住所の構造化 (revised from http://kbyanc.blogspot.com/2007/07/parsing-japanese-addresses.html)

transduce_jpaddress.py

import re
import unicodedata
_address_re = re.compile(r'(京都府|東京都|.+?[道府県])?(.+郡)?(.+?[市町村])?(.+?区)?(.*)')
 
def transduce_jpaddress(addrstr):
	addrstr_n = kansuji_to_arabic_numeral(addrstr)
    m = _address_re.match(addrstr_n)
    if m is not None:
        (prefecture, county, city, ward, address) = m.groups()
        address = address.strip()
        # 東京都 is both a city and a prefecture.
        if prefecture == '東京都' and city is None:
            city = prefecture
        return {'prefecture': prefecture,
                'county': county,
                'city': city, 
                'ward': ward, 
                'address': address}
    return addrstr