Hi, OK this one is a little embarrasing. I've been doing this for years and I just wonder if I'm wrong....
Say you have an exit link on your site, e.g. /leave.php, which accepts a "url" get arg. You use this page to record stats/whatever and then issue a Location: header to take the user to the correct location. Fairly standard yeah? Well I've been doing something like: $url = 'http://colin.guthr.ie/'; echo '<a href="/leave.php?url='.urlencode($url).'">Click</a>'; The logic in /leave.php does not need to call urldecode as it's done automatically by PHP. This has worked well for me in the browsers I've used (IE, FF etc.) Recently, though, when using google webmaster tools I noticed that I was getting a lot of 404's and this ultimately stemmed from the double urlencoding of these url paramaters whereby the % signs used to encode characters like / as %2F were encoded themselves leading to %252F. PHP would automatically urldecode this to %2F but that still leaves me with an encoded variable. Ugg. So my question is, is the google bot just getting it wrong? Is it reading the link and seeing a % and encoding it? Or is it finding a page somewhere randomly on the interweb which has incorrectly double encoded it and going from there? It doesn't give you an referrer info which makes tracking down such errors pretty tricky.... :( I could just call urldecode manually, but I'm curious as to why I should need to. Anyone fought with this before? Col -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
