Regular Expression Experts?

[MelBridgeman]

Hey anyone in tune with regular expressions?

I am trying to replace the content attribute of HTML meta tags..

The issue is is that meta tags isnt closed by </meta> and the xpression i have created is actually working, but replacing the entire contents past the meta tag

I am using PHP and the eregi_replace function, for the title tags I am using
<title[^>]*>.*</title[^>]*> to replace the contents of the tags and it works great....Anyone have a idea how to parse out tags that don't have a terminate tag?

I am not a regular expression guru, just trying to go off examples on the net.

[MelBridgeman]

[nfotiu]

What have you tried?

<meta[^>]*> should work, no?

[MelBridgeman]

It does not work, i think it is because some of my meta tags are structured like this

<meta name="keywords" content="<!--#include virtual="/includes/filename.html"-->">

basically i want to completely replace this entire tag with something else

sometimes they are structured like this

<meta name="keywords" content="<!--#include virtual="/includes/filename.html"-->"/>

the combination i tried similar to what you posted actually changed it correctly, but replaced everything after it too...

[c.t.ner]

Sorry, I'm not familiar with regular expression? Do you mean Expression Engine? (Just curious)

[missdpuck]

My regular expressions are "Don't call me babe" and the Bronx Cheer. Sorry.

[MelBridgeman]

Originally Posted by c.t.ner

Sorry, I'm not familiar with regular expression? Do you mean Expression Engine? (Just curious)

http://en.wikipedia.org/wiki/Regular_expression

[photon]

Sorry wish I knew something about them. My interactions with them usually entail a furious google search until I find what I need which I promptly forget.

[MelBridgeman]

Ya my problem is that i am expecting google to find me the answer, instead of using what i find and applying it...which is what i should be doing

[llama64]

Originally Posted by MelBridgeman

Ya my problem is that i am expecting google to find me the answer, instead of using what i find and applying it...which is what i should be doing

Isn't that what all programmers do?



Regular Expressions make my brain hurt. If you do find a solution though, could you post it? I'd be curious to see why it isn't working.

[maverickstruth]

Try something like... this?

<([A-Z][A-Z0-9]*)\b[^>]*>(.*?)</\1>

(Got it from http://www.regular-expressions.info/examples.html, but basically all it's doing is finding a tag, then matching everything between that tag and it's closing tag using the \1 backreference)

Gah, just reread. You've got that part working already.

What about... (warning, this is really messy and probably won't work)

<([A-Z][A-Z0-9]*)(\b[^>]*)?(<!--(\b[^>]*)?>)?(\b[^>]*)?>

< to capture the beginning of the tag
([A-Z][A-Z0-9]*) to capture the tag itself (note, you may need to include lower case here)
(\b[^>]*)? to capture anything that does not include a >
(<!--(\b[^>]*)?>)? to capture zero or more words after <!-- but before >, which may never occur
(\b[^>]*)? to capture zero or more words after the html comment
> to capture the end of the tag

[sclitheroe]

Would a tool like this help: http://gskinner.com/RegExr/

Personally I would write a small AWK script to parse the text. AWK is a fantastic tool to have in your arsenal for when you need to do text manipulation. It's less obtuse than Perl, but still really powerful for working with text.

[MelBridgeman]

Originally Posted by sclitheroe

Would a tool like this help: http://gskinner.com/RegExr/

Personally I would write a small AWK script to parse the text. AWK is a fantastic tool to have in your arsenal for when you need to do text manipulation. It's less obtuse than Perl, but still really powerful for working with text.

nice link man
well try out some suggestions

[Shazam]

Try

<meta\b[^>]*>(.*?)[^/]>

Should match all meta tags that aren't closed XHTML style, but not ones that are.

[nfotiu]

If the meta tag is always one line, and the only thing on the line, this would work:

<meta[^$]*>$

Or maybe alter a little to account for leading or trailing white space.

[Shazam]

Originally Posted by nfotiu

If the meta tag is always one line, and the only thing on the line, this would work:

<meta[^$]*>$

Or maybe alter a little to account for leading or trailing white space.

This works better than mine. Change to

<meta[^$]*[^/]>$

If you don't want to catch valid meta tags.