doi2bib

#!/usr/bin/env python
# This script takes a DOI as a command line argument or a redirected file
# with a list of DOIs and uses doi.org to generate a .bib file which is
# output to the screen.
#
# Timothy Crone (tjcrone@gmail.com)

#imports
import subprocess, re, sys
from select import select
from unicode_dict import unicode_dict

# read input dois
dois = []
if len(sys.argv) == 2:
  dois.append(sys.argv[1].strip())
else:
  timeout = 0.1
  rlist, _, _ = select([sys.stdin], [], [], timeout)
  if rlist:
    for line in sys.stdin:
      if line[0] != '#':
        dois.append(line.strip())

#error check dois
if len(dois) == 0:
  sys.stdout.write("Pass a DOI as a single argument, or pipe in a file with a list of DOIs.\n")
  raise SystemExit

# get value function
def get_value(bib, key):
  try:
    #value = re.sub('}$', '', bib.split('%s={' % key)[1].encode("utf-8").split('}, ')[0]).decode("utf-8")
    value = bib.split('%s={' % key)[1].split('}, ')[0]
    for i in re.sub('[ -~]', '', value):
      value = value.replace(i, unicode_dict[i])
  except:
    value = ''
  return value

# loop through dois
for doi in dois:
  #better:
  #curl -LH "Accept: text/bibliography" "https://api.crossref.org/v1/works/http://dx.doi.org/10.1029/2001JB000918.xml"

  # get raw bib entry
  cmd = ('curl -s -LH "Accept: text/bibliography; style=bibtex" "http://dx.doi.org/%s"' % doi)
  p = subprocess.Popen(cmd, stdout=subprocess.PIPE, shell=True)
  bib = p.communicate()[0].decode("utf-8").strip()[:-1]
  #print(bib.encode("utf-8").strip())
  #raise SystemExit

  # extract components of entry
  try:
    entry = re.sub('}$', '', bib.split('{',1)[0].split('@')[1]).capitalize()
  except:
    entry = False
  author = get_value(bib, 'author')
  journal = get_value(bib, 'journal').replace('&', '{\&}')
  year = get_value(bib, 'year')
  month = get_value(bib, 'month').capitalize()
  volume = get_value(bib, 'volume')
  number = get_value(bib, 'number')
  pages = get_value(bib, 'pages')
  doi = get_value(bib, 'DOI')
  publisher = get_value(bib, 'publisher').replace('&', '{\&}')
  title = get_value(bib, 'title').replace('&', '{\&}')

  # simplify author for key
  author_simple = author.split(', ',1)[0].replace('}','').replace('{','')
  author_simple = re.sub('\\\.','',author_simple)
  author_simple = re.sub(' *','',author_simple)
  if author_simple.isupper():
    author_simple = author_simple.lower().capitalize()

  # reorder author
  reauthor = []
  if len(author) > 0:
    for i in author.split(' and '):
      if i == u'et al.':
        name = i
      else:
        name = i.split(',', 1)[1].strip() + ' ' + i.split(',', 1)[0].strip()
      reauthor.append(name)
    author = ' and '.join(reauthor)

  # remove periods from author initials
  author = author.replace('.',' ')
  author = author.replace('  ',' ')

  # enclose consecutive and title caps in braces
  if len(publisher) > 0:
    publisher = re.sub('([A-Z]{2,})', r'{\1}', publisher)
  if len(title) > 0:
    title = title[0] + re.sub('([A-Z]{1,})', r'{\1}', title[1:])

  # print entry
  if entry:
    sys.stdout.write('@%s{%s%s,\n' % (entry, author_simple, year))
    sys.stdout.write('  author       = {%s},\n' % author)
    sys.stdout.write('  title        = {%s},\n' % title)
    sys.stdout.write('  journal      = {%s},\n' % journal)
    sys.stdout.write('  year         = {%s},\n' % year)
    sys.stdout.write('  volume       = {%s},\n' % volume)
    sys.stdout.write('  number       = {%s},\n' % number)
    if "n/a" not in pages:
      sys.stdout.write('  pages        = {%s},\n' % pages)
    sys.stdout.write('  doi          = {%s},\n' % doi)
    sys.stdout.write('  publisher    = {%s},\n' % publisher)
    sys.stdout.write('}\n')
    sys.stdout.write('\n')