Gabor Hojtsy wrote: >> Which the great problem with OO after all? Render, search, filenames? >> http://www.docbook.org/tdg/en/html/classsynopsis.html > > Rendering, what to put into a file, crosslinks, relating OO docs to > functional docs, etc... > > Goba
RENDERING The bigger problem for OO docs is rendering then, rigth? It's because isn't no space in methodsynopsis to put a class name. In my previous post I say to use a script to build class.Z.xml from refentry's ID. So, the resulting XML should be something as: <reference id="ref.dir"> &reference.dir.functions; &reference.dir.classes; </reference> &reference.dir.functions; and functions.xml remains the same. &reference.dir.classes; and the new classes.xml should be: &reference.dir.classes.Directory; witch points to generated file class.Diretory.xml: <classsynopsis> <ooclass><classname>Directory</classname></ooclass> &class.Directory.construct.1; &class.Directory.method.read; &class.Directory.method.close; [...] </classsynopsis> The important thing here is that now the methodsynopsis appear inside of an classsynopsys! (but between then appear refentry and others ref* tags). It is here that I can be wrong, but: in a XML reder it's easy to build a stylesheet witchs and <method>class.Diretctory.method.close</method> and renders to {$first_up_occurrence_classsynopsis->ooclass->text}"->"{methodsynopsis->meth odname->text} Yes ... ? -------------------------------------------- Well, still remains the problem to separate the functions of the global space from object methods. Easy ;) All global functions may be documented by funcsynopsys and object methods by methodsynopsis (now that they appear INSIDE of classsynopsis) []s André Ć