bug-guile
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Documentation with GCC3.1, (was: build error on solaris)


From: Martin Grabmueller
Subject: Re: Documentation with GCC3.1, (was: build error on solaris)
Date: Fri, 23 Mar 2001 10:34:56 +0100

> From: Neil Jerram <address@hidden>
> Date: 23 Mar 2001 09:07:32 +0000
> 
> >>>>> "Neil" == Neil Jerram <address@hidden> writes:
> 
> >>>>> "Alexander" == Alexander Klimov <address@hidden> writes:
>     Alexander> The change which broke guile documentation is that
>     Alexander> [...]
> 
>     Neil> I'm planning to try fixing this, mainly this Friday.
> 
> Hmm...  This is harder than I thought to fix elegantly.  Two reasons
> suggest that it would be _wrong_ to try to use Guile to replace the
> Awk step here:
> 
> (1) Guile startup time being quite slow.
> 
> (2) Regexps are an optional feature when configuring Guile, so it
>     seems unwise to assume that regexp and string matching procedures
>     would be available.
> 
> Any thoughts or ideas?

(1) We could change the build stuff to fire up Guile once, and snarf
    all .x and .doc files in one run.  That would probably break
    incremental building, though.

(2) It's not only that we can't depend on regexp functions, we cannot
    depend that Guile is available at all, I think.

A possibility would be to distribute the .x and .doc files, and only
regenerate them if the corresponding .c files are changed.  I f that's
a good idea, I don't know.

What about running some snarfing script (awk or guile) over the C
sources instead of the output of the C preprocessor?  The format of
the C files (comments, SCM_DEFINEs etc.) is well suited for parsing.
(not that I tried it already, though...)

Regards,
  'martin



reply via email to

[Prev in Thread] Current Thread [Next in Thread]