S3Iterator Strips / from Directory

[amarcionek]

I've noticed that the S3Iterator returns directory elements with an S3Path that is missing the / at the end of the key. With the results from the listobjects call, the code intentionally splits each returned object on the / and then feeds those elements to the S3Path. S3Path in turn build a URL, but then if would remove a / at the end if it exists.

What I don't understand is why? If a key ends with a / then its the only way to access that object.