For other reviewers: note that this fixes a latent problem that only becomes evident if you add another rst document for something else that also has a values in its namespace, such as the planned docs/library/string.templatelib.rst from #17557.

When adding the above file to e.g. df05cae, and running make -C micropython/docs html, this produces the error:

~/micropython/docs/library/btree.rst:154: WARNING: more than one target found for 'any' cross-reference 'values()': could be :py:meth:`btree.btree.values` or :py:attr:`string.templatelib.Template.values`
~/micropython/docs/library/btree.rst:159: WARNING: more than one target found for 'any' cross-reference 'values()': could be :py:meth:`btree.btree.values` or :py:attr:`string.templatelib.Template.values`

Noting the change to the rendered output.

The current document renders as:

Constants¶

btree.INCL¶

A flag for keys(), values(), items() methods to specify that scanning should be inclusive of the end key.

btree.DESC¶

A flag for keys(), values(), items() methods to specify that scanning should be in descending direction of keys.

While the modified document renders as:

Constants¶

btree.INCL¶

A flag for btree.keys(), btree.values(), btree.items() methods to specify that scanning should be inclusive of the end key.

btree.DESC¶

A flag for btree.keys(), btree.values(), btree.items() methods to specify that scanning should be in descending direction of keys.

In my experience in parsing the .rst to create the MicroPython stub files, there are several different practices used wrt notation of method and class names.

I personally think the 2nd one is more explicit and clear.