#!/usr/bin/env python# -*- coding: UTF-8 -*import time

from selenium import webdriverfrom selenium.webdriver.support.ui import WebDriverWaitfrom selenium.common.exceptions import NoSuchElementException, TimeoutExceptionfrom selenium.webdriver.common.action_chains import ActionChains

def scroll_page(url, browser='chrome'):if browser=='chrome':driver = webdriver.Chrome()else:driver = webdriver.Firefox() #右键异常 ActionChains(driver).context_click(e).perform()# driver = webdriver.PhantomJS()  #可截长图！可是加载长网页时间超长？driver.set_window_size(1200, 900)driver.get(url) # Load pagestart = time.time()print driver.title    driver.execute_script("""(function () {var y = 0;var step = 1000;  //100window.scroll(0, 0);function f() {if (y < document.body.scrollHeight) {y += step;window.scroll(0, y);setTimeout(f, 100); //100  递归循环调用} else {if(document.title.indexOf("scroll-done") < 0){  //-1 找不到,还没执行下文的driver.execute_scripty -= step;window.scroll(0, y);setTimeout(f, 100); //100                    //window.scroll(0, 0);//document.title += "scroll-done";}//else{//window.scroll(0, 0);//}}}setTimeout(f, 1000);  //1000})();""")#拉到最后 出现元素“没有更多了”WebDriverWait(driver, 500).until(lambda x: x.find_element_by_xpath('//div[@style="text-align:center"]/em'))  #或者引发 TimeoutException#停止上面的js到终点再上拉 循环driver.execute_script("""(function () {function f() {document.title += " scroll-done";}setTimeout(f, 1000);})();""")# <div class="js-infinite-layout"># <div class="js-infinite-item"># Jetbrains公司正式发布Pycharm 5rst =  driver.find_element_by_xpath('//div[@class="js-infinite-item"][last()]//div[@class="header"]').text #不能在path内部写入 text()print rstif rst != u'Jetbrains公司正式发布Pycharm 5':raise RuntimeError('wrong!!!')print time.time()-startprint driver.title# time.sleep(3)  #执行js也需要时间WebDriverWait(driver, 10).until(lambda x: x.title == u'编程派 | Coding Python scroll-done')  #或者引发 TimeoutExceptionprint driver.title  return driver

import win32conimport win32guiimport win32processimport psutil

def get_hwnds_for_pid (pid):def callback (hwnd, hwnds):if win32gui.IsWindowVisible(hwnd) and win32gui.IsWindowEnabled(hwnd):_, found_pid = win32process.GetWindowThreadProcessId(hwnd)if found_pid == pid:hwnds.append(hwnd)return Truehwnds = []win32gui.EnumWindows(callback, hwnds)return hwndsdef set_process_foreground(pid_part_name, pid_window_text):pids_target = []for pid in psutil.pids():p = psutil.Process(pid)if pid_part_name in p.name().lower(): #'chrome.exe'  'EXCEL.EXE'
            pids_target.append(pid)for pid in pids_target:for hwnd in get_hwnds_for_pid(pid):# 92292 chrome.exe 137328 编程派 | Coding Python - Google Chrome# EXCEL.EXE 857830 Microsoft Excel - Book1.xlsx# 90644 firefox.exe 595556 编程派 | Coding Python scroll-done - Mozilla Firefoxif pid_window_text.encode('gbk') in win32gui.GetWindowText(hwnd):  print pid, psutil.Process(pid).name(), hwnd, win32gui.GetWindowText(hwnd)win32gui.SetForegroundWindow(hwnd)returnraise RuntimeError('process not found')

from autopy import key, mouse
def save_result(driver): time_for_filename = time.strftime('%H%M%S')with open('%s.html'%time_for_filename,'wb') as f:f.write(driver.page_source.encode('utf-8'))# e =  driver.find_element_by_xpath('//img[@src="/static/images/logo.png"]')# ActionChains(driver).context_click(e).perform()  # time.sleep(1.5)# esc 退出右键菜单# key.tap(key.K_ESCAPE)# driver.set_window_position(0,0)# window_position = driver.get_window_position() #  {u'x': 10, u'y': 10}# mouse.move(int(window_position['x']+150), int(window_position['y']+150)) #右键再左键 避免错误点开链接# mouse.click(mouse.RIGHT_BUTTON)     # mouse.click(mouse.LEFT_BUTTON)# 关键是浏览器位于foreground
    set_process_foreground(driver.name,driver.title)key.tap('s', key.MOD_CONTROL)time.sleep(1.5)    key.type_string(time_for_filename)time.sleep(0.5)      key.tap(key.K_RETURN)   driver.save_screenshot('%s.png'%time_for_filename)# driver.close()if __name__ == "__main__":browser = 'chrome'# browser = 'firefox'url = "http://codingpy.com"driver = scroll_page(url,browser)save_result(driver)print 'All DONE'