Re: replacing munge-texi with Perl

[Rik]

Søren Hauberg wrote:
> tor, 10 02 2011 kl. 13:23 -0800, skrev Rik:
>> Michael D Godfrey wrote:
>>> On 12/16/2009 04:05 PM, Rik wrote:
>>>> I think we should be open-minded about how munge-texi is implemented. 
>>>> It is currently written in C++, but I find doing any text processing in
>>>> C/C++ to be awkward.  We already require Perl for building, so if you
>>>> think it might be easier to provide the functionality in another
>>>> language then consider that.
>>>>
>>>>   
>>> I agree with this.  My quick inspection of munge-texi.cc finds that it just
>>> passes the @example ... @end example code through.  And, using C++
>>> for more text manipulation is not, I think, a very good idea.  So, a rewrite
>>> of munge-texi.cc into Perl may be a good start.  Or, should Octave be
>>> considered?
>>>
>>> Michael
>>>
>> This exchange was from 2009, but I thought I would go ahead and implement
>> it now that 3.4.0 is out.  Any objections?
> 
> Just curious but why not rewriting it into Octave code?
> 
Mostly because Perl is very good for text and I know the language well
enough for this to be an easy task.  munge-texi could probably be replaced
by something in awk as well, but I don't know the language.  Awk would not
increase our dependency on Perl for which jwe expressed a concern.

Octave will be requiring the PCRE library going forward which may make this
a simple enough task to be rewritten in Octave.  I'll see if I can do it in
10 minutes.  I certainly can do it in 5 in Perl.

--Rik