Use pass_name from pennylane transform

[albi3ro]

Context:

Trying to improve our integration between pennylane transforms and catalyst passes.

See #2206 for where we are able to update all built-in catalyst passes to be transforms instead. That PR helps validate we can make this change with no breaking behaviour.

Going forward, we can setup a plan to get rid of the catalyst-specific pass handling infrastructure.

Also note that this PR targets the PennyLane branch, so we can confirm we won't need anymore changes there to get catalyst working correctly before approving and merging.

Description of the Change:

Uses the transform.pass_name property as the higher priority source of the pass_name.

Updates the pass_pipeline from the quantum_kernel_p primitive to be a tuple of TransformContainer objects instead of Pass, but leaves in the logic to handle tuple[Pass,...] for safer backwards compatibility.

Also makes using transforms with pass names by splitting a QNode's transform program into "things at the start of the program that are tape transforms" and "things at the end of the program that are MLIR passes". This will allow us to get rid of PassPipelineWrapper use for both systems and soften the change from old frontend to new frontend.

Benefits:

Unified handling of transforms across pennylane and catalyst.

Possible Drawbacks:

Related GitHub Issues:

Depends on PennyLaneAI/pennylane#8539 [sc-103775]

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 97.56%. Comparing base (a452617) to head (ce51804).
⚠️ Report is 4 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #2149   +/-   ##
=======================================
  Coverage   97.55%   97.56%           
=======================================
  Files          93       93           
  Lines       11184    11219   +35     
  Branches     1065     1073    +8     
=======================================
+ Hits        10911    10946   +35     
  Misses        208      208           
  Partials       65       65           

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[paul0403]

Don't forget changelog (we can probably just add this one to the single transform primitive one)

[mudit2812]

This looks great to me! I'll let the compilation team be the final approvers here

[sengthai]

LGTM 👍