Add support for mistune 3.1.0

[fcollonval]

Fixes #2198

[fcollonval]

Unsure about the mypy error, should I just ignore them?

nbconvert/filters/markdown_mistune.py:106: error: Cannot override class variable (previously declared on base class "BlockParser") with instance variable  [misc]
nbconvert/filters/markdown_mistune.py:106: note: Error code "misc" not covered by "type: ignore" comment
nbconvert/filters/markdown_mistune.py:148: error: Cannot override class variable (previously declared on base class "InlineParser") with instance variable  [misc]
nbconvert/filters/markdown_mistune.py:486: error: Argument 4 to "__init__" of "Markdown" has incompatible type "Iterable[Callable[[Markdown], None]]"; expected "Optional[Iterable[Plugin]]"  [arg-type]

[abbasidaniyal]

Hey @fcollonval!
Thanks for raising this PR. This patch seems to fix the issue for me.

[bollwyvl]

mypy error

Yeah, we should fix them if we can.

106 and 148 vs upstream:

DEFAULT_RULES: ClassVar[Iterable[str]]  = 

For 486 needs a bit of a forensics to figure out why we have our own type there...

[bollwyvl]

Ah, the Plugin thing is just a protocol. I guess we could backfill that:

    try:
        from mistune.plugins import Plugin
    except ImportError:
        class Plugin(Protocol):
            def __call__(self, markdown: "Markdown") -> None: ...

But really much of it could be hidden behind TYPE_CHECKING...

[fcollonval]

Thanks a lot for the hints @bollwyvl

[bollwyvl]

So a quick glance a the CI doesn't tell me whether we're getting coverage against older versions of mistune.

For the offending link:

http://damianavila.github.io/blog/posts/one-line-deployment-of-your-site-to-gh-pages.html
# changed to
https://damianavila.github.io/blog/posts/one-line-deployment-of-your-site-to-gh-pages

[bollwyvl]

Here's a thing from nbformat:

  test_minimum_versions:
    name: Test Minimum Versions
    timeout-minutes: 20
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - name: Base Setup
        uses: jupyterlab/maintainer-tools/.github/actions/base-setup@v1
        with:
          dependency_type: minimum
      - name: Run the unit tests
        run: |
          hatch run test:nowarn || hatch run test:nowarn --lf

We should probably get that in there and and least see what we can see.

[krassowski]

It looks like test_minimum_versions is already here? I can see it running: https://github.com/jupyter/nbconvert/actions/runs/12580785345/job/35063360841?pr=2199

nbconvert/.github/workflows/tests.yml

Lines 104 to 127 in 6e5fdb3

	test_minimum_versions:
	name: Test Minimum Versions
	timeout-minutes: 20
	runs-on: ubuntu-latest
	steps:
	- uses: actions/checkout@v4
	- name: Base Setup
	uses: jupyterlab/maintainer-tools/.github/actions/base-setup@v1
	with:
	dependency_type: minimum
	- name: Install dependencies
	run: |
	sudo apt-get update
	sudo apt-get install texlive-plain-generic inkscape texlive-xetex latexmk
	sudo apt-get install xvfb x11-utils libxkbcommon-x11-0 libxcb-xinerama0 python3-pyqt5
	# pandoc is not up to date in the ubuntu repos, so we install directly
	wget https://github.com/jgm/pandoc/releases/download/2.9.2.1/pandoc-2.9.2.1-1-amd64.deb && sudo dpkg -i pandoc-2.9.2.1-1-amd64.deb
	- name: Run tests
	run: |
	# Attempt to work around https://github.com/pypa/pip/issues/12781
	PIP_CONSTRAINT= hatch env run -e test -- pip install 'pip>=24.2'
	xvfb-run --auto-servernum hatch run test:nowarn || xvfb-run --auto-servernum hatch run test:nowarn --lf

[krassowski]

Looks good to me:

[fcollonval]

Thanks @krassowski for pushing this. I arrived at the same conclusion. So I pushed a new commit to add a test for mistune 3.0 ... hopefully.

[krassowski]

Thank you @fcollonval!

[bollwyvl]

Huzzah! Happy New Year!

[Heitapval]

I installed nbconvert from pypi.

"jupyter nbconvert --to html jarvi.ipynb" gives the error

MathBlockParser' object has no attribute 'parse_axt_heading'. Did you mean: 'parse_atx_heading

[krassowski]

@Heitapval and which version jupyter nbconvert --version shows?

[Heitapval]

@Heitapval and which version jupyter nbconvert --version shows?

The version is 7.16.4

[krassowski]

The version is 7.16.4

The fix was released in nbconvert v7.16.5 as mentioned before in #2198 (comment) - you need to upgrade nbconvert or find out why pip installed an older version.

[Heitapval]

The version is 7.16.4

The fix was released in nbconvert v7.16.5 as mentioned before in #2198 (comment) - you need to upgrade nbconvert or find out why pip installed an older version.

Thanks a lot. I got version 7.16.6

I have created a virtual environment with --system-site-packages. I had installed jupyterlab including nbconvert also to my opensuse. I have upgraded with pip3.13 install --upgrade --no-cache-dir -r <req_file>. Usually pip has seemed to upgrade to the latest version when opensuse has an older version. Maybe there were some dependencies.

I uninstalled nbconvert from my opensuse (which uninstalled a lot of other packages.)

pip --force-reinstall might have worked as well