fopen/fsockopen question

Dear All,

I want to write a simple PHP code to verify whether a given list of
URLs is broken or not. The URLs given have various formats, for example,

http://www.afro.com/history/history.html
http://www.worldhistorycompass.com/index.htm
http://www.afup.org/article.php3?id_article=242

fsockopen(...) only accepts target format (like www.yahoo.com etc), so I
simply choose fopen (url,'r') to do the test.

This method seems working well for majority of URLs I have. However,
when the following site seems causing problem,

$site = fopen("http://www.marketingpower.com/live/topics12.php",'r');

my browser seems working very hard on this link and after a long time,
the browser stops (not continuing on the next command. I am using
netscape on linux).

Initially I thought probably fopen doesn't like http://xxxx/xxx.php type
of format, however, links like:

http://www.afup.org/article.php3?id_article=242

works just fine.

So I am confused. Can anybody give me a hint on this?

Thanks.

re: fopen/fsockopen question

Shaoyong Wang wrote:
[color=blue]
> Dear All,
>
> I want to write a simple PHP code to verify whether a given list of
> URLs is broken or not. The URLs given have various formats, for example,
>
> http://www.afro.com/history/history.html
> http://www.worldhistorycompass.com/index.htm
> http://www.afup.org/article.php3?id_article=242
>
> fsockopen(...) only accepts target format (like www.yahoo.com etc), so I
> simply choose fopen (url,'r') to do the test.
>
> This method seems working well for majority of URLs I have. However,
> when the following site seems causing problem,
>
> $site = fopen("http://www.marketingpower.com/live/topics12.php",'r');
>
> my browser seems working very hard on this link and after a long time,
> the browser stops (not continuing on the next command. I am using
> netscape on linux).
>
> Initially I thought probably fopen doesn't like http://xxxx/xxx.php type
> of format, however, links like:
>
> http://www.afup.org/article.php3?id_article=242
>
> works just fine.
>
> So I am confused. Can anybody give me a hint on this?
>
> Thanks.[/color]

I've tried to use nc to open the site you mentioned. It seems it does a
whole lot of redirecting and setting cookies. Might be that's what's
giving the URL wrapper trouble opening the page properly.

If that's the case, you will have no other choice than to write your own
HTTP request handler, which handles redirects and cookies. Which isn't
too hard, but it's a lot of work :).

-Rico

re: fopen/fsockopen question

Thanks.

This link seems the only problematic one so far. I can just skip this one.
Notice fopen can't stop at this link, even I set max_execution_time in
php.ini, no idea.
I am using PHP 4.2.2

re: fopen/fsockopen question

On Thu, 27 May 2004 13:11:59 +0200, Rico Huijbers
<E.A.M.Huijbers@REMOVEstudent.tue.nl> wrote:
[color=blue]
>Shaoyong Wang wrote:
>[color=green]
>> Dear All,
>>
>> I want to write a simple PHP code to verify whether a given list of
>> URLs is broken or not. The URLs given have various formats, for example,
>>
>> http://www.afro.com/history/history.html
>> http://www.worldhistorycompass.com/index.htm
>> http://www.afup.org/article.php3?id_article=242
>>
>> fsockopen(...) only accepts target format (like www.yahoo.com etc), so I
>> simply choose fopen (url,'r') to do the test.
>>
>> This method seems working well for majority of URLs I have. However,
>> when the following site seems causing problem,
>>
>> $site = fopen("http://www.marketingpower.com/live/topics12.php",'r');
>>
>> my browser seems working very hard on this link and after a long time,
>> the browser stops (not continuing on the next command. I am using
>> netscape on linux).
>>
>> Initially I thought probably fopen doesn't like http://xxxx/xxx.php type
>> of format, however, links like:
>>
>> http://www.afup.org/article.php3?id_article=242
>>
>> works just fine.
>>
>> So I am confused. Can anybody give me a hint on this?
>>
>> Thanks.[/color]
>
>I've tried to use nc to open the site you mentioned. It seems it does a
>whole lot of redirecting and setting cookies. Might be that's what's
>giving the URL wrapper trouble opening the page properly.
>
>If that's the case, you will have no other choice than to write your own
>HTTP request handler, which handles redirects and cookies. Which isn't
>too hard, but it's a lot of work :).[/color]

couldn't you just do something like this?:

$hostname = preg_replace('/http:\/\/([^\/]*).*/','$1',$url);
$conn_id = fsockopen('tcp://' . gethostbyname($hostname), 80,
$errno, $errstr);
fputs($conn_id,"GET $url HTTP/1.1\r\nHost: $hostname\r\n\r\n");

where $url is the url in question, and then where you do whatever you
would normally do?

re: fopen/fsockopen question

Shaoyong Wang <wangsha@cis.ohio-state.edu> wrote in message news:<c94tpl$7vg$1@news.cis.ohio-state.edu>...[color=blue]
> Thanks.
>
> This link seems the only problematic one so far. I can just skip this one.
> Notice fopen can't stop at this link, even I set max_execution_time in
> php.ini, no idea.
> I am using PHP 4.2.2[/color]

1. If possible, use cURL functions instead. It is much handy
2. Check the user_agent setting of your ini file. Some sites may
lock weird/unknown user agents.

--
| Just another PHP saint |
Email: rrjanbiah-at-Y!com

re: fopen/fsockopen question

Shaoyong Wang wrote:
[color=blue]
> Dear All,
>
> I want to write a simple PHP code to verify whether a given list of
> URLs is broken or not. The URLs given have various formats, for example,
>
> http://www.afro.com/history/history.html
> http://www.worldhistorycompass.com/index.htm
> http://www.afup.org/article.php3?id_article=242
>
> fsockopen(...) only accepts target format (like www.yahoo.com etc), so I
> simply choose fopen (url,'r') to do the test.
>
> This method seems working well for majority of URLs I have. However,
> when the following site seems causing problem,
>
> $site = fopen("http://www.marketingpower.com/live/topics12.php",'r');
>
> my browser seems working very hard on this link and after a long time,
> the browser stops (not continuing on the next command. I am using
> netscape on linux).
>
> Initially I thought probably fopen doesn't like http://xxxx/xxx.php type
> of format, however, links like:
>
> http://www.afup.org/article.php3?id_article=242
>
> works just fine.
>
> So I am confused. Can anybody give me a hint on this?
>
> Thanks.
>[/color]

just use fopen

what you have to decide on are how many conditions do you have to deal with.

1) success!
2) a 404 response (file not found)
3) connection refused (port not listening)
4) timeout (server is down or not resolvable)