Make String.prototype.matchAll return a lazy iterator#206
Merged
Conversation
Replace the eager array-backed iterator with a lazy TGocciaRegExpMatchAllIteratorValue that computes one match per .next() call, matching the ES2026 spec behavior. The new iterator stores a cloned RegExp and advances FSearchIndex on demand. Both RegExp.prototype[Symbol.matchAll] and the String.prototype.matchAll fallback path now return this lazy iterator. Closes #175 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Contributor
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 5 minutes and 10 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (3)
📝 WalkthroughWalkthroughThe PR converts Changes
Sequence DiagramsequenceDiagram
participant Consumer as Iterator Consumer
participant Iterator as RegExpMatchAllIterator
participant Regex as RegExp Object
participant Engine as Match Engine
Consumer->>Iterator: Create iterator<br/>(regex, input, global)
activate Iterator
Iterator->>Regex: Clone regex
note over Iterator: Store state:<br/>SearchIndex = 0<br/>Global flag
loop For Each next() call
Consumer->>Iterator: next()
Iterator->>Engine: MatchRegExpObject<br/>(input, SearchIndex)
activate Engine
Engine-->>Iterator: Match found<br/>at [start, end]
deactivate Engine
Iterator->>Iterator: Update SearchIndex<br/>to end position
Iterator-->>Consumer: {value: match, done: false}
end
Note over Iterator: No more matches<br/>or done condition met
Iterator-->>Consumer: {value: undefined, done: true}
deactivate Iterator
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Possibly related PRs
Poem
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Contributor
Benchmark Results274 benchmarks Interpreted: 🟢 15 improved · 🔴 133 regressed · 126 unchanged · avg -1.8% arraybuffer.js — Interp: 🟢 2, 🔴 5, 7 unch. · avg -0.5% · Bytecode: 🟢 12, 🔴 1, 1 unch. · avg +8.7%
arrays.js — Interp: 🟢 3, 🔴 6, 10 unch. · avg -0.3% · Bytecode: 🟢 19 · avg +8.7%
async-await.js — Interp: 🔴 6 · avg -2.3% · Bytecode: 🟢 6 · avg +11.1%
classes.js — Interp: 🔴 12, 19 unch. · avg -1.3% · Bytecode: 🟢 11, 🔴 4, 16 unch. · avg +2.2%
closures.js — Interp: 🔴 7, 4 unch. · avg -2.0% · Bytecode: 🟢 7, 4 unch. · avg +3.7%
collections.js — Interp: 🟢 1, 🔴 7, 4 unch. · avg -2.1% · Bytecode: 🟢 10, 🔴 2 · avg +11.6%
destructuring.js — Interp: 🔴 19, 3 unch. · avg -3.4% · Bytecode: 🟢 17, 5 unch. · avg +6.4%
fibonacci.js — Interp: 🟢 1, 🔴 5, 2 unch. · avg -2.2% · Bytecode: 🟢 8 · avg +5.9%
for-of.js — Interp: 🔴 3, 4 unch. · avg -1.4% · Bytecode: 🟢 5, 2 unch. · avg +4.2%
helpers/bench-module.js — Interp: 0 · Bytecode: 0
iterators.js — Interp: 🔴 11, 9 unch. · avg -2.6% · Bytecode: 🟢 16, 🔴 1, 3 unch. · avg +6.1%
json.js — Interp: 🔴 5, 15 unch. · avg -1.2% · Bytecode: 🟢 20 · avg +10.2%
jsx.jsx — Interp: 🟢 5, 🔴 1, 15 unch. · avg +0.5% · Bytecode: 🟢 19, 2 unch. · avg +5.8%
modules.js — Interp: 🔴 6, 3 unch. · avg -2.4% · Bytecode: 🟢 2, 🔴 3, 4 unch. · avg +0.7%
numbers.js — Interp: 🟢 1, 🔴 6, 4 unch. · avg -2.6% · Bytecode: 🟢 10, 🔴 1 · avg +8.8%
objects.js — Interp: 🟢 1, 🔴 1, 5 unch. · avg -0.7% · Bytecode: 🟢 6, 1 unch. · avg +4.5%
promises.js — Interp: 🔴 9, 3 unch. · avg -3.2% · Bytecode: 🟢 11, 1 unch. · avg +6.3%
regexp.js — Interp: 🟢 1, 🔴 3, 7 unch. · avg -1.3% · Bytecode: 🟢 8, 🔴 1, 2 unch. · avg +3.5%
strings.js — Interp: 🔴 8, 3 unch. · avg -3.6% · Bytecode: 🟢 10, 1 unch. · avg +9.5%
typed-arrays.js — Interp: 🔴 13, 9 unch. · avg -2.3% · Bytecode: 🟢 11, 🔴 6, 5 unch. · avg +1.6%
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 Timing
Measured on ubuntu-latest x64. |
![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 the current code and only fix it if needed.
