From: Hans Aberg (haberg@math.su.se)
Date: Wed Oct 10 2007 - 03:45:50 CDT
On 10 Oct 2007, at 06:30, Philippe Verdy wrote:
> Once again, the DFA argument does not matter at all. It absolutely
> does not
> influence the language or its algebra. But you are still missing
> the point:
> the order of matches DOES matter in all practical uses: whatever a
> regexp
> will match, according to its algebra, this will be according to it
> only a
> unordered set, that the client application will process in an ordered
> fashion;
You need to define what grammars you algebra elements should produce;
then the language complement exists, and if the grammar is RE, and RE
for that complement language exists.
One can define different symbols for different types of matches,
which then will correspond to different grammar constructs. I am not
interested in that part.
Hans Åberg
This archive was generated by hypermail 2.1.5 : Wed Oct 10 2007 - 03:48:26 CDT