1. 程式人生 > >更新一下我對於python爬蟲的更改

更新一下我對於python爬蟲的更改

 

#使用build_opener()是為了讓python程式模仿瀏覽器進行訪問
opener = urllib.request.build_opener()
opener.addheaders = [('User-agent', 'Mozilla/5.0')]

 

你是否好奇標識瀏覽器身份的User-Agent,為什麼每個瀏覽器都有Mozilla字樣?

  Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/27.0.1453.94 Safari/537.36

  Mozilla/5.0 (Linux; U; Android 4.1.2; zh-tw; GT-I9300 Build/JZO54K) AppleWebKit/534.30 (KHTML, like Gecko) Version/4.0 Mobile Safari/534.30

  Mozilla/5.0 (Windows NT 6.1; WOW64; rv:20.0) Gecko/20100101 Firefox/20.0 Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.0; Trident/5.0)

  故事還得從頭說起,最初的主角叫NCSA Mosaic,簡稱Mosaic(馬賽克),是1992年末位於伊利諾伊大學厄巴納-香檳分校的國家超級計算機應用中心(National Center for Supercomputing Applications,簡稱NCSA)開發,並於1993年釋出的一款瀏覽器。它自稱“NCSA_Mosaic/2.0(Windows 3.1)”,Mosaic可以同時展示文字和圖片,從此瀏覽器變得有趣多了。

  然而很快就出現了另一個瀏覽器,這就是著名的Mozilla,中文名稱摩斯拉。一說 Mozilla = Mosaic + Killer,意為Mosaic殺手,也有說法是 Mozilla = Mosaic & Godzilla,意為馬賽克和哥斯拉,而Mozilla最初的吉祥物是隻綠色大蜥蜴,後來更改為紅色暴龍,跟哥斯拉長得一樣。

  但Mosaic對此非常不高興,於是後來Mozilla更名為Netscape,也就是網景。Netscape自稱“Mozilla/1.0(Win3.1)”,事情開始變得更加有趣。網景支援框架(frame),由於大家的喜歡框架變得流行起來,但是Mosaic不支援框架,於是網站管理員探測user agent,對Mozilla瀏覽器傳送含有框架的頁面,對非Mozilla瀏覽器傳送沒有框架的頁面。

  後來網景拿微軟尋開心,稱微軟的Windows是“沒有除錯過的硬體驅動程式”。微軟很生氣,後果很嚴重。此後微軟開發了自己的瀏覽器,這就是Internet Explorer,並希望它可以成為Netscape Killer。IE同樣支援框架,但它不是Mozilla,所以它總是收不到含有框架的頁面。微軟很鬱悶很快就沉不住氣了,它不想等到所有的網站管理員都瞭解IE並且給IE傳送含有框架的頁面,它選擇宣佈IE是相容Mozilla,並且模仿Netscape稱IE為“Mozilla/1.22(compatible; MSIE 2.0; Windows 95)”,於是IE可以收到含有框架的頁面了,所有微軟的人都嗨皮了,但是網站管理員開始暈了。

  因為微軟將IE和Windows捆綁銷售,並且把IE做得比Netscape更好,於是第一次瀏覽器血腥大戰爆發了,結果是Netscape以失敗退出歷史舞臺,微軟更加嗨皮。但沒想到Netscape居然以Mozilla的名義重生了,並且開發了Gecko,這次它自稱為“Mozilla/5.0(Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826”。

  Gecko是一款渲染引擎並且很出色。Mozilla後來變成了Firefox,並自稱“Mozilla/5.0 (Windows; U; Windows NT 5.1; sv-SE; rv:1.7.5) Gecko/20041108 Firefox/1.0”。Firefox效能很出色,Gecko也開始攻城略地,其他新的瀏覽器使用了它的程式碼,並且將它們自己稱為“Mozilla/5.0 (Macintosh; U; PPC Mac OS X Mach-O; en-US; rv:1.7.2) Gecko/20040825 Camino/0.8.1”,以及“Mozilla/5.0 (Windows; U; Windows NT 5.1; de; rv:1.8.1.8) Gecko/20071008 SeaMonkey/1.0”,每一個都將自己裝作Mozilla,而它們全都使用Gecko。

  Gecko很出色,而IE完全跟不上它,因此user agent探測規則變了,使用Gecko的瀏覽器被髮送了更好的程式碼,而其他瀏覽器則沒有這種待遇。Linux的追隨者對此很難過,因為他們編寫了Konqueror,它的引擎是KHTML,他們認為KHTML和Gecko一樣出色,但卻因為不是Gecko而得不到好的頁面,於是Konqueror為得到更好的頁面開始將自己偽裝成“like Gecko”,並自稱為“Mozilla/5.0 (compatible; Konqueror/3.2; FreeBSD) (KHTML, like Gecko)”。自此user agent變得更加混亂。

