handle encoding of base64Binary Literals

[delocalizer]

Fixes #1257

Proposed Changes

• add base64 encoding for base64Binary datatype, by analogy to hexlify/unhexlify for hexBinary
• (cosmetic) remove use of redundant lambda when handling decode

tests pass

[coveralls]

Coverage decreased (-0.006%) to 75.486% when pulling 31a3d16 on delocalizer:fix/base64Binary into a8d435b on RDFLib:master.

[white-gecko]

Thank you very much for your contribution :-)

This is a case I have not encountered so far. Could you please provide a test that shows what was failing before and that it is fixed now?

[white-gecko]

Will there be any interference with #1222 ?

[delocalizer]

The test is by analogy with the existing test_hex_binary (minus the integer stuff that relies on hex-specific features of format builtin and int constructor)
it fails on master but passes with this PR

[delocalizer]

Will there be any interference with #1222 ?

I don't believe so, the issue is not with the serialization per se, rather that encoding was not defined for xsd:base64Binary type so the decoded value was being used as the lexical. Now I think about it, a more descriptive summary of this PR might be "make handling of xsd:base64Binary type consistent with the existing handling of xsd:hexBinary."

[aucampia]

@white-gecko @ashleysommer this PR is still pending on your reviews

[nicholascar]

This PR has had a couple of reviews, passes all tests and introduces more tests, so I'm merging. It's pretty low-risk PR too (just an updated Literal type).