, content, rels, &mut blocks, 0);
+        self.read_blocks(&mut inner, content, rels, &mut blocks, self.chunks.depth);
        self.group_fill = group;
        self.flush_pending(&mut inner, &mut blocks);
        self.doc.add_part(PartKind::TextBox, blocks)
@@ -295,11 +295,41 @@ impl Reader<'_> {
        }
        let kind = kind?;
        let mut extra = ShapeExtra { cap, ..Default::default() };
+        // A horizontal rule (`o:hr`, a pasted `<hr>`): across `o:hrpct` tenths of a percent of its
+        // line, or its own width (`o:hrpct="0"`); a width of 0 with no percentage is the whole
+        // line. `o:hralign` places a shorter one (Word, measured: 50 % centred, left, right).
+        // Only a rule in the text, shown: a positioned or hidden one stays a plain shape, which
+        // keeps its placement and visibility when saved.
+        if attr("o:hr").is_some_and(|v| matches!(v.trim(), "t" | "true" | "1")) && float.wrap == Wrap::Inline && !float.hidden {
+            let tenths = attr("o:hrpct").and_then(|v| v.trim().parse::<f32>().ok()).filter(|p| p.is_finite() && *p > 0.0);
+            let pct = tenths.map_or(if w > 0.0 { 0.0 } else { 100.0 }, |t| (t / 10.0).min(100.0));
+            let align = match attr("o:hralign").map(|a| a.trim().to_ascii_lowercase()).as_deref() {
+                Some("center") => wordcraft_doc::props::Align::Center,
+                Some("right") => wordcraft_doc::props::Align::Right,
+                _ => wordcraft_doc::props::Align::Left,
+            };
+            let standard = attr("o:hrstd").is_some_and(|v| matches!(v.trim(), "t" | "true" | "1"));
+            extra.hr = Some(wordcraft_doc::connector::HorizontalRule { pct, align, standard });
+        }
        if kind == ShapeKind::Line {
-            // Arrowheads (`v:stroke` startarrow / endarrow): any but `none` is drawn.
-            let arrow = |name: &str| shape.child("v:stroke").and_then(|s| s.attr(name)).is_some_and(|v| v != "none");
-            extra.arrow_start = arrow("startarrow");
-            extra.arrow_end = arrow("endarrow");
+            // Arrowheads (`v:stroke` startarrow / endarrow, with their width and length): any but
+            // `none` is drawn.
+            use wordcraft_doc::connector::{ArrowKind, ArrowSize, ArrowStyle};
+            let stroke = shape.child("v:stroke");
+            let arrow = |end: &str| {
+                let s = stroke?;
+                let kind = s.attr(&format!("{end}arrow")).filter(|v| *v != "none")?;
+                let size = |what: &str| ArrowSize::from_vml(s.attr(&format!("{end}arrow{what}")).unwrap_or("medium"));
+                Some(ArrowStyle { kind: ArrowKind::from_vml(kind), width: size("width"), length: size("length") })
+            };
+            if let Some(style) = arrow("start") {
+                extra.arrow_start = true;
+                extra.start_style = style;
+            }
+            if let Some(style) = arrow("end") {
+                extra.arrow_end = true;
+                extra.end_style = style;
+            }
        }
        if shape.local() == "line" {
            // A line runs `from` → `to` (points, offsets from where its style places it). A preset
@@ -33,7 +33,7 @@ const MAX_STORIES: usize = 50_000;

 /// Read a Word 2003 XML Document.
 pub fn read(bytes: &[u8]) -> Result<Document, DocxError> {
-    crate::read::read_opened(&convert(bytes)?)
+    crate::read::read_opened(&convert(bytes)?, None)
 }

 /// Parts the converted package gets relationships for.
@@ -382,11 +382,10 @@ fn line_dash(w: &mut W, dash: &wordcraft_doc::line_dash::LineDash) {
 /// A shape's `a:effectLst` (ECMA-376 Part 1 §20.1.8.26), children in schema order.
 /// A line's arrowheads (`a:headEnd`, `a:tailEnd`, ECMA-376 §20.1.8.38 / §20.1.8.57).
 fn arrow_ends(w: &mut W, extra: &wordcraft_doc::connector::ShapeExtra) {
-    if extra.arrow_start {
-        w.empty("a:headEnd", &[("type", "triangle")]);
-    }
-    if extra.arrow_end {
-        w.empty("a:tailEnd", &[("type", "triangle")]);
+    for (on, tag, style) in [(extra.arrow_start, "a:headEnd", extra.start_style), (extra.arrow_end, "a:tailEnd", extra.end_style)] {
+        if on {
+            w.empty(tag, &[("type", style.kind.ooxml()), ("w", style.width.ooxml()), ("len", style.length.ooxml())]);
+        }
    }
 }

@@ -115,6 +115,9 @@ impl Writer<'_> {
            // ink keep their original source and related parts.
            InlineObject::Graphic { .. } => self.write_graphic(w, o, props, rels, depth),
            InlineObject::Image { .. } => self.write_image(w, o, props, rels, depth),
+            InlineObject::Shape { extra, float, .. } if extra.hr.is_some() && float.wrap == Wrap::Inline && !float.hidden => {
+                self.write_hr(w, o, props)
+            }
            InlineObject::Shape { .. } => self.write_shape(w, o, props, rels, depth),
            InlineObject::Group { .. } => self.write_group(w, o, props, rels, depth),
            InlineObject::Ruby { base, ruby, align, size, raise, base_size, lang, props: ruby_props } => {
@@ -277,6 +280,74 @@ impl Writer<'_> {
        self.rev_close(w, props);
    }

+    /// A horizontal rule goes back as the VML rule it was read from: DrawingML has none, and a
+    /// fixed-size shape would lose its line-wide width.
+    fn write_hr(&mut self, w: &mut W, o: &InlineObject, props: &CharProps) {
+        let InlineObject::Shape { w: sw, h: sh, fill, stroke, stroke_width, extra, float, .. } = o else { return };
+        let Some(hr) = extra.hr else { return };
+        let align = match hr.align {
+            wordcraft_doc::props::Align::Center => "center",
+            wordcraft_doc::props::Align::Right => "right",
+            _ => "left",
+        };
+        let width = if hr.pct > 0.0 { "0".to_string() } else { format!("{:.2}pt", sw.max(0.0)) };
+        let mut style = format!("width:{width};height:{:.2}pt", sh.max(0.0));
+        let spin = float.spin();
+        if spin.deg != 0.0 {
+            style.push_str(&format!(";rotation:{}", spin.deg));
+        }
+        match (spin.flip_h, spin.flip_v) {
+            (true, true) => style.push_str(";flip:x y"),
+            (true, false) => style.push_str(";flip:x"),
+            (false, true) => style.push_str(";flip:y"),
+            (false, false) => {}
+        }
+        let pct = format!("{}", (hr.pct * 10.0).round().max(0.0) as u32);
+        let fill_hex = fill.as_ref().map(|f| format!("#{}", f.main_color().hex()));
+        let stroke_hex = stroke.map(|c| format!("#{}", c.hex()));
+        let weight = format!("{stroke_width:.2}pt");
+        let mut attrs: Vec<(&str, &str)> = vec![
+            ("xmlns:v", "urn:schemas-microsoft-com:vml"),
+            ("xmlns:o", "urn:schemas-microsoft-com:office:office"),
+            ("style", &style),
+            ("o:hrpct", &pct),
+            ("o:hralign", align),
+            ("o:hr", "t"),
+        ];
+        if hr.standard {
+            attrs.push(("o:hrstd", "t"));
+        }
+        if let Some(f) = &fill_hex {
+            attrs.push(("fillcolor", f));
+        } else {
+            attrs.push(("filled", "f"));
+        }
+        match &stroke_hex {
+            Some(s) => {
+                attrs.push(("strokecolor", s));
+                attrs.push(("strokeweight", &weight));
+            }
+            None => attrs.push(("stroked", "f")),
+        }
+        self.rev_open(w, props);
+        w.open("w:r", &[]);
+        self.rpr(w, props);
+        w.open("w:pict", &[]);
+        // A see-through fill keeps its opacity, which `fillcolor` can't say.
+        match fill.as_ref() {
+            Some(wordcraft_doc::fill::Fill::Solid { transparency, .. }) if *transparency > 0.0 => {
+                let opacity = format!("{:.4}", (1.0 - transparency / 100.0).clamp(0.0, 1.0));
+                w.open("v:rect", &attrs);
+                w.empty("v:fill", &[("opacity", &opacity)]);
+                w.close("v:rect");
+            }
+            _ => w.empty("v:rect", &attrs),
+        }
+        w.close("w:pict");
+        w.close("w:r");
+        self.rev_close(w, props);
+    }
+
    fn write_shape(&mut self, w: &mut W, o: &InlineObject, props: &CharProps, rels: &mut PartRels, depth: usize) {
        let InlineObject::Shape { kind, w: sw, h: sh, float, effects, freeform, extra, .. } = o else { return };
        self.rev_open(w, props);
@@ -0,0 +1,365 @@
+//! Alternative format chunks (`w:altChunk`, ECMA-376 §17.17.2.1): content in another format that
+//! is brought in where the chunk stands. All markup here is synthetic.
+
+use std::io::Write;
+
+use wordcraft_doc::{Block, Document, InlineObject, StoryRef};
+use wordcraft_docx::{AltChunk, ReadOptions, read, read_with};
+
+const NS: &str = r#"xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships""#;
+const ROOT_RELS: &str = r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"><Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="word/document.xml"/></Relationships>"#;
+const DOCX_MAIN: &str = "application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml";
+
+fn p(text: &str) -> String {
+    format!(r#"<w:p><w:r><w:t xml:space="preserve">{text}</w:t></w:r></w:p>"#)
+}
+
+/// A package whose body is `body`, with the chunk `afchunk.{ext}` (content type `ct`) as `rIdC`.
+fn package(body: &str, chunk: Option<(&str, &str, &[u8])>) -> Vec<u8> {
+    package_styled(body, chunk, None)
+}
+
+/// [`package`] with a styles part holding `styles`.
+fn package_styled(body: &str, chunk: Option<(&str, &str, &[u8])>, styles: Option<&str>) -> Vec<u8> {
+    let doc = format!(r#"<w:document {NS}><w:body>{body}<w:sectPr/></w:body></w:document>"#);
+    let mut files: Vec<(String, Vec<u8>)> = vec![("_rels/.rels".into(), ROOT_RELS.into()), ("word/document.xml".into(), doc.into_bytes())];
+    let mut types = String::from(
+        r#"<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types"><Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/><Default Extension="xml" ContentType="application/xml"/>"#,
+    );
+    let mut rels = String::new();
+    if let Some((ext, ct, bytes)) = chunk {
+        types.push_str(&format!(r#"<Default Extension="{ext}" ContentType="{ct}"/>"#));
+        rels.push_str(&format!(r#"<Relationship Id="rIdC" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/aFChunk" Target="/word/afchunk.{ext}"/>"#));
+        files.push((format!("word/afchunk.{ext}"), bytes.to_vec()));
+    }
+    if let Some(styles) = styles {
+        rels.push_str(
+            r#"<Relationship Id="rIdS" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/styles" Target="styles.xml"/>"#,
+        );
+        files.push(("word/styles.xml".into(), format!(r#"<w:styles {NS}>{styles}</w:styles>"#).into_bytes()));
+    }
+    if !rels.is_empty() {
+        files.push((
+            "word/_rels/document.xml.rels".into(),
+            format!(r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">{rels}</Relationships>"#).into_bytes(),
+        ));
+    }
+    types.push_str(r#"<Override PartName="/word/document.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml"/></Types>"#);
+    files.push(("[Content_Types].xml".into(), types.into_bytes()));
+    let mut zw = zip::ZipWriter::new(std::io::Cursor::new(Vec::new()));
+    for (name, bytes) in files {
+        zw.start_file(name, zip::write::SimpleFileOptions::default()).unwrap();
+        zw.write_all(&bytes).unwrap();
+    }
+    zw.finish().unwrap().into_inner()
+}
+
+const CHUNK: &str = r#"<w:altChunk r:id="rIdC"/>"#;
+
+fn body_lines(d: &Document) -> Vec<String> {
+    d.plain_text(StoryRef::Body).split('\n').map(str::to_string).collect()
+}
+
+#[test]
+fn a_chunk_comes_in_through_the_importer_where_it_stands() {
+    let html = b"<p>Chunk one</p><p>Chunk two</p>";
+    let bytes = package(&format!("{}{CHUNK}{}", p("Before"), p("After")), Some(("htm", "application/xhtml+xml", html)));
+    let seen = std::cell::RefCell::new(Vec::new());
+    let import = |c: &AltChunk| {
+        seen.borrow_mut().push((c.content_type.to_string(), c.part.to_string(), c.bytes.to_vec()));
+        Some(Document::from_text("Chunk one\nChunk two"))
+    };
+    let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+    assert_eq!(body_lines(&d), ["Before", "Chunk one", "Chunk two", "After"]);
+    assert_eq!(*seen.borrow(), [("application/xhtml+xml".to_string(), "word/afchunk.htm".to_string(), html.to_vec())]);
+    // Without an importer the chunk is left out.
+    assert_eq!(body_lines(&read(&bytes).unwrap()), ["Before", "After"]);
+}
+
+#[test]
+fn a_bookmark_opened_before_a_chunk_at_the_start_of_the_body_is_kept() {
+    let body = format!(r#"<w:bookmarkStart w:id="0" w:name="Here"/>{CHUNK}<w:bookmarkEnd w:id="0"/>{}"#, p("After"));
+    let bytes = package(&body, Some(("htm", "application/xhtml+xml", b"<p>Chunk</p>")));
+    let import = |_: &AltChunk| Some(Document::from_text("Chunk"));
+    let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+    assert_eq!(body_lines(&d), ["Chunk", "After"]);
+    let marks = d.bookmarks();
+    assert_eq!(marks.len(), 1);
+    assert_eq!((marks[0].0.as_str(), marks[0].1.path.last(), marks[0].1.off), ("Here", 0, 0));
+}
+
+#[test]
+fn a_bookmark_opened_between_a_table_and_a_chunk_starts_the_chunk() {
+    let table = format!("<w:tbl><w:tr><w:tc>{}</w:tc></w:tr></w:tbl>", p("Cell"));
+    let body = format!(r#"{table}<w:bookmarkStart w:id="0" w:name="Here"/>{CHUNK}<w:bookmarkEnd w:id="0"/>{}"#, p("After"));
+    let bytes = package(&body, Some(("htm", "application/xhtml+xml", b"<p>Chunk</p>")));
+    let import = |_: &AltChunk| Some(Document::from_text("Chunk"));
+    let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+    let marks = d.bookmarks();
+    assert_eq!(marks.len(), 1);
+    assert_eq!((marks[0].1.path.clone(), marks[0].1.off), (wordcraft_doc::Path::top(1), 0), "not in the table's cell");
+}
+
+#[test]
+fn permission_ranges_of_chunks_and_their_host_stay_apart() {
+    let inner = package(r#"<w:p><w:permStart w:id="0" w:ed="them"/><w:r><w:t>Inner</w:t></w:r><w:permEnd w:id="0"/></w:p>"#, None);
+    // The host's range 0 is open across two chunks whose own ranges are 0 too.
+    let body = format!(
+        r#"<w:p><w:permStart w:id="0" w:edGrp="everyone"/><w:r><w:t>Host</w:t></w:r></w:p>{CHUNK}{CHUNK}<w:p><w:r><w:t>After</w:t></w:r><w:permEnd w:id="0"/></w:p>"#
+    );
+    let d = read(&package(&body, Some(("docx", DOCX_MAIN, &inner)))).unwrap();
+    assert_eq!(body_lines(&d), ["Host", "Inner", "Inner", "After"]);
+    let check = |d: &Document| {
+        let ranges = d.perm_ranges();
+        let mut ids: Vec<u32> = ranges.iter().map(|r| r.id).collect();
+        ids.sort_unstable();
+        ids.dedup();
+        assert_eq!(ids.len(), 3, "{ranges:?}");
+        let host = ranges.iter().find(|r| r.group == "everyone").unwrap();
+        assert_eq!((host.start.path.last(), host.end.path.last()), (0, 3));
+        assert!(ranges.iter().filter(|r| r.editor == "them").all(|r| r.start.path == r.end.path));
+    };
+    check(&d);
+    check(&read(&wordcraft_docx::write(&d).unwrap()).unwrap());
+}
+
+/// A package whose body is `body`, with story parts (relationship type, part name, XML) and the
+/// Word document `chunk` as `rIdC` from the main part and from each story part.
+fn package_with_stories(body: &str, stories: &[(&str, &str, String)], chunk: Option<&[u8]>) -> Vec<u8> {
+    let rel = |id: &str, ty: &str, target: &str| {
+        format!(r#"<Relationship Id="{id}" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/{ty}" Target="{target}"/>"#)
+    };
+    let rels =
+        |inner: String| format!(r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">{inner}</Relationships>"#);
+    let chunk_rel = if chunk.is_some() { rel("rIdC", "aFChunk", "/word/afchunk.docx") } else { String::new() };
+    let doc = format!(r#"<w:document {NS}><w:body>{body}<w:sectPr/></w:body></w:document>"#);
+    let mut main_rels = chunk_rel.clone();
+    let mut files: Vec<(String, Vec<u8>)> = vec![("_rels/.rels".into(), ROOT_RELS.into()), ("word/document.xml".into(), doc.into_bytes())];
+    for (i, (ty, name, xml)) in stories.iter().enumerate() {
+        main_rels.push_str(&rel(&format!("rIdS{i}"), ty, name));
+        files.push((format!("word/{name}"), xml.clone().into_bytes()));
+        files.push((format!("word/_rels/{name}.rels"), rels(chunk_rel.clone()).into_bytes()));
+    }
+    files.push(("word/_rels/document.xml.rels".into(), rels(main_rels).into_bytes()));
+    if let Some(c) = chunk {
+        files.push(("word/afchunk.docx".into(), c.to_vec()));
+    }
+    let mut zw = zip::ZipWriter::new(std::io::Cursor::new(Vec::new()));
+    for (name, bytes) in files {
+        zw.start_file(name, zip::write::SimpleFileOptions::default()).unwrap();
+        zw.write_all(&bytes).unwrap();
+    }
+    zw.finish().unwrap().into_inner()
+}
+
+/// A paragraph `text` that comment `id` (in the file) is anchored around.
+fn commented(id: u32, text: &str) -> String {
+    format!(
+        r#"<w:p><w:commentRangeStart w:id="{id}"/><w:r><w:t>{text}</w:t></w:r><w:commentRangeEnd w:id="{id}"/><w:r><w:commentReference w:id="{id}"/></w:r></w:p>"#
+    )
+}
+
+/// A comments part: (file id, author, content).
+fn comments_part(list: &[(u32, &str, String)]) -> (&'static str, &'static str, String) {
+    let body: String = list.iter().map(|(id, who, content)| format!(r#"<w:comment w:id="{id}" w:author="{who}">{content}</w:comment>"#)).collect();
+    ("comments", "comments.xml", format!("<w:comments {NS}>{body}</w:comments>"))
+}
+
+/// The authors of the comments anchored in the stories whose text contains `text`, and whether
+/// every comment has text of its own and a distinct id.
+fn anchored(d: &Document, text: &str) -> Vec<String> {
+    let mut stories: Vec<&[std::sync::Arc<Block>]> = vec![&d.body];
+    stories.extend(d.parts.values().map(|p| p.blocks.as_slice()));
+    let mut out = Vec::new();
+    for blocks in stories {
+        for b in blocks {
+            let Block::Para(p) = &**b else { continue };
+            if !p.text.contains(text) {
+                continue;
+            }
+            for o in &p.objects {
+                if let InlineObject::CommentStart { id } = o {
+                    out.push(d.comments.get(id).map_or("?".to_string(), |c| c.author.clone()));
+                }
+            }
+        }
+    }
+    out
+}
+
+/// A Word document with one comment by "Them" around "Inner".
+fn commented_chunk() -> Vec<u8> {
+    package_with_stories(&commented(0, "Inner"), &[comments_part(&[(0, "Them", p("Theirs"))])], None)
+}
+
+#[test]
+fn a_chunk_in_a_footnote_keeps_its_comments_beside_the_documents() {
+    let notes = format!(r#"<w:footnotes {NS}><w:footnote w:id="1"><w:altChunk r:id="rIdC"/></w:footnote></w:footnotes>"#);
+    let body = format!(r#"{}<w:p><w:r><w:footnoteReference w:id="1"/></w:r></w:p>"#, commented(0, "Host"));
+    let stories = [("footnotes", "footnotes.xml", notes), comments_part(&[(0, "Us", p("Ours"))])];
+    let d = read(&package_with_stories(&body, &stories, Some(&commented_chunk()))).unwrap();
+    assert_eq!(d.comments.len(), 2, "{:?}", d.comments);
+    assert_eq!(anchored(&d, "Host"), ["Us"]);
+    assert_eq!(anchored(&d, "Inner"), ["Them"]);
+}
+
+#[test]
+fn a_chunk_in_a_comment_keeps_its_comments_beside_the_documents() {
+    let inside = format!(r#"{}<w:altChunk r:id="rIdC"/>"#, p("Ours"));
+    let body = format!("{}{}", commented(0, "Host"), commented(1, "Second"));
+    let stories = [comments_part(&[(0, "Us", inside), (1, "Us too", p("Ours too"))])];
+    let d = read(&package_with_stories(&body, &stories, Some(&commented_chunk()))).unwrap();
+    assert_eq!(d.comments.len(), 3, "{:?}", d.comments);
+    assert_eq!(anchored(&d, "Host"), ["Us"]);
+    assert_eq!(anchored(&d, "Second"), ["Us too"]);
+    assert_eq!(anchored(&d, "Inner"), ["Them"]);
+}
+
+#[test]
+fn a_wordprocessing_chunk_is_read_without_an_importer_and_flows_in_its_section() {
+    let inner = package(&format!("{}{}", p("Inner one"), p("Inner two")), None);
+    let bytes = package(&format!("{}{CHUNK}{}", p("Before"), p("After")), Some(("docx", DOCX_MAIN, &inner)));
+    let d = read(&bytes).unwrap();
+    assert_eq!(body_lines(&d), ["Before", "Inner one", "Inner two", "After"]);
+    // The inner document's own section doesn't come along.
+    assert_eq!(d.sections().len(), 1);
+}
+
+/// Styles a matchSrc chunk brings under new ids get distinct ones: with `A` and `A1` ours and
+/// `A` to `A10` the chunk's, its `A` and `A1` both renamed don't land on one id (the second then
+/// lost, its paragraphs taking the first's formatting), and keep their formatting through a save.
+#[test]
+fn match_src_gives_each_renamed_style_its_own_id() {
+    let style = |id: &str, half_points: u32| {
+        format!(r#"<w:style w:type="paragraph" w:styleId="{id}"><w:name w:val="{id}"/><w:rPr><w:sz w:val="{half_points}"/></w:rPr></w:style>"#)
+    };
+    let ours: String = ["A", "A1"].iter().map(|id| style(id, 22)).collect();
+    let mut theirs = style("A", 20) + &style("A1", 28);
+    for n in 2..=10 {
+        theirs += &style(&format!("A{n}"), 16);
+    }
+    let para = |id: &str, text: &str| format!(r#"<w:p><w:pPr><w:pStyle w:val="{id}"/></w:pPr><w:r><w:t>{text}</w:t></w:r></w:p>"#);
+    let inner = package_styled(&(para("A", "ten") + &para("A1", "fourteen")), None, Some(&theirs));
+    let body = format!(r#"{}<w:altChunk r:id="rIdC"><w:altChunkPr><w:matchSrc/></w:altChunkPr></w:altChunk>"#, p("Ours"));
+    let d = read(&package_styled(&body, Some(("docx", DOCX_MAIN, &inner)), Some(&ours))).unwrap();
+    let check = |d: &Document| {
+        let look = |text: &str| {
+            d.body
+                .iter()
+                .find_map(|b| match &**b {
+                    Block::Para(p) if p.text == text => {
+                        Some((p.props.style.clone(), d.styles.resolve_char(p.props.style.as_deref(), p.props_of_char(0)).size))
+                    }
+                    _ => None,
+                })
+                .unwrap()
+        };
+        let ((ten, ten_size), (fourteen, fourteen_size)) = (look("ten"), look("fourteen"));
+        assert_ne!(ten, fourteen);
+        assert_eq!((ten_size, fourteen_size), (10.0, 14.0));
+    };
+    check(&d);
+    check(&read(&wordcraft_docx::write(&d).unwrap()).unwrap());
+}
+
+#[test]
+fn match_src_keeps_a_word_chunks_own_formatting() {
+    // Both define Normal and Subtitle differently, and the chunk's defaults differ from ours.
+    let ours = r#"<w:docDefaults><w:rPrDefault><w:rPr><w:sz w:val="22"/></w:rPr></w:rPrDefault><w:pPrDefault><w:pPr><w:spacing w:after="200"/></w:pPr></w:pPrDefault></w:docDefaults><w:style w:type="paragraph" w:default="1" w:styleId="Normal"><w:name w:val="Normal"/><w:pPr><w:spacing w:after="248" w:line="248" w:lineRule="exact"/></w:pPr></w:style><w:style w:type="paragraph" w:styleId="Subtitle"><w:name w:val="Subtitle"/><w:rPr><w:i/></w:rPr></w:style>"#;
+    let theirs = r#"<w:docDefaults><w:rPrDefault><w:rPr><w:sz w:val="24"/></w:rPr></w:rPrDefault></w:docDefaults><w:style w:type="paragraph" w:default="1" w:styleId="Normal"><w:name w:val="Normal"/></w:style><w:style w:type="paragraph" w:styleId="Subtitle"><w:name w:val="Subtitle"/><w:basedOn w:val="Normal"/></w:style>"#;
+    let inner = package_styled(
+        &format!(r#"{}<w:p><w:pPr><w:pStyle w:val="Subtitle"/></w:pPr><w:r><w:t>Sub</w:t></w:r></w:p>"#, p("Plain")),
+        None,
+        Some(theirs),
+    );
+    for match_src in [true, false] {
+        let pr = if match_src { "<w:altChunkPr><w:matchSrc/></w:altChunkPr>" } else { "" };
+        let body = format!(r#"{}<w:altChunk r:id="rIdC">{pr}</w:altChunk>"#, p("Ours"));
+        let d = read(&package_styled(&body, Some(("docx", DOCX_MAIN, &inner)), Some(ours))).unwrap();
+        let look = |i: usize| match &*d.body[i] {
+            Block::Para(p) => {
+                let para = d.styles.resolve_para(&p.props);
+                let chr = d.styles.resolve_char(p.props.style.as_deref(), p.props_of_char(0));
+                (para.space_after, chr.size, chr.italic)
+            }
+            b => panic!("{b:?}"),
+        };
+        // Ours stays ours either way.
+        assert_eq!(look(0), (12.4, 11.0, false));
+        if match_src {
+            // The chunk's own Normal (its 12 pt, Word's no space after) and its Subtitle (not italic).
+            assert_eq!(look(1), (0.0, 12.0, false));
+            assert_eq!(look(2), (0.0, 12.0, false));
+        } else {
+            // Without it the chunk takes our styles (our Subtitle stands on our defaults).
+            assert_eq!(look(1), (12.4, 11.0, false));
+            assert_eq!(look(2), (10.0, 11.0, true));
+        }
+    }
+}
+
+#[test]
+fn nested_wordprocessing_chunks_stop_at_a_depth() {
+    let mut bytes = package(&p("Deepest"), None);
+    for level in 0..8 {
+        bytes = package(&format!("{}{CHUNK}", p(&format!("Level {level}"))), Some(("docx", DOCX_MAIN, &bytes)));
+    }
+    let lines = body_lines(&read(&bytes).unwrap());
+    assert_eq!(lines.first().map(String::as_str), Some("Level 7"));
+    assert!(!lines.iter().any(|l| l == "Deepest"), "{lines:?}");
+    assert!(lines.len() >= 4, "{lines:?}");
+}
+
+#[test]
+fn a_chunks_pictures_come_with_it() {
+    let bytes = package(&format!("{}{CHUNK}", p("Before")), Some(("htm", "text/html", b"<p>x</p>")));
+    let import = |_: &AltChunk| {
+        let mut d = Document::new();
+        let key = d.add_media(vec![1, 2, 3], "png");
+        let mut para = wordcraft_doc::Paragraph::new();
+        let image = InlineObject::Image { media: key, w: 10.0, h: 10.0, alt: String::new(), float: Default::default(), crop: [0.0; 4], ole: None };
+        para.insert_object(0, image, &Default::default()).unwrap();
+        d.body = vec![std::sync::Arc::new(Block::Para(para))];
+        Some(d)
+    };
+    let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+    let keys: Vec<String> = d
+        .body
+        .iter()
+        .filter_map(|b| match &**b {
+            Block::Para(p) => p.objects.iter().find_map(|o| match o {
+                InlineObject::Image { media, .. } => Some(media.clone()),
+                _ => None,
+            }),
+            _ => None,
+        })
+        .collect();
+    assert_eq!(keys.len(), 1);
+    assert_eq!(d.media.get(&keys[0]).map(|m| m.as_slice()), Some(&[1u8, 2, 3][..]));
+}
+
+#[test]
+fn a_file_brings_in_a_bounded_number_of_chunks() {
+    let bytes = package(&CHUNK.repeat(300), Some(("htm", "text/html", b"<p>x</p>")));
+    let calls = std::cell::Cell::new(0);
+    let import = |_: &AltChunk| {
+        calls.set(calls.get() + 1);
+        Some(Document::from_text("x"))
+    };
+    let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+    assert_eq!(calls.get(), 256);
+    assert_eq!(body_lines(&d).len(), 256);
+}
+
+#[test]
+fn hostile_chunks_never_panic() {
+    for (ext, ct, bytes) in [("docx", DOCX_MAIN, &b"PK\x03\x04 not a zip"[..]), ("htm", "text/html", b""), ("bin", "", b"\x00\xff")] {
+        let bytes = package(&format!("{CHUNK}{CHUNK}{}", p("x")), Some((ext, ct, bytes)));
+        let import = |_: &AltChunk| None;
+        let d = read_with(&bytes, &ReadOptions { chunks: Some(&import), ..Default::default() }).unwrap();
+        assert_eq!(body_lines(&d), ["x"]);
+    }
+    // A chunk naming a missing part or another relationship type is left out.
+    let bytes = package(&format!(r#"<w:altChunk r:id="nope"/><w:altChunk/>{}"#, p("y")), None);
+    assert_eq!(body_lines(&read(&bytes).unwrap()), ["y"]);
+}
@@ -48,6 +48,76 @@ fn read_body(body: &str) -> Document {
    wordcraft_docx::read(&docx(body, &[], &[])).unwrap()
 }

+/// Markup compatibility replaces an AlternateContent wrapper with one selected branch,
+/// including where a table expects rows or cells (ECMA-376 Part 3).
+#[test]
+fn table_alternate_content_keeps_selected_rows_and_cells() {
+    let cell = |text: &str| format!("<w:tc><w:p><w:r><w:t>{text}</w:t></w:r></w:p></w:tc>");
+    let row = |text: &str| format!("<w:tr>{}</w:tr>", cell(text));
+    let table =
+        |rows: &str, columns: usize| format!(r#"<w:tbl><w:tblGrid>{}</w:tblGrid>{rows}</w:tbl>"#, r#"<w:gridCol w:w="2400"/>"#.repeat(columns));
+    let alternate = |chosen: &str, fallback: &str, requires: &str| {
+        format!(
+            r#"<mc:AlternateContent><mc:Choice Requires="{requires}">{chosen}</mc:Choice><mc:Fallback>{fallback}</mc:Fallback></mc:AlternateContent>"#
+        )
+    };
+    for (requires, use_choice) in [("wps", true), ("alias", true), ("unknown", false), ("wps unknown", false)] {
+        for cells in [false, true] {
+            let choice = if cells { cell("Chosen cell") } else { row("Chosen row") };
+            let fallback = if cells { cell("Fallback cell") } else { row("Fallback row") };
+            let selected = if use_choice { &choice } else { &fallback };
+            let middle = alternate(&choice, &fallback, requires);
+            let (actual, expected) = if cells {
+                (
+                    format!("<w:tr>{}{middle}{}</w:tr>", cell("Before"), cell("After")),
+                    format!("<w:tr>{}{selected}{}</w:tr>", cell("Before"), cell("After")),
+                )
+            } else {
+                (format!("{}{middle}{}", row("Before"), row("After")), format!("{}{selected}{}", row("Before"), row("After")))
+            };
+            let columns = if cells { 3 } else { 1 };
+            let expected = read_body(&table(&expected, columns));
+            let wrapped = format!(
+                r#"<w:customXml xmlns:alias="http://schemas.microsoft.com/office/word/2010/wordprocessingShape" xmlns:unknown="urn:original:unsupported">{}</w:customXml>"#,
+                table(&actual, columns)
+            );
+            let mut actual = read_body(&wrapped);
+            for round in 0..=2 {
+                assert_eq!(actual.body, expected.body, "requires={requires}, cells={cells}, round={round}");
+                if round < 2 {
+                    actual = wordcraft_docx::read(&wordcraft_docx::write(&actual).unwrap()).unwrap();
+                }
+            }
+        }
+    }
+}
+
+#[test]
+fn table_alternate_content_keeps_nested_row_and_cell_controls() {
+    let wrap =
+        |tag: &str, content: &str| format!(r#"<w:sdt><w:sdtPr><w:tag w:val="{tag}"/></w:sdtPr><w:sdtContent>{content}</w:sdtContent></w:sdt>"#);
+    let alternate =
+        |content: &str| format!(r#"<mc:AlternateContent><mc:Choice Requires="wps">{content}</mc:Choice><mc:Fallback/></mc:AlternateContent>"#);
+    let cell = "<w:tc><w:p><w:r><w:t>Controlled cell</w:t></w:r></w:p></w:tc>";
+    let oracle_cell = wrap("cell", cell);
+    let actual_cell = alternate(&wrap("cell", &alternate(cell)));
+    let oracle_row = wrap("row", &format!("<w:tr>{oracle_cell}</w:tr>"));
+    let actual_row = wrap("row", &alternate(&format!("<w:tr>{actual_cell}</w:tr>")));
+    let expected = read_body(&format!("<w:tbl>{oracle_row}</w:tbl>"));
+    let mut actual = read_body(&format!("<w:tbl>{}</w:tbl>", alternate(&actual_row)));
+    for round in 0..=2 {
+        assert_eq!(actual.body, expected.body, "round={round}");
+        let table = actual.body.iter().find_map(|b| b.as_table()).unwrap();
+        assert_eq!(table.rows[0].controls.open[0].tag, "row");
+        assert_eq!(table.rows[0].controls.close, 1);
+        assert_eq!(table.rows[0].cells[0].controls.open[0].tag, "cell");
+        assert_eq!(table.rows[0].cells[0].controls.close, 1);
+        if round < 2 {
+            actual = wordcraft_docx::read(&wordcraft_docx::write(&actual).unwrap()).unwrap();
+        }
+    }
+}
+
 #[test]
 fn repeated_style_ids_use_last() {
    let styles = format!(
@@ -2475,6 +2545,52 @@ fn vml_lines_and_rectangles() {
    ));
 }

