Elsevier parser should extract metadata from `simple-head` too

[michamos]

In the Elsevier parser, we extract several pieces of metadata from the head field (for example title). However, it seems that some documents lack a head field and have a simple-head instead.

Example file: https://gist.github.com/michamos/a6e2ded5b3e32e3ba96e4d1c80501ada