Solving bug #2397 , I think it’s clear that we want the following 2 examples to store files in the same physical location:

<s3://bucket/prefix>

and

<s3://bucket/prefix/>

. i.e. all objects stored in

<s3://bucket/prefix/>

But what about cases when the user provides something strange like

<s3://bucket/prefix///>

, should we also store it under

<s3://bucket/prefix/>

, or is

<s3://bucket/prefix///>

appropriate in this case..?

What do other systems that use object store URIs do (e.g. Spark, Presto, Hive, AWS CLI) when requesting to write to such a path? if there’s no “standard” behavior for this I vote to simply not accept it, since the meaning is ambiguous.

Turns out Spark ignores multiple slashes

<s3://bucket/prefix///>

is equivalent to

<s3://bucket/prefix/>

. So will remove trailing slashes in a similar manner to avoid errors for this very peculiar case.