+/// An arrowhead's kind, width and length (DrawingML `a:headEnd`/`a:tailEnd` `type`/`w`/`len`, VML
+/// `startarrow`/`startarrowwidth`/`startarrowlength`) are read and saved: they set the size Word
+/// draws it at. An end without `w`/`len` is medium.
+#[test]
+fn arrowheads_keep_their_kind_and_size() {
+    use wordcraft_doc::connector::{ArrowKind, ArrowSize, ArrowStyle};
+    let style = |kind, width, length| ArrowStyle { kind, width, length };
+    let body = [
+        anchored_wsp(
+            "Line",
+            1270000,
+            0,
+            r#"<a:prstGeom prst="line"><a:avLst/></a:prstGeom><a:ln w="12700"><a:solidFill><a:srgbClr val="000000"/></a:solidFill><a:headEnd type="stealth" w="sm" len="lg"/><a:tailEnd type="diamond" w="lg" len="sm"/></a:ln>"#,
+            "",
+            "<wps:cNvCnPr/>",
+        ),
+        anchored_wsp("Plain", 1270000, 0, r#"<a:prstGeom prst="line"><a:avLst/></a:prstGeom><a:ln><a:solidFill><a:srgbClr val="000000"/></a:solidFill><a:tailEnd type="triangle"/></a:ln>"#, "", "<wps:cNvCnPr/>"),
+        r#"<w:r><w:pict><v:line style="position:absolute" from="10pt,20pt" to="40pt,20pt"><v:stroke startarrow="classic" startarrowwidth="narrow" startarrowlength="long" endarrow="oval" endarrowwidth="wide" endarrowlength="short"/></v:line></w:pict></w:r>"#.to_string(),
+    ]
+    .concat();
+    let ends = |d: &Document| -> Vec<(bool, ArrowStyle, bool, ArrowStyle)> {
+        paras(d)[0]
+            .objects
+            .iter()
+            .map(|o| match o {
+                InlineObject::Shape { extra, .. } => (extra.arrow_start, extra.start_style, extra.arrow_end, extra.end_style),
+                o => panic!("{o:?}"),
+            })
+            .collect()
+    };
+    let expected = vec![
+        (true, style(ArrowKind::Stealth, ArrowSize::Small, ArrowSize::Large), true, style(ArrowKind::Diamond, ArrowSize::Large, ArrowSize::Small)),
+        (false, ArrowStyle::default(), true, ArrowStyle::default()),
+        (true, style(ArrowKind::Stealth, ArrowSize::Small, ArrowSize::Large), true, style(ArrowKind::Oval, ArrowSize::Large, ArrowSize::Small)),
+    ];
+    let d = read_body(&format!("<w:p>{body}</w:p>"));
+    assert_eq!(ends(&d), expected);
+    let bytes = wordcraft_docx::write(&d).unwrap();
+    let mut xml = String::new();
+    let mut zip = zip::ZipArchive::new(std::io::Cursor::new(bytes.as_slice())).unwrap();
+    std::io::Read::read_to_string(&mut zip.by_name("word/document.xml").unwrap(), &mut xml).unwrap();
+    assert!(xml.contains(r#"<a:headEnd type="stealth" w="sm" len="lg"/>"#), "{xml}");
+    assert!(xml.contains(r#"<a:tailEnd type="diamond" w="lg" len="sm"/>"#), "{xml}");
+    assert_eq!(ends(&wordcraft_docx::read(&bytes).unwrap()), expected);
+}
+
 /// The right arrow's shapetype as Word writes it (ECMA-376 Part 4 `v:shapetype`, `o:spt` 13).
 const VML_RIGHT_ARROW_TYPE: &str = r#"<v:shapetype id="_x0000_t13" coordsize="21600,21600" o:spt="13" adj="16200,5400" path="m@0,l@0@1,0@1,0@2@0@2@0,21600,21600,10800xe"><v:stroke joinstyle="miter"/><v:formulas><v:f eqn="val #0"/><v:f eqn="val #1"/></v:formulas><v:path o:connecttype="custom"/></v:shapetype>"#;

@@ -3976,6 +4092,30 @@ fn equation_in_alternate_content_is_read() {
    assert!(!p.plain_text().contains("[picture]"));
 }

+/// A fill's `a:alpha` is its opacity: a translucent white wash over a picture stayed opaque and
+/// hid the picture.
+#[test]
+fn a_shape_fill_keeps_its_transparency() {
+    let fill =
+        r#"<a:prstGeom prst="rect"><a:avLst/></a:prstGeom><a:solidFill><a:srgbClr val="FFFFFF"><a:alpha val="30980"/></a:srgbClr></a:solidFill>"#;
+    let d = read_body(&format!("<w:p>{}</w:p>", anchored_wsp("Wash", 1270000, 635000, fill, "", "<wps:cNvSpPr/>")));
+    let transparency = |d: &Document| match &paras(d)[0].objects[0] {
+        InlineObject::Shape { fill: Some(wordcraft_doc::fill::Fill::Solid { color, transparency }), .. } => (Some(*color), *transparency),
+        o => panic!("{o:?}"),
+    };
+    const FILL_ALPHA: u32 = 30_980;
+    const ALPHA_SCALE: f32 = 100_000.0;
+    // Canonical Fill stores the percentage computed from opacity, including f32 rounding.
+    let expected = 100.0 - (FILL_ALPHA as f32 / ALPHA_SCALE) * 100.0;
+    assert_eq!(transparency(&d), (Some(Rgb(255, 255, 255)), expected));
+    let bytes = wordcraft_docx::write(&d).unwrap();
+    let parts = unzip(&bytes);
+    let xml = String::from_utf8_lossy(parts.get("word/document.xml").unwrap());
+    assert!(xml.contains(&format!("<a:alpha val=\"{FILL_ALPHA}\"/>")), "{xml}");
+    let again = wordcraft_docx::read(&bytes).unwrap();
+    assert_eq!(transparency(&again), (Some(Rgb(255, 255, 255)), expected));
+}
+
 /// #1500: gradient fills and transparency on shapes (the cases in the issue), from the shape's
 /// own properties, the theme's fill styles (`a:fillRef` 3 and 1003, as Word's cover pages use),
 /// a group's fill (`a:grpFill`), an outline gradient and VML's `v:fill`. Each is read as the
@@ -4172,3 +4312,62 @@ fn vml_textbox_style_anchor_and_fit() {
    assert_eq!(tb.body.anchor, wordcraft_doc::props::VAlign::Center);
    assert!(tb.body.auto_fit);
 }
+
+/// A VML horizontal rule (`o:hr`, a web page's `<hr>` pasted into Word) is read as one: the whole
+/// line when it has no width and no `o:hrpct`, a share of it (`o:hrpct`, tenths of a percent), or
+/// its own width (`o:hrpct="0"`), placed by `o:hralign`. It saves back as the same VML rule.
+#[test]
+fn vml_horizontal_rules_are_read_and_kept() {
+    use wordcraft_doc::connector::HorizontalRule;
+    use wordcraft_doc::props::Align;
+    let hr = |attrs: &str, style: &str| format!(r#"<w:p><w:r><w:pict><v:rect style="{style}" {attrs}/></w:pict></w:r></w:p>"#);
+    let body = [
+        hr(r#"o:hralign="center" o:hrstd="t" o:hr="t""#, "width:0;height:1.5pt"),
+        hr(r#"o:hrpct="500" o:hralign="right" o:hr="t""#, "width:0;height:1.5pt"),
+        hr(r#"o:hrpct="0" o:hralign="center" o:hr="t""#, "width:100pt;height:3pt"),
+        hr("", "width:100pt;height:3pt"),
+    ]
+    .concat();
+    let rules = |d: &Document| -> Vec<(Option<HorizontalRule>, f32)> {
+        paras(d)
+            .iter()
+            .filter_map(|p| p.objects.iter().find_map(|o| if let InlineObject::Shape { extra, w, .. } = o { Some((extra.hr, *w)) } else { None }))
+            .collect()
+    };
+    let d = read_body(&body);
+    let got = rules(&d);
+    assert_eq!(
+        got,
+        [
+            (Some(HorizontalRule { pct: 100.0, align: Align::Center, standard: true }), 0.0),
+            (Some(HorizontalRule { pct: 50.0, align: Align::Right, standard: false }), 0.0),
+            (Some(HorizontalRule { pct: 0.0, align: Align::Center, standard: false }), 100.0),
+            (None, 100.0),
+        ]
+    );
+    let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap();
+    assert_eq!(rules(&again)[..3], got[..3]);
+    // A rule's rotation and flips are kept.
+    let spin = |d: &Document| paras(d)[0].objects.iter().find_map(|o| o.frame()).map(|(_, _, f)| f.spin());
+    let d = read_body(&hr(r#"o:hrpct="500" o:hr="t""#, "width:0;height:1.5pt;rotation:45;flip:x y"));
+    assert_eq!(spin(&d), Some(wordcraft_geom::Spin::new(45.0, true, true)));
+    let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap();
+    assert_eq!(spin(&again), spin(&d));
+    // A see-through rule stays see-through.
+    let fill = |d: &Document| paras(d)[0].objects.iter().find_map(|o| if let InlineObject::Shape { fill, .. } = o { fill.clone() } else { None });
+    let d = read_body(
+        r##"<w:p><w:r><w:pict><v:rect style="width:0;height:1.5pt" o:hr="t" fillcolor="#ff0000"><v:fill opacity="0.25"/></v:rect></w:pict></w:r></w:p>"##,
+    );
+    let Some(wordcraft_doc::fill::Fill::Solid { transparency, .. }) = fill(&d) else { panic!("{:?}", fill(&d)) };
+    assert!((transparency - 75.0).abs() < 0.01, "{transparency}");
+    let again = wordcraft_docx::read(&wordcraft_docx::write(&d).unwrap()).unwrap();
+    assert_eq!(fill(&again), fill(&d));
+    // A positioned or hidden rule stays a plain shape, keeping its placement and visibility.
+    let body = [
+        hr(r#"o:hralign="center" o:hr="t""#, "position:absolute;margin-left:10pt;margin-top:20pt;width:100pt;height:2pt"),
+        hr(r#"o:hralign="center" o:hr="t""#, "visibility:hidden;width:0;height:2pt"),
+    ]
+    .concat();
+    let d = read_body(&body);
+    assert_eq!(rules(&d).iter().map(|r| r.0).collect::<Vec<_>>(), [None, None]);
+}
@@ -1,3 +1,4 @@
+mod alt_chunks;
 mod bidi;
 mod bidi_runs;
 mod cell_margin_inheritance;
@@ -41,6 +41,9 @@ js-sys = "0.3"

 [target.'cfg(not(target_arch = "wasm32"))'.dependencies]
 getrandom = { workspace = true }
+# Language packs (CP4.3): downloaded, verified and unpacked by the suite's font store.
+craft-font-store = { workspace = true }
+craft-font-catalog = { workspace = true }

 [target.'cfg(unix)'.dependencies]
 # O_NONBLOCK, so a path swapped for a named pipe can't block `open` (#3666).
@@ -28,6 +28,7 @@ pub mod moves;
 pub mod objects;
 pub mod options;
 pub mod outline;
+pub mod packs;
 pub mod page;
 pub mod para;
 pub mod paste;
@@ -84,6 +85,7 @@ pub fn registry() -> Registry {
    v.extend(review::specs());
    v.extend(file::specs());
    v.extend(options::specs());
+    v.extend(packs::specs());
    v.extend(design::specs());
    v.extend(references::specs());
    v.extend(fields::specs());
@@ -134,7 +134,8 @@ pub fn specs() -> Vec<CommandSpec> {
        CommandSpec::new("insert.textFromFile", "Text from File", "Insert › Text › Object", |s, v| {
            let path = p::req_str(v, "path")?;
            let doc = crate::io::open_path(std::path::Path::new(path)).map_err(CmdError::Failed)?;
-            let mut frag = doc.body_fragment();
+            // The file's pictures, lists and styles come along (its own keys and ids mean nothing here).
+            let mut frag = s.doc.adopt(doc);
            super::paste::unlink_sections(&mut frag.blocks);
            let at = delete_selection(s)?;
            let end = s.doc.insert_fragment(&at, &frag)?;
@@ -235,7 +235,7 @@ pub fn current(s: &Session) -> Value {
        "userTemplatesFolder": pr.user_templates_folder,
        "dictionaries": wordcraft_proof::installed_dictionaries()
            .into_iter()
-            .map(|(lang, dic)| json!({"language": lang.code(), "file": dic.map(|d| d.to_string_lossy().into_owned())}))
+            .map(|(lang, dic)| json!({"language": lang.tag(), "file": dic.map(|d| d.to_string_lossy().into_owned())}))
            .collect::<Vec<_>>(),
    })
 }
@@ -0,0 +1,423 @@
+//! Language packs as commands (CP4.3): list, install and remove, for the CLI, MCP and the control
+//! channel. They open nothing and never turn downloading on: an install needs the font store's
+//! consent setting, which only a person in the app turns on (File › Options › Language).
+//! See [`crate::packs`].
+
+use serde_json::{Value, json};
+
+use crate::packs;
+use crate::{CmdError, CmdResult, CommandSpec, Session, p};
+
+pub fn specs() -> Vec<CommandSpec> {
+    vec![
+        CommandSpec::new("proof.packs", "Language Packs", "File › Options › Language", list)
+            .params(r#"{"refresh"?: bool (fetch the pinned index again; only while downloading is on; in the app it runs in the background: see "refreshing")} — opens nothing. Returns {"packs": [{"locale", "name", "size" (bytes), "licence" (SPDX), "class", "source" (where the data comes from), "archive" (the download), "spelling", "hyphenation", "installed"}], "installed": [{"locale", "name", "licence"}], "installing": [{"locale", "done", "total"} (bytes)], "refreshing": bool, "errors": {locale or "" (the list): message} (the last failure of each), "index_override": string|null (a test index in use instead of the pinned one), "network": bool (downloading on), "available": bool (a font store to install into)}"#)
+            .pure(),
+        CommandSpec::new("proof.installPack", "Install Language Pack", "File › Options › Language", install)
+            .params(r#"{"locale": string (a pack's locale from proof.packs, like nl or pt-pt)} — downloads, verifies (sha256) and installs it; spelling and hyphenation in that language use it at once. Needs downloading on (only the person in the app can turn it on). Opens nothing. In the app it runs in the background and returns {"started": true}: follow it with proof.packs ("installing", then "installed" or "errors"); elsewhere it returns when installed."#)
+            .pure()
+            .when(available),
+        CommandSpec::new("proof.removePack", "Remove Language Pack", "File › Options › Language", remove)
+            .params(r#"{"locale": string} — removes the installed pack. Opens nothing."#)
+            .pure()
+            .when(available),
+    ]
+}
+
+fn available(_: &Session) -> Option<&'static str> {
+    if cfg!(target_arch = "wasm32") { Some("Language packs aren't available in the browser") } else { None }
+}
+
+fn locale_param(v: &Value) -> Result<String, CmdError> {
+    let raw = p::str(v, "locale").ok_or_else(|| CmdError::Params("`locale` (a language tag like nl) is required".into()))?;
+    packs::parse_locale(raw)
+}
+
+fn list(_: &mut Session, v: &Value) -> CmdResult {
+    if p::bool(v, "refresh") == Some(true) {
+        packs::start_refresh().map_err(CmdError::Failed)?;
+        if let Some(e) = packs::error("") {
+            return Err(CmdError::Failed(e));
+        }
+    }
+    let index = packs::available();
+    let installed = packs::installed();
+    let backend = packs::backend();
+    let rows: Vec<Value> = index
+        .iter()
+        .map(|p| {
+            json!({
+                "locale": p.locale, "name": p.name, "size": p.size, "licence": p.licence, "class": p.class,
+                "source": p.source_link(), "archive": p.url, "spelling": p.spelling, "hyphenation": p.hyphenation,
+                "installed": installed.iter().any(|i| i.locale == p.locale),
+            })
+        })
+        .collect();
+    Ok(json!({
+        "packs": rows,
+        "installed": installed.iter().map(|i| json!({"locale": i.locale, "name": i.name, "licence": i.licence})).collect::<Vec<_>>(),
+        "installing": packs::installing(),
+        "refreshing": packs::fetching(),
+        "errors": packs::errors().into_iter().map(|(l, e)| (l, Value::String(e))).collect::<serde_json::Map<String, Value>>(),
+        "network": backend.as_ref().is_some_and(|b| b.network_on()),
+        // The packs come from a test index (`CRAFT_LANGUAGE_PACK_INDEX`), not the pinned one.
+        "index_override": packs::index_override(),
+        "available": backend.is_some(),
+    }))
+}
+
+fn install(s: &mut Session, v: &Value) -> CmdResult {
+    let locale = locale_param(v)?;
+    if packs::is_installed(&locale) {
+        return Ok(json!({"locale": locale, "installed": true, "already": true}));
+    }
+    if packs::start_install(&locale).map_err(CmdError::Failed)? {
+        return Ok(json!({"locale": locale, "started": true}));
+    }
+    let done =
+        packs::installed().into_iter().find(|p| p.locale == locale).ok_or_else(|| CmdError::Failed(format!("the {locale} pack didn't install")))?;
+    // Spelling and hyphenation in that language change.
+    s.relayout();
+    Ok(json!({"locale": done.locale, "name": done.name, "licence": done.licence, "installed": true}))
+}
+
+fn remove(s: &mut Session, v: &Value) -> CmdResult {
+    let locale = locale_param(v)?;
+    packs::remove(&locale).map_err(CmdError::Failed)?;
+    s.relayout();
+    Ok(json!({"locale": locale, "installed": false}))
+}
+
+#[cfg(test)]
+mod tests {
+    use std::sync::Arc;
+    use std::sync::atomic::Ordering;
+
+    use super::*;
+    use crate::packs::{Backend, FixedBackend};
+
+    /// Tests that set the process-wide backend take turns.
+    static LOCK: std::sync::Mutex<()> = std::sync::Mutex::new(());
+
+    /// Puts the process back to no backend and no background thread when dropped, even when
+    /// the test panicked.
+    pub(crate) struct Reset;
+
+    impl Drop for Reset {
+        fn drop(&mut self) {
+            packs::set_background(None);
+            packs::set_backend(None);
+        }
+    }
+
+    fn with_backend<T>(b: FixedBackend, f: impl FnOnce(&mut Session, &FixedBackend) -> T) -> T {
+        let _g = LOCK.lock().unwrap_or_else(std::sync::PoisonError::into_inner);
+        let _reset = Reset;
+        let b = Arc::new(b);
+        packs::set_backend(Some(b.clone() as Arc<dyn Backend>));
+        let mut s = Session::new(crate::sample::sample_document());
+        f(&mut s, &b)
+    }
+
+    #[test]
+    fn lists_installs_and_removes_without_opening_anything() {
+        with_backend(FixedBackend::sample(), |s, b| {
+            let r = s.run("proof.packs", &json!({})).unwrap();
+            assert_eq!(r["packs"].as_array().unwrap().len(), 6);
+            assert_eq!(r["network"], false);
+            assert_eq!(r["index_override"], Value::Null, "the pinned index");
+            let nl = r["packs"].as_array().unwrap().iter().find(|p| p["locale"] == "nl").unwrap();
+            assert_eq!(nl["licence"], "BSD-3-Clause");
+            assert_eq!(nl["installed"], false);
+            // Downloading is off: an agent can't install, and can't turn it on.
+            let e = s.run("proof.installPack", &json!({"locale": "nl"})).unwrap_err();
+            assert!(e.to_string().contains("downloading is turned off"), "{e}");
+            assert!(packs::installed().is_empty());
+            b.network.store(true, Ordering::Relaxed);
+            let r = s.run("proof.installPack", &json!({"locale": "NL"})).unwrap();
+            assert_eq!(r["locale"], "nl");
+            assert!(packs::is_installed("nl"));
+            let again = s.run("proof.installPack", &json!({"locale": "nl"})).unwrap();
+            assert_eq!(again["already"], true);
+            let r = s.run("proof.packs", &json!({})).unwrap();
+            assert_eq!(r["installed"][0]["locale"], "nl");
+            s.run("proof.removePack", &json!({"locale": "nl"})).unwrap();
+            assert!(!packs::is_installed("nl"));
+            assert!(s.run("proof.removePack", &json!({"locale": "nl"})).is_err(), "not installed");
+            assert!(s.ui_requests.is_empty(), "no dialog opens");
+        });
+    }
+
+    /// The bar's missing-pack cache keys on `index_generation`, so a refresh that returns a list
+    /// of the same length must still change it (CodeRabbit on #5208).
+    #[test]
+    fn a_refresh_of_the_same_length_still_changes_the_index_generation() {
+        with_backend(FixedBackend::sample(), |s, b| {
+            b.network.store(true, Ordering::Relaxed);
+            let n = packs::available_len();
+            let before = packs::index_generation();
+            s.run("proof.packs", &json!({"refresh": true})).unwrap();
+            assert_eq!(packs::available_len(), n, "same length");
+            assert_ne!(packs::index_generation(), before);
+        });
+    }
+
+    #[test]
+    fn unknown_and_hostile_locales_are_errors() {
+        with_backend(FixedBackend::sample(), |s, b| {
+            b.network.store(true, Ordering::Relaxed);
+            for bad in [json!({}), json!({"locale": 5}), json!({"locale": "../../etc"}), json!({"locale": "x".repeat(10_000)}), json!({"locale": ""})]
+            {
+                assert!(s.run("proof.installPack", &bad).is_err(), "{bad}");
+                assert!(s.run("proof.removePack", &bad).is_err(), "{bad}");
+            }
+            let e = s.run("proof.installPack", &json!({"locale": "tlh"})).unwrap_err();
+            assert!(e.to_string().contains("no language pack for tlh"), "{e}");
+            assert!(packs::installed().is_empty());
+        });
+    }
+
+    #[test]
+    fn an_agent_backend_cannot_turn_downloading_on() {
+        let b = FixedBackend { agent: true, ..FixedBackend::sample() };
+        assert!(b.enable_network().is_err());
+        assert!(!b.network_on());
+    }
+
+    #[test]
+    fn without_a_store_nothing_is_available() {
+        let _g = LOCK.lock().unwrap_or_else(std::sync::PoisonError::into_inner);
+        let _reset = Reset;
+        packs::set_backend(None);
+        let mut s = Session::new(crate::sample::sample_document());
+        let r = s.run("proof.packs", &json!({})).unwrap();
+        assert_eq!(r["packs"], json!([]));
+        assert_eq!(r["available"], false);
+        assert!(s.run("proof.installPack", &json!({"locale": "nl"})).is_err());
+        assert!(s.run("proof.packs", &json!({"refresh": true})).is_err());
+    }
+
+    #[test]
+    fn the_bar_offers_only_languages_with_a_pack() {
+        with_backend(FixedBackend::sample(), |s, b| {
+            s.run("document.setText", &json!({"text": "Hallo wereld\nDzień dobry\nGuten Tag\nHello\nWitam"})).unwrap();
+            assert!(packs::missing(s).is_empty(), "all English");
+            for (para, lang) in [(0, "nl-NL"), (1, "pl-PL"), (2, "de-DE"), (4, "pl")] {
+                s.run("caret.set", &json!({"pos": {"block": para, "off": 0}})).unwrap();
+                s.run("select.paragraph", &json!({})).unwrap();
+                s.run("review.language", &json!({"lang": lang})).unwrap();
+            }
+            let langs = packs::document_languages(s);
+            assert!(langs.iter().any(|l| l == "nl-NL"), "{langs:?}");
+            let missing: Vec<String> = packs::missing(s).into_iter().map(|p| p.locale).collect();
+            // German has no pack: never offered.
+            assert_eq!(missing, ["nl", "pl"]);
+            b.network.store(true, Ordering::Relaxed);
+            s.run("proof.installPack", &json!({"locale": "pl"})).unwrap();
+            let missing: Vec<String> = packs::missing(s).into_iter().map(|p| p.locale).collect();
+            assert_eq!(missing, ["nl"]);
+            // Text marked "no language" (`zxx`) never asks for a pack.
+            assert!(!packs::proofable("zxx"));
+        });
+    }
+
+    #[test]
+    fn tags_map_to_the_most_specific_pack() {
+        let mut packs = FixedBackend::sample().packs;
+        packs.push(Pack { locale: "pt-pt".into(), ..packs[0].clone() });
+        packs.push(Pack { locale: "pt".into(), ..packs[0].clone() });
+        assert_eq!(packs::pack_for_tag("pt-PT", &packs).map(|p| p.locale.as_str()), Some("pt-pt"));
+        assert_eq!(packs::pack_for_tag("pt_BR", &packs).map(|p| p.locale.as_str()), Some("pt"));
+        assert_eq!(packs::pack_for_tag("nl-BE", &packs).map(|p| p.locale.as_str()), Some("nl"));
+        assert_eq!(packs::pack_for_tag("en-GB", &packs), None);
+        assert_eq!(packs::pack_for_tag("", &packs), None);
+        assert!(packs::valid_locale("sr-latn"));
+        assert!(!packs::valid_locale("sr--latn") && !packs::valid_locale("-x") && !packs::valid_locale("x/y"));
+    }
+
+    /// An install or a removal in one window proofs every open document again, parked ones too:
+    /// their layouts are keyed on the dictionaries' generation, not only their own revision.
+    #[test]
+    fn installing_proofs_parked_documents_again() {
+        with_backend(FixedBackend::sample(), |s, b| {
+            let mut parked = Session::new(crate::sample::sample_document());
+            let before = parked.layout();
+            assert!(std::sync::Arc::ptr_eq(&before, &parked.layout()), "unchanged: the same layout");
+            b.network.store(true, Ordering::Relaxed);
+            s.run("proof.installPack", &json!({"locale": "pl"})).unwrap();
+            assert!(!std::sync::Arc::ptr_eq(&before, &parked.layout()), "the parked document is laid out (and proofed) again");
+            let after = parked.layout();
+            s.run("proof.removePack", &json!({"locale": "pl"})).unwrap();
+            assert!(!std::sync::Arc::ptr_eq(&after, &parked.layout()));
+        });
+    }
+
+    /// A backend whose installs wait until the test lets them finish.
+    struct Slow {
+        inner: FixedBackend,
+        go: std::sync::Mutex<std::sync::mpsc::Receiver<()>>,
+    }
+
+    impl Backend for Slow {
+        fn index(&self) -> Result<Vec<Pack>, String> {
+            self.inner.index()
+        }
+        fn fetch_index(&self, c: &packs::Cancel) -> Result<Vec<Pack>, String> {
+            self.inner.fetch_index(c)
+        }
+        fn network_on(&self) -> bool {
+            self.inner.network_on()
+        }
+        fn enable_network(&self) -> Result<(), String> {
+            self.inner.enable_network()
+        }
+        fn install(&self, pack: &Pack, cancel: &packs::Cancel, progress: &mut dyn FnMut(u64, u64)) -> Result<(), String> {
+            progress(1, pack.size);
+            let _ = self.go.lock().unwrap().recv();
+            self.inner.install(pack, cancel, progress)
+        }
+        fn remove(&self, locale: &str) -> Result<(), String> {
+            self.inner.remove(locale)
+        }
+        fn packs_dir(&self) -> Option<std::path::PathBuf> {
+            None
+        }
+        fn installed(&self) -> Vec<packs::Installed> {
+            self.inner.installed()
+        }
+    }
+
+    /// In the app an install runs on a thread, whoever asks: the command returns at once, and
+    /// `proof.packs` shows it running, then installed. A removal of the same pack meanwhile is
+    /// refused.
+    #[test]
+    fn in_the_app_installs_run_in_the_background() {
+        let _g = LOCK.lock().unwrap_or_else(std::sync::PoisonError::into_inner);
+        let _reset = Reset;
+        let (tx, rx) = std::sync::mpsc::channel();
+        let slow = Arc::new(Slow { inner: FixedBackend::sample(), go: std::sync::Mutex::new(rx) });
+        slow.inner.network.store(true, Ordering::Relaxed);
+        packs::set_backend(Some(slow.clone() as Arc<dyn Backend>));
+        let notified = Arc::new(std::sync::atomic::AtomicUsize::new(0));
+        let n = notified.clone();
+        packs::set_background(Some(Arc::new(move || {
+            n.fetch_add(1, Ordering::Relaxed);
+        })));
+        let mut s = Session::new(crate::sample::sample_document());
+        let before = packs::finished();
+        let r = s.run("proof.installPack", &json!({"locale": "pl"})).unwrap();
+        assert_eq!(r["started"], true);
+        let list = s.run("proof.packs", &json!({})).unwrap();
+        assert_eq!(list["installing"][0]["locale"], "pl");
+        let e = s.run("proof.removePack", &json!({"locale": "pl"})).unwrap_err();
+        assert!(e.to_string().contains("being installed"), "{e}");
+        tx.send(()).unwrap();
+        for _ in 0..500 {
+            if packs::finished() != before {
+                break;
+            }
+            std::thread::sleep(std::time::Duration::from_millis(10));
+        }
+        assert_ne!(packs::finished(), before);
+        assert!(notified.load(Ordering::Relaxed) > 0, "the window is told");
+        let list = s.run("proof.packs", &json!({})).unwrap();
+        assert_eq!(list["installing"], json!([]));
+        assert_eq!(list["installed"][0]["locale"], "pl");
+        assert!(s.ui_requests.is_empty(), "no dialog opens");
+    }
+
+    /// A pack read from its folder proofs and hyphenates its language's runs, and leaves the
+    /// English runs to English: a tiny Lithuanian pack (`ProofLang::Pack`, no rules of its own).
+    #[test]
+    fn an_installed_pack_proofs_and_hyphenates_its_runs() {
+        let _g = LOCK.lock().unwrap_or_else(std::sync::PoisonError::into_inner);
+        let _reset = Reset;
+        let dir = std::env::temp_dir().join(format!("wordcraft-pack-fixture-{}", std::process::id()));
+        let lt = dir.join("lt");
+        std::fs::create_dir_all(&lt).unwrap();
+        std::fs::write(lt.join("pack.toml"), "locale = \"lt\"\nname = \"Lietuvių\"\nlicence = \"MIT\"\nclass = \"Embeddable\"\n").unwrap();
+        std::fs::write(lt.join("spelling.aff"), "SET UTF-8\n").unwrap();
+        std::fs::write(lt.join("spelling.dic"), "2\nlabas\nrytas\n").unwrap();
+        // One break, laba-rytas; the other letters only make up the patterns' alphabet.
+        std::fs::write(lt.join("hyphenation.pat"), "a1r b l s t y z\n").unwrap();
+        wordcraft_proof::packs::set_packs_dir(Some(dir.clone()));
+        let text = "labas rytaz labarytas typography beautifull";
+        let mut s = Session::new(wordcraft_doc::Document::from_text(text));
+        let eng = text.find("typography").unwrap();
+        s.doc.format_range(&wordcraft_doc::Pos::body(0, 0), &wordcraft_doc::Pos::body(0, eng), &|c| c.lang = Some("lt-LT".into())).unwrap();
+        s.doc.format_range(&wordcraft_doc::Pos::body(0, eng), &wordcraft_doc::Pos::body(0, text.len()), &|c| c.lang = Some("en-US".into())).unwrap();
+        s.doc.settings.auto_hyphenation = true;
+        s.relayout();
+        let issues = s.run("review.issues", &json!({})).map(|v| v.to_string()).unwrap_or_default();
+        let layout = s.layout();
+        let ends: Vec<usize> = layout
+            .pages
+            .iter()
+            .flat_map(|p| &p.items)
+            .filter_map(|i| if let wordcraft_layout::Placed::Lines { para, .. } = i { Some(para) } else { None })
+            .flat_map(|para| para.hyph_after.iter().filter_map(|&k| para.clusters.get(k as usize).map(|c| c.end)).collect::<Vec<_>>())
+            .collect();
+        wordcraft_proof::packs::set_packs_dir(None);
+        let _ = std::fs::remove_dir_all(&dir);
+        let laba = text.find("labarytas").unwrap() + "laba".len();
+        assert!(ends.contains(&laba), "the pack's pattern breaks laba-rytas: {ends:?}");
+        assert!(issues.contains("rytaz"), "the Lithuanian misspelling, from the pack: {issues}");
+        assert!(issues.contains("beautifull"), "the English one: {issues}");
+        assert!(!issues.contains("\"labas\"") && !issues.contains("typography"), "{issues}");
+    }
+
+    /// Japanese, Chinese and Thai have no Hunspell dictionary anywhere: without a pack for them
+    /// in the list the status bar doesn't say "(no dictionary)" and the bar offers nothing; once
+    /// the list has one, both do.
+    #[test]
+    fn a_language_with_no_pack_and_no_dictionary_asks_for_nothing() {
+        with_backend(FixedBackend::sample(), |s, _| {
+            for tag in ["ja-JP", "zh-CN", "th-TH"] {
+                assert_eq!(packs::missing_dictionary(Some(tag)), None, "{tag}");
+            }
+            s.run("document.setText", &json!({"text": "日本語のテキスト"})).unwrap();
+            s.run("select.all", &json!({})).unwrap();
+            s.run("review.language", &json!({"lang": "ja-JP"})).unwrap();
+            assert!(packs::missing(s).is_empty(), "no Japanese pack: nothing offered");
+        });
+        let mut b = FixedBackend::sample();
+        let ja = Pack { locale: "ja".into(), name: "日本語".into(), ..b.packs[0].clone() };
+        b.packs.push(ja);
+        with_backend(b, |s, _| {
+            assert!(packs::missing_dictionary(Some("ja-JP")).is_some(), "a Japanese pack is available");
+            s.run("document.setText", &json!({"text": "日本語のテキスト"})).unwrap();
+            s.run("select.all", &json!({})).unwrap();
+            s.run("review.language", &json!({"lang": "ja-JP"})).unwrap();
+            assert_eq!(packs::missing(s).into_iter().map(|p| p.locale).collect::<Vec<_>>(), ["ja"]);
+        });
+    }
+
+    /// Swiss German and European Portuguese are pack languages in craft-proof (`de-ch`, `pt-pt`:
+    /// their spelling is their own) but hyphenate and check grammar as German and Portuguese.
+    #[test]
+    fn swiss_german_and_portugal_portuguese_keep_their_languages_rules() {
+        use wordcraft_proof::ProofLang;
+        let de_ch = ProofLang::from_tag(Some("de-CH"));
+        let pt_pt = ProofLang::from_tag(Some("pt-PT"));
+        assert_eq!(de_ch.tag(), "de-ch");
+        assert_eq!(pt_pt.tag(), "pt-pt");
+        assert!(packs::proofable("de-CH") && packs::proofable("pt-PT"));
+        for word in ["Zusammenarbeit", "Rechtschreibung"] {
+            let german = wordcraft_proof::hyphen::word_breaks_in(word, ProofLang::De).0;
+            assert!(!german.is_empty(), "German hyphenates {word}");
+            assert_eq!(wordcraft_proof::hyphen::word_breaks_in(word, de_ch).0, german, "Swiss German hyphenates {word} as German");
+        }
+        for word in ["computador", "informação"] {
+            assert_eq!(
+                wordcraft_proof::hyphen::word_breaks_in(word, pt_pt).0,
+                wordcraft_proof::hyphen::word_breaks_in(word, ProofLang::Pt).0,
+                "{word}"
+            );
+        }
+        let messages = |lang| wordcraft_proof::check_grammar_in("Ele disse que que viria .", lang).into_iter().map(|i| i.message).collect::<Vec<_>>();
+        assert!(!messages(ProofLang::Pt).is_empty());
+        assert_eq!(messages(pt_pt), messages(ProofLang::Pt), "Portugal text gets the Portuguese grammar checks");
+    }
+
+    use crate::packs::Pack;
+}
@@ -374,6 +374,9 @@ fn line_numbering_options() {
 /// there breaks the word.
 #[test]
 fn hyphenation_options_and_manual_hyphenation() {
+    if crate::tests::no_language_data("hyphenation_options_and_manual_hyphenation") {
+        return;
+    }
    let mut s =
        Session::new(wordcraft_doc::Document::from_text(&"a extraordinarily be uncharacteristic of responsibilities the considerations ".repeat(12)));
    let r = s.run("layout.hyphenation", &json!({"zone": 36, "caps": false, "limit": 3})).unwrap();
@@ -8,7 +8,6 @@
 //! program's HTML or RTF; Merge Formatting keeps only its emphasis (#687).
 //! Plain text has no formatting to keep, so Paste inserts it as Keep Text Only does.

