Given the following HTML content (limited to the absolute minimum I require):
How would I be able to extract Page Title using Regex?
-
Are you only grabbing titles or are you going to be parsing out more from the document? If so, use an HTML parser.David B– David B2012-09-10 16:00:34 +00:00Commented Sep 10, 2012 at 16:00
-
1You may look at this answerLuca Martini– Luca Martini2012-09-10 16:02:05 +00:00Commented Sep 10, 2012 at 16:02
-
Wow :O Happened to have missed that. So should I use an HTML parser, and if so, which one?Daan– Daan2012-09-10 16:07:27 +00:00Commented Sep 10, 2012 at 16:07
-
It depends on what language you want to use. The main reason for an HTML parser is the malformed nature of HTML/XML.David B– David B2012-09-10 16:09:37 +00:00Commented Sep 10, 2012 at 16:09
-
1The language is C# (if that's what you mean). I still feel that an HTML parser is overkill in my situation. What if we assume that the pattern is always exactly this way, can't I better use regex?Daan– Daan2012-09-10 16:15:29 +00:00Commented Sep 10, 2012 at 16:15
Add a comment
|
1 Answer
As others have commented, regular expressions may not be suitable for a bullet-proof method. E.g. using regex, it would be difficult to check if the <title> tag were part of a quoted string within the HTML. That's a recurring response on StackOverflow for questions like this. But personally, I think you've got a point that a parser would be overkill for such a simple extraction. If you're looking for a method that works most of the time, one of the following should surfice.
Option 1: Lookbehind / lookahead
(?<=<title[\s\n]*>[\s\n]*)(.(?![\s\n]*</title[\s\n]*>))*
This uses lookbehind and lookahead for the tags - .NET has a sophisticated regex engine that allows for infinite repetition so you can even check for whitespace/return characters between the tag name and end brace (see this answer).
Option 2: Capturing group
<title[\s\n]*>[\s\n]*(.*)[\s\n]*</title[\s\n]*>
Similar but slightly simpler - the whole regex match includes the start and end tags. The first (and only) capturing group (.*) captures the bit that is of interest in between.
Visualisation:
