Re: Silence checkdoc for symbols designating major and minor modes

[Philip Kaludercic]

Eli Zaretskii <eliz@gnu.org> writes:

>> From: Philip Kaludercic <philipk@posteo.net>
>> Date: Tue, 27 Jul 2021 21:50:43 +0000
>> 
>> an issue I have been having when documenting code is that checkdoc wants
>> me to disambiguate symbols referring to major or minor modes. That
>> usually means adding "function", "command" "variable", "option" or
>> "symbol" but in my experience these usually do not fit. In fact, most of
>> the time it is obvious what is meant when referring to a major/minor
>> mode.

First of all, I have to correct myself: This only applies to minor
modes, because they are both bound and fbound.

>> 
>>         ... Foo when `bar-mode' is active ...
>> 
>> The patch I added below silences the disambiguation request when a
>> symbol ends in "-mode". Alternatively, one could also add a "minor mode"
>> and "major mode" to the list of accepted keywords, but I think that
>> would sound to repetitive:
>> 
>>         ... Foo when minor mode `bar-mode' is active ...
>> 
>> Opinions?
>
> First, the code, if installed, should have a comment explaining why
> these strings are being exempted.

Of course, this was just a first suggestion.

> More generally, could you please show an example of a doc string and
> the warning(s) it triggers?  I'm not sure I understand the problem
> well enough to make up my mind.  (Do all of our existing modes suffer
> from this problem?)

I noticed this when updating the rcirc docstrings. For example, I still
have the issue here:

        (defcustom rcirc-omit-responses
          '("JOIN" "PART" "QUIT" "NICK")
          "Responses which will be hidden when `rcirc-omit-mode' is enabled."
          :type '(repeat string))

triggering this response:

        Disambiguate rcirc-omit-mode by preceding w/ 
function,command,variable,option or symbol.

Browsing through the repository, I also discovered the same issue in
turn-on-eldoc-mode's docstring. There are other examples in third-party
code I remember, but couldn't find right now.

> Thanks.

-- 
        Philip Kaludercic