-use std::collections::BTreeMap;
 use std::sync::Arc;

 use serde_json::{Value, json};
@@ -401,35 +400,6 @@ pub(crate) fn each_para(b: &mut Block, depth: usize, f: &mut dyn FnMut(&mut Para
    }
 }

-/// Point a tracked formatting change at its renumbered revision (dropped when it has none).
-fn remap_change<T>(c: &mut Option<Box<wordcraft_doc::props::PropChange<T>>>, rev: &dyn Fn(&mut Option<u32>)) {
-    let mut r = c.as_ref().map(|c| c.rev);
-    rev(&mut r);
-    match (r, c.as_mut()) {
-        (Some(v), Some(x)) => x.rev = v,
-        _ => *c = None,
-    }
-}
-
-/// [`remap_change`] for the table, row and cell formatting changes in `b` (bounded depth).
-fn remap_table_changes(b: &mut Block, depth: usize, rev: &dyn Fn(&mut Option<u32>)) {
-    if depth > MAX_DEPTH {
-        return;
-    }
-    if let Block::Table(t) = b {
-        remap_change(&mut t.props.fmt_change, rev);
-        for row in &mut t.rows {
-            remap_change(&mut row.props.fmt_change, rev);
-            for cell in &mut row.cells {
-                remap_change(&mut cell.props.fmt_change, rev);
-                for cb in &mut cell.blocks {
-                    remap_table_changes(Arc::make_mut(cb), depth + 1, rev);
-                }
-            }
-        }
-    }
-}
-
 /// Section breaks brought in from another document keep their page, but their header and
 /// footer ids are that document's: clear them (and the different first page that went with
 /// them) so the sections link to the previous section's headers and footers, as in Word.
@@ -448,81 +418,12 @@ pub(crate) fn unlink_sections(blocks: &mut [Block]) {
 /// Turn an imported document into a fragment of the session's document: bring over the media,
 /// lists and styles it refers to (renumbered so they can't collide with ours).
 fn adopt(s: &mut Session, src: Document) -> Fragment {
-    let mut media = BTreeMap::new();
-    for (key, data) in &src.media {
-        let ext = key.rsplit_once('.').map(|(_, e)| e).unwrap_or("png");
-        media.insert(key.clone(), s.doc.add_media(data.to_vec(), ext));
-    }
-    let next_abs = s.doc.numbering.abstracts.iter().map(|a| a.id.saturating_add(1)).max().unwrap_or(0);
-    let mut abs_map = BTreeMap::new();
-    for (i, a) in src.numbering.abstracts.iter().enumerate() {
-        let id = next_abs.saturating_add(i as u32);
-        abs_map.insert(a.id, id);
-        s.doc.numbering.abstracts.push(wordcraft_doc::numbering::AbstractNum { id, ..a.clone() });
-    }
-    let next_num = s.doc.numbering.nums.iter().map(|n| n.id.saturating_add(1)).max().unwrap_or(1).max(1);
-    let mut num_map = BTreeMap::new();
-    for (i, n) in src.numbering.nums.iter().enumerate() {
-        let id = next_num.saturating_add(i as u32);
-        num_map.insert(n.id, id);
-        let abstract_id = abs_map.get(&n.abstract_id).copied().unwrap_or(n.abstract_id);
-        s.doc.numbering.nums.push(wordcraft_doc::numbering::Num { id, abstract_id, ..n.clone() });
-    }
-    for st in &src.styles.styles {
-        if s.doc.styles.get(&st.id).is_none() {
-            s.doc.styles.upsert(st.clone());
-        }
-    }
-    // Tracked changes the source carries (RTF revisions) point at its own revisions: add them
-    // after ours and renumber.
-    let rev_base = u32::try_from(s.doc.revisions.len()).unwrap_or(u32::MAX);
-    let src_revs = src.revisions.len();
-    s.doc.revisions.extend(src.revisions.iter().cloned());
-    let rev = |i: &mut Option<u32>| {
-        *i = i.filter(|v| (*v as usize) < src_revs).and_then(|v| rev_base.checked_add(v));
-    };
-    let mut blocks: Vec<Block> = src.body.iter().map(|b| (**b).clone()).collect();
+    let mut frag = s.doc.adopt(src);
    // 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 {
+    for b in &mut frag.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()
-                && n.num != 0
-            {
-                n.num = num_map.get(&n.num).copied().unwrap_or(0);
-            }
-            for r in &mut p.runs {
-                rev(&mut r.props.ins);
-                rev(&mut r.props.del);
-                remap_change(&mut r.props.fmt_change, &rev);
-            }
-            rev(&mut p.mark.ins);
-            rev(&mut p.mark.del);
-            remap_change(&mut p.mark.fmt_change, &rev);
-            remap_change(&mut p.props.fmt_change, &rev);
-            let mut n = p.props.num_change.as_ref().map(|c| c.rev);
-            rev(&mut n);
-            match (n, p.props.num_change.as_mut()) {
-                (Some(v), Some(c)) => c.rev = v,
-                _ => p.props.num_change = None,
-            }
-            for o in &mut p.objects {
-                if let wordcraft_doc::InlineObject::Image { media: key, .. } = o
-                    && let Some(k) = media.get(key.as_str())
-                {
-                    *key = k.clone();
-                }
-            }
-        });
-    }
-    for b in &mut blocks {
-        remap_table_changes(b, 0, &rev);
-    }
-    unlink_sections(&mut blocks);
-    // Text box stories the imported shapes show; pasting gives each a new story of its own.
-    let parts = src.parts.into_iter().filter(|(_, p)| p.kind == wordcraft_doc::PartKind::TextBox).collect();
-    Fragment { blocks, parts }
+    unlink_sections(&mut frag.blocks);
+    frag
 }
@@ -80,6 +80,7 @@ pub fn specs() -> Vec<CommandSpec> {
        CommandSpec::new("review.changes", "Reviewing Pane", "Review › Tracking", list_changes).pure(),
        CommandSpec::new("review.spelling", "Spelling & Grammar", "Review › Proofing", next_issue)
            .key(super::mac_or("F7 / Alt+F7 / Mod+Alt+L", "F7 / Alt+F7"))
+            .params(r#"{"wrap"?: bool (default true; false stops at the end of the current story)}"#)
            .pure(),
        CommandSpec::new("review.issues", "Proofing Issues", "Review › Proofing", all_issues).pure(),
        CommandSpec::new("review.suggestions", "Spelling Suggestions", "Review › Proofing", suggestions).params(r#"{"pos"?: Pos}"#).pure(),
@@ -55,7 +55,16 @@ fn para_issues(s: &Session, story: StoryRef, path: &wordcraft_doc::Path) -> Vec<

 /// The issue under a position.
 pub(super) fn issue_at(s: &Session, at: &Pos) -> Option<(Pos, Pos, wordcraft_proof::Issue)> {
-    para_issues(s, at.story, &at.path).into_iter().find(|(a, b, _)| a.off <= at.off && at.off <= b.off)
+    let issues = para_issues(s, at.story, &at.path);
+    let selected = s.sel.ordered();
+    // F7 can select a spelling issue inside a larger grammar issue. Keep that
+    // exact target for either selection direction before falling back to a point.
+    if at == &s.sel.focus
+        && let Some(issue) = issues.iter().find(|(a, b, _)| a == &selected.0 && b == &selected.1)
+    {
+        return Some(issue.clone());
+    }
+    issues.into_iter().find(|(a, b, _)| a.off <= at.off && at.off <= b.off)
 }

 fn issue_json(s: &Session, a: &Pos, b: &Pos, i: &wordcraft_proof::Issue) -> Value {
@@ -64,20 +73,28 @@ fn issue_json(s: &Session, a: &Pos, b: &Pos, i: &wordcraft_proof::Issue) -> Valu
    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(),
+        "message": i.message, "suggestions": sugg, "lang": i.lang.tag(),
    })
 }

 /// F7: select the next issue after the caret and return it with suggestions.
-pub(super) fn next_issue(s: &mut Session, _: &Value) -> CmdResult {
+pub(super) fn next_issue(s: &mut Session, v: &Value) -> CmdResult {
    let caret = s.sel.ordered().1;
    let story = caret.story;
    let paths = s.doc.para_paths(story);
    let start = paths.iter().position(|p| *p == caret.path).unwrap_or(0);
-    for k in 0..paths.len() {
-        let Some(path) = paths.get((start + k) % paths.len().max(1)) else { break };
+    let wrap = p::bool(v, "wrap").unwrap_or(true);
+    // Scan forward first, including the rest of the caret's paragraph. A second pass
+    // optionally visits the earlier part; never select the first typo repeatedly in
+    // a single-paragraph story. Dialogs disable wrapping so Ignore Once can finish.
+    for (path, before) in paths
+        .iter()
+        .skip(start)
+        .map(|path| (path, false))
+        .chain(paths.iter().take(if wrap { start.saturating_add(1) } else { 0 }).map(|path| (path, true)))
+    {
        for (a, b, i) in para_issues(s, story, path) {
-            if k == 0 && a.off < caret.off && start + k < paths.len() && paths.len() > 1 {
+            if *path == caret.path && (a.off < caret.off) != before {
                continue;
            }
            let r = issue_json(s, &a, &b, &i);
@@ -103,8 +120,13 @@ pub(super) fn all_issues(s: &mut Session, _: &Value) -> CmdResult {
 }

 pub(super) fn suggestions(s: &mut Session, v: &Value) -> CmdResult {
-    let at = v.get("pos").and_then(crate::cmd::parse_pos).unwrap_or_else(|| s.sel.focus.clone());
-    match issue_at(s, &at) {
+    // Explicit positions are point lookups, even when they equal the selection focus.
+    // Only an implicit or unparseable position keeps the exact selected issue.
+    let issue = match v.get("pos").and_then(crate::cmd::parse_pos) {
+        Some(at) => para_issues(s, at.story, &at.path).into_iter().find(|(a, b, _)| a.off <= at.off && at.off <= b.off),
+        None => issue_at(s, &s.sel.focus),
+    };
+    match issue {
        Some((a, b, i)) => Ok(issue_json(s, &a, &b, &i)),
        None => Ok(Value::Null),
    }
@@ -1058,3 +1058,66 @@ fn rtf_tracked_changes_accept_and_reject_after_opening_and_round_trip() {
        }
    }
 }
+
+#[cfg(test)]
+mod spelling_navigation_tests {
+    use super::*;
+    use wordcraft_doc::{Document, Pos, StoryRef};
+
+    use crate::Selection;
+
+    #[test]
+    fn spelling_review_dialog_navigation_moves_forward_and_wraps_only_when_requested() {
+        for text in ["The brwn fox and brwn dog.", "The brwn fox.\nThe brwn dog."] {
+            let mut s = Session::new(Document::from_text(text));
+            let first = s.run("review.spelling", &json!({"wrap": false})).unwrap();
+            assert_eq!(first["text"], "brwn");
+            let first_selection = s.sel.clone();
+            let second = s.run("review.spelling", &json!({"wrap": false})).unwrap();
+            assert_eq!(second["text"], "brwn");
+            assert_ne!(s.sel, first_selection);
+            assert_eq!(s.run("review.spelling", &json!({"wrap": false})).unwrap()["done"], true);
+            assert_eq!(s.run("review.spelling", &json!({})).unwrap()["start"], first["start"]);
+        }
+    }
+    #[test]
+    fn spelling_review_dialog_navigation_resolves_selected_spelling_before_overlapping_grammar() {
+        for reverse in [false, true] {
+            let mut s = Session::new(Document::from_text("A sentense sentense."));
+            s.sel = Selection::caret(Pos::body(0, 11));
+            let selected = s.run("review.spelling", &json!({"wrap": false})).unwrap();
+            assert_eq!(selected["kind"], "spelling");
+            assert_eq!(selected["start"]["off"], 11);
+            assert_eq!(selected["text"], "sentense");
+            let issues = s.run("review.issues", &json!({})).unwrap();
+            assert!(issues.as_array().unwrap().iter().any(|i| i["kind"] == "grammar" && i["text"] == "sentense sentense"));
+            if reverse {
+                std::mem::swap(&mut s.sel.anchor, &mut s.sel.focus);
+            }
+            assert_eq!(s.run("review.suggestions", &json!({})).unwrap(), selected);
+            let before_lookup = s.sel.clone();
+            let focus = s.sel.focus.off;
+            let explicit = s.run("review.suggestions", &json!({"pos": {"block": 0, "off": focus}})).unwrap();
+            assert_eq!(explicit["kind"], "grammar", "an explicit focus is a point lookup in either selection direction");
+            assert_eq!(explicit["text"], "sentense sentense");
+            assert_eq!(explicit["start"]["off"], 2);
+            assert_eq!(explicit["end"]["off"], 19);
+            assert_eq!(s.sel, before_lookup, "suggestion lookups do not move the selection");
+            for pos in [serde_json::Value::Null, json!({"off": "invalid"})] {
+                assert_eq!(
+                    s.run("review.suggestions", &json!({"pos": pos})).unwrap(),
+                    selected,
+                    "unparseable positions preserve implicit selection lookup"
+                );
+            }
+            assert_eq!(s.run("review.suggestions", &json!({})).unwrap(), selected);
+            let elsewhere = s.run("review.suggestions", &json!({"pos": {"block": 0, "off": 5}})).unwrap();
+            assert_eq!(elsewhere["kind"], "spelling");
+            assert_eq!(elsewhere["start"]["off"], 2, "an explicit other position keeps point lookup");
+            s.run("review.applySuggestion", &json!({"text": "sentence"})).unwrap();
+            assert_eq!(s.doc.plain_text(StoryRef::Body), "A sentense sentence.");
+            s.run("edit.undo", &json!({})).unwrap();
+            assert_eq!(s.doc.plain_text(StoryRef::Body), "A sentense sentense.");
+        }
+    }
+}
@@ -1,6 +1,8 @@
 use super::{
-    Arc, Block, CharProps, CmdError, CmdResult, Document, Paragraph, RevisionKind, Selection, Session, StoryRef, Value, json, p, para_block,
+    Arc, Block, CharProps, CmdError, CmdResult, Document, InlineObject, Paragraph, RevisionKind, Selection, Session, StoryRef, Value, json, p,
+    para_block,
 };
+use wordcraft_doc::para::OBJ;

 /// Compare and Combine take the same settings.
 pub(super) 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)}"#;
@@ -64,7 +66,13 @@ pub(super) fn compare(s: &mut Session, v: &Value) -> CmdResult {
    let author = p::str(v, "author").map(str::trim).filter(|a| !a.is_empty()).map(str::to_string).unwrap_or_else(|| s.author.clone());
    let date = crate::cmd::now_iso();
    let old: Vec<Arc<Block>> = s.doc.body.clone();
-    let new: Vec<Arc<Block>> = revised.body.clone();
+    let new: Vec<Arc<Block>> = if revised.parts.values().any(|p| p.kind == wordcraft_doc::PartKind::TextBox) {
+        // The revised document's text boxes get stories of their own in the result, so its
+        // boxes never show a story of the original's (or none at all).
+        s.doc.adopt_fragment(&revised.body_fragment()).into_iter().map(Arc::new).collect()
+    } else {
+        revised.body.clone()
+    };
    s.doc.revisions.push(wordcraft_doc::Revision { kind: RevisionKind::Insert, author: author.clone(), date: date.clone(), move_name: None });
    let ins = (s.doc.revisions.len() - 1) as u32;
    s.doc.revisions.push(wordcraft_doc::Revision { kind: RevisionKind::Delete, author: author.clone(), date: date.clone(), move_name: None });
@@ -86,6 +94,16 @@ pub(super) fn compare(s: &mut Session, v: &Value) -> CmdResult {
                let (a, b) = (old.get(i), new.get(j));
                // The revised block, with its formatting changes tracked.
                if let (Some(Block::Para(pa)), Some(Block::Para(pb))) = (a.map(|x| &**x), b.map(|x| &**x)) {
+                    if !same_objects(&s.doc, pa, pb) {
+                        // Same text, but objects (text boxes, pictures…) differ: diffed, so the
+                        // original's objects aren't lost.
+                        out.push(para_block(word_diff(&s.doc, pa, pb, ins, del, &opts, &mut budget)));
+                        changes += 1;
+                        i += 1;
+                        j += 1;
+                        k += 1;
+                        continue;
+                    }
                    let mut p = pb.clone();
                    if opts.formatting && pa.text == pb.text {
                        let before = s.doc.revisions.len();
@@ -111,7 +129,9 @@ pub(super) fn compare(s: &mut Session, v: &Value) -> CmdResult {
                for x in 0..dels.max(inss) {
                    let (a, b) = (if x < dels { old.get(i + x) } else { None }, if x < inss { new.get(j + x) } else { None });
                    match (a.map(|x| &**x), b.map(|x| &**x)) {
-                        (Some(Block::Para(pa)), Some(Block::Para(pb))) => out.push(para_block(word_diff(pa, pb, ins, del, &opts, &mut budget))),
+                        (Some(Block::Para(pa)), Some(Block::Para(pb))) => {
+                            out.push(para_block(word_diff(&s.doc, pa, pb, ins, del, &opts, &mut budget)))
+                        }
                        _ => {
                            if let Some(x) = a {
                                out.push(Arc::new(mark_block(x, &|c| c.del = Some(del), 0)));
@@ -308,30 +328,97 @@ fn flatten_region(out: &mut Vec<Op>, region: &[(Op, usize, Kind)]) {
    out.extend(std::iter::repeat_n(Op::Ins, inss));
 }

-/// Words with the spaces after them ("The ", "cat "), or single characters.
-fn tokens(s: &str, characters: bool) -> Vec<String> {
-    if characters {
-        return s.chars().map(String::from).collect();
+/// A token of a changed paragraph: a word (or character), or an inline object without text
+/// of its own (a text box anchor, a picture, a bookmark…), which is compared and kept whole.
+enum Tok<'a> {
+    Text(String),
+    Obj(&'a InlineObject),
+}
+
+/// What a comparison sees of a token: the text as the options say, an object with the text of
+/// the text boxes it shows instead of their story ids (each document numbers its own stories).
+#[derive(PartialEq)]
+enum Key {
+    Text(String),
+    Obj(Box<InlineObject>, Vec<String>),
+}
+
+fn obj_key(doc: &Document, o: &InlineObject) -> Key {
+    let texts = o.text_boxes().into_iter().map(|id| doc.plain_text(StoryRef::Part(id))).collect();
+    let mut o = o.clone();
+    for story in o.text_box_slots() {
+        *story = None;
+    }
+    Key::Obj(Box::new(o), texts)
+}
+
+/// Whether two paragraphs hold the same objects, as compared.
+fn same_objects(doc: &Document, a: &Paragraph, b: &Paragraph) -> bool {
+    a.objects.len() == b.objects.len() && a.objects.iter().zip(&b.objects).all(|(x, y)| x == y || obj_key(doc, x) == obj_key(doc, y))
+}
+
+/// Words with the spaces after them ("The ", "cat "), or single characters, and the objects
+/// between them. Objects with text (fields, equations…) count as their text.
+fn tokens(p: &Paragraph, characters: bool) -> Vec<Tok<'_>> {
+    fn push(v: &mut Vec<Tok<'_>>, cur: &mut String, c: char, characters: bool) {
+        cur.push(c);
+        if characters || c == ' ' {
+            v.push(Tok::Text(std::mem::take(cur)));
+        }
    }
    let mut v = Vec::new();
    let mut cur = String::new();
-    for c in s.chars() {
-        cur.push(c);
-        if c == ' ' {
-            v.push(std::mem::take(&mut cur));
+    let mut objects = p.objects.iter();
+    for c in p.text.chars() {
+        if c != OBJ {
+            push(&mut v, &mut cur, c, characters);
+            continue;
+        }
+        match objects.next() {
+            Some(o) if o.plain_text().is_empty() => {
+                if !cur.is_empty() {
+                    v.push(Tok::Text(std::mem::take(&mut cur)));
+                }
+                v.push(Tok::Obj(o));
+            }
+            Some(o) => o.plain_text().chars().for_each(|c| push(&mut v, &mut cur, c, characters)),
+            None => {
+                // A marker without its object (a broken paragraph): kept as a visible
+                // placeholder, so the loss shows in the result instead of vanishing.
+                log::warn!("compare: an object marker without its object");
+                push(&mut v, &mut cur, char::REPLACEMENT_CHARACTER, characters);
+            }
        }
    }
    if !cur.is_empty() {
-        v.push(cur);
+        v.push(Tok::Text(cur));
    }
    v
 }

-fn word_diff(a: &Paragraph, b: &Paragraph, ins: u32, del: u32, opts: &CompareOptions, budget: &mut u64) -> Paragraph {
+/// Add a token at the end of `p` with `props`.
+fn put(p: &mut Paragraph, t: Option<&Tok<'_>>, props: &CharProps) {
+    let at = p.len();
+    match t {
+        Some(Tok::Text(t)) => {
+            let _ = p.insert_text(at, t, props);
+        }
+        Some(Tok::Obj(o)) => {
+            let _ = p.insert_object(at, (*o).clone(), props);
+        }
+        None => {}
+    }
+}
+
+fn word_diff(doc: &Document, a: &Paragraph, b: &Paragraph, ins: u32, del: u32, opts: &CompareOptions, budget: &mut u64) -> Paragraph {
    let (at, bt) = (a.plain_text(), b.plain_text());
    let chars = opts.characters && at.chars().count().max(bt.chars().count()) <= MAX_CHAR_DIFF;
-    let (ta, tb) = (tokens(&at, chars), tokens(&bt, chars));
-    let ops = lcs_ops(&ta.iter().map(|t| opts.key(t)).collect::<Vec<_>>(), &tb.iter().map(|t| opts.key(t)).collect::<Vec<_>>(), budget);
+    let (ta, tb) = (tokens(a, chars), tokens(b, chars));
+    let key = |t: &Tok<'_>| match t {
+        Tok::Text(t) => Key::Text(opts.key(t)),
+        Tok::Obj(o) => obj_key(doc, o),
+    };
+    let ops = lcs_ops(&ta.iter().map(key).collect::<Vec<_>>(), &tb.iter().map(key).collect::<Vec<_>>(), budget);
    // Only at character granularity: a matched whole word almost never happens by coincidence,
    // but a matched single character very often does (digits, punctuation), so only there does
    // a short interior match need folding back into its surrounding edit.
@@ -341,20 +428,19 @@ fn word_diff(a: &Paragraph, b: &Paragraph, ins: u32, del: u32, opts: &CompareOpt
    let base = a.runs.first().map(|r| r.props.clone()).unwrap_or_default();
    let (mut i, mut j) = (0, 0);
    for op in ops {
-        let at = p.len();
        match op {
            Op::Same => {
                // Equal as compared (perhaps up to case or spacing): the revised text.
-                let _ = p.insert_text(at, tb.get(j).map(String::as_str).unwrap_or(""), &base);
+                put(&mut p, tb.get(j), &base);
                i += 1;
                j += 1;
            }
            Op::Del => {
-                let _ = p.insert_text(at, ta.get(i).map(String::as_str).unwrap_or(""), &CharProps { del: Some(del), ..base.clone() });
+                put(&mut p, ta.get(i), &CharProps { del: Some(del), ..base.clone() });
                i += 1;
            }
            Op::Ins => {
-                let _ = p.insert_text(at, tb.get(j).map(String::as_str).unwrap_or(""), &CharProps { ins: Some(ins), ..base.clone() });
+                put(&mut p, tb.get(j), &CharProps { ins: Some(ins), ..base.clone() });
                j += 1;
            }
        }
@@ -80,6 +80,86 @@ fn compare_at_word_level_still_matches_a_whole_unchanged_word() {
    );
 }

+#[test]
+fn compare_keeps_text_boxes_of_the_original() {
+    // Text boxes are anchored in body paragraphs: one after changed text, one alone in an
+    // otherwise empty paragraph, one after text the revised document keeps. Their anchors must
+    // survive the compare (marked deleted, since the revised text has no boxes), so no text is lost.
+    for show_in in ["original", "revised", "new"] {
+        let mut s = Session::new(Document::from_text("Body one.\n\nKeep."));
+        let mut boxes = Vec::new();
+        for (para, off, text) in [(0usize, "Body one.".len(), "Alpha boxword gamma."), (1, 0, "Vml onlyword here."), (2, "Keep.".len(), "Kept box.")]
+        {
+            s.run("caret.set", &json!({"pos": Pos::body(para, off)})).unwrap();
+            let id = s.run("insert.textBox", &json!({"text": text})).unwrap()["story"].as_u64().unwrap() as u32;
+            boxes.push((id, text));
+        }
+        s.run("caret.set", &json!({"pos": Pos::body(0, 0)})).unwrap();
+        s.run("review.compare", &json!({"text": "Totally new first.\nSecond new line.\nKeep.", "showIn": show_in})).unwrap();
+        for (id, text) in &boxes {
+            let anchor = s.doc.text_box_anchor(*id).unwrap_or_else(|| panic!("{show_in}: the box \"{text}\" lost its anchor"));
+            assert_eq!(s.doc.plain_text(StoryRef::Part(*id)), *text);
+            let p = s.doc.para(anchor.story, &anchor.path).unwrap();
+            assert!(
+                p.props_of_char(anchor.off - wordcraft_doc::para::OBJ.len_utf8()).del.is_some(),
+                "{show_in}: the box \"{text}\" is not marked deleted"
+            );
+        }
+        s.run("review.rejectAll", &json!({})).unwrap();
+        assert_eq!(s.doc.plain_text(StoryRef::Body), "Body one.\n\nKeep.");
+        for (id, text) in &boxes {
+            assert!(s.doc.text_box_anchor(*id).is_some(), "{show_in}: rejecting lost the box \"{text}\"");
+            assert_eq!(s.doc.plain_text(StoryRef::Part(*id)), *text);
+        }
+    }
+}
+
+#[test]
+fn compare_with_a_revised_file_keeps_its_text_boxes_backed() {
+    // Text boxes in both documents, the revised one from a file: every anchor in the result
+    // shows a text box story of the result, and both documents' boxes are there, marked.
+    fn with_boxes(text: &str, boxes: &[(usize, usize, &str)]) -> Session {
+        let mut s = Session::new(Document::from_text(text));
+        for (para, off, text) in boxes {
+            s.run("caret.set", &json!({"pos": Pos::body(*para, *off)})).unwrap();
+            s.run("insert.textBox", &json!({"text": text})).unwrap();
+        }
+        s
+    }
+    /// Each text box anchored in the body: its story's text, and whether its anchor is deleted
+    /// or inserted.
+    fn boxes(d: &Document) -> Vec<(String, bool, bool)> {
+        let mut out = Vec::new();
+        for path in d.para_paths(StoryRef::Body) {
+            let p = d.para(StoryRef::Body, &path).unwrap();
+            for (off, o) in p.objects_with_offsets() {
+                for id in o.text_boxes() {
+                    let part = d.parts.get(&id).unwrap_or_else(|| panic!("an anchor shows story {id}, which the result lacks"));
+                    assert_eq!(part.kind, wordcraft_doc::PartKind::TextBox, "story {id}");
+                    let c = p.props_of_char(off);
+                    out.push((d.plain_text(StoryRef::Part(id)), c.del.is_some(), c.ins.is_some()));
+                }
+            }
+        }
+        out.sort();
+        out
+    }
+    let rev = with_boxes("Body one.\nNew middle.\nKeep.", &[(1, "New middle.".len(), "Revised box."), (2, "Keep.".len(), "Kept box.")]);
+    let path = std::env::temp_dir().join(format!("wc-compare-boxes-{}.docx", std::process::id()));
+    crate::io::save_path(&path, &rev.doc).unwrap();
+    let t = |t: &str, del, ins| (t.to_string(), del, ins);
+    for show_in in ["original", "revised", "new"] {
+        for (then, left) in [("review.acceptAll", "Revised box."), ("review.rejectAll", "Original box.")] {
+            let mut s = with_boxes("Body one.\nOld middle.\nKeep.", &[(1, "Old middle.".len(), "Original box."), (2, "Keep.".len(), "Kept box.")]);
+            s.run("review.compare", &json!({"path": path.to_string_lossy(), "showIn": show_in})).unwrap();
+            assert_eq!(boxes(&s.doc), [t("Kept box.", false, false), t("Original box.", true, false), t("Revised box.", false, true)], "{show_in}");
+            s.run(then, &json!({})).unwrap();
+            assert_eq!(boxes(&s.doc), [t("Kept box.", false, false), t(left, false, false)], "{show_in} {then}");
+        }
+    }
+    let _ = std::fs::remove_file(&path);
+}
+
 #[test]
 fn compare_bounds_total_diff_work() {
    // Many long, slightly different paragraphs: each pair alone fits one comparison table,
@@ -18,6 +18,8 @@ pub const OPEN_EXTS: &[&str] = &[
    "markdown",
    "html",
    "htm",
+    "mht",
+    "mhtml",
    "rtf",
    "odt",
    "tex",
@@ -201,7 +203,7 @@ fn load_embedded_fonts(doc: &Document) {
 fn read_bytes(name: &str, bytes: &[u8], password: Option<&str>, page: &wordcraft_doc::SectionProps) -> Result<(Document, bool), String> {
    let ext = ext_of(name);
    if matches!(ext.as_str(), "docx" | "docm" | "dotx" | "dotm" | "doc" | "dot") && wordcraft_docx::is_encrypted(bytes) {
-        let mut doc = wordcraft_docx::read_with_password(bytes, password).map_err(|e| e.to_string())?;
+        let mut doc = read_docx(bytes, password).map_err(|e| e.to_string())?;
        opened(&mut doc);
        return Ok((doc, true));
    }
@@ -230,6 +232,42 @@ fn read_bytes(name: &str, bytes: &[u8], password: Option<&str>, page: &wordcraft
    Ok((doc, false))
 }