  這時更有Opera跳出來說“毫無疑問,我們應該讓使用者來決定他們想讓我們偽裝成哪個瀏覽器。”於是Opera乾脆建立了選單項讓使用者自主選擇讓Opera瀏覽器變成“Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; en) Opera 9.51”,或者“Mozilla/5.0 (Windows NT 6.0; U; en; rv:1.8.1) Gecko/20061208 Firefox/2.0.0 Opera 9.51”, 或者“Opera/9.51 (Windows NT 5.1; U; en)”。

  後來蘋果開發了Safari瀏覽器,並使用KHTML作為渲染引擎,但蘋果加入了許多新的特性,於是蘋果從KHTML另闢分支稱之為WebKit,但它又不想拋棄那些為KHTML編寫的頁面,於是Safari自稱為“Mozilla/5.0 (Macintosh; U; PPC Mac OS X; de-de) AppleWebKit/85.7 (KHTML, like Gecko) Safari/85.5”,這進一步加劇了user agent的混亂局面。

  因為微軟十分忌憚Firefox,於是IE重灌上陣,這次它自稱為“Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.0) ”,並且渲染效果同樣出色,但是需要網站管理員的指令它這麼做才行。

  再後來,谷歌開發了Chrome瀏覽器,Chrome使用Webkit作為渲染引擎,和Safari之前一樣,它想要那些為Safari編寫的頁面,於是它偽裝成了Safari。於是Chrome使用WebKit,並將自己偽裝成Safari,WebKit偽裝成KHTML,KHTML偽裝成Gecko,最後所有的瀏覽器都偽裝成了Mozilla,這就是為什麼所有的瀏覽器User-Agent裡都有Mozilla。Chrome自稱為“Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US) AppleWebKit/525.13 (KHTML, like Gecko) Chrome/0.2.149.27 Safari/525.13”。

  因為以上這段歷史,現在的User-Agent字串變得一團糟,幾乎根本無法彰顯它最初的意義。追根溯源,微軟可以說是這一切的始作俑者,但後來每一個人都在試圖假扮別人,最終把User-Agent搞得混亂不堪。

  一句話結論:因為網站開發者可能會因為你是某瀏覽器(這裡是 Mozilla),所以輸出一些特殊功能的程式程式碼(這裡指好的特殊功能),所以當其它瀏覽器也支援這種好功能時,就試圖去模仿 Mozilla 瀏覽器讓網站輸出跟 Mozilla 一樣的內容,而不是輸出被閹割功能的程式程式碼。大家都為了讓網站輸出最好的內容,都試圖假裝自己是 Mozilla 一個已經不存在的瀏覽器……

 

  附:各大瀏覽器誕生年表:

  根據考證,世界上第一個瀏覽器是WorldWideWeb,誕生於1990年。 而第一個普及的,是由NCSA開發的Mosaic瀏覽器。

  1993年1月23日:Mosaic Alpha,同年3月推出0.1版本,它實際上也是Netscape與IE的共同祖先。

  1994年12月:Netscape 1.0(Mozilla/1.0

  1994年:Opera,基於Elektra排版引擎(已成為歷史)。

  1995年8月16日:Internet Explorer 1.0(Mozilla/1.22

  1996年10月14日:Kongqueror瀏覽器,基於KHTML引擎,主要應用於 Linux 和 BSD家族的作業系統。

  1998年,Windows98橫空出世,捆綁了IE4.0,Netscape最終被徹底地打敗!

  1998年11月24日,美國線上AOL宣佈收購Netscape公司。之後將所有Netscape的瀏覽器原始程式碼都送給了開源社群自行開發,併成立開源社群Mozilla(mozilla.org)。

  2000年11月:網景公司正式釋出Netscape 6.0,基於新一代的排版引擎Gecko。

  2001年10月,微軟釋出Windows XP,因為忌憚Firefox的效能優勢,XP系統捆綁釋出了Internet Explorer 6.0(Mozilla/4.0)。

  2003年1月7日:Safari beta,基於KHTML排版引擎。

  2003年6月,Opera7.0,啟用Presto排版引擎。

  2003年7月15日,美國線上AOL解散了網景公司Netscape,而Mozilla基金會亦在當天成立,它就是現在Firefox瀏覽器的管理和推進組織,從此FireFox成了削弱IE市場份額的一隻重要棋子,Netscape的借屍還魂術宣告成功。

