Forms question.

ucs308

This may take a couple of interations to fully understand what I am doing. But in short I have a page (http) that is included using fopen, on the included page I can not get access to the $_SESSION I can see them in the page that includes the page referred to above, but not on the included page.

Any ideas? Do I need to post more info.

jimson

u can't

www.test.com/a.php

$a = 25;

www.java.com/b.php

include('http://www.test.com/a.php');
++$a;
echo $a;

i don't think you will have $a = 26.

ucs308

I thought that $_SESSION were of global scope. And I assumed Global scope would be for the aggregated page.

Oddly enought I can access $_POST Vars.

Quest

a.php - " 25 "// in page

$num_array = file("a.php");
$data = explode("\r\n", $num_array);

$data[0]++;

echo $data[0];

a.php - " $a = 25 " //in page

include("a.php")

a$++;

echo $a;

Weedpacket

Originally posted by ucs308
I thought that $_SESSION were of global scope. And I assumed Global scope would be for the aggregated page.

Oddly enought I can access $_POST Vars.

That is odd, because:

[man]include[/man]
....you can specify the file to be included using a URL insted of a local pathname. If the target server interprets the target file as PHP code, variables may be passed to the included file using an URL request string as used with HTTP GET. This is not strictly speaking the same thing as including the file and having it inherit the parent file's variable scope; the script is actually being run on the remote server and the result is then being included into the local script. [Emphasis mine]

ucs308

I looked at this a little closer and I see that I am getting multiple session id's for each page load. In short the page that I "include" in my main page is being given a different session id. I verfied this by looking at the session files being created on the server, and setting a var with an index of the page name, to yes.

My pages are structured as follows.

page.php
Common Header
Body (included using the fopen code below)
Common Footer

The body is included using the following code.

$pagecontent = 'http://xxx.xxx.xxx.xxx:2035/contact.php';
if ($pagecontent != NULL) {				 
  $fileHandle = fopen($pagecontent,'r') or die($php_errormsg);
  while (! feof ($fileHandle)) {
    $thePage .=fread($fileHandle,1048576);
  }
  fclose ($fileHandle);
} else {
  $pagecontent ="";
}		

Session vars are shared for any load of page.php regardless of which body is included. But the included body has a different session Id and as expected different session vars.

So my question becomes why does the included page get a new session id?

Weedpacket

Originally posted by ucs308
So my question becomes why does the included page get a new session id?

Like the manual excerpt I gave said "the script is actually being run on the remote server and the result is then being included into the local script."
The two scripts are being run as entirely separate requests - separate sessions, separate session IDs.

ucs308

OK. I understand that "it" is treating them separately. But I don't understand why. There is no remote server. There is only one server.

When I tack the PHPSESSID var onto the URL I get a 403. Or the browser hangs.

Is there something wrong with my configuration? Or is the way PHP is designed.

stolzyboy

but you are including a page from an external server(a domain/ip other than yours), right, so you do have two servers, yours and that domain/ip

or am i reading this thread wrong

ucs308

I am developing this on single server. IIS5. Using class C ip address. I uses the following to load a page..

http://111.111.111.111:9000/page.php?page=contact

Or say..

http://111.111.111.111:9000/page.php?page=support.

Inside Page.php I am using a switch to build a url as follows.

http://111.111.111.111:9000/contact.php.

and then including this page using fopen, or include, they both yeild the same results.

The session id for page.php is relativley static/stable. The session id for the include/fopened page changes each time the page is loaded.

If were loading this froma remote server or local machine that is different, or even just a different web server, I would expect the session id to change. BUT I am not. This is all on the same box. The same web server.

This happens for all pages loaded this way.

Weedpacket

But you're still using the http protocol to request it from the web server. It's an entirely separate connection, an entirely separate request, and since you're not sending the session ID as part of the request, an entirely separate session.

If they're all on the same server why are you even using http URLs to locate the files? Why aren't you referring to them by the filename? Why ask the web server for the file instead of asking the filesystem?

ucs308

There is no special reason for using http: versus just referring to them as files. I assumed that the underlying layer would figure it out.

I changed the code to refer to files instead of URL's and the session stuff works as expected.

Was it unreasonable to expect whatever layer handles sessions to spot that it was the same server, same user, same connection and by definition the same session. I get a little confused with all the actors involved in these technologies.

I did try adding the session id to the URL but that failed to work, giving me a 403 error, which I think is telling me I am not authorised.

Thanks for your insights. PHP is new to me, and your probing has been very helpful.

Weedpacket

Glad I could help. Since you were wondering, while a reasonable thing to guess, expecting the underlying layer to sort things out is in fact infeasible.

The reason is that URLs don't necessarily refer to files on the web server. URLs are (as the name says) "resource locators". In other words, if you give the the service a URL, it will (all being well) respond with the resource at that "location".

There is, however, no guarantee that just because there is a resource called "foo.html" there is necessarily a file with that name anywhere on the server. There may be some script that generates the foo.html page whenever its requested.

Similarly, a URL like http://www.example.com/news/archives/2001/05/11/page3.html doesn't necessarily mean that sitting on the server is a directory called news containing a directory called archives containing a directory called 2001 ... containing a file called "page3.html". In fact, that's highly unlikely to be the case. What is much more likely is that the web server will receive that URL, and transform it into something else, say, http://www.example.com/news.php?y=2001&m=05&d=11&p=3 - and news.php would be pulling stuff out of a database.

The practical upshot is, if you ask for something via http, the request has to be sent to the http server, because the http server has to figure out out which resource goes with which URL (and that in a nutshell is all an http server does for a living.)