Skip to content

Conversation

algomaster99
Copy link
Collaborator

During inference, the errors are often duplicated in the same group ID. This will ensure that benchmark has unique group IDs and which roughly implies unique source repositories.

@algomaster99 algomaster99 requested a review from wbxyz September 3, 2025 23:18
@algomaster99
Copy link
Collaborator Author

algomaster99 commented Sep 3, 2025

@wbxyz should we do this in the SQL query instead? Although the tables don't seem to have individual group and artifact columns.

@algomaster99
Copy link
Collaborator Author

I added another benchmark mavenTop500UniqueGroups. I am not sure what is the ratio of Gradle, Maven, and other build tools. But running this should give us distinct categories of failures that we should prioritize.

@algomaster99 algomaster99 requested a review from msuozzo September 4, 2025 15:46
@algomaster99 algomaster99 changed the title Group by GroupID for maven packages Add benchmark for grouping by GroupID for maven packages Sep 4, 2025
}
close(pkgs)
}()
seenGroups := make(map[string]struct{})
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If this were a map[string]bool then you can assign true as the value and just do something like if seenGroups[g] { ...

if len(psp.Versions) >= 5 {
continue
}
nameParts := strings.SplitN(p.Package, ":", 2)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can this be combined with the group split above?

log.Fatal(err.Error())
}
query := client.Query(`
SELECT
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this SQL needs to be updated following some of the other PRs

@algomaster99
Copy link
Collaborator Author

Achieved by #772.

@algomaster99 algomaster99 deleted the push-yzsxuuppsrtq branch September 19, 2025 19:49
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants