From octave-maintainers-request at bevo dot che dot wisc dot edu Fri Feb 20 21:49:53 2004
Subject: Re: TRACING
From: JD Cole <listguy at transientresearch dot com>
To: Jens Ruecknagel <Jens dot Ruecknagel at stud dot tu-ilmenau dot de>
CC: octave-maintainers at bevo dot che dot wisc dot edu
Date: Fri, 20 Feb 2004 21:19:17 -0800

It's quite possible that that variable has gone away, there have been 
some big changse to Octave internals since I wrote that code. You might 
try comparing that file, i.e. pt-emit-cplusplus.cc, to the 2.1.55 of 
Octave's pt-pr-code.cc---that file was the basis of 
pt-emit-cplusplus.cc. Basically pt-pr-code gives you a good overview of 
variables contained in each class of the parse tree, variable that will 
be usefull in creating the code converter we've been talking about.

Another thing you could do to get a better feel for overall Octave code 
is to use a code doc system like Doxygen to generate nicely format 
documentation of the code. Especially valuable is the way in which 
doxygen displays class dependencies. Check out www.doxygen.org, if you 
need help, get back to me.

JD

Jens Ruecknagel wrote:

> Hi JD Cole,
>
> Thanks to you I am tracing through Octave and your .oct File - so the 
> debugger looks up which header file a function comes from.
>
> I Compiled Octave with -O0 and your oct file with -O0 (had to screw 
> around in mkoctfile) - but now it works
>
> By the way - somehow your code does not compile with octave 2.54 - so 
> I downgraded to 2.50 - thats good enougth for right now
> in 2.54 the compiler complains about a not defined symbol in 
> compile.cc f->functionname - didn`t want to look around for it.
> I think there is no need for it right now.
>
> Jens
>