Fancier parser/translator classes
It would be great to ease the addition of more parsers (or translators) to allow addition of new entries based on other inputs.
I had already made a note about these translators
in #45 (closed) when discussing the ISBN parser implementation.
Now that this is merged, I really think it would be great to extract the different parsers into separate classes in their own module. In this fashion the addition of new method is easier and cleaner (in terms of the open-closed principle) and could in theory even be done by an advanced end-user. Furthermore, in this fashion we could also provide methods to import from other tools such as papis, Zotero or Mendeley using the same interface.
Since this is essentially a quite significant refactoring, I will add this to the 3.0
(although it may rather become 3.X
) milestone.