Bug 312596 – [JSC] Intl.Segmenter `isWordLike` returns false for numeric segments due to off-by-one in UBRK_WORD_NONE_LIMIT check

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Alistair]

Intl.Segmenter with granularity:"word" reports isWordLike:false for numeric segments.

  for (const s of new Intl.Segmenter("en", {granularity:"word"}).segment("abc 123 v2"))
    print(s.segment, s.isWordLike);
  // JSC: abc true / 123 false / v2 false
  // V8:  abc true / 123 true  / v2 true

IntlSegmentDataObject.h checks `ruleStatus <= UBRK_WORD_NONE_LIMIT`, but ICU's
UWordBreak ranges are half-open (UBRK_WORD_NONE_LIMIT == UBRK_WORD_NUMBER == 100),
so the comparison should be `<`.

[Alistair]

https://github.com/WebKit/WebKit/pull/62999

[EWS]

Committed 311507@main (0145d169cd9c): <https://commits.webkit.org/311507@main>

Reviewed commits have been landed. Closing PR #62999 and removing active labels.

[Radar WebKit Bug Importer]

<rdar://problem/175057894>