  2003年,蘋果推出了採用WebKit引擎的瀏覽器 Safari。

  2004年11月9日:Mozilla基金會發布了Firefox的第一個正式版本1.0(Mozilla/5.0),基於Gecko排版引擎。

  2005年,蘋果宣佈將WebKit完全開源(之前僅有從KHTML直接搬來的WebCore及JavaScriptCore是開源的)。

  2008年9月2日:Chrome,基於Chromium引擎,WebKit的一個分支。。

  2010年4月,蘋果公司宣佈了其瀏覽器引擎Webkit的最新專案Webkit2。

  2013年2月,Opera宣佈將用WebKit替代當前的核心瀏覽器引擎Presto。

  2013年4月3日,谷歌宣佈將與蘋果的開源瀏覽器核心Webkit分道揚鑣,在Chromium專案中研發Blink渲染引擎,內置於Chrome瀏覽器之中。

  2013年4月,Opera又宣佈放棄WEBKIT,跟隨GOOGLE的新開發的Blink引擎。

  2015年6月7日,Windows10預覽版,啟用Microsoft Edge瀏覽器名稱,WebCore是EdegHTML,JSCore是Chakra。

  2016年1月12日起,微軟停止為IE8/9/10這些舊版本的IE瀏覽器提供技術支援,Internet Explorer正式退出。

 

 

   except urllib.error.HTTPError:
       print('urllib.error.HTTPError')
       print ("HTTPError錯誤時間 : %s" % time.ctime())
       time.sleep(60)
       os.system("python pachongcsdn2.py")# 執行命令 重新開啟一個視窗

       
# 發生異常之後延時重新開始執行下面的語句       
    except urllib.error.URLError:
        print('urllib.error.URLError')
        print ("URLError錯誤時間 : %s" % time.ctime())
        time.sleep(60)

 

 

首先介紹一下  random  模組:

Python中的random模組用於生成隨機數

1. Beautiful Soup的簡介

簡單來說,Beautiful Soup是python的一個庫,最主要的功能是從網頁抓取資料。官方解釋如下:

      Beautiful Soup提供一些簡單的、python式的函式用來處理導航、搜尋、修改分析樹等功能。它是一個工具箱,通過解析文件為使用者提供需要抓取的資料,因為簡單,所以不需要多少程式碼就可以寫出一個完整的應用程式。

Beautiful Soup自動將輸入文件轉換為Unicode編碼,輸出文件轉換為utf-8編碼。你不需要考慮編碼方式,除非文件沒有指定一個編碼方式,這時,Beautiful Soup就不能自動識別編碼方式了。然後,你僅僅需要說明一下原始編碼方式就可以了。

Beautiful Soup已成為和lxml、html6lib一樣出色的python直譯器,為使用者靈活地提供不同的解析策略或強勁的速度。

2. Beautiful Soup 安裝

Beautiful Soup 3 目前已經停止開發,推薦在現在的專案中使用Beautiful Soup 4,不過它已經被移植到BS4了,也就是說匯入時我們需要 import bs4 。所以這裡我們用的版本是 Beautiful Soup 4.3.2 (簡稱BS4),另外據說 BS4 對 Python3 的支援不夠好,不過我用的是 Python2.7.7,如果有小夥伴用的是 Python3 版本,可以考慮下載 BS3 版本。

 

可以利用 pip 或者 easy_install 來安裝,以下兩種方法均可

 

1

easy_install beautifulsoup4

 

 

1

pip install beautifulsoup4

如果想安裝最新的版本,請直接下載安裝包來手動安裝,也是十分方便的方法。在這裡我安裝的是 Beautiful Soup 4.3.2

Beautiful Soup 3.2.1Beautiful Soup 4.3.2

下載完成之後解壓

執行下面的命令即可完成安裝

 

1

sudo python setup.py install

然後需要安裝 lxml

 

1

easy_install lxml

 

 

1

pip install lxml

另一個可供選擇的解析器是純Python實現的 html5lib , html5lib的解析方式與瀏覽器相同,可以選擇下列方法來安裝html5lib:

 

1

easy_install html5lib

 

 

1

pip install html5lib

Beautiful Soup支援Python標準庫中的HTML解析器,還支援一些第三方的解析器,如果我們不安裝它,則 Python 會使用 Python預設的解析器,lxml 解析器更加強大,速度更快,推薦安裝。

<thead”>

 

解析器 使用方法 優勢 劣勢
Python標準庫 BeautifulSoup(markup, “html.parser”)
  • Python的內建標準庫
  • 執行速度適中
  • 文件容錯能力強
  • Python 2.7.3 or 3.2.2)前 的版本中文件容錯能力差
lxml HTML 解析器 BeautifulSoup(markup, “lxml”)
  • 速度快
  • 文件容錯能力強
  • 需要安裝C語言庫
lxml XML 解析器 BeautifulSoup(markup, [“lxml”, “xml”])BeautifulSoup(markup, “xml”)
  • 速度快
  • 唯一支援XML的解析器
  • 需要安裝C語言庫
html5lib BeautifulSoup(markup, “html5lib”)
  • 最好的容錯性
  • 以瀏覽器的方式解析文件
  • 生成HTML5格式的文件
  • 速度慢
  • 不依賴外部擴充套件

