[JSC] Use UNumberFormatter unconditionally
Created attachment 407564 [details] Patch
Created attachment 407565 [details] Patch
Comment on attachment 407565 [details] Patch Hm, maybe, we should check `OS(DARWIN)`.
Comment on attachment 407565 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=407565&action=review > Source/JavaScriptCore/runtime/IntlNumberFormat.cpp:43 > +// They later become stable APIs. > +#if U_ICU_VERSION_MAJOR_NUM < 64 > +static constexpr UNumberFormatFields UNUM_MEASURE_UNIT_FIELD = static_cast<UNumberFormatFields>(11); > +static constexpr UNumberFormatFields UNUM_COMPACT_FIELD = static_cast<UNumberFormatFields>(12); The comment here seems a bit misleading -- these symbols don't even exist as draft API before that point, right? > Source/JavaScriptCore/runtime/IntlNumberFormat.cpp:46 > +static_assert(static_cast<int32_t>(UNUM_COMPACT_FIELD ) == 12); Whoops, style checker is rightly complaining about the extra space here. > Source/JavaScriptCore/runtime/IntlNumberFormat.h:37 > +#define U_HIDE_DRAFT_API Hmm, seems a bit odd that we #ifdef U_HIDE_DRAFT_API and then re-#define it unconditionally. > JSTests/test262/expectations.yaml:1638 > +test/intl402/NumberFormat/prototype/format/engineering-scientific-de-DE.js: So are these all now failing for different reasons that we can address separately then?
<rdar://problem/68417474>