Http request open POST instead of GET

[Zhris]

Here is a section of my code:

function fAccess() {

	var message=document.getElementById('message').value;
	document.getElementById('message').value = "";
	document.getElementById('message').focus();
    var url="bottom_execute.pl?Message="+message+"&Random="+Math.random();
    xhr.open("GET",url,false);
    xhr.send(null);
    document.getElementById("DisplayA").innerHTML=xhr.responseText;

}

The problem I have is that sometimes the message will contain URL's which may be in the format: "

http://domain.com/index.php?key=value&key2=value2"

They are then sent to the php script in the format: "script.php?message=

http://domain.com/index.php?key=value&key2=value2"

However the script reads this format like:
script.php
?
message=

http://domain.com/index.php?key=value

&
key2=value2

Instead of:
script.php
?
message=

http://domain.com/index.php?key=value&key2=value2

key2=value2 is no longer part of message.

I have tried replacing xhr.open("GET",url,false); with xhr.open("POST",url,false); assuming this would have fixed the problem, however message is not sent to the php script. Can this be fixed in the javascript or is this a php issue?

Any help will be much appreciated.

Thanks,

Chris

 

[BabyJeffy]

It's something easily fixed by escaping the ampersands in the initial URL. You would do this in javascript, and then would need to also un-escape them in your script server-side.

var param = "[URL unfurl="true"]http://www.tek-tips.com/viewthread.cfm?qid=1511561&page=1";[/URL]
escapedParam = [!]escape(param)[/!];
// escapedParam is now "http%3A//[URL unfurl="true"]www.tek-tips.com/viewthread.cfm%3Fqid%3D1511561%26page%3D1"[/URL]
var url = "[URL unfurl="true"]http://www.example.com/index.php?message="[/URL] + escapedParam;

In php you might do this:

$param = $_GET['message'];
$unescapedParam = urldecode($param);

Hope this is workable for you.

Cheers,
Jeff

[tt]Visit my blog [!]@[/!]

http://www.coderambler.com/

Visit Code Couch [!]@[/!]

http://www.codecouch.com/

[/tt]

Make sure your web page and css validates properly against the doctype you have chosen - before you attempt to debug a problem!

FAQ216-6094

 

[Ascalonian]

it is a URL issue I believe...

message will read everything up to the next & sign.

I would recommend escaping the url for message and then unescape it in the php script.

so...

var oldMessage = '[URL unfurl="true"]http://domain.com/index.php?key=value&key2=value2';[/URL]

var newMessage = escape(oldMessage);

Then in your PHP script, I believe the function is called urldecode to get the URL back how it should be.

If that doesn't work, I hope this gives you some direction at least.

 

[Zhris]

Thank you for your response(s),

I've been working with PHP all day and had it in my mind at the time of posting. I'm actually working with a Perl backend script but anyhow decoding isn't a problem (in fact when I first tested I didn't even need to decode :s).

Escaping works very nicely. However the information (message) is stored to a plain txt file and it seems as if a newline character is included to the end after encoding. I have tried using message = str.replace(/^\s+|\s+$/g, ''); to remove the newline after trying other numerous trimming methods, however none work.

Any suggestions?

Chris

 

[Zhris]

Ok I should have tried it before, but I have it working now. I removed the newline character using Perl's chomp() which I had previously assumed wouldn't work due to the possiblity that javascript/perl newline characters would be different (don't ask me why!).

Still interested to know the reason why I don't need to decode the string again in the Perl script?

Thank you both for your help, its very much appreciated.

Chris

 

[Zhris]

I've just realised it creates a new line if processed in Internet Explorer (wrong) but doesn't in Firefox (correct). I'm working on it but any suggestions in the mean time would be appreciated.

Chris