On Wed, 4 Nov 2026 18:33:23 GMT, Justin Lu <[email protected]> wrote:
>> Naoto Sato has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Addresses review comments
>
> make/jdk/src/classes/build/tools/generatecharacter/SpecialCaseMap.java line
> 64:
>
>> 62: SpecialCaseMap item = parse(line.trim());
>> 63: if (item != null &&
>> 64: item.getCharSource() >> 16 == plane) {
>
> I presume it is intentional that we no longer end early when the item's plane
> is greater than the passed plane? Is it worth describing the expectations in
> the method javadoc? i.e., the items in the file are no longer ordered by
> plane.
Yes, it was intentional. I am not sure whether to explain it in the comment, as
it was an incorrect assumption on the order of the input lines, which was never
claimed to be guaranteed. The modified code does what it is expected to do. So
I'd just leave it as it is.
> src/java.base/share/classes/java/lang/StringLatin1.java line 400:
>
>> 398: if (cf1 != cf2 ||
>> 399: // Special case for Sharp-S
>> 400: cp1 == 0x00DF) {
>
> Instead of hard-coding against "Sharp-S", can check the chars on `cp2` to see
> if it is not `1`? If that works, it would support future Unicode update
> additions that share this same behavior better.
Good point. Changed the code to look at the length difference, instead of code
point checking
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/33705#discussion_r4472433551
PR Review Comment: https://git.openjdk.org/jdk/pull/33705#discussion_r4472446795
