Posted By

tamuratetsuya on 02/19/09

Tagged

python

Versions (?)

Last Edited at 02/19/09 08:09am

Statistics

Viewed 1550 times

Favorited by 0 user(s)

Related snippets

äººå·¥ç„¡è„³ã®ã‚½ãƒ¼ã‚¹ã€€ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚³ãƒ³ãƒ†ã‚¹ãƒˆå‹‰å¼·ä¼šã«ã¦

/ Published in: Python

ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚³ãƒ³ãƒ†ã‚¹ãƒˆã®å‹‰å¼·ä¼šã§ç´¹ä»‹ã—ãŸGoogle App Engineã§å‹•ãpythonã‚³ãƒ¼ãƒ‰ã§ã™
Djangoãƒ•ãƒ¬ãƒ¼ãƒ ãƒ¯ãƒ¼ã‚¯ã‚’å‰æã«ã—ã¦ã„ã¾ã™ã€‚
ã‚ã¾ã‚Šã‚³ãƒ¼ãƒ‰ã¨ã—ã¦ã¯å‚è€ƒã«ãªã‚‰ãªã„ã‹ã‚‚çŸ¥ã‚Œã¾ã›ã‚“ãŒã€ã“ã‚Œã ã‘ã§ã‚‚ç°¡å˜ãªäººå·¥ç„¡è„³ãŒã§ãã‚‹ã‚“ã§ã™ã

Highlighted TextExpand

#!-*- coding:utf-8 -*-
import sys
import os
import string
import urllib
import urllib2
import math
import htmlentitydefs
import re
from django.shortcuts import render_to_response
from google.appengine.api import users
from django.http import HttpResponseRedirect
from BeautifulSoup import BeautifulSoup
from google.appengine.api import urlfetch
from base64 import b64encode

#åˆæœŸãƒšãƒ¼ã‚¸
def Mainpage(request):
    return render_to_response('input.html',)

#è³ªå•ã‚’ã‚‚ã‚‰ã£ã¦å›žç”ã‚’å‡ºã™ã¾ã§ã®ãƒ¡ã‚½ãƒƒãƒ‰
def getAnswer(request):
    search_text = ''
    from_user = 'username'
    try:
        search_text=request.POST['query']
        search_text=search_text.encode('utf-8')
    except:pass
    if search_text == '':
        search_text,from_user = GetresponseTwitter()
        #print >>sys.stderr,'search_text=%s'%search_text
        search_text = search_text.encode('utf-8')
    to_ = ''
    from_ = ''
    
    #è³ªå•æ–‡ã‚’ãã®ã¾ã¾æ¤œç´¢ã«ã‹ã‘ã‚‹
    if search_text!='' :
        url = 'http://pcod.no-ip.org/yats/search?query=%s&rss' % (search_text)
        xml_search2=urlfetch.fetch(url).content
        entry = BeautifulSoup(xml_search2)('entry')
        words=[]
        titledict=dict()
        answer = ''
        for en in entry:
            #print >>sys.stderr,en('summary')[0]
            words=en('summary')[0](text = True)[0].split(' ')
            update_before = en('updated')[0](text = True)[0]
            #print >>sys.stderr,status_id
            try:
                regexp = re.compile('username')
                if regexp.search(en('summary')[0](text = True)[0])== None:#tam_botã®ç™ºè¨€ã¯ç„¡è¦–
                    from_=words[0].split('@')[1]
                    to_=words[2].split('@')[1]
                    
