On Sat, Oct 26, 2002 at 01:18:07PM -0400, Steven Lembark wrote: > Take a look at Damian Conway's text formatter (Format::Simple > or something like that). It will format and re-order roman > numeral lists and has all the regexes for it hard-coded.
Excellent suggestion. Damian would appear to be the king of dynamically generated code and regexes. As I pointed out to someone else on this list (but forgot to cc: fwp), the ability to properly parse Roman Numbers is so context-sensitive that each regex would be different. However, it is certainly within the realm of possibility to dynamically generate a regular expression to avoid catching what your scan has determined to be actual English words (eg. since 'mix' is a verb, design the regular expression to avoid catching 'mix' in this sentence position.) Brian -- Brian Richardson Programmer/Analyst Cubik Software Design - "Of course it runs NetBSD!" brian at cubik dot ca | http://www.cubik.ca/