Add support for pgf.documentclass rcParam #30063
Conversation
|
Thanks! It would be nice with a bit of documentation and some test for this. Docs: https://matplotlib.org/stable/users/explain/text/pgf.html#custom-preamble would be a place to mention it. And it can be worth having a release note that this is a new feature since more people can be interested in it. Test: I quickly browsed the available tests for some inspiration, but I think the best way, if possible, is to add a test where the resulting PGF-file is checked to contain the correct documentclass. |
lib/matplotlib/__init__.py
Outdated
| @@ -958,6 +958,7 @@ def rc_params_from_file(fname, fail_on_error=False, use_default_template=True): | |||
| # Strip leading comment. | |||
| transform=lambda line: line[1:] if line.startswith("#") else line, | |||
| fail_on_error=True) | |||
| rcParamsDefault["pgf.documentclass"] = "article" | |||
There was a problem hiding this comment.
The default value should go in https://github.com/matplotlib/matplotlib/blob/main/lib/matplotlib/mpl-data/matplotlibrc
| _DOCUMENTCLASS = r"\documentclass{article}" | ||
| _DOCUMENTCLASS = ( | ||
| rf"\documentclass{{{mpl.rcParams.get('pgf.documentclass', 'article')}}}" | ||
| ) |
There was a problem hiding this comment.
I think the reason that this is a variable is that it should be possible to override the document class. (Although in an undocumented way.)
Since it is now possible to do that in a more regular way, it would make sense to rewrite the code to not use the variable any more. It is used in three locations though, so maybe one should simply try to factor out the pre-amble code in some way. Which may lead to that your current solution can stay...
33e34c3 to
6d48325
Compare
Since AFM is now private, we can delete unused methods without deprecation. Additionally, add `AFM.get_glyph_name` so that the PostScript mathtext code doesn't need to special case AFM files.
Deprecate point_at_t and document that a BezierSegment can be called
* Remove apply_theta_transforms argument * Improve formatting * Rename xxxxxx-DS.rst to 30004-DS.rst * Delete extra line
Co-authored-by: hannah <story645@gmail.com>
Co-authored-by: hannah <story645@gmail.com>
…izer Update to docs with regards to colorbar and colorizer
…rties DOC: Document the properties of Normalize
DOC: Clarify that types in docstrings do not use formal type annotation syntax
I'm not sure why this was added originally, but it seems unnecessary. Fortunately, the waste was probably avoided by LTO.
…tplotlib#30130) When the subplot tool already exists, then it is shown, but never returned. This seems to be an accident.
CI: Skip jobs on forks
PR summary
This PR adds support for configuring the LaTeX
\documentclass{...}used by the PGF backend via a newrcParam:pgf.documentclass.Why is this change necessary?
Currently, the PGF backend uses a hardcoded LaTeX document class (
article) when rendering PGF figures. This causes layout inconsistencies (such as in #29978) when the.pgffile is included in documents using other classes likeIEEEtran,acmart, etc.What problem does it solve?
It allows users to specify the LaTeX document class used during PGF rendering via:
This provides a clean and documented alternative to the current workaround of modifying the internal
_DOCUMENTCLASSvariable inbackend_pgfshowed in #29978.What is the reasoning for this implementation?
The new rcParam follows Matplotlib’s configuration pattern and is registered via
rcsetup.py. It defaults to'article', ensuring backward compatibility. The PGF backend retrieves the class usingmpl.rcParams.get(...).Doubts:
matplotlibrcbecause I'm unsure of its exact placement or formatting style (or even if Ia should add it). I'd appreciate feedback on whether it should be included there explicitly.Please advise on both.
PR checklist