Litstudy

Added
- Add support for loading CSV files exported from Scopus (see PR 45, Thanks tleedepriest!)

Changed
Removed
Fixed
- Fix incorrect return type of `load_ris_file` (fixes 34)
- Fix passing session as non-positional argument in `refine_semanticscholar`, see PR 35. (Thanks martinuray!)
- Fix incorrect filtering in `Corpus` when building corpus from docs (fixes 38)
- Fix error when calling `fetch_crossref` and `refine_crossref` with `session=None` as argument (fixes 40)
- Fix KeyError when loading CSV files from IEEE that have an incorrect format (fixed 55)
- Fix bug in nlp.py due to argument `square_distances` being removed in newer versions of sklearn, see PR 58 (Thanks rjavierch!)

Fixed
- Fix wrong argument in call to `matplotlib.pyplot.grid(...)` due to change in their API
- Fix semanticscholar backend not retrieving papers correctly

Added
- Add `load_csv` function
- Add `search_crossref` function

Fixed
- Fix issue where CSV files could not be parsed due to BOM marker

Fixed

- Fix bug in the semantic scholar backend that did not fetch papers correctly
- Fix bug in `fetch_crossref` where document title was not extracted correctly

Changed
- Remove dependency on fa2. The version of fa2 on pip is broken under Python 3.9+.

Fixed
- `litstudy` now works under Python 3.9+.

Added
- Support for the arXiv API (Thanks ksilo!)

Changed
- Made project compatible with Python 3.6