模拟浏览器抓屏功能

gistfile1.txt

#!/usr/bin/env python
# -*- coding: utf-8 -*-
import sys, thread
import gtk, webkit
import ctypes
import warnings
import urllib
import json
from sys import stdin
from time import sleep

use_proxy = False

libgobject = ctypes.CDLL('/usr/lib/x86_64-linux-gnu/libgobject-2.0.so.0')
libwebkit = ctypes.CDLL('/usr/lib/x86_64-linux-gnu/libsoup-2.4.so.1')
libsoup = ctypes.CDLL('/usr/lib/x86_64-linux-gnu/libsoup-2.4.so.1')
libwebkit = ctypes.CDLL('/usr/lib/x86_64-linux-gnu/libwebkitgtk-1.0.so')

session = libwebkit.webkit_get_default_session()

if use_proxy:
    proxy_uri = libsoup.soup_uri_new('http://web-proxy.oa.com:8080') # set your proxy url
    libgobject.g_object_set(session, "proxy-uri", proxy_uri, None)

cookiejar = libsoup.soup_cookie_jar_text_new('/home/daweiqian/temp/baidu_login/cookies.txt',False)
libsoup.soup_session_add_feature(session, cookiejar)

gtk.gdk.threads_init() # suggested by Nicholas Herriot for Ubuntu Koala
win = gtk.Window(gtk.WINDOW_TOPLEVEL)
view = webkit.WebView()
win.add(view)
win.set_size_request(1000,600)
win.show_all()
win.connect("destroy", lambda q: gtk.main_quit())

keyword_list =["张碧晨","王思聪","王紫绮","郭美美","grace","贞操妹妹","王林","张雨绮","王自如","王珂","曹格"]
keyword_index = -1


def finished_loading(view, frame):
    print "in loading finished with index %d" % keyword_index 
    sleep(10)
    view.execute_script('oldtitle=document.title;document.title=document.documentElement.innerHTML;')
    html = view.get_main_frame().get_title()
    view.execute_script('document.title=oldtitle;')
    width, height = win.get_size()
    pixbuf = gtk.gdk.Pixbuf(gtk.gdk.COLORSPACE_RGB, False, 8, width, height)
    screenshot = pixbuf.get_from_drawable(win.window, win.get_colormap(),0, 0, 0, 0, width, height)
    screenshot.save(keyword_list[keyword_index] + '.png', 'png')
    open_page()

view.connect("load-finished", finished_loading)

def open_page():
    global keyword_index
    keyword_index = keyword_index + 1
    if keyword_index >= len(keyword_list):
        gtk.main_quit()
        return 
    keyword = keyword_list[keyword_index]
    search_param = {'tpl':"trend", 'word':keyword.decode("utf-8").encode("gb2312")}
    url = 'http://index.baidu.com/?' + urllib.urlencode(search_param)
    view.open(url)

open_page()

gtk.main()