+/// Read a Word package (decrypted with `password` if it is encrypted), bringing in its
+/// alternative format chunks (`w:altChunk`) as Word does when it opens one: see [`import_chunk`].
+pub(crate) fn read_docx(bytes: &[u8], password: Option<&str>) -> Result<Document, wordcraft_docx::DocxError> {
+    wordcraft_docx::read_with(bytes, &wordcraft_docx::ReadOptions { password, chunks: Some(&import_chunk) })
+}
+
+/// An alternative format chunk in a format the docx reader leaves to the format importers, as a
+/// document: an MHT web archive, HTML, RTF or plain text, told by its content type, its part's
+/// extension or what it holds. Its pictures fit the text width, as when the file is opened.
+fn import_chunk(c: &wordcraft_docx::AltChunk) -> Option<Document> {
+    use wordcraft_formats::{html, mht, rtf, txt};
+    let full = c.content_type.to_ascii_lowercase();
+    let ct = full.split(';').next().unwrap_or("").trim();
+    let charset = full.split(';').skip(1).find_map(|p| p.trim().strip_prefix("charset=")).map(|v| v.trim_matches('"'));
+    let ext = ext_of(c.part);
+    let b = c.bytes;
+    let mut doc = if matches!(ct, "message/rfc822" | "application/x-mimearchive")
+        || ct.starts_with("multipart/")
+        || matches!(ext.as_str(), "mht" | "mhtml")
+        || mht::is_mht(b)
+    {
+        mht::import(b)
+    } else if ct.contains("rtf") || ext == "rtf" || b.starts_with(b"{\\rtf") {
+        rtf::import(b).ok()?
+    } else if ct.contains("html") || matches!(ext.as_str(), "html" | "htm" | "xhtml") {
+        html::import_page(&html::decode_as(b, charset), &|_| None)
+    } else if ct == "text/plain" || ext == "txt" {
+        txt::import(b)
+    } else {
+        return None;
+    };
+    // shortcut: pictures fit the chunk's default page, not the section it lands in; fit them there if wide pictures in chunks overflow narrow sections.
+    fit_pictures(&mut doc);
+    Some(doc)
+}
+
 /// 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)
@@ -270,7 +308,7 @@ pub(crate) fn fit_pictures(doc: &mut Document) {
 /// Does the file hold a different document format than its Word or RTF extension says? An RTF
 /// file saved as `.doc` (as accounting systems and report generators often save their
 /// documents), a `.docx` package renamed `.doc`, a Word 97-2003 file renamed `.docx`, a Word XML
-/// Document (2003 or Flat OPC) or a web page saved as `.doc` (#2853): Word opens
+/// Document (2003 or Flat OPC) or a web page (HTML or MHT) saved as `.doc` (#2853): Word opens
 /// each by what it holds, so they go through [`open_by_content`] instead of failing.
 fn signature_contradicts(ext: &str, bytes: &[u8]) -> bool {
    let expected = match ext {
@@ -289,6 +327,8 @@ fn signature_contradicts(ext: &str, bytes: &[u8]) -> bool {
        "xml"
    } else if looks_like_web_page(bytes) {
        "html"
+    } else if wordcraft_formats::mht::is_mht(bytes) {
+        "mht"
    } else {
        return false;
    };
@@ -346,9 +386,9 @@ fn open_by_content(
    page: &wordcraft_doc::SectionProps,
 ) -> Result<(Document, bool), String> {
    let (mut doc, encrypted) = if bytes.starts_with(OLE_SIG) && wordcraft_docx::is_encrypted(bytes) {
-        (wordcraft_docx::read_with_password(bytes, password).map_err(|e| e.to_string())?, true)
+        (read_docx(bytes, password).map_err(|e| e.to_string())?, true)
    } else if bytes.starts_with(ZIP_SIG) {
-        (wordcraft_docx::read(bytes).map_err(|e| format!("{name}: not a document WordCraft can open ({e})"))?, false)
+        (read_docx(bytes, None).map_err(|e| format!("{name}: not a document WordCraft can open ({e})"))?, false)
    } else if bytes.starts_with(OLE_SIG) {
        open_binary_doc(bytes, password)
            .map_err(|e| if needs_password(&e) || wrong_password(&e) { e } else { format!("{name}: not a document WordCraft can open ({e})") })?
@@ -359,6 +399,9 @@ fn open_by_content(
    } else if matches!(ext, "doc" | "dot" | "docx" | "docm" | "dotx" | "dotm" | "rtf") && looks_like_web_page(bytes) {
        let opened = crate::io_ext::open("html", bytes, page).unwrap_or_else(|| Err(format!("{name}: can't open web pages")));
        (opened?, false)
+    } else if matches!(ext, "doc" | "dot" | "docx" | "docm" | "dotx" | "dotm" | "rtf") && wordcraft_formats::mht::is_mht(bytes) {
+        let opened = crate::io_ext::open("mht", bytes, page).unwrap_or_else(|| Err(format!("{name}: can't open web archives")));
+        (opened?, false)
    } else if wordcraft_formats::txt::is_text(bytes) {
        (wordcraft_formats::txt::import_on(bytes, page), false)
    } else if ext.is_empty() {
@@ -6,7 +6,7 @@ use wordcraft_doc::Document;
 /// start on `page`.
 pub fn open(ext: &str, bytes: &[u8], page: &wordcraft_doc::SectionProps) -> Option<Result<Document, String>> {
    match ext {
-        "docx" | "docm" | "dotx" | "dotm" => Some(wordcraft_docx::read(bytes).map_err(|e| e.to_string())),
+        "docx" | "docm" | "dotx" | "dotm" => Some(crate::io::read_docx(bytes, None).map_err(|e| e.to_string())),
        "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())),
@@ -16,6 +16,7 @@ mod io_ext;
 pub mod keymap;
 pub(crate) mod locale;
 pub mod math_gallery;
+pub mod packs;
 pub(crate) mod prefs;
 pub mod print;
 pub mod recovery;
@@ -0,0 +1,730 @@
+//! Language packs (CP4.3): spelling dictionaries and hyphenation patterns for languages beyond
+//! the embedded English, one pack per locale, downloaded and verified by the font store
+//! (craft-libs `craft-font-store`, item kind `language-pack`) and read by `craft-proof` from the
+//! store's `language-packs/` folder.
+//!
+//! The commands (`proof.packs`, `proof.installPack`, `proof.removePack`), Options › Language,
+//! the missing-pack bar and About › Language data all go through this module. It never asks
+//! anything: an install needs the font store's single consent setting (downloading on), which only
+//! a person in the app can turn on ([`Backend::enable_network`], called from an Install click).
+//! Agents (MCP, the CLI, the control channel) install only while it is on.
+//!
+//! The [`Backend`] is set once at start-up by the desktop app and the CLI ([`use_font_store`]).
+//! Without one (tests, the web build, `ui_shot`) nothing is available and nothing is downloaded.
+
+use std::collections::BTreeSet;
+use std::path::PathBuf;
+use std::sync::atomic::{AtomicBool, Ordering};
+use std::sync::{Arc, Mutex, PoisonError, RwLock};
+
+use serde::Serialize;
+
+use crate::{CmdError, Session};
+
+#[cfg(not(target_arch = "wasm32"))]
+mod store;
+
+/// Longest locale tag taken, in bytes (BCP 47 tags in practice stay under 20).
+pub const MAX_LOCALE_LEN: usize = 35;
+/// Most packs an index may list that are read.
+const MAX_PACKS: usize = 1000;
+/// Paragraphs of a document looked at for its languages (the missing-pack bar).
+const MAX_PARAGRAPHS_SCANNED: usize = 20_000;
+/// Distinct languages of a document collected.
+const MAX_DOC_LANGUAGES: usize = 64;
+
+/// A pack the pinned index lists.
+#[derive(Clone, Debug, PartialEq, Eq, Serialize)]
+pub struct Pack {
+    /// Lowercase BCP 47 (`nl`, `pt-pt`).
+    pub locale: String,
+    /// The language's own name (`Nederlands`).
+    pub name: String,
+    /// The craft-i18n commit the pack was built from.
+    pub version: String,
+    /// The archive.
+    pub url: String,
+    /// The upstream folder the data comes from, at its pinned commit (the index's optional
+    /// `source`), shown before install; the archive when the index gives none.
+    pub source: Option<String>,
+    /// Bytes of the archive.
+    pub size: u64,
+    /// SPDX expression.
+    pub licence: String,
+    /// `Embeddable` or `Download-only`.
+    pub class: String,
+    /// Has a spelling dictionary.
+    pub spelling: bool,
+    /// Has hyphenation patterns.
+    pub hyphenation: bool,
+}
+
+/// An installed pack.
+#[derive(Clone, Debug, PartialEq, Eq, Serialize)]
+pub struct Installed {
+    pub locale: String,
+    pub name: String,
+    pub licence: String,
+    pub path: PathBuf,
+}
+
+/// Stops a download. On the desktop it is the font store's own flag, which a download watches
+/// even while the server sends nothing.
+#[derive(Clone, Debug, Default)]
+pub struct Cancel(#[cfg(not(target_arch = "wasm32"))] pub(crate) craft_font_store::Cancel, #[cfg(target_arch = "wasm32")] Arc<AtomicBool>);
+
+impl Cancel {
+    pub fn new() -> Self {
+        Self::default()
+    }
+    #[cfg(not(target_arch = "wasm32"))]
+    pub fn cancel(&self) {
+        self.0.cancel();
+    }
+    #[cfg(not(target_arch = "wasm32"))]
+    pub fn is_cancelled(&self) -> bool {
+        self.0.is_cancelled()
+    }
+    #[cfg(target_arch = "wasm32")]
+    pub fn cancel(&self) {
+        self.0.store(true, Ordering::Relaxed);
+    }
+    #[cfg(target_arch = "wasm32")]
+    pub fn is_cancelled(&self) -> bool {
+        self.0.load(Ordering::Relaxed)
+    }
+}
+
+/// Where packs come from: the font store on the desktop, a fixed list in tests and `ui_shot`.
+pub trait Backend: Send + Sync {
+    /// The pinned index as last fetched (no request); empty when it never was.
+    fn index(&self) -> Result<Vec<Pack>, String>;
+    /// Fetch the pinned index (a request; only while downloading is on).
+    fn fetch_index(&self, cancel: &Cancel) -> Result<Vec<Pack>, String>;
+    /// Whether the font store may download.
+    fn network_on(&self) -> bool;
+    /// Turn downloading on. Only from the person's own action in the app (an Install click):
+    /// the commands never call it, and a backend for an agent refuses.
+    fn enable_network(&self) -> Result<(), String>;
+    /// Download, verify and unpack `pack` into the packs folder. `progress` gets (done, total) bytes.
+    fn install(&self, pack: &Pack, cancel: &Cancel, progress: &mut dyn FnMut(u64, u64)) -> Result<(), String>;
+    /// Remove the installed pack for `locale`.
+    fn remove(&self, locale: &str) -> Result<(), String>;
+    /// The installed-packs folder (`<font store>/language-packs`).
+    fn packs_dir(&self) -> Option<PathBuf>;
+    /// The packs installed now.
+    fn installed(&self) -> Vec<Installed> {
+        installed_from_proof()
+    }
+    /// The test index in use instead of the pinned one (`CRAFT_LANGUAGE_PACK_INDEX`), if any.
+    fn index_override(&self) -> Option<String> {
+        None
+    }
+}
+
+static BACKEND: RwLock<Option<Arc<dyn Backend>>> = RwLock::new(None);
+/// The index last read, so the bar and the list don't read the store's cache every frame.
+static INDEX: Mutex<Option<Vec<Pack>>> = Mutex::new(None);
+/// Bumped whenever [`INDEX`] changes.
+static INDEX_GENERATION: std::sync::atomic::AtomicU64 = std::sync::atomic::AtomicU64::new(0);
+
+fn set_index(index: Option<Vec<Pack>>) {
+    *INDEX.lock().unwrap_or_else(PoisonError::into_inner) = index;
+    INDEX_GENERATION.fetch_add(1, Ordering::Relaxed);
+}
+
+/// Changes whenever the list of available packs does (a fetch, another backend).
+pub fn index_generation() -> u64 {
+    INDEX_GENERATION.load(Ordering::Relaxed)
+}
+
+/// Use `backend` for packs from now on (`None`: no packs), and point proofing at its folder.
+pub fn set_backend(backend: Option<Arc<dyn Backend>>) {
+    let dir = backend.as_ref().and_then(|b| b.packs_dir());
+    *BACKEND.write().unwrap_or_else(PoisonError::into_inner) = backend;
+    set_index(None);
+    jobs().errors.clear();
+    wordcraft_proof::packs::set_packs_dir(dir);
+}
+
+/// Use the suite's font store for packs (`<font store>/language-packs`), as the person in the app
+/// (`user`: an Install click may turn downloading on) or as an agent (the CLI, headless MCP: it
+/// never can). The pack index is the pinned one, or `CRAFT_LANGUAGE_PACK_INDEX` (a URL or a path,
+/// for testing packs before they are published), which the font store reads itself. Returns
+/// whether there is a store folder.
+#[cfg(not(target_arch = "wasm32"))]
+pub fn use_font_store(user: bool) -> bool {
+    match store::FontStoreBackend::new(user) {
+        Some(b) => {
+            set_backend(Some(Arc::new(b)));
+            true
+        }
+        None => false,
+    }
+}
+
+/// [`use_font_store`] with the store in `dir` instead of the person's (`ui_shot`, tests).
+#[cfg(not(target_arch = "wasm32"))]
+pub fn use_font_store_in(dir: PathBuf, user: bool) {
+    set_backend(Some(Arc::new(store::FontStoreBackend::in_dir(dir, user))));
+}
+
+/// The backend in use.
+pub fn backend() -> Option<Arc<dyn Backend>> {
+    BACKEND.read().unwrap_or_else(PoisonError::into_inner).clone()
+}
+
+fn installed_from_proof() -> Vec<Installed> {
+    wordcraft_proof::packs::installed().into_iter().map(|p| Installed { locale: p.locale, name: p.name, licence: p.licence, path: p.path }).collect()
+}
+
+/// The test source packs come from instead of the pinned index (`CRAFT_LANGUAGE_PACK_INDEX`):
+/// Options › Language says so.
+pub fn index_override() -> Option<String> {
+    backend().and_then(|b| b.index_override())
+}
+
+/// The packs installed now (none without a backend).
+pub fn installed() -> Vec<Installed> {
+    backend().map(|b| b.installed()).unwrap_or_default()
+}
+
+/// The available packs: the index as last read, read from the store's cache the first time.
+pub fn available() -> Vec<Pack> {
+    let Some(b) = backend() else { return Vec::new() };
+    let mut cached = INDEX.lock().unwrap_or_else(PoisonError::into_inner);
+    if let Some(i) = cached.as_ref() {
+        return i.clone();
+    }
+    let index = clean_index(b.index().unwrap_or_default());
+    *cached = Some(index.clone());
+    drop(cached);
+    INDEX_GENERATION.fetch_add(1, Ordering::Relaxed);
+    index
+}
+
+/// How many packs [`available`] lists, without copying them (the window asks every frame).
+pub fn available_len() -> usize {
+    if let Some(i) = INDEX.lock().unwrap_or_else(PoisonError::into_inner).as_ref() {
+        return i.len();
+    }
+    available().len()
+}
+
+/// Fetch the index again (downloading must be on).
+pub fn refresh(cancel: &Cancel) -> Result<Vec<Pack>, String> {
+    let b = backend().ok_or_else(no_store)?;
+    if !b.network_on() {
+        return Err(network_off());
+    }
+    let index = clean_index(b.fetch_index(cancel)?);
+    set_index(Some(index.clone()));
+    Ok(index)
+}
+
+/// Drop packs with a malformed locale, duplicates and English (embedded), and cap the list.
+fn clean_index(mut packs: Vec<Pack>) -> Vec<Pack> {
+    packs.truncate(MAX_PACKS);
+    let mut seen = BTreeSet::new();
+    packs.retain(|p| valid_locale(&p.locale) && !is_embedded(&p.locale) && seen.insert(p.locale.clone()));
+    packs.sort_by(|a, b| a.locale.cmp(&b.locale));
+    packs
+}
+
+/// English is embedded in every build; it never comes as a pack.
+fn is_embedded(locale: &str) -> bool {
+    locale == "en" || locale.starts_with("en-")
+}
+
+/// A lowercase BCP 47-shaped tag: letters, digits and `-`, starting with a letter.
+pub fn valid_locale(locale: &str) -> bool {
+    !locale.is_empty()
+        && locale.len() <= MAX_LOCALE_LEN
+        && locale.starts_with(|c: char| c.is_ascii_lowercase())
+        && locale.chars().all(|c| c.is_ascii_lowercase() || c.is_ascii_digit() || c == '-')
+        && !locale.contains("--")
+        && !locale.ends_with('-')
+}
+
+/// A locale from a command's params: trimmed, lowercased, `_` as `-`.
+pub fn parse_locale(raw: &str) -> Result<String, CmdError> {
+    let l = raw.trim().replace('_', "-").to_ascii_lowercase();
+    if valid_locale(&l) { Ok(l) } else { Err(CmdError::Params(format!("`locale` must be a language tag like nl or pt-PT, not {raw:?}"))) }
+}
+
+fn no_store() -> String {
+    if cfg!(target_arch = "wasm32") {
+        "Language packs aren't available in the browser".into()
+    } else {
+        "there is no font store folder for language packs".into()
+    }
+}
+
+fn network_off() -> String {
+    "downloading is turned off: turn it on in WordCraft (File › Options › Language, or an Install button there)".into()
+}
+
+/// The pack for `locale` in the index.
+pub fn find(locale: &str) -> Option<Pack> {
+    available().into_iter().find(|p| p.locale == locale)
+}
+
+/// Whether a pack is installed for `locale`.
+pub fn is_installed(locale: &str) -> bool {
+    installed().iter().any(|p| p.locale == locale)
+}
+
+/// Install the pack for `locale` (downloading must be on): the commands and the Install button.
+/// Proofing picks it up at once.
+pub fn install(locale: &str, cancel: &Cancel, progress: &mut dyn FnMut(u64, u64)) -> Result<Installed, String> {
+    let b = backend().ok_or_else(no_store)?;
+    let pack = find(locale).ok_or_else(|| format!("there is no language pack for {locale}"))?;
+    if !b.network_on() {
+        return Err(network_off());
+    }
+    if let Some(dir) = b.packs_dir() {
+        crate::file_access::check_write(&dir.to_string_lossy())?;
+    }
+    b.install(&pack, cancel, progress)?;
+    reload(&b);
+    b.installed().into_iter().find(|p| p.locale == locale).ok_or_else(|| format!("the {locale} pack didn't install"))
+}
+
+/// Remove the installed pack for `locale`.
+pub fn remove(locale: &str) -> Result<(), String> {
+    let b = backend().ok_or_else(no_store)?;
+    if jobs().installs.iter().any(|(l, ..)| l == locale) {
+        return Err(format!("the {locale} pack is being installed: cancel it first"));
+    }
+    if !b.installed().iter().any(|p| p.locale == locale) {
+        return Err(format!("no language pack is installed for {locale}"));
+    }
+    if let Some(dir) = b.packs_dir() {
+        crate::file_access::check_write(&dir.to_string_lossy())?;
+    }
+    b.remove(locale)?;
+    reload(&b);
+    Ok(())
+}
+
+/// What the app does with a long pack request (an install or a fetch of the list): `None` runs it
+/// in the caller (the CLI, headless MCP, tests); `Some(notify)` runs it on a thread, so the window
+/// keeps drawing whoever asked (the person, the control channel, MCP), and calls `notify` on each
+/// change (the window repaints). Set once by the desktop app.
+pub type Notify = Arc<dyn Fn() + Send + Sync>;
+
+static BACKGROUND: RwLock<Option<Notify>> = RwLock::new(None);
+
+/// Installs and fetches running on a thread, and how the finished ones ended.
+#[derive(Default)]
+struct Jobs {
+    /// Running installs by locale: (bytes done, bytes in all, its Cancel).
+    installs: Vec<(String, u64, u64, Cancel)>,
+    fetching: bool,
+    /// The last failure of each locale's install ("" for the list), until it is tried again.
+    errors: Vec<(String, String)>,
+    /// Bumped as each job ends: the window relays out and shows the result.
+    finished: u64,
+}
+
+static JOBS: Mutex<Jobs> = Mutex::new(Jobs { installs: Vec::new(), fetching: false, errors: Vec::new(), finished: 0 });
+
+fn jobs() -> std::sync::MutexGuard<'static, Jobs> {
+    JOBS.lock().unwrap_or_else(PoisonError::into_inner)
+}
+
+/// Run long pack requests on a thread (`Some`: the desktop app), or in the caller (`None`).
+pub fn set_background(notify: Option<Notify>) {
+    *BACKGROUND.write().unwrap_or_else(PoisonError::into_inner) = notify;
+}
+
+fn background() -> Option<Notify> {
+    if cfg!(target_arch = "wasm32") {
+        return None;
+    }
+    BACKGROUND.read().unwrap_or_else(PoisonError::into_inner).clone()
+}
+
+/// Whether installs and fetches run on a thread ([`set_background`]).
+pub fn runs_in_background() -> bool {
+    background().is_some()
+}
+
+/// A running install: (locale, bytes done, bytes in all).
+#[derive(Clone, Debug, PartialEq, Eq, Serialize)]
+pub struct Progress {
+    pub locale: String,
+    pub done: u64,
+    pub total: u64,
+}
+
+/// The installs running now.
+pub fn installing() -> Vec<Progress> {
+    jobs().installs.iter().map(|(l, d, t, _)| Progress { locale: l.clone(), done: *d, total: *t }).collect()
+}
+
+/// Whether the list is being fetched.
+pub fn fetching() -> bool {
+    jobs().fetching
+}
+
+/// The last failure of `locale`'s install ("" for the list), until it is tried again.
+pub fn error(locale: &str) -> Option<String> {
+    jobs().errors.iter().find(|(l, _)| l == locale).map(|(_, e)| e.clone())
+}
+
+/// Every recorded failure, (locale or "" for the list, message).
+pub fn errors() -> Vec<(String, String)> {
+    jobs().errors.clone()
+}
+
+/// Counts finished jobs: when it changes, the window lays the documents out again.
+pub fn finished() -> u64 {
+    jobs().finished
+}
+
+/// Stop `locale`'s running install.
+pub fn cancel(locale: &str) {
+    if let Some((_, _, _, c)) = jobs().installs.iter().find(|(l, ..)| l == locale) {
+        c.cancel();
+    }
+}
+
+fn set_error(locale: &str, e: Option<String>) {
+    let mut j = jobs();
+    j.errors.retain(|(l, _)| l != locale);
+    if let Some(e) = e {
+        j.errors.push((locale.to_string(), e));
+    }
+}
+
+/// Start installing `locale` ([`install`]): on a thread when [`runs_in_background`], else now.
+/// What's wrong up front (no pack, downloading off, outside the write root) is an error at once.
+/// Returns whether it is still running (`false`: done, or it already was running).
+pub fn start_install(locale: &str) -> Result<bool, String> {
+    let b = backend().ok_or_else(no_store)?;
+    let pack = find(locale).ok_or_else(|| format!("there is no language pack for {locale}"))?;
+    if !b.network_on() {
+        return Err(network_off());
+    }
+    if let Some(dir) = b.packs_dir() {
+        crate::file_access::check_write(&dir.to_string_lossy())?;
+    }
+    let Some(notify) = background() else {
+        let r = install(locale, &Cancel::new(), &mut |_, _| {});
+        set_error(locale, r.as_ref().err().cloned());
+        return r.map(|_| false);
+    };
+    let cancel = Cancel::new();
+    {
+        let mut j = jobs();
+        if j.installs.iter().any(|(l, ..)| l == locale) {
+            return Ok(true);
+        }
+        j.installs.push((locale.to_string(), 0, pack.size, cancel.clone()));
+        j.errors.retain(|(l, _)| l != locale);
+    }
+    let key = locale.to_string();
+    let locale = locale.to_string();
+    let started = spawn(move || {
+        let progress_notify = notify.clone();
+        let r = install(&locale, &cancel, &mut |done, total| {
+            if let Some(job) = jobs().installs.iter_mut().find(|(l, ..)| *l == locale) {
+                job.1 = done;
+                job.2 = total;
+            }
+            progress_notify();
+        });
+        let mut j = jobs();
+        j.installs.retain(|(l, ..)| *l != locale);
+        if let Err(e) = r {
+            j.errors.push((locale, e));
+        }
+        j.finished = j.finished.wrapping_add(1);
+        drop(j);
+        notify();
+    });
+    if let Err(e) = started {
+        jobs().installs.retain(|(l, ..)| *l != key);
+        return Err(e);
+    }
+    Ok(true)
+}
+
+/// Start fetching the list again ([`refresh`]): on a thread when [`runs_in_background`], else now.
+/// Returns whether it is still running.
+pub fn start_refresh() -> Result<bool, String> {
+    let b = backend().ok_or_else(no_store)?;
+    if !b.network_on() {
+        return Err(network_off());
+    }
+    let Some(notify) = background() else {
+        let r = refresh(&Cancel::new());
+        set_error("", r.as_ref().err().cloned());
+        return r.map(|_| false);
+    };
+    {
+        let mut j = jobs();
+        if j.fetching {
+            return Ok(true);
+        }
+        j.fetching = true;
+        j.errors.retain(|(l, _)| !l.is_empty());
+    }
+    let started = spawn(move || {
+        let r = refresh(&Cancel::new());
+        let mut j = jobs();
+        j.fetching = false;
+        if let Err(e) = r {
+            j.errors.push((String::new(), e));
+        }
+        j.finished = j.finished.wrapping_add(1);
+        drop(j);
+        notify();
+    });
+    if let Err(e) = started {
+        jobs().fetching = false;
+        return Err(e);
+    }
+    Ok(true)
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+fn spawn(f: impl FnOnce() + Send + 'static) -> Result<(), String> {
+    std::thread::Builder::new().name("language-pack".into()).spawn(f).map(|_| ()).map_err(|e| format!("couldn't start the download: {e}"))
+}
+
+#[cfg(target_arch = "wasm32")]
+fn spawn(_: impl FnOnce() + Send + 'static) -> Result<(), String> {
+    Err(no_store())
+}
+
+/// Have proofing look at the packs folder again (its dictionaries are reloaded).
+fn reload(b: &Arc<dyn Backend>) {
+    if wordcraft_proof::packs::packs_dir() == b.packs_dir() {
+        wordcraft_proof::packs::reload();
+    } else {
+        wordcraft_proof::packs::set_packs_dir(b.packs_dir());
+    }
+}
+
+/// The full licence texts for About › Language data: the embedded English data, then each
+/// installed pack.
+pub fn licence_texts() -> Vec<(String, String)> {
+    let mut texts = wordcraft_proof::packs::licence_texts();
+    // SCOWL's notices (with the Ispell BSD licence of its affix files), when English spells from
+    // the embedded SCOWL data; after the Moby and supplement entries.
+    if let Some(notice) = wordcraft_proof::scowl::notice() {
+        let at = texts.len().min(2);
+        texts.insert(at, (SCOWL_TITLE.into(), notice.into()));
+    }
+    texts
+}
+
+/// About › Language data's title for SCOWL's notices.
+pub const SCOWL_TITLE: &str = "English spelling dictionaries (SCOWL)";
+
+/// The pack locale for a document language tag: the whole tag (`pt-pt`), else its language
+/// (`nl` for `nl-NL`), whichever the index has. `None` for English and tags with no pack.
+pub fn pack_for_tag<'a>(tag: &str, packs: &'a [Pack]) -> Option<&'a Pack> {
+    let tag = tag.trim().replace('_', "-").to_ascii_lowercase();
+    let base = tag.split('-').next().unwrap_or("");
+    if base.is_empty() || base == "en" {
+        return None;
+    }
+    packs.iter().find(|p| p.locale == tag).or_else(|| packs.iter().find(|p| p.locale == base))
+}
+
+/// The languages (`w:lang`, resolved through the styles) of the body text, in first-seen order.
+pub fn document_languages(s: &Session) -> Vec<String> {
+    let mut out: Vec<String> = Vec::new();
+    for path in s.doc.para_paths(wordcraft_doc::StoryRef::Body).into_iter().take(MAX_PARAGRAPHS_SCANNED) {
+        let Some(p) = s.doc.para(wordcraft_doc::StoryRef::Body, &path) else { continue };
+        if p.text.trim().is_empty() {
+            continue;
+        }
+        let style = p.props.style.as_deref();
+        for (range, c) in p.run_ranges() {
+            let Some(text) = p.text.get(range) else { continue };
+            if !text.chars().any(char::is_alphabetic) {
+                continue;
+            }
+            let rc = s.doc.styles.resolve_char_in(style, None, c);
+            if rc.no_proof {
+                continue;
+            }
+            let tag = rc.lang.unwrap_or_else(|| "en-US".into());
+            if !out.contains(&tag) {
+                out.push(tag);
+                if out.len() >= MAX_DOC_LANGUAGES {
+                    return out;
+                }
+            }
+        }
+    }
+    out
+}
+
+/// Whether proofing reads a pack for text in `tag`: every language craft-proof can name, its own
+/// (Polish…) and any other (`ProofLang::Pack`, read from the pack alone); not `und`, `zxx` or a
+/// malformed tag. The one place the bar's language rule lives.
+pub fn proofable(tag: &str) -> bool {
+    wordcraft_proof::ProofLang::from_tag(Some(tag)) != wordcraft_proof::ProofLang::Other
+}
+
+/// The Hunspell dictionary `tag`'s language needs and doesn't have, for the status bar's
+/// "(no dictionary)": craft-proof's [`wordcraft_proof::missing_dictionary`], except for a
+/// language craft-proof has no rules of its own for (`ProofLang::Pack`: Japanese, Chinese,
+/// Thai…) when no pack for it is available either, as no Hunspell dictionary may exist at all.
+pub fn missing_dictionary(tag: Option<&str>) -> Option<&'static str> {
+    let name = wordcraft_proof::missing_dictionary(tag)?;
+    let lang = wordcraft_proof::ProofLang::from_tag(tag);
+    if lang.pack_locale().is_some() && pack_for_tag(tag.unwrap_or(""), &available()).is_none() {
+        return None;
+    }
+    Some(name)
+}
+
+/// The packs the document's text asks for that are available and not installed: what the
+/// missing-pack bar offers. Never a language without a pack, nor one proofing can't use.
+pub fn missing(s: &Session) -> Vec<Pack> {
+    let packs = available();
+    if packs.is_empty() {
+        return Vec::new();
+    }
+    let installed = installed();
+    let mut out: Vec<Pack> = Vec::new();
+    for tag in document_languages(s) {
+        if proofable(&tag)
+            && let Some(p) = pack_for_tag(&tag, &packs)
+            && !installed.iter().any(|i| i.locale == p.locale)
+            && !out.iter().any(|o| o.locale == p.locale)
+        {
+            out.push(p.clone());
+        }
+    }
+    out
+}
+
+/// A backend with a fixed list and installs that only record themselves: tests and `ui_shot`.
+#[derive(Default)]
+pub struct FixedBackend {
+    pub packs: Vec<Pack>,
+    pub network: AtomicBool,
+    /// Agents can't turn the network on.
+    pub agent: bool,
+    pub installed: Mutex<Vec<Installed>>,
+}
+
+impl Pack {
+    /// Where the data comes from: the upstream source when the index names it, else the archive.
+    pub fn source_link(&self) -> &str {
+        self.source.as_deref().filter(|s| !s.is_empty()).unwrap_or(&self.url)
+    }
+}
+
+impl FixedBackend {
+    /// A few packs shaped like the published ones, for screenshots and tests.
+    pub fn sample() -> Self {
+        let pack = |locale: &str, name: &str, size: u64, licence: &str, class: &str| Pack {
+            locale: locale.into(),
+            name: name.into(),
+            version: "0000000".into(),
+            url: format!("https://github.com/storytold/craft-i18n/releases/download/language-packs-1/{locale}.tar.zst"),
+            source: Some("https://github.com/LibreOffice/dictionaries".into()),
+            size,
+            licence: licence.into(),
+            class: class.into(),
+            spelling: true,
+            hyphenation: true,
+        };
+        FixedBackend {
+            packs: vec![
+                pack("ca", "Català", 1_480_000, "GPL-2.0-or-later OR LGPL-2.1-or-later", "Download-only"),
+                pack("el", "Ελληνικά", 2_310_000, "MPL-1.1 OR GPL-2.0-or-later OR LGPL-2.1-or-later", "Embeddable"),
+                pack("fr", "Français", 1_120_000, "MPL-2.0", "Embeddable"),
+                pack("nl", "Nederlands", 1_950_000, "BSD-3-Clause", "Embeddable"),
+                pack("pl", "Polski", 4_620_000, "GPL-2.0-or-later OR LGPL-2.1-or-later OR MPL-1.1 OR Apache-2.0 OR CC-BY-4.0", "Embeddable"),
+                pack("sk", "Slovenčina", 2_040_000, "GPL-2.0-only OR LGPL-2.1-only OR MPL-1.1", "Download-only"),
+            ],
+            ..Default::default()
+        }
+    }
+}
+
+impl Backend for FixedBackend {
+    fn index(&self) -> Result<Vec<Pack>, String> {
+        Ok(self.packs.clone())
+    }
+    fn fetch_index(&self, _: &Cancel) -> Result<Vec<Pack>, String> {
+        Ok(self.packs.clone())
+    }
+    fn network_on(&self) -> bool {
+        self.network.load(Ordering::Relaxed)
+    }
+    fn enable_network(&self) -> Result<(), String> {
+        if self.agent {
+            return Err("only the user can turn on downloading".into());
+        }
+        self.network.store(true, Ordering::Relaxed);
+        Ok(())
+    }
+    fn install(&self, pack: &Pack, cancel: &Cancel, progress: &mut dyn FnMut(u64, u64)) -> Result<(), String> {
+        if cancel.is_cancelled() {
+            return Err("cancelled".into());
+        }
+        progress(pack.size, pack.size);
+        let mut i = self.installed.lock().unwrap_or_else(PoisonError::into_inner);
+        i.retain(|p| p.locale != pack.locale);
+        i.push(Installed { locale: pack.locale.clone(), name: pack.name.clone(), licence: pack.licence.clone(), path: PathBuf::new() });
+        Ok(())
+    }
+    fn remove(&self, locale: &str) -> Result<(), String> {
+        self.installed.lock().unwrap_or_else(PoisonError::into_inner).retain(|p| p.locale != locale);
+        Ok(())
+    }
+    fn packs_dir(&self) -> Option<PathBuf> {
+        None
+    }
+    fn installed(&self) -> Vec<Installed> {
+        self.installed.lock().unwrap_or_else(PoisonError::into_inner).clone()
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    /// CP4.4 (`cargo xtask language`): a build embeds English Language data and nothing it
+    /// wasn't asked to. A Language build names the locales it embeds in `CRAFT_PROOF_EMBED`
+    /// (`en` alone: English only); the default embeds the carry-over locales too, until Language
+    /// packs are published and pinned (craft-proof's `DEFAULT_EMBED`; TODO(CP4.4): English only
+    /// then).
+    /// About › Language data shows SCOWL's notices (the Ispell BSD licence among them) whenever
+    /// English spells from the embedded SCOWL data.
+    #[test]
+    fn about_shows_the_scowl_notice_when_scowl_is_embedded() {
+        let texts = super::licence_texts();
+        match wordcraft_proof::scowl::notice() {
+            Some(notice) => {
+                let entry = texts.iter().find(|(t, _)| t == super::SCOWL_TITLE).map(|(_, x)| x.as_str());
+                assert_eq!(entry, Some(notice));
+                assert!(notice.contains("Ispell") || notice.contains("ispell"), "the Ispell notice is part of it");
+            }
+            None => assert!(texts.iter().all(|(t, _)| t != super::SCOWL_TITLE), "no SCOWL data: no SCOWL entry"),
+        }
+    }
+
+    #[test]
+    fn the_default_build_embeds_only_english() {
+        const CARRY_OVER: [&str; 5] = ["de", "pl", "pt", "it", "ru"];
+        let named: Vec<String> = std::env::var("CRAFT_PROOF_EMBED")
+            .unwrap_or_default()
+            .split(',')
+            .map(|l| l.trim().to_ascii_lowercase())
+            .filter(|l| !l.is_empty())
+            .collect();
+        let allowed = |base: &str| base == "en" || if named.is_empty() { CARRY_OVER.contains(&base) } else { named.iter().any(|n| n == base) };
+        let embedded = wordcraft_proof::packs::embedded_locales();
+        let extra: Vec<&&str> = embedded.iter().filter(|l| !allowed(l.split('-').next().unwrap_or(l))).collect();
+        assert!(extra.is_empty(), "embedded beyond English and {named:?} (or the carry-over by default): {extra:?}");
+    }
+}
@@ -0,0 +1,104 @@
+//! The font store as the packs [`Backend`](super::Backend): its folder, its consent setting, the
+//! pinned pack index and verified installs (craft-libs `craft-font-store`, kind `language-pack`).
+
+use std::path::PathBuf;
+use std::sync::{Mutex, PoisonError};
+
+use craft_font_catalog::FontEnv;
+use craft_font_store::{FontStore, Network, PackEntry, PackIndex, PackIndexPin};
+
+use super::{Cancel, Pack};
+
+/// The pack index this build trusts: craft-i18n's first Language pack release, language-packs-1
+/// (61 packs, built from craft-i18n `4504e288`), pinned by the sha256 of its `index.json`.
+/// `CRAFT_LANGUAGE_PACK_INDEX` (read by the font store itself) points WordCraft at a local or test
+/// index instead, for trying packs before they are released.
+pub const PACK_INDEX: PackIndexPin = PackIndexPin {
+    url: "https://github.com/storytold/craft-i18n/releases/download/language-packs-1/index.json",
+    sha256: "531465b50126d0594bb89a9c4512f7ae67d581259636040a410190641ec1428c",
+};
+
+pub(super) struct FontStoreBackend {
+    store: FontStore,
+    /// The index last read or fetched; installs take their entry (url, size, sha256) from it.
+    index: Mutex<Option<PackIndex>>,
+}
+
+impl FontStoreBackend {
+    pub(super) fn new(user: bool) -> Option<Self> {
+        let env = FontEnv::from_process();
+        let store = if user { FontStore::for_user(&env)? } else { FontStore::for_agent(&env)? };
+        Some(Self { store, index: Mutex::new(None) })
+    }
+
+    /// A store in `dir`.
+    pub(super) fn in_dir(dir: PathBuf, user: bool) -> Self {
+        let caller = if user { craft_font_store::Caller::User } else { craft_font_store::Caller::Agent };
+        Self { store: FontStore::new(dir, caller), index: Mutex::new(None) }
+    }
+
+    fn keep(&self, index: PackIndex) -> Vec<Pack> {
+        let packs = index.packs.iter().map(pack).collect();
+        *self.index.lock().unwrap_or_else(PoisonError::into_inner) = Some(index);
+        packs
+    }
+}
+
+fn pack(e: &PackEntry) -> Pack {
+    Pack {
+        locale: e.locale.clone(),
+        name: e.name.clone(),
+        version: e.version.clone(),
+        url: e.url.clone(),
+        // TODO(CP4): the index's optional `source` once craft-font-store's PackEntry carries it.
+        source: None,
+        size: e.size,
+        licence: e.licence.clone(),
+        class: e.class.clone(),
+        spelling: e.spelling,
+        hyphenation: e.hyphenation,
+    }
+}
+
+impl super::Backend for FontStoreBackend {
+    fn index(&self) -> Result<Vec<Pack>, String> {
+        Ok(self.store.cached_language_pack_index(&PACK_INDEX).map(|i| self.keep(i)).unwrap_or_default())
+    }
+
+    fn fetch_index(&self, cancel: &Cancel) -> Result<Vec<Pack>, String> {
+        let index = self.store.fetch_language_pack_index(&PACK_INDEX, &cancel.0).map_err(|e| e.to_string())?;
+        Ok(self.keep(index))
+    }
+
+    fn network_on(&self) -> bool {
+        self.store.network() == Network::On
+    }
+
+    fn enable_network(&self) -> Result<(), String> {
+        self.store.enable_network().map_err(|e| e.to_string())
+    }
+
+    fn install(&self, pack: &Pack, cancel: &Cancel, progress: &mut dyn FnMut(u64, u64)) -> Result<(), String> {
+        let index = {
+            let held = self.index.lock().unwrap_or_else(PoisonError::into_inner).clone();
+            match held {
+                Some(i) => i,
+                None => self.store.cached_language_pack_index(&PACK_INDEX).ok_or("the list of Language packs hasn't been fetched")?,
+            }
+        };
+        let entry = index.pack(&pack.locale).ok_or_else(|| format!("there is no language pack for {}", pack.locale))?.clone();
+        self.store.install_language_pack(&index, &entry, &cancel.0, &mut |p| progress(p.done, p.total)).map(|_| ()).map_err(|e| e.to_string())
+    }
+
+    fn remove(&self, locale: &str) -> Result<(), String> {
+        self.store.remove_language_pack(locale).map_err(|e| e.to_string())
+    }
+
+    fn index_override(&self) -> Option<String> {
+        craft_font_store::language_pack_index_overridden()
+    }
+
+    fn packs_dir(&self) -> Option<PathBuf> {
+        Some(self.store.language_packs_dir())
+    }
+}
@@ -27,7 +27,10 @@ impl Session {
        let deleted = (hide_deleted, self.prefs.markup.hidden_deletions(), self.revisions_in_balloons());
        let outline = (mode == ViewMode::Outline).then(|| self.view.outline.view());
        let shown = (self.view.marks || self.prefs.marks.hidden, self.view.field_codes);
-        let (rev, proofing, numeral) = (self.rev, self.view.proofing, self.view.numeral);
+        // A dictionary or Language pack installed, removed or edited (in this document's window or
+        // any other) proofs it again, open or parked.
+        let rev = (self.rev, wordcraft_proof::dictionary_generation());
+        let (proofing, numeral) = (self.view.proofing, self.view.numeral);
        let fits = |e: &Option<LayoutEntry>| {
            e.as_ref().is_some_and(|(r, w, h, m, sp, _, pf, o, hd, n, t, sh)| {
                *r == rev
@@ -74,10 +77,8 @@ impl Session {
        };
        let l = Arc::new(wordcraft_layout::layout(&self.doc, &mut self.cache, &opts));
        let [newest, _] = std::mem::take(&mut self.layout);
-        self.layout = [
-            Some((self.rev, width, height, mode, spacing, l.clone(), self.view.proofing, outline, deleted, self.view.numeral, typing, shown)),
-            newest,
-        ];
+        self.layout =
+            [Some((rev, width, height, mode, spacing, l.clone(), self.view.proofing, outline, deleted, self.view.numeral, typing, shown)), newest];
        l
    }

