A font of a pan-CJK family lists the code pages of every language it
covers. With no `meta` languages and no CJK family name, the first group
in menu order won, so a Simplified Chinese face such as "… Noto Sans CJK
SC" was listed under Japanese.
When a font lists several CJK code pages, the region tag in its family
name (JP; SC, CN, GB; TC, HK, TW, MO, HC; KR — a whole word in capitals)
now decides, if the font lists that language. Fonts with one CJK code
page, or none, are grouped as before.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013phQ3ekq6uGBEWD6FjFYG2
Cut, Duplicate, Paste in Place, the Arrange commands and pasting text
run a second command inside themselves, and both recorded an undo
step: the second Undo did nothing. Only the outer command records now.
A command that fails after changing the document records its step
too; a script that stopped at a bad step left changes that could not
be undone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Conflicts:
- crates/fonts/src/fontdb.rs, vertical.rs: next's em cache
((top, bottom), declared) with FontFace::declared_em_box and the ICF
cache/icf_box kept; the branch's base_em_box accessor dropped (same
value). em_box's fallback chain takes the branch's let em form.
- crates/compose/src/lib.rs: next's cjk_icf_box and the branch's
alignment_em_box, which reads declared_em_box.
- crates/compose/src/tests.rs: both tests. next applies character
alignment only under the Japanese composers (5dc8383a), so the branch's
smaller_cjk_run_aligns_to_the_line_em_box_centre runs under japanese();
next's cjk_em_center_alignment_moves_small_characters measures centres
on the alignment em box (declared, else 0.88/-0.12 em).
Conflicts:
- crates/pdf/src/text.rs run_text: the branch's two generated_text
branches first, then #392's kashida branch (no text), then the cmap
fallback for other inserted glyphs.
- crates/compose/src/lib.rs: next's kashida PlacedGlyph (with dx) gets
generated_text: None.
- crates/compose/src/shape.rs hyphen_after: next's rendered_char/moji
resets and generated_text = None.
- crates/compose/src/ruby.rs: next's ruby module (#363); the branch's
edit to the old ruby set() has nothing to apply to.
- crates/pdf/src/tests.rs: tests from both.
Also, so the merge compiles: next's ruby (ruby.rs annotate) and kenten
(kenten.rs shape_mark) PlacedGlyph literals take generated_text: None;
they map back through the font like other inserted glyphs. Kanji page
numbers, composite-font note numbers and fallback-face label segments
go through shape_segment's replacement and list labels through
prepend_label, so they carry their generated text.
next's shatai path draws a slanted glyph on its own through tagged_run,
which opens a Span; inside #392's right-to-left line span that nested
marked content and krilla panicked ("can't start marked content twice")
in tagged exports. While a line span is open, the glyph is drawn with
run like the line's other runs. Test:
shatai_glyphs_in_right_to_left_lines_stay_in_the_line_span (panicked
before).
Asset::same_file compares pdf_crop (#362) with type, page and bytes: one
PDF page shown with two crop boxes keeps two assets, so add_asset,
adopt_asset, insert_asset (IDML/INDD import) and the load/save merge
never point a placement at another crop. Test in
identical_files_share_one_asset.
The copy clipping (edit.rs clip_doc) keeps the assets used_assets finds
in it, so graphics anchored in copied text, graphic table cells (#379),
notes and media posters keep their files; it collected only graphics on
the copied items before. Test:
copied_text_frames_keep_the_graphics_anchored_in_their_text.
Conflicts:
- crates/engine/src/cmd/{datamerge/fill.rs, file.rs, mathexpr.rs,
media.rs, table.rs}: the branch's add_asset calls, with next's Asset
fields (pdf_crop, pdf_box) through ..Default::default(); file.place
keeps next's PdfCrop-aware asset (crop_pdf) and adds it with add_asset;
the data merge graphic keeps next's wrap field.
- crates/format/src/lib.rs: load runs next's upgrade(doc, version), then
merge_duplicate_assets (zip and legacy JSON); tests from both, with the
round-trip test placing its asset (saves drop unused ones).
- crates/idml/src/import.rs graphic(): the branch's linked_assets reuse and
insert_asset, with next's PDF page and PdfCrop on the asset and the
graphic's text wrap (#389 integration) in the shared graphic closure.
linked_assets is keyed by (path, page, crop), so placements of one PDF
showing different pages or crop boxes keep separate assets.
Also, so the merge compiles: Asset/Graphic literals in the branch's tests
(crates/doc/src/assets.rs, format, engine links.rs) take next's pdf_crop,
pdf_box and wrap fields.
