Repository navigation
fix(publish): bound resource reader allocations - #3
Merged
Merged
Conversation
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Resource publication read every object with the full runtime size limit. On Windows, repeatedly allocating a 128 MiB read buffer for small blobs made a roughly 59,000-object validation take about an hour.
Read in bounded 1 MiB chunks until EOF or the first excess byte. The maximum-size check, path checks, content hashing, and decoding remain in place. A local 1,058,591-byte sample returned identical bytes and hashes; 25 complete reader calls took 0.055 seconds instead of 1.167 seconds. This is a sample measurement, not a full-catalogue benchmark.
Validation: 39 tests passed in
test_resource_publication.pyandtest_verified_representations.py, including empty, exact-limit, over-limit, multi-chunk, and short-read cases.