ANN: Regular Associations

Hi,

The idea is inspired by James Clark's approach to Clark's schema
location approach for nXML,
http://groups.yahoo.com/group/emacs-nxml-mode/message/259.

I was going to implement it for my needs, but realized that a more
consistent solution would be more appropriate. The original proposal
lists a number of features which can be used to determine a document's
type; however, the exact set of chosen features, while can be practical 
for some applications, can become a certain obstacle for others.

In fact, regular expressions can be used. For URIs, that is, for files' names,
there are string regular expressions, in any of the many available syntaxes;
for XML, that is, for files' contents, the regular expressions language is
Relax NG. 

One does not need to specify a complete syntax for either URIs or contents
to choose between document types:
- /.*\.rng/ is a pattern for the name of a file with Relax NG grammar inside;

-
    namespace rng = "http://relaxng.org/ns/structure/1.0"
    start = element rng:* {any}
    any = (element * {any}|attribute * {text}|text)*

  is a Relax NG pattern for the contents of a file with Relax NG grammar inside.

ARX, http://davidashen.net/rnv.html , is a tool that uses a list of regular
(string and XML) patterns to associate files with symbolic names (of 
grammars, stylesheets etc.). 

David Tolpin
http://davidashen.net/