The merge of idml-feathers left overlay() in crates/idml/src/import.rs and
imports_and_round_trips_text_wrap_side_and_contour_type in tests.rs
without their closing braces (both sides added code at the same place);
this closes them.
Test: feathers_come_from_the_object_style_unless_its_category_turns_them_off
(gradient and directional feathers come from the style through
with_style; EnableGradientFeather="false" in
ObjectStyleObjectEffectsCategorySettings drops the style's gradient
feather; the item's own Applied="false" wins).
Conflicts:
- crates/idml/src/import.rs, free functions: next's effect_flag/overlay
(#389) and the branch's finite/gradient_feather_in both kept. The
feather reads sit inside the TransparencySetting block, which next reads
through with_style, so object style feathers are inherited and their
EnableGradientFeather/EnableDirectionalFeather category flags apply.
- crates/idml/src/tests.rs: tests from both.
A graphic element that names an AppliedObjectStyle reads its
TextWrapPreference through with_style, so mode, side, offsets and
ContourOption come from the style where the graphic sets none. Test:
text_wrap_side_and_contour_come_from_the_style_for_frames_and_placed_graphics.
Also: swatch_ref's doc comment sits on swatch_ref again.
Conflicts:
- crates/idml/src/import.rs, style chain helper: next's object_style_chain
(trims BasedOn, ends with [None], cycle-safe) kept; the branch's
style_chain dropped. with_style and attr_or_style use object_style_chain.
- import.rs corners: next's corner import over the object style chain kept
(per-corner then all-corners at each level, InDesign default radius);
the branch's second corner pass dropped, so corners are read once.
- import.rs text wrap: with_style(e, "TextWrapPreference") feeds next's
shared text_wrap() reader (side, ContourOption, offsets) instead of the
branch's inline TextWrap build.
- import.rs TextFramePreference: next's chain-aware text_frame_options
with column rule colour kept; BaselineFrameGridOption read through
with_style.
- crates/idml/src/tests.rs: test modules from both.
- release.yml: new linux-riscv64 job (cross-compiled on ubuntu-24.04 against
ports.ubuntu.com riscv64 libs, tar.gz only, CLI smoke-tested under QEMU),
ported from GridCraft; the draft release job needs it, so SHA256SUMS covers it.
- packaging/linux/package.sh: CROSS_ARCH/CROSS_TARGET/CROSS_COMPILE/EMULATOR
support, as in GridCraft.
- cargo xtask version set 0.6.0; README, hardware parity and ROADMAP log note
the new platform.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
- install_fonts no longer waits for the system font catalog on the first
frame: the built-in faces are installed at once, the scan runs in the
background, and the installed CJK faces are added when it is done.
- The installed faces are read once and shared: changing the interface
language reorders them without copying the font files into egui again.
- fonts: line_metrics no longer sits between feature's doc comment and
feature.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013phQ3ekq6uGBEWD6FjFYG2
install_fonts caches system_ui_fonts in a OnceLock; the face bytes are
leaked once and borrowed (Cow::Borrowed) by each FontData, so a language
change reorders the faces without re-reading font files or copying them.
SystemFont.bytes is a Cow<'static, [u8]>.
Also: line_metrics no longer sits between feature's doc comment and
feature (crates/fonts), and the canvas texture test's frame closure
drops an unneeded mut.
Lines with right-to-left text are wrapped in a marked span whose
ActualText is the line's text in reading order, so viewers that order
text by position (Preview, PDFKit) no longer return it reversed.
A base letter and its marks (Hebrew points) map glyph by glyph when
each glyph stands for one of the cluster's characters, and are drawn in
text order, so ToUnicode-only extraction reads each letter once with
its points after it. Marks stay in their run, placed by glyph offsets
(compensated for skew), instead of starting runs of their own.
Kashidas inserted by justification carry no text.
Fixes#392.
Conflict: crates/ui-egui/src/theme.rs install_fonts: the PR's
system_ui_fonts/add_system_fallbacks replace next's #126 system-font loop
(fallback_for per CJK sample char into system-ui-{id} faces when
CRAFT_FONTS is empty), so installed faces are not loaded twice. No #126
helper in theme.rs is left unused; FontDb::fallback_for stays (compose,
images use it).
Placing a file that the document already holds reuses its asset (same
bytes, type and PDF page), in file.place, place.load, media and poster
placement, graphic table cells, math, data merge image fields, IDML
import (a linked file is read once) and paste/duplicate between
documents (which also maps ids that collide in the target).
