反爬蟲字體混淆

Win7下的python: 通過amd64的二進制文件安裝, 位置在 C:\Users\Milton\AppData\Local\Programs\Python\Python37\ 

安裝pip3

C:\Users\Milton\AppData\Local\Programs\Python\Python37\Scripts\easy_install.exe pip

通過pip3安裝fonttools

C:\Users\Milton\AppData\Local\Programs\Python\Python37\Scripts\pip3.7.exe install fonttools

代碼例子, 貓眼的字體反爬蟲已經升級了, 不再是簡單的順序關系, 這樣取到的值依然是錯的. 需要根據根據各個字體里面的字體定義TTGlyph.contour的值去判斷具體數字.

import requests
import re
import os
from fontTools.ttLib import TTFont


class MaoYan(object):
    def __init__(self):
        self.url = 'https://maoyan.com/films/42964'
        self.headers = {
            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36"
        }

    # 發送請求獲得響應
    def get_html(self, url):
        response = requests.get(url, headers=self.headers)
        return response.content

    # 創建 self.font 屬性
    def create_font(self, font_file):
        # 列出已下載文件
        file_list = os.listdir('./fonts')
        # 判斷是否已下載
        if font_file not in file_list:
            # 未下載則下載新庫
            print('不在字體庫中, 下載:', font_file)
            url = 'http://vfile.meituan.net/colorstone/' + font_file
            new_file = self.get_html(url)
            with open('./fonts/' + font_file, 'wb') as f:
                f.write(new_file)

        # 打開字體文件，創建 self.font屬性
        self.font = TTFont('./fonts/' + font_file)
        self.font.saveXML('./fonts/' + font_file + '.xml')

    # 把獲取到的數據用字體對應起來，得到真實數據
    def modify_data(self, data):
        print(data);
        # 獲取 GlyphOrder 節點
        gly_list = self.font.getGlyphOrder()
        # 前兩個不是需要的值，截掉
        gly_list = gly_list[2:]
        # 枚舉, number是下標，正好對應真實的數字，gly是亂碼
        for number, gly in enumerate(gly_list):
            # 把 gly 改成網頁中的格式
            gly = gly.replace('uni', '&#x').lower() + ';'
            # 如果 gly 在字符串中，用對應數字替換
            if gly in data:
                data = data.replace(gly, str(number))
        # 返回替換后的字符串
        return data

    def start_crawl(self):
        html = self.get_html(self.url).decode('utf-8')

        # 正則匹配字體文件
        font_file = re.findall(r'vfile\.meituan\.net\/colorstone\/(\w+\.woff)', html)[0]
        print(font_file);
        self.create_font(font_file)

        # 正則匹配星級
        star = re.findall(r'<span class="index-left info-num ">\s+<span class="stonefont">(.*?)</span>\s+</span>', html)[0]
        star = self.modify_data(star)

        # 正則匹配評論的人數
        people = ''.join(re.findall(r'''<span class='score-num'><span class="stonefont">(.*?萬)</span>(人評分)</span>''', html)[0])
        people = self.modify_data(people)

        # 正則匹配累計票房
        ticket_number = ''.join(re.findall(r'''<span class="stonefont">(.*?)</span><span class="unit">(億)</span>''', html)[0])
        ticket_number = self.modify_data(ticket_number)

        print('用戶評分: %s 星' % star)
        print('評分人數: %s' % people)
        print('累計票房: %s' % ticket_number)


if __name__ == '__main__':
    maoyan = MaoYan()
    maoyan.start_crawl()

.