Newbie: Extracting links / iframes from HTML

[MarkusJ]

Hi, I am still learning PHP and given some HTML I am trying to extract all links and iframes from the HTML and append them to a different string.

 

I am still learning PHP so I am not sure if how I am checking that the returned array has values (isset) or if I should be appending strings together (.) is correct

 

The code that I have so far is

function GetLinksIFrames($content)
{
       $innerContent ='';
       $regex_pattern_links = "/<a href=\"(.*)\">(.*)<\/a>/";

	preg_match_all($regex_pattern_links,$content,$matches);

		for ($i = 0; $i < count($matches); $i++) 
		{
	  		if(isset($matches[0][$i]))// Is this correct?
			{
				$innerContent = $innerContent.$matches[0][$i]." "; // Is this how to append a result to an existing string?
			}
		}

	$regex_pattern_iframe = "/<iframe src=\"(.*)\">(.*)<\/iframe>/";

	preg_match_all($regex_pattern_iframe,$content,$matches);

		for ($i = 0; $i < count($matches); $i++) 
		{
			if(isset($matches[0][$i]))
			{
				$innerContent = $innerContent.$matches[0][$i]." ";
			}
		}

	return $innerContent;
}

Any help appreciated

 

Thanks
Mark

[Irate]

If you **want**, you can use JavaScript's DOM methods for that.

 

document.links and window.frames respectively do just that.

 

If you are looking for the PHP version, look up non-greedy repititions.

 

<a href="some/path/link.php">A link here.</a><a href="another/path/link.php">Another link here.</a>

 

The above line would be matched by your RegExp

[MarkusJ]

Thanks for the feedback

 

If I can ask as direct php question

if(isset($matches[0][$i]))// Is this correct?
            {
                $innerContent = $innerContent.$matches[0][$i]." "; // Is this how to append a result to an existing string?
            }

Is the above the best way to check for a null reference and to append a string to itself?

 

Thanks!

[.josh]

Technically you should check if $matches[0] exists before checking if $matches[0][$1] exists:

 

if(isset($matches[0])&&isset($matches[0][$i]))

But this is just to avoid a warning (assuming your error level is set to report warnings); the logic itself would have worked as-is.

 

And yes, that is how you append a string to another string (concatenation).

 

Sidenote: In your patterns, you use (.*) in several places. This is a greedy match, and it will yield unexpected results. You should use a non-greedy match instead: (.*?)

 

Also, I would like to point out that regex isn't really the best method for parsing html. For example, if the links and iframes have any other attributes, or use single quotes instead of double quotes, spaced differently, or any number of other things, your regex would fail, since it doesn't account for any of that. What you should instead use is a DOM parser.

[MarkusJ]

Thanks for the help