Add text-based fastpaths into LS token matching #41924
Merged
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
typescript-bot
added
Author: Team
For Uncommitted Bug
weswigham
commented
Dec 10, 2020
| @@ -1391,7 +1391,23 @@ namespace ts { | |||
| return isInsideJsxElementTraversal(getTokenAtPosition(sourceFile, position)); | |||
| } | |||
|
|
|||
| export function findPrecedingMatchingToken(token: Node, matchingTokenKind: SyntaxKind, sourceFile: SourceFile) { | |||
| export function findPrecedingMatchingToken(token: Node, matchingTokenKind: SyntaxKind.OpenBraceToken | SyntaxKind.OpenParenToken | SyntaxKind.OpenBracketToken, sourceFile: SourceFile) { | |||
There was a problem hiding this comment.
The changes to this function were able to bring the test runtime from ~36s to ~10s on my machine. Just looking for the likely token position using lastIndexOf gives a great improvement,
weswigham
commented
Dec 10, 2020
Co-authored-by: Daniel Rosenwasser <DanielRosenwasser@users.noreply.github.com>
Co-authored-by: Daniel Rosenwasser <DanielRosenwasser@users.noreply.github.com>
Co-authored-by: Daniel Rosenwasser <DanielRosenwasser@users.noreply.github.com>
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Outdated suggestions cannot be applied.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.
This takes the runtime of the
excessivelyLargeArrayLiteralCompletions.tsfourslashtest down from ~36s on my machine (near the timeout limit) to 1s.A bit of history: The test in question was added recently, in #40953, as a perf-monitoring test. The initial fix did make the perf much better (going from "spin forever" to 11s is pretty good), however its perf has been creeping back up over the last two months, to the point where on @andrewbranch 's machine it started timing out. So I took a look again to see what was going on. The low-level operation that we spend all the time doing is matching tokens, and that's what I'd algorithmically optimized in #40953. In this PR, I took a look from the top down - the reason we're matching tokens, at all, is to determine if the location we're looking for completions at is a type argument position (which in turn filters the completion list) - to do so, we match all braces backwards from the cursor until we run into a
<which isn't contained within a set of matched braces. Unfortunately, in the test given, there are thousands of braces to match, so even though the token finding within the source file is pretty efficient(tm), we have a lot of work to do, still, since we essentially are forced to find almost every other token in the file. And that's where this PR comes in: in two phases, we use textual heuristics to quickly stop or speed up that token matching process.