@@ -258,7 +258,7 @@ impl Session {
        }
        self.rev = rev;
        for slot in &mut self.layout {
-            if slot.as_ref().is_some_and(|l| l.0 != rev) {
+            if slot.as_ref().is_some_and(|l| l.0.0 != rev) {
                *slot = None;
            }
        }
@@ -56,12 +56,12 @@ impl EditSnapshot {
 /// block.
 pub(super) type RunSnapshot = (Document, Selection, Vec<Arc<Undo>>, Vec<Arc<Undo>>, bool, u64, Option<ColumnBlock>);

-/// A layout made and what it was made for: (revision, width, read height, view, read line
+/// A layout made and what it was made for: ((revision, proofing dictionaries' generation), width, read height, view, read line
 /// spacing, layout, proofing, outline, (deletions left out, deletions marked Hidden, body
 /// deletions in balloons), numeral display, the word being typed, (hidden text shown, field
 /// codes shown)).
 pub(super) type LayoutEntry = (
-    u64,
+    (u64, u64),
    f32,
    f32,
    ViewMode,
@@ -1,6 +1,6 @@
-// Reuse our hand-written Czech proof fixture; its install helper resolves this parent import.
+// A copy of craft-proof's hand-written Czech fixture (`crates/proof/src/cs/test_dict.rs` in
+// craft-libs, which a dependency's tests can't share); its install helper resolves this parent import.
 use wordcraft_proof::cs::install_dictionary;
-#[path = "../../proof/src/cs/test_dict.rs"]
 mod czech_test_dict;

 use serde_json::json;
@@ -18,6 +18,22 @@ mod revision_checks;
 mod typing;
 mod wordart;

+/// True, after saying so on stderr, when this build lacks the Language data the proofing tests
+/// need (craft-proof built without `CRAFT_I18N_DIR`). The probes are craft-proof's public
+/// `english_source` (the embedded English dictionary) and its American/British variant table
+/// (`colour` → `color`): a craft-i18n checkout without some of the files embeds only the rest,
+/// so both are asked. Without them English still spells from craft-proof's built-in word list,
+/// but there are no hyphenation patterns and no variant table, so the tests that check those
+/// return early, as craft-proof's own do; with the data (CI, releases) they all run.
+pub(crate) fn no_language_data(test: &str) -> bool {
+    let missing = wordcraft_proof::english_source(wordcraft_proof::ProofLang::En) != Some(wordcraft_proof::EnglishSource::Embedded)
+        || wordcraft_proof::en_variants::preferred("colour", wordcraft_proof::ProofLang::En).as_deref() != Some("color");
+    if missing {
+        eprintln!("{test}: skipped, needs the Language data (set CRAFT_I18N_DIR to a craft-i18n checkout)");
+    }
+    missing
+}
+
 fn s() -> Session {
    Session::new(wordcraft_doc::Document::new())
 }
@@ -8665,6 +8681,9 @@ fn pairs(v: &[(&str, &str)]) -> Vec<(String, String)> {
 /// it and Change fixes it.
 #[test]
 fn english_dialects_through_the_review_commands() {
+    if no_language_data("english_dialects_through_the_review_commands") {
+        return;
+    }
    let mut s = s();
    run(&mut s, "document.setText", json!({"text": "The colour of the center."}));
    // US English by default: the British spelling is flagged, the American one isn't.
@@ -8698,6 +8717,9 @@ fn english_dialects_through_the_review_commands() {
 /// and is proofed the same way when it comes back.
 #[test]
 fn english_dialects_survive_a_docx_round_trip() {
+    if no_language_data("english_dialects_survive_a_docx_round_trip") {
+        return;
+    }
    use std::io::Read;
    let mut s = s();
    run(&mut s, "document.setText", json!({"text": "Colour and color. Centre and center."}));
@@ -8726,6 +8748,9 @@ fn english_dialects_survive_a_docx_round_trip() {
 /// system, #1045); another region's language leaves US English.
 #[test]
 fn new_documents_follow_the_regions_english() {
+    if no_language_data("new_documents_follow_the_regions_english") {
+        return;
+    }
    let mut s = s();
    s.system_region = wordcraft_geom::Region::from_tag("en_GB.UTF-8");
    assert_eq!(s.region_english().as_deref(), Some("en-GB"));
@@ -11599,3 +11624,140 @@ fn export_png_rejects_a_page_number_that_does_not_fit() {
    assert!(!target.exists(), "no file is written for an invalid page");
    let _ = std::fs::remove_dir_all(&dir);
 }
+
+/// A Word package whose body holds `body` and, as `rIdC`, the chunk part `word/afchunk.{ext}`.
+#[cfg(not(target_arch = "wasm32"))]
+fn docx_with_chunk(body: &str, ext: &str, content_type: &str, chunk: &[u8]) -> Vec<u8> {
+    use std::io::Write;
+    let files: [(String, Vec<u8>); 4] = [
+        (
+            "[Content_Types].xml".into(),
+            format!(r#"<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types"><Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/><Default Extension="{ext}" ContentType="{content_type}"/><Override PartName="/word/document.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml"/></Types>"#).into_bytes(),
+        ),
+        ("_rels/.rels".into(), br#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"><Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="word/document.xml"/></Relationships>"#.to_vec()),
+        (
+            "word/_rels/document.xml.rels".into(),
+            format!(r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"><Relationship Id="rIdC" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/aFChunk" Target="afchunk.{ext}"/></Relationships>"#).into_bytes(),
+        ),
+        (
+            "word/document.xml".into(),
+            format!(r#"<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"><w:body>{body}</w:body></w:document>"#).into_bytes(),
+        ),
+    ];
+    let mut zw = zip::ZipWriter::new(std::io::Cursor::new(Vec::new()));
+    for (name, bytes) in files.iter().map(|(n, b)| (n.as_str(), b.as_slice())).chain([(format!("word/afchunk.{ext}").as_str(), chunk)]) {
+        zw.start_file(name, zip::write::SimpleFileOptions::default()).unwrap();
+        zw.write_all(bytes).unwrap();
+    }
+    zw.finish().unwrap().into_inner()
+}
+
+/// A Single File Web Page under a Word or RTF name opens as one, like the web pages of #2853.
+#[test]
+fn a_single_file_web_page_saved_as_doc_opens_by_content() {
+    let mht = "MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=\"B\"\r\n\r\n--B\r\nContent-Type: text/html; charset=\"utf-8\"\r\n\r\n<p>Izvje\u{161}taj</p>\r\n--B--\r\n";
+    for name in ["report.doc", "report.docx", "report.rtf"] {
+        let doc = crate::io::open_bytes(name, mht.as_bytes()).unwrap_or_else(|e| panic!("{name}: {e}"));
+        assert_eq!(doc.plain_text(StoryRef::Body), "Izvje\u{161}taj", "{name}");
+    }
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+#[test]
+fn alt_chunks_open_in_place_as_word_brings_them_in() {
+    let png = wordcraft_formats::model::base64_encode(&tiny_png());
+    let mht = format!(
+        "MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=\"B\"\r\n\r\n--B\r\nContent-Location: file:///C:/doc.htm\r\n\
+         Content-Transfer-Encoding: quoted-printable\r\nContent-Type: text/html; charset=\"utf-8\"\r\n\r\n\
+         <style>p.c {{ text-align: center }}</style><p class=3D\"c\">=C4=8Clanak 1.</p><p><img src=3D\"doc_files/a.png\"></p>\r\n\
+         --B\r\nContent-Location: file:///C:/doc_files/a.png\r\nContent-Transfer-Encoding: base64\r\nContent-Type: image/png\r\n\r\n{png}\r\n--B--\r\n"
+    );
+    let p = |t: &str| format!(r#"<w:p><w:r><w:t>{t}</w:t></w:r></w:p>"#);
+    let chunk = r#"<w:altChunk r:id="rIdC"/>"#;
+    let bytes = docx_with_chunk(&format!("{}{chunk}{}", p("Before"), p("After")), "mht", "message/rfc822", mht.as_bytes());
+    let doc = crate::io::open_bytes("chunk.docx", &bytes).unwrap();
+    let text = doc.plain_text(StoryRef::Body);
+    assert!(text.starts_with("Before\nČlanak 1.\n") && text.ends_with("After"), "{text:?}");
+    let para = doc.para(StoryRef::Body, &wordcraft_doc::Path::top(1)).unwrap();
+    assert_eq!(para.props.align, Some(wordcraft_doc::Align::Center));
+    // Web page text: Times New Roman unless the page says otherwise.
+    assert_eq!(para.props_of_char(0).font.as_deref(), Some("Times New Roman"));
+    // The picture came along, its bytes under the key it shows.
+    let key = doc
+        .body
+        .iter()
+        .find_map(|b| match &**b {
+            wordcraft_doc::Block::Para(p) => p.objects.iter().find_map(|o| match o {
+                wordcraft_doc::InlineObject::Image { media, .. } => Some(media.clone()),
+                _ => None,
+            }),
+            _ => None,
+        })
+        .expect("the chunk's picture");
+    assert_eq!(doc.media.get(&key).map(|m| m.to_vec()), Some(tiny_png()));
+
+    // HTML (in its content type's charset), RTF and plain-text chunks come in through their
+    // importers too, and an MHT chunk is known by what it holds.
+    for (ext, ct, data, want) in [
+        ("htm", "text/html; charset=windows-1250", &b"<p>\xc8lanak chunk</p>"[..], "Članak chunk"),
+        ("rtf", "application/rtf", br"{\rtf1\ansi Rich chunk\par}", "Rich chunk"),
+        ("txt", "text/plain", b"Plain chunk", "Plain chunk"),
+        ("bin", "", mht.as_bytes(), "Članak 1."),
+    ] {
+        let bytes = docx_with_chunk(&format!("{chunk}{}", p("End")), ext, ct, data);
+        let text = crate::io::open_bytes("chunk.docx", &bytes).unwrap().plain_text(StoryRef::Body);
+        assert!(text.starts_with(want) && text.ends_with("End"), "{ext}: {text:?}");
+    }
+}
+
+#[cfg(not(target_arch = "wasm32"))]
+#[test]
+fn text_from_file_brings_the_files_pictures() {
+    let dir = scratch_dir("text-from-file-pictures");
+    let mut src = Session::new(wordcraft_doc::Document::new());
+    let list = src.doc.numbering.add_list(wordcraft_doc::numbering::ListKind::Bullet);
+    // A list item, then a paragraph (a single paragraph would take the caret's formatting).
+    src.doc.split_paragraph(&Pos::body(0, 0)).unwrap();
+    src.doc.para_mut(StoryRef::Body, &wordcraft_doc::Path::top(0)).unwrap().props.numbering =
+        Some(wordcraft_doc::props::NumRef { num: list, level: 0 });
+    let key = src.doc.add_media(tiny_png(), "png");
+    let image =
+        wordcraft_doc::InlineObject::Image { media: key, w: 20.0, h: 20.0, alt: String::new(), float: Default::default(), crop: [0.0; 4], ole: None };
+    src.doc.para_mut(StoryRef::Body, &wordcraft_doc::Path::top(0)).unwrap().insert_object(0, image, &Default::default()).unwrap();
+    let path = dir.join("pictures.docx");
+    std::fs::write(&path, wordcraft_docx::write(&src.doc).unwrap()).unwrap();
+
+    let mut s = s();
+    // A picture and a list of our own first, so the file's key and list id would point at ours.
+    let ours = s.doc.add_media(vec![9, 9, 9], "png");
+    let our_list = s.doc.numbering.add_list(wordcraft_doc::numbering::ListKind::Numbered);
+    run(&mut s, "insert.textFromFile", json!({"path": path.to_string_lossy()}));
+    let _ = std::fs::remove_dir_all(&dir);
+    let shown: Vec<String> = s
+        .doc
+        .body
+        .iter()
+        .filter_map(|b| match &**b {
+            wordcraft_doc::Block::Para(p) => p.objects.iter().find_map(|o| match o {
+                wordcraft_doc::InlineObject::Image { media, .. } => Some(media.clone()),
+                _ => None,
+            }),
+            _ => None,
+        })
+        .collect();
+    assert_eq!(shown.len(), 1);
+    assert_ne!(shown[0], ours);
+    assert_eq!(s.doc.media.get(&shown[0]).map(|m| m.to_vec()), Some(tiny_png()));
+    let listed: Vec<u32> = s
+        .doc
+        .body
+        .iter()
+        .filter_map(|b| match &**b {
+            wordcraft_doc::Block::Para(p) => p.props.numbering.map(|n| n.num),
+            _ => None,
+        })
+        .collect();
+    assert_eq!(listed.len(), 1);
+    assert_ne!(listed[0], our_list);
+    assert!(s.doc.numbering.level(listed[0], 0).is_some_and(|l| l.format == wordcraft_doc::section::NumFormat::Bullet));
+}
@@ -3631,3 +3631,82 @@ fn backspace_cannot_join_into_locked_content_control() {
    assert!(s.run("text.backspace", &json!({})).is_err());
    assert_eq!(s.doc, before);
 }
+
+// English spelling from SCOWL 2026 (CP2): each English reads its own SCOWL dictionary, with
+// `en-variants.txt` deciding the other dialect's spellings. Typed text, as a person types it.
+
+/// The spelling words typed text gets in `lang` (the whole document's language), with the first
+/// suggestion of each.
+fn typed_spelling_in(lang: &str, keys: &str) -> Vec<(String, String)> {
+    let mut s = s();
+    run(&mut s, "review.language", json!({"lang": lang, "default": true}));
+    typ(&mut s, keys);
+    let issues = s.run("review.issues", &json!({})).unwrap();
+    issues
+        .as_array()
+        .unwrap()
+        .iter()
+        .filter(|i| i["kind"] == "spelling")
+        .map(|i| (i["text"].as_str().unwrap().to_string(), i["suggestions"][0].as_str().unwrap_or("").to_string()))
+        .collect()
+}
+
+/// #40: the British spellings the issue lists are right in en-GB, and the American ones are
+/// flagged with the British form first.
+#[test]
+fn issue_40_british_spellings_are_right_in_en_gb() {
+    if super::no_language_data("issue_40_british_spellings_are_right_in_en_gb") {
+        return;
+    }
+    let none: Vec<(String, String)> = Vec::new();
+    let text = "The judgement on the defence licence put the programme in the catalogue. ";
+    assert_eq!(typed_spelling_in("en-GB", text), none);
+    assert_eq!(typed_spelling_in("en-GB", "Its colour and centre, organised and organized, travelled. "), none);
+    let flagged = typed_spelling_in("en-GB", "The color of the center. ");
+    assert_eq!(flagged, [("color".to_string(), "colour".to_string()), ("center".to_string(), "centre".to_string())]);
+}
+
+/// #520: common modern words are right (SCOWL plus our supplement), and real misspellings of
+/// them are still caught.
+#[test]
+fn issue_520_common_words_are_right() {
+    if super::no_language_data("issue_520_common_words_are_right") {
+        return;
+    }
+    let none: Vec<(String, String)> = Vec::new();
+    assert_eq!(typed_spelling_in("en-US", "My desktop email has downloads, a smartphone and a website. "), none);
+    // (`recieved` would be AutoCorrected as it is typed.)
+    let flagged = typed_spelling_in("en-US", "My desktpo emial. ");
+    assert_eq!(flagged.iter().map(|f| f.0.as_str()).collect::<Vec<_>>(), ["desktpo", "emial"]);
+    assert_eq!(flagged[0].1, "desktop");
+}
+
+/// #1281 and #1045: Canadian and Australian English (and the other British-spelling Englishes)
+/// proof by their own rules; ordinals are words in every English.
+#[test]
+fn issues_1281_1045_english_dialects_proof_typed_text() {
+    if super::no_language_data("issues_1281_1045_english_dialects_proof_typed_text") {
+        return;
+    }
+    let none: Vec<(String, String)> = Vec::new();
+    // Canada: British -our and -re, American -ize.
+    assert_eq!(typed_spelling_in("en-CA", "We organize the colour in the centre. "), none);
+    assert_eq!(
+        typed_spelling_in("en-CA", "We organise the color. "),
+        [("organise".to_string(), "organize".to_string()), ("color".to_string(), "colour".to_string())]
+    );
+    // Australia, New Zealand, India: British spelling.
+    for lang in ["en-AU", "en-NZ", "en-IN"] {
+        assert_eq!(typed_spelling_in(lang, "We analyse the colour of the theatre. "), none, "{lang}");
+        assert_eq!(typed_spelling_in(lang, "We analyze it. "), [("analyze".to_string(), "analyse".to_string())], "{lang}");
+    }
+    // US English flags British spellings with the American form first.
+    assert_eq!(typed_spelling_in("en-US", "The colour. "), [("colour".to_string(), "color".to_string())]);
+    // SCOWL's ordinals (COMPOUNDRULE) are words, even where numbers are checked.
+    let mut s = s();
+    run(&mut s, "file.options", json!({"ignoreNumbers": false}));
+    typ(&mut s, "The 21st and 112th times, not the 21th. ");
+    let issues = s.run("review.issues", &json!({})).unwrap();
+    let words: Vec<&str> = issues.as_array().unwrap().iter().filter(|i| i["kind"] == "spelling").filter_map(|i| i["text"].as_str()).collect();
+    assert_eq!(words, ["21th"]);
+}
@@ -35,7 +35,7 @@ pub(crate) use names::face_names;
 pub use names::{LocalName, base_style, name_language_matches, system_font_dirs};
 use names::{MAX_FILE_FACES, enumerate_faces, make_face};
 #[cfg(test)]
-use names::{file_face_names, font_faces, mini_font, name};
+use names::{MAX_NAME_CHARS, file_face_names, font_faces, mini_font, name};
 #[cfg(test)]
 pub(crate) use names::{last_resort_face, legacy_family};

@@ -61,8 +61,16 @@ pub(super) struct CatalogEntry {
 /// The longest name a document's font table gives an embedded font that faces take on.
 pub(super) const MAX_FAMILY_NAME: usize = 128;

+/// The longest name read from a font's `name` table, in characters. A record can hold 64K
+/// characters, and each face copies its family into its localized names and into every named
+/// instance, so a small font could claim hundreds of megabytes of names. Real names run to a
+/// few dozen characters.
+pub(super) const MAX_NAME_CHARS: usize = 128;
+
 pub(super) fn name(font: &skrifa::FontRef<'_>, ids: &[StringId]) -> Option<String> {
-    ids.iter().find_map(|id| font.localized_strings(*id).english_or_first().map(|s| s.to_string()).filter(|s| !s.is_empty()))
+    ids.iter().find_map(|id| {
+        font.localized_strings(*id).english_or_first().map(|s| s.chars().take(MAX_NAME_CHARS).collect::<String>()).filter(|s| !s.is_empty())
+    })
 }

 /// The family's names in other languages: localized typographic family names (name ID 16) stand
@@ -70,7 +78,6 @@ pub(super) fn name(font: &skrifa::FontRef<'_>, ids: &[StringId]) -> Option<Strin
 /// font has no typographic names). Bounded: a damaged font can list any number of records.
 fn local_names(f: &skrifa::FontRef<'_>, family: &str, id1: Option<&str>) -> Vec<LocalName> {
    const MAX_NAMES: usize = 64;
-    const MAX_CHARS: usize = 128;
    let typographic = f.localized_strings(StringId::TYPOGRAPHIC_FAMILY_NAME).next().is_some();
    let legacy = if typographic { id1 } else { Some(family) };
    let mut out: Vec<LocalName> = Vec::new();
@@ -81,7 +88,7 @@ fn local_names(f: &skrifa::FontRef<'_>, family: &str, id1: Option<&str>) -> Vec<
            if lang == "en" || lang.starts_with("en-") {
                continue;
            }
-            let name: String = s.chars().take(MAX_CHARS).collect::<String>().trim().to_string();
+            let name: String = s.chars().take(MAX_NAME_CHARS).collect::<String>().trim().to_string();
            if name.is_empty() || name.eq_ignore_ascii_case(english) || out.iter().any(|n| n.lang == lang && n.name == name) {
                continue;
            }
@@ -245,15 +245,24 @@ fn postscript_names_select_faces() {
 /// A variable font holding just `name` and `fvar` tables: family "Var Test", one `wght` axis
 /// and `instances` named instances, each with its own style name.
 fn variable_font(instances: u16) -> Vec<u8> {
-    let mut names: Vec<(u16, String)> = vec![(1, "Var Test".into()), (2, "Regular".into())];
-    names.extend((0..instances).map(|i| (256 + i, format!("Style {i}"))));
+    variable_font_named("Var Test", &[], instances)
+}
+
+/// [`variable_font`] with family `family`, and family names in other languages `local`
+/// (Windows language ids).
+fn variable_font_named(family: &str, local: &[(u16, &str)], instances: u16) -> Vec<u8> {
+    let mut names: Vec<(u16, u16, String)> = vec![(0x409, 1, family.into()), (0x409, 2, "Regular".into())];
+    names.extend(local.iter().map(|(lang, n)| (*lang, 1, n.to_string())));
+    names.extend((0..instances).map(|i| (0x409, 256 + i, format!("Style {i}"))));
+    // Records sorted by language, then name id.
+    names.sort_by_key(|(lang, id, _)| (*lang, *id));
    let mut name = vec![0, 0];
    name.extend_from_slice(&(names.len() as u16).to_be_bytes());
    name.extend_from_slice(&(6 + 12 * names.len() as u16).to_be_bytes());
    let mut strings = Vec::new();
-    for (id, text) in &names {
+    for (lang, id, text) in &names {
        let utf16: Vec<u8> = text.encode_utf16().flat_map(u16::to_be_bytes).collect();
-        for v in [3, 1, 0x409, *id, utf16.len() as u16, strings.len() as u16] {
+        for v in [3, 1, *lang, *id, utf16.len() as u16, strings.len() as u16] {
            name.extend_from_slice(&v.to_be_bytes());
        }
        strings.extend(utf16);
@@ -357,6 +366,27 @@ fn the_draw_that_turns_hinting_off_is_redrawn_unhinted_for_the_outline_cache() {
    assert_eq!(db.outline(&face, gid).elements(), tripping.path.elements());
 }

