axiom-developer
[Top][All Lists]
Advanced

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

RE: [Axiom-developer] Test of a WEB style


From: C Y
Subject: RE: [Axiom-developer] Test of a WEB style
Date: Fri, 4 May 2007 04:17:58 -0700 (PDT)

--- Bill Page <address@hidden> wrote:

> On May 4, 2007 12:52 AM C Y wrote:
> > 
> > I have created a sample file of a format for pamphlets that can
> > be generated without needing the noweb style file (obviously
> > without many of the major features.)  
> > 
>
http://portal.axiom-developer.org/Members/starseeker/cl-web-v0.5-hyperreftes
> t.pdf/download
> 
> Cliff, I do really like the "look and feel" of this example!
> It looks pretty and it seems to work.

Thanks!

> > This file makes use of the hyperref and listings packages in
> > LaTeX (I believe both are included standard with teTeX).  It
> > has hyperlinks from chunk references to chunks.  To generate
> > it, the commands were:
> > 
> > latex cl-web-v0.5-hyperreftest.tex
> > bibtex cl-web-v0.5-hyperreftest
> > latex cl-web-v0.5-hyperreftest.tex
> > latex cl-web-v0.5-hyperreftest.tex
> > latex cl-web-v0.5-hyperreftest.tex
> > dvips cl-web-v0.5-hyperreftest.dvi -o cl-web-v0.5-hyperreftest.ps
> > ps2pdf cl-web-v0.5-hyperreftest.ps
> > 
> > (thanks again to Bill for figuring out these steps and the
> > hyperref options for endpamphlet).
> 
> I presume the above steps (because there are so many) include
> the processing for the embedded graphviz commands used to
> generate the finite-state diagram? If so, very nice.

No, not yet actually - I didn't want to introduce the build requirement
of graphviz into the Axiom source distribution yet (where this is
hopefully headed eventually).  I added an extra latex or two to make
sure all the references got resolved, but two after the bibtex file
might be enough.  The dvips and ps2pdf option works better for the
tangle diagram (pdflatex doesn't like it, even after I convert it -
probably need to play with bounding boxes or something...)

I think that probably CAN be done without too much trouble, but it
would mean graphviz would be needed to build it.  What do we want to do
about that?  Implementing the graphviz algorithms in Lisp is going to
be a bit more involved than noweb...

> What does the actual pamphlet source file look like? Specifically,
> how are chunks delimited? Can you give us a link to the source
> .pamphlet file?

Right now, they would be delimited just like a noweb pamphlet.  Chunk
delimiters aren't so critical actually, unless we want NO
pre-processing steps before latexing a pamphlet file.  There isn't a
pamphlet source file that gets translated to this form yet, but once
weave is written in Lisp existing pamphlets should be translatable to
this without major (or any, for that matter) adjustment.

> > If it is of interest the "Included in ..." tail can be added
> > to identify chunks containing the current chunk, although I 
> > didn't do that here because I'm generating this one by hand
> > as a test case.
> 
> Ah. By "generating this one by hand" do you mean that this is
> not (yet) generated by some operation equivalent to 'noweave'?

Right.  I want to know what I need to generate before I start on the
machinery to do so.

> So you do not (yet) have the source that would produce the above
> result? Or are you just talking about the "Included in ..."?

The regular pamphlet source should produce this result - what I don't
have yet is the routines that will build the LaTeX from the pamphlet
structure.  The TeX file I did is here, FWIW:

http://portal.axiom-developer.org/Members/starseeker/cl-web-v0.5-hyperreftest.tex/download

> > Feedback appreciated - is this something people think would
> > be good to have?  If so, I believe it can be autogenerated
> > from cl-web requiring only packages already present in teTeX.
> 
> How much work remains to be done to do this autogeneration?
> (autogeneration == noweave?)

It SHOULD be a case of adding some information recording to the chunk
structure and scan-for-chunks function, creating a utility to convert
strings and chars to the correct output stream format, and writing the
weave function to make the correct calls.  There may be a couple other
gotchas to deal with once I actually start but that's the core of what
needs to be done.

Cheers,
CY

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 




reply via email to

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