On 9/11/07, Alain Matthes <alain.matthes@mac.com> wrote:

But why conditionally include grammars ??

Because then you could base which grammars are included based on which packages are included in each document. Suppose that this actually worked:

{    name = 'meta.package.latex';
            match = '^\usepackage{lstlisting}'
            patterns = ( { include = 'text.tex.latex.lstlisting'; } );
        },

or even better
{    name = 'meta.package.latex';
            match = '^\usepackage{(lstlisting)}'
            patterns = ( { include = 'text.tex.latex.$1'; } );
        },

I'm not a specialist of script, regex etc.... but in a first time
it's possible to build some grammars like

text.tex
text.tex.math
text.tex.beamer
text.tex.listings
text.tex.tikz etc....

And to finish the text.tex.latex by some include like

include = ' text.tex.listings'; ?

Each user can modify the list of the package grammars !

I think this is where we differ. I don't think we want each user to be modifying the latex grammar to decide which packages to include. For one thing it will be different for each different project. So, I"m either going to end up including everything all the time, or I'm going to drive myself crazy modifying the latex grammar every time I work on a different writing project. Plus most users don't want to go anywhere near editing the grammar file!

About listings, why make a grammar to this package ?

Because lstlistings is the best package I've found for including source code, and since I'm writing CS textbooks, I do that a lot. If you can recommend a better one I would be happy to look at it! But really, most of the grammar for the package is only concerned with then command and then including the grammar for the particular source code that is inside the lstlistings environment.

There is a way to turn around the difficulties.

\lstinputlisting[key=value list]{hfile name}

example :

\usepackage{listings}
\lstset{language=C}
...
\begin{document}
\lstinputlisting{main.c}
\end{document}

or you can make that

\lstinputlisting[first=10,last=50]{main.c}

I know about this feature of lstlisting but I like to have the code right there to look at as I'm writing. I guess its a writing style issue but I tried your method for a while and I did not like it at all.

It is the only way for me to use listings because I'm french (I need
only to
include latex code but with accents) and it's impossible
to use listings with utf8 actually, so I need to save file with
latin1 encoding
and to use two encodings latin1 and utf8 is the same file ! (very bad)

I don't understand about the problem with utf8. My latex source files are all encoded as utf8 documents and I don't have any problem with lstlistings.

I suppose that you need to include java code or python code but there
are scopes
to verify the syntax of these languages. So why waste time with a
difficult grammar and not
include the sources of codes.

Again, the lstlisting grammar includes source.java or source.python. In fact by my calculations the number of lines in the latex grammar dedicated to the listing package is only 47 lines out of 494 total.

Regards Alain
______________________________________________________________________
For new threads USE THIS: textmate@lists.macromates.com
(threading gets destroyed and the universe will collapse if you don't)
http://lists.macromates.com/mailman/listinfo/textmate