On Tue, Aug 14, 2012 at 12:33:58AM +0200, Thien-Thi Nguyen wrote: > > I think so. > > However, as mentioned in another post, whatever format we choose should > support indexing and incremental loading (similar to the way Emacs keeps > track of and displays docstrings). With that in mind, i realize now > that SXML alone is unsuitable since it doesn't even support the integer > data type! (I can't imagine building an index w/o integers.) Maybe i'm > just ignorant -- XML experts please correct me if i'm missing something. > > Ideally, if i were to write a renderer, i would want five sections in > the file: MAGIC, OPTIONS, METADATA, INDEX, DATA. In this layout, SXML > could work fine for DATA. The other sections require numbers.
Another useful information that is kept in the tree that is done from the parsed Texinfo in perl is a reference to other elements of the tree. For example, a @node has a reference to the following sectioning command and vice-versa, an @end has a reference to the @-command that begun the block, a @quotation with @author holds a list of references to those commands, while the @author has a reference to the @titlepage or @quotation. Would this be doable in an sexpr? As for the information that may be wanted in a renderer, looking at what is available now, there is the index informations (index entries, merged indices), informations on floats (for @listoffloats), informations on labels (@node, @anchor, @float with label) especially for internal @*ref, a possibility to access the tree of some @-commands, like @copying, @documentdescription, @top, @settitle..., and some misc informations, like the succession of @direntry and @dircategory, the input file name. -- Pat
