Le 14/03/2014 01:39, Zdenek Wagner a écrit : > 2014-03-13 19:09 GMT+01:00 François Patte > <francois.pa...@mi.parisdescartes.fr>: >> Bonjour, >> >> When writing a Sanskrit text, there are (at least) two ways: roman >> transliteration, or devanagari. >> >> In a transliterated text, there are much more blank spaces between >> "words" than in a devanagari text. I would like to be able to have only >> one file for a text and choose either transliteration or devanagari. >> >> Is it possible to have some active character which separates "words" in >> the file and this character will give a blank space, if I choose >> tranliteration, and will give a null space if I choose devanagari, but a >> null space which allows ligature if necessary. >> >> For instance, in the file we have: tasmād_dhetor_bijasya and we get, >> after compilation: >> >> tasmād dhetor bījasya (in transliteration) >> >> or >> >> तस्माद्धेतोर्बीजस्य (in devanagari) >> >> (I have chosen underscore as the active character, but it is not the >> best idea...) >> > The TECkit mapping is applied after all macros are expanded, just > before the horizontal list is built. You have to decide in advance > whether _ will expand to \space or to nothing. Thus you can do the > following: > > \documentclass[a4paper,11pt]{report} > > \usepackage{xltxtra} > > \setmainfont[Mapping=tex-text]{% > FreeSerif% > } > > \usepackage{polyglossia}% I prefer polyglossia to babel > \setdefaultlanguage{sanskrit} > > \newfontfamily\sanskritfont[% > Script=Devanagari,Language=Sanskrit,AutoFakeSlant=0.2]{FreeSerif} > > \begin{document} > > \catcode`\_=13 > \def\MyText{tat_sat śrī_bhagavān_uvāca}% _ not expanded here > > % Sanskrit is the default language, no need to change the font > > % As default, we have the transliteration, we have to define _ > \def_{\space} > > \MyText > > {% Devanagari, redefine _ and add mapping > \def_{} > \addfontfeature{Mapping=RomDev} > \MyText > > } > > % Transliteration again > \MyText > > \end{document} > >
Thank you! I must use babel instead of polyglossia because the features concerning French language are very poor in polyglossia when we compare with those in babel. -- François Patte UFR de mathématiques et informatique Laboratoire CNRS MAP5, UMR 8145 Université Paris Descartes 45, rue des Saints Pères F-75270 Paris Cedex 06 Tél. +33 (0)1 8394 5849 http://www.math-info.univ-paris5.fr/~patte
signature.asc
Description: OpenPGP digital signature
-------------------------------------------------- Subscriptions, Archive, and List information, etc.: http://tug.org/mailman/listinfo/xetex