예제 #1
0
def similarity2(a, b):
    """Calculates similarity of a multi-word strings."""
    alist = filter(bool, _split_words_re.split(a.lower()))
    blist = filter(bool, _split_words_re.split(b.lower()))
    total = 0
    score = 0.0
    if len(alist) > len(blist):
        alist, blist = blist, alist
    for a in alist:
        ms = 0.0
        mp = None
        for position, b in enumerate(blist):
            s = astrcmp(a, b)
            if s > ms:
                ms = s
                mp = position
        if mp is not None:
            score += ms
            if ms > 0.6:
                del blist[mp]
        total += 1
    total += len(blist) * 0.4
    if total:
        return score / total
    else:
        return 0
예제 #2
0
def similarity(a1, b1):
    """Calculates similarity of single words."""
#    return astrcmp(a1, b1)
    """Calculates "smart" similarity of strings ``a`` and ``b``."""
    a2 = normalize(a1)
    if a2:
        b2 = normalize(b1)
    else:
        b2 = ""
    return astrcmp(a2, b2)