【原创】python反爬策略--字体反爬（政策网）

字体反爬，也是一种常见的反爬技术，这些网站采用了自定义的字体文件，在浏览器上正常显示，但是爬虫抓取下来的数据要么就是乱码，要么就是变成其他字符。下面我们通过其中一种方式解决字体反爬。

1 字体反爬案例

来源网站：查策网_https://www.chacewang.com/chanye/news。
我们可以看到网站展示的时间日期与html中的时间日期不一致，每次刷新网页，html中的时间都会变化，与实际不一致。

2 使用环境

python3.6 + windows 10专业版 + pycharm 2019.1.3专业版

3 安装python第三方库

pip3 install requests ==2.25.1
pip3 install fontTools ==4.28.5

4 查看woff文件

刷新网页，在network中可以查看到一个woff文件，woff文件未字体加密文件，我们需要解析文件内容，得出加密字体与实际字体间的关系。

5 woff文件解决字体反爬全过程

5.1 调用第三方库

import requests
from fontTools.ttLib import TTFont 

5.2 请求woff链接下载woff文件到本地

# 下载woff
url = "https://web.chace-ai.com/media/fonts/fFz9g4IQsHDyEaXI.woff?version=9.666110168223948"
headers = {
'accept':'*/*',
'accept-encoding':'gzip, deflate, br',
'accept-language':'zh-CN,zh;q=0.9',
'origin':'https://www.chacewang.com',
'referer':'https://www.chacewang.com/',
'user-agent':'Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.93 Safari/537.36',
}
resp = requests.get(url=url, headers=headers)
save_woff = 'demo.woff'
with open(save_woff, "wb") as f:f.write(resp.content)f.close()

