You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Halfwidth corner brackets (山田 「タロー」 タロウ) aren't read as a nickname #597
The full-width Japanese corner brackets 「」 are a default nickname delimiter (#273), but their halfwidth forms 「」 (U+FF62/U+FF63) are not. The bracketed word stays in the name as a middle name, brackets and all:
The bracket also costs the name its order. The unclassified 「 keeps the text from being wholly East Asian script, so 山田 「タロー」 タロウ reads given 山田, family タロウ. With the pair added, the clause is extracted and the rest reads family-first, as 山田 タロウ does since #594:
Why it matters. Halfwidth brackets come from the same source as the halfwidth katakana #594 just classified: legacy JIS X 0201 data, where 「」 were the only corner brackets available. The PR #595 review found the gap.
Proposal. Add ("「", "」") to DEFAULT_NICKNAME_DELIMITERS, beside 「」.
Rationale: the halfwidth pair is the same punctuation as 「」 in a narrower encoding, and appears in no non-CJK text.
Boundaries:John 「Jack」 Smith would read nickname Jack like John 「Jack」 Smith already does. Measured today, it reads middle 「Jack」.
Nothing else needs a halfwidth twin. Unicode has no halfwidth form of 『』 or of the full-width parentheses.
Accepted consequences: this changes a default, so it's a release-log-classified behavior change. Every 「…」 input moves, so the differential needs a case row and a ledger rule.
Activity
Read halfwidth katakana as katakana (#594) (#595)
Merge pull request #599 from derek73/fix/issue-597-halfwidth-corner-b…
Read halfwidth corner brackets as a nickname delimiter (derek73#597)
Fix what the review of derek73#597 found