vectorized note and svara converters properly

[bmcfee]

Reference Issue

Fixes #1553

What does this implement/fix? Explain your changes.

This PR replaces the optional recursion in midi_to_note (and in _svara_*) by a proper numpy vectorization. This simplifies both the implementation of these functions, and allows them to operate more cleanly on arbitrary array shapes.

It also introduces a function librosa.util.decorators.vectorize, which operates similarly to np.vectorize but preserves scalar inputs. (Numpy vectorize, as noted in #1553, maps everything to arrays, including scalars.)

Example usage:

With a multi-dimensional array (list of lists), we now produce a proper ndarray output:

In [11]: librosa.midi_to_svara_c([[22, 23, 24], [25, 26, 27]], Sa=60, mela=1)
Out[11]: 
array([['N₂', 'N₃', 'S'],
       ['R₁', 'G₁', 'G₂']], dtype='<U2')

with a scalar input, we produce a scalar output as exepcted:

In [12]: librosa.midi_to_svara_c(22, Sa=60, mela=1)
Out[12]: 'N₂'

This is a very slightly API-breaking change, in that we now produce ndarrays instead of lists. I'm okay with doing this without a full deprecation cycle, but we should definitely highlight it in the release notes.

Any other comments?

A few things I'd like to improve in this PR:

1. The vectorize helper is not a proper decorator. It doesn't necessarily need to be, but it would be a bit more elegant if it was. I just couldn't get my head through the extra level of indirection involved in making a decorator factory.
2. Tests should be expanded to cover the new use cases.

[codecov]

Codecov Report

Base: 98.75% // Head: 98.76% // Increases project coverage by +0.00% 🎉

Coverage data is based on head (c7f4349) compared to base (6562758).
Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1572   +/-   ##
=======================================
  Coverage   98.75%   98.76%           
=======================================
  Files          32       32           
  Lines        4030     4041   +11     
=======================================
+ Hits         3980     3991   +11     
  Misses         50       50           

Flag	Coverage Δ
unittests	98.76% <100.00%> (+<0.01%)	⬆️

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

Impacted Files	Coverage Δ
librosa/core/convert.py	99.09% <100.00%> (-0.01%)	⬇️
librosa/util/decorators.py	100.00% <100.00%> (ø)
librosa/filters.py	98.24% <0.00%> (ø)

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

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[bmcfee]

I think this one is ready to - anyone want to check it over?

[bmcfee]

I'd like to merge this one today, unless there are objections.