buffer: consolidate encoding parsing

[mscdex]

This consolidates encoding parsing for buffer at both the JS and C++ levels (only fill() still parses the encoding twice). Doing so not only cleans things up but also improves performance in some cases.

Various relevant benchmark results:

 confidence improvement accuracy (*) (**) (***)
buffers/buffer-bytelength.js n=10000000 len=16 encoding='base64' * 4.03 % ±3.40% ±4.53% ±5.90%
buffers/buffer-bytelength.js n=10000000 len=16 encoding='buffer' *** 17.47 % ±1.09% ±1.46% ±1.90%
buffers/buffer-bytelength.js n=10000000 len=16 encoding='utf8' *** 14.05 % ±2.17% ±2.92% ±3.87%
buffers/buffer-tostring.js n=10000000 len=64 args=0 encoding='ascii' *** 1.05 % ±0.58% ±0.78% ±1.03%
buffers/buffer-tostring.js n=10000000 len=64 args=0 encoding='hex' *** 2.09 % ±0.76% ±1.02% ±1.34%
buffers/buffer-tostring.js n=10000000 len=64 args=0 encoding='latin1' *** 1.30 % ±0.49% ±0.65% ±0.85%
buffers/buffer-tostring.js n=10000000 len=64 args=0 encoding='utf8' *** 2.31 % ±0.98% ±1.31% ±1.72%
buffers/buffer-tostring.js n=10000000 len=64 args=1 encoding='ascii' *** 8.75 % ±1.22% ±1.64% ±2.17%
buffers/buffer-tostring.js n=10000000 len=64 args=1 encoding='hex' *** 5.16 % ±0.71% ±0.94% ±1.23%
buffers/buffer-tostring.js n=10000000 len=64 args=1 encoding='latin1' *** 7.52 % ±1.41% ±1.88% ±2.46%
buffers/buffer-tostring.js n=10000000 len=64 args=1 encoding='UCS-2' *** 7.93 % ±1.40% ±1.86% ±2.43%
buffers/buffer-tostring.js n=10000000 len=64 args=1 encoding='utf8' *** 9.29 % ±1.19% ±1.60% ±2.11%
buffers/buffer-tostring.js n=10000000 len=64 args=3 encoding='ascii' *** 9.34 % ±0.89% ±1.19% ±1.57%
buffers/buffer-tostring.js n=10000000 len=64 args=3 encoding='hex' *** 4.83 % ±0.60% ±0.80% ±1.04%
buffers/buffer-tostring.js n=10000000 len=64 args=3 encoding='latin1' *** 10.68 % ±1.50% ±2.01% ±2.66%
buffers/buffer-tostring.js n=10000000 len=64 args=3 encoding='UCS-2' *** 4.50 % ±1.19% ±1.58% ±2.07%
buffers/buffer-tostring.js n=10000000 len=64 args=3 encoding='utf8' *** 9.86 % ±0.84% ±1.12% ±1.48%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='' ** 1.87 % ±1.20% ±1.60% ±2.09%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='ascii' *** 8.84 % ±2.47% ±3.30% ±4.31%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='hex' *** 1.24 % ±0.56% ±0.75% ±0.98%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='latin1' *** 8.56 % ±1.02% ±1.36% ±1.77%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='utf16le' *** 5.39 % ±0.68% ±0.91% ±1.18%
buffers/buffer-write-string.js n=10000000 len=2048 args='' encoding='utf8' *** 2.59 % ±1.09% ±1.45% ±1.89%
buffers/buffer-from.js n=4000000 len=100 source='string' *** 3.36 % ±1.69% ±2.24% ±2.92%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='@' *** 13.54 % ±4.29% ±5.71% ±7.43%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='</i> to the Caterpillar' * 0.57 % ±0.48% ±0.63% ±0.82%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='aaaaaaaaaaaaaaaaa' * 0.44 % ±0.43% ±0.58% ±0.75%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='Alice' *** 18.63 % ±4.71% ±6.27% ±8.16%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='Gryphon' *** 3.35 % ±1.91% ±2.55% ±3.31%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='neighbouring pool' * 0.48 % ±0.46% ±0.61% ±0.80%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='ucs2' search='SQ' *** 3.65 % ±1.69% ±2.25% ±2.92%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='utf8' search='@' *** 16.95 % ±3.99% ±5.31% ±6.92%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='utf8' search='Alice' *** 18.55 % ±5.48% ±7.30% ±9.52%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='utf8' search='Gryphon' ** 3.03 % ±1.78% ±2.36% ±3.08%
buffers/buffer-indexof.js n=100000 type='buffer' encoding='utf8' search='Ou est ma chatte?' * 0.78 % ±0.66% ±0.88% ±1.14%
buffers/buffer-indexof.js n=100000 type='string' encoding='ucs2' search='@' *** 14.38 % ±5.38% ±7.16% ±9.32%
buffers/buffer-indexof.js n=100000 type='string' encoding='ucs2' search='Alice' *** 27.09 % ±5.71% ±7.60% ±9.91%
buffers/buffer-indexof.js n=100000 type='string' encoding='ucs2' search='SQ' *** 3.63 % ±2.07% ±2.75% ±3.58%
buffers/buffer-indexof.js n=100000 type='string' encoding='utf8' search='@' *** 12.08 % ±4.53% ±6.03% ±7.85%
buffers/buffer-indexof.js n=100000 type='string' encoding='utf8' search='Alice' *** 15.36 % ±5.71% ±7.60% ±9.90%
buffers/buffer-indexof.js n=100000 type='string' encoding='utf8' search='Gryphon' * 1.88 % ±1.87% ±2.48% ±3.23%
buffers/buffer-indexof.js n=100000 type='string' encoding='utf8' search='Ou est ma chatte?' * 0.88 % ±0.78% ±1.04% ±1.36%
buffers/buffer-indexof.js n=100000 type='string' encoding='utf8' search='SQ' * 3.60 % ±3.04% ±4.04% ±5.27%

Checklist

• make -j4 test (UNIX), or vcbuild test (Windows) passes
• commit message follows commit guidelines

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/25112/

[Trott]

@nodejs/buffer This could use some reviews.

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/25173/

[bnoordhuis]

Nice work, some impressive speedups!

[bnoordhuis]

Maybe add a comment that for performance reasons it's imperative that all the entries have the same shape / hidden map? (I.e., users of an ops object shouldn't go poly- or megamorphic.)

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/25208/ ✔️

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/25242/