Repository navigation
test(latex): let the arXiv corpus test re-check failures and read chosen shards - #4969
Merged
PeterStaar-IBM merged 2 commits intoOct 22, 2026
Conversation
The arXiv corpus test skipped every file in the manifest and counted the failure stages of earlier runs. A file that failed once was never converted again, so the test kept failing after the fix until the manifest was deleted. Add --recheck-failures to convert those files again. The new manifest entry replaces the earlier one, and the failure count now comes from the latest entry of each file. Signed-off-by: Daniel Nguyen <danielnguyenh07@gmail.com>
The corpus test always streamed the latex config from shard 0, which holds the arXiv papers of 2007 and 2008, and it wrote its manifest, report, and saved files to the repository root. Add --corpus-shard N (0-79, repeatable) to stream only those Parquet files, and --corpus-dir DIR to write the output files there. Several runs can now cover different shards in parallel, each in its own folder. Signed-off-by: Daniel Nguyen <danielnguyenh07@gmail.com>
Contributor
|
✅ DCO Check Passed Thanks @DanielNg0729, all your commits are properly signed off. 🎉 |
Contributor
Merge Protections🟢 Merge protection satisfied — ready to merge. Show 1 satisfied protection🟢 Enforce conventional commitMake sure that we follow https://www.conventionalcommits.org/en/v1.0.0/
|
github-actions
Bot
added
the
ai:ci-needs-care
Contributor
AI triage (advisory)Checked commit Warning The AI analysis did not finish. Only the deterministic checks are shown. CI safety: needs maintainer care before you approve CI.
Possibly related pull requests
The |
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
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
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.
Hi maintainers, this is a small follow-up to the arXiv corpus test from #4817, so it can check a fix and go past the 2007-2008 papers. Thank you!
Summary
This PR makes these changes:
--recheck-failuresconverts again the files that failed in an earlier run. The new manifest entry replaces the old one, and the failure count now comes from the latest entry of each file. Before, a failed file was skipped on every rerun and its old failure still counted, so the test kept failing after the fix until the manifest was deleted.--corpus-shard N(0-79, repeatable) streams only those Parquet files of thelatexconfig. Without it the stream always starts at shard 0, the 2007-2008 papers.--corpus-dir DIRwrites the manifest, report, and saved files to that folder instead of the repository root, so several runs can cover different shards in parallel.Checklist: