PERF: Fix reference leak in read_hdf

[lithomas1]

• closes BUG: Excessive memory usage loading Dataframe with mixed data types from HDF5 file saved in "table" format #37441 (Replace xxxx with the GitHub issue number)
• Tests added and passed if fixing a bug or adding a new feature
• All code checks passed.
• Added type annotations to new arguments/methods/functions.
• Added an entry in the latest doc/source/whatsnew/vX.X.X.rst file if fixing a bug or adding a new feature.

With my really hacky benchmark, I get,

       before           after         ratio
     [bb18847e]       [b5ddc810]
     <main>           <fix-hdf-memusage>
-            6.6M                0     0.00  io.hdf.HDF.mem_read_hdf_index('table')

SOME BENCHMARKS HAVE CHANGED SIGNIFICANTLY.
PERFORMANCE INCREASED.

[lithomas1]

Not a fan myself of how this is written.

asv's memory benchmarks use pympler.asizeof internally, but with a view of a numpy array, it'll only give us the size of the view.

peakmem avoids this problem by tracking RSS, but that doesn't catch this improvement, since this only improves the memory usage at the end.

[mroeschke]

Could we assert this is None in a unit test instead?

[lithomas1]

Yeah, that's probably better.

[mroeschke]

Thanks @lithomas1