On Tue, 15 Nov 2022 22:23:27 GMT, Roger Riggs <rri...@openjdk.org> wrote:
>> Jim Laskey has updated the pull request incrementally with one additional
>> commit since the last revision:
>>
>> Wrong line separator
>
> src/java.base/share/classes/java/lang/AbstractStringBuilder.java line 1826:
>
>> 1824:
>> 1825: // Used by StringConcatHelper via JLA.
>> 1826: long mix(long lengthCoder) {
>
> `mix` is not a very descriptive name (in all of the places it is used).
> Perhaps a name that captures the combination of coder and length.
>
> `coderAndLength()` or similar
>
> The implementation here can do the wrong thing if the `lengthCoder` already
> includes the bit for UTF16
> and the coder for the AbstractStringBuilder is LATIN1. Or is intentional that
> it will upgrade to UTF16?
> That could definately use a comment.
The goal here is to take the contents of a `StringBuilder/StringBuffer `without
first converting to a string when referenced by `StringConcatFactory`. So the
terminology is a carry over from `StringConcatFactory` and `StringConcatHelper`.
I agree `mix` is weird but historically, it is what it is. Maybe a separate
task to rename is in order.
The implementation is deliberate. If something in the list of things being
concatenated is UTF16, then everything being concatenated is upgraded to UTF16.
I'll add commentary.
-------------
PR: https://git.openjdk.org/jdk/pull/10889
