Repository navigation
Should Japanese 旧姓 (former surname) populate the maiden field? (山田（旧姓：佐藤）) #309
Description
Activity
Punting in favour of #329, which covers this and generalizes it.
Two findings changed the shape of the ask.
The Japanese-specific part belongs in the default vocabulary, not the pack.
旧姓satisfies the rule that already admitted the Cyrillic markers — "native-script entries cannot collide with Latin-script names, which is what makes them safe to enable by default." Matching is whole-token,_normalize("旧姓")is旧姓, and neither character appears in any shipped surname, title, suffix, conjunction, particle or bound-given vocabulary. It sits alongsideурожденнаяandgeboren, not behind a Japanese-data declaration —locales.JAexists for things that need to know the data is Japanese, and a Han-script marker can only ever match Han text.Measured with it added:
山田花子 旧姓 佐藤→ family山田花子, maiden佐藤;山田 花子 旧姓 佐藤→ given花子, family山田, maiden佐藤.But the form this issue's title uses still does not work, and that half is language-agnostic:
Parser(policy=Policy(maiden_delimiters={('（','）')})) 山田（旧姓：佐藤） -> maiden '旧姓：佐藤' # marker and colon ride along
extract_delimitedassigns bracketed contentRole.MAIDENwholesale beforeclassifyruns, so the marker inside is never tagged and_group's consuming rule never fires.néehas the identical asymmetry —Jane Smith née Jonesgives maidenJones,Jane Smith (née Jones)gives maidennée Jones.So the vocabulary and the extraction fix are worth little apart, and are filed together as #329. Reopen if that lands and Japanese still needs something specific.
Worth recording separately: patching
maiden_delimitersinlocales.JAis not an option either way.（）is already a nickname delimiter, and_extractresolves the collision by exclusion — "a pair listed in maiden_delimiters is dropped from the effective nickname set" — so it would trade fullwidth-paren nicknames for maiden names wholesale. Measured:山田 太郎（マイケル・ジャクソン）goes from nickname to maiden. That collision is a real ambiguity in the writing system; the only thing separating the two conventions is the marker inside the brackets.- added a commit that references this issue
on Aug 3, 2026 - added a commit that references this issue
on Aug 22, 2026
Japanese documents record a pre-marriage surname as 旧姓, conventionally parenthesized with a colon: 山田花子（旧姓：佐藤）. Today the fullwidth parens are nickname delimiters (#273), so 旧姓：佐藤 presumably lands in
nicknameas literal text. The maiden-marker mechanism (#274: née, geb.) routes a following name tomaiden, but it operates on tokens in the main stream — this convention puts the marker inside a delimited region, plus a colon.Settling this needs: whether marker recognition should reach into extracted regions (a mechanism change), whether 旧姓 alone (spaced, unparenthesized: 佐藤 旧姓 山田?) occurs enough to matter, and what the colon does to tokenization. Filed as a question rather than a plan — the answer may be "document as unsupported."