Fix Cython 3.0 regression with time_loc_dups #55915
Merged
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Outdated suggestions cannot be applied.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.
re #55179
From discussion in cython/cython#1807 (comment) it looks like Cython prior to 3.0 would always use the sequence protocol for indexing with an integral value. However, Python prefers the object protocol first if available, and Cython switched to match that logic with 3.0
NumPy arrays implement both the sequence and the mapping protocol. In cases where we have untyped arrays that fall back to Python calls we will see a performance regression since this will now route through the mapping space
The changes in this PR are not meant to be an exhaustive review of the codebase, rather just a quick POC to reset the time_loc_dups benchmark