bug#54131: 29.0.50; Flyspell incorrectly reports first word in Python f-string

[Visuwesh <visuweshm@gmail.com>]

[வியாழன், பிப்ரவரி 24 2022] Eli Zaretskii wrote:

>> From: Lars Ingebrigtsen <larsi@gnus.org>
>> Cc: p.stephani2@gmail.com,  54131@debbugs.gnu.org
>> Date: Thu, 24 Feb 2022 10:12:58 +0100
>> 
>> Eli Zaretskii <eliz@gnu.org> writes:
>> 
>> > For program sources, flyspell-prog-mode relies on font-lock faces to
>> > tell it where strings and comments are, so perhaps this part doesn't
>> > work for some reason (and perhaps the root cause is in font-lock of
>> > python-mode, not in flyspell.el per se).
>> 
>> I didn't know about flyspell-prog-mode, but it does indeed do the right
>> thing here.  (I've now mentioned the mode in the flyspell-mode doc
>> string.)
>
> That's strange, because the OP explicitly invoked flyspell-prog-mode,
> so how come it didn't work for Philipp, but did for you?
>
>> But `M-$' does not do the right thing -- if you `M-$' on f'hello, it
>> tries to correct the whole thing.
>
> Maybe we should have a key binding for
> ispell-comment-or-string-at-point, and/or maybe
> ispell-comments-and-strings and flyspell-prog-mode should rebind M-$
> to ispell-comment-or-string-at-point.

I misspell parts of a symbol quite often so rebinding M-$ to a command
that NOOPs when the point is not over a comment/string would end up
annoying, IMO.  I am not sure how common this usecase is though.