+#[test]
+fn long_names_are_cut_before_faces_copy_them() {
+    // A name record can hold 64K characters, and every named instance copies the family and its
+    // localized names: the names are cut when read, so the copies stay small.
+    let long = "W".repeat(14_000);
+    let local = "Ж".repeat(14_000);
+    let font = variable_font_named(&long, &[(0x419, &local)], 32);
+    let faces = font_faces(&skrifa::FontRef::new(&font).unwrap(), 0);
+    assert_eq!(faces.len(), 32);
+    for f in &faces {
+        assert_eq!(f.family.chars().count(), MAX_NAME_CHARS);
+        assert!(f.legacy.chars().count() <= MAX_NAME_CHARS + " Style 31".len(), "{}", f.legacy.len());
+        assert_eq!(f.local.len(), 1);
+        assert_eq!(f.local[0].name.chars().count(), MAX_NAME_CHARS);
+        assert_eq!(f.local[0].english, f.family);
+    }
+    // Short names are kept whole.
+    let faces = font_faces(&skrifa::FontRef::new(&variable_font_named("Var Test", &[(0x419, "Вар Тест")], 2)).unwrap(), 0);
+    assert_eq!((faces[0].family.as_str(), faces[0].local[0].name.as_str()), ("Var Test", "Вар Тест"));
+}
+
 #[test]
 fn named_instances_and_collection_members_are_capped() {
    // #3716: a file's faces are capped before they are built, however the file multiplies them.
@@ -8,18 +8,22 @@
 //! (`rb`/`rt`/`rp`: furigana and other phonetic guides, #1681), and the CSS
 //! properties `color`, `background-color`, `font-weight`, `font-style`, `font-size`,
 //! `font-family`, `text-decoration`, `vertical-align`, `text-align`, `white-space` and page
-//! breaks. Scripts and styles are skipped; nesting is capped.
+//! breaks, and on paragraphs `margin`, `text-indent` and `line-height`. CSS comes from `style`
+//! attributes and from `style` elements' rules with simple selectors (`p`, `.note`, `p.note`),
+//! applied in cascade order (specificity, then source order, then the `style` attribute).
+//! Scripts are skipped; nesting is capped.
 //!
 //! Export writes clean semantic HTML with inline CSS, pictures as `data:` URIs and ruby as
 //! `<ruby>`.

 use std::sync::Arc;

+use wordcraft_doc::props::LineSpacing;
 use wordcraft_doc::props::VAlign;
 use wordcraft_doc::{Align, Document};

 use crate::model::{
-    self, Cell, FBlock, FTable, Flow, Fmt, Inline, Kind, ListInfo, Meta, Para, base64_encode, data_uri, make_img, mime_of, parse_color,
+    self, Cell, FBlock, FTable, Flow, Fmt, Inline, Kind, ListInfo, Meta, Para, Spacing, base64_encode, data_uri, make_img, mime_of, parse_color,
 };

 mod builder;
@@ -32,9 +36,11 @@ mod math_ruby;

 pub use builder::ImageLoader;
 pub use css::css_len_pt;
-pub use entity::{decode, decode_entity, unescape};
+pub use entity::{decode, decode_as, decode_entity, unescape};
 pub use export::{esc, export, export_flow, export_fragment, inlines_html};
-pub use import::{import, import_on, import_with, parse, parse_with};
+#[cfg(test)]
+use import::parse_in;
+pub use import::{PAGE_FONT, import, import_on, import_page, import_with, parse, parse_with};

 #[cfg(test)]
 mod tests;
@@ -1,6 +1,9 @@
-use super::css::{apply_style, css_len_pt};
-use super::import::collapse_ws;
-use super::{Align, Arc, Cell, FBlock, FTable, Flow, Fmt, Inline, Kind, ListInfo, Meta, Para, VAlign, data_uri, make_img, model, parse_color};
+use super::css::{apply_style, css_len_pt, strip_important};
+use super::import::{PAGE_FONT, PAGE_HEADING_SIZES, collapse_ws};
+use super::{
+    Align, Arc, Cell, FBlock, FTable, Flow, Fmt, Inline, Kind, LineSpacing, ListInfo, Meta, Para, Spacing, VAlign, data_uri, make_img, model,
+    parse_color,
+};

 /// Open elements kept at most (deeper start tags are ignored).
 const MAX_STACK: usize = 256;
@@ -28,6 +31,12 @@ struct El {
    name: String,
    kind: ElKind,
    fmt: Fmt,
+    /// Margins, indents and line height its CSS states (blocks only; they don't inherit; on a
+    /// page's spans, their margins).
+    spacing: Spacing,
+    /// A page's `p` or heading: margins its CSS leaves out (or writes invalidly) are HTML's
+    /// automatic spacing.
+    auto_margins: bool,
    align: Option<Align>,
    preserve: bool,
    page_break: bool,
@@ -77,20 +86,37 @@ fn apply_font_attributes(attrs: &[(String, String)], fmt: &mut Fmt) {
    }
 }

-/// Whether an element's `border` attribute or `style` declares a visible border.
-fn declares_border(attrs: &[(String, String)], is_table: bool) -> bool {
+/// Whether an element's `border` attribute or its CSS declares a visible border.
+fn declares_border(attrs: &[(String, String)], css: &str, is_table: bool) -> bool {
    if is_table && attr(attrs, "border").is_some_and(|b| b.trim().parse::<u32>().map(|n| n > 0).unwrap_or(b.trim().is_empty())) {
        return true;
    }
-    let Some(style) = attr(attrs, "style") else { return false };
-    style.split(';').any(|decl| {
-        let Some((name, value)) = decl.split_once(':') else { return false };
-        let name = name.trim().to_ascii_lowercase();
-        if !matches!(name.as_str(), "border" | "border-top" | "border-right" | "border-bottom" | "border-left") {
-            return false;
+    // Each side's last declaration decides (the `style` attribute's come after the sheet's).
+    let mut sides = [false; 4];
+    for (name, value) in decls(css) {
+        let set = match name.as_str() {
+            "border" => 0..4,
+            "border-top" => 0..1,
+            "border-right" => 1..2,
+            "border-bottom" => 2..3,
+            "border-left" => 3..4,
+            _ => continue,
+        };
+        let value = value.to_ascii_lowercase();
+        let on = !value.is_empty() && !value.split_whitespace().any(|tok| matches!(tok, "none" | "hidden" | "0" | "0px" | "0pt"));
+        if let Some(s) = sides.get_mut(set) {
+            s.fill(on);
        }
-        let value = value.trim().to_ascii_lowercase();
-        !value.is_empty() && !value.split_whitespace().any(|tok| matches!(tok, "none" | "hidden" | "0" | "0px" | "0pt"))
+    }
+    sides.contains(&true)
+}
+
+/// A declaration block's declarations, at most [`MAX_DECLS`]: (property in lower case, value
+/// without `!important`).
+pub(super) fn decls(css: &str) -> impl Iterator<Item = (String, &str)> {
+    css.split(';').take(MAX_DECLS).filter_map(|d| {
+        let (k, v) = d.split_once(':')?;
+        Some((k.trim().to_ascii_lowercase(), strip_important(v)))
    })
 }

@@ -116,6 +142,12 @@ pub(super) struct Builder<'r> {
    pre_tail: bool,
    meta: Meta,
    title: Option<String>,
+    /// Rules of the page's `style` elements.
+    pub(super) sheet: Sheet,
+    /// Read as Word opens a web page ([`import_page`]), not as a pasted fragment.
+    page: bool,
+    /// What is left of [`MAX_CSS_WORK`].
+    css_work: usize,
 }

 fn is_void(n: &str) -> bool {
@@ -196,12 +228,310 @@ fn closes_p(n: &str) -> bool {
    )
 }

+/// The paragraph spacing a block's CSS states: `margin` (and its sides) as space before / after
+/// and left / right indents, `text-indent` as the first-line indent, `line-height` as line
+/// spacing (a number or percentage: a multiple; a length: exactly). `auto` and values that don't
+/// parse state nothing.
+fn block_spacing(css: &str) -> Spacing {
+    let mut sp = Spacing::default();
+    // A length with a unit, or 0: a bare number, a percentage or a keyword states nothing.
+    let len = |v: &str| match v.trim() {
+        "0" => Some(0.0),
+        v if v.ends_with('%') || v.parse::<f32>().is_ok() || !v.starts_with(|c: char| c.is_ascii_digit() || matches!(c, '.' | '-' | '+')) => None,
+        v => css_len_pt(v).filter(|x| x.abs() <= 1584.0),
+    };
+    for (k, v) in decls(css) {
+        match k.as_str() {
+            "margin" => {
+                let parts: Vec<Option<f32>> = v.split_whitespace().map(len).collect();
+                let [top, right, bottom, left] = match parts.as_slice() {
+                    [a] => [*a; 4],
+                    [a, b] => [*a, *b, *a, *b],
+                    [a, b, c] => [*a, *b, *c, *b],
+                    [a, b, c, d] => [*a, *b, *c, *d],
+                    _ => continue,
+                };
+                sp.before = top.map(|x| x.max(0.0)).or(sp.before);
+                sp.after = bottom.map(|x| x.max(0.0)).or(sp.after);
+                sp.left = left.or(sp.left);
+                sp.right = right.or(sp.right);
+            }
+            "margin-top" => sp.before = len(v).map(|x| x.max(0.0)).or(sp.before),
+            "margin-bottom" => sp.after = len(v).map(|x| x.max(0.0)).or(sp.after),
+            "margin-left" => sp.left = len(v).or(sp.left),
+            "margin-right" => sp.right = len(v).or(sp.right),
+            "text-indent" => sp.first = len(v).or(sp.first),
+            "line-height" => {
+                let lv = v.to_ascii_lowercase();
+                let multiple = |m: f32| (m.is_finite() && m > 0.0).then(|| LineSpacing::Multiple(m.clamp(0.25, 10.0)));
+                sp.line = if let Some(pc) = lv.strip_suffix('%') {
+                    pc.trim().parse::<f32>().ok().and_then(|p| multiple(p / 100.0))
+                } else if let Ok(m) = lv.parse::<f32>() {
+                    multiple(m)
+                } else if lv == "normal" {
+                    Some(LineSpacing::Multiple(1.0))
+                } else {
+                    len(&lv).filter(|x| *x > 0.0).map(LineSpacing::Exactly)
+                }
+                .or(sp.line);
+            }
+            _ => {}
+        }
+    }
+    sp
+}
+
+/// A rule of a `style` element with a simple selector: an element name, classes or both.
+#[derive(Clone, Debug)]
+struct Rule {
+    tag: Option<String>,
+    classes: Vec<String>,
+    /// Classes count ten, an element name one.
+    specificity: usize,
+    /// Shared by the selectors of one block.
+    decls: Arc<str>,
+    /// How many declarations `decls` holds.
+    count: usize,
+}
+
+/// Most rules kept from a page's style elements.
+const MAX_RULES: usize = 10_000;
+/// Most declarations read for one element (its rules' and its `style` attribute's).
+const MAX_DECLS: usize = 512;
+/// Most declarations read from one `style` attribute.
+const INLINE_DECLS: usize = 64;
+/// Rule matching (rules looked at plus bytes gathered) one page may spend.
+const MAX_CSS_WORK: usize = 64 << 20;
+/// Most bytes of declarations gathered from the rules one element matches.
+const MAX_RULE_CSS: usize = 64 * 1024;
+/// Most of an element's classes looked up.
+const MAX_CLASSES: usize = 32;
+
+/// A page's style rules, found by the element name or a class they need.
+#[derive(Default)]
+pub(super) struct Sheet {
+    rules: Vec<Rule>,
+    /// Rules naming classes, by their first class.
+    by_class: std::collections::HashMap<String, Vec<usize>>,
+    /// Rules naming only an element, by its name.
+    by_tag: std::collections::HashMap<String, Vec<usize>>,
+    /// `*` rules.
+    any: Vec<usize>,
+}
+
+impl Sheet {
+    fn push(&mut self, r: Rule) {
+        let i = self.rules.len();
+        match (r.classes.first(), &r.tag) {
+            (Some(c), _) => self.by_class.entry(c.clone()).or_default().push(i),
+            (None, Some(t)) => self.by_tag.entry(t.clone()).or_default().push(i),
+            (None, None) => self.any.push(i),
+        }
+        self.rules.push(r);
+    }
+}
+
+/// Read the rules of a style sheet into `out`. At-rules (`@media`, `@page`…) and rules whose
+/// selectors aren't simple (descendants, pseudo-classes, attributes…) are skipped.
+pub(super) fn parse_sheet(css: &str, out: &mut Sheet) {
+    // Comments go, as do the `<!-- -->` old pages hide their styles in.
+    let mut text = String::with_capacity(css.len());
+    let mut rest = css;
+    while let Some(i) = rest.find("/*") {
+        text.push_str(rest.get(..i).unwrap_or(""));
+        rest = rest.get(i + 2..).and_then(|r| r.find("*/").and_then(|j| r.get(j + 2..))).unwrap_or("");
+    }
+    text.push_str(rest);
+    let text = text.replace("<!--", " ").replace("-->", " ");
+    let mut rest = text.as_str();
+    while let Some(open) = rest.find('{') {
+        // A statement at-rule before it (`@charset "x";`, `@import url(a);`) isn't a selector.
+        let selectors = rest.get(..open).unwrap_or("").rsplit(';').next().unwrap_or("").trim();
+        let body = rest.get(open + 1..).unwrap_or("");
+        // The block's end, counting nested braces (an at-rule's own rules).
+        let mut depth = 1usize;
+        let mut end = body.len();
+        for (i, c) in body.char_indices() {
+            match c {
+                '{' => depth += 1,
+                '}' => {
+                    depth -= 1;
+                    if depth == 0 {
+                        end = i;
+                        break;
+                    }
+                }
+                _ => {}
+            }
+        }
+        let decls = body.get(..end).unwrap_or("");
+        rest = body.get(end + 1..).unwrap_or("");
+        if selectors.starts_with('@') || decls.contains('{') || decls.len() > MAX_RULE_CSS {
+            continue;
+        }
+        let (count, decls) = (decls.split(';').count(), Arc::<str>::from(decls));
+        for sel in selectors.split(',') {
+            if out.rules.len() >= MAX_RULES {
+                return;
+            }
+            if let Some((tag, classes)) = simple_selector(sel.trim()) {
+                let specificity = classes.len() * 10 + usize::from(tag.is_some());
+                out.push(Rule { tag, classes, specificity, decls: decls.clone(), count });
+            }
+        }
+    }
+}
+
+/// `p`, `.note`, `p.note.wide` or `*`: (element name, classes). Class names may start with a
+/// digit (`.000003`): browsers drop such a rule, but Word's HTML import applies it, and
+/// converters write them.
+fn simple_selector(sel: &str) -> Option<(Option<String>, Vec<String>)> {
+    if sel == "*" {
+        return Some((None, Vec::new()));
+    }
+    let mut parts = sel.split('.');
+    let tag = parts.next()?;
+    let tag = if tag.is_empty() {
+        None
+    } else if tag.chars().all(|c| c.is_ascii_alphanumeric()) && tag.starts_with(|c: char| c.is_ascii_alphabetic()) {
+        Some(tag.to_ascii_lowercase())
+    } else {
+        return None;
+    };
+    let ident = |c: &str| !c.is_empty() && c.chars().all(|ch| ch.is_alphanumeric() || ch == '_' || ch == '-');
+    let classes: Vec<String> = parts.map(str::to_string).collect();
+    if classes.iter().any(|c| !ident(c)) || (tag.is_none() && classes.is_empty()) {
+        return None;
+    }
+    Some((tag, classes))
+}
+
+impl Builder<'_> {
+    /// An element's CSS: the declarations of the style sheet rules it matches, in cascade order,
+    /// then its `style` attribute (later declarations win).
+    fn css_of(&mut self, name: &str, attrs: &[(String, String)]) -> String {
+        // The `style` attribute's first declarations always fit in what is read of the result.
+        let inline: Vec<&str> = attr(attrs, "style").unwrap_or("").split(';').take(INLINE_DECLS).collect();
+        let inline = inline.join(";");
+        let sheet = &self.sheet;
+        if sheet.rules.is_empty() || self.css_work == 0 {
+            return important_last(&inline);
+        }
+        let mut classes: Vec<&str> = attr(attrs, "class").map(|c| c.split_whitespace().take(MAX_CLASSES).collect()).unwrap_or_default();
+        classes.sort_unstable();
+        classes.dedup();
+        let mut candidates: Vec<usize> = sheet.any.clone();
+        candidates.extend(sheet.by_tag.get(name).into_iter().flatten());
+        for c in &classes {
+            candidates.extend(sheet.by_class.get(*c).into_iter().flatten());
+        }
+        let mut hits: Vec<(usize, usize)> = candidates
+            .iter()
+            .filter_map(|&i| sheet.rules.get(i).map(|r| (i, r)))
+            .filter(|(_, r)| r.tag.as_deref().is_none_or(|t| t == name) && r.classes.iter().all(|c| classes.contains(&c.as_str())))
+            .map(|(i, r)| (r.specificity, i))
+            .collect();
+        hits.sort_unstable();
+        let mut css = String::new();
+        let mut count = 0;
+        for (_, i) in hits {
+            let Some(r) = sheet.rules.get(i) else { continue };
+            if css.len() + r.decls.len() > MAX_RULE_CSS || count + r.count > MAX_DECLS - INLINE_DECLS {
+                break;
+            }
+            css.push_str(&r.decls);
+            css.push(';');
+            count += r.count;
+        }
+        // A page whose rules cost too much to apply reads on without them.
+        self.css_work = self.css_work.saturating_sub(candidates.len() + css.len());
+        css.push_str(&inline);
+        important_last(&css)
+    }
+}
+
+/// `css`'s declarations in order, the `!important` ones after the others: they win over normal
+/// ones wherever they come from (CSS Cascade 5 §6.4), an inline one over the sheet's.
+fn important_last(css: &str) -> String {
+    let (important, normal): (Vec<&str>, Vec<&str>) = css.split(';').filter(|d| !d.trim().is_empty()).partition(|d| strip_important(d) != d.trim());
+    normal.into_iter().chain(important).collect::<Vec<_>>().join(";")
+}
+
+/// The borders a cell's CSS gives it (`border`, `border-top`…: style, width and colour in any
+/// order; `windowtext` is the automatic colour). `None` when it states none.
+fn cell_borders(css: &str) -> Option<wordcraft_doc::props::Borders> {
+    use wordcraft_doc::props::{Border, BorderStyle, Borders};
+    let mut b = Borders::default();
+    let mut any = false;
+    for (k, v) in decls(css) {
+        let sides: &[u8] = match k.as_str() {
+            "border" => b"tlbr",
+            "border-top" => b"t",
+            "border-left" => b"l",
+            "border-bottom" => b"b",
+            "border-right" => b"r",
+            _ => continue,
+        };
+        // CSS's default width is `medium`.
+        let (mut style, mut width, mut color) = (None, 2.25_f32, None);
+        for tok in v.split_whitespace() {
+            let lt = tok.to_ascii_lowercase();
+            match lt.as_str() {
+                "none" | "hidden" => style = Some(BorderStyle::None),
+                "solid" => style = Some(BorderStyle::Single),
+                "double" => style = Some(BorderStyle::Double),
+                "dotted" => style = Some(BorderStyle::Dotted),
+                "dashed" => style = Some(BorderStyle::Dashed),
+                "thin" => width = 0.75,
+                "medium" => width = 2.25,
+                "thick" => width = 3.75,
+                "windowtext" => {}
+                _ => match css_len_pt(tok).filter(|_| tok.starts_with(|c: char| c.is_ascii_digit() || c == '.')) {
+                    Some(w) => width = w.clamp(0.0, 12.0),
+                    None => color = parse_color(tok).or(color),
+                },
+            }
+        }
+        // CSS draws no border without a style, or at no width.
+        let style = if width > 0.0 { style.unwrap_or(BorderStyle::None) } else { BorderStyle::None };
+        let border = Border { style, width, color, space: 0.0 };
+        for side in sides {
+            let slot = match side {
+                b't' => &mut b.top,
+                b'l' => &mut b.left,
+                b'b' => &mut b.bottom,
+                _ => &mut b.right,
+            };
+            *slot = Some(border);
+        }
+        any = true;
+    }
+    any.then_some(b)
+}
+
+/// Most columns given widths in one table.
+const MAX_TABLE_COLS: usize = 63;
+
+/// A cell's width, points: its CSS `width`, else its `width` attribute (pixels); percentages
+/// and widths that don't parse give none.
+fn cell_width(css: &str, attrs: &[(String, String)]) -> Option<f32> {
+    let from_css = decls(css).filter(|(k, v)| k == "width" && !v.ends_with('%')).filter_map(|(_, v)| css_len_pt(v));
+    let from_attr = || {
+        attr(attrs, "width")
+            .map(str::trim)
+            .filter(|v| !v.ends_with('%'))
+            .and_then(|v| v.trim_end_matches("px").trim().parse::<f32>().ok())
+            .map(|px| px * 0.75)
+    };
+    from_css.last().or_else(from_attr).filter(|w| w.is_finite() && *w > 1.0 && *w <= 1584.0)
+}
+
 fn attr<'a>(attrs: &'a [(String, String)], name: &str) -> Option<&'a str> {
    attrs.iter().find(|(k, _)| k == name).map(|(_, v)| v.as_str())
 }

 impl<'r> Builder<'r> {
-    pub(super) fn new(images: ImageLoader<'r>) -> Builder<'r> {
+    pub(super) fn new(images: ImageLoader<'r>, page: bool) -> Builder<'r> {
        Builder {
            images,
            containers: vec![Vec::new()],
@@ -215,11 +545,19 @@ impl<'r> Builder<'r> {
            pre_tail: false,
            meta: Meta::default(),
            title: None,
+            sheet: Sheet::default(),
+            page,
+            css_work: MAX_CSS_WORK,
        }
    }

    fn fmt(&self) -> Fmt {
-        self.stack.last().map(|e| e.fmt.clone()).unwrap_or_default()
+        match self.stack.last() {
+            Some(e) => e.fmt.clone(),
+            // A web page's text is Times New Roman 12 pt unless its CSS says otherwise.
+            None if self.page => Fmt { font: Some(PAGE_FONT.into()), size: Some(12.0), ..Fmt::default() },
+            None => Fmt::default(),
+        }
    }

    fn preserve(&self) -> bool {
@@ -237,6 +575,15 @@ impl<'r> Builder<'r> {
            let mut list_level: Option<(bool, usize)> = None;
            let mut li_at: Option<usize> = None;
            let mut lists = 0usize;
+            let mut spaced = false;
+            // A page's first span stating margins (see below).
+            let span = self
+                .stack
+                .iter()
+                .rev()
+                .take_while(|e| e.kind == ElKind::Inline)
+                .find(|e| e.spacing.before.is_some() || e.spacing.after.is_some())
+                .map(|e| (e.spacing.before, e.spacing.after));
            for (i, e) in self.stack.iter().enumerate().rev() {
                match e.kind {
                    ElKind::Cell => {
@@ -270,6 +617,27 @@ impl<'r> Builder<'r> {
                if p.align.is_none() && e.align.is_some() && e.kind != ElKind::Inline {
                    p.align = e.align;
                }
+                // Margins belong to the paragraph's own block, not to those around it.
+                if !spaced && e.kind != ElKind::Inline {
+                    p.spacing = e.spacing;
+                    if self.page {
+                        // Word opens a page's paragraphs with single lines; a margin a span
+                        // states replaces the one its paragraph states, and one the paragraph
+                        // leaves out is automatic spacing.
+                        let (top, bottom) = span.unwrap_or_default();
+                        let sp = &mut p.spacing;
+                        sp.before_auto = sp.before.is_none() && e.auto_margins;
+                        sp.after_auto = sp.after.is_none() && e.auto_margins;
+                        sp.before = sp.before.map(|own| top.unwrap_or(own));
+                        sp.after = sp.after.map(|own| bottom.unwrap_or(own));
+                        sp.line = sp.line.or(Some(LineSpacing::Multiple(1.0)));
+                    }
+                    spaced = true;
+                }
+            }
+            // Page text without a block of its own (straight in a cell) is single-spaced too.
+            if self.page && !spaced {
+                p.spacing.line = Some(LineSpacing::Multiple(1.0));
            }
            if let (Some((o, _)), Some(li)) = (list_level, li_at) {
                p.list = Some(ListInfo { ordered: o, level: lists.saturating_sub(1).min(8) as u8, start: 1 });
@@ -310,7 +678,11 @@ impl<'r> Builder<'r> {
            }
        }
        if p.inlines.len() == 1 && p.text() == "\u{00A0}" {
-            // `<p>&nbsp;</p>`: an empty paragraph on purpose.
+            // `<p>&nbsp;</p>`: an empty paragraph on purpose, whose line is as high as the
+            // space's formatting makes it.
+            if let Some(Inline::Text(_, f)) = p.inlines.first() {
+                p.mark = Some(Box::new(f.clone()));
+            }
            p.inlines.clear();
            self.para_cont = false;
            self.container().push(FBlock::Para(p));
@@ -482,6 +854,10 @@ impl<'r> Builder<'r> {
                    let mut ft = FTable { rows, widths: Vec::new(), borderless };
                    if !ft.rows.is_empty() {
                        ft.insert_covered();
+                        // Columns keep the widths their cells give when every column has one.
+                        if ft.cols() <= MAX_TABLE_COLS {
+                            ft.widths_from_cells();
+                        }
                        self.container().push(FBlock::Table(ft));
                    }
                }
@@ -625,13 +1001,15 @@ impl<'r> Builder<'r> {
        true
    }

-    fn element_style(&self, name: &str, attrs: &[(String, String)], kind: ElKind) -> ElementStyle {
+    /// `css`: the element's declarations in cascade order ([`Builder::css_of`]).
+    fn element_style(&self, name: &str, attrs: &[(String, String)], kind: ElKind, css: &str) -> ElementStyle {
        let mut fmt = self.fmt();
        let mut align = None;
        let mut preserve = self.preserve();
        let mut page_break = false;
        let mut page_break_after = false;
        match name {
+            // A header cell is bold, as in browsers and Word, unless its own CSS says otherwise.
            "b" | "strong" | "th" => fmt.bold = true,
            "i" | "em" | "cite" | "var" | "dfn" | "address" => fmt.italic = true,
            "u" | "ins" => fmt.underline = true,
@@ -662,9 +1040,16 @@ impl<'r> Builder<'r> {
        if name == "center" {
            align = Some(Align::Center);
        }
-        if let Some(style) = attr(attrs, "style") {
-            apply_style(style, &mut fmt, &mut align, &mut preserve, &mut page_break, &mut page_break_after);
+        if self.page
+            && let ElKind::Heading(n) = kind
+        {
+            // A page's headings are bold, black (unless they inherit a colour) and sized as
+            // browsers show them.
+            fmt.bold = true;
+            fmt.color.get_or_insert(wordcraft_doc::Rgb(0, 0, 0));
+            fmt.size = PAGE_HEADING_SIZES.get(usize::from(n.clamp(1, 6)) - 1).copied();
        }
+        apply_style(css, &mut fmt, &mut align, &mut preserve, &mut page_break, &mut page_break_after);
        if kind == ElKind::Pre {
            preserve = true;
        }
@@ -703,9 +1088,10 @@ impl<'r> Builder<'r> {
        if kind != ElKind::Inline {
            self.flush();
        }
-        let ElementStyle { mut fmt, align, preserve, page_break, page_break_after } = self.element_style(name, attrs, kind);
+        let css = self.css_of(name, attrs);
+        let ElementStyle { mut fmt, align, preserve, page_break, page_break_after } = self.element_style(name, attrs, kind, &css);
        match kind {
-            ElKind::Table => self.tables.push(TableB { rows: Vec::new(), row: None, bordered: declares_border(attrs, true) }),
+            ElKind::Table => self.tables.push(TableB { rows: Vec::new(), row: None, bordered: declares_border(attrs, &css, true) }),
            ElKind::Tr => {
                if let Some(t) = self.tables.last_mut()
                    && let Some(r) = t.row.take()
@@ -716,7 +1102,7 @@ impl<'r> Builder<'r> {
            ElKind::Cell => {
                let tb = self.find_open(|e| e.kind == ElKind::Tr, |e| e.kind == ElKind::Table).is_none();
                if let Some(t) = self.tables.last_mut()
-                    && declares_border(attrs, false)
+                    && declares_border(attrs, &css, false)
                {
                    t.bordered = true;
                }
@@ -728,6 +1114,7 @@ impl<'r> Builder<'r> {
                    t.row = Some(Vec::new());
                }
                let num = |k: &str| attr(attrs, k).and_then(|v| v.trim().parse::<u32>().ok()).map(|v| v.clamp(1, 63));
+                let colspan = num("colspan").unwrap_or(1);
                let shading = attr(attrs, "bgcolor").and_then(parse_color).or(fmt.background);
                let mut valign = attr(attrs, "valign").and_then(|v| match v.trim().to_ascii_lowercase().as_str() {
                    "top" => Some(VAlign::Top),
@@ -735,36 +1122,35 @@ impl<'r> Builder<'r> {
                    "bottom" => Some(VAlign::Bottom),
                    _ => None,
                });
-                if let Some(st) = attr(attrs, "style") {
-                    for decl in st.split(';') {
-                        if let Some((k, v)) = decl.split_once(':')
-                            && k.trim().eq_ignore_ascii_case("vertical-align")
-                        {
-                            match v.trim().to_ascii_lowercase().as_str() {
-                                "top" => valign = Some(VAlign::Top),
-                                "middle" | "center" => valign = Some(VAlign::Center),
-                                "bottom" => valign = Some(VAlign::Bottom),
-                                _ => {}
-                            }
-                        }
+                for (_, v) in decls(&css).filter(|(k, _)| k == "vertical-align") {
+                    match v.to_ascii_lowercase().as_str() {
+                        "top" => valign = Some(VAlign::Top),
+                        "middle" | "center" => valign = Some(VAlign::Center),
+                        "bottom" => valign = Some(VAlign::Bottom),
+                        _ => {}
                    }
                }
                self.cells.push(Cell {
                    blocks: Vec::new(),
-                    colspan: num("colspan").unwrap_or(1),
+                    colspan,
                    rowspan: attr(attrs, "rowspan").and_then(|v| v.trim().parse::<u32>().ok()).map(|v| v.clamp(1, 1000)).unwrap_or(1),
                    covered: false,
                    header: name == "th",
                    shading,
-                    borders: None,
+                    borders: cell_borders(&css),
                    valign,
+                    width: if colspan == 1 { cell_width(&css, attrs) } else { None },
                });
                fmt.background = None;
                self.containers.push(Vec::new());
            }
            _ => {}
        }
-        self.stack.push(El { name: name.to_string(), kind, fmt, align, preserve, page_break, page_break_after, used: false });
+        // The page's own margins (`body`, `html`) aren't its paragraphs'.
+        let spacing =
+            if (kind != ElKind::Inline || self.page) && !matches!(name, "body" | "html") { block_spacing(&css) } else { Spacing::default() };
+        let auto_margins = self.page && (name == "p" || matches!(kind, ElKind::Heading(_)));
+        self.stack.push(El { name: name.to_string(), kind, fmt, spacing, auto_margins, align, preserve, page_break, page_break_after, used: false });
    }

    pub(super) fn end(&mut self, name: &str) {
@@ -1,3 +1,5 @@
+use super::builder::decls;
+use super::import::PAGE_FONT;
 use super::{Align, Fmt, model, parse_color};

 /// A CSS length (or font-size keyword) in points; `None` when it is not one.
@@ -33,7 +35,7 @@ pub fn css_len_pt(v: &str) -> Option<f32> {
    }
 }

-fn strip_important(v: &str) -> &str {
+pub(super) fn strip_important(v: &str) -> &str {
    let s = v.trim();
    if let Some(pos) = s.rfind('!') {
        let after = s[pos + 1..].trim();
@@ -53,10 +55,7 @@ pub(super) fn apply_style(
    page_break: &mut bool,
    page_break_after: &mut bool,
 ) {
-    for decl in css.split(';').take(64) {
-        let Some((k, v)) = decl.split_once(':') else { continue };
-        let k = k.trim().to_ascii_lowercase();
-        let v = strip_important(v);
+    for (k, v) in decls(css) {
        let lv = v.to_ascii_lowercase();
        match k.as_str() {
            "font-weight" => f.bold = lv.starts_with("bold") || lv.parse::<u32>().is_ok_and(|w| w >= 600),
@@ -94,6 +93,14 @@ pub(super) fn apply_style(
                    f.code = true;
                } else if !fam.is_empty() && !matches!(fam.to_ascii_lowercase().as_str(), "serif" | "sans-serif" | "inherit" | "initial") {
                    f.font = model::kept_font(&fam).or(f.font.take());
+                } else if f.font.is_some() {
+                    // A generic family replaces a face the text has (a page's Times New Roman);
+                    // text without one keeps the document's font.
+                    match fam.to_ascii_lowercase().as_str() {
+                        "serif" => f.font = Some(PAGE_FONT.into()),
+                        "sans-serif" => f.font = Some("Arial".into()),
+                        _ => {}
+                    }
                }
            }
            "vertical-align" => {
@@ -274,17 +274,39 @@ fn sniff_charset(head: &str) -> Option<&str> {

