Regex driving me crazy too greedy

xpro · Jul 3, 2010

Hello

I'm trying to extract some data from a webpage and it should create 2 separate matches, but the regex keeps on going and it makes it as one match. What can I do so that it can stop at the first "</from>" it finds?

Best Regards!

The content

Code:

<messageInfo mid="1_449_ALlVimIAABd1TC5hsgc16zIYJFY" toEmail="Idelle" subject="Re: free ipod" mimeType="multipart/alternative" xapparentlyto="[email protected]" receivedDate="1278108078" size="7408"><flags isReplied="0" isFlagged="0" isRead="1" isDraft="0" isForwarded="0" isHam="0" isSpam="0" hasAttachment="0" inAddressBook="0" isRecent="1"/><from><name>LaVonda Bland</name><email>[email protected]</email></from><inboxservices><name>Retro</name><value>Y</value></inboxservices><inboxservices><name>SgrnP</name><value>N</value></inboxservices><inboxservices><name>d_t</name><value>1278108081</value></inboxservices><inboxservices><name>s_ip</name><value>74.6.228.93</value></inboxservices><inboxservices><name>showStationery=</name><value/></inboxservices><inboxservices><name>url</name><value>craigslist.org,mailto:[email protected],yahoo.com,mailto:[email protected]</value></inboxservices></messageInfo><messageInfo mid="1_22_ALVVimIAABo+TC1qtAhqDS72Xas" toEmail="[email protected]" subject="Welcome to Yahoo!" mimeType="text/html" receivedDate="1278044852" size="720"><flags isReplied="0" isFlagged="0" isRead="0" isDraft="0" isForwarded="0" isHam="0" isSpam="0" hasAttachment="0" inAddressBook="0" isRecent="1"/><from><name>Yahoo!</name><email>[email protected]</email></from>

My regex

Code:

<messageInfo mid=\"(.*?)\" toEmail=\".*?\" subject=\"(.*?)\" mimeType=\".*?\" xapparentlyto=\".*?\" receivedDate=\".*?\" size=\".*?\"><flags isReplied=\"0\" isFlagged=\"0\" isRead=\"0\" isDraft=\"0\" isForwarded=\"0\" isHam=\"0\" isSpam=\"0\" hasAttachment=\"0\" inAddressBook=\"0\" isRecent=\"1\"/><from><name>(.*?)</name><email>(.*?)</email></from>

madblacker · Jul 4, 2010

I had these same types of problems before.. basically, regex isn't ideal for parsing a web page, you need to parse it using some sort of HTML parser and then just use regex to parse whats left after the initial parse... it depends on what language you are using as to which parser to use..

voyevoda · Jul 9, 2010

I found the problem. You're trying to use a regular expression to parse XML! This is bad.

Your regular expression isn't even valid. You aren't escaping the forward slashes.

http://rubular.com/r/TTeyFAqD5y

The matches look correct to me.

bhmailer · Aug 21, 2010

Try this tool "The Regex Coach"; lets you test your regular expressions easily

anty · Aug 21, 2010

you specified

Code:

isRead="0"

but in your sample there's only code with isRead="1" to match. That's why there's no output.

The correct way would be to correct the content of the website to make it valid XML, then use XPath to access the data you want.

I usually catch characters until a character occurs that is now allowed, which is really easy: They either can't have '"' or '<' characters in HTML code.

Here's the code for your example with my method (isRead is still set to "0"):

Code:

<messageInfo mid=\"([^\"]*)\" toEmail=\"[^\"]*\" subject=\"([^\"]*)\" mimeType=\"[^\"]*\" xapparentlyto=\"[^\"]*\" receivedDate=\"[^\"]*\" size=\"[^\"]*\"><flags isReplied=\"0\" isFlagged=\"0\" isRead=\"0\" isDraft=\"0\" isForwarded=\"0\" isHam=\"0\" isSpam=\"0\" hasAttachment=\"0\" inAddressBook=\"0\" isRecent=\"1\"/><from><name>([^<]*)</name><email>([^<]*)</email></from>

Regex driving me crazy too greedy

xpro

Regular Member

madblacker

Regular Member

voyevoda

Regular Member

bhmailer

Newbie

anty

Newbie

Main Menu

Marketplace

Making Money

BlackHat World