CP1.3: review — no-data proofing described right, NOTICE names craft-i18n's hyphenation.pat files, Flatpak dictionaries folder, printing checkout without credentials, data gate asks the variant table too

CodeRabbit on #5202:
- AGENTS.md, ATTRIBUTION.md, README.md, docs/architecture.md: without
  CRAFT_I18N_DIR English spelling falls back to an installed Hunspell English
  or craft-proof's built-in Moby list, and only pattern hyphenation is off
  (Bulgarian, Turkish and Russian hyphenate by rule); they said English
  wasn't spell-checked. (CLAUDE.md carries the same sentence; left for the
  maintainer, as agents don't edit it.)
- NOTICE: craft-i18n's patterns are language-data/<lang>/hyphenation.pat,
  not en-us.pat, pl.pat, pt.pat, de.pat, it.pat.
- docs/spelling-dictionaries.md: the Flatpak's dictionaries folder is
  ~/.var/app/ai.storyteller.wordcraft/data/wordcraft/dictionaries; the host's
  folders are outside its sandbox.
- printing.yml: the repository checkout doesn't persist credentials.
- tests.rs no_language_data: also probes the American/British variant table,
  so a craft-i18n checkout missing some files skips the dialect tests
  instead of failing them.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FypgCX8LsYYE8fsj4byBm2

.github/workflows/printing.yml

@@ -43,6 +43,8 @@ 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).

AGENTS.md

@@ -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`). 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 isn't spell-checked and hyphenation is off, 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.
+- 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

ATTRIBUTION.md

@@ -61,4 +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 is not spell-checked and hyphenation is off |
+| 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 |

NOTICE

@@ -26,11 +26,11 @@ list (author, source, license) is in ATTRIBUTION.md.
   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
-  (craft-i18n language-data/en-us/en-us.pat) were trained from it and are released as CC0-1.0.
-- Polish hyphenation patterns (craft-i18n language-data/pl/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.
+- 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 (craft-i18n language-data/pt/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,
@@ -56,7 +56,7 @@ 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 (craft-i18n language-data/de/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:
 
@@ -78,7 +78,7 @@ 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 (craft-i18n language-data/it/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:
 

README.md

@@ -143,7 +143,9 @@ 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 in English with no English spell-checking and no hyphenation:
+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

crates/engine/src/tests.rs

@@ -18,15 +18,16 @@ mod revision_checks;
 mod typing;
 mod wordart;
 
-/// True, after saying so on stderr, when this build has no Language data embedded (craft-proof
-/// built without `CRAFT_I18N_DIR`). The probe is craft-proof's public `english_source`: the
-/// embedded en-us word list is there exactly when the craft-i18n data was embedded, since
-/// `CRAFT_I18N_DIR` brings all of it or none. Without it English still spells from craft-proof's
-/// built-in word list, but there are no hyphenation patterns and no American/British variant
-/// table, so the tests that check those return early, as craft-proof's own do; with the data (CI,
-/// releases) they all run.
+/// 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);
+    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)");
     }

docs/architecture.md

@@ -22,7 +22,7 @@ crate name links to its source tree.
 | L1 | [`doc`](../crates/doc/src) | 23,719 | Document model: stories, paragraphs, runs, tables, sections, styles, numbering, fields, notes, comments, revisions and equations; edit primitives and bidi |
 | L1 | [`fonts`](../crates/fonts/src) | 7,088 | Font database, metrics, outlines, HarfRust shaping, Word font substitution and OpenType MATH |
 | L1 | [`metafile`](../crates/metafile/src) | 3,614 | Shared Windows Metafile parsing and drawing primitives used by format readers and renderers |
-| L1 (external) | [`craft-proof`](https://github.com/storytold/craft-libs/tree/main/crates/proof) (package alias `wordcraft-proof`) | — | Proofing moved to craft-libs (CP1, 2026-11): fifteen spelling dictionaries: English (US, GB, CA), Polish, Portuguese, Italian, Croatian, Turkish, Russian, Czech, Indonesian and Spanish (Spain, Mexico, Latin America, Uruguay); language modules provide spelling, suggestions and grammar, while Bulgarian and German have hyphenation only. Its word lists and patterns are craft-i18n's `language-data/`, embedded from `CRAFT_I18N_DIR` at the commit in `craft-i18n.rev` (CI and releases set `CRAFT_I18N_REQUIRED=1`); without them English is not spell-checked and hyphenation is off |
+| L1 (external) | [`craft-proof`](https://github.com/storytold/craft-libs/tree/main/crates/proof) (package alias `wordcraft-proof`) | — | Proofing moved to craft-libs (CP1, 2026-11): fifteen spelling dictionaries: English (US, GB, CA), Polish, Portuguese, Italian, Croatian, Turkish, Russian, Czech, Indonesian and Spanish (Spain, Mexico, Latin America, Uruguay); language modules provide spelling, suggestions and grammar, while Bulgarian and German have hyphenation only. Its word lists and patterns are craft-i18n's `language-data/`, embedded from `CRAFT_I18N_DIR` at the commit in `craft-i18n.rev` (CI and releases set `CRAFT_I18N_REQUIRED=1`); without them English spelling falls back to an installed Hunspell English or craft-proof's built-in Moby list, and pattern hyphenation is off (Bulgarian, Turkish and Russian hyphenate by rule) |
 | L2 | [`docbin`](../crates/docbin/src) | 8,815 | Word 97–2003 `.doc` reader (FIB, piece table, FKPs, SPRMs, stylesheet, lists, tables, notes, pictures) over a CFB container |
 | L2 | [`docx`](../crates/docx/src) | 37,376 | OOXML WordprocessingML reader/writer (Transitional and Strict), OMML, custom properties and preserved VBA parts |
 | L2 | [`formats`](../crates/formats/src) | 23,058 | TXT, Markdown, HTML, RTF, ODT and LaTeX import/export through the shared document model |

