Dealing with extra '/' in URL - php

Question

So I have a custom made site that uses this type of input:
example.com/?id=4e2dc982
Or this would also work:
example.com/index.php?id=4e2dc982
But now I've started seeing hits in my log from GoogleBot trying to retrieve this for some reason:
example.com/index.php/?id=4e2dc982
The worse thing is that this actually works, it pulls the page with the right GET parameter, but because of the extra '/' all the links and references don't work. When it tries to load "image.jpg" instead of loading the proper "example.com/image.jpg" it tries to load "example.com/index.php/image.jpg". How do I best fix this? I know I could go back and replace every link to use absolute paths but that's silly. The link with an extra '/' shouldn't work in the first place.
Update:
I found the fix, but still don't know why this is even allowed. I went to:
http://ca1.php.net/manual-lookup.php?pattern=test
And tried to see if the following was possible, and sure enough it works:
http://ca1.php.net/manual-lookup.php/?pattern=test
But their page doesn't break. So I looked at it and found out why:
<base href="http://ca1.php.net/manual-lookup.php" />
So basically, ANY PHP script seems to accept an extra /, but if you didn't code all your links to have absolute paths, or use a base tag, your site will be screwed up whenever someone adds an extra '/'.

Answer 1

It must be linked from somewhere, which you need to figure out from where. You can use google site search to (i.e. site:yoursie) may be to figure out.
One suggestion for now is to use canonical tag
http://googlewebmastercentral.blogspot.com.au/2009/02/specify-your-canonical.html

Answer 2

I think that one of the things you could actually do is getting the header or browser agent (although some browsers don't send this), you could possibly do it. Then if the header contains anything like Google, do not allow the bot to crawl the page, else redirect the user to the site.
Below is an example:
$browser = $_SERVER['HTTP_USER_AGENT'];
checkbrowser($browser); //Calls checkbrowser(); with the browser version.
function checkbrowser($analyze) {
$searchwords = array("bot","google","crawler");
$matches = array();
$matchFound = preg_match_all(
"/\b(" . implode($searchwords,"|") . ")\b/i",
$analyze,
$matches
);
if ($matchFound) {
$words = array_unique($matches[0]);
foreach($words as $word) {
if($word == "bot") {
echo "Sorry, bots are not allowed to crawl this specific page.";
die(); //Terminate the script and leave the bot with that message so it cannot crawl.
}
}
}
}
This is how I often do it, but I utilize this method for different things. You can modify the function by changing the $searchwords to something that fits you best.