Fixes docstrings for unicode

rep_script.py

#!/usr/bin/env python
import re
import sys
from builtins import open

squote_exp = re.compile(r"(?P<ms>u?'(?:[^'\n\r\\]|(?:'')|(?:\\x[0-9a-fA-F]+)|(?:\\.))*')",
                        re.UNICODE)

def add_upref(m):
    ms = m.group('ms')
    if ms[0] == 'u':
        return ms
    return 'u' + ms


def add_flag(matchobj):
    firstline = matchobj.group('firstline')
    secondline = matchobj.group('secondline')
    flag = '\n'

    if '"""' in secondline or firstline.endswith('\\'):
        return firstline + flag + secondline + '\n'

    if '+IGNORE_UNICODE' not in firstline and '+SKIP' not in firstline:
        flag = ' +IGNORE_UNICODE' + flag
        if '# doctest:' not in firstline:
            flag = ' # doctest:' + flag

    secondline = squote_exp.sub(add_upref, secondline)

    return firstline + flag + secondline + '\n'

def main():
    from argparse import ArgumentParser, RawTextHelpFormatter
    parser = ArgumentParser(description='Regex doctests fixer',
                            formatter_class=RawTextHelpFormatter)

    parser.add_argument('input_file', action='store', help='the file to be fixed')
    args = parser.parse_args()

    with open(args.input_file, 'r') as myfile:
        data=''.join(myfile.read())

    flag_expr = re.compile(r'(?P<firstline>(?:\ *\>{3}\ .*))(?:\n)(?P<secondline>(?:\ )*(?<!\>{3}\ )([\w\d\[\]\{\}]*(?<!u)[\'"]{1}.*[\'"]{1}.*\n?)+)\n', re.UNICODE)

    with open(args.input_file, 'w') as myfile:
        myfile.write(flag_expr.sub(add_flag, data))


if __name__ == '__main__':
    main()