Implement Unicode RegExp simple case folding#640
Merged
Conversation
Add generated Unicode CaseFolding.txt simple/common mappings to the RegExp Unicode resource and use them for Unicode-aware ignoreCase matching. This expands compiled character sets for /iu and /iv, canonicalizes Unicode backreference comparisons, and keeps non-Unicode /i behavior separate. Covers the ECMAScript simple-folding distinction that maps ſ to s, K to k, and ẞ to ß without applying full expansions such as ß to ss.
|
The latest updates on your projects. Learn more about Vercel for GitHub. 1 Skipped Deployment
|
Contributor
|
ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughImplements Unicode simple case folding for RegExp case-insensitive matching: adds UCD parsing and embedded pair caches, exposes folding/canonicalization APIs, refactors compiler char-class emission to apply folding, updates VM backreference comparisons to use Unicode-aware canonicalization, adds tests, and updates docs. ChangesUnicode Simple Case Folding Implementation
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related PRs
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. Comment |
Contributor
Benchmark Results407 benchmarks Interpreted: 🟢 55 improved · 🔴 69 regressed · 283 unchanged · avg +0.4% arraybuffer.js — Interp: 🔴 6, 8 unch. · avg -0.5% · Bytecode: 🔴 12, 2 unch. · avg -7.5%
arrays.js — Interp: 🔴 1, 18 unch. · avg -0.5% · Bytecode: 🔴 18, 1 unch. · avg -9.8%
async-await.js — Interp: 6 unch. · avg -1.5% · Bytecode: 🔴 4, 2 unch. · avg -11.6%
async-generators.js — Interp: 2 unch. · avg -0.9% · Bytecode: 🔴 1, 1 unch. · avg -12.0%
base64.js — Interp: 10 unch. · avg -0.4% · Bytecode: 🟢 7, 🔴 3 · avg +1.9%
classes.js — Interp: 🟢 1, 🔴 3, 27 unch. · avg -0.4% · Bytecode: 🔴 20, 11 unch. · avg -7.2%
closures.js — Interp: 🔴 3, 8 unch. · avg -1.1% · Bytecode: 🔴 11 · avg -12.7%
collections.js — Interp: 12 unch. · avg -1.2% · Bytecode: 🔴 12 · avg -10.6%
csv.js — Interp: 🟢 4, 9 unch. · avg +1.1% · Bytecode: 🔴 12, 1 unch. · avg -9.3%
destructuring.js — Interp: 🟢 1, 🔴 5, 16 unch. · avg -1.3% · Bytecode: 🔴 21, 1 unch. · avg -10.3%
fibonacci.js — Interp: 🔴 3, 5 unch. · avg -1.5% · Bytecode: 🔴 8 · avg -11.6%
float16array.js — Interp: 🟢 1, 🔴 4, 27 unch. · avg -0.4% · Bytecode: 🟢 4, 🔴 27, 1 unch. · avg -5.9%
for-of.js — Interp: 🟢 3, 4 unch. · avg +1.1% · Bytecode: 🔴 7 · avg -11.4%
generators.js — Interp: 4 unch. · avg -2.1% · Bytecode: 🔴 4 · avg -8.0%
iterators.js — Interp: 🟢 24, 18 unch. · avg +3.2% · Bytecode: 🔴 40, 2 unch. · avg -8.3%
json.js — Interp: 🟢 1, 19 unch. · avg +1.1% · Bytecode: 🔴 20 · avg -9.9%
jsx.jsx — Interp: 🔴 18, 3 unch. · avg -5.3% · Bytecode: 🔴 19, 2 unch. · avg -8.9%
modules.js — Interp: 9 unch. · avg -0.2% · Bytecode: 🔴 9 · avg -14.2%
numbers.js — Interp: 11 unch. · avg -1.2% · Bytecode: 🔴 11 · avg -8.7%
objects.js — Interp: 7 unch. · avg -0.9% · Bytecode: 🔴 7 · avg -9.6%
promises.js — Interp: 🟢 2, 🔴 1, 9 unch. · avg -0.5% · Bytecode: 🔴 11, 1 unch. · avg -8.5%
regexp.js — Interp: 🟢 1, 🔴 2, 8 unch. · avg -17.2% · Bytecode: 🟢 6, 🔴 5 · avg -13.7%
strings.js — Interp: 🟢 3, 🔴 2, 14 unch. · avg -0.4% · Bytecode: 🔴 19 · avg -10.9%
tsv.js — Interp: 🔴 3, 6 unch. · avg -2.6% · Bytecode: 🔴 8, 1 unch. · avg -6.6%
typed-arrays.js — Interp: 🟢 10, 12 unch. · avg +26.8% · Bytecode: 🔴 18, 4 unch. · avg -9.5%
uint8array-encoding.js — Interp: 🟢 2, 🔴 8, 8 unch. · avg -4.3% · Bytecode: 🟢 5, 🔴 9, 4 unch. · avg +7.8%
weak-collections.js — Interp: 🟢 2, 🔴 10, 3 unch. · avg -1.6% · Bytecode: 🟢 1, 🔴 13, 1 unch. · avg -21.7%
Deterministic profile diffDeterministic profile diff: no significant changes. Measured on ubuntu-latest x64. Benchmark ranges compare cached main-branch min/max ops/sec with the PR run; overlapping ranges are treated as unchanged noise. Percentage deltas are secondary context. |
Contributor
Suite TimingTest Runner (interpreted: 9,307 passed; bytecode: 9,307 passed)
MemoryGC rows aggregate the main thread plus all worker thread-local GCs. Test runner worker shutdown frees thread-local heaps in bulk; that shutdown reclamation is not counted as GC collections or collected objects.
