The project has had a long and convoluted history. My log goes as far as May 2012 when I seem to have rewrote the parsing routines from scratch. Even that was repeated in November of the same year. In June 2014 I decided to use a dict instead of a list as the main object class. The parsing routines were simplified in February 2016, and in June of that year I made major changes to them to increase their robustness.
In August 2016 I switched to using PyPI-compatible version numbering scheme in order to prepare for PyPI packaging, and made minor changes to the code in order to ensure Python 2 compatibility. The latter appears to have still caused issues since in July of the next year (2017) I had to do more changes towards it.
1 July 2019 was declared as the first “public” version (“1.0.0”), yet I continued to make bug fixes. During the next two weeks I made several semi-major changes to the code so the “1.0.0” moniker was clearly premature.
Trying to simplify code.
Completely rewritten parsers and lots of other internal changes, including the possibility to use "..." for the BibTeX fields. Python 2 compatibility hacks discarded. f-strings utilized.
bibkeys updated.
Finally there.