On Mon, Dec 06, 2004 at 04:59:54PM +0100, Nicolas Pettiaux (AEL) wrote:
On Mon, 6 Dec 2004 00:19:33 +0100, Nicolas François wrote
> On Sun, Dec 05, 2004 at 08:46:53PM +0100, Martin Quinson wrote:
> > this is a very decent start, please commit it. Just rename it to LaTeX.pm
> > before. The case is important, and Latex is different.
>
> Do you rather TeX or LaTeX?
definitively LaTeX. It is more flexible.
Sorry. I wasn't clear. It was about the module name.
If the parser can parse LaTeX, it should probably also be able to parse
TeX.
Maybe TeX (alone) is not used anymore, and it's better to name the module
LaTeX.
I'm asking this because renaming a file in the CVS repository is not
really convenient.
> Any drawback in converting \\ to \newline?
I don't think so ... provided \newlines is accepted by LaTex (which I
don't remember and can"t find now).
\newline is mentioned in the JoliManuelPourLatex, but it is not clear if
it is equivalent.
\newline may cause an issue in the tabular environment.
maybe \newline doesn't support the optional argument that \\ does.
So I will try to do without this conversion.
> * Is there any problem if a space is added at the end of
lines?
no
That's cool.
> * Any idea on how to test the results? (diffing dvi?)
human reading could work ?
It will be the fall-back test mode.
What I'm willing to do is to parse a (La)TeX document with po4a and do a
so called po4a-normalisation (i.e. coping the string to translate on the
translated string), and then verify that the so built document is
identical to the original one.
But I know there will be some differences (for example, line re-wrapped,
spaces at the beginning or end of a line).
So I'm looking for a way to compare the graphical representation of the
LaTeX source.
I had a look to your book in the train. It gave me some ideas, so a longer
mail will follow.
--
Nekral