Serchilo talk:Command syntax

From Serchilo

i have a question about the regex syntax. i tried the following command, which didn't match:

   regexp   = ^sbb (\S*) (\S*)$
   syntax   = sbb {from} {to}
   url      = http://fahrplan.sbb.ch/bin/query.exe/d?S={$1|iso-8859-1}&Z={$2|iso-8859-1}&start=1

what kind of regex can be used?

The regex is working, but I think the command is blocked by the regular (i.e. without regular expressions =)) sbb commands. This is arguably a bug, as there is no one-argument sbb command. Kevang 02:42, 9 March 2008 (CET)

the following did work though:

   regexp   = ^sbb ([a-z]*) ([a-z]*)$
   syntax   = sbb {from} {to}
   url      = http://fahrplan.sbb.ch/bin/query.exe/d?S={$1|iso-8859-1}&Z={$2|iso-8859-1}&start=1

the problem seems to be coming from the \S caillou

Yes, please use [a-z] instead of \S. The regexp has to be according to the MySQL-Regexps. --Jorges 23:37, 9 March 2008 (CET)

Is it possible to have whitespaces in an argument convertet to a particular charakter before its handed over to the destination cgi? in Particular: the library of congress uses queries like this: http://www.loc.gov/fedsearch/metasearch/?cclquery={%item1}+{%item2}

obviously i could use n-number of query-item, but then the command user would need to seperate them with commata. instead I would like functonality similar to the google-commands where you just enter a list of terms seperated by whitespace. Unlike google the library of congress can't work with those whitespace-seperator so i am looking for a way to expand them.