Talk:Execute a Markov algorithm: Difference between revisions

Line 152:
: You'd think BNF would consistently be either minimal or greedy, but I couldn't find anything with Wikipedia or Google. I guess we just shouldn't consider that case. There are a number of related flaws in the ruleset syntax; there's no way to end a pattern with whitespace, for example. —[[User:Underscore|Underscore]] ([[User talk:Underscore|Talk]]) 17:56, 20 December 2009 (UTC)
:: I wouldn't call this a “production” syntax. One could upgrade it by putting double quotes around each half (making it more like the syntax used on wikipedia) but then you'd have to worry about how to embed a double quote in the string. There's also no way to specify a rule that only matches at the start and/or the end of the string. All in all, a mess. But there's enough there to show of the general power of this class of string replacement algorithm, and none of the examples need the problematic sequence, so perhaps it is best to just leave it ambiguous. (I just don't view the code to parse the ruleset as being the most important part of this task.) –[[User:Dkf|Donal Fellows]] 07:51, 21 December 2009 (UTC)
 
::: +1 on leaving as-is. --[[User:Paddy3118|Paddy3118]] 08:47, 21 December 2009 (UTC)
 
== Task addition ==
Anonymous user