Add device preprocessing to from_plxpr

[mudit2812]

Context:

• This PR adds support for device preprocessing with program capture.

Description of the Change:

• Add a new catalyst.from_plxpr.device_utils submodule, which contains a function for generating a device preprocessing pipeline based on device capabilities and the execution config.
  • For transforms that don't have a pass_name, the transform is processed as follows:
    • a dummy MLIR transform, empty, is used. This transform has a key string argument that is used to indicate to devs which transform this pass is standing in for.
    • If any dummy transforms are added to the pipeline, a warning is raised to let the user know that some dummy transforms are present.
      • Note that this is only true for dummy transforms added conditionally. Transforms that are always added which don't have native implementations do not trigger this warning.
• Create a "device"pass pipeline that is added to the quantum_kernel_p binding in from_plxpr.
• Add a skip_preprocess argument to qjit, trace_from_pennylane, from_plxpr, and WorkflowInterpreter, which can be used to skip appending device preprocessing transforms to the pass pipeline.
• Add a _preprocess_warn=True argument to from_plxpr and WorkflowInterpreter which can be used to toggle the dummy transform warnings off. This argument is private and not exposed by qjit. I added it as a testing convenience.
• The preprocessing function also does MCM related validation and preprocessing.

Benefits:

• Device preprocessing available with program capture.

Possible Drawbacks:

• Depending on the device being used (and hence, which preprocessing transforms are used), the dummy transform warning can get annoying.

Related GitHub Issues:

[sc-111493] [sc-112073]

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 96.91%. Comparing base (2fd67bd) to head (ee2688f).
⚠️ Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2557      +/-   ##
==========================================
+ Coverage   96.90%   96.91%   +0.01%     
==========================================
  Files         155      156       +1     
  Lines       16974    17065      +91     
  Branches     1636     1659      +23     
==========================================
+ Hits        16448    16539      +91     
  Misses        390      390              
  Partials      136      136              

☔ 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]

Amazing PR! This is a big step towards the device capabilities mliestone in capture! 💯 🥇

[albi3ro]

changelog?

[paul0403]

🥳

[albi3ro]

Looks great! Thanks so much for getting this all done. 🚀