Re: Integration of Info manuals in programming modes

[Eli Zaretskii]

> From: Visuwesh <visuweshm@gmail.com>
> Cc: arne_bab@web.de,  rms@gnu.org,  spwhitton@spwhitton.name,
>   jb@jeremybryant.net,  emacs-devel@gnu.org
> Date: Thu, 05 Dec 2024 16:58:22 +0530
> 
> [வியாழன் டிசம்பர் 05, 2024] Eli Zaretskii wrote:
> 
> >> From: Visuwesh <visuweshm@gmail.com>
> >> Cc: arne_bab@web.de,  rms@gnu.org,  spwhitton@spwhitton.name,
> >>   jb@jeremybryant.net,  emacs-devel@gnu.org
> >> Date: Thu, 05 Dec 2024 15:19:13 +0530
> >> 
> >> [வியாழன் டிசம்பர் 05, 2024] Eli Zaretskii wrote:
> >> 
> >> >> but I do not understand how info-lookup works really.  C-h S RET on
> >> >> 'read' jumps to the FGET page which is not we want.
> >> >
> >> > Look at the index node you mentioned: what other menu item that
> >> > matches "read" there is more appropriate?  And what did you expect to
> >> > see instead?
> >> 
> >> I would rather see a complain about no matches.  If someone could figure
> >> out how to reject non-exact matches, that would be do the job.
> >
> > That's a double-edge sword: rejecting non-exact matches could miss
> > valid hits.
> 
> I agree, dismissing a invalid match is not too much trouble.
> 
> > However, maybe the regexp should only match keywords in upper-case,
> > and ignore-case should not be used?
> 
> But Fortran is a case-insensitive language and recent style-guides ask
> you to use lower case lettering for everything.  Other code that I've
> seen use upper case lettering for keywords, not for functions and
> subroutines so we cannot opt for this route unfortunately.

What is important in this context is not whether Fortran is
case-insensitive, but how the index of the manual spells the keywords.