create file (image) based on header

[christiang]

Hi guys, I'd like to use Alexia thumbnail service. It's an interesting
service, you pass an url to its cgi and you get a thumbnail of that
url. Of course they don't have the screenshots of all the websites so I
have to check if the image returned is fine or not. This is not that
problematic, I can read the headers and understand if it is a jpg or a
gif (the error one). The problem that I have is that sometimes although
they have the screenshot in the database they return the error one. For
different reasons I thought to backup once on a while the thumbs that I
need.

So the procedure should be something like this:

- query the cgi
- check if it is a correct image
- if it is, backup it

I'm having problem on backup it basically. To read and check the header
I do:

$a_url = parse_url($url);
if (!isset($a_url['port'])) $a_url['port'] = 80;
$errno = 0;
$errstr = '';
$timeout = 30;
if(isset($a_url['host']) &&
$a_url['host']!=gethostbyname($a_url['host'])){
$fid = fsockopen($a_url['host'], $a_url['port'], $errno, $errstr,
$timeout);
if (!$fid) return false;
$page = isset($a_url['path'])  ?$a_url['path']:'';
$page .= isset($a_url['query'])?'?'.$a_url['query']:'';
fputs($fid, 'HEAD '.$page.' HTTP/1.0'."\r\n".'Host:
'.$a_url['host']."\r\n\r\n");
$head = fread($fid, 4096);
if(!preg_match('[Content-Type: image/gif]', $head) &&
preg_match('#^HTTP/.*\s+[200|302]+\s#i', $head)){
$foundThumb = true;
// here I should to create the file using the $fid resource
}
fclose($fid);
}
 

Since sometimes the service fails I have to use the same stream I check
when creating the new file and I tried in many way to read the bites
and "fput" them in the new file but without success. How could I reuse
$fid to create the new file? I don't know so much fsockopen and I
didn't understand in which way I'm handling the file, it's quite weird
to me to open a socket to a server and then get the headers setting the
header in the socket. If it is like that how to make me pass the stream
and create the image without doing another call to the cgi with the
risk of having the error one?

Any help would be welcome and let me know if something is not clear.
Thanks, chr

[christiang]

If I use fopen($url) I'm able to copy the stream:

$buffer = fopen($url,"r");
if ($buffer)
{
$head = fread($buffer, 4096);
print("first: ".preg_match('[Content-Type: image/gif]', $head));
print("second: ".preg_match('#^HTTP/.*\s+[200|302]+\s#i', $head));
//print($head."<br>");
if(!preg_match('[Content-Type: image/gif]', $head) &&
preg_match('#^HTTP/.*\s+[200|302]+\s#i', $head)){
print("found image!<br>");
$foundThumb = true;
$content = $head;
while (!feof($buffer))
{
$content .= fread($buffer, 8192);
}
$ourFileHandle = fopen($this->thumbs_folder."/".$name, 'w');
fputs($ourFileHandle,$content);
fclose($ourFileHandle);
}
fclose($buffer);
}

but then my regex don't work anymore. I tried to create put together
the stream in my previous code:

if(!preg_match('[Content-Type: image/gif]', $head) &&
preg_match('#^HTTP/.*\s+[200|302]+\s#i', $head)){
print("found image!<br>");
$foundThumb = true;
$content = $head;
while (!feof($fid))
{
$content .= fread($fid, 8192);
}
$ourFileHandle = fopen($this->thumbs_folder."/".$name, 'w');
fputs($ourFileHandle,$fid);
fclose($ourFileHandle);
}

but the files are corrupted (ridiculous small size). Any clue?

[Janwillem Borleffs]

ch********@lateral.net wrote:

but the files are corrupted (ridiculous small size). Any clue?

Send some details like where this service can be found, so we can try it for
ourselfs.
JW

[christiang]

$url =
"http://pthumbnails.alexa.com/image_server.cgi?size=small&url=http://www.apple.com";

thanks, chr