Conversation
|
If you execute |
|
Just a short heads-up: I will only be able to review this properly on 5th of October, as I'm on vacation without my notebook (just phone). |
|
@MartinThoma Black reformatted tests/test_page.py. Have a nice vacation! |
Codecov ReportBase: 94.10% // Head: 94.16% // Increases project coverage by
Additional details and impacted files@@ Coverage Diff @@
## main #1373 +/- ##
==========================================
+ Coverage 94.10% 94.16% +0.06%
==========================================
Files 28 28
Lines 5071 5074 +3
Branches 1052 1052
==========================================
+ Hits 4772 4778 +6
+ Misses 177 176 -1
+ Partials 122 120 -2
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
|
Wow, that was a nice catch and good fix ❤️ Thank you for adding a test - without it, it would have been super hard for me to see that this does the right 🙏 |
Bug Fixes (BUG): - td matrix (#1373) - Cope with cmap from #1322 (#1372) Robustness (ROB): - Cope with str returned from get_data in cmap (#1380) Full Changelog: 2.11.0...2.11.1
This PR contains a bugfix which introduces a matrix multiplication [1 0 0 1 tx ty] * tm_matrix when processing operator Td (see 5.3.1 Text-Positioning Operators, p. 406). This is important in cases where the text matrix is not like [1 0 0 1 e f].
I added an ASCII-sample pdf file which contains such a text matrix:
The test file ST.2019.pdf of #1142 contains such text matrices:
Therefore I adjusted test iss_1142.
After generating a svg export of Sample_Td-matrix.pdf I found a problem at the visitor-calls of #1252: One may get the wrong text matrix when an operator like Tm is executed together with not yet written text-output. I didn't want to mix that problem with the Td-bug therefore I created this PR first of all.