Fix v-flag bugs #85
Merged
Fix v-flag bugs #85
Conversation
This file contains 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
JLHwung
commented
Sep 19, 2023
tests/fixtures/unicode-set.js
Outdated
| @@ -105,6 +105,8 @@ const unicodeSetFixtures = [ | |||
| }, | |||
| { | |||
| pattern: '[^[a-z][f-h]]', | |||
| matches: ["A", "\u{12345}"], | |||
| nonMatches: ["a", "z"], | |||
| expected: '(?:(?![a-z])[\\s\\S])', | |||
There was a problem hiding this comment.
The current transpiled result does not match "\u{12345}".
JLHwung
commented
Sep 20, 2023
| ); | ||
| const negativeSet = UNICODE_SET.clone().remove(singleChars); | ||
| const bmpOnly = regenerateContainsAstral(negativeSet); | ||
| update(characterClassItem, negativeSet.toString({ bmpOnly: bmpOnly })); |
There was a problem hiding this comment.
If the regenerate set spans from code points before surrogate to astral sets, toString({ bmpOnly: false }) returns much more verbose results while toString({ bmpOnly: false }) is already correct: I think it should be fixed in regenerate later.
const regenerate = require('regenerate');
const set = regenerate().addRange(0xd000, 0x10000);
console.log(set.toString());
// [\uD000-\uD7FF\uE000-\uFFFF]|\uD800\uDC00|[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?:[^\uD800-\uDBFF]|^)[\uDC00-\uDFFF]
console.log(set.toString({ bmpOnly: true }));
// [\uD000-\uFFFF]|\uD800\uDC00The latter is apparently correct as it matches lone surrogates as well as U+10000. The former seems like [\uD000-\uFFFF]|\uD800\uDC00 is passed to the bmp pass again.
JLHwung
commented
Sep 20, 2023
| expected: '(?:[\\0-JL-\\uD7FF\\uE000-\\uFFFF]|[\\uD800-\\uDBFF][\\uDC00-\\uDFFF]|[\\uD800-\\uDBFF](?![\\uDC00-\\uDFFF])|(?:[^\\uD800-\\uDBFF]|^)[\\uDC00-\\uDFFF])', | ||
| matches: ["k", "\u212a", "\u{12345}", "\uDAAA", "\uDDDD"], | ||
| nonMatches: ["K"], | ||
| expected: '(?:[\\0-JL-\\uFFFF]|[\\uD800-\\uDBFF][\\uDC00-\\uDFFF])', |
There was a problem hiding this comment.
They are now much shorter and easier to reason about. I also added matches tests so that we are confident that transpiled result is correct.
mathiasbynens
approved these changes
Sep 20, 2023
JLHwung
force-pushed
the
fix-v-flag-bugs
branch
from
23feaf3 to
8501e5d
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
In this PR we reuse the unicode fixtures for the v-flag tests, based the observation that
/.../uand/.../vshould yield the same result unless set/string properties features are involved.We also introduce the
matchesandnonMatchesproperties to the v-flag fixture runner: They includes the strings that the transpiled regex is supposed to match / reject. It is useful when the transpiled regex is too verbose for proper comprehension.This PR includes commits from #84, I will rebase once that PR is merged.This is a draft PR as I still haven't figured out how to avoid double-bmpify regex strings: In the negative set notation we extract single code points from theUNICODE_SET, which yields surrogate stuffs in the output, but then it was bmp-ified again in theregenerate, yielding longer than necessary results, though it seems correct.