detection of a robot in php

Hello everybody :)

A friend recently showed me an odd thing while playing with the command
wget under linux, I don't know why... But the result has surprised me :
$ wget http://www.prizee.com/parole.php
--02:35:29-- http://www.prizee.com/parole.php
=> `parole.php'
Resolution de www.prizee.com... 213.186.63.5
Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
requete HTTP transmise, en attente de la reponse...302 Found
Emplacement: /index.php?joueur=1 [suivant]
--02:35:30-- http://www.prizee.com/index.php?joueur=1
=> `index.php?joueur=1.1'
Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
requete HTTP transmise, en attente de la reponse...200 OK
Longueur: non specifie [text/html]

[ <=>

] 12,521
--.--K/s

02:35:30 (103.57 KB/s) - ? index.php?joueur=1.1 a sauvegarde [12521]


Then, he obtains an http error code (302) which redirect him on the
index page of the site.
With a browser like firefox, ie, safari we get the good page without
any redirection.
After that, I've made some tests. I tried to change the user agent
string with wget to identify it as mozilla, but I have the same result
(redirection). I tried links (command line browser) and curl but same
problem.
Here is the result of curl command :

$ curl -v http://www.prizee.com/parole.php
* About to connect() to www.prizee.com port 80
* Trying 213.186.63.5... connected
* Connected to www.prizee.com (213.186.63.5) port 80
> GET /parole.php HTTP/1.1
> User-Agent: curl/7.15.1 (i686-pc-linux-gnu) libcurl/7.15.1 GnuTLS/1.2.10 zlib/1.2.3 libidn/0.5.15
> Host: www.prizee.com
> Accept: */*
>
< HTTP/1.1 302 Found
< Date: Wed, 09 Aug 2006 00:02:57 GMT
< Server: Apache/1.3.33 (Unix) PHP/4.3.10
< X-Powered-By: PHP/4.3.10
< X-Accelerated-By: PHPA/1.3.3r2
< Expires: Mon, 26 Jul 1997 05:00:00 GMT
< Last-Modified: Wed, 09 Aug 2006 00:02:59 GMT
< Cache-Control: no-cache, must-revalidate
< Pragma: no-cache
< Set-Cookie: COOKIEis_accepted=1; path=/; domain=.prizee.com
< Location: /index.php?joueur=1
< Connection: close
< Transfer-Encoding: chunked
< Content-Type: text/html
* Closing connection #0


So, my question is : How we can detect the use of a command line tool
on a web site ? Like the site above. Thank you for your answers.

Sorry for my bad english, i'm french ;)

giminik@gmail.com wrote:

> Hello everybody :)
>
> A friend recently showed me an odd thing while playing with the
> command wget under linux, I don't know why... But the result has
> surprised me : $ wget http://www.prizee.com/parole.php
> --02:35:29-- http://www.prizee.com/parole.php
> => `parole.php'
> Resolution de www.prizee.com... 213.186.63.5
> Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
> requete HTTP transmise, en attente de la reponse...302 Found
> Emplacement: /index.php?joueur=1 [suivant]
> --02:35:30-- http://www.prizee.com/index.php?joueur=1
> => `index.php?joueur=1.1'
> Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
> requete HTTP transmise, en attente de la reponse...200 OK
> Longueur: non specifie [text/html]
>
> [ <=>
>
> ] 12,521
> --.--K/s
>
> 02:35:30 (103.57 KB/s) - ? index.php?joueur=1.1 a sauvegarde [12521]
>
>
> Then, he obtains an http error code (302) which redirect him on the
> index page of the site.
> With a browser like firefox, ie, safari we get the good page without
> any redirection.
> After that, I've made some tests. I tried to change the user agent
> string with wget to identify it as mozilla, but I have the same result
> (redirection). I tried links (command line browser) and curl but same
> problem.
> Here is the result of curl command :
>
> $ curl -v http://www.prizee.com/parole.php
> * About to connect() to www.prizee.com port 80
> * Trying 213.186.63.5... connected
> * Connected to www.prizee.com (213.186.63.5) port 80
>> GET /parole.php HTTP/1.1
>> User-Agent: curl/7.15.1 (i686-pc-linux-gnu) libcurl/7.15.1
>> GnuTLS/1.2.10 zlib/1.2.3 libidn/0.5.15 Host: www.prizee.com
>> Accept: */*
>>
> < HTTP/1.1 302 Found
> < Date: Wed, 09 Aug 2006 00:02:57 GMT
> < Server: Apache/1.3.33 (Unix) PHP/4.3.10
> < X-Powered-By: PHP/4.3.10
> < X-Accelerated-By: PHPA/1.3.3r2
> < Expires: Mon, 26 Jul 1997 05:00:00 GMT
> < Last-Modified: Wed, 09 Aug 2006 00:02:59 GMT
> < Cache-Control: no-cache, must-revalidate
> < Pragma: no-cache
> < Set-Cookie: COOKIEis_accepted=1; path=/; domain=.prizee.com
> < Location: /index.php?joueur=1
> < Connection: close
> < Transfer-Encoding: chunked
> < Content-Type: text/html
> * Closing connection #0
>
>
> So, my question is : How we can detect the use of a command line tool
> on a web site ? Like the site above. Thank you for your answers.

