Skip cached releases with proven dependency conflicts - #21839
Draft
charliermarsh wants to merge 1 commit into
Draft
charliermarsh wants to merge 1 commit into
charliermarsh wants to merge 1 commit into
Conversation
uv test inventory changesThis PR changes the tests when compared with the
|
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
When consecutive registry releases depend on a range ruled out by a fixed requirement, we currently select and reject each release separately, even when we have already fetched its metadata. Look ahead at that metadata after repeated conflicts and teach PubGrub about additional releases whose declared dependencies are proven incompatible. Follow up to two learned dependency edges only when their incompatibilities cover the entire intermediate range; otherwise use the existing search. The lookahead doesn't fetch more metadata and skips packages with potentially different sources.
For example, if the project pins
A==1and several releases ofBeach requireA>=2, we can exclude thoseBreleases together. A release that also allowsA==1still goes through normal selection.Across 141 successful public-project resolutions, the outputs were byte-for-byte identical; four tried fewer distinct package/version pairs and none tried more. Distinct package counts did not change:
docextra)Two independent warm runs of the checked-in Home Assistant fixture measured median solve times of 104.2 → 100.9 ms and 102.4 → 97.7 ms. I did not establish a repeatable timing improvement on the other projects.