bug#57551: 29.0.50; hide-show in python-mode selects the current block imperfectly

[Dima Kogan]

Hi. I've been using the recently-committed hide-show functionality from:

  https://debbugs.gnu.org/cgi/bugreport.cgi?bug=56635

It's working well, but I just found a problem. Python lacks explicit
block-end characters that a language like C has. This creates some
ambiguity about what is intended, and I think emacs makes the wrong
choice in at least one spot. Let's say we have a tst.py:

def f():
    if 1:
        11
    elif 2:
        22
    elif 3:
        33
    elif 4:
        44
    else:
        5

If I want to collapse the "elif 2:" block, I move the point to somewhere
inside that block, and (hs-hide-block). Placing the point at the end of
the "elif 2:" line puts us unambiguously inside that block, so this does
the job.

But what if the point is at the start of the "elif 2:" line? Currently
emacs considers this to be INSIDE this block also, so (hs-hide-block)
there also hides the "elif 2" block, but should it? Would it not make
more sense if the START of the "elif 2:" line was considered in the "def
f():" block, but outside all the "if" blocks? Then (hs-hide-block) at
the start of that line would collapse the whole "def f():" block.
Currently the only point where this can be done is at the end of the
"def f():" line.

This also raises a question about what should happen if we
(hs-hide-block) with the point on the "i" in any of the "elif" lines.
Probably this should be outside the "elif" blocks too? Maybe?

Thanks!