comparing data by difference ratio

test.py

from difflib import SequenceMatcher
from mturk.models import Hit
from message.models import Email

email_ids = [m.email_id for m in Hit.objects.all()]
emails = [e.subject for e in Email.objects.filter(id__in=email_ids)]

res = {}
for e in emails[:100]:
    f = False
    if not res:
        res[e] = [e]
        continue
    for k in res.keys():
        pa = SequenceMatcher(None, e, k)
        if pa.ratio() >= 0.5:
            res.setdefault(k, [e])
            res[k].append(e)
            f = True
            break
    if not f:
        res[e] = [e]

for l in res.values():
    print l[0]