I tried both Firefox and Konqueror and they both redirected me to the
second page, so there doesn't appear to be anything different between
using wget and using a graphical browser, at least to me.

You can't detect the use of a command line tool if they set the user
agent correctly. For example:

wget --user-agent="Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)"

followed by the url will tell the website you're using IE on Windows XP.

--
Chris Hope | www.electrictoolbox.com | www.linuxcdmall.com

giminik@gmail.com wrote:

> Hello everybody :)
>
> A friend recently showed me an odd thing while playing with the command
> wget under linux, I don't know why... But the result has surprised me :
> $ wget http://www.prizee.com/parole.php
> --02:35:29-- http://www.prizee.com/parole.php
> => `parole.php'
> Resolution de www.prizee.com... 213.186.63.5
> Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
> requete HTTP transmise, en attente de la reponse...302 Found
> Emplacement: /index.php?joueur=1 [suivant]
> --02:35:30-- http://www.prizee.com/index.php?joueur=1
> => `index.php?joueur=1.1'
> Connexion vers www.prizee.com|213.186.63.5|:80...connecte.
> requete HTTP transmise, en attente de la reponse...200 OK
> Longueur: non specifie [text/html]
>
> [ <=>
>
> ] 12,521
> --.--K/s
>
> 02:35:30 (103.57 KB/s) - ? index.php?joueur=1.1 a sauvegarde [12521]
>
>
> Then, he obtains an http error code (302) which redirect him on the
> index page of the site.
> With a browser like firefox, ie, safari we get the good page without
> any redirection.
> After that, I've made some tests. I tried to change the user agent
> string with wget to identify it as mozilla, but I have the same result
> (redirection). I tried links (command line browser) and curl but same
> problem.
> Here is the result of curl command :
>
> $ curl -v http://www.prizee.com/parole.php
> * About to connect() to www.prizee.com port 80
> * Trying 213.186.63.5... connected
> * Connected to www.prizee.com (213.186.63.5) port 80
> > GET /parole.php HTTP/1.1
> > User-Agent: curl/7.15.1 (i686-pc-linux-gnu) libcurl/7.15.1 GnuTLS/1.2.10 zlib/1.2.3 libidn/0.5.15
> > Host: www.prizee.com
> > Accept: */*
> >
> < HTTP/1.1 302 Found
> < Date: Wed, 09 Aug 2006 00:02:57 GMT
> < Server: Apache/1.3.33 (Unix) PHP/4.3.10
> < X-Powered-By: PHP/4.3.10
> < X-Accelerated-By: PHPA/1.3.3r2
> < Expires: Mon, 26 Jul 1997 05:00:00 GMT
> < Last-Modified: Wed, 09 Aug 2006 00:02:59 GMT
> < Cache-Control: no-cache, must-revalidate
> < Pragma: no-cache
> < Set-Cookie: COOKIEis_accepted=1; path=/; domain=.prizee.com
> < Location: /index.php?joueur=1
> < Connection: close
> < Transfer-Encoding: chunked
> < Content-Type: text/html
> * Closing connection #0
>
>
> So, my question is : How we can detect the use of a command line tool
> on a web site ? Like the site above. Thank you for your answers.
>
> Sorry for my bad english, i'm french ;)

probably just redirects for linux users and not ms by checking the
agent-type.

Flamer.

Thank for your answers.
I found the problem. It was a session cookie problem.
I've just used the wget option : --keep-session-cookies with
--load-cookies to solve the problem.