Re: [NTG-context] algorithms style : need \starttyping with TeX and math feature

3 Jan 2005

      Le 25 déc. 04, à 20:03, John Culleton a écrit :
...
On Friday 24 December 2004 09:23, Mojca Miklavec wrote:
...
Maurice Diamantini wrote:
...
Well I agree that it's no more "typing" any more!
So a special option for \setuplines would be more appropriate?
Exactly, \starttyping does exactly that what you don't want: it 
disables
math and expansion of TeX commands.
...
Also, perhaps my request is simply an option of \startlines
which keep not only the lines break, but also the first
spaces of each line (in a verbatim font fors these space)
Happy new year !
Thanks to Mojca and John, I use \startlines for the
algorithms environment.

In the "Algo" environment, I defined some short names command
to help dealing with indent spacing (\0, \1, ..., \5)
and with itemize like indent (i.e. space with the
same length as "- "

Here is the environment I defined. It could be better
(to keep the use of it event simpler)
But as the "père Noël" Hans give me a "Christmas present"
(see next post) It is no more usefull!

     \definestartstop[Algo][
        before={
             \start
             \tx
             \def\H{\phantom{- }}% H for Hyphen (espace correspondant à 
"- "
             \def\0{\hskip0em}  % indent 0
             \def\1{\hskip2em}  % indent 1
             \def\2{\hskip4em}  % indent 2
             \def\3{\hskip6em}  % indent 3
             \def\4{\hskip8em}  % indent 5
             \def\5{\hskip10em} % indent 6
             \setupinterlinespace[small]
             \startframedtext[
                  width=\makeupwidth,
                  framecolor=blue,
             ]
             \startlines
        },
        after={
             \stoplines
             \stopframedtext
             \stop
        },
     ]

Just a "french like" algorithm

\startAlgo
\0  Entrée :
\1   - nom de l'instance à traiter (e.g. ucp8, ...)
\1   - LDS\_Limit Nombre maximum de discordances (e.g. 1 pour LDS\_1, 
...)
\0  Sortie :
\1   - affichage des meilleures solutions primale et duale
\0  Algorithme :
\1   - initialisation :
\2      - LDS\_Counter := 0 (nombre de discordances courantes)
\2      - lecture de l'instance
\2      - propagation initiale
\1   - relaxation lagrangienne principale avec primalisations interne
\1\H   $\Rightarrow$  on connait bestDualValue,  bestPrimalValue et les 
coûts réduits
\1   - explorer le noeud racine (i.e. toutes les variables étant libres)
\1   - afficher les résultats
\stopAlgo

Thanks you very much !

Maurice