in reply to Re: A new chatterbox flavour
in thread A new chatterbox flavour

Duh, of course. Thanks for catching that. Doing it at the comparison point is wrong though, the real fix is to de-escape the URI encoding for cookie data when it's parsed. That means simply:

--- pmchatter.html.old  2004-12-19 17:05:59.000000000 +0100
+++ pmchatter.html      2004-12-19 17:05:13.000000000 +0100
@@ -17,3 +17,3 @@
                        var kv = record[ i ].split( /=/ );
-                       this.data[ kv[ 0 ] ] = kv[ 1 ];
+                       this.data[ decodeURIComponent( kv[ 0 ] ) ] = d
+ecodeURIComponent( kv[ 1 ] );
                }
@@ -63,3 +63,3 @@
        if( userpass.length ) {
-               req.setRequestHeader( 'Cookie', 'userpass=' + userpass
+ );
+               req.setRequestHeader( 'Cookie', 'userpass=' + encodeUR
+IComponent( userpass ) );
        }

[download]

I updated the root node.

Update: it wasn't quite as trivial — need to re-escape the cookie to send it, of course.

Makeshifts last the longest.

Replies are listed 'Best First'.
Re^3: A new chatterbox flavour
by The Mad Hatter (Priest) on Dec 19, 2004 at 16:16 UTC
    Good point. I hadn't really studied it much other than looking for where the login "failed" for me.
[reply]

      But then I forgot to re-escape the cookie… patch and root node updated.

      Makeshifts last the longest.

[reply]

        And it's still broken. :-)

        I just got around to using your updated version instead of my hackery, and noticed it didn't work. Here's a fix: 

        --- pmchat.html.old     2004-12-19 20:12:15.000000000 -0500
+++ pmchat.html 2004-12-19 20:07:59.000000000 -0500
@@ -15,7 +15,7 @@
                var record = str.split( /\s*;\s*/ );
                for( var i = 0; i < record.length; ++i ) {
                        var kv = record[ i ].split( /=/ );
-                       this.data[ decodeURIComponent( kv[ 0 ] ) ] = d
+ecodeURIComponent( kv[ 1 ] );
+                       this.data[ decodeURIComponent( kv[ 0 ].replace
+(/%25/g, '%') ) ] = decodeURIComponent( kv[ 1 ].replace(/%25/g, '%') 
+);
                }
        },
        get: function( key ) {

        [download]
        PM seems to double-encode or something -- at the least, percents (%) are replaced with %25 in the already URL encoded string. It doesn't seem to care if you do the reverse or not when re-encoding it.

[reply]
[d/l]

In Section Perl Monks Discussion