【VSRC唯科普】简易图像识别与文字处理（第11/14篇）

# encoding=utf8
import time
from urllib.request import urlretrieve
import subprocess
from selenium import webdriver

driver = webdriver.Chrome(executable_path='C:\chromedriver.exe')
driver.get(
    "https://www.amazon.com/War-Peace-Leo-Nikolayevich-Tolstoy/dp/1427030200"
)
time.sleep(2)

ele = driver.find_element_by_id('sitbLogoImg')
driver.execute_script("arguments[0].click();", ele)
time.sleep(5)

imageList = set()
while "pointer" in driver.find_element_by_id('sitbReaderRightPageTurner').get_attribute("style"):
    driver.find_element_by_id("sitbReaderRightPageTurner").click()
    time.sleep(2)
    pages = driver.find_elements_by_xpath("//div[@class='pageImage']/div/img")
    for page in pages:
        image = page.get_attribute("src")
        imageList.add(image)
        if len(imageList) > 3:  
            break
driver.quit()

for image in sorted(imageList):
    urlretrieve(image,"page.jpg")
    p = subprocess.Popen(["D:\\Program Files\\Tesseract-OCR\\tesseract.exe","page.jpg","page"],stdout=subprocess.PIPE,stderr=subprocess.PIPE)
    p.wait()
    f = open("page.txt","r")
    print (f.read())

我们通过一个在线的带验证码的评论表单 http://www.pythonscraping.com/humans-only ，来演示如何使用机器人破解验证码。

实现代码如下：

# encoding=utf8
from urllib.request import urlretrieve
from urllib.request import urlopen
from bs4 import BeautifulSoup
import subprocess
import requests
from PIL import Image
from PIL import ImageOps


def cleanImage(imagePath):
    image = Image.open(imagePath)
    image = image.point(lambda x: 0 if x <143 else 255)
    borderImage = ImageOps.expand(image, border=20, fill='white')
    borderImage.save(imagePath)


html = urlopen('http://www.pythonscraping.com/humans-only')
bsObj = BeautifulSoup(html,"html.parser")

imageLocation = bsObj.find("img",{'title':"Image CAPTCHA"})["src"]
formBuildId = bsObj.find("input",{"name":"form_build_id"})["value"]
captchaSid = bsObj.find("input",{"name":"captcha_sid"})["value"]
captchaToken = bsObj.find("input",{"name":"captcha_token"})["value"]

captchaUrl = "http://pythonscraping.com"+imageLocation
urlretrieve(captchaUrl,"captcha.jpg")
cleanImage("captcha.jpg")
p = subprocess.Popen(["D:\\Program Files\\Tesseract-OCR\\tesseract.exe","captcha.jpg","captcha"],stdout=subprocess.PIPE,stderr=subprocess.PIPE)
p.wait()
f = open("captcha.txt","r")

captchaResponse = f.read().replace(" ","").replace("\n","")
print ("Captcha solution attempt: "+captchaResponse)

if len(captchaResponse) == 5:
    params = {"captcha_token":captchaToken,"captcha_sid":captchaSid,
              "form_id":"comment_node_page_form","form_build_id":formBuildId,
              "captcha_response":captchaResponse,"name":"Ryan Mitchell",
              "subject":"I come to seek the Grail",
              "comment_body[und][0][value]":
              "...and I am definitely not a bot"}
    r = requests.post ("http://www.pythonscraping.com/comment/reply/10",data=params)
    responseObj = BeautifulSoup(r.text)
    if responseObj.find("div",{"class":"messages"}) is not None:
        print(responseObj.find("div",{"class":"messages"}).get_text())
    else:
        print ("There was a problem reading the CAPTCHA correctly!")

1、https://www.w3school.com.cn/xpath/xpath_intro.asp