Revert AVX512-optimized is_ascii (#22)

[bonega]

Reverts the custom AVX512 is_ascii() from #22 (commit 65ad00d).

That implementation was a workaround for poor codegen from stdlib's is_ascii() under -C target-cpu=native on AVX512 CPUs (~30x slowdown). The optimization has since been contributed to and merged into Rust std, so the workaround is no longer needed — falling back to stdlib is_ascii() now gets the fast path for free.

Changes

• Remove src/simd.rs
• Restore stdlib is_ascii() calls in decoder/complete.rs, decoder/incomplete.rs, and encoder.rs
• Drop the simd module from lib.rs

The nix.yml fail-threshold bump (120% → 130%) from #22 is kept.

Build verified locally with cargo build.

[github-actions]

Yore Benchmarks

Details

Benchmark suite	Current: a94753a	Previous: 65ad00d	Ratio
decode_checked/mostly_ascii/8	30 ns/iter (± 10)	34 ns/iter (± 12)	0.88
decode_checked/mostly_ascii/64	62 ns/iter (± 7)	67 ns/iter (± 8)	0.93
decode_checked/mostly_ascii/256	142 ns/iter (± 13)	147 ns/iter (± 42)	0.97
decode_checked/mostly_ascii/512	243 ns/iter (± 20)	240 ns/iter (± 20)	1.01
decode_checked/mostly_ascii/1024	446 ns/iter (± 29)	419 ns/iter (± 29)	1.06
decode_checked/mostly_ascii/2048	887 ns/iter (± 37)	911 ns/iter (± 123)	0.97
decode_checked/mostly_ascii/4096	1971 ns/iter (± 110)	1998 ns/iter (± 71)	0.99
decode_checked/ascii/8	10 ns/iter (± 0)	9 ns/iter (± 0)	1.11
decode_checked/ascii/64	9 ns/iter (± 0)	8 ns/iter (± 0)	1.13
decode_checked/ascii/256	11 ns/iter (± 0)	10 ns/iter (± 0)	1.10
decode_checked/ascii/512	14 ns/iter (± 0)	16 ns/iter (± 0)	0.88
decode_checked/ascii/1024	27 ns/iter (± 0)	26 ns/iter (± 0)	1.04
decode_checked/ascii/2048	48 ns/iter (± 0)	52 ns/iter (± 0)	0.92
decode_checked/ascii/4096	93 ns/iter (± 0)	92 ns/iter (± 1)	1.01
decode_checked/extended/8	34 ns/iter (± 1)	37 ns/iter (± 3)	0.92
decode_checked/extended/64	85 ns/iter (± 3)	87 ns/iter (± 2)	0.98
decode_checked/extended/256	254 ns/iter (± 21)	266 ns/iter (± 11)	0.95
decode_checked/extended/512	483 ns/iter (± 9)	493 ns/iter (± 8)	0.98
decode_checked/extended/1024	909 ns/iter (± 6)	910 ns/iter (± 10)	1.00
decode_checked/extended/2048	1768 ns/iter (± 15)	1765 ns/iter (± 42)	1.00
decode_checked/extended/4096	3489 ns/iter (± 34)	3481 ns/iter (± 45)	1.00
decode_lossy/all_bad/8	39 ns/iter (± 2)	48 ns/iter (± 5)	0.81
decode_lossy/all_bad/64	62 ns/iter (± 1)	64 ns/iter (± 1)	0.97
decode_lossy/all_bad/256	174 ns/iter (± 2)	178 ns/iter (± 3)	0.98
decode_lossy/all_bad/512	329 ns/iter (± 10)	333 ns/iter (± 10)	0.99
decode_lossy/all_bad/1024	626 ns/iter (± 34)	646 ns/iter (± 27)	0.97
decode_lossy/all_bad/2048	1223 ns/iter (± 16)	1272 ns/iter (± 18)	0.96
decode_lossy/all_bad/4096	2440 ns/iter (± 24)	2491 ns/iter (± 41)	0.98
decode_lossy/mostly_ascii/8	39 ns/iter (± 14)	48 ns/iter (± 19)	0.81
decode_lossy/mostly_ascii/64	64 ns/iter (± 6)	66 ns/iter (± 7)	0.97
decode_lossy/mostly_ascii/256	143 ns/iter (± 13)	154 ns/iter (± 17)	0.93
decode_lossy/mostly_ascii/512	231 ns/iter (± 20)	246 ns/iter (± 20)	0.94
