# Public domain username sanitizer by Lenny Domnitser, 2007
# Found at http://domnit.org/software
#
# "It's less dumb about Unicode than most of what you see on the web"
# -- self-proclaimed
#
# This would be a good place to list modifications to the original code:
# * _____

import re
import unicodedata

def username(name, min_length=3, max_length=30):
    '''username(name) -> (what_you_put_in_the_db, what_you_display)'''
    name = unicode(name)
    name = name.strip()
    if min_length and len(name) < min_length or \
       max_length and len(name) > max_length:
        raise ValueError('illegal username: bad length', 1)
    canonical_name = []
    only_script = ''
    for c in name:
        cat0 = unicodedata.category(c)[0]
        char_name = unicodedata.name(c)
        if cat0 == 'L': # letters
            if cat0 == 'L':
                script = char_name.split(' ', 1)[0]
                if only_script and only_script != script:
                    raise ValueError('illegal username: mixed scripts', 2)
                only_script = script
            canonical_name.append(c)
        elif cat0 == 'N':
            if c in '0123456789':
                simple_digits = c
            else:
                try:
                    digit_name = re.findall(r'(?:DIGIT|NUMBER) ([A-Z]+)',
                                            char_name)[0]
                except IndexError:
                    raise ValueError('illegal username: illegal characters', 3)
                simple_digits = {'ZERO': '0', 'ONE': '1', 'TWO': '2',
                                 'THREE': '3', 'FOUR': '4', 'FIVE': '5',
                                 'SIX': '6', 'SEVEN': '7', 'EIGHT': '8',
                                 'NINE': '9', 'TEN': '10', 'ELEVEN': '11',
                                 'TWELVE': '12', 'THIRTEEN': '13',
                                 'FOURTEEN': '14', 'FIFTEEN': '15',
                                 'SIXTEEN': '16', 'SEVENTEEN': '17',
                                 'EIGHTEEN': '18', 'NINETEEN': '19',
                                 'TWENTY': '20'}[digit_name]
            canonical_name.append(simple_digits)
        elif cat0 == 'P' or c == ' ': # numbers, punctuation, space
            pass
        else:
            raise ValueError('illegal username: illegal characters', 3)
    canonical_name = ''.join(canonical_name)
    canonical_name = canonical_name.lower()
    return canonical_name, name