@scottjones03 Would you be willing to add a that exercises the issue you found with saving pdfs? If I understood what you said in #23276 correctly, making a figure with a sub-figure and then saving as a pdf to a io.StringIO buffer would be enough?

@scottjones03 Would you be willing to add a that exercises the issue you found with saving pdfs? If I understood what you said in #23276 correctly, making a figure with a sub-figure and then saving as a pdf to a io.StringIO buffer would be enough?

Sorry, slight confused, Do you want me to summarise reproducibility steps for the bug so it can be QAd?

Sorry, slight confused, Do you want me to summarise reproducibility steps for the bug so it can be QAd?

Yes. I think this needs a test (probably in https://github.com/matplotlib/matplotlib/blob/main/lib/matplotlib/tests/test_figure.py) that looks something like

from io import StringIO
def test_subfigure_pdf():
    fig = ...  # minimal code to make a broken figure
    buffer = StringIO()
    fig.savefig(buffer, format='pdf')

so that this bug can not come back.

I also changed the title to something I think is a bit more accurate as to what is being done here. We (for better or worse) use "deprecate" as a term of art for the process by which we remove public API from the library, where as here we are "just" not using a part of the public API internally. A deprecation will get a lot more scrutiny than an internal refactor so the adjusted title will hopefully put reviewers in the right state of mind when they click on the PR.

Sorry, slight confused, Do you want me to summarise reproducibility steps for the bug so it can be QAd?

Yes. I think this needs a test (probably in https://github.com/matplotlib/matplotlib/blob/main/lib/matplotlib/tests/test_figure.py) that looks something like

from io import StringIO
def test_subfigure_pdf():
    fig = ...  # minimal code to make a broken figure
    buffer = StringIO()
    fig.savefig(buffer, format='pdf')

so that this bug can not come back.

Done

I'm not merging to give @QuLogic a chance to check (based on the review request, although I think it is OK).

@scottjones03 it is possible to do "Add suggestion to batch" and then do a single commit. Not a problem since we should squash the commits before merging, but often a bit easier.

Owee, I'm MrMeeseeks, Look at me.

There seem to be a conflict, please backport manually. Here are approximate instructions:

1. Checkout backport branch and update it.

git checkout v3.5.x
git pull

2. Cherry pick the first parent branch of the this PR on top of the older branch:

git cherry-pick -x -m1 047254dd0c5fcf602a19b245e3b14a38d1af1614

3. You will likely have some merge/cherry-pick conflict here, fix them and commit:

git commit -am 'Backport PR #23278: Remove internal use of get/set dpi'

4. Push to a named branch:

git push YOURFORK v3.5.x:auto-backport-of-pr-23278-on-v3.5.x

5. Create a PR against branch v3.5.x, I would have named this PR:

"Backport PR #23278 on branch v3.5.x (Remove internal use of get/set dpi)"

And apply the correct labels and milestones.

Congratulations — you did some good work! Hopefully your backport PR will be tested by the continuous integration and merged soon!

Remember to remove the Still Needs Manual Backport label once the PR gets merged.

If these instructions are inaccurate, feel free to suggest an improvement.

drat, I forgot to squash this (realized about a second after I clicked merge) 🤦🏻

backport is a bit gnarly.