Commit 18c4dd0
src: replace custom ASCII validation with simdutf one
PR-URL: #46271
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Reviewed-By: Richard Lau <rlau@redhat.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Darshan Sen <raisinten@gmail.com>1 parent 94421b4 commit 18c4dd0
1 file changed
Lines changed: 3 additions & 55 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
@@ -25,6 +25,7 @@ | |||
| 25 | 25 | #include "env-inl.h" | |
| 26 | 26 | #include "node_buffer.h" | |
| 27 | 27 | #include "node_errors.h" | |
| 28 | + #include "simdutf.h" | ||
| 28 | 29 | #include "util.h" | |
| 29 | 30 | ||
| 30 | 31 | #include <climits> | |
@@ -467,60 +468,6 @@ Maybe<size_t> StringBytes::Size(Isolate* isolate, | |||
| 467 | 468 | UNREACHABLE(); | |
| 468 | 469 | } | |
| 469 | 470 | ||
| 470 | - | ||
| 471 | - | ||
| 472 | - | ||
| 473 | - static bool contains_non_ascii_slow(const char* buf, size_t len) { | ||
| 474 | - for (size_t i = 0; i < len; ++i) { | ||
| 475 | - if (buf[i] & 0x80) | ||
| 476 | - return true; | ||
| 477 | - } | ||
| 478 | - return false; | ||
| 479 | - } | ||
| 480 | - | ||
| 481 | - | ||
| 482 | - static bool contains_non_ascii(const char* src, size_t len) { | ||
| 483 | - if (len < 16) { | ||
| 484 | - return contains_non_ascii_slow(src, len); | ||
| 485 | - } | ||
| 486 | - | ||
| 487 | - const unsigned bytes_per_word = sizeof(uintptr_t); | ||
| 488 | - const unsigned align_mask = bytes_per_word - 1; | ||
| 489 | - const unsigned unaligned = reinterpret_cast<uintptr_t>(src) & align_mask; | ||
| 490 | - | ||
| 491 | - if (unaligned > 0) { | ||
| 492 | - const unsigned n = bytes_per_word - unaligned; | ||
| 493 | - if (contains_non_ascii_slow(src, n)) | ||
| 494 | - return true; | ||
| 495 | - src += n; | ||
| 496 | - len -= n; | ||
| 497 | - } | ||
| 498 | - | ||
| 499 | - | ||
| 500 | - #if defined(_WIN64) || defined(_LP64) | ||
| 501 | - const uintptr_t mask = 0x8080808080808080ll; | ||
| 502 | - #else | ||
| 503 | - const uintptr_t mask = 0x80808080l; | ||
| 504 | - #endif | ||
| 505 | - | ||
| 506 | - const uintptr_t* srcw = reinterpret_cast<const uintptr_t*>(src); | ||
| 507 | - | ||
| 508 | - for (size_t i = 0, n = len / bytes_per_word; i < n; ++i) { | ||
| 509 | - if (srcw[i] & mask) | ||
| 510 | - return true; | ||
| 511 | - } | ||
| 512 | - | ||
| 513 | - const unsigned remainder = len & align_mask; | ||
| 514 | - if (remainder > 0) { | ||
| 515 | - const size_t offset = len - remainder; | ||
| 516 | - if (contains_non_ascii_slow(src + offset, remainder)) | ||
| 517 | - return true; | ||
| 518 | - } | ||
| 519 | - | ||
| 520 | - return false; | ||
| 521 | - } | ||
| 522 | - | ||
| 523 | - | ||
| 524 | 471 | static void force_ascii_slow(const char* src, char* dst, size_t len) { | |
| 525 | 472 | for (size_t i = 0; i < len; ++i) { | |
| 526 | 473 | dst[i] = src[i] & 0x7f; | |
@@ -634,7 +581,8 @@ MaybeLocal<Value> StringBytes::Encode(Isolate* isolate, | |||
| 634 | 581 | } | |
| 635 | 582 | ||
| 636 | 583 | case ASCII: | |
| 637 | - if (contains_non_ascii(buf, buflen)) { | ||
| 584 | + if (simdutf::validate_ascii_with_errors(buf, buflen).error) { | ||
| 585 | + // The input contains non-ASCII bytes. | ||
| 638 | 586 | char* out = node::UncheckedMalloc(buflen); | |
| 639 | 587 | if (out == nullptr) { | |
| 640 | 588 | *error = node::ERR_MEMORY_ALLOCATION_FAILED(isolate); | |
0 commit comments