Clean up QuantumScript._par_info

[albi3ro]

Note that PR branches off of #3097 and must be merged after it.

The QuantumTape._par_info (now QuantumScript._par_info) was a dictionary mapping from integers starting at zero to a dictionary. This modifies that data structure to be a simple list instead.

This required several changes through the code base:

• Switching tape._par_info.items() to enumerate(tape._par_info)
• Switching tape._par_info.values() into tape._par_info.

[codecov]

Codecov Report

Merging #3185 (5fc88ef) into master (f5d1e68) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master    #3185      +/-   ##
==========================================
- Coverage   99.69%   99.69%   -0.01%     
==========================================
  Files         275      275              
  Lines       24014    24012       -2     
==========================================
- Hits        23940    23938       -2     
  Misses         74       74              

Impacted Files	Coverage Δ
pennylane/circuit_graph.py	93.06% <100.00%> (ø)
pennylane/gradients/gradient_transform.py	100.00% <100.00%> (ø)
pennylane/gradients/parameter_shift_cv.py	99.59% <100.00%> (ø)
pennylane/tape/qscript.py	100.00% <100.00%> (ø)
pennylane/tape/tape.py	97.58% <100.00%> (+0.01%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[AlbertMitjans]

Just one comment:

The docstring of QuantumScript._update_trainable_params should be updated:

def _update_trainable_params(self):
    """Set the trainable parameters

    Sets:
        _trainable_params (list[int]): Script parameter indices of trainable parameters

    self._par_info.keys() is assumed to be sorted and up to date when calling
    this method. This assumes that self._par_info was created in a sorted manner,
    as in _update_par_info.

    Call `_update_par_info` before `_update_trainable_params`
    """

[antalszava]

Looks good! 👏