Basic uscode citation-to-neoid function.

getcitations.py

import re


def citations(text):
    rgx = u'(\\d+)\\s+U\\.?S\\.?C\\.?\\s*\xa7*\\s\\s*([\\d\\w\\-\\\u2013\\.\u2013]+)(\\([\\d\\w\\-\\\u2013\\.\u2013()]+\\))*'
    matches = []
    for match in re.finditer(rgx, text):
        title, section, path = match.groups()
        section = section.strip(u'\u2013.- ')
        start, end = match.span()
        ident = '/us/usc/t%s/s%s' % (title, section)
        if path:
            pathsegs = path.strip('()')
            pathsegs = re.split(r'[ ()]', pathsegs)
            pathsegs = filter(None, pathsegs)
            ident += '/' + '/'.join(pathsegs)

        matches.append((start, end, ident))

    return matches