Implement caching in AiidaNodeViewWidget

[edan-bainglass]

Moving node view caching mechanism implemented by @superstar54 in the QE app (aiidalab/aiidalab-qe#921) into AiidaNodeViewWidget

[edan-bainglass]

@AndresOrtegaGuerrero this should resolve your issue. Essentially, when @superstar54 implemented caching of node views in the QE app, we stopped using AWB's AiidaNodeViewWidget, which supports non-widget views. This PR moves the mechanism into the widget, which will now be used in the QE app, which will re-enable non-widget views.

[codecov]

Codecov Report

Attention: Patch coverage is 76.47059% with 4 lines in your changes missing coverage. Please review.

Project coverage is 83.48%. Comparing base (9601736) to head (16799f2).
Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
aiidalab_widgets_base/viewers.py	76.47%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #686      +/-   ##
==========================================
+ Coverage   83.24%   83.48%   +0.23%     
==========================================
  Files          18       18              
  Lines        3588     3602      +14     
==========================================
+ Hits         2987     3007      +20     
+ Misses        601      595       -6     

Flag	Coverage Δ
python-3.11	83.48% <76.47%> (+0.23%)	⬆️
python-3.9	83.51% <76.47%> (+0.23%)	⬆️

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[danielhollas]

Moving caching mechanism implemented by @superstar54 in the QE app into the AiiDANodeViewWidget

Can you point to the original PR in QeApp. It's not super clear what is meant by caching in this context.

[superstar54]

Nice!

[superstar54]

Hi @danielhollas , It here: aiidalab/aiidalab-qe#921

Sorry that I didn't notice your message when approving this PR.

[superstar54]

Hi @edan-bainglass , as suggested by @yakutovicha , could you add a test for the caching?

[edan-bainglass]

Yep. I'll add a test for this and the loader from #685

[danielhollas]

I am a bit wary of the changes in this PR since the cache introduced here is unbounded. So for example for a large workchain, a user clicking around the node tree can blow up the memory of the Python process (I don't know if that is a real concern or node, but I assume some viewers may consume nontrivial amount of memory.

At the very least we need to provide a way to clear the cache, probably in the reset method. Or better yet, use one of pythons stdlib lru_cache decorators and wrap the logic in a cached function call somehow.