A tool for cleaning up bibtex databases to enforce uniform and unique
identifiers.  The command must include an input bibtex file(s) and options to
specify th desired output.  Each entry is processed to use a new identifier,
and duplicated entries are excluded.  The new identifier can be specified using
python's dictionary substitution.  The following variables are available:

  bid: the original bibtex identifier
  year: the year of publication
  lname: the lowercase last name of the first author
  name: the capitalized last name of the first author
  uname: the uppercase last name of the first author
  lword: the lowercase first word of the title
  word: the capitalized first word of the title
  uword: the uppercase first word of the title
  lthree: the lowercase three letter identifier of the title
  three: the capitalized three letter identifier of the title
  uthree: the uppercase three letter identifier of the title

All non-ASCII and non-alphabetic characters are removed, and title-derived
variables omit a number of stop works (e.g. "a", "an", "the", etc.).  As an
example, the identifier pattern "%(lname)s%(year)s%(word)s" produces a format
similar to Google Scholar:

  @article{langmead2009ultrafast,
    title={Ultrafast and memory-efficient alignment of 
      short DNA sequences to the human genome},
    author={Langmead, Ben and Trapnell, Cole and Pop, Mihai 
      and Salzberg, Steven L and others},
    journal={Genome biol},
    volume={10},
    number={3},
    pages={R25},
    year={2009}
  }

This tool is based on the very nice package Pybtex (http://pybtex.org).

Author: Ryan Cabeen, cabeen@gmail.com