BUG: Fix incorrect tm_matrix in call to visitor_text

[troethe]

Supply the old tm_matrix when flushing out text to the visitor_text
in crlf_space_check. The new one might already be changed and
unrelated to the current text.

Also add a test for the tm_matrix and cm_matrix that are given to
visitor_text when extracting text.
The test computes the coordinates of three letters in different
parts of a test page based on the matrices and checks, if they are
roughly where they should be.

Fixes #2059

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.10% 🎉

Comparison is base (c04a6bb) 94.17% compared to head (b7544be) 94.28%.
Report is 16 commits behind head on main.

❗ Current head b7544be differs from pull request most recent head 2e42262. Consider uploading reports for the commit 2e42262 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2060      +/-   ##
==========================================
+ Coverage   94.17%   94.28%   +0.10%     
==========================================
  Files          41       41              
  Lines        7333     7365      +32     
  Branches     1442     1451       +9     
==========================================
+ Hits         6906     6944      +38     
+ Misses        266      262       -4     
+ Partials      161      159       -2     

Files Changed	Coverage Δ
pypdf/_page.py	93.65% <100.00%> (+0.04%)	⬆️
pypdf/_text_extraction/__init__.py	93.27% <100.00%> (+0.05%)	⬆️

... and 3 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[pubpub-zz]

@troethe
can you complete your PR with a test (reusing your pdf is a good idea) in order to ensure the issue will not come back.

[troethe]

@pubpub-zz : All right, I added a new test on a simplified test document where the correct positioning is a bit more apparent. Let me know if you have any feedback.

Edit: Also note that my original PR was wrong, because I passed the tm_prev matrix as tm_matrix to the visitor, even though tm_prev wasn't actually the tm_matrix but tm_matrix * cm_matrix. I fixed this and updated the PR.

[MartinThoma]

@troethe From my side we are good to go :-) Please adjust the top comment of the PR in such a way that it can be the commit message of the squash-merge. For example:

Supply the old tm_matrix when flushing out `text` to the `visitor_text`
in `crlf_space_check`.

Check the coordinates computed from the tm_matrix and cm_matrix
that are given to `visitor_text` when extracting text.
The test computes the coordinates of three letters in different
parts of a test page based on the matrices given to the `visitor_text`
function and checks, if they are roughly where they should be.

Fixes #2059

One last question:

A preliminary fix to #2059

What do you mean by "preliminary"? Doesn't this fix the mentioned issue?

[MartinThoma]

@pubpub-zz I guess this PR is fine from your perspective?

[troethe]

What do you mean by "preliminary"? Doesn't this fix the mentioned issue?

Yes, it does, but when I originally submitted this, I didn't quite understand all of the code yet and I had not thoroughly tested it as shown by the fact that it ended up not being quite right. I guess I should have called it a WIP/draft instead.

[MartinThoma]

Ok, then please let me know when you think I can merge :-)

[troethe]

@MartinThoma Ok, I think it's good now!
Sorry for those last unnecessary pushes. I misread and thought you wanted me to update the top commit's message instead of the PR's top comment.

[MartinThoma]

No problem :-)

Thank you for the contribution 🙏 I will make a release that contains it today

[MartinThoma]

If you want, I can add you to https://pypdf.readthedocs.io/en/latest/meta/CONTRIBUTORS.html :-)

[troethe]

If you want, I can add you to https://pypdf.readthedocs.io/en/latest/meta/CONTRIBUTORS.html :-)

Sure, why not. :) Thank you!