                    url='http://pcod.no-ip.org/yats/search?query=user:'+to_+'%20%40'+from_+'%20&lang=ja&rss'
                    #print >>sys.stderr,url
                    xml_search=urlfetch.fetch(url).content
                    title1to2 = BeautifulSoup(xml_search)('entry')
                    for x in title1to2:
                        #titledict[x('updated')[0](text = True)[0]]=x('summary')[0].contents[0]
                        update_after=x('updated')[0](text = True)[0]
                        if  update_after > update_before:
                            #print >>sys.stderr,'update_before=%s update_after=%s'%(update_before,update_after)
                            tmp_answer=x('summary')[0](text = True)[0].split(' ')[3]
                            regexp1 = re.compile('@')
                            if regexp1.search(tmp_answer)==None and tmp_answer!='':
                                print >>sys.stderr,'answer=%s'%tmp_answer
                                answer = tmp_answer
                                break#ä¸€å€‹ã§ã‚‚ã¿ã¤ã‹ã‚Œã°break
                    if tmp_answer != '':
                        break
            except:pass
        #print >>sys.stderr,'answer=%s'%answer
        
        
        if answer != '':#æ¤œç´¢çµæžœãŒã‚ã£ãŸå ´åˆ
            post = '@%s %s'%(from_user,answer)
        else:#æ¤œç´¢çµæžœãŒãªã‹ã£ãŸå ´åˆ
            #############æ–‡ç« ã ã‘ã§å›žç”ãŒå‡ºã›ãªã„å ´åˆ#############
            question_word=GetKeitaiso_mecapi(search_text).encode('utf-8')
            q = {'query':question_word,'lang' : 'ja'}            
            url = 'http://pcod.no-ip.org/yats/search?rss&' + urllib.urlencode(q)
            xml=urlfetch.fetch(url).content
            #print >>sys.stderr,xml
            entry = BeautifulSoup(xml)('entry')
            for en in entry:
                tmp_answer=en('summary')[0](text = True)[0].split(' : ')[1].rstrip()
                regexp = re.compile('tam_bot')
                if regexp.search(en('summary')[0](text = True)[0])== None:#tam_botã®ç™ºè¨€ã¯ç„¡è¦–
                    regexp1 = re.compile('@')
                    if regexp1.search(tmp_answer)==None and tmp_answer!='' and search_text.rstrip()!=tmp_answer.rstrip().encode('utf-8'):
                        answer = tmp_answer
                        #print >>sys.stderr,'tmp_answer=%s'%tmp_answer
                        break
            post = '@%s %s'%(from_user,answer)
            
            #########æœ€å¾Œã¾ã§ã‚ã‹ã‚‰ãªã„å ´åˆ##############
            if answer == '':
                post = u'@%s %s'%(from_user,unicode(search_text,'utf-8'))
    
    #print >>sys.stderr,'post=%s'%post
    Twitter(post)        
    return render_to_response('outanswer.html',{'search_text':search_text,'post':post})

#è³ªå•æ–‡ç« ã‚’å½¢æ…‹ç´ è§£æžã‹ã‘ã¦å˜èªžã‚’å–ã‚Šå‡ºã™
def GetKeitaiso_mecapi(question):
    question_word=''
    word_rank=dict()
    i=0
    q={'format':'xml','sentence':question,'response':'','filter':'uniq'}
    url='http://mimitako.net/api/mecapi.cgi?'+ urllib.urlencode(q)
    xml_question=urlfetch.fetch(url).content
    #print >>sys.stderr,BeautifulSoup(xml_question)('mecabresult')
    ngword=[u'ä½•',u'èª°']
    for mlist in BeautifulSoup(xml_question)('mecabresult'):
        for m in mlist('word'):
            term=m('feature')[0].contents[0].split(',')
            word=m('surface')[0]
            word=tr(u'!"#$%&()=-^~\|[]{}@:`*;+,.<>/\?_ã€‚ã€ãƒ»', u'                                   ',word)
            word=string.replace(word, ' ', '')
            #print >>sys.stderr,'%s %s %s %s %s %s %s'%(term[0],term[1],term[2],term[3],term[4],term[5],word)
            if (term[0]==u'åè©ž' or term[0]==u'å‹•è©ž' or term[0]==u'å½¢å®¹è©ž') and word != '' and word not in ngword:
                cntwd = int(countWord(word))
                #print >>sys.stderr,'%s %s %s'%(term[0],word,cntwd)
                if i<3 and cntwd<100000000:
                    question_word = u'%sã€€%s'%(question_word,word)
                    print >>sys.stderr,'%s %s %s'%(term[0],word,cntwd)
                    i=i+1
    return question_word

