Comments on: Parsing a URL with a Regex http://lostechies.com/chadmyers/2010/11/20/parsing-a-url-with-a-regex/ Software development, testing, and techie life Thu, 08 Mar 2012 22:19:00 +0000 hourly 1 http://wordpress.org/?v=3.4.2 By: Speednet http://lostechies.com/chadmyers/2010/11/20/parsing-a-url-with-a-regex/#comment-1213 Speednet Sun, 21 Nov 2010 13:31:24 +0000 /blogs/chad_myers/archive/2010/11/19/parsing-a-url-with-a-regex.aspx#comment-1213 For some reason, John Gruber tries to match balanced parens in the URL. It's great that he correctly matches parens, but there is absolutely no requirement for parens to be "balanced" in a URL. e.g., this URL is fine: http://www.test.com/this(page.htm Also, there is totally no need to look for "www" or "www\d+". Lots of sites have dropped the "www". Better just to code it like this: (?:[-a-z0-9]{1,63}\.)+ Or, if your regex engine supports lookbacks, you can avoid the invalid syntax of a hyphen right before each dot: (?:[-a-z0-9]{1,63}(?<!-)\.)+ (Note: by placing the hyphen as the first char inside the brackets, we can avoid escaping it, making it more readable.) Another tip: top-level domains are currently up to 6 characters in length (i.e., ".museum"), so it should be changed to [a-z]{2,6}. -Todd For some reason, John Gruber tries to match balanced parens in the URL. It’s great that he correctly matches parens, but there is absolutely no requirement for parens to be “balanced” in a URL. e.g., this URL is fine: http://www.test.com/this(page.htm

Also, there is totally no need to look for “www” or “www\d+”. Lots of sites have dropped the “www”.

Better just to code it like this: (?:[-a-z0-9]{1,63}\.)+

Or, if your regex engine supports lookbacks, you can avoid the invalid syntax of a hyphen right before each dot: (?:[-a-z0-9]{1,63}(?

(Note: by placing the hyphen as the first char inside the brackets, we can avoid escaping it, making it more readable.)

Another tip: top-level domains are currently up to 6 characters in length (i.e., “.museum”), so it should be changed to [a-z]{2,6}.

-Todd

]]> By: bob http://lostechies.com/chadmyers/2010/11/20/parsing-a-url-with-a-regex/#comment-1212 bob Sat, 20 Nov 2010 10:41:18 +0000 /blogs/chad_myers/archive/2010/11/19/parsing-a-url-with-a-regex.aspx#comment-1212 @Aidin : ok thanks, got it ;) @Aidin : ok thanks, got it ;)

]]> By: Aidan http://lostechies.com/chadmyers/2010/11/20/parsing-a-url-with-a-regex/#comment-1211 Aidan Sat, 20 Nov 2010 02:27:46 +0000 /blogs/chad_myers/archive/2010/11/19/parsing-a-url-with-a-regex.aspx#comment-1211 @bob The problem is identifying where a URL occurs within an arbitrary string of text. The RFC regex breaks a *known* URL into its constituent parts. @bob The problem is identifying where a URL occurs within an arbitrary string of text. The RFC regex breaks a *known* URL into its constituent parts.

]]> By: bob http://lostechies.com/chadmyers/2010/11/20/parsing-a-url-with-a-regex/#comment-1210 bob Sat, 20 Nov 2010 01:18:21 +0000 /blogs/chad_myers/archive/2010/11/19/parsing-a-url-with-a-regex.aspx#comment-1210 the regex for parsing a url is in the rfc http://tools.ietf.org/html/rfc3986#page-50 or am I missing something? the regex for parsing a url is in the rfc http://tools.ietf.org/html/rfc3986#page-50 or am I missing something?

]]>