Pattern matching: find the first jpg or png image in html file

[bipster]

I wrote the following code to extract the first jpg or png image url from an HTML file:

$img="";
$pattern1 = '#<img\s+.+?\s*/?>#ise'; 
preg_match($pattern1, $content, $match1);
foreach($match1 as $i){
  $img=$i;
}         

$url="";
$pattern2 = '#http://[a-z\d&#231;&#241;_(),.%&\-+/\s]+(png|jpe?g)#i';
preg_match($pattern2, $img, $match2);
foreach($match2 as $i){
  if (strncmp("http://",$i,7)==0 && strlen($i)>7) {
    $url=$i;
  }
}         

As you can see I'm a newbie at regular expressions. Is there a simpler way to do the same?

Thanks in advance

[TeraTask]

I'd say to use strpos to find the <img start. Then use it again to find src= and extract to the first white space or closing quote if it exists at the start (again using strpos). strpos is faster than preg_match. stripos can be used too if you have PHP 5 and want a case-insensitive match.

Take a look:

$img_start = stripos($content,'<img');
if ($img_start !== false) {
  $src_start = stripos($content,'src=',$img_start);
  if ($src_start !== false) {
    //Determine if the character after the = is a single or double quote
    $src_delimiter = ' ';
    if ($content[$src_start + 5] == '"') {
      $src_delimiter = '"';
    } else if ($content[$src_start + 5] == "'") {
      $src_delimiter = "'";
    }
    $url_end = strpos($content,$src_delimiter,$src_start + 5);
    if ($url_end !== false) {
      $url = substr($content,$src_start + 5, $url_end - $src_start - 6);
      echo 'Your First Image: '.$url;
    } else {
      //Unable to find end of URL.
    }
  } else {
    //Could not find src attribute
  }
} else {
  //Image not found
}

[Weedpacket]

It would be a lot simpler if XPath had an "ends-with" function to go with "starts-with"...

$doc = new DOMDocument;

$doc->loadHTML($variable_containing_HTML_document);

$xpath = new DOMXPath($doc);
$query = "//img[
	substring(@src, string-length(@src)-3)='.jpg' or
	substring(@src, string-length(@src)-4)='.jpeg' or
	substring(@src, string-length(@src)-3)='.png'
][1]/@src";

$entries = $xpath->query($query);

foreach ($entries as $entry) {
    echo $entry->value,"\n";
}

[TeraTask]

Would XPath require the parsing of the whole document into its structure before being able to execute such a query?

[Weedpacket]

Yes. Does yours take into account the possibility that there might be whitespace surrounding the '=' between the attribute name and its value, and does it find the first "jpeg or png"?

[TeraTask]

I do not understand the hostility. I did not know the answer to my question. I guessed, but I have never used that class. It was a sincere question -- thank you for the answer.

Mine does have the limitations you rhetorically asked.

As a good friend of mine likes to say, "Remember to keep smiling"

[TeraTask]

Oh, one more -- again, sincere -- question: Does XPath fail if the document doesn't conform to HTML? XHTML?

[TeraTask]

I did some research and found http://us2.php.net/manual/en/domdocument.loadhtml.php which says

Unlike loading XML, HTML does not have to be well-formed to load.

So, there's the answer. 🙂

[Weedpacket]

Oh, well, I ditched the XPath anyway. That query was too ugly for me to stomach.

$first_img = null;
foreach($doc->getElementsByTagName('img') as $img)
{
	if(preg_match('/\.(jpe?|pn)g$/i', $src = $img->getAttribute('src')))
	{
		$first_img = $src;
		break;
	}
}
echo $first_img;