more regex fun

OK, so I have a DCR with a replicant of N number of fields and I want to collect the ones between the IntroText.



 while ($txt =~ m|\G<IntroText>(.*?)</IntroText>|img) {$outtxt.=" \n IntroText ".$1;}

  $txt =~ m|<IntroText>(.*?)</IntroText>|img;
  $outtxt.=" \n IntroText 1 ".$1;

If I feed it values of rep1 rep2 rep3 inside the desired tags (over many lines) the first regex returns nothing. The second gives me 3 rep1.

Any ideas ?

Find more posts tagged with

Comments

ISCBorisB

You may want to try

while ($txt =~ m/<IntroText>(.*?)<\/IntroText>/cgims) {
   print "$1\n";
}

nipper

So that worked, I thought you needed the \G flag.

Thanks

ISCBorisB

So that worked, I thought you needed the \G flag.

Thanks

It is possible to make it work with \G. Much easier with 'c' modifier though ('c' continues the previous match if used with 'g' modifier)

Adam Stoller

It is possible to make it work with \G. Much easier with 'c' modifier though ('c' continues the previous match if used with 'g' modifier)

I think the major difference was the addition of the 's' modifier - which according to Perl Best Practices (Damian Conway) you should almost always use the 'ms' modifiers together.

Adam Stoller

Also - have you considered using TeamSite::XMLnode and the get_inner_xml() method?

ISCBorisB

I think the major difference was the addition of the 's' modifier - which according to Perl Best Practices (Damian Conway) you should almost always use the 'ms' modifiers together.

Granted, for simple testing 'gims' will work just like 'cgims'. I like to add 'c' though to avoid 'beginning of the string' resetting.
If that happens and you have a match - it's an infinite loop that is very hard to find

nipper

Also - have you considered using TeamSite::XMLnode and the get_inner_xml() method?

Normally I do. However in this case I am writing a MetaTagger preprocessor to parse out 80% of a DCR and use the text fields, called description or longdescription or introtext or whatever. That way I can look at all of them quickly in a regex. After some help of course. :-)

Thanks

ANdy