Image link counting & listing from a remote page

Ozz

I need to count and list image links from a remote page.
I tried this code:

$url = "http://www.siteurl.com/page.html";
$file = fopen ("$url", "r");
if (!($file)) 
{
echo "Not exist!";
exit;
}
while (!feof ($file)) {
   $line = fgets ($file, 1024);
   if (eregi ("HREF=\"(.*).jpg\">", $line, $out)) {
       $image = $out[1].".jpg";
       $pos = strpos($image, "http://");
if ($pos === false) {
   $image_url = dirname($url)."/".$image;
}

echo $image_url;
}
}

But it's not working stable on different sites especially only works if the image link codes are on one line.

Can anyone suggest any code for listing image links?

Thanks in advance.

bpat1434

<?php

$site = 'http://photobucket.com/images/recent/'; // The web page you want to link images from

$html = file_get_contents($site); // Try this first, much easier

if(!$html || empty($html)) {
  $html = '';
  $fp = fopen($site, "r");
  while(!feof($fp)) {
    $html .= fgets($fp, 1024);
  }
  fclose($fp);
}

// Now, $html holds all the html on that page.  Now we'll use a regular expression
// and get all our images out
$regex = '/<img.*?src="(.*?)\.jpg".*?>/i';
preg_match_all($regex, $html, $matches);
if(!$matches || empty($matches))
  die('An error occured with the regex.');
else {
  $base_site = substr($site, 0, (strrpos($site, '/'))); // Makes $base_site = 'http://www.site.com/'
  array_shift($matches);
  foreach($matches[0] AS $key=>$img) {
	$http = (substr($img, 0, 7) == 'http://') ? true : false;
    if(!$http) {
      echo '<img src="'.$base_site.'/'.$img.'.jpg" />';
    }
    else
      echo '<img src="'.$img.'.jpg" />';
  }
}

?>

That seems to work.

Ozz

Thanks it works just as what I need.

bpat1434

Good, glad I could help. But do you understand how/why it works?

Ozz

Yes because your code gets whole page with file_get_contents and I'm really suck with reg-ex and also with a help of preg_match_all.

bpat1434

The regex looks for any <img> tag and extracts anything inside of the "src" attribute. The ".*?" means any character (non-greedy). Non-greedy meaning that (in our case) when we get to a ".jpg", that should not be added as part of the match, but rather marks the end of it. And then we allow any text character (once again, non-greedy) until the closing of the <img> tag. The modifier "i" means that it's case insensitive, so it would match
<img src="">
<IMG src="">
<IMG SRC="">
Or any variation upon that.

The preg_match_all function is easier to use than eregi, sometimes faster, and much more powerful. It uses PERL as it's base for the Regular Expression (Perl RegEx). You can have a more intricate regex, and get faster results.

When the array comes back, it comes back with array[0][1] being filled with the entire matching text (so all the <img> tags that were matched in one string) and array[1][n] (where "n" is a number between 0 and the number of matches) contains each individual match (so each url in the src attribute).

Hope that helps explain the stuff. And as for learning regex, this tutorial is a good one, I used it a lot when learning. I also would use it, post here, and have Weedpacket or someone show me a better way (hence the non-greedy stuff). So it just takes time and practice.

Ozz

Great explanation bpat1434. I'll work on that tutorial too.
It's really hard to find a real helper in these 0-1's world.
Thank you very much indeed.

MarkR

I thought I should mention that regular expressions cannot be used to correctly parse HTML.

Regular expressions aren't powerful enough to parse HTML. They won't work in the general case.

Use DOMDocument::loadHTML instead, which will correctly build a DOM tree from the document. Then interrogate the DOM to find the img elements you're interested in. This has the advantage of actually parsing the HTML rather than using some regular expressions which may or may not match some part of a HTML page which might or might not be an image element.

Bear in mind that parsing HTML is difficult:

HTML comments exist and should be ignored even if they contain valid-looking tags
attributes can be enclosed in double OR single quotes, or indeed none
It's legal for attributes to contain HTML tags themselves etc ( <a name="<img src='qwer.jpg'"> is legal)
Attribute values can contain entities etc, which need to be interpreted correctly.
Script and style elements need to be treated specially
Plus many other idiosyncrasies

Mark

bpat1434

I thought I should mention that regular expressions cannot be used to correctly parse HTML.

I don't disagree, but I would say it could be better said if you said:
RegEx should not be used to parse HTML.

RegEx won't work 100% of the time; however, it can work for simple things, such as this. If there were a better way (RSS / XML) I'd have gone that route; however, there really isn't, so this is what we're stuck with.

chewan

bpat1434 wrote:
Good, glad I could help. But do you understand how/why it works?

How could you help me with this please?

these messages are from savedata.php I have my form in another page (adduser.php)

Notice: Undefined index: username in C:\EasyPHP 2.0b1\www\database\addUser+login\savedata.php on line 16

Notice: Undefined index: password in C:\EasyPHP 2.0b1\www\database\addUser+login\savedata.php on line 17

thanks in advance

bpat1434

Sure, but you're going to have to give me some reference as to what you're talking about. The "Undefined index" error means that you're referencing some array in savedata.php that doesn't have the indes "username" or "password" set yet. So look in savedata.php for something like:

$some_var['username']

Around line 16, and on line 17 something like:

$some_var['password']

But, without code... no-one can help. And if this isn't really part of this topic, please create your own thread.