爬蟲--今日頭條ajax動态

2017年11月30日 運作通過

# -*-coding:utf-8-*-
import json
import os
from json import JSONDecodeError
from urllib.parse import urlencode
import requests
from bs4 import BeautifulSoup
from requests.exceptions import ConnectionError
import re
from hashlib import md5

def get_page_index(offset, keyword):
    data = {
        'autoload': 'true',
        'count': ,
        'cur_tab': ,
        'format': 'json',
        'keyword': keyword,
        'offset': offset,
    }
    params = urlencode(data)
    base = 'http://www.toutiao.com/search_content/'
    url = base + '?' + params
    try:
        response = requests.get(url)
        if response.status_code == :
            return response.text
        return None
    except ConnectionError:
        print('Error occurred')
        return None


def download_image(url):
    print('Downloading', url)
    try:
        response = requests.get(url)
        if response.status_code == :
            save_image(response.content)
        return None
    except ConnectionError:
        return None


def save_image(content):
    file_path = '{0}/images/{1}.{2}'.format(os.getcwd(), md5(content).hexdigest(), 'jpg')
    if not os.path.exists(file_path):
        with open(file_path, 'wb') as f:
            f.write(content)
            f.close()


def parse_page_index(text):
    try:
        data = json.loads(text)
        if data and 'data' in data.keys():
            for item in data.get('data'):
                yield item.get('article_url')
    except JSONDecodeError:
        pass


def get_page_detail(url):
    try:
        response = requests.get(url)
        if response.status_code == :
            return response.text
        return None
    except ConnectionError:
        print('Error occurred')
        return None


def parse_page_detail(html, url):
    soup = BeautifulSoup(html, 'lxml')
    result = soup.select('title')
    title = result[].get_text() if result else ''
    images_pattern = re.compile('BASE_DATA.galleryInfo = (.*?)</script>', re.S)
    result = re.search(images_pattern, html)
    if result:
        data = result.group()
        gallery = data[data.index('JSON.parse("')+:data.index('siblingList')-]
        images = json.loads(json.loads('"' + gallery + '"'))
        sub_images = images["sub_images"]
        images = [item['url'] for item in sub_images]
        for image in images:
            download_image(image)

if __name__ == '__main__':
    offset = [offset for offset in range(,,)]
    KEYWORD = "小黃人"
    for off in offset:
        text = get_page_index(offset, KEYWORD)
        urls = parse_page_index(text)
        for url in urls:
            html = get_page_detail(url)
            result = parse_page_detail(html, url)
           

此代碼基本為 https://github.com/Germey/TouTiao中代碼

修改了原正規表達式提取所有組圖url部分。

（因為網站變更，原來的已經比對不到了）

本來想直接比對 gallery: JSON.parse後的内容，使用如下比對不到

re.compile('JSON.parse (.*?)，', re.S)
           

是以改成取BASE_DATA.galleryInfo = {} 整個内容。再拿JSON.parse的部分。取來的内容直接json.loads會報錯，可能是有轉義的字元\導緻。使用如下解決：

json.loads(json.loads('"' + gallery + '"'))
           

還是使用scrapy下載下傳圖檔友善