Fix 200% untar progress issue by avoiding a second read of file #17708
+42
−23
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
AbrilRBS
reviewed
Feb 5, 2025
memsharded
reviewed
Feb 5, 2025
There was a problem hiding this comment.
Please share real timings of unzipping some large files, with and without this change.
UPDATE: Sorry I didn't see the OP clearly, now it is clear, thanks!
Then @AbrilRBS your comment
Looks great! Seems like performance-wise, both methods will end up calling extract_one in the tarfile for each member, so there would be no extra overhead there, we'll just need to address my comment
Not fully clear, what do you mean?
memsharded
reviewed
Feb 5, 2025
Comment on lines
-367
to
-369
| if pattern: | ||
| members = list(filter(lambda m: fnmatch(m.name, pattern), | ||
| tarredgzippedFile.getmembers())) |
There was a problem hiding this comment.
Doesn't this case read a third time the file, potentially causing 300%?
If the getmembers() is doing a file read IO, this seems the case?
There was a problem hiding this comment.
Nop, getmembers have a static variable / cache mechanism.
The maximum reads that tarfile could perform on a targz are two!
See code:
def getmembers(self):
"""Return the members of the archive as a list of TarInfo objects. The
list has the same order as the members in the archive.
"""
self._check()
if not self._loaded: # if we want to obtain a list of
self._load() # all members, we first have to
# scan the whole archive.
return self.members
AbrilRBS
approved these changes
Feb 5, 2025
|
I did not check the code, but if I understand correctly, the uncompress status is updated every second. I just tried this branch and indeed it works until 99%, but Qt is huge, then, each step is repeated, but it's fine IMO: |
Yes... I think I can add a final log which explicitly marks the 100% just for the TOC
That is a very good point, the progress printer is managed by a |
uilianries
approved these changes
Feb 5, 2025
There was a problem hiding this comment.
@perseoGI thank you for clarifying the feature. It looks to me now.
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.
Changelog: Fix: Improve untar performance
Docs: omit
Calling
TarFile.getmembers()forces a complete IO read of the tar file.The same happens when calling
TarFile.extractall(members=members)for the first time.This PR aims to fix the reported issue of having untar progress exceeding 100% occasionated by calling
seek(0)twice, one in the first read querying all files and a second for the actual uncompression.Details:
Perform the decompression in a single IO loop by iterating over compressed files without actually reading them, filtering and uncompressing when needed one at a time.
Benchmarks:
develop2 untar code
New changes
The results show a 25% impact on uncompression efficiency
Closes #17697
developbranch, documenting this one.