#Twitterã¸ã®postãƒ¡ã‚½ãƒƒãƒ‰
def Twitter(post):
    #basic auth
    username='username'
    password='password'
    url = 'http://twitter.com/statuses/update.json'
    payload = urllib.urlencode({"status": post.encode("utf-8")})
    base64string =b64encode("%s:%s" % (username, password))
    headers = {"Authorization": "Basic %s" % base64string}
    json_result = urlfetch.fetch(url, payload=payload, method=urlfetch.POST, headers=headers).content
    return 1

#Twitterã‹ã‚‰ã‚ã‚‹ãƒ¦ãƒ¼ã‚¶ã®ç›´è¿‘ã®è³ªå•ã‚’ã¨ã£ã¦ãã‚‹
def GetresponseTwitter():
    #basic auth
    username='username'
    password='password'

url = 'http://twitter.com/statuses/replies.xml'
    base64string =b64encode("%s:%s" % (username, password))
    headers = {"Authorization": "Basic %s" % base64string}
    xml_result=urlfetch.fetch(url, payload='', method=urlfetch.POST, headers=headers).content
    result_comment=htmlentity2unicode(BeautifulSoup(xml_result)('text')[0].contents[0].split(' ')[1])
    from_user=BeautifulSoup(xml_result)('screen_name')[0].contents[0]
    return result_comment,from_user

#googleã®æ¤œç´¢çµæžœæ•°ã‚’è¦‹ã‚‹
def countWord(word):
    q = {#'q':'\"'+question+'\"',
        'q':word.encode('utf-8'),
        'v' : 1.0,
        'hl' : 'ja',
        'rsz': 'small'
    }
    url=u'http://ajax.googleapis.com/ajax/services/search/web?'+ urllib.urlencode(q)
    json_result=urlfetch.fetch(url).content
    json_result = string.replace(json_result, 'null', '"null"')
    try:
        dict_result=eval('dict('+json_result+')')        
        return dict_result['responseData']['cursor']['estimatedResultCount']
    except:
        return 0

def tr(pattern, repl, string):
    m = dict(zip(pattern, repl))
    return ''.join((m.get(c,c) for c in string))

#æ•°å€¤æ–‡å—å‚ç…§â†’æ–‡å—ã¸
def htmlentity2unicode(text):
    # æ£è¦è¡¨ç¾ã®ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«
    reference_regex = re.compile(u'&(#x?[0-9a-f]+|[a-z]+);', re.IGNORECASE)
    num16_regex = re.compile(u'#x\d+', re.IGNORECASE)
    num10_regex = re.compile(u'#\d+', re.IGNORECASE)
    
    result = u''
    i = 0
    while True:
        # å®Ÿä½“å‚ç…§ or æ–‡å—å‚ç…§ã‚’è¦‹ã¤ã‘ã‚‹
        match = reference_regex.search(text, i)
        if match is None:
            result += text[i:]
            break
        
        result += text[i:match.start()]
        i = match.end()
        name = match.group(1)
        
        # å®Ÿä½“å‚ç…§
        if name in htmlentitydefs.name2codepoint.keys():
            result += unichr(htmlentitydefs.name2codepoint[name])
        # æ–‡å—å‚ç…§
        elif num16_regex.match(name):
            # 16é€²æ•°
            result += unichr(int(u'0'+name[1:], 16))
        elif num10_regex.match(name):
            # 10é€²æ•°
            result += unichr(int(name[1:]))
        
    return result

URL: http://www.team-lab.com/news/index.php?itemid=469

Report this snippet Tweet