Confusing return type in tokenize.untokenize docs

[tomasr8]

Documentation

The docs for tokenize.untokenize are somewhat confusing
regarding the return type of the function. The second paragraphs says:

The reconstructed script is returned as a single string.

while the third paragraph clarifies that it can be both str and bytes:

cpython/Doc/library/tokenize.rst

Lines 100 to 102 in 6c52ada

	It returns bytes, encoded using the :data:`~token.ENCODING` token, which
	is the first token sequence output by :func:`.tokenize`. If there is no
	encoding token in the input, it returns a str instead.

I think we should remove the first sentence (The reconstructed script is returned as a single string.)
because it's wrong and if someone is just skimming the docs they might not realize that the full
explanation is only in the next paragraph.

Linked PRs

• gh-129020: Remove ambiguous sentence from tokenize.untokenize docs #129021
• [3.13] gh-129020: Remove ambiguous sentence from tokenize.untokenize docs (GH-129021) #129035
• [3.12] gh-129020: Remove ambiguous sentence from tokenize.untokenize docs (GH-129021) #129036