Benchmarks (interpreted: 407; bytecode: 407)
MemoryGC rows aggregate the main thread plus all worker thread-local GCs. Benchmark runner performs explicit between-file collections, so collection and collected-object counts can be much higher than the test runner.
Measured on ubuntu-latest x64. |
Contributor
test262 Conformance
Areas closest to 100%
Per-test deltas (+8 / -0)Newly passing (8):
Steady-state failures are non-blocking; regressions vs the cached main baseline (lower total pass count, or any PASS → non-PASS transition) fail the conformance gate. Measured on ubuntu-latest x64, bytecode mode. Areas grouped by the first two test262 path components; minimum 25 attempted tests, areas already at 100% excluded. Δ vs main compares against the most recent cached |
Handle non-Unicode RegExp ignoreCase with generated simple uppercase mappings, filtered by the ECMAScript non-ASCII-to-ASCII guard. Also handle /iu property escapes with complement-before-canonicalization semantics for \P{...}, and decode UTF-8 pattern literals consistently for non-u RegExp source text.
coderabbitai
Bot
added
new feature
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@source/units/Goccia.RegExp.Compiler.pas`:
- Around line 454-459: The negated-path currently only runs the
complement-before-folding branch when ANegated and FModifier.IgnoreCase and
FUnicode and not FUnicodeSets, which leaves top-level \P{...} in Unicode-sets
mode using the old negate-after-folding logic; change the condition in the block
that creates FoldRanges/ReduceUnicodeSimpleCaseFoldClosed/EmitRawCharClassRanges
so it triggers whenever ANegated and FModifier.IgnoreCase and FUnicode (i.e.,
remove the not FUnicodeSets guard) so the complement-before-folding path is
applied in `/iv` as well; update the relevant code around FoldRanges,
CharRangesToUnicodeRanges, ReduceUnicodeSimpleCaseFoldClosed,
EmitRawCharClassRanges and UnicodeRangesToCharRanges and add a small `/iv`
regression test mirroring the `/iu` cases.
In `@source/units/Goccia.RegExp.UnicodeData.pas`:
- Around line 767-770: The empty stub ExpandRegExpNonUnicodeCaseFolding must
implement the non-embedded ASCII case-fold expansion so ignore-case matching
works when GOCCIA_REGEXP_EMBEDDED_UCD is off: iterate the incoming ARanges
(TUnicodePropertyRangeArray) and for any ASCII letters add their opposite-case
codepoint ranges (e.g., map 'a'-'z' to 'A'-'Z' and vice versa) or expand
single-letter ranges to include the ASCII counterpart, merging/normalizing
overlaps so classes remain sorted/unique; update the routine used by
EmitCharMatch to call this so /i and character-class forms ([a] vs a) behave
identically—note RegExpCanonicalizeCodePoint handles backreferences only, so
keep this ASCII-only expansion here and preserve array normalization semantics
used elsewhere.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: afd38f83-f6f2-41e4-96f5-dc59362c3c05
⛔ Files ignored due to path filters (2)
source/generated/Generated.UnicodeData.pasis excluded by!**/generated/**source/generated/Generated.UnicodeData.resis excluded by!**/generated/**
📒 Files selected for processing (6)
docs/built-ins.mdscripts/generate-unicode-data.jssource/units/Goccia.RegExp.Compiler.passource/units/Goccia.RegExp.UnicodeData.passource/units/Goccia.RegExp.VM.pastests/built-ins/RegExp/unicode.js
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
/iuand/iv.CaseFolding/Simple, expand compiled character sets through simple folds, and canonicalize Unicode backreference comparisons.ßdoes not matchSS.Testing
Verification run:
./build.pas testrunner && ./build/GocciaTestRunner tests/built-ins/RegExp --asi./build/GocciaTestRunner tests/built-ins/RegExp --asi --mode=bytecode./build.pas loader && ./build/GocciaScriptLoader /tmp/goccia-regexp-615-fixed.js --asi./format.pas --check./fixtures/ffi/build.sh && ./build/GocciaTestRunner tests --asi --unsafe-ffi