docs/spelling-dictionaries.md

@@ -1,6 +1,6 @@
 # Installing spelling dictionaries
 
-> **Last reviewed:** 2026-10-16 · **Last updated:** 2026-11-01 · **Change:** trivial (WordCraft's own dictionaries folder is searched on macOS and Linux too, as on Windows, so a dictionary dropped there works without a setting; previously minor: ISO 8859-9 Turkish dictionaries are read, #3692; previously minor: Spanish: `es_ES`, `es_MX`, `es_419` and `es_UY` with their fallbacks, #3360, #3553; previously trivial: every Hunspell language shows (no dictionary), not only Polish and Portuguese, and Indonesian is listed, #3150; previously trivial: ISO 8859-1 dictionaries are read, #2824)
+> **Last reviewed:** 2026-10-16 · **Last updated:** 2026-11-02 · **Change:** trivial (the Flatpak's own dictionaries folder; previously trivial: WordCraft's own dictionaries folder is searched on macOS and Linux too, as on Windows, so a dictionary dropped there works without a setting; previously minor: ISO 8859-9 Turkish dictionaries are read, #3692; previously minor: Spanish: `es_ES`, `es_MX`, `es_419` and `es_UY` with their fallbacks, #3360, #3553; previously trivial: every Hunspell language shows (no dictionary), not only Polish and Portuguese, and Indonesian is listed, #3150; previously trivial: ISO 8859-1 dictionaries are read, #2824)
 
 English spelling works out of the box. Polish, Russian, Portuguese, Czech, Italian, Croatian,
 Turkish, Indonesian and Spanish spelling use a standard **Hunspell** dictionary (an `.aff` and a `.dic` file) that
@@ -64,6 +64,12 @@ The first folder holding a matching pair wins:
 | macOS | `~/Library/Application Support/WordCraft/dictionaries`, `~/Library/Spelling`, `/Library/Spelling`, `/opt/homebrew/share/hunspell`, `/usr/local/share/hunspell`, `/Applications/LibreOffice.app/Contents/Resources/extensions/dict-*` |
 | Linux and BSD | `$XDG_DATA_HOME/wordcraft/dictionaries` (default `~/.local/share/wordcraft/dictionaries`), `~/.local/share/hunspell`, `/usr/share/hunspell`, `/usr/share/myspell`, `/usr/share/myspell/dicts`, `/usr/local/share/hunspell`, `/usr/local/share/myspell`, LibreOffice's `share/extensions/dict-*` under `/usr/lib/libreoffice` or `/opt/libreoffice` |
 
+The Flatpak runs in a sandbox: its `$XDG_DATA_HOME` is
+`~/.var/app/ai.storyteller.wordcraft/data`, so its own folder is
+`~/.var/app/ai.storyteller.wordcraft/data/wordcraft/dictionaries`, and it can't see the host's
+`/usr/share/hunspell` or `~/.local/share/hunspell` (the paths above are for the other builds).
+Put dictionaries in that folder.
+
 The web version can't read these folders, so these languages aren't spell-checked there.
 
 ## Example: Brazilian Portuguese (`pt_BR`)
@@ -140,6 +146,7 @@ Never commit dictionary files to this repository: they stay on the user's system
 
 | Date | Change | Summary |
 |---|---|---|
+| 2026-11-02 | trivial | The Flatpak's own dictionaries folder (`~/.var/app/ai.storyteller.wordcraft/data/wordcraft/dictionaries`); the host's folders are outside its sandbox |
 | 2026-10-22 | minor | Spanish (#3360, #3553): the file names for Spain, Mexico, Latin America and Uruguay and their fallbacks |
 | 2026-10-22 | trivial | Italian, Croatian, Turkish, Russian, Czech and Indonesian without a dictionary now show **(no dictionary)** (or **(dictionary unreadable)**) too, as this page already said; before, only Polish and Portuguese did. Indonesian is listed with its file names (#3150) |
 | 2026-11-01 | trivial | WordCraft's own `dictionaries` folder under the user's application data is searched on macOS (`~/Library/Application Support/WordCraft/dictionaries`) and Linux/BSD (`$XDG_DATA_HOME/wordcraft/dictionaries`) too, before the system folders, as it was on Windows (craft-proof; ported from storytold/craft-libs#31 by brzuchal) |