Fredrik Lundh wrote: > same benchmark as above, on slightly slower hardware.
I've posted updated benchmarks for a variety of event-driven/incremental parsers here: http://online.effbot.org/2005_01_01_archive.htm#20050122 Amara is surprisingly slow; I haven't looked at the code, but it must surely be possibly to process more than 25k per second on a 3GHz PC with that interface... </F> _______________________________________________ XML-SIG maillist - XML-SIG@python.org http://mail.python.org/mailman/listinfo/xml-sig