# -*- coding: utf-8 -*-
#登陆知乎抓取个人收藏 然后保存为word
import sys
reload(sys) 
sys.setdefaultencoding('utf-8')
import urllib
import urllib2
import cookielib
import string
import re
from bs4 import BeautifulSoup
from docx import Document
from docx import *
from docx.shared import Inches
from sys import exit
import os
 
#这儿是因为在公司上网的话需要使用socket代理
#import socks
#import socket
#socks.setdefaultproxy(socks.PROXY_TYPE_SOCKS5,"127.0.0.1",8088)
#socket.socket =socks.socksocket
 
loginurl='http://www.zhihu.com/login'
 
headers = {'User-Agent' : 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/34.0.1847.116 Safari/537.36',} 
 
postdata={
 '_xsrf': 'acab9d276ea217226d9cc94a84a231f7',
 'email': '',
 'password': '',
 'rememberme':'y'  
}
 
if not os.path.exists('myimg'):
  os.mkdir('myimg')
if os.path.exists('123.docx'):
  os.remove('123.docx')
if os.path.exists('checkcode.gif'):
  os.remove('checkcode.gif')
 
mydoc=Document()
questiontitle=''
#----------------------------------------------------------------------
def dealimg(imgcontent):
  soup=BeautifulSoup(imgcontent)
  try:
    for imglink in soup.findAll('img'):
      if imglink is not None :
        myimg= imglink.get('src')
        #print myimg
        if myimg.find('http')>=0:
          imgsrc=urllib2.urlopen(myimg).read()
          imgnamere=re.compile(r'http\S*/')
          imgname=imgnamere.sub('',myimg)
          #print imgname
          with open(u'myimg'+'/'+imgname,'wb') as code:
            code.write(imgsrc)
            mydoc.add_picture(u'myimg/'+imgname,width=Inches(1.25))
  except:
    pass
  strinfo=re.compile(r'<noscript>[\s\S]*</noscript>')
  imgcontent=strinfo.sub('',imgcontent)
  strinfo=re.compile(r'<img class[\s\S]*</>')
  imgcontent=strinfo.sub('',imgcontent)
  #show all
  strinfo=re.compile(r'<a class="toggle-expand[\s\S]*</a>')
  imgcontent=strinfo.sub('',imgcontent)
 
  strinfo=re.compile(r'<a class=" wrap external"[\s\S]*rel="nofollow noreferrer" target="_blank">')
  imgcontent=strinfo.sub('',imgcontent)
  imgcontent=imgcontent.replace('<i class="icon-external"></i></a>','')
 
 
  imgcontent=imgcontent.replace('</b>','').replace('</p>','').replace('<p>','').replace('<p>','').replace('<br>','')
  return imgcontent
   
 
 
 
 
def enterquestionpage(pageurl):
  html=urllib2.urlopen(pageurl).read()
  soup=BeautifulSoup(html)
  questiontitle=soup.title.string
  mydoc.add_heading(questiontitle,level=3)
  for div in soup.findAll('div',{'class':'fixed-summary zm-editable-content clearfix'}):
    #print div
    conent=str(div).replace('<div class="fixed-summary zm-editable-content clearfix">','').replace('</div>','')
     
    conent=conent.decode('utf-8')
    conent=conent.replace('<br/>','\n')
     
    conent=dealimg(conent)
    ###这一块弄得太复杂了 有时间找找看有没有处理html的模块
    conent=conent.replace('<div class="fixed-summary-mask">','').replace('<blockquote>','').replace('<b>','').replace('<strong>','').replace('</strong>','').replace('<em>','').replace('</em>','').replace('</blockquote>','')
    mydoc.add_paragraph(conent,style='BodyText3')
    """file=open('222.txt','a')
    file.write(str(conent))
    file.close()"""
     
 
def entercollectpage(pageurl):
  html=urllib2.urlopen(pageurl).read()
  soup=BeautifulSoup(html)
  for div in soup.findAll('div',{'class':'zm-item'}):
    h2content=div.find('h2',{'class':'zm-item-title'})
    #print h2content
    if h2content is not None:
      link=h2content.find('a')
      mylink=link.get('href')
      quectionlink='http://www.zhihu.com'+mylink
      enterquestionpage(quectionlink)
      print quectionlink    
 
 
 
def loginzhihu():
  postdatastr=urllib.urlencode(postdata)
  '''
  cj = cookielib.LWPCookieJar()
  cookie_support = urllib2.HTTPCookieProcessor(cj)
  opener = urllib2.build_opener(cookie_support,urllib2.HTTPHandler)
  urllib2.install_opener(opener)
  '''
  h = urllib2.urlopen(loginurl)
  request = urllib2.Request(loginurl,postdatastr,headers)
  request.get_origin_req_host
  response = urllib2.urlopen(request)
  #print response.geturl()
  text = response.read()
 
 
  collecturl='http://www.zhihu.com/collections'
  req=urllib2.urlopen(collecturl)
  if str(req.geturl())=='http://www.zhihu.com/?next=%2Fcollections':
    print 'login fail!'
    return
  txt=req.read()
 
  soup=BeautifulSoup(txt)
  count=0
  divs =soup.findAll('div',{'class':'zm-item'})
  if divs is None:
    print 'login fail!'
    return
  print 'login ok!\n'
  for div in divs:
     
    link=div.find('a')
    mylink=link.get('href')
    collectlink='http://www.zhihu.com'+mylink
    entercollectpage(collectlink)
    print collectlink
    #这儿是当时做测试用的,值获取一个收藏
    #count+=1
    #if count==1:
    #  return
     
 
def getcheckcode(thehtml):
  soup=BeautifulSoup(thehtml)
  div=soup.find('div',{'class':'js-captcha captcha-wrap'})
  if div is not None:
    #print div
    imgsrc=div.find('img')
    imglink=imgsrc.get('src')
    if imglink is not None:
      imglink='http://www.zhihu.com'+imglink
 
      imgcontent=urllib2.urlopen(imglink).read()
      with open('checkcode.gif','wb') as code:
        code.write(imgcontent)
      return True
    else:
      return False
  return False
 
 
if __name__=='__main__':
   
  import getpass
  username=raw_input('input username:')
  password=getpass.getpass('Enter password: ') 
   
  postdata['email']=username
  postdata['password']=password
  postdatastr=urllib.urlencode(postdata)
  cj = cookielib.LWPCookieJar()
  cookie_support = urllib2.HTTPCookieProcessor(cj)
  opener = urllib2.build_opener(cookie_support,urllib2.HTTPHandler)
  urllib2.install_opener(opener)
 
  h = urllib2.urlopen(loginurl)
  request = urllib2.Request(loginurl,postdatastr,headers)
  response = urllib2.urlopen(request)
  txt = response.read()
 
  if getcheckcode(txt):
    checkcode=raw_input('input checkcode:')
    postdata['captcha']=checkcode
    loginzhihu()
    mydoc.save('123.docx')
  else:
    loginzhihu()
    mydoc.save('123.docx')
 
  print 'the end'
  raw_input()