使用requests爬取梨視訊、bilibili視訊、汽車之家，bs4遍歷文件樹、搜尋文件樹，css選擇器

阿新 • • 發佈：2022-05-09

今日內容概要

使用requests爬取梨視訊
requests+bs4爬取汽車之家
bs4遍歷文件樹
bs4搜尋文件樹
css選擇器

內容詳細

1、使用requests爬取梨視訊

# 模擬傳送http請求的庫：requests---》只能傳送http請求----》沒有解析庫--》re、bs4、lxml
# requests-html：傳送請求+解析xml

# 視訊m3u8格式，分段---》會員試看6分鐘---》只加載了6分鐘
# 收費視訊：視訊解析


### 完整的視訊檔案儲存到本地

# re 解析想要的資料
# import requests
# res=requests.get("https://www.pearvideo.com/")
# print(res.text)


### 爬取到的視訊檔案放到本地 video 目錄下
# https://www.pearvideo.com/category_loading.jsp?reqType=5&categoryId=8&start=24

import requests
import re

res = requests.get('https://www.pearvideo.com/category_loading.jsp?reqType=5&categoryId=8&start=24')
# print(res.text)

# 解析出頁面中所有的視訊地址
video_list = re.findall('<a href="(.*?)" class="vervideo-lilink actplay">', res.text)
# print(video_list)

for video in video_list:
    video_url = 'https://www.pearvideo.com/' + video
    video_id = video_url.split('_')[-1]
    header = {
        'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/99.0.4844.74 Safari/537.36',
        'Referer': video_url
    }
    # 第一層反扒是加refer
    res_video = requests.get('https://www.pearvideo.com/videoStatus.jsp?contId=%s&mrd=0.7113776105084832' % video_id,
                             headers=header)
    mp4_url = res_video.json()['videoInfo']['videos']['srcUrl']

    # 第二層反扒是把不能播放地址變成能播放地址
    mp4_url = mp4_url.replace(mp4_url.split('/')[-1].split('-')[0], 'cont-%s' % video_id)
    print(mp4_url)

    # 下載到本地
    res_video_detail = requests.get(mp4_url)
    with open('./video/%s.mp4' % video_id, 'wb') as f:
        for line in res_video_detail.iter_content(1024):
            f.write(line)

# 單執行緒下載，速度不快，全是io操作，開啟多執行緒能夠顯著提高速度---》使用多執行緒全站下載視訊
# 執行緒池整站爬取


# 不能播放的地址
# https://video.pearvideo.com/mp4/third/20220314/    1652060493892      -10097838-231626-hd.mp4  # 不正常地址
# https://video.pearvideo.com/mp4/third/20220314/    cont-1754713       -10097838-231626-hd.mp4  # 正常地址


# mp4_url='https://video.pearvideo.com/mp4/third/20220314/      1652060493892   -10097838-231626-hd.mp4'
# mp4_url=mp4_url.replace(mp4_url.split('/')[-1].split('-')[0],'cont-%s'%video_id)

2、爬取bilibili視訊

# 爬取的b站視訊儲存到本地 是分為兩個檔案：
	視訊檔案
	音訊檔案
    
# 通過第三方軟體做了整合之後才是完整的視訊

# 視訊去水印--》fmmpeg--》加水印，拼接裁剪，摳圖，轉碼。。。
# 裝上使用python來呼叫處理

# 匯入requests模組，模擬傳送請求
import requests
# 匯入json
import json
# 匯入re
import re

# 定義請求頭
headers = {
    'Accept': '*/*',
    'Accept-Language': 'en-US,en;q=0.5',
    'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_12_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/63.0.3239.84 Safari/537.36'
}


# 正則表示式，根據條件匹配出值
def my_match(text, pattern):
    match = re.search(pattern, text)
    print(match.group(1))
    print()
    return json.loads(match.group(1))


def download_video(old_video_url, video_url, audio_url, video_name):
    headers.update({"Referer": old_video_url})
    print("開始下載視訊：%s" % video_name)

    video_content = requests.get(video_url, headers=headers)
    print('%s視訊大小：' % video_name, video_content.headers['content-length'])

    audio_content = requests.get(audio_url, headers=headers)
    print('%s音訊大小：' % video_name, audio_content.headers['content-length'])

    # 下載視訊開始
    received_video = 0
    with open('%s_video.mp4' % video_name, 'ab') as output:
        while int(video_content.headers['content-length']) > received_video:
            headers['Range'] = 'bytes=' + str(received_video) + '-'
            response = requests.get(video_url, headers=headers)
            output.write(response.content)
            received_video += len(response.content)
    # 下載視訊結束

    # 下載音訊開始
    audio_content = requests.get(audio_url, headers=headers)
    received_audio = 0
    with open('%s_audio.mp4' % video_name, 'ab') as output:
        while int(audio_content.headers['content-length']) > received_audio:
            # 視訊分片下載
            headers['Range'] = 'bytes=' + str(received_audio) + '-'
            response = requests.get(audio_url, headers=headers)
            output.write(response.content)
            received_audio += len(response.content)
    # 下載音訊結束
    return video_name


if __name__ == '__main__':
    # 換成你要爬取的視訊地址
    url = 'https://www.bilibili.com/video/BV17F411M7pQ'

