[SOLVED]perl regex matches twice in one line of text; i want first mat

brisbin33 · 2010-06-03 01:43:51

echo 'anything<h1>something</h1>anything<h1>somethingelse</h1>anything' | perl -pe 's/.*\<h1\>(.*)\<\/h1\>.*/\1/g'
somethingelse

i want "something" (i.e. the first match left-to-right). is there a regex that will accomplish this?

the end goal is use in php's preg_replace function, but it's easier to test/play with using the above.

Last edited by brisbin33 (2010-06-03 14:24:24)

Trent · 2010-06-03 02:50:18

Oops, was too quick to respond. I don't think I understand the question. Let me ponder for a while...

Last edited by Trent (2010-06-03 02:51:52)

Daenyth · 2010-06-03 03:01:56

Why are you parsing html with regex? Don't do that!

http://stackoverflow.com/questions/5907 … ml-why-not

Trent · 2010-06-03 12:31:26

You want a non-greedy qualifier (?) on the first .* And the /g is unnecessary if you want only the first match.

s/.*?\<h1\>(.*)\<\/h1\>.*/\1/

rson451 · 2010-06-03 12:46:05

This works, but Daenyth is right here..

echo 'anything<h1>something</h1>anything<h1>somethingelse</h1>anything' | perl -pe 's/.*?\<h1\>(.*?)\<\/h1\>.*/\1/'

brisbin33 · 2010-06-03 13:55:57

@Trent, i could only get it to work by adding a second ? inside the ( )

@rson, nicely done, see above .

@Daenyth, i'll be looking into DOMs and whatnot, but in my simple situation, this regex will work well.

Last edited by brisbin33 (2010-06-03 14:09:18)

Trent · 2010-06-03 17:41:33

Yep, that's right. I think I assumed it was already non-greedy. Despite the obvious lack of qualifier. I have no excuse.

brisbin33 · 2010-06-03 18:04:51

just as additional info for those that want it. i looked into PHP5's out of the box DOM stuff. it's quite easy to use. but you can't extract from commented tags (which i need).

http://simplehtmldom.sourceforge.net/manual.htm is a sweet drop-in script that works really well (i'll be using it in general going forward, Daenyth ).

i found it b/c their manual specifically mentions a special feature where $html->find('comment') will return all comments, within which i could then $comment->find('my custom tag').

sadly, it doesn't work. no output. i made a simple test script and emailed the writer of this software, we'll see if i'm dumb or it needs fixing.

cheers!

Arch Linux

#1 2010-06-03 01:43:51

[SOLVED]perl regex matches twice in one line of text; i want first mat

#2 2010-06-03 02:50:18

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#3 2010-06-03 03:01:56

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#4 2010-06-03 12:31:26

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#5 2010-06-03 12:46:05

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#6 2010-06-03 13:55:57

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#7 2010-06-03 17:41:33

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

#8 2010-06-03 18:04:51

Re: [SOLVED]perl regex matches twice in one line of text; i want first mat

Board footer