Re: Spell checking of the commentary section by checkdoc.el.

[Lute Kamstra]

"Stefan Monnier" <monnier+gnu/emacs/address@hidden> writes:

>> That seems like a reasonable change to me.  It may accidentally
>> pick up todo lists, and other miscellany comments too, but they may
>> need spell checking too.
>
> Shouldn't we use lisp-mnt.el functionality instead ?
> (or add/fix it if it's missing/broken)

Currently, lisp-mnt.el has the functions lm-commentary-mark, which
returns the beginning of the commentary section, and lm-commentary,
which returns the comment section as a string.  checkdoc.el cannot use
the latter function as it uses ispell for spell checking and, as I
understand it, ispell only works on buffers.  So checkdoc.el
additionally needs the end of the commentary section (as a buffer
position).  We could introduce lm-commentary-end for this purpose and
let lm-commentary use it (as well).  The way lm-commentary now
determines the end of the commentary section is not very robust,
though.  It depends on the presence of a ChangeLog/History or Code
header.  Maybe we should let lm-commentary-end fall back to my method
of finding the end of the commentary section (i.e., skipping over all
whitespace and comment lines) in case the aforementioned headers are
absent?


Lute.