Saving merges identical assets into the first and leaves out assets
nothing references; the open document keeps them for undo. Opening a
file saved with one copy per placement shares the identical bytes as
they are read and merges the assets. Package and Copy Links write only
used files, and links.list lists only used assets.
Refs #391
patch_texture writes full-resolution dirty rects with set_partial; when
upload() has shrunk the canvas texture to max_texture_side (a viewport
wider than the GPU limit), the edit takes a full render instead of a
partial write past the texture's bounds (a wgpu validation error). Test:
edits_patch_the_canvas_only_at_full_resolution.
The content ghost proxy goes through widgets::fit_texture (capped at 4096)
instead of skipping graphics larger than the limit.
patch_texture computed the damaged region in view pixels and wrote it
with set_partial into the canvas texture upload() had sampled down to
the GPU limit: a viewport wider than max_texture_side put the region in
the wrong place or outside the texture (a wgpu validation error on the
first edit). Such a texture is now rendered whole.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013phQ3ekq6uGBEWD6FjFYG2
Conflict: crates/ui-egui/src/canvas.rs draw_tool_overlays: the PR's
Overlay::Path arm (draw_path_overlay) and next's Overlay::ContentGhost
arm both kept; draw_path_overlay sits before draw_content_ghost.
The PR was rebased onto upstream main; its code commits are unchanged
(range-diff shows only ROADMAP lines moved into the new docs commit), so
next already carries them.
Conflicts: crates/ui-egui/src/{dialogs.rs,i18n.rs,i18n/it.rs,lib.rs}:
next's versions (the old PR code plus next's own dialogs, findFont and
numberingSection work). ROADMAP.md and docs/target-app-parity.md take the
PR's New Document entries.
Import reads GradientFeatherSetting (Applied, Type, Angle, Length,
GradientStart, first and last OpacityGradientStop) and
DirectionalFeatherSetting (Applied and the four widths) on page items.
Export writes both when the effect is on.
- apps/designcraft/Cargo.toml: the macOS target keeps objc2 0.6 / objc2-app-kit 0.3 (IME) and
adds fmv-macos-events =0.1.0, which uses the same objc2 and objc2-app-kit versions.
- apps/designcraft/src/main.rs: the shared Arc<gpu::Startup>, catch_unwind and restart flow
stay; the Apple-event handlers are installed before run_native (outside catch_unwind, held
until the window ends), the app drains the inbox each frame, and documents that arrived are
noted on the Startup. The open dialog keeps the INDD/INDT filter and ends with All files.
The open-filter test checks both; the save-dialog test is added.
- apps/designcraft/src/gpu.rs: Startup::opening records the documents the OS handed this run;
a start again after a graphics failure (both finish and the first-frame watch) passes them
as file arguments, since Finder doesn't send them twice.
- apps/designcraft/src/apple_events.rs: poll returns the documents it handed the app, and
queued() returns those still in the inbox when the window ends.
- crates/engine/src/cmd/file.rs: file.open decides by content first (IDML by its mimetype
entry, then a DesignCraft package or JSON), then an InDesign file by its signature or
.indd/.indt extension, then a .idml or .designcraft extension for the reader's own error;
anything else is "not a DesignCraft or IDML file". INDD stays in file.openBytes. Save
commands add .designcraft as on file-open-os; the params docs describe all of it.
- crates/engine/src/cmd/mod.rs: re-exports from both sides (is_document,
with_document_extension and layout::chapter_numbering).
A page item takes every value it lacks from its applied object style and
the styles that one is based on, attribute by attribute: corner shapes and
radii, TransparencySetting (blending, drop shadow, feather), text wrap,
text frame options, the baseline frame grid and frame fitting. Nested
elements merge level by level. Effects whose category the applied style
turns off come from the item alone.
Export writes the column rule attributes on every frame, so a frame keeps
its own rule settings over those of its object style on import.
Fixes#389.
Generated glyphs (page numbers, section markers, list labels) carry the
text they were shaped from, so a figure the font substitutes under small
caps or other OpenType features extracts as the digit instead of U+FFFD.
Character alignment (em box top, centre, bottom) measured each glyph's em
box from its font's ascender/descender proportions, which puts a Latin
font's em box centre too low, so a smaller CJK run in a Latin line sat
below the line's em box centre. Alignment now uses the font's BASE em box,
else 0.88 em above and 0.12 em below the baseline.
The system menu bar was built once at launch, so Edit › Interface
Language changed the panels but left the menus in the old language
until the next launch. It is rebuilt when the language changes.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_013phQ3ekq6uGBEWD6FjFYG2
aiongg