I'm using this regex to find <script> tags:
<script (.|\n)*>(.|\n)*?</script>
The problem is, it matches the ENTIRE string below, not just each tag separately:
<script src="crap2.js"></script><script src="crap2.js"></script>
views:
905answers:
7
+10
A:
I don't think anything else needs to be said other than http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454#1732454.
JSBangs
2009-11-17 17:49:39
That is probably the best answer I've seen to any question!
Andy
2009-11-17 17:54:55
That’s rather a comment than an answer.
Gumbo
2009-11-17 17:58:23
Should be wiki.
Michael Myers
2009-11-17 18:01:10
This is a terrible answer. Look, I'm not trying to use regex to parse <b>XHTML</b>. I'm trying to match the <b>string</b> <script ...></script>. That is perfectly within the capabilities of regex.
JamesBrownIsDead
2009-11-17 18:03:41
JamesBrownIsDead, except that you need to care for case, whitespaces, HTML comments, strings inside embedded scripts, `<pre>` regions... Parsing HTML is a solved problem.
Svante
2009-11-17 18:19:48
Again, I'm not parsing HTML.
JamesBrownIsDead
2009-11-17 18:28:51
You *are* parsing HTML. If you weren't, there wouldn't be <script> tags in it.
Carl Smotricz
2009-11-17 19:41:22
+8
A:
You really would be better off using the DOM to process HTML for this reason and all sorts of others.
Andy
2009-11-17 17:50:02
If you're not processing HTML, why did you tag your question as HTML-related?
TrueWill
2009-11-17 18:21:30
+4
A:
change your first * to *?
This is the non-greedy 'match all', so it will match the smallest set of characters before the next '>'.
TheSean
2009-11-17 17:50:45
while i agree with JS Bangs' link, im pretty sure this will fix his problem
Galen
2009-11-17 17:57:15
If someone comes to a gunfight with a dull knife, will sharpening it fix his problem?
Svante
2009-11-17 18:05:19
@TheSean: And I guess with "bullets", you mean things like javascript strings containing '</script>'? Basically, you are *assuming* there are no bullets. But if you value your life: Run if you see a gun pointed at you!
soulmerge
2009-11-18 08:12:27
A:
try to exclude any '<' from the content
<script (.|\n)*>(.|\n|[^<])*?</script>
Pierre
2009-11-17 17:50:45
Even if it's technically not valid valid HTML, people often write code like: `<script>if(a < b) { /* code */ }</script>`
intgr
2009-11-17 17:52:50
You're not excluding `<` from the content with `(.|\n|[^<])*?`. The negated character class will never be reached when an occurrence of a `<` is stumbled upon since the `.` meta character already matches it. In fact, the only character will be `\r` (carriage feed) that `[^<]` is going to match.
Bart Kiers
2009-11-17 18:20:14
A:
<script[\s\S]*?>[\s\S]*?</script>
This matches most common situations, but it's very important to consider JS Bangs answer.
Rubens Farias
2009-11-17 17:53:09
+2
A:
I'll keep posting links to my previous answers until this question type has been wiped from this planet's surface (hopefully in 10 years or so): Don't user regular expressions for irregular languages like html or xml. Use a parser instead.
soulmerge
2009-11-17 17:55:04
Any regular expression you create will match a closing script tag in your javascript, for example, so: Yes, you *are* parsing a language.
soulmerge
2009-11-18 04:49:54
Another approach: You are parsing XML, which *is* a language. (or a sub-set of XML - XML documents must have a single root node, which your string doesn't)
soulmerge
2009-11-18 04:51:06
+7
A:
Also see this week's Coding Horror: Parsing Html The Cthulhu Way, inspired by the epic answer by @bobince that @JS Bangs links to.
Bill Karwin
2009-11-17 17:56:34