Inline comments:
In `@units/Goccia.Builtins.GlobalRegExp.pas`:
- Around line 444-447: The code treats sticky (PROP_STICKY) as enabling repeated
matchAll iteration, but per spec only global (PROP_GLOBAL) should; change the
determination of IsGlobal used when constructing the iterator (see RegexClone,
GetRegExpBooleanProperty, PROP_GLOBAL, PROP_STICKY and
TGocciaRegExpMatchAllIteratorValue.Create) so it only checks PROP_GLOBAL (remove
or ignore PROP_STICKY here) and leave sticky handling to the regex matching
logic elsewhere.
In `@units/Goccia.Values.Iterator.RegExp.pas`:
- Around line 40-49: The constructor TGocciaRegExpMatchAllIteratorValue
currently resets FSearchIndex := 0 and thus discards the cloned regex's
preserved lastIndex; instead, when the iterator was created from a cloned RegExp
for global or sticky patterns (FGlobal or sticky flag on the cloned object),
initialize FSearchIndex from the cloned RegExp's lastIndex property (the value
preserved by CloneRegExpObject / the ARegExp passed into the constructor) rather
than unconditionally zero—leave the zero initialization only for
non-global/non-sticky regexes; update TGocciaRegExpMatchAllIteratorValue.Create
to read the cloned object's lastIndex and assign it to FSearchIndex when
appropriate.
🪄 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: defaults
Review profile: CHILL
Plan: Pro
Run ID: 396e3dfe-0824-438f-82d5-c73eca83887c
📒 Files selected for processing (7)
docs/built-ins.mddocs/language-restrictions.mdtests/built-ins/RegExp/prototype/symbol-matchAll.jstests/built-ins/String/prototype/matchAll.jsunits/Goccia.Builtins.GlobalRegExp.pasunits/Goccia.Values.Iterator.RegExp.pasunits/Goccia.Values.StringObjectValue.pas
💤 Files with no reviewable changes (1)
- docs/language-restrictions.md
…onour cloned lastIndex Per ES2026 §22.2.6.9, the internal `global` boolean in CreateRegExpStringIterator is derived from flags containing "g" — sticky (y) affects match positioning but does not trigger repeated iteration. Also initialise FSearchIndex from the cloned regex's lastIndex so that /a/g with lastIndex=2 starts matching at index 2 as the spec requires. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
frostney
added
enhancement
new feature
4 tasks
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
TGocciaRegExpMatchAllIteratorValuethat computes one match per.next()call, aligning with ES2026 §22.2.6.9 spec behavior.RegExp.prototype[Symbol.matchAll]and theString.prototype.matchAllfallback path now return the lazy iterator.Goccia.Values.Iterator.RegExp.pasfollows the established lazy iterator pattern (same asTGocciaLazyMapIteratorValueetc.), with proper GCMarkReferences,ToStringTag, and support for both global and non-global modes.Closes #175
Testing
🤖 Generated with Claude Code
Summary by CodeRabbit
Documentation
String.prototype.matchAll()returns a lazy iterator computing matches on demand per ES2026 specification.Tests
for..ofloops.