Extract data from fml text

fml_extract.py

# -*- coding: utf-8 -*-
import codecs

sql = 'insert into fml(id, author, gender, address, content, date) values(%s, \'%s\', \'%s\', \'%s\', \'%s\', \'%s\');\n'

def getPrefix(tx):
    return tx[tx.rfind('-') + 1: tx.rfind('.')]

def getDate(tx):
    return tx[tx.rfind('/') + 1:tx.rfind('-')].replace('-', '/')

def getAuthor(tx):
    l = tx.find(u'（')
    if l != -1:
        return tx[:l].strip()
    l = tx.find(u'(')
    if l != -1:
        return tx[:l].strip()
    return tx[:tx.find(' - ')].strip()

def getGender(tx):
    tx = tx[:tx.find(' - ')]
    if tx.find(u'匿') != -1:
        return u'un'
    elif tx.find(u'女') != -1:
        return u'fm'
    else:
        return u'ma'

def getAddr(tx):
    return tx[tx.find('- ') + 2:]

def extra(src):    
    txt = codecs.open(src, 'r' , 'utf-8').read()
    txt = unicode(txt)
    p = txt.split(u'###\r\n')
    pre = getPrefix(src)
    i = 99
    of = open('sql/%s.txt' %pre, 'w')
    for x in p:
        l = x.split('\r\n')
        ui = pre + str(i)
        au = getAuthor(l[0])
        ge = getGender(l[0])
        ad = getAddr(l[0])
        co = l[1].strip()
        da = getDate(src)
        print 'id: ' + ui
        print 'author: ' + au
        print 'gender: ' + ge
        print 'addr: ' + ad
        print 'content: ' + co
        print 'date: ' + da
        print ''
        i -= 1
        cmd =  (sql %(ui, au, ge, ad, co, da)).encode('utf-8')    
        of.write(cmd)
    of.close()

ss = open('path.txt').read().split('\n')
for s in ss:
    extra('data/' + s)