3

I am trying to write a regex which match the first ending form tag.

  <form.*name="loginForm".*>[^~]*</form>

The above regex matches till the second from ends i.e till line 8. but I want a regex that matches the immediate ending from tag in the below example it should match line 5. 

<html>
<body>
<form method = "post" name="loginForm" >
<input type="text" name="userName"/>
</form>
<form method = "post" name="signupForm" >
<input type="text" name="userName"/>
</form>
</body>
</html>
  • Thanks for the quick response and suggestion. I found my answer. Special thanks to Guffa. –  Sep 22 '09 at 14:57

3 Answers3

11

Just make the pattern non-greedy so that it matches the smallest possible amount of characters instead of the largest possible:

<form[^>]*name="loginForm"[^>]*>[^~]*?</form>

Edit:
Changed .* to [^>]* in the form tag, so that it doesn't match outside the tag.

Guffa
  • 687,336
  • 108
  • 737
  • 1,005
  • 4
    Will fail for nested tags. Not sure that forms will ever be nested, but using a regex to parse HTML is still a bad idea, even if it works in some select cases. – Chris Lutz Sep 22 '09 at 06:07
  • @Guffa: Then you should make all quantifiers non-greedy. – Gumbo Sep 22 '09 at 06:15
  • The question didn't ask for the matching form tag, it asked for the first, which might not be the best thing to do, but this is a valid and useful trick sometimes. – Kevin Peterson Sep 22 '09 at 06:17
  • @Chris: Form tags can't be nested. – Guffa Sep 22 '09 at 07:01
  • @Kevin: The first ending tag is the matching tag. – Guffa Sep 22 '09 at 07:04
  • @Gumbo: Good point, however they don't have to be non-greedy, they just have to be kept inside the tag. – Guffa Sep 22 '09 at 07:22
  • @Gumbo: Actually, if the expression before the name attribute would be just non-greedy, it fails if the loginForm is not the first form in the code... – Guffa Sep 22 '09 at 07:37
2

Use a real parser like DOMDocument, SimpleXML or SimpleHTMLDOM. Regular expressions are not suitable for parsing non-regular languages like HTML.

Gumbo
  • 643,351
  • 109
  • 780
  • 844
  • 3
    +1 Was going to say it, but I had to make the question presentable first. Let's link to the explanation of why (http://stackoverflow.com/questions/701166/can-you-provide-some-examples-of-why-it-is-hard-to-parse-xml-and-html-with-a-rege) and the example of parsers (http://stackoverflow.com/questions/773340/can-you-provide-an-example-of-parsing-html-with-your-favorite-parser) for completeness. – Chris Lutz Sep 22 '09 at 05:42
  • 1
    A regular expression works just fine for parsing a string like this. There is definitely no need for it to be a regular language to be parsed by a regular expression. – Guffa Sep 22 '09 at 09:53
2

You should NOT use regular expressions, but parse it with DOM:

Javascript:

var forms = document.getElementsByTagName('form');
forms[0] // is the first form element.

PHP:

$dom = new DOMDocument();
$dom->loadHTML( $html );
$forms = $dom->getElementsByTagName('form');
$first = $forms->item(0); // reference to first form

You can use minidom and ElementTree for Python.

meder omuraliev
  • 183,342
  • 71
  • 393
  • 434