 /// Decode HTML bytes: BOM, then a `charset` in the first 2 KB, else UTF-8 with a legacy fallback.
 pub fn decode(b: &[u8]) -> String {
+    decode_as(b, None)
+}
+
+/// [`decode`] in the `charset` its container states (a MIME part's, a content type's), when it
+/// states one. Single-byte Windows code pages (`windows-1250`…, `iso-8859-1`, `us-ascii` as 1252)
+/// are decoded as such, other labels by their WHATWG decoders.
+pub fn decode_as(b: &[u8], charset: Option<&str>) -> String {
    if b.starts_with(&[0xFF, 0xFE]) || b.starts_with(&[0xFE, 0xFF]) || b.starts_with(&[0xEF, 0xBB, 0xBF]) {
        return crate::txt::decode(b);
    }
    let head = String::from_utf8_lossy(b.get(..b.len().min(2048)).unwrap_or(b)).to_ascii_lowercase();
-    if let Some(label) = sniff_charset(&head)
-        && let Some(s) = crate::txt::decode_as(b, label)
-    {
-        return s;
+    let named = charset.map(str::to_ascii_lowercase).or_else(|| sniff_charset(&head).map(str::to_string));
+    let code_page = match named.as_deref().map(str::trim) {
+        Some("iso-8859-1" | "latin1" | "us-ascii") => Some(1252),
+        Some(cs) => cs
+            .strip_prefix("windows-")
+            .or_else(|| cs.strip_prefix("cp"))
+            .and_then(|n| n.parse::<u16>().ok())
+            .filter(|cp| wordcraft_doc::encoding::SINGLE_BYTE_CODE_PAGES.contains(cp)),
+        None => None,
+    };
+    // Other labels (ISO 8859, KOI8, Shift_JIS, GBK, Big5, EUC-KR…) by their WHATWG decoders; a
+    // UTF-16 one without a BOM means UTF-8, as in browsers.
+    let other = named
+        .as_deref()
+        .and_then(|l| encoding_rs::Encoding::for_label(l.trim().as_bytes()))
+        .filter(|e| ![encoding_rs::UTF_8, encoding_rs::UTF_16LE, encoding_rs::UTF_16BE].contains(e));
+    match (code_page, other) {
+        (Some(cp), _) => b.iter().map(|&c| wordcraft_doc::encoding::windows_byte(cp, c)).collect(),
+        (None, Some(enc)) => enc.decode_without_bom_handling(b).0.into_owned(),
+        (None, None) => match std::str::from_utf8(b) {
+            Ok(s) => s.to_string(),
+            Err(_) => crate::txt::decode(b),
+        },
    }
-    if let Ok(s) = std::str::from_utf8(b) {
-        return s.to_string();
-    }
-    crate::txt::decode(b)
 }
@@ -1,4 +1,4 @@
-use super::builder::{Builder, ImageLoader};
+use super::builder::{Builder, ImageLoader, parse_sheet};
 use super::lexer::{Lexer, Tok};
 use super::math_ruby::{read_math, read_ruby};
 use super::{Document, Flow, Inline, model};
@@ -10,6 +10,10 @@ pub fn parse(s: &str) -> Flow {

 /// [`parse`], loading pictures that aren't `data:` URIs through `images`.
 pub fn parse_with(s: &str, images: ImageLoader) -> Flow {
+    parse_in(s, images, false)
+}
+
+pub(super) fn parse_in(s: &str, images: ImageLoader, page: bool) -> Flow {
    // HTML normalizes input newlines before tokenization, not after decoding character references.
    let normalized = if s.contains('\r') {
        let mut out = String::with_capacity(s.len());
@@ -23,8 +27,10 @@ pub fn parse_with(s: &str, images: ImageLoader) -> Flow {
    } else {
        std::borrow::Cow::Borrowed(s)
    };
-    let mut lx = Lexer { s: &normalized, pos: 0 };
-    let mut b = Builder::new(images);
+    // The lexer's offsets (a style element's text is cut out by them) are into the normalized text.
+    let s: &str = &normalized;
+    let mut lx = Lexer { s, pos: 0 };
+    let mut b = Builder::new(images, page);
    let mut guard = 0usize;
    while let Some(t) = lx.next() {
        guard += 1;
@@ -52,7 +58,11 @@ pub fn parse_with(s: &str, images: ImageLoader) -> Flow {
                    continue;
                }
                if matches!(name.as_str(), "script" | "style" | "noscript" | "template" | "textarea") && !self_close {
-                    lx.skip_past_end_tag(&name);
+                    let from = lx.pos;
+                    let to = lx.skip_past_end_tag(&name);
+                    if name == "style" {
+                        parse_sheet(s.get(from..to).unwrap_or(""), &mut b.sheet);
+                    }
                    continue;
                }
                b.start(&name, &attrs, self_close);
@@ -91,9 +101,25 @@ pub fn import_with(s: &str, images: ImageLoader) -> Document {
    model::to_doc(&parse_with(s, images))
 }

-/// [`import_with`] onto `page` (its size and margins), which tables size to: HTML gives no page
+/// A web page opened as a document, as Word opens one (a file, or a document's HTML
+/// `w:altChunk`): its text is Times New Roman 12 pt on single lines, headings bold and black at
+/// the sizes browsers give them, and a `p` or heading whose CSS leaves a margin out (or writes it
+/// invalidly) gets HTML's automatic spacing there; a margin a span states replaces the one its
+/// paragraph states (both as Word's import does). [`import`] reads a pasted fragment, which takes
+/// the document's own formatting instead (but its CSS still applies).
+pub fn import_page(s: &str, images: ImageLoader) -> Document {
+    model::begin_import();
+    model::to_doc(&parse_in(s, images, true))
+}
+
+/// [`import_page`] onto `page` (its size and margins), which tables size to: HTML gives no page
 /// size, so the caller picks it (the region's paper, #1381).
 pub fn import_on(s: &str, images: ImageLoader, page: &wordcraft_doc::SectionProps) -> Document {
    model::begin_import();
-    model::to_doc_in(&parse_with(s, images), page.clone())
+    model::to_doc_in(&parse_in(s, images, true), page.clone())
 }
+
+/// The font of a web page's text when its CSS names none.
+pub const PAGE_FONT: &str = "Times New Roman";
+/// Heading sizes on a web page, `h1`–`h6`, points.
+pub(super) const PAGE_HEADING_SIZES: [f32; 6] = [24.0, 18.0, 13.5, 12.0, 10.0, 7.5];
@@ -34,7 +34,9 @@ impl Lexer<'_> {
    }

    /// Skip past the next matching end tag `</tag...>` (ASCII case-insensitive), or to the end.
-    pub(super) fn skip_past_end_tag(&mut self, tag: &str) {
+    /// Skip past the end tag of raw text element `tag`; returns where that end tag starts (the
+    /// end of the input when there is none).
+    pub(super) fn skip_past_end_tag(&mut self, tag: &str) -> usize {
        let prefix = format!("</{tag}");
        let p = prefix.as_bytes();
        while self.pos < self.s.len() {
@@ -42,17 +44,19 @@ impl Lexer<'_> {
            let found = hay.windows(p.len().max(1)).position(|w| w.eq_ignore_ascii_case(p));
            let Some(k) = found else {
                self.pos = self.s.len();
-                return;
+                return self.pos;
            };
            let after_tag = k + p.len();
            let next_byte = hay.get(after_tag);
            if next_byte.is_none_or(|b| b.is_ascii_whitespace() || *b == b'>' || *b == b'/') {
+                let at = self.pos + k;
                self.pos += after_tag;
                self.skip_past(">");
-                return;
+                return at;
            }
            self.pos += k + 2;
        }
+        self.s.len()
    }

    pub(super) fn next(&mut self) -> Option<Tok> {
@@ -7,6 +7,27 @@ fn numeric_entity_rejects_sign() {
    assert_eq!(unescape("a&#65;&#x41;b"), "aAAb");
 }

+#[test]
+fn important_declarations_win_the_cascade() {
+    let color = |html: &str| fmt_of(paras(&parse(html))[0]).color;
+    let red = Some(wordcraft_doc::Rgb(255, 0, 0));
+    let blue = Some(wordcraft_doc::Rgb(0, 0, 255));
+    assert_eq!(color("<style>p { color: red }</style><p style='color: blue'>x</p>"), blue);
+    assert_eq!(color("<style>p { color: red !important }</style><p style='color: blue'>x</p>"), red);
+    assert_eq!(color("<style>p { color: red !important }</style><p style='color: blue !important'>x</p>"), blue);
+    assert_eq!(color("<p style='color: red !important; color: blue'>x</p>"), red);
+}
+
+#[test]
+fn a_declared_charset_decodes_the_page() {
+    // ISO 8859-2 (Central European): 0xA9 Š, 0xB9 š; named by the caller or the page.
+    assert_eq!(decode_as(&[0xA9, b' ', 0xB9], Some("iso-8859-2")), "Š š");
+    assert!(decode_as(b"<meta charset=iso-8859-2><p>\xA9</p>", None).contains("<p>Š</p>"));
+    assert_eq!(decode_as(&[0x8A], Some("windows-1252")), "Š");
+    // A page that says UTF-16 without a BOM is read as UTF-8, as browsers do.
+    assert_eq!(decode_as("é".as_bytes(), Some("utf-16")), "é");
+}
+
 fn texts(f: &Flow) -> Vec<String> {
    f.blocks
        .iter()
@@ -33,6 +54,9 @@ fn borderless_tables_get_no_grid_style() {
    assert_eq!(table_style("<table border=1><tr><td>a</table>").as_deref(), Some("TableGrid"));
    assert_eq!(table_style("<table style='border:1px solid #000'><tr><td>a</table>").as_deref(), Some("TableGrid"));
    assert_eq!(table_style("<table><tr><td style='border:1px solid #999'>a</table>").as_deref(), Some("TableGrid"));
+    // The last declaration decides: a `style` attribute's over the style sheet's.
+    assert_eq!(table_style("<style>table { border: 1px solid }</style><table><tr><td>a</table>").as_deref(), Some("TableGrid"));
+    assert_eq!(table_style("<style>table { border: 1px solid }</style><table style='border:none'><tr><td>a</table>").as_deref(), Some(""));
 }

 #[test]
@@ -522,3 +546,171 @@ fn empty_comments_do_not_swallow_the_rest() {
        assert_eq!(doc.plain_text(Default::default()), "ab\nc", "{html}");
    }
 }
+
+fn paras(f: &Flow) -> Vec<&Para> {
+    f.blocks.iter().filter_map(|b| if let FBlock::Para(p) = b { Some(p) } else { None }).collect()
+}
+fn fmt_of(p: &Para) -> Fmt {
+    p.inlines.iter().find_map(|i| if let Inline::Text(_, f) = i { Some(f.clone()) } else { None }).unwrap_or_default()
+}
+
+#[test]
+fn style_sheet_rules_apply_in_cascade_order() {
+    let f = parse(
+        "<style><!-- p { color: #00ff00; font-size: 10pt } .big { font-size: 20pt } p.big { font-weight: bold } \
+         /* a comment { } */ @media print { p { color: #ff0000 } } div p, li:hover { font-style: italic } .000003 { text-align: center } --></style>\
+         <p>plain</p><p class='big x'>big</p><p class=big style='font-size: 8pt'>inline wins</p><p class=000003>digits</p>",
+    );
+    let ps = paras(&f);
+    let (plain, big, inline, digits) = (fmt_of(ps[0]), fmt_of(ps[1]), fmt_of(ps[2]), fmt_of(ps[3]));
+    assert_eq!((plain.color, plain.size, plain.bold, plain.italic), (Some(wordcraft_doc::Rgb(0, 255, 0)), Some(10.0), false, false));
+    assert_eq!((big.size, big.bold), (Some(20.0), true));
+    assert_eq!(inline.size, Some(8.0));
+    assert_eq!(ps[3].align, Some(Align::Center));
+    assert_eq!(digits.size, Some(10.0));
+}
+
+#[test]
+fn paragraph_css_states_spacing_indents_and_line_height() {
+    let f = parse(
+        "<style>p.a { margin: 6pt 1in 3pt; text-indent: -18pt; line-height: 150% }</style>\
+         <p class=a>a</p><p style='margin-top: 14,4px; margin-bottom: auto; line-height: 20pt'>b</p><p>c</p>",
+    );
+    let ps = paras(&f);
+    assert_eq!(
+        ps[0].spacing,
+        Spacing {
+            before: Some(6.0),
+            after: Some(3.0),
+            left: Some(72.0),
+            right: Some(72.0),
+            first: Some(-18.0),
+            line: Some(LineSpacing::Multiple(1.5)),
+            ..Spacing::default()
+        }
+    );
+    // Values that don't parse (a comma decimal, `auto`) state nothing.
+    assert_eq!(ps[1].spacing, Spacing { line: Some(LineSpacing::Exactly(20.0)), ..Spacing::default() });
+    assert_eq!(ps[2].spacing, Spacing::default());
+}
+
+#[test]
+fn a_page_opens_as_word_opens_one() {
+    let html = "<style>p.k { margin-bottom: 13px } span.z { margin: 0in }</style>\
+                <p>first</p><h1>Head</h1><p class=k><span class=z>kept</span></p><p class=k>own</p><div>bare</div><p>&nbsp;</p>";
+    let f = parse_in(html, &|_| None, true);
+    let ps = paras(&f);
+    // Times New Roman 12 pt on single lines; headings bold and black at a browser's sizes.
+    let first = fmt_of(ps[0]);
+    assert_eq!((first.font.as_deref(), first.size), (Some(PAGE_FONT), Some(12.0)));
+    assert_eq!(ps[0].spacing.line, Some(LineSpacing::Multiple(1.0)));
+    let cell = parse_in("<table><tr><td>cell</td></tr></table>", &|_| None, true);
+    let FBlock::Table(t) = &cell.blocks[0] else { panic!() };
+    let FBlock::Para(cp) = &t.rows[0][0].blocks[0] else { panic!() };
+    assert_eq!(cp.spacing.line, Some(LineSpacing::Multiple(1.0)));
+    let head = fmt_of(ps[1]);
+    assert_eq!((head.bold, head.size, head.color), (true, Some(24.0), Some(wordcraft_doc::Rgb(0, 0, 0))));
+    // A colour the heading inherits stays.
+    // A generic family names a face on a page; pasted text keeps the document's font.
+    let generic = parse_in("<p style='font-family: sans-serif'>a</p><p style='font-family: serif'>b</p>", &|_| None, true);
+    let fonts: Vec<Option<String>> = paras(&generic).iter().map(|p| fmt_of(p).font.map(|f| f.to_string())).collect();
+    assert_eq!(fonts, [Some("Arial".to_string()), Some(PAGE_FONT.to_string())]);
+    assert_eq!(fmt_of(paras(&parse("<p style='font-family: sans-serif'>a</p>"))[0]).font, None);
+    let inherited = parse_in("<body style='color: #000080'><h1>Head</h1></body>", &|_| None, true);
+    assert_eq!(fmt_of(paras(&inherited)[0]).color, Some(wordcraft_doc::Rgb(0, 0, 0x80)));
+    // Automatic spacing where a p or heading leaves its margins out (layout drops it at the
+    // top of the page and of a cell).
+    let auto = |p: &Para| (p.spacing.before_auto, p.spacing.after_auto, p.spacing.before, p.spacing.after);
+    assert_eq!(auto(ps[0]), (true, true, None, None));
+    assert_eq!(auto(ps[1]), (true, true, None, None));
+    // A span's margin replaces the one its paragraph states; the one it leaves out stays automatic.
+    assert_eq!(auto(ps[2]), (true, false, None, Some(0.0)));
+    assert_eq!(auto(ps[3]), (true, false, None, Some(9.75)));
+    // Other blocks get no automatic spacing.
+    assert_eq!(auto(ps[4]), (false, false, None, None));
+    // The document says so, and the empty paragraph's mark keeps its size.
+    let d = model::to_doc(&f);
+    let props = |i: usize| match &*d.body[i] {
+        wordcraft_doc::Block::Para(p) => p.clone(),
+        b => panic!("{b:?}"),
+    };
+    assert_eq!((props(0).props.space_before_auto, props(0).props.space_after_auto), (Some(true), Some(true)));
+    assert_eq!((props(2).props.space_after_auto, props(2).props.space_after), (None, Some(0.0)));
+    assert_eq!(props(5).mark.size, Some(12.0));
+    // `<p>&nbsp;</p>` keeps the size of its space for its line.
+    assert!(ps[5].inlines.is_empty());
+    assert_eq!(ps[5].mark.as_ref().and_then(|m| m.size), Some(12.0));
+    // A fragment (a paste) keeps the document's own formatting.
+    let pasted = parse(html);
+    assert_eq!(fmt_of(paras(&pasted)[0]).font, None);
+    assert_eq!(paras(&pasted)[0].spacing, Spacing::default());
+}
+
+#[test]
+fn cells_give_their_columns_widths() {
+    let f = parse(
+        "<style>td.n { width: 42px }</style><table border=1><tr><td colspan=3>t</td></tr><tr><td class=n>1</td><td width=200>a</td><td style='width: 2in'>b</td></tr></table>",
+    );
+    let FBlock::Table(t) = &f.blocks[0] else { panic!("{:?}", f.blocks) };
+    assert_eq!(t.widths, [31.5, 150.0, 144.0]);
+    // A column without a width leaves them all to the layout.
+    let f = parse("<table><tr><td width=100>a</td><td>b</td></tr></table>");
+    let FBlock::Table(t) = &f.blocks[0] else { panic!() };
+    assert!(t.widths.is_empty());
+    // A cell beside a cell spanning rows from above gives its own column its width.
+    let f =
+        parse("<table><tr><td rowspan=2>a</td><td width=200>b</td></tr><tr><td width=300>c</td></tr><tr><td width=100>d</td><td>e</td></tr></table>");
+    let FBlock::Table(t) = &f.blocks[0] else { panic!() };
+    assert_eq!(t.widths, [75.0, 150.0]);
+}
+
+#[test]
+fn cells_take_their_css_borders() {
+    use wordcraft_doc::props::{Border, BorderStyle};
+    let f = parse(
+        "<style>td.b { border-top: solid windowtext 1px; border-bottom: 2pt double #ff0000; border-left: none; border-right: 0px solid; vertical-align: bottom }</style><table><tr><td class=b>a</td><td>b</td></tr></table>",
+    );
+    let FBlock::Table(t) = &f.blocks[0] else { panic!() };
+    // The style sheet's other cell properties apply too.
+    assert_eq!(t.rows[0][0].valign, Some(wordcraft_doc::props::VAlign::Bottom));
+    let b = t.rows[0][0].borders.expect("borders");
+    assert_eq!(b.top, Some(Border::single(0.75)));
+    assert_eq!(b.bottom, Some(Border { style: BorderStyle::Double, width: 2.0, color: Some(wordcraft_doc::Rgb(255, 0, 0)), space: 0.0 }));
+    assert_eq!(b.left.map(|l| l.style), Some(BorderStyle::None));
+    // A border of no width is none (OOXML has no zero-width line).
+    assert_eq!(b.right.map(|r| r.style), Some(BorderStyle::None));
+    assert_eq!(t.rows[0][1].borders, None);
+}
+
+#[test]
+fn style_sheet_edge_cases() {
+    let f = parse_in(
+        "<style>@charset \"utf-8\"; @import url(a.css); p.a { color: #0000ff } body { margin: 20px } p.b { margin: 10 5% medium }</style>\
+         <body><p class=a>a</p>bare<p class=b>b</p></body>",
+        &|_| None,
+        true,
+    );
+    let ps = paras(&f);
+    // A rule right after statement at-rules still applies.
+    assert_eq!(fmt_of(ps[0]).color, Some(wordcraft_doc::Rgb(0, 0, 255)));
+    // The body's margins aren't the paragraphs' indents.
+    assert_eq!((ps[1].spacing.left, ps[1].spacing.before), (None, None));
+    // Bare numbers, percentages and keywords state no margin: automatic spacing stays.
+    assert_eq!((ps[2].spacing.before_auto, ps[2].spacing.left), (true, None));
+    // The `style` attribute wins even over a sheet with more declarations than are read.
+    let many = (0..600).map(|i| format!("x-{i}: 1")).collect::<Vec<_>>().join(";");
+    let f = parse(&format!("<style>p {{ {many}; color: #ff0000 }}</style><p style='color: #00ff00'>c</p>"));
+    assert_eq!(fmt_of(paras(&f)[0]).color, Some(wordcraft_doc::Rgb(0, 255, 0)));
+}
+
+#[test]
+fn hostile_style_sheets_never_panic() {
+    // One declaration block shared by many selectors isn't copied for each.
+    let selectors = vec!["p"; 20_000].join(",");
+    let huge = format!("<style>{selectors} {{ color: red; {} }}</style><p>x</p>", "x: y;".repeat(200_000));
+    let _ = parse_in(&huge, &|_| None, true);
+    let many = ".a { color: red } ".repeat(20_000);
+    for css in ["{", "}", "p {", "@media { p { x } ", "p { color: red", "/* open", ".{}", "p..a{}", "\u{0}.é{x:y}", many.as_str()] {
+        let _ = parse_in(&format!("<style>{css}</style><p class='a é'>x</p><td style='width: 1e40px'>y</td>"), &|_| None, true);
+    }
+}
@@ -430,6 +430,8 @@ pub fn escape_text(s: &str) -> String {
            '>' => o.push_str("\\textgreater{}"),
            '|' => o.push_str("\\textbar{}"),
            '\u{A0}' => o.push('~'),
+            // TeX collapses a run of spaces into one; a control space keeps each further one.
+            ' ' if prev == ' ' => o.push_str("\\ "),
            // An opening bracket here would become an optional item or line-break argument.
            '[' if !line_has_text => o.push_str("{[}"),
            // Keep TeX from joining `--`, ``` `` ``` and `''` into dashes and quotes.
@@ -485,7 +485,11 @@ impl<'a> Parser<'a> {

    fn command(&mut self, name: &str) {
        if let Some(sym) = text_symbol(name) {
-            if name == "item" {
+            if name == " " {
+                // A control space is never collapsed into the space before it.
+                let f = self.fmt();
+                self.para.push_text(sym, &f);
+            } else if name == "item" {
                self.item();
            } else if matches!(name, "centering" | "raggedleft" | "raggedright") {
                let a = match name {
@@ -1,6 +1,7 @@
 //! WordCraft import/export for the "other" text formats: plain text, Markdown (CommonMark subset
-//! plus GFM tables and strikethrough), HTML, RTF, OpenDocument Text and LaTeX — and, read-only,
-//! spreadsheet cells for mail-merge recipient lists ([`sheet`]: .xlsx/.xlsm/.xls/.ods).
+//! plus GFM tables and strikethrough), HTML (and MHT web archives, read-only), RTF, OpenDocument
+//! Text and LaTeX — and, read-only, spreadsheet cells for mail-merge recipient lists ([`sheet`]:
+//! .xlsx/.xlsm/.xls/.ods).
 //!
 //! Every format maps to a small flow model ([`model::Flow`]: paragraphs with a kind, list
 //! membership, alignment and formatted spans; tables of cells), which [`model::to_doc`] and
@@ -14,6 +15,7 @@ pub mod html;
 pub mod latex;
 pub mod markdown;
 pub mod mathml;
+pub mod mht;
 pub mod model;
 pub mod odt;
 pub mod rtf;
@@ -22,8 +24,12 @@ pub mod txt;

 use wordcraft_doc::Document;

+/// Most entries an ODT or spreadsheet package may declare, as many as a Word package
+/// (`wordcraft_docx::package::MAX_ENTRIES`); checked before `zip` parses them (#3660).
+pub const MAX_ZIP_ENTRIES: u64 = 20_000;
+
 /// Extensions this crate handles.
-pub const EXTENSIONS: &[&str] = &["txt", "text", "md", "markdown", "html", "htm", "xhtml", "rtf", "odt", "tex", "latex", "ltx"];
+pub const EXTENSIONS: &[&str] = &["txt", "text", "md", "markdown", "html", "htm", "xhtml", "mht", "mhtml", "rtf", "odt", "tex", "latex", "ltx"];

 /// Import `bytes` in the format named by `ext` (no dot, any case). `None` = not our format.
 pub fn import(ext: &str, bytes: &[u8]) -> Option<Result<Document, String>> {
@@ -31,7 +37,8 @@ pub fn import(ext: &str, bytes: &[u8]) -> Option<Result<Document, String>> {
    let r = match ext.as_str() {
        "txt" | "text" => Ok(txt::import(bytes)),
        "md" | "markdown" => Ok(markdown::import(&txt::decode(bytes))),
-        "html" | "htm" | "xhtml" => Ok(html::import(&html::decode(bytes))),
+        "html" | "htm" | "xhtml" => Ok(html::import_page(&html::decode(bytes), &|_| None)),
+        "mht" | "mhtml" => Ok(mht::import(bytes)),
        "rtf" => rtf::import(bytes),
        "odt" => odt::import(bytes),
        "tex" | "latex" | "ltx" => Ok(latex::import(bytes)),
@@ -44,7 +51,7 @@ pub fn import(ext: &str, bytes: &[u8]) -> Option<Result<Document, String>> {
 }

 /// [`import`] onto `page` (its size and margins) for the formats whose files give no page size:
-/// plain text, Markdown and HTML start on it, and their tables size to it (#1381). The other
+/// plain text, Markdown, HTML and MHT start on it, and their tables size to it (#1381). The other
 /// formats keep the page their file gives.
 pub fn import_on(ext: &str, bytes: &[u8], page: &wordcraft_doc::SectionProps) -> Option<Result<Document, String>> {
    let ext = ext.trim_start_matches('.').to_ascii_lowercase();
@@ -52,6 +59,7 @@ pub fn import_on(ext: &str, bytes: &[u8], page: &wordcraft_doc::SectionProps) ->
        "txt" | "text" => txt::import_on(bytes, page),
        "md" | "markdown" => markdown::import_on(&txt::decode(bytes), page),
        "html" | "htm" | "xhtml" => html::import_on(&html::decode(bytes), &|_| None, page),
+        "mht" | "mhtml" => mht::import_on(bytes, page),
        _ => return import(&ext, bytes),
    };
    doc.ensure_nonempty();
@@ -380,9 +380,16 @@ pub(super) fn write_blocks(blocks: &[FBlock], out: &mut Vec<String>) {
                            // The bullet marker can turn a short run of hyphens into a rule.
                            if thematic_break(&format!("{marker} {body}")) { format!("\\{body}") } else { body }
                        };
-                        out.push(format!("{pad}{marker} {body}").trim_end().to_string());
+                        let line = format!("{pad}{marker} {body}").trim_end().to_string();
+                        // A list inside a block quote stays in it: every line of the item, its
+                        // continuation lines included, takes the quote marker.
+                        let line = if p.kind == Kind::Quote { format!("> {}", line.replace('\n', "\n> ")) } else { line };
+                        out.push(line);
                        prev_list = true;
-                        let next_is_list = matches!(blocks.get(i + 1), Some(FBlock::Para(q)) if q.list.is_some());
+                        // A list going into or out of a block quote is a new block: without the
+                        // blank line, an unquoted item would be read back into the quote.
+                        let quoted = |k: Kind| k == Kind::Quote;
+                        let next_is_list = matches!(blocks.get(i + 1), Some(FBlock::Para(q)) if q.list.is_some() && quoted(q.kind) == quoted(p.kind));
                        if !next_is_list {
                            out.push(String::new());
                        }
@@ -558,6 +558,25 @@ fn ordered_list_keeps_start_number() {
    assert_eq!(export(&import("5. a\n6. b")).trim_end(), "5. a\n6. b");
 }