    # 傳送請求，拿回資料
    res = requests.get(url, headers=headers)

    # 視訊詳情json
    playinfo = my_match(res.text, '__playinfo__=(.*?)</script><script>')

    # 視訊內容json
    initial_state = my_match(res.text, r'__INITIAL_STATE__=(.*?);\(function\(\)')

    # 視訊分多種格式，直接取解析度最高的視訊 1080p
    video_url = playinfo['data']['dash']['video'][0]['baseUrl']

    # 取出音訊地址
    audio_url = playinfo['data']['dash']['audio'][0]['baseUrl']
    video_name = initial_state['videoData']['title']
    print('視訊名字為：video_name')
    print('視訊地址為：', video_url)
    print('音訊地址為：', audio_url)
    download_video(url, video_url, audio_url, video_name)

3、requests+bs4爬取汽車之家

### 爬取到的圖片都會儲存到本地

### 也可以在pycharm直接打印出 新聞地址 和 圖片地址 單獨訪問

import requests
# pip3 install beautifulsoup4
from bs4 import BeautifulSoup

res = requests.get('https://www.autohome.com.cn/news/1/#liststart')
# print(res.text)

# html.parser bs4預設的解析庫
soup = BeautifulSoup(res.text, 'html.parser')

# 使用bs4的查詢
ul_list = soup.find_all(name='ul', class_='article')
# print(len(ul_list))

for ul in ul_list:
    # 找ul標籤下所有的li標籤
    li_list = ul.find_all(name='li')

    for li in li_list:
        h3 = li.find(name='h3')

        if h3:
            title = h3.text  # 獲取h3標籤的文字內容
            desc = li.find(name='p').text
            img = li.find(name='img')['src']

            if not img.startswith('http'):
                img = 'https:' + img
            url = 'https:' + li.find('a')['href']
            print('''
            新聞標題：%s
            新聞摘要：%s
            新聞圖片：%s
            新聞地址：%s
            ''' % (title, desc, img, url))

            # 把圖片儲存到本地
            res_img = requests.get(img)
            img_name = img.split('/')[-1]
            with open('./img/%s' % img_name, 'wb') as f:
                for line in res_img.iter_content(1024):
                    f.write(line)

            # 把資料存到資料庫 pymysql寫入資料庫--》建庫建表--》cursor.exec(insert ..)-->commit

4、bs4遍歷文件樹(快速定位查詢)

from bs4 import BeautifulSoup

html_doc = """
<html><head><title>The Dormouse's story</title></head>
<body>
<p class="title" id="id_p">lqz<b>The Dormouse's story</b></p>

<p class="story">Once upon a time there were three little sisters; and their names were
<a href="http://example.com/elsie" class="sister" id="link1">Elsie</a>,
<a href="http://example.com/lacie" class="sister" id="link2">Lacie</a> and
<a href="http://example.com/tillie" class="sister" id="link3">Tillie</a>;
and they lived at the bottom of a well.</p>

<p class="story">...</p>
"""

### html.parser 內建的，速度一般,容錯能力強
# lxml    第三方的，速度快,容錯能力強
# soup = BeautifulSoup(html_doc, 'html.parser')

# pip3 install lxml
soup = BeautifulSoup(html_doc, 'lxml')

# print(soup.prettify())  # 對html進行美化 美化成完整的書寫格式


# 1 遍歷文件樹之  .  遍歷    速度快
print(soup.title)
print(soup.body.p)
print(soup.body.p.b)


# 2、獲取標籤的名稱
print(soup.title.name)  # title
print(soup.body.name)  # body


# 3、獲取標籤的屬性
print(soup.body.p)  # 找 p標籤 速度快些
print(soup.p['class'])  # ['title']  因為class可能有多個，所以是列表
print(soup.p['id'])  # id_p
print(soup.p.attrs)  # {'class': ['title'], 'id': 'id_p'}  所有屬性放到字典中


# 4、獲取標籤的內容--文字內容
print(soup.p.text)  # lqzThe Dormouse's story 當前標籤和子子孫的文字內容拼到一起
print(soup.p.string)  # None  當前標籤只有文字或只有一個子有文字才拿出來，如果有多個子子孫孫，返回None
print(list(soup.p.strings))  # ['lqz', "The Dormouse's story"]  把子子孫孫的文字內容放到generator


# 5、巢狀選擇
# 可以連續點巢狀選擇
print(soup.head.title.string)  # The Dormouse's story


# 6、子節點、子孫節點
print(soup.p.contents)  # p下所有子節點，放到列表中
print(list(soup.p.children))  # 得到一個迭代器,包含p下所有子節點,跟contents本質一樣，只是節約記憶體
print(list(soup.p.descendants))  # 獲取子孫節點,p下所有的標籤都會選擇出來  子子孫孫

for i, child in enumerate(soup.p.children):
    print(i, child)
for i, child in enumerate(soup.p.descendants):
    print(i, child)


