Lists Home |
Date Index |
I think the very fact that relax uses a non-xml lexical structure, and the
obvious lexical clumsiness of XSLT (of which its xml infoset makes great
sense), identifies an opportunity to improve the lexical structure of XML.
Historically, I doubt XML was intended by its designers to represent
schemas, programming language, etc or really intended for users to spend
much time editing with an editor or looking at as a text file. But we all
do these things.
Maybe XML 1.3 should acknowledge XML use is larger than imagined, and define
an alternative (preferred) lexical syntax.
<A v="3" j="8" >
could be more succiently serialized into a more human and machine readable
format. Haskell and python are excellent examples of making code more
indentation instead of seperator tokens.
A b=3 j=8
or if you must have delimiters,
(K (A b=3 j=8 (C) =Cool))
This does mean that every application might need two parsers.