5.3 查看woff文件内容，可以通过以下两种方式

5.3.1 通过woff文件解析器，得到结果

5.3.2 将woff文件转为xml文件，此操作也便于我们对字体反爬结果分析

# # woff转xml文件
font = TTFont(save_woff)
save_xml = 'demo.xml'
font.saveXML(save_xml)  # 转换为xml文件

• 得到的xml文件如下：
  
• 一般来说，字体反爬是两个方式表现：
  形式一：GlyphID id与 GlyphID name形成对应关系。
  
  形式二：通过TTGlyph name的xy坐标信息形成文字
  
• 通过对比，我们发现形式一得出的结果明显不是我们想要的，因此我们对形式二进行处理。

5.4 获取所有坐标，通过matplotlib展示结果并对结果进行处理。

以此类推得出所有坐标的信息：

5.5 建立字体反爬后与原字体间对应关系

# 字体反爬坐标信息
dict_data = {'[(96, 74), (158, 38), (231, 38), (333, 38), (445, 198), (445, 346), (444, 345), (443, 346), (393, 251), (275, 251), (179, 251), (54, 382), (54, 487), (54, 599), (191, 742), (300, 742), (413, 742), (541, 565), (541, 395), (541, 185), (376, -42), (231, -42), (151, -42), (96, -16), (151, 500), (151, 423), (229, 334), (298, 334), (358, 334), (439, 415), (439, 474), (439, 555), (359, 663), (292, 663), (231, 663), (151, 570)]':'9','[(75, 98), (153, 38), (251, 38), (329, 38), (421, 115), (421, 179), (421, 323), (215, 323), (151, 323), (151, 403), (212, 403), (394, 403), (394, 538), (394, 663), (255, 663), (175, 663), (104, 609), (104, 699), (179, 742), (278, 742), (375, 742), (493, 641), (493, 560), (493, 411), (341, 369), (341, 367), (423, 358), (519, 260), (519, 186), (519, 84), (371, -42), (249, -42), (141, -42), (75, -1)]':'3','[(518, -29), (57, -29), (57, 54), (277, 274), (368, 365), (430, 472), (430, 528), (430, 593), (358, 662), (289, 662), (188, 662), (96, 576), (96, 673), (185, 742), (304, 742), (407, 742), (525, 632), (525, 538), (525, 468), (449, 332), (343, 227), (170, 58), (170, 56), (518, 56)]':'2','[(92, 88), (168, 38), (251, 38), (329, 38), (425, 125), (425, 271), (328, 351), (235, 351), (191, 351), (115, 344), (115, 730), (488, 730), (488, 645), (204, 645), (204, 430), (247, 433), (268, 433), (388, 433), (523, 311), (523, 205), (523, 94), (379, -42), (252, -42), (145, -42), (92, -10)]':'5','[(503, 632), (444, 663), (380, 663), (281, 663), (162, 491), (161, 343), (163, 343), (217, 447), (335, 447), (433, 447), (551, 319), (551, 213), (551, 101), (415, -42), (310, -42), (194, -42), (63, 139), (63, 305), (63, 506), (236, 742), (377, 742), (457, 742), (503, 720), (167, 219), (167, 144), (246, 38), (312, 38), (375, 38), (454, 130), (454, 202), (454, 281), (380, 368), (311, 368), (247, 368), (167, 280)]':'6','[(557, 168), (460, 168), (460, -29), (367, -29), (367, 168), (14, 168), (14, 230), (349, 730), (460, 730), (460, 244), (557, 244), (367, 244), (367, 562), (367, 596), (369, 640), (367, 640), (360, 621), (338, 581), (113, 244)]':'4','[(544, 696), (247, -29), (148, -29), (430, 645), (51, 645), (51, 730), (544, 730)]':'7','[(49, 335), (49, 536), (183, 742), (309, 742), (550, 742), (550, 354), (550, 162), (414, -42), (292, -42), (176, -42), (49, 151), (147, 340), (147, 38), (301, 38), (452, 38), (452, 345), (452, 663), (304, 663), (147, 663)]':'0','[(219, 370), (84, 433), (84, 553), (84, 637), (211, 742), (310, 742), (401, 742), (518, 644), (518, 565), (518, 439), (377, 371), (377, 369), (543, 309), (543, 165), (543, 70), (406, -42), (284, -42), (183, -42), (57, 69), (57, 157), (57, 302), (219, 368), (421, 554), (421, 605), (358, 664), (303, 664), (252, 664), (181, 602), (181, 555), (181, 460), (300, 410), (421, 461), (294, 325), (153, 268), (153, 162), (153, 107), (236, 37), (366, 37), (446, 107), (446, 159), (446, 269)]':'8','[(532, -29), (97, -29), (97, 54), (267, 54), (267, 630), (93, 579), (93, 668), (363, 746), (363, 54), (532, 54)]':'1',
}
# 对应实际字体信息
data_num_list = [9,3,2,5,6,4,7,0,8,1]

