perf(lexer): use SIMD to skip single line comment

[Boshen]

Speeding up single line comment parsing should save a few millisecond.

On the other hand, skipping multi-line comments in chunks will give us a huge improvement.

[github-actions]

Parser Benchmark Results

group                    main                                   pr
-----                    ----                                   --
parser/babylon.max.js    1.00    126.3±1.23ms    81.7 MB/sec    1.00    126.7±0.75ms    81.5 MB/sec
parser/d3.js             1.00     16.2±0.05ms    33.8 MB/sec    1.01     16.3±0.10ms    33.5 MB/sec
parser/lodash.js         1.00      4.6±0.09ms   112.1 MB/sec    1.00      4.6±0.08ms   111.9 MB/sec
parser/pdf.js            1.00      9.2±0.07ms    43.7 MB/sec    1.01      9.3±0.09ms    43.1 MB/sec
parser/typescript.js     1.00    128.7±0.70ms    74.7 MB/sec    1.00    128.4±0.97ms    74.9 MB/sec

[Boshen]

This is not adding much performance. Closing.