regex help

[skyline]

I'm trying to get both the link url and img url from this string:

<a href="linkurl"><img src="imgurl" alt="alttext" width="346" height="260" border="1"></a>

what would the regex be for this to get all links and their images on page? I would like to ignore the other attributes both infront and behind for the link and the image tags

[SKDevelopment]

You would need to use preg_match_all(). I am giving you a very simple working example. Probably you would need to make something a little bit more complicated.

<?php
// Sample string:
$str = <<< TEXT
<a href="linkurl1"><img src="imgurl1" alt="alttext" width="346" height="260" border="1"></a>
<a href="linkurl2"><img src="imgurl2" alt="alttext" width="346" height="260" border="1"></a>
<a href="linkurl3"><img src="imgurl3" alt="alttext" width="346" height="260" border="1"></a>
TEXT;
// Extracting URL's to the array $m
preg_match_all("/<a href=\"([^\"]*)\"><img src=\"([^\"]*)\".*?><\/a>/si",$str,$m);
// Displaying the result
echo '<pre>' . htmlspecialchars(print_r($m[1],true)) . '</pre>'; // link URL's
echo '<pre>' . htmlspecialchars(print_r($m[2],true)) . '</pre>'; // image URL's
?>

[BurakUeda]

Another simple example here:

$reg = "/(a +href ?= ?|img +src ?= ?)(\"|\')?(http:\/\/)?([\w-_]*\.?)*[\/\w\.?]*/i";

preg_match_all($reg, $string, $match);

foreach($match[0] as $key => $val){
    $urls[$key] = str_replace(array("=", "src", "a href", "img", " ", "'", "\""), "", $val);
}
echo "<pre>".htmlspecialchars(print_r($urls, true))."</pre>"; 

works for:

<a href="http://www.someurl-website.com/subfolder/index.php"><img src="http://domain.something.com/images/imgurl.gif" alt="alttext" width="346" height="260" border="1"></a>
<a href="www.someurl-website.com/subfolder/index.php"><img src="http://domain.something.com/images/imgurl.gif" alt="alttext" width="346" height="260" border="1"></a>
<a href="someurl-website.com/subfolder/index.php"><img src="http://something.com/images/imgurl.gif" alt="alttext" width="346" height="260" border="1"></a>
<a href="/subfolder/index.php"><img src="something.com/images/imgurl.gif" alt="alttext" width="346" height="260" border="1"></a>
<a href="index.php"><img src="images/imgurl.gif" alt="alttext" width="346" height="260" border="1"></a>

Won't work for:

<a target = "_blank" href = "linkhere"><img width = "" alt = "" src="linkhere" /></a>

link and img tags must begin with <a href and <img src

[skyline]

Thanks guys,

If i use eg this one:

"/<a href=\"([^\"]*)\"><img src=\"([^\"]*)\".*?><\/a>/si"

how can i make it case insensitive? Some of the links are eg <a HREF

[SKDevelopment]

This regexp is already case-insensitive because of i-modifier

"/<a href=\"([^\"]*)\"><img src=\"([^\"]*)\".*?><\/a>/si"

[skyline]

ah yes, actually the problem is when this example crops up:

<a HREF="imglink"><img name="" src="imgurl" width="140" height="140" border="0"></a>

basically i need the regex to ignore everything other than the src for both the a tag and the inbetween img tag

[SKDevelopment]

Try this (not checked):

"/<a [^>]*?href=\"([^\"]*)\"><img [^>]*?src=\"([^\"]*)\".*?><\/a>/si"

[skyline]

ok that was working brilliantly until i hit some examples where there was other stuff before the closing a tag:

eg:

<a HREF="link"><img src="image" width="360" height="270" border="0"><br>

                  </a>

in this case there is a <br> tag and also some character returns etc (presumabley /n/r ??)

so how can i make it so that anything else can appear after the close of the img tag, but before the close a tag?

[SKDevelopment]

E.g. you could try this:

"/<a [^>]*?href=\"([^\"]*)\"><img [^>]*?src=\"([^\"]*)\".*?>.*?<\/a>/si"

[skyline]

ok thanks

[Jake Arkinstall]

To be honest, when it comes to something like this where attributes can be in different orders, and tags can vary on the inside, shouldn't you be taking advantage of PHP's DOM capabilities?

[skyline]

To be honest, when it comes to something like this where attributes can be in different orders, and tags can vary on the inside, shouldn't you be taking advantage of PHP's DOM capabilities?

I did briefly look at that, but its not something i have any experience with so i thought i'd leave it! This seems to work at the moment at least...

[Jake Arkinstall]

Yeah, however if you do come across a different URL, you may run into problems.

[skyline]

Another issue:

sometimes there are spaces and carriage returns between the <a> and the <img tag...

eg:

			<A HREF="link">
				<IMG SRC="image" WIDTH=163 HEIGHT=62 BORDER=0 ALT=""></A>

How can i allow just /n and /s and /r between the opeing a and the img tag?

[SKDevelopment]

I think the simplest would be to use \s*

"/<a [^>]*?href=\"([^\"]*)\">\s*<img [^>]*?src=\"([^\"]*)\".*?>.*?<\/a>/si"

It would allow 0 or more of so called whitespace characters: \r, \n, \t, space, formfeed (and probably vertical tab) between <a> and <img> opening tags.