[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: parts of TREX for generic matching
I was talking to Uche Ogbuji today about RDF inference languages and the topic came up of how to express complex matches. It occurred to me that TREX provides a good framework for generic matching. If you look at the patterns: the following are specific to validating XML (matching requires a notion of attributes and child elements): - element - attribute - empty - anyString - string - data - mixed - group - interleave - oneOrMore - zeroOrMore whereas the following are generic matching constructs: - notAllowed - optional - choice - concur - grammar - ref (and we can add difference here) If one introduces the notion of an ordered sequence, one can move the following patterns across to being generic: - group - interleave - oneOrMore - zeroOrMore Furthermore, if conditions on string matching are relaxed (no pun intended) one can use the generic constructs plus the following to do string regular expressions: - anyString - string - data I just thought I'd throw these idea out to the TREX TC. If nothing else, I might write up a spec on generic matching constructs that borrows heavily from TREX. James
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC