fix: Chunked Upload V2 not working on Nextcloud 30 and later

[DrZoidberg09]

Previously the chunk number started with 0, which leads to errors. Simple fix: starting chunk number with 1

Fixes: #319

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 94.68%. Comparing base (2cbf584) to head (77ebaa3).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #324   +/-   ##
=======================================
  Coverage   94.68%   94.68%           
=======================================
  Files          46       46           
  Lines        5364     5365    +1     
=======================================
+ Hits         5079     5080    +1     
  Misses        285      285           

Files with missing lines	Coverage Δ
nc_py_api/files/files.py	100.00% <100.00%> (ø)

[bigcat88]

sorry for such a big delay with the review, I have a small question - in which standard or where can I read that the chunk number should start with one?

and is it possible to get an example of code that does not work without this fix, and with it it starts to work(i would like to add it to test suite if possible)?

(I could not repeat the error from the linked issue unfortunately)

[DrZoidberg09]

Thanks for looking at the PR!

It is not perfectly documented. It is mentioned in the example here: https://docs.nextcloud.com/server/latest/developer_manual/client_apis/WebDAV/chunking.html

However, the real issue comes here:

Used your lib to upload with v2 to the dev instance that is a normal Nextcloud instance using local storage. However, when using it with the production instance that uses s3 storage backend, it gives a bad request in return, if you go over the 5 mb chunk size where this lib automatically uses v2. Also described in the linked issue.

When starting the chunk number with 1, everything works fine also in the s3 backend Nextcloud instance.

As in v2 the chunks are directly uploaded to s3 (which is a huge advantage), I assume that the issue is rather the passthrough of the chunks from upload directly s3 rather than the pure upload.

Hence, it would nice if you could merge this.

[bigcat88]

Your explanations are clear enough, thanks for the detailed description of why it won't fail in tests without this.