G'day all. Quoting Branimir Maksimovic <[EMAIL PROTECTED]>:
> After seeing that your program is fastest (I've also tried one from > http://haskell.org/hawiki/RunTimeCompilation but perhaps I'm not > that good in converting to search replace?) You probably did it right, but you could post your version to the list if you want me to take a look. When I wrote the RunTimeCompilation code, it wasn't intended to be a shining example of efficiency, merely an illustration. Remember that it's doing TWO things: compiling the pattern to code, and then performing the search. The compilation phase is likely to be much slower than the search, so the speedup (if any!) would only be realised the SECOND time that you searched a string using the same pattern. (Assuming you re-used the compiled match code, of course!) Cheers, Andrew Bromage _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
