Improve doc build performance #44
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
duposyl
changed the title
jorgepiloto
force-pushed
the
CI/doc-build-performance
branch
from
6d1fd60 to
496c2cc
Compare
- Removing matrix build using multiple python versions. - Commenting out numpydoc checks. - Restoring tox configuration to initial state.
duposyl
changed the title
jorgepiloto
approved these changes
Nov 30, 2022
Member
|
Thanks a lot for such a detailed description, @duposyl. Let us keep track of that issue. Merging 🚀 |
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Outdated suggestions cannot be applied.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.
The documentation build has been taking hours on GitHub while it was taking minutes on my local machine. The issue was traced back to the version of Python. Building the doc using Python 3.8 (which is what I had locally) is fast. Building it with Python 3.9 or 3.10 is very slow. This was verified by running the doc build on GitHub using these different versions:
After further investigation, the issue was traced to the numpydoc docstring validation:
The
numpydoc_validation_checksoption in the Sphinx configuration is what causes the slowdown. The call toinspect.getsourcelinesat https://github.com/numpy/numpydoc/blob/main/numpydoc/validate.py#L213 gets much slower starting with Python >= 3.9, and is what is causing the bad doc build performance.So, in order to have decent performance during CI, as a workaround, we decided to turn off the numpydoc validation for the time being. A separate issue (#47) has been created to revisit this in the future when we are ready to dig into the numpydoc warnings.
Note that this change also turns on the parallel option for Sphinx, as this helps independently with performance a little bit too.