Parsing vocabularies with trailing and without trailing slash result in different stat results.

[earthquakesan]

Bug report initially comes from Pierre-Yves Vandenbussche (LOV use case).

When checking vocabularies information on your website I noticed that
there are some "redundancies". For instance the very same vocabulary [1]
is present two times ending with a "/" or not [2]. Whereas this has no
influence on how we use your data, this could be misleading for some
users. I guess this shouldn't be hard to fix.

[1] http://lov.okfn.org/dataset/lov/details/vocabulary_postcode.html
[2] http://stats.lod2.eu/vocabularies?search=http%3A%2F%2Fdata.ordnancesurvey.co.uk%2Fontology%2Fpostcode