# 7、父節點、祖先節點
print(soup.a.parent)  # 獲取a標籤的父節點
print(list(soup.a.parents))  # 找到a標籤所有的祖先節點，父親的父親，父親的父親的父親...

# 8、兄弟節點
print(soup.a.next_sibling)  # 下一個兄弟
print(soup.a.previous_sibling)  # 上一個兄弟

print(list(soup.a.next_siblings))  # 下面的兄弟們=>生成器物件
print(soup.a.previous_siblings)  # 上面的兄弟們=>生成器物件



### 重點記憶：
	. 遍歷
	取屬性 [] attrrs.get()
	取文字  text  string  strings

5、bs4搜尋文件樹

from bs4 import BeautifulSoup

html_doc = """
<html><head><title>The Dormouse's story</title></head>
<body>
<p class="title" id="id_p">lqz<b>The Dormouse's story</b></p>

<p class="story">Once upon a time there were three little sisters; and their names were
<a href="http://example.com/elsie" class="sister" id="link1">Elsie</a>,
<a href="http://example.com/lacie" class="sister" id="link2">Lacie</a> and
<a href="http://example.com/tillie" class="sister" id="link3">Tillie</a>;
and they lived at the bottom of a well.</p>

<p class="story">...</p>
"""

soup = BeautifulSoup(html_doc, 'lxml')

##### 五種過濾器: 字串、正則表示式、列表、True、方法
# find：找到第一個      find_all：找所有

# 1、字串  --->value值是字串
res = soup.find_all(name='p')
# res = soup.find(id='id_p')
# res = soup.find_all(class_='story')
# res = soup.find_all(name='p', class_='story')  # and條件
# res = soup.find(name='a', id='link2').text
# res = soup.find(name='a', id='link2').attrs.get('href')
# res = soup.find(attrs={'id': 'link2', 'class': 'sister'}).attrs.get('href')
print(res)


# 2、正則表示式--->value是正則表示式
import re

# res = soup.find_all(name=re.compile('^b'))
# res = soup.find_all(href=re.compile('^http'))
res = soup.find_all(class_=re.compile('^s'))
print(res)


# 3、列表  value值是列表
# res = soup.find_all(name=['body', 'a'])
# res = soup.find_all(class_=['sister', 'story'])
res = soup.find_all(id=['link2', 'link3'])
print(res)


# 4、True   value值是True
# res = soup.find_all(name=True)
# res = soup.find_all(id=True)
res = soup.find_all(href=True)
print(res)


# 5、方法
def has_class_but_no_id(tag):
    return tag.has_attr('class') and not tag.has_attr('id')

print(soup.find_all(name=has_class_but_no_id))  # 有class但是沒有id的標籤



### 補充總結：
# 1 html頁面中，只要有的東西，通過bs4都可以解析出來

# 2 遍歷文件樹+搜尋文件樹混用
def has_class_but_no_id(tag):
    return tag.has_attr('class') and not tag.has_attr('id')

print(soup.find(name=has_class_but_no_id).a.text)


# 3 find_all的其他引數limit:限制取幾條  recursive：是否遞迴查詢
def has_class_but_no_id(tag):
    return tag.has_attr('class') and not tag.has_attr('id')

res = soup.find_all(name=has_class_but_no_id, limit=1)
print(res)

res = soup.find_all(name='a', recursive=False)  # 不遞迴查詢,速度快，只找一層
print(res)

6、css選擇器

### css，xpath選擇器是通用的---》基本所有的解析庫(bs4,lxml,pyquery,selenium的解析庫)--->都支援css選擇器-->css在前端通用

from bs4 import BeautifulSoup

html_doc = """
<html><head><title>The Dormouse's story</title></head>
<body>
<p class="title" id="id_p">lqz<b>The Dormouse's story</b></p>

<p class="story">Once upon a time there were three little sisters; and their names were
<a href="http://example.com/elsie" class="sister" id="link1">Elsie</a>,
<a href="http://example.com/lacie" class="sister" id="link2">Lacie</a> and
<a href="http://example.com/tillie" class="sister" id="link3">Tillie</a>;
and they lived at the bottom of a well.</p>

<p class="story">...</p>
"""
soup = BeautifulSoup(html_doc, 'lxml')

# soup.select()  # 找所有
# soup.select_one()  # 找一個

'''
div  找div標籤
div>a  找div下的緊鄰的a
div a  找div下的子子孫孫的a
.sister  找類名為sister的標籤
#id_p    找id為id_p的標籤
'''

# res = soup.select('#id_p')
# res = soup.select('.sister')
res = soup.select_one('.story>a').attrs.get('href')
print(res)


# 終極大招
import requests

response = requests.get('https://www.runoob.com/cssref/css-selectors.html')
soup = BeautifulSoup(response.text, 'lxml')
res = soup.select_one('#content > table > tbody > tr:nth-child(2) > td:nth-child(3)').text  # 找到標籤 右鍵 copy--selector
print(res)

# 只要頁面中有的通過bs4都能解析出來