On Sun, 17 Jan 2021 14:56:40 GMT, Peter Levart <plev...@openjdk.org> wrote:
>> Claes Redestad has updated the pull request incrementally with one
>> additional commit since the last revision:
>>
>> Simplify lookupCharset
>
> This looks good.
> Are you planning to do similar things for encoding too?
I already approved the changes and they are OK. Maybe for a followup: just
noticing after the fact that logic for `newStringUTF8NoRepl(....)` vs. `new
String(...., StringCoding.UTF_8)` differ in handling unmappabale characters,
which is by the spec, but also the constructor contains special handling of
input that only contains non-negative bytes:
if (charset == UTF_8) {
if (COMPACT_STRINGS && !StringCoding.hasNegatives(bytes, offset,
length)) {
this.value = Arrays.copyOfRange(bytes, offset, offset + length);
this.coder = LATIN1;
return;
...while `newStringUTF8NoRepl(....)` does not contain this optimization. I
guess ZipCoder could benefit from that optimization too since paths are mostly
ASCII only. So WDYT of this additional simplification/consolidation of UTF-8
decoding:
https://github.com/plevart/jdk/commit/0b8b12c998e4ed451588442205ebe8f7423db7d8
-------------
PR: https://git.openjdk.java.net/jdk/pull/2102