5.6 得到结果

结果一一对应，完美撒花

6 完整代码如下

# -*- coding:utf-8 _*-import os
import requests
from fontTools.ttLib import TTFont
import matplotlib.pyplot as plt
import re# 下载woff
url = "https://web.chace-ai.com/media/fonts/fFz9g4IQsHDyEaXI.woff?version=9.666110168223948"
headers = {
'accept':'*/*',
'accept-encoding':'gzip, deflate, br',
'accept-language':'zh-CN,zh;q=0.9',
'origin':'https://www.chacewang.com',
'referer':'https://www.chacewang.com/',
'user-agent':'Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/96.0.4664.93 Safari/537.36',
}
resp = requests.get(url=url, headers=headers)
save_woff = 'demo.woff'
with open(save_woff, "wb") as f:f.write(resp.content)f.close()
# # woff转xml文件
font = TTFont(save_woff)
save_xml = 'demo.xml'
font.saveXML(save_xml)  # 转换为xml文件# # 读取woff文件
font = TTFont('demo.woff')dict_data = {'[(96, 74), (158, 38), (231, 38), (333, 38), (445, 198), (445, 346), (444, 345), (443, 346), (393, 251), (275, 251), (179, 251), (54, 382), (54, 487), (54, 599), (191, 742), (300, 742), (413, 742), (541, 565), (541, 395), (541, 185), (376, -42), (231, -42), (151, -42), (96, -16), (151, 500), (151, 423), (229, 334), (298, 334), (358, 334), (439, 415), (439, 474), (439, 555), (359, 663), (292, 663), (231, 663), (151, 570)]':'9','[(75, 98), (153, 38), (251, 38), (329, 38), (421, 115), (421, 179), (421, 323), (215, 323), (151, 323), (151, 403), (212, 403), (394, 403), (394, 538), (394, 663), (255, 663), (175, 663), (104, 609), (104, 699), (179, 742), (278, 742), (375, 742), (493, 641), (493, 560), (493, 411), (341, 369), (341, 367), (423, 358), (519, 260), (519, 186), (519, 84), (371, -42), (249, -42), (141, -42), (75, -1)]':'3','[(518, -29), (57, -29), (57, 54), (277, 274), (368, 365), (430, 472), (430, 528), (430, 593), (358, 662), (289, 662), (188, 662), (96, 576), (96, 673), (185, 742), (304, 742), (407, 742), (525, 632), (525, 538), (525, 468), (449, 332), (343, 227), (170, 58), (170, 56), (518, 56)]':'2','[(92, 88), (168, 38), (251, 38), (329, 38), (425, 125), (425, 271), (328, 351), (235, 351), (191, 351), (115, 344), (115, 730), (488, 730), (488, 645), (204, 645), (204, 430), (247, 433), (268, 433), (388, 433), (523, 311), (523, 205), (523, 94), (379, -42), (252, -42), (145, -42), (92, -10)]':'5','[(503, 632), (444, 663), (380, 663), (281, 663), (162, 491), (161, 343), (163, 343), (217, 447), (335, 447), (433, 447), (551, 319), (551, 213), (551, 101), (415, -42), (310, -42), (194, -42), (63, 139), (63, 305), (63, 506), (236, 742), (377, 742), (457, 742), (503, 720), (167, 219), (167, 144), (246, 38), (312, 38), (375, 38), (454, 130), (454, 202), (454, 281), (380, 368), (311, 368), (247, 368), (167, 280)]':'6','[(557, 168), (460, 168), (460, -29), (367, -29), (367, 168), (14, 168), (14, 230), (349, 730), (460, 730), (460, 244), (557, 244), (367, 244), (367, 562), (367, 596), (369, 640), (367, 640), (360, 621), (338, 581), (113, 244)]':'4','[(544, 696), (247, -29), (148, -29), (430, 645), (51, 645), (51, 730), (544, 730)]':'7','[(49, 335), (49, 536), (183, 742), (309, 742), (550, 742), (550, 354), (550, 162), (414, -42), (292, -42), (176, -42), (49, 151), (147, 340), (147, 38), (301, 38), (452, 38), (452, 345), (452, 663), (304, 663), (147, 663)]':'0','[(219, 370), (84, 433), (84, 553), (84, 637), (211, 742), (310, 742), (401, 742), (518, 644), (518, 565), (518, 439), (377, 371), (377, 369), (543, 309), (543, 165), (543, 70), (406, -42), (284, -42), (183, -42), (57, 69), (57, 157), (57, 302), (219, 368), (421, 554), (421, 605), (358, 664), (303, 664), (252, 664), (181, 602), (181, 555), (181, 460), (300, 410), (421, 461), (294, 325), (153, 268), (153, 162), (153, 107), (236, 37), (366, 37), (446, 107), (446, 159), (446, 269)]':'8','[(532, -29), (97, -29), (97, 54), (267, 54), (267, 630), (93, 579), (93, 668), (363, 746), (363, 54), (532, 54)]':'1',
}
data_num_list = [9,3,2,5,6,4,7,0,8,1]result_list = []
for i in range(0,10):x = []y = []val = []for _data in font['glyf']['{}'.format(i)].coordinates:_x = _data[0]_y = _data[1]x.append(_x)y.append(_y)_val = (_x,_y)val.append(_val)_result = dict_data.get(str(val))result_list.append(_result)
# print(result_list)
_result = result_list[1:] + list(result_list[0])
print(_result)

# ps： 这是我针对此次字体反爬的解决方法，欢迎各位大佬给出意见或提问咨询