Always consider metadata as utf-8. #22
Merged
Conversation
This file contains 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
This is documented near the beginning of https://packaging.python.org/en/latest/specifications/core-metadata/ and warehouse currently serves these as `content-type: application/octet-stream` with no encoding, so requests spends a surprising amount of time on what are generally lower-ascii files trying to guess an encoding.
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #22 +/- ##
==========================================
+ Coverage 94.80% 95.80% +1.00%
==========================================
Files 10 10
Lines 616 620 +4
Branches 112 112
==========================================
+ Hits 584 594 +10
+ Misses 30 22 -8
- Partials 2 4 +2 ☔ View full report in Codecov by Sentry. |
jwodder
requested changes
Feb 17, 2024
src/pypi_simple/client.py
Outdated
| @@ -383,7 +383,7 @@ def get_package_metadata( | |||
| r.raise_for_status() | |||
| digester.update(r.content) | |||
| digester.finalize() | |||
| return r.text | |||
| return r.content.decode("utf-8") | |||
There was a problem hiding this comment.
Suggested change
| return r.content.decode("utf-8") | |
| return r.content.decode("utf-8", "surrogateescape") |
Because we don't want to crash on the inevitable case where someone somehow created a wheel with non-UTF-8 metadata.
thatch
force-pushed
the
fetch-metadata-as-bytes
branch
from
23fa4e8 to
4fdb8f6
Compare
|
I moved the bulk of the code to a bytes api, because that's really what I was after. Added surrogateescape as requested, and there are tests. |
|
This is now released in pypi-simple 1.5.0. Thank you for the contribution! |
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.
This is documented near the beginning of
https://packaging.python.org/en/latest/specifications/core-metadata/ and warehouse currently serves these as
content-type: application/octet-streamwith no encoding, so requests spends a surprising amount of time on what are generally lower-ascii files trying to guess an encoding.I think this is 100% safe -- files that can't be understood as utf-8 appear to not work in pip, which reads them as bytes, saves them as bytes and importlib always uses utf-8. A better API for pypi-simple might be to return bytes, but that's more disruptive.
Sample debug log pydantic-2.6.1-py3-none-any.whl.metadata showing more than 300ms spent trying to load the thing, only to give up and treat it as utf-8 anyway.
You might not notice this in a normal application, but I'm fetching these in a loop, with cachecontrol as the session, with a hot cache.
Here are traces for this change (the green block is
get_package_metadataand the total time is visible top-right):Before:
After: