identify_timezone

identify_timezone.py

#-*-coding:utf8-*-
from study.models import *
from BeautifulSoup import BeautifulSoup
import re,datetime

exclude_words = [u'1:1', u'명', u'문장', u'시간', u'PT', u'세', u'일', u'월']

time_regex = [ur'((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*((?:저녁|아침)(?:시간)?|명|AM|PM|오(?:전|후))?\s*((?:-|–|~|부터|또는))\s*((?:저녁|아침)(?:시간?)?|AM|PM|오(?:전|후))?\s*([0-2]?\d)\s*([시:])\s*((?:[0-5]?\d분?|반))?\s*((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?',
              ur'((?:저녁|아침)(?:시간)?|오(?:전|후))\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*((?:-|–|~|부터|또는))\s*((?:저녁|아침)(?:시간?)?|오(?:전|후))?\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?',
              ur'((?:저녁|아침)(?:시간)?|오(?:전|후))?\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*((?:-|–|~|부터|또는))\s*((?:저녁|아침)(?:시간?)?|오(?:전|후))\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?',   
              ur'([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*(AM|PM)\s*((?:-|–|~|부터|또는))\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*(AM|PM)?',
              ur'([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*(AM|PM)?\s*((?:-|–|~|부터|또는))\s*([0-2]?\d)\s*([시:])?\s*((?:[0-5]?\d분?|반))?\s*(AM|PM)',
              ur'((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?\s*([0-2]?\d)\s*([시:]\s*(?:[0-5]?\d분|반))\s*((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?',
              ur'((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?\s*([0-2]?\d시간?)\s*((?:저녁|아침)(?:시간)?|AM|PM|오(?:전|후))?']

def get_clean_content(study):
    content = BeautifulSoup(re.sub(r'<br\s*(>)*(/>)*','\n',study.content)).text.replace('&nbsp','')
    content = re.sub(r'&#8211;','-', content)
    return re.sub(ur'^(?:01[016789]{1}|02|0[3-9]{1}[0-9]{1})[-.]?[0-9]{3,4}[-.\s]?[0-9]{4}$', '', content, re.M)

def get_study_list():
    return Study.objects.filter(write_time = datetime.datetime(year=2013, month=5, day=03))

def identify_time_zone(time):
    time = re.sub(ur'\s*','',time)
    hour = re.findall(ur'[0-2]?\d',time)
    if hour:
        hour = hour[0]
        try:
            hour = int(hour)
        except ValueError:
            return 'etc'
        if  u'오전' in time or u'AM' in time or u'am' in time or u'아침' in time or (hour >= 8 and hour < 12):
             return 'ante meridiem'
        elif ((u'오후' in time or u'PM' in time or u'pm' in time or u'저녁' in time) and (hour >= 0 and hour < 6) or (hour >= 12 and hour < 18)) or (hour >= 0 and hour < 6) or (hour >= 12 and hour < 18):
            return 'post meridiem'
        elif ((u'오후' in time or u'PM' in time or u'pm' in time or u'저녁' in time) and  (hour >= 6 and hour < 8) or (hour >= 18 and hour < 24)) or (hour >= 6 and hour < 8) or (hour >= 18 and hour < 24):
            return 'evening'
        else:
            return 'etc'
    
def get_time(study):
    global study_count
    content = get_clean_content(study)
    for regex in time_regex:
        is_real_time = None
        found_time = re.findall(regex, content, re.I)
        for time in found_time:
            is_real_time = True
            for word in exclude_words:
                time = ''.join(time)
                if word in time:
                    is_real_time = False
                    break
            if is_real_time:
                study_count +=1
                time_zone = identify_time_zone(time)
                print 'Study ID : %s'%study.id
                print 'Time : %s'%time
                print 'Meridiem : %s'%time_zone
                break
        if found_time and is_real_time:
            break
        

def test1():
    global study_count
    study_count = 0
    study_list = get_study_list()  
    for study in study_list:
        get_time(study)
    percent = (float(study_count) / float(len(study_list))) * 100.0
    print "study_count : %d  Percent : %.2f"%(study_count, percent)               
    
def test2(start, end):
    global study_count
    study_count = 0
    study_list = get_study_list()  
    for study in study_list[start:end]:
        get_time(study)
    percent = (float(study_count) / float((end-1)-start)) * 100.0
    print "study_count : %d  Percent : %.2f"%(study_count, percent)