bug#81638: 32.0.50; flyspell-prog-text-faces type and semantics

>> > And if all of the above is not enough, and you still insist, I could
>> > agree to adding nil to the valid values (but then please suggest a
>> > suitable :tag),
>> BTW, there can be other good reasons to use a value not covered by the
>> current `:type`, e.g. it can make a lot of sense to put `sh-heredoc` or
>> `perl-heredoc` in there.
> AFAIU, those are not normally human-readable text, so I'm not sure
> what sense you are seeing in doing that.

I'd be surprised if I'm the only one who has shell scripts where
heredocs contain natural language text.  I agree it might not be
desirable to add those elements globally, but they do make sense
file-locally.

[ BTW, another face I've seen used in there is things like
  `font-lock-function-name-face` so as to spell check identifier names
  at the point where they're chosen but not when merely referring to
  a function defined elsewhere (where presumably the compiler will
  detect typos).  ]

>> >> But I wonder if we should, in addition, generalize
>> >> `flyspell-generic-progmode-verify' so that members of
>> >> `flyspell-prog-text-faces' are allowed to be functions (test
>> >> predicates).
>> > If you want to do that, it is better (and easier) to override
>> > flyspell-generic-progmode-verify with your own function.
>> You mean set `flyspell-generic-check-word-predicate`?
> That's one way, yes.  But if someone wants to use flyspell-prog-mode,
> the variable is already set.

`flyspell-prog-mode` doesn't do anything else than set
`flyspell-generic-check-word-predicate`, so if you want a different
predicate you're better off using `flyspell-mode`, indeed.


=== Stefan