Arnaud Quirin comments

Results 6 comments of


                                            Arnaud Quirin

Memory leaks when uncompressing multi-volume archives

Checking this a bit more, and it seems to me that the issues might reside here: https://github.com/miurahr/py7zr/blob/6b253d199c35ff3d846d4ef4a073f271b917b267/py7zr/py7zr.py#L588 Filling the dict in the loop inside the `_extract` function might prevent to...

refactor: drop read/readall method, and add feature to stream data into user object

Does this also solves the memory error described here? - https://github.com/miurahr/py7zr/issues/579#issue-2190900415 - https://github.com/miurahr/py7zr/issues/575#issue-2165043877 I have tried the code in the first link, but I still have a memory error. >...

refactor: drop read/readall method, and add feature to stream data into user object

@miurahr I have tested your updated PR using your new streaming technique, but unfortunately I am still reporting a huge memory consumption (2Gb) for multipart 7z files containing a lot...

refactor: drop read/readall method, and add feature to stream data into user object

> What your `Py7zIO` class? `from py7zr.io import WriterFactory, Py7zIO` > Could you share Python profiling result? https://docs.python.org/3/library/profile.html Attached. [test_multi_stats.zip](https://github.com/user-attachments/files/17403629/test_multi_stats.zip) I do not have `tox`, let me know if the...

refactor: drop read/readall method, and add feature to stream data into user object

Ok thanks for the insights. But is it possible to add an API parameter to set the memory limit? Relying on an external context manager looks convoluted to me especially...

refactor: drop read/readall method, and add feature to stream data into user object

> It is nonsense. > Your code does not return an object of Py7zIO **interface**. Sorry, it is not non-sense, it was just a leftover from your code sample that...