decode_lossy/mostly_ascii/1024	408 ns/iter (± 24)	431 ns/iter (± 30)	0.95
decode_lossy/mostly_ascii/2048	777 ns/iter (± 35)	810 ns/iter (± 46)	0.96
decode_lossy/mostly_ascii/4096	1500 ns/iter (± 49)	1571 ns/iter (± 49)	0.95
encode_checked/mostly_ascii/8	41 ns/iter (± 16)	43 ns/iter (± 17)	0.95
encode_checked/mostly_ascii/64	128 ns/iter (± 5)	139 ns/iter (± 2)	0.92
encode_checked/mostly_ascii/256	442 ns/iter (± 9)	481 ns/iter (± 41)	0.92
encode_checked/mostly_ascii/512	872 ns/iter (± 16)	939 ns/iter (± 50)	0.93
encode_checked/mostly_ascii/1024	1687 ns/iter (± 34)	2005 ns/iter (± 109)	0.84
encode_checked/mostly_ascii/2048	3438 ns/iter (± 54)	4090 ns/iter (± 116)	0.84
encode_checked/mostly_ascii/4096	6672 ns/iter (± 166)	7311 ns/iter (± 309)	0.91
encode_checked/ascii/8	10 ns/iter (± 0)	9 ns/iter (± 0)	1.11
encode_checked/ascii/64	9 ns/iter (± 0)	8 ns/iter (± 0)	1.13
encode_checked/ascii/256	12 ns/iter (± 0)	10 ns/iter (± 1)	1.20
encode_checked/ascii/512	16 ns/iter (± 0)	14 ns/iter (± 0)	1.14
encode_checked/ascii/1024	25 ns/iter (± 0)	24 ns/iter (± 0)	1.04
encode_checked/ascii/2048	50 ns/iter (± 0)	49 ns/iter (± 0)	1.02
encode_checked/ascii/4096	93 ns/iter (± 4)	89 ns/iter (± 0)	1.04
encode_checked/extended/8	54 ns/iter (± 0)	53 ns/iter (± 0)	1.02
encode_checked/extended/64	190 ns/iter (± 5)	202 ns/iter (± 1)	0.94
encode_checked/extended/256	789 ns/iter (± 9)	705 ns/iter (± 7)	1.12
encode_checked/extended/512	1366 ns/iter (± 24)	1375 ns/iter (± 19)	0.99
encode_checked/extended/1024	2710 ns/iter (± 26)	2726 ns/iter (± 20)	0.99
encode_checked/extended/2048	5389 ns/iter (± 334)	5400 ns/iter (± 49)	1.00
encode_checked/extended/4096	10698 ns/iter (± 559)	10749 ns/iter (± 34)	1.00
encode_lossy/all_bad/8	49 ns/iter (± 1)	54 ns/iter (± 12)	0.91
encode_lossy/all_bad/64	215 ns/iter (± 8)	247 ns/iter (± 1)	0.87
encode_lossy/all_bad/256	801 ns/iter (± 7)	877 ns/iter (± 141)	0.91
encode_lossy/all_bad/512	1572 ns/iter (± 12)	1717 ns/iter (± 24)	0.92
encode_lossy/all_bad/1024	3119 ns/iter (± 218)	3412 ns/iter (± 64)	0.91
encode_lossy/all_bad/2048	6186 ns/iter (± 39)	6764 ns/iter (± 131)	0.91
encode_lossy/all_bad/4096	12311 ns/iter (± 113)	13475 ns/iter (± 61)	0.91

This comment was automatically generated by workflow using github-action-benchmark.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Yore Benchmarks'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.10.

Benchmark suite	Current: a94753a	Previous: 65ad00d	Ratio
decode_checked/ascii/8	10 ns/iter (± 0)	9 ns/iter (± 0)	1.11
decode_checked/ascii/64	9 ns/iter (± 0)	8 ns/iter (± 0)	1.13
encode_checked/ascii/8	10 ns/iter (± 0)	9 ns/iter (± 0)	1.11
encode_checked/ascii/64	9 ns/iter (± 0)	8 ns/iter (± 0)	1.13
encode_checked/ascii/256	12 ns/iter (± 0)	10 ns/iter (± 1)	1.20
encode_checked/ascii/512	16 ns/iter (± 0)	14 ns/iter (± 0)	1.14
encode_checked/extended/256	789 ns/iter (± 9)	705 ns/iter (± 7)	1.12

This comment was automatically generated by workflow using github-action-benchmark.