Remove Kdenlive adapter for split out

[jlskuz]

Summarize your change.

As part of the ongoing outlining process of contrib adapters, this removes the Kdenlive adapter from the core repository. As discussed with the Kdenlive team (of which @vpinon, the original author, is part), the code has been moved to the KDE GitLab https://invent.kde.org/multimedia/kdenlive-opentimelineio

Related #1378

See also discussion in #1390

NOTE: while removing the section about AAF from docs/tutorials/otio-plugins.md is theoretically not part of this PR, it is required to make the tests pass.

[meshula]

It's fantastic to see the adapter hosted at kde.org!

What's the nature of the test failure stemming from the AAF documentation block?

[jlskuz]

It is due to docs/tutorials/otio-plugins.md being auto-generated in sync with the code. The section about AAF was still in the doc, while the code as been (re)moved. See the first paragraph of docs/tutorials/otio-plugins.md:

This document is automatically generated by running the
autogen_plugin_documentation command, or by running make plugin-model. It
is part of the unit tests suite and should be updated whenever the schema
changes. If it needs to be updated, run: make doc-plugins-update and this
file should be regenerated.

This is the (shortened) output of the test:

======================================================================
FAIL: test_plugin_documentation (test_serialized_schema.PluginDocumentationTester)
Verify that the plugin manifest matches what is checked into the
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/julius/Programmierung/github/OpenTimelineIO/tests/test_serialized_schema.py", line 67, in test_plugin_documentation
    self.assertMultiLineEqual(
AssertionError: '# Pl[614 chars]\n- `opentimelineio/adapters/builtin_adapters.[17532 chars]\n\n' != '# Pl[614 chars]\n- `site-packages/otio_moviemaker_adapter/plu[2903 chars]\n\n'
[...]  
 : 
 The schema has changed and the autogenerated documentation in /home/julius/Programmierung/github/OpenTimelineIO/docs/tutorials/otio-plugins.md needs to be updated.  run: `make doc-plugins-update`

----------------------------------------------------------------------

[meshula]

I'm approving, but I'd like to get a response from more of the TSC since this is a structural change to the project.

[meshula]

Since the AAF plugin hasn't yet moved to its own repo, it would be good to get the plugin authors to have a look at the documentation issue, and perhaps resolve the AAF documentation bug as a separate PR.

[meshula]

Note this PR which moves the AAF adapter to its own repo:

#1348

[ssteinbach]

The AAF adapter is still in the repo, you shouldn't have to delete those docs. The CI shows errors where that doc is missing:
https://github.com/AcademySoftwareFoundation/OpenTimelineIO/actions/runs/3164460045/jobs/5152761379#step:9:360

So I'm wondering if maybe some stale local state led to needing to delete it locally? They definitely should be there still. I think everything else about this PR is good to go.

[markreidvfx]

@jlskuz I can replicate the docs behavior if I mess up my local install of pyaaf2. Does python -c "import aaf2" report a error?

[codecov-commenter]

Codecov Report

Merging #1438 (24eaed2) into main (531edf5) will decrease coverage by 7.72%.
The diff coverage is n/a.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1438      +/-   ##
==========================================
- Coverage   86.38%   78.65%   -7.73%     
==========================================
  Files         201      201              
  Lines       20854    21758     +904     
  Branches     2298     4421    +2123     
==========================================
- Hits        18014    17114     -900     
- Misses       2244     2399     +155     
- Partials      596     2245    +1649     

Flag	Coverage Δ
py-unittests	78.65% <ø> (-7.73%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
...elineio/opentimelineio-bindings/otio_anyVector.cpp	30.76% <0.00%> (-69.24%)	⬇️
...entimelineio-bindings/otio_serializableObjects.cpp	24.54% <0.00%> (-67.03%)	⬇️
...timelineio/opentime-bindings/opentime_bindings.cpp	33.33% <0.00%> (-66.67%)	⬇️
...eio/opentimelineio-bindings/otio_anyDictionary.cpp	33.33% <0.00%> (-66.67%)	⬇️
...imelineio/opentime-bindings/opentime_timeRange.cpp	32.39% <0.00%> (-61.36%)	⬇️
...ineio/opentime-bindings/opentime_timeTransform.cpp	42.30% <0.00%> (-57.70%)	⬇️
src/opentimelineio/stackAlgorithm.cpp	28.12% <0.00%> (-54.23%)	⬇️
...lineio/opentime-bindings/opentime_rationalTime.cpp	45.04% <0.00%> (-48.79%)	⬇️
src/opentimelineio/composition.h	40.74% <0.00%> (-48.15%)	⬇️
...melineio/opentimelineio-bindings/otio_bindings.cpp	51.58% <0.00%> (-46.71%)	⬇️
... and 60 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 531edf5...24eaed2. Read the comment docs.

[jlskuz]

@jlskuz I can replicate the docs behavior if I mess up my local install of pyaaf2. Does python -c "import aaf2" report a error?

That command worked just fine. It turned out that make tests and make doc-plugins-update got confused likely by the Kdenlive adapter and the AAF adapter I had installed separately from their new repos. After I uninstalled them and reset my OpenTimelineIO git clone, it worked properly.

I just force pushed the fix, it should be okay now.

[ssteinbach]

Thank you!