URL's With Spaces

Anon

Ok, I run windows with Apache+PHP+MySQL.

My problem--I need for my script to be able able to recieve URL's that have spaces in them, and there's no way for me to urlencode them.

an example would be

http://myserver.com/myscript.php3?myname=Scott Sipe&a=b

There's nothing I can do about the space, and in this example myname=Scott only. Is there a way to fix this?

I am going to start reading apache docs if there are no PHP side suggestions?

Thanks,
Scott Sipe
scott.sipe@duke.edu

Anon

There is nothing PHP can do about invalid URLs. What broken Web client would send such a thing?

Anon

The web client:
AIM/30 (Mozilla 1.24b; Windows; I; 32-bit)

to be exact, from AOL Instant Messenger.

I know sites can handle the space in the URL, but I can't figure it out. www.subprofile.com (which is down right now) manages it somehow.

I've checked all the apache environmental variables and none of them have the username in it, so I'm really not quite sure what to do.

About the only thing I CAN think of doing, is parsing the apache log file, but that would be an absolute last resort.

thanks
Scott

Anon

From Apache's point of view, everything in a GET that follows a legal URL must be protocol. Give the following a try.

Reconstruct the query string:

$evil_query=$QUERY_STRING . $SERVER_PROTOCOL;

Then kill everything beginning with "HTTP/" and parse the remaining mess.

Anon

Ahah, SERVER_PROTOCOL contained the rest of what I needed, I can combine and parse the two as required..thanks a lot!!

Scott

nightninjacom

i'm trying to solve this problem also, and i don't understand what you both just conversed... could you explain in a little further detail how to do this, if its not to much trouble? thanks

Anon

Basically, spaces are NOT legal in URLs. Apparently some clients will send them anyway, forming illegal HTTP requests.

HTTP requests actually look like this:

GET /pub/WWW/TheProject.html HTTP/1.1
Host: www.w3.org

If you send an illegal space, it will look like this:

GET /pub/WWW/The Project.html HTTP/1.1
Host: www.w3.org

Apache handles the error by processing the request as if it is HTTP/1.1, and stuffing the unknown bogus protocol into $SERVER_PROTOCOL, where you can retrieve the rest of the URL.

You then can reconstruct the URI and figure out what to do.

nightninjacom

i'm very new at this stuff and i don't understand how to put it into the $SERVER_PROTOCOL and put the URL back together... i don't wanna be a nuisance(sp!) sorry