Bit more complicated PHP question

[Porsche_maniak]

Hi guys !
I am not so good with php as you previously may know...
So i can't achieve the following and i hope you can help me with some script.

I want to get the contents from .gz file (which is a txt file) and search in this .txt file for match with '[url=http://rapidshare , [url=http://megaupload ,[url=http://hotfile , and .torrent '.

If there is match , then show rapidshare.png for [url=http://rapidshare
If there is match , then show megaupload.png for [url=http://megaupload

and so on...

Thanks !

Porsche_maniak Reviewed by Porsche_maniak on 27th May 2010. Bit more complicated PHP question Hi guys ! I am not so good with php as you previously may know... So i can't achieve the following and i hope you can help me with some script. I want to get the contents from .gz file (which is a txt file) and search in this .txt file for match with '[url=http://rapidshare , [url=http://megaupload ,[url=http://hotfile , and .torrent '. If there is match , then show rapidshare.png for [url=http://rapidshare If there is match , then show megaupload.png for [url=http://megaupload Rating: 5

[JmZ]

$file = gzfile('yourfile.gz');
$file = implode("\n",$file);

$matches = array(
    array('rapidshare.png', '[url=http://rapidshare'),
    array('megaupload.png', '[url=http://megaupload')
);

foreach($matches as $match) {
    if(strpos($file, $match[1]) !== false) {
        $image = $match[0];
        break;
    }
}

Something like this maybe.

This is assuming your gz file is purely a gzipped text file, not a tar.gz for example.

[Porsche_maniak]

thanks jmz but i have many .gz archives.

[litewarez]

Firstly you need to break it up.

• Archive Opening (extraction into memory) | gzopen
• Parsing the contents (Extract links from tags) | preg_match_all
• Compile the output.

So i would start by making a base class to work with

class GzipDownloads
{
   protected $gzfile = null;
   protected $gzcontents = null;

   function __construct($file)
   {
        if(!file_exists($file))
       {
           trigger_error('Unable to open ' . $file,E_USER_ERROR); //Die here
       }
       $this->gzfile = gzopen($file);
       $this->gzcontents = gzread($this->gzfile,filesize($file));
   }

   function getMeta()
   {
      preg_match_all('/\[url\=(.*?).*?\].*?\[.*?\]/is',$this->gzcontent,$matches);
      $mata = array();
      foreach($matches as $match)
      {
          //Url segment should be [1]
          if(preg_match('/http:\/\//',$match[1]))
          {
               $usegments = parse_url($match[1]);
               if($usegment['host'])
               {
                   $host = str_replace(array('.com','.net','.co.uk','.org'),'',$usegment['host']); //Remove tld
                   $meta[$host] = true;
               }
          }
       }
       return $mata;
   }
} 

Before i can build the getMeta() method witch will hold the links and other statuses, i need to examine the contents of the text file to look for similarities

-- Updated CODE AND Read below

$gzd = new GzipDownloads('my.file.txt.gz');
$meta = $gzd->getMeta();
foreach($meta as $host)
{
    if(file_exists('images/' . $host . '.png'))
    {
        //images/rapidshare.png exists so show it here!
        //$host will be the domain name without any tld so rapidhsare links will be rapidshare, hotfile.com/../../../ will be hotfile
    }
} 

The above is all example and untested.

[JmZ]

Firstly you need to break it up.

• Archive Opening (extraction into memory) | gzopen
• Parsing the contents (Extract links from tags) | preg_match_all
• Compile the output.

So i would start by making a base class to work with
Before i can build the getMeta() method witch will hold the links and other statuses, i need to examine the contents of the text file to look for similarities

He doesn't need regex, just to check if it contains rs, mu or whatever else.
Also, gzfile() should be fine for any gzipped text file.

Porsche_maniak: the code I posted will read one gzipped text file and check which hosts it contains. This seems like what you are asking for.

If you wish to handle multiple files, simply read each one in a loop or use a class and have an object per file.

Regardless of how you read it in though, that is how you should match certain strings.

[litewarez]

Update my post, please check!

Edit:
I really don't know why people run from PCRE :/ puzzles me, i mean PHP /mysql can handle 100s / 1000s of queries per second yet people try so hard to get them down to like 3 and that :/

[JmZ]

Again, im not trying to say your code is wrong lite, but regex is useless here.

He needs to check for matches, he doesn't need the matches.
Strpos is much, much faster.

Also, since all he wants is to read the files in and check for string matches, such a complex object won't be needed.

[litewarez]

Yea i know dood, just saying using my code you only need to have an image in the folder and it will match for that aswell.

so if a new host came out called terahost.com, he just needs to add the terahost.png to the dir and its found,

[JmZ]

Yes but the point is, it is excessive and slower than it could be.

All you need is to read each gz in, check for matches with strpos and use the correct image. It is a very, very simple problem with an equally simple solution.

[Porsche_maniak]

litewarez really tnx for your effort,but i think that JmZ is right...

@ JmZ
how do i read each one in a loop ?

@litewarez
Hmm sounds interesting...