Fixing labeler again

[germa89]

This thing breaks way too much in my opinion.

But finally, it should close #1337.

[jorgepiloto]

Try to add the following, just in case.

[jorgepiloto]

If that doesn't work, we can always use the demo worklows.

[codecov]

Codecov Report

Merging #1343 (12e02e5) into main (efc02a9) will not change coverage.
The diff coverage is n/a.

❗ Current head 12e02e5 differs from pull request most recent head 9bb254d. Consider uploading reports for the commit 9bb254d to get more accurate results

@@           Coverage Diff           @@
##             main    #1343   +/-   ##
=======================================
  Coverage   75.91%   75.91%           
=======================================
  Files          43       43           
  Lines        6813     6813           
=======================================
  Hits         5172     5172           
  Misses       1641     1641           

[github-actions]

Please add one of the following labels to add this contribution to the Release Notes 👇

• BUG
• NEW FEATURE
• Enhancement
• Documentation
• Examples
• Maintenance
• CI/CD

[jorgepiloto]

It is working now 😄

[germa89]

yep! I'm lucky sometimes.

But I think it should not comment if the bot add some labels

[germa89]

This if:

if: toJSON(github.event.pull_request.labels.*.name) == '[]'

is not working fine.

[germa89]

Pinging @PProfizi for keeping him on the loop.

Also @RobPasMue because surely he can add something to this and probably he is looking for some fun stuff. xD

[germa89]

I can't believe it took 5 runs before I got all the tests to pass (I guess Github is not very stable lastly).

[germa89]

Any ideas on this:

But I think it should not comment if the bot add some labels

@jorgepiloto @RobPasMue

[RobPasMue]

I've heard this is kind of an ugly hack but... It may work. So, we could set up an of condition for the commenter similar to this code block https://github.com/orgs/community/discussions/27125#discussioncomment-3254718

[RobPasMue]

In case there already labels --> Avoid commenting.

Targeting specific bot actions is going to be more difficult I guess...

[akaszynski]

I can't believe it took 5 runs before I got all the tests to pass (I guess Github is not very stable lastly).

I've seen this as well. The issue isn't GH, it's anything that requires IO is much less forgiving in regards to buffers and caching. I've seen similar issues with race conditions on pyvista. It's just a sign we need to fix MAPDL having to write to disk for each and every command (unless muted).

[germa89]

I've heard this is kind of an ugly hack but... It may work. So, we could set up an of condition for the commenter similar to this code block https://github.com/orgs/community/discussions/27125#discussioncomment-3254718

What about this?

https://github.com/orgs/community/discussions/27125#discussioncomment-3254720

[germa89]

I'm just going to merge, so I can test this is working.

[germa89]

Testing:

• No manual label, no significant files/directories changes. Expected: Bot comment.
  testasdfasdf #1381 🆗

• No manual label. Changing marked file. Expected: Bot labeling but not commenting.
  myasdftest2 #1382 🆗

Both works.... Green light 🟢