Niklas Broberg wrote:
.. We came up with the idea of HaRP: Haskell Regular Patterns.
taking data and patterns as typed trees obviously is the right thing (tm). in principle, such patterns describe regular tree languages (if we disregard nested data types for the moment - they give CF languages) and they could be translated into finite tree automata, which leads to fast matching algorithms (but cost some preprocessing time). you want to always return exactly one match (or none)? this forces you to make ad-hoc decisions (longest/shortest match etc.) when defining the semantics. I think a lot of similar issues are being discussed w.r.t. XML processing. see http://citeseer.nj.nec.com/388107.html and many others. -- -- Johannes Waldmann, Tel/Fax: (0341) 3076 6479 / 6480 -- ------ http://www.imn.htwk-leipzig.de/~waldmann/ ---------