+/// A list inside a block quote stays in the quote: its items were imported as quoted list
+/// paragraphs but written back without `>`, so the list moved out of the quote.
+#[test]
+fn a_list_inside_a_quote_stays_quoted() {
+    let src = "> Intro\n>\n> - one\n> - two\n>   - nested\n> 1. first\n\nAfter.\n";
+    let want = "> Intro\n\n> - one\n> - two\n>     - nested\n> 1. first\n\nAfter.";
+    let md = export(&import(src));
+    assert_eq!(md.trim_end(), want);
+    assert_eq!(export(&import(&md)).trim_end(), want, "and stays so when read back");
+    // An item's continuation line is quoted too.
+    assert_eq!(export(&import("> - a\\\n>   b\n")).trim_end(), "> - a\\\n>   b");
+    // A list outside a quote is unchanged.
+    assert_eq!(export(&import("- one\n- two\n")).trim_end(), "- one\n- two");
+    // A list after a quoted list stays outside the quote.
+    let md = export(&import("> - a\n\n2. b\n"));
+    assert_eq!(md.trim_end(), "> - a\n\n2. b");
+    assert_eq!(export(&import(&md)).trim_end(), "> - a\n\n2. b", "and stays so when read back");
+}
+
 #[test]
 fn lines_after_line_break_stay_in_paragraph() {
    use wordcraft_doc::{CharProps, Document, Paragraph, StoryRef, para_block};
@@ -0,0 +1,444 @@
+//! MHT / MHTML (Single File Web Page): a web page and its pictures in one MIME message (RFC 2557
+//! `multipart/related`, RFC 2045 transfer encodings). The page is the part the message's `start`
+//! parameter names, else its first HTML part; it is read by [`crate::html`], with `img` sources
+//! found among the other parts by `Content-Location` (also relative to the page's) or `cid:`
+//! `Content-ID`, and the page opens as Word opens one ([`crate::html::import_page`]). Word keeps
+//! HTML `w:altChunk`s in this form too.
+
+use std::collections::HashMap;
+use std::sync::Arc;
+
+use wordcraft_doc::Document;
+
+/// Most parts read from one message (nested multiparts included).
+const MAX_PARTS: usize = 10_000;
+/// Deepest `multipart` nesting followed.
+const MAX_NESTING: usize = 8;
+/// Longest header block looked at for one part.
+const MAX_HEADER: usize = 64 * 1024;
+/// Longest boundary (RFC 2046 §5.1.1).
+const MAX_BOUNDARY: usize = 70;
+/// Picture bytes one page may load, counting every `img` that names a part.
+const PICTURE_BUDGET: usize = 1 << 30;
+
+/// One leaf part of the message: its headers (names lower-cased) and decoded body.
+struct Part {
+    content_type: String,
+    location: Option<String>,
+    id: Option<String>,
+    /// Shared by every `img` that names it.
+    body: Arc<Vec<u8>>,
+}
+
+impl Part {
+    fn mime(&self) -> String {
+        self.content_type.split(';').next().unwrap_or("").trim().to_ascii_lowercase()
+    }
+}
+
+/// Does `bytes` look like a MIME message (an `.mht` file)?
+pub fn is_mht(bytes: &[u8]) -> bool {
+    let head = String::from_utf8_lossy(bytes.get(..bytes.len().min(4096)).unwrap_or(bytes)).to_ascii_lowercase();
+    head.starts_with("mime-version:") || head.starts_with("content-type: multipart/") || head.contains("\ncontent-type: multipart/related")
+}
+
+/// Import an MHT file. A message without an HTML part reads as its text (an empty document
+/// when there's none).
+pub fn import(bytes: &[u8]) -> Document {
+    import_on(bytes, &wordcraft_doc::SectionProps::default())
+}
+
+/// [`import`] onto `page`: like HTML, an MHT file gives no page size (#1381).
+pub fn import_on(bytes: &[u8], page: &wordcraft_doc::SectionProps) -> Document {
+    let mut parts = Vec::new();
+    let (headers, body) = split_headers(bytes);
+    let start = param(header(&headers, "content-type").unwrap_or(""), "start").map(|s| s.trim_matches(['<', '>']).to_string());
+    collect(&headers, body, &mut parts, 0);
+    let root = start
+        .as_deref()
+        .and_then(|s| parts.iter().position(|p| p.id.as_deref() == Some(s)))
+        .or_else(|| parts.iter().position(|p| matches!(p.mime().as_str(), "text/html" | "application/xhtml+xml")));
+    let Some(root) = root.and_then(|i| parts.get(i)) else {
+        let text = parts.iter().find(|p| p.mime() == "text/plain").map(|p| p.body.as_slice()).unwrap_or(&[]);
+        return crate::txt::import_on(text, page);
+    };
+    let html = crate::html::decode_as(&root.body, param(&root.content_type, "charset"));
+    let base = root.location.as_deref().unwrap_or("");
+    // Parts by `Content-Location` (lower case) and `Content-ID`; the first of each wins.
+    let mut by_location = HashMap::new();
+    let mut by_id = HashMap::new();
+    for p in &parts {
+        if let Some(l) = &p.location {
+            by_location.entry(l.to_ascii_lowercase()).or_insert(p);
+        }
+        if let Some(i) = &p.id {
+            by_id.entry(i.as_str()).or_insert(p);
+        }
+    }
+    let left = std::cell::Cell::new(PICTURE_BUDGET);
+    let loader = |src: &str| -> Option<Arc<Vec<u8>>> {
+        let src = src.trim();
+        let part = match src.get(..4).filter(|s| s.eq_ignore_ascii_case("cid:")).and(src.get(4..)) {
+            Some(cid) => by_id.get(cid),
+            // By its location as written, or resolved against the page's.
+            None => by_location.get(&src.to_ascii_lowercase()).or_else(|| by_location.get(&resolve(base, src).to_ascii_lowercase())),
+        }?;
+        left.set(left.get().checked_sub(part.body.len())?);
+        Some(part.body.clone())
+    };
+    crate::html::import_on(&html, &loader, page)
+}
+
+/// `reference` resolved against the page's location `base`, as RFC 2557 §5 asks (RFC 3986
+/// §5.2.2): a reference with a scheme stays as it is, `//host/…` takes the base's scheme, `/…`
+/// its scheme and host, an empty path its path and query, anything else is merged with its path
+/// (§5.2.3); `.` and `..` segments go (§5.2.4) and fragments are dropped. Nothing is fetched.
+fn resolve(base: &str, reference: &str) -> String {
+    let (b, r) = (UriParts::of(base), UriParts::of(reference));
+    let t = if r.scheme.is_some() {
+        UriParts { path: remove_dots(&r.path), ..r }
+    } else if r.authority.is_some() {
+        UriParts { scheme: b.scheme, path: remove_dots(&r.path), ..r }
+    } else if r.path.is_empty() {
+        UriParts { query: r.query.or(b.query), ..b }
+    } else if r.path.starts_with('/') {
+        UriParts { path: remove_dots(&r.path), query: r.query, ..b }
+    } else {
+        let merged = if b.authority.is_some() && b.path.is_empty() {
+            format!("/{}", r.path)
+        } else {
+            format!("{}{}", b.path.rfind('/').and_then(|i| b.path.get(..=i)).unwrap_or(""), r.path)
+        };
+        UriParts { path: remove_dots(&merged), query: r.query, ..b }
+    };
+    t.to_string()
+}
+
+/// A URI's parts (RFC 3986 §3): `scheme:`, `//authority`, path and `?query`; the fragment left out.
+struct UriParts<'a> {
+    scheme: Option<&'a str>,
+    authority: Option<&'a str>,
+    path: String,
+    query: Option<&'a str>,
+}
+
+impl<'a> UriParts<'a> {
+    fn of(uri: &'a str) -> Self {
+        let uri = uri.split('#').next().unwrap_or("");
+        // A one-letter scheme is a Windows drive (`C:/…`), not a scheme.
+        let scheme = uri.split_once(':').map(|(k, _)| k).filter(|k| {
+            k.len() > 1 && k.starts_with(|c: char| c.is_ascii_alphabetic()) && k.chars().all(|c| c.is_ascii_alphanumeric() || "+-.".contains(c))
+        });
+        let rest = scheme.and_then(|k| uri.get(k.len() + 1..)).unwrap_or(uri);
+        let (authority, rest) = match rest.strip_prefix("//") {
+            Some(r) => {
+                let end = r.find(['/', '?']).unwrap_or(r.len());
+                (r.get(..end), r.get(end..).unwrap_or(""))
+            }
+            None => (None, rest),
+        };
+        let (path, query) = match rest.split_once('?') {
+            Some((p, q)) => (p, Some(q)),
+            None => (rest, None),
+        };
+        UriParts { scheme, authority, path: path.to_string(), query }
+    }
+}
+
+impl std::fmt::Display for UriParts<'_> {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        if let Some(s) = self.scheme {
+            write!(f, "{s}:")?;
+        }
+        if let Some(a) = self.authority {
+            write!(f, "//{a}")?;
+        }
+        f.write_str(&self.path)?;
+        if let Some(q) = self.query {
+            write!(f, "?{q}")?;
+        }
+        Ok(())
+    }
+}
+
+/// `path` without its `.` and `..` segments (RFC 3986 §5.2.4); `..` stops at the root.
+fn remove_dots(path: &str) -> String {
+    let mut out: Vec<&str> = Vec::new();
+    let segments: Vec<&str> = path.split('/').collect();
+    for (i, seg) in segments.iter().enumerate() {
+        match *seg {
+            "." => {}
+            ".." => {
+                if out.len() > 1 || out.first().is_some_and(|f| !f.is_empty()) {
+                    out.pop();
+                }
+            }
+            s => out.push(s),
+        }
+        if i + 1 == segments.len() && matches!(*seg, "." | "..") {
+            out.push("");
+        }
+    }
+    out.join("/")
+}
+
+/// Split a part into its unfolded header lines and its body (after the first empty line).
+fn split_headers(b: &[u8]) -> (Vec<(String, String)>, &[u8]) {
+    // Whichever blank line comes first ends the headers.
+    let blank = [find_bytes(b, b"\r\n\r\n").map(|i| (i, 4)), find_bytes(b, b"\n\n").map(|i| (i, 2))].into_iter().flatten().min();
+    let (head, body) = match blank {
+        Some((i, n)) => (b.get(..i).unwrap_or(&[]), b.get(i + n..).unwrap_or(&[])),
+        // Headers only, or a body without them.
+        None if b.first().is_some_and(|c| c.is_ascii_alphabetic()) && find_bytes(b, b":").is_some() => (b, &[][..]),
+        None => (&[][..], b),
+    };
+    let head = String::from_utf8_lossy(head.get(..head.len().min(MAX_HEADER)).unwrap_or(head));
+    let mut headers: Vec<(String, String)> = Vec::new();
+    for line in head.lines() {
+        if line.starts_with([' ', '\t']) {
+            if let Some((_, v)) = headers.last_mut() {
+                v.push(' ');
+                v.push_str(line.trim());
+            }
+        } else if let Some((k, v)) = line.split_once(':') {
+            headers.push((k.trim().to_ascii_lowercase(), v.trim().to_string()));
+        }
+    }
+    (headers, body)
+}
+
+fn header<'a>(h: &'a [(String, String)], name: &str) -> Option<&'a str> {
+    h.iter().find(|(k, _)| k == name).map(|(_, v)| v.as_str())
+}
+
+/// A `; name=value` parameter of a header value (quotes removed), name case-insensitive.
+fn param<'a>(value: &'a str, name: &str) -> Option<&'a str> {
+    value.split(';').skip(1).find_map(|p| {
+        let (k, v) = p.split_once('=')?;
+        k.trim().eq_ignore_ascii_case(name).then(|| v.trim().trim_matches('"'))
+    })
+}
+
+/// Gather the leaf parts of a (possibly multipart) entity.
+fn collect(headers: &[(String, String)], body: &[u8], out: &mut Vec<Part>, depth: usize) {
+    if out.len() >= MAX_PARTS {
+        return;
+    }
+    let ct = header(headers, "content-type").unwrap_or("text/plain").to_string();
+    if ct.trim_start().to_ascii_lowercase().starts_with("multipart/") {
+        let Some(boundary) = param(&ct, "boundary").filter(|b| !b.is_empty() && b.len() <= MAX_BOUNDARY) else { return };
+        if depth >= MAX_NESTING {
+            return;
+        }
+        for chunk in multipart_bodies(body, boundary.as_bytes()) {
+            let (h, b) = split_headers(chunk);
+            collect(&h, b, out, depth + 1);
+            if out.len() >= MAX_PARTS {
+                return;
+            }
+        }
+        return;
+    }
+    let enc = header(headers, "content-transfer-encoding").unwrap_or("").trim().to_ascii_lowercase();
+    let body = match enc.as_str() {
+        "quoted-printable" => quoted_printable(body),
+        "base64" => crate::model::base64_decode(&String::from_utf8_lossy(body)).unwrap_or_default(),
+        _ => body.to_vec(),
+    };
+    let clean = |s: &str| s.trim().trim_matches(['<', '>']).to_string();
+    out.push(Part {
+        content_type: ct,
+        location: header(headers, "content-location").map(clean),
+        id: header(headers, "content-id").map(clean),
+        body: Arc::new(body),
+    });
+}
+
+/// The bodies between `--boundary` lines, up to the closing `--boundary--`.
+fn multipart_bodies<'a>(body: &'a [u8], boundary: &[u8]) -> Vec<&'a [u8]> {
+    let mut delim = b"--".to_vec();
+    delim.extend_from_slice(boundary);
+    let mut out = Vec::new();
+    // Delimiters sit at line starts.
+    let mut starts = Vec::new();
+    let mut i = 0;
+    while let Some(k) = body.get(i..).and_then(|rest| find_bytes(rest, &delim)) {
+        let at = i + k;
+        if at == 0 || body.get(at - 1) == Some(&b'\n') {
+            starts.push(at);
+            if starts.len() > MAX_PARTS {
+                break;
+            }
+        }
+        i = at + delim.len();
+    }
+    for w in starts.windows(2) {
+        let (Some(&a), Some(&b)) = (w.first(), w.get(1)) else { continue };
+        let Some(seg) = body.get(a + delim.len()..b) else { continue };
+        if seg.starts_with(b"--") {
+            break;
+        }
+        // The rest of the delimiter line, then the part; the line break before the next
+        // delimiter belongs to it.
+        let seg = match find_bytes(seg, b"\n") {
+            Some(n) => seg.get(n + 1..).unwrap_or(&[]),
+            None => &[],
+        };
+        let seg = seg.strip_suffix(b"\r\n").or_else(|| seg.strip_suffix(b"\n")).unwrap_or(seg);
+        out.push(seg);
+        if out.len() >= MAX_PARTS {
+            break;
+        }
+    }
+    out
+}
+
+fn find_bytes(hay: &[u8], needle: &[u8]) -> Option<usize> {
+    if needle.is_empty() {
+        return None;
+    }
+    hay.windows(needle.len()).position(|w| w == needle)
+}
+
+/// Decode quoted-printable (RFC 2045 §6.7): `=XX` bytes and `=` soft line breaks; anything
+/// malformed is kept as it stands.
+fn quoted_printable(b: &[u8]) -> Vec<u8> {
+    let hex = |c: u8| (c as char).to_digit(16).and_then(|d| u8::try_from(d).ok());
+    let mut out = Vec::with_capacity(b.len());
+    let mut i = 0;
+    while let Some(&c) = b.get(i) {
+        if c != b'=' {
+            out.push(c);
+            i += 1;
+            continue;
+        }
+        match (b.get(i + 1).copied(), b.get(i + 2).copied()) {
+            (Some(b'\r'), Some(b'\n')) => i += 3,
+            (Some(b'\n'), _) => i += 2,
+            (Some(h), Some(l)) if let (Some(h), Some(l)) = (hex(h), hex(l)) => {
+                out.push((h << 4) | l);
+                i += 3;
+            }
+            _ => {
+                out.push(c);
+                i += 1;
+            }
+        }
+    }
+    out
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use wordcraft_doc::{Block, InlineObject};
+
+    const PNG_1X1: &str = "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8z8BQDwAEhQGAhKmMIQAAAABJRU5ErkJggg==";
+
+    fn message(html_part: &str) -> String {
+        format!(
+            "MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=\"----=_B\"\r\n\r\n\
+             ------=_B\r\n{html_part}\r\n\
+             ------=_B\r\nContent-Location: file:///C:/page_files/dot.png\r\nContent-Transfer-Encoding: base64\r\nContent-Type: image/png\r\n\r\n{PNG_1X1}\r\n\
+             ------=_B--\r\n"
+        )
+    }
+
+    #[test]
+    fn reads_the_quoted_printable_page_and_its_pictures() {
+        let page = "Content-Location: file:///C:/page.htm\r\nContent-Transfer-Encoding: quoted-printable\r\nContent-Type: text/html; charset=\"utf-8\"\r\n\r\n\
+                    <html><body><p class=3D\"x\">Dobar dan, gospo=C4=91o, ovo je =\r\njedan red.</p><p><img src=3D\"page_files/dot.png\" width=3D\"10\" height=3D\"10\"></p></body></html>";
+        let d = import(message(page).as_bytes());
+        let text = d.plain_text(wordcraft_doc::StoryRef::Body);
+        assert!(text.contains("Dobar dan, gospođo, ovo je jedan red."), "{text:?}");
+        let pictures =
+            d.body.iter().filter(|b| matches!(&***b, Block::Para(p) if p.objects.iter().any(|o| matches!(o, InlineObject::Image { .. })))).count();
+        assert_eq!(pictures, 1);
+        assert_eq!(d.media.len(), 1);
+    }
+
+    #[test]
+    fn pictures_named_relative_to_the_page_resolve_as_uris() {
+        // The page in `reports/`, its picture in `images/` (RFC 2557 §5: relative URIs resolve
+        // against the page's Content-Location).
+        let m = |src: &str| {
+            format!(
+                "MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=b\r\n\r\n\
+                 --b\r\nContent-Location: https://example.test/reports/page.html\r\nContent-Type: text/html\r\n\r\n<p>x</p><img src=\"{src}\">\r\n\
+                 --b\r\nContent-Location: https://example.test/images/logo.png\r\nContent-Transfer-Encoding: base64\r\nContent-Type: image/png\r\n\r\n{PNG_1X1}\r\n--b--\r\n"
+            )
+        };
+        for src in [
+            "../images/logo.png",
+            "/images/logo.png",
+            "./../images/./logo.png",
+            "//example.test/images/logo.png",
+            "https://example.test/images/logo.png",
+        ] {
+            assert_eq!(import(m(src).as_bytes()).media.len(), 1, "{src}");
+        }
+        // Nothing past the root: `..` there stays at the root.
+        assert_eq!(import(m("../../../images/logo.png").as_bytes()).media.len(), 1);
+        assert_eq!(import(m("images/logo.png").as_bytes()).media.len(), 0);
+        assert_eq!(resolve("https://h/a/b/c.html", "../d/./e.png#f"), "https://h/a/d/e.png");
+        assert_eq!(resolve("file:///C:/x/page.htm", "page_files/dot.png"), "file:///C:/x/page_files/dot.png");
+        assert_eq!(resolve("", "dot.png"), "dot.png");
+        // A host with no path, and a base whose query holds slashes (RFC 3986 §5.2.3: the path
+        // merges, not the query; an authority with an empty path gets its `/`).
+        assert_eq!(resolve("https://example.test", "logo.png"), "https://example.test/logo.png");
+        assert_eq!(resolve("https://example.test/page.html?redirect=/other/", "logo.png"), "https://example.test/logo.png");
+        assert_eq!(resolve("https://h/a/p.html?q", "?r"), "https://h/a/p.html?r");
+        assert_eq!(resolve("https://h/a/p.html?q", "x.png?v=1"), "https://h/a/x.png?v=1");
+        assert_eq!(resolve("https://h/a/p.html", "mailto:x@y"), "mailto:x@y");
+        for page in ["https://example.test", "https://example.test/page.html?redirect=/other/"] {
+            let m = format!(
+                "MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=b\r\n\r\n\
+                 --b\r\nContent-Location: {page}\r\nContent-Type: text/html\r\n\r\n<p>x</p><img src=\"logo.png\">\r\n\
+                 --b\r\nContent-Location: https://example.test/logo.png\r\nContent-Transfer-Encoding: base64\r\nContent-Type: image/png\r\n\r\n{PNG_1X1}\r\n--b--\r\n"
+            );
+            assert_eq!(import(m.as_bytes()).media.len(), 1, "{page}");
+        }
+    }
+
+    #[test]
+    fn the_part_charset_decides_the_text() {
+        let page = "Content-Type: text/html; charset=windows-1250\r\nContent-Transfer-Encoding: quoted-printable\r\n\r\n<p>=E8lanak</p>";
+        let d = import(message(page).as_bytes());
+        assert_eq!(d.plain_text(wordcraft_doc::StoryRef::Body).trim(), "članak");
+    }
+
+    #[test]
+    fn the_start_part_is_the_page_and_cid_names_pictures() {
+        let m = format!(
+            "MIME-Version: 1.0\nContent-Type: multipart/related; type=\"text/html\"; start=\"<page>\"; boundary=b\n\n\
+             --b\nContent-Type: text/html\n\n<p>Not this one</p>\n\
+             --b\nContent-ID: <page>\nContent-Type: text/html\n\n<p>This one</p><img src=\"cid:dot\">\n\
+             --b\nContent-ID: <dot>\nContent-Transfer-Encoding: base64\nContent-Type: image/png\n\n{PNG_1X1}\n--b--\n"
+        );
+        let d = import(m.as_bytes());
+        assert_eq!(d.plain_text(wordcraft_doc::StoryRef::Body).trim(), "This one");
+        assert_eq!(d.media.len(), 1);
+        // Without a page, a message reads as its text.
+        let d = import(b"MIME-Version: 1.0\nContent-Type: multipart/mixed; boundary=b\n\n--b\nContent-Type: text/plain\n\nJust text\n--b--\n");
+        assert_eq!(d.plain_text(wordcraft_doc::StoryRef::Body).trim(), "Just text");
+        assert!(is_mht(m.as_bytes()) && !is_mht(b"<html><p>x</p>"));
+    }
+
+    #[test]
+    fn hostile_messages_never_panic() {
+        for m in [
+            &b""[..],
+            b"MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=\"\"\r\n\r\n",
+            b"MIME-Version: 1.0\r\nContent-Type: multipart/related; boundary=x\r\n\r\n--x\r\n--x--",
+            b"MIME-Version: 1.0\nContent-Type: multipart/related; boundary=x\n\n--x\nContent-Transfer-Encoding: quoted-printable\nContent-Type: text/html\n\n=\n=Z=4",
+            b"Content-Type: text/html\r\nContent-Transfer-Encoding: base64\r\n\r\n!!!!",
+        ] {
+            let _ = import(m);
+        }
+        let mut deep = String::from("MIME-Version: 1.0\r\n");
+        for i in 0..64 {
+            deep.push_str(&format!("Content-Type: multipart/related; boundary=b{i}\r\n\r\n--b{i}\r\n"));
+        }
+        let _ = import(deep.as_bytes());
+    }
+}
@@ -148,6 +148,7 @@ fn flow_paras_in(doc: &Document, p: &Paragraph, tracked: bool) -> Vec<Para> {
        mark_ins: None,
        mark_del: None,
        fmt_change: None,
+        mark: None,
    };
    let in_code = kind == Kind::Code;
    let mut k = 0usize;
@@ -179,6 +180,7 @@ fn flow_paras_in(doc: &Document, p: &Paragraph, tracked: bool) -> Vec<Para> {
                        mark_ins: None,
                        mark_del: None,
                        fmt_change: None,
+                        mark: None,
                    };
                    let prev = std::mem::replace(&mut out, next);
                    if prev.inlines.is_empty() {
@@ -320,6 +322,7 @@ fn flow_table(doc: &Document, t: &Table, depth: usize, tracked: bool) -> FTable
                shading: c.props.shading,
                borders: c.props.borders,
                valign: c.props.valign,
+                width: None,
            });
        }
        rows.push(cells);
@@ -100,6 +100,12 @@ fn para_props(p: &Para, list_num: Option<u32>) -> ParaProps {
    out.indent_left = sp.left;
    out.indent_right = sp.right;
    out.indent_first = sp.first;
+    if sp.before_auto {
+        out.space_before_auto = Some(true);
+    }
+    if sp.after_auto {
+        out.space_after_auto = Some(true);
+    }
    if p.page_break {
        out.page_break_before = Some(true);
    }
@@ -177,6 +183,9 @@ impl<'a> Builder<'a> {
            let old = para_props(&c.old, old_num);
            out.props.fmt_change = PropChange::boxed(self.revision(RevisionKind::Format, &c.mark), old);
        }
+        if let Some(f) = &p.mark {
+            out.mark = self.budget(char_props(f));
+        }
        if p.keep_next {
            out.props.keep_next = Some(true);
        }
@@ -220,6 +220,9 @@ pub struct Spacing {
    pub right: Option<f32>,
    /// First-line indent; negative = hanging.
    pub first: Option<f32>,
+    /// HTML's automatic space before / after (see `ParaProps::space_before_auto`).
+    pub before_auto: bool,
+    pub after_auto: bool,
 }

 #[derive(Clone, Debug, Default, PartialEq)]
@@ -244,6 +247,8 @@ pub struct Para {
    /// Tracked formatting change of the paragraph: the paragraph's formatting before it (kind,
    /// list, alignment, page break, spacing and borders of `old`; its inlines are unused).
    pub fmt_change: Option<Box<FmtChange<Para>>>,
+    /// Formatting of an empty paragraph's mark (what sizes its line), when the source gives it.
+    pub mark: Option<Box<Fmt>>,
 }

 impl Para {
@@ -337,11 +342,14 @@ pub struct Cell {
    /// The cell's own borders (RTF `\clbrdr…`), drawn whatever the table's style.
    pub borders: Option<wordcraft_doc::props::Borders>,
    pub valign: Option<VAlign>,
+    /// The width, points, the source gives a cell spanning one column (HTML): see
+    /// [`FTable::widths_from_cells`].
+    pub width: Option<f32>,
 }

 impl Default for Cell {
    fn default() -> Self {
-        Cell { blocks: Vec::new(), colspan: 1, rowspan: 1, covered: false, header: false, shading: None, borders: None, valign: None }
+        Cell { blocks: Vec::new(), colspan: 1, rowspan: 1, covered: false, header: false, shading: None, borders: None, valign: None, width: None }
    }
 }

@@ -372,6 +380,22 @@ impl FTable {
    pub fn cols(&self) -> usize {
        self.rows.iter().map(|r| r.iter().map(|c| c.colspan.clamp(1, 63) as usize).sum::<usize>()).max().unwrap_or(0)
    }
+    /// Grid column widths from the cells that give one (after [`Self::insert_covered`], so a
+    /// cell's column counts those spanned from rows above), when every column has one.
+    pub fn widths_from_cells(&mut self) {
+        let mut widths = vec![None; self.cols()];
+        for row in &self.rows {
+            let mut g = 0;
+            for c in row {
+                if let (Some(w), Some(slot)) = (c.width, widths.get_mut(g)) {
+                    slot.get_or_insert(w);
+                }
+                g += c.colspan.clamp(1, 63) as usize;
+            }
+        }
+        self.widths = widths.into_iter().collect::<Option<Vec<f32>>>().unwrap_or_default();
+    }
+
    /// After an importer that gives `rowspan` but no placeholder cells (HTML): insert covered
    /// cells into the rows below each spanning cell.
    pub fn insert_covered(&mut self) {
@@ -115,6 +115,7 @@ fn parse_full(bytes: &[u8]) -> Result<(Flow, SectionProps), String> {

 /// [`parse_full`], plus whether change tracking is on (`None`: the document has no tracked changes).
 fn parse_tracked(bytes: &[u8]) -> Result<(Flow, SectionProps, Option<bool>, Option<String>, page_parts::Parts), String> {
+    wordcraft_doc::zip_guard::check(bytes, crate::MAX_ZIP_ENTRIES).map_err(|e| format!("ODT: {e}"))?;
    let mut z = zip::ZipArchive::new(Cursor::new(bytes)).map_err(|e| format!("ODT: not a valid package ({e})"))?;
    if let Some(mt) = read_entry(&mut z, "mimetype") {
        let mt = String::from_utf8_lossy(&mt);
@@ -261,5 +261,13 @@ fn spacing(st: &St) -> model::Spacing {
    } else {
        Some(LineSpacing::AtLeast(pt(sl)))
    };
-    model::Spacing { before: gap(st.sb), after: gap(st.sa), line, left: st.li.map(pt), right: st.ri.map(pt), first: st.fi.map(pt) }
+    model::Spacing {
+        before: gap(st.sb),
+        after: gap(st.sa),
+        line,
+        left: st.li.map(pt),
+        right: st.ri.map(pt),
+        first: st.fi.map(pt),
+        ..Default::default()
+    }
 }
@@ -201,6 +201,7 @@ struct Package<'a> {

 impl<'a> Package<'a> {
    fn open(bytes: &'a [u8]) -> Result<Self, String> {
+        wordcraft_doc::zip_guard::check(bytes, crate::MAX_ZIP_ENTRIES)?;
        let zip = zip::ZipArchive::new(Cursor::new(bytes)).map_err(|e| format!("not a valid spreadsheet package ({e})"))?;
        Ok(Package { zip, total: 0 })
    }
@@ -401,6 +402,14 @@ mod tests {
    use super::*;
    use std::io::Write;

+    #[test]
+    fn a_package_declaring_too_many_entries_is_refused() {
+        // #3660: refused before `zip` reserves and parses them.
+        let bomb = wordcraft_doc::zip_guard::zip64_entry_count_sample(crate::MAX_ZIP_ENTRIES + 1);
+        assert!(sheet_names(&bomb).is_err_and(|e| e.contains("declares")));
+        assert!(crate::odt::parse(&bomb).is_err_and(|e| e.contains("declares")));
+    }
+
    #[test]
    fn negative_zero_formats_as_zero() {
        assert_eq!(number(-0.0), "0");
@@ -1011,3 +1011,11 @@ fn html_title_outside_head_is_metadata_not_body() {
        assert_eq!(d.core.title, "T", "{html}");
    }
 }
+
+#[test]
+fn latex_export_keeps_repeated_spaces() {
+    let tex = String::from_utf8(crate::export("tex", &Document::from_text("a  b   c")).unwrap().unwrap()).unwrap();
+    assert!(!tex.contains("  "), "{tex}");
+    let d = crate::import("tex", tex.as_bytes()).unwrap().unwrap();
+    assert_eq!(para_texts(&d), vec!["a  b   c".to_string()]);
+}
@@ -6,8 +6,8 @@ use super::page::PageFrame;
 use super::placed::Placed;
 use super::table_flow::below;
 use super::{
-    Anchor, Arc, Block, Blocks, Border, Document, Float, FloatAlign, HashMap, ParaLayout, Paragraph, Path, Rect, SectionProps, StoryRef, TableFloat,
-    TextDirection, Wrap, WrapText, para, table,
+    Anchor, Arc, Block, Blocks, Border, Document, Float, FloatAlign, HashMap, ParaLayout, Paragraph, Path, Rect, StoryRef, TableFloat, TextDirection,
+    Wrap, WrapText, para, table,
 };

 /// Cell identity is independent of optional table-style formatting.
@@ -281,7 +281,8 @@ pub(crate) fn layout_box(ctx: &mut Ctx, input: BoxInput<'_>) -> BoxLayout {
                        continue;
                    }
                    if let Some(pin) = ctx.pin.as_ref().filter(|pin| in_cell && pin.story == story && float.outside_cell && page_anchored(float)) {
-                        let page = PageFrame { sect: &pin.sect, origin: (0.0, 0.0), number: pin.number, mirror: pin.mirror };
+                        let page =
+                            PageFrame { sect: &pin.sect, margin_top: pin.margin_top, origin: (0.0, 0.0), number: pin.number, mirror: pin.mirror };
                        pinned.insert(oi, float_rect(Some(page), (0.0, width), para_top, w, h, float));
                        continue;
                    }
@@ -460,15 +461,22 @@ fn box_float_table_rect(
            _ => None,
        }
    };
-    let page_y = |s: &SectionProps| match f.v_rel {
-        Anchor::Page => Some((0.0, s.page_h)),
-        Anchor::Margin => Some((s.margin_top, s.text_height())),
-        _ => None,
+    let page_y = |fr: &PageFrame| {
+        let s = fr.sect;
+        match f.v_rel {
+            Anchor::Page => Some((0.0, s.page_h)),
+            Anchor::Margin => {
+                let mut reference = s.clone();
+                reference.margin_top = fr.margin_top.unwrap_or(s.margin_top);
+                Some((reference.margin_top, reference.text_height()))
+            }
+            _ => None,
+        }
    };
    let h_area = frame.and_then(|fr| page_x(&fr).map(|(a, e)| (a - fr.origin.0, e))).unwrap_or((0.0, width));
    // In a cell an offset from the page or margin has nothing to measure from: the table stands
    // where the text is.
-    let (v_area, y_off) = match frame.and_then(|fr| page_y(fr.sect).map(|(a, e)| (a - fr.origin.1, e))) {
+    let (v_area, y_off) = match frame.and_then(|fr| page_y(&fr).map(|(a, e)| (a - fr.origin.1, e))) {
        Some(area) => (area, f.y),
        None if matches!(f.v_rel, Anchor::Page | Anchor::Margin) => ((y, 0.0), 0.0),
        None => ((y, 0.0), f.y),
@@ -621,15 +629,17 @@ pub(crate) fn push_para(
    let left = x + li;
    let right = x + width - ri;
    let (top, bottom) = (y - e.top - e.up, y + h + e.bottom);
+    // The outer edges of the side borders.
+    let sides = pl.rp.borders.unwrap_or_default();
+    let (lx, rx) = (left - border_room(sides.left), right + border_room(sides.right));
    if let Some(c) = pl.rp.shading {
        let pad = pl.rp.borders.as_ref().map(|b| b.left.map(|l| l.space).unwrap_or(4.0)).unwrap_or(0.0);
-        items.push(Placed::Fill { rect: Rect::new(left - pad, top, right - left + pad * 2.0, bottom - top), color: c });
+        let (sl, sr) = (lx.min(left - pad), rx.max(right + pad));
+        items.push(Placed::Fill { rect: Rect::new(sl, top, sr - sl, bottom - top), color: c });
    }
    items.push(Placed::Lines { story, path: Path(path.to_vec()), para: pl.clone(), l0, l1, x, y, turn: TextDirection::Horizontal });
    if let Some(b) = &pl.rp.borders {
-        let sp = |o: &Option<Border>| o.map(|b| b.space).unwrap_or(0.0);
-        let (lx, rx) = (left - sp(&b.left), right + sp(&b.right));
-        // The top and bottom lines lie beyond their distance from the text.
+        // All four lines lie beyond their distance from the text; top and bottom reach the corners.
        if let Some(t) = b.top.filter(|t| e.top > 0.0 && t.is_visible()) {
            let ty = y - t.space - t.width / 2.0;
            items.push(Placed::Rule { x0: lx, y0: ty, x1: rx, y1: ty, border: t });
@@ -639,10 +649,12 @@ pub(crate) fn push_para(
            items.push(Placed::Rule { x0: lx, y0: by, x1: rx, y1: by, border: t });
        }
        if let Some(t) = b.left.filter(Border::is_visible) {
-            items.push(Placed::Rule { x0: lx, y0: top, x1: lx, y1: bottom, border: t });
+            let bx = lx + t.width / 2.0;
+            items.push(Placed::Rule { x0: bx, y0: top, x1: bx, y1: bottom, border: t });
        }
        if let Some(t) = b.right.filter(Border::is_visible) {
-            items.push(Placed::Rule { x0: rx, y0: top, x1: rx, y1: bottom, border: t });
+            let bx = rx - t.width / 2.0;
+            items.push(Placed::Rule { x0: bx, y0: top, x1: bx, y1: bottom, border: t });
        }
    }
    y + h
@@ -89,6 +89,8 @@ struct ParaEnvInput<'a> {
 pub(crate) struct Pin {
    pub(crate) story: StoryRef,
    pub(crate) sect: SectionProps,
+    /// Rendered header Margin boundary; absent in footers and height probes.
+    pub(crate) margin_top: Option<f32>,
    pub(crate) number: u32,
    pub(crate) mirror: bool,
    pub(crate) behind: Vec<Placed>,
@@ -108,7 +110,14 @@ pub(crate) fn page_anchored(f: &Float) -> bool {
 /// Run `f`, a layout of header or footer `id` for page `number` of `sect`, with the floats in its
 /// table cells that are placed on the page put there (see [`Pin`]). Returns `f`'s result and those
 /// floats' items, in page coordinates: behind the header's text, and in front.
-pub(crate) fn with_pin<T>(ctx: &mut Ctx, id: u32, sect: &SectionProps, number: u32, f: impl FnOnce(&mut Ctx) -> T) -> (T, Vec<Placed>, Vec<Placed>) {
+pub(crate) fn with_pin<T>(
+    ctx: &mut Ctx,
+    id: u32,
+    sect: &SectionProps,
+    number: u32,
+    margin_top: Option<f32>,
+    f: impl FnOnce(&mut Ctx) -> T,
+) -> (T, Vec<Placed>, Vec<Placed>) {
    let tables = ctx.doc.parts.get(&id).is_some_and(|p| p.blocks.iter().any(|b| matches!(**b, Block::Table(_))));
    if !tables {
        return (f(ctx), Vec::new(), Vec::new());
@@ -116,6 +125,7 @@ pub(crate) fn with_pin<T>(ctx: &mut Ctx, id: u32, sect: &SectionProps, number: u
    let pin = Pin {
        story: StoryRef::Part(id),
        sect: sect.clone(),
+        margin_top,
        number,
        mirror: ctx.mirror,
        behind: Vec::new(),
@@ -23,6 +23,8 @@ mod paint;

 pub(crate) use glyphs::author_order;
 pub use glyphs::{CHECK_BOX_INSET, check_box, text_color, upright_in_vertical};
+#[cfg(test)]
+pub(crate) use lines::{leader_pitch, leader_xs};
 pub use lines::{mark_text_width, section_break_label};
 pub(crate) use objects::inline_rect;
 pub use objects::seg_path;
--- a/Show More
+++ b/Show More