[derek73]

Closes #751.

This moves the last 44 words out of nameparser/vocab/und.toml into the files of the languages whose names they occur in. That leaves und.toml with only bon and dí: two particles with no record of which names they come from, whose removal is #766's question. A comment at the top of the file says so.

No compiled set moves. tools/vocab/build.py --check reports _compiled.py current, tools/differential/snapshot.py matches, and the word-order check passes.

Where they went

Each word went where its own record put it:

• Conjunctions follow Provide constants in non-Latin scripts (Cyrillic, Greek, Arabic, Hebrew) #269, Josep Carod i Rovira reads family Rovira — Catalan and Polish link surnames with i, which is not a conjunction #397 and their comments:
  • en: and, of, the
  • fr: et
  • de: und
  • es: y
  • pt and it: e
  • ca and pl: i
  • ru and bg: и
  • uk: і, й, та
  • el: και
  • ar: و
  • mul: &, a symbol no language owns
• Bound given names all go to ar.toml, both the Latin transliterations and the Arabic script, as Classify the particles in und.toml by language and kind #749 and Classify the titles in und.toml by language and kind #750 sent a transliteration to its source language. abu, أبو and ابو join the teknonymic readings already in that file.
• Birth markers follow the old module docstring (Recognize maiden-name markers like "née" and "geb." (Jane Smith née Jones) #274):
  • fr and en: née, né, nee
  • de: geb, geborene
  • nl: geboren, which the docstring calls Dutch
  • cs: rozená
  • da and nb: født
  • nn: fødd (Nynorsk)
  • sv: född
  • ru: the five Russian spellings
  • pl: z domu (parse("Maria Kowalska z domu Nowak") cannot reach the Polish maiden marker — markers do not compose the way suffixes do #434)
  • ja: 旧姓 (Should Japanese 旧姓 (former surname) populate the maiden field? (山田（旧姓：佐藤）) #309)

ca.toml and nn.toml are new files.

Comments became fields

• The single-letter-connective comment block is gone. Its general reasoning is already in decisions/P8 and rules/P3. Its per-word facts are now notes on e and i, and its keep-outs, y and the Cyrillic и, і and й, are excluded = { may_be_initial = … }.
• The Provide constants in non-Latin scripts (Cyrillic, Greek, Arabic, Hebrew) #269 comment blocks became an issue on each word, plus notes on і, و and عبد. abd's comment is now its note in ar.toml.
• Spellings carry same_as, for example abdel → abdul, й → і, nee → née, and the е spellings → the ё ones.
• Gender is recorded where the form marks it.
• Two keep-outs that were only prose in decisions/vocabulary-exclusions are now exclusions:
  • roz in cs.toml, removed 2026-08-26. I measured it again: with roz listed, Rosalind Roz Smith reads maiden 'Smith' and takes Rosalind for the family.
  • the Scandinavian f. in da.toml, which collides with the initial F.

Docs

• decisions/vocabulary-model has a new part-2 entry.
• decisions/vocabulary-exclusions' conjunctions_ambiguous and maiden-marker blocks now point at the queries.
• nameparser/vocab/README.md's und paragraphs and the RFC's phase table are updated.

Review

Two docs-review rounds, the second on the fix commit:

• The first found that the intro claimed every maiden-marker refusal had become data, that nee was placed against the docstring the entry cites, and that the en.toml née/né had no gender. All three are fixed.
• The second found a README total that the fix commit had made stale. That count is now gone.

Suite: 22,122 passed. mypy and ruff are clean.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.05%. Comparing base (85ac416) to head (9fa0300).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #780   +/-   ##
=======================================
  Coverage   99.05%   99.05%           
=======================================
  Files          47       47           
  Lines        4642     4642           
=======================================
  Hits         4598     4598           
  Misses         44       44           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.