railties/test_parser: Cache prism work done in definition_for #53936
+18
−15
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
jsharpify
force-pushed
the
jsharpify/prism-parsing
branch
from
9861e49 to
1d95d38
Compare
nvasilevski
approved these changes
Dec 13, 2024
There was a problem hiding this comment.
Makes sense!
definition_for is called repeatedly and performs a lot of work that really adds up.
Occasionally there would be an opportunity to avoid calling definition_for repeatedly in the first place, by either changing the code or moving memoization at a higher level. However this doesn't seem to be the case, I briefly looked over the code and it does look like definition_for is expected to be called multiple times so memoizing parse results makes perfect sense
jsharpify
force-pushed
the
jsharpify/prism-parsing
branch
from
1d95d38 to
fa5e5b5
Compare
`definition_for` is used to extract the line numbers of a test, which is relevant when filtering tests to a line number or range, eg my_test.rb:1-2. The filter itself operates by, for each test, comparing the lines the test spans against the filter. In the current code, this will call `Prism.parse_file` and then walk the parse result until the matching definition/call is found. For files with a large number of tests, this is fairly onerous and can dominate the test runtime. Fortunately, we can parse the file once and cache the starts and ends of every def/call node. This beings the `definition_for` runtime down from seconds to low milliseconds. We also see a reduction in gc time that is probably attributable to not repeatedly parsing the file. For a test run in an 240 test file that went from 11s of parsing to 20ms, gc went from 4.5-2.2 seconds. Finally, for any user that has written a reporter or otherwise hooked into the filters, this should provide speedup as well. Co-authored-by: Stath Stivaros <stath.stivaros@shopify.com>
jsharpify
force-pushed
the
jsharpify/prism-parsing
branch
from
fa5e5b5 to
fbb37c0
Compare
rafaelfranca
added a commit
that referenced
this pull request
Dec 13, 2024
railties/test_parser: Cache prism work done in definition_for
rafaelfranca
added a commit
that referenced
this pull request
Dec 13, 2024
railties/test_parser: Cache prism work done in definition_for
|
J |
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 PR applies to the Prism implementation of
definition_for.Motivation
In files with lots of tests,
definition_foris called repeatedly and performs a lot of work that really adds up. The work is reusable across many tests, so let's do it.Details
definition_foris used to extract the line numbers of a test, which is relevant when filtering tests to a line number or range, eg my_test.rb:1-2.The filter itself operates by, for each test, comparing the lines the test spans against the filter. In the existing code, this will call
Prism.parse_fileand then walk the parse result until the matching definition/call is found.For files with a large number of tests, this is fairly onerous and can dominate the test runtime. Fortunately, we can parse the file once and cache the starts and ends of every def/call node.
This brings the
definition_forruntime down from seconds to low milliseconds.We also see a reduction in gc time that is probably attributable to not repeatedly parsing the file.
For a test run in an 240 test file that went from 11s of
definition_forto 20ms, gc went from 4.5 to 2.2 seconds.Finally, for any user that has written a test reporter or otherwise hooked into the filters, this should provide speedup as well.
Checklist
[Fix #issue-number]