"Eric M. Ludlam" <eric@...> writes:
>>Great! Maybe some part of your work could be used to provide Semantic
>>facilities for an 'xpath-mode'? What do you think? Eric?
> [ ... ]
> That certainly makes sense. An xpath mode to bind the grammar against
> would be a good thing, and the editing and completion help such a
> parser can provide would be useful to all xpath authors.
Hm, I still don't understand how this will work. There are no and
there probably never will be any XPATH files. XPATH is something like
an URL. There is no URL mode. URLs appear in HTML files, though.
XPATHs appear in XML (and XSL, which is also XML) files. Thus, the
major mode of such files will be xml-mode, sgml-mode, html-mode,
xhtml-mode, or something along these lines. If you really want to
support XPATH in XML documents, for example, it seems to me that you
need to add the XPATH BNF to a more general XSL BNF, and redo the
entire thing. Since the two of you are so enthusiastic, however, I
get the feeling that I am missing something. What is it? Some hidden
feature of Semantic which allows you to mix and match major modes?
Something even more powerful? :)