perl regular expression

rbroen

Hi, I'm a newbie to this forum and switched from VBscript to PhP some six months ago.

PhP rocks! I really like the preg_match_all() function and have read alle the basic manuals, but i'm having problem building the expression for:

Er was een een <b>knappe</b> prins die op zoek was naar een <b><i>mooie</i> prinses</b> in een <b>donker</b> bos.

I would like to retreive the text between the bold tags. My expression is:

/<b>(.*)<\/b>/

But this results in:

array(2) {
  [0]=>
  array(1) {
    [0]=>
    string(93) "<b>knappe</b> prins die op zoek was naar een <b><i>mooie</i> prinses</b> in een <b>donker</b>"
  }
  [1]=>
  array(1) {
    [0]=>
    string(86) "knappe</b> prins die op zoek was naar een <b><i>mooie</i> prinses</b> in een <b>donker"
  }
}

instead of:

array(2) {
  [0]=>
  array(3) {
    [0]=>
    string(13) "<b>knappe</b>"
    [1]=>
    string(27) "<b><i>mooie</i> prinses</b>"
    [2]=>
    string(13) "<b>donker</b>"
  }
  [1]=>
  array(3) {
    [0]=>
    string(6) "knappe"
    [1]=>
    string(20) "<i>mooie</i> prinses"
    [2]=>
    string(6) "donker"
  }
}

Who can help me ajust this expression?

onyx

Ok from adapting some code on php.net

preg_match_all ("|<b>(.*)</b>|U", "<b>example:</b><div align=left><b>this</b> is a test</div> <b>hello</b>", $out, PREG_SET_ORDER);
print $out[0][1] . ", " . $out[1][1] . ", " . $out[2][1];

Take a look at:

http://uk2.php.net/manual/en/function.preg-match-all.php

rbroen

Thanks Sam, you are a big help!

It was the U that did it.

U = PCRE_UNGREEDY - This modifier inverts the "greediness" of the
quantifiers so that they are not greedy by default,
but become greedy if followed by "?". It is not
compatible with Perl. It can also be set by
a (?U) modifier setting within the pattern.

What's up with that 'not compatible with Perl' stuff?

I'm using the following expression, it will work with other tags than just "b". I prefer the / start-end. Do you always use | or just because it seemed logical here?

/<([\w]+)>(.*)<\/\1>/U[code]

BSTRhino

I would have used "|<b>(.*?)</b>|i" as my pattern.

The 'i' means case insensitive, and the question mark makes the * metacharacter non-greedy, so it will only take as little as possible to generate a match.