3. 開啟Beautiful Soup 之旅

在這裡先分享官方文件連結,不過內容是有些多,也不夠條理,在此本文章做一下整理方便大家參考。

官方文件

4. 建立 Beautiful Soup 物件

首先必須要匯入 bs4 庫

 

1

from bs4 import BeautifulSoup

我們建立一個字串,後面的例子我們便會用它來演示

Python

 

1

2

3

4

5

6

7

8

9

10

11

html = """

<html><head><title>The Dormouse's story</title></head>

<body>

<p class="title" name="dromouse"><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>

"""

建立 beautifulsoup 物件

Python

 

1

soup = BeautifulSoup(html)

另外,我們還可以用本地 HTML 檔案來建立物件,例如

 

1

soup = BeautifulSoup(open('index.html'))

上面這句程式碼便是將本地 index.html 檔案開啟,用它來建立 soup 物件

下面我們來列印一下 soup 物件的內容,格式化輸出

 

1

print soup.prettify()

 

 

1

2

3

4

5

<html>

<head>

  <title>

   The Dormouse's story

  </title>

以上便是輸出結果,格式化打印出了它的內容,這個函式經常用到,小夥伴們要記好咯。

5. 四大物件種類

Beautiful Soup將複雜HTML文件轉換成一個複雜的樹形結構,每個節點都是Python物件,所有物件可以歸納為4種:

  • Tag
  • NavigableString
  • BeautifulSoup
  • Comment

下面我們進行一一介紹

(1)Tag

Tag 是什麼?通俗點講就是 HTML 中的一個個標籤,例如

 

1

<title>The Dormouse's story</title>

 

 

1

<a class="sister" href="http://example.com/elsie" id="link1">Elsie</a>

上面的 title a 等等 HTML 標籤加上裡面包括的內容就是 Tag,下面我們來感受一下怎樣用 Beautiful Soup 來方便地獲取 Tags

下面每一段程式碼中註釋部分即為執行結果

 

1

2

print soup.title

#<title>The Dormouse's story</title>

 

 

1

2

print soup.head

#<head><title>The Dormouse's story</title></head>

 

 

1

2

print soup.a

#<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>

 

 

1

2

print soup.p

#<p class="title" name="dromouse"><b>The Dormouse's story</b></p>

我們可以利用 soup加標籤名輕鬆地獲取這些標籤的內容,是不是感覺比正則表示式方便多了?不過有一點是,它查詢的是在所有內容中的第一個符合要求的標籤,如果要查詢所有的標籤,我們在後面進行介紹。

我們可以驗證一下這些物件的型別

 

1

2

print type(soup.a)

#<class 'bs4.element.Tag'>

對於 Tag,它有兩個重要的屬性,是 name 和 attrs,下面我們分別來感受一下

name

 

1

2

3

4

print soup.name

print soup.head.name

#[document]

#head

soup 物件本身比較特殊,它的 name 即為 [document],對於其他內部標籤,輸出的值便為標籤本身的名稱。

attrs

 

1

2

print soup.p.attrs

#{'class': ['title'], 'name': 'dromouse'}

在這裡,我們把 p 標籤的所有屬性列印輸出了出來,得到的型別是一個字典。

如果我們想要單獨獲取某個屬性,可以這樣,例如我們獲取它的 class 叫什麼

 

1

2

print soup.p['class']

#['title']

還可以這樣,利用get方法,傳入屬性的名稱,二者是等價的

 

1

2

print soup.p.get('class')

#['title']

我們可以對這些屬性和內容等等進行修改,例如

 

1

2

3

soup.p['class']="newClass"

print soup.p

#<p class="newClass" name="dromouse"><b>The Dormouse's story</b></p>

還可以對這個屬性進行刪除,例如

 

1

2

3

del soup.p['class']

print soup.p

#<p name="dromouse"><b>The Dormouse's story</b></p>

不過,對於修改刪除的操作,不是我們的主要用途,在此不做詳細介紹了,如果有需要,請檢視前面提供的官方文件

(2)NavigableString

既然我們已經得到了標籤的內容,那麼問題來了,我們要想獲取標籤內部的文字怎麼辦呢?很簡單,用 .string 即可,例如

 

1

2

print soup.p.string

#The Dormouse's story

這樣我們就輕鬆獲取到了標籤裡面的內容,想想如果用正則表示式要多麻煩。它的型別是一個 NavigableString,翻譯過來叫 可以遍歷的字串,不過我們最好還是稱它英文名字吧。

來檢查一下它的型別

 

1

2

print type(soup.p.string)

#<class 'bs4.element.NavigableString'>

 

(3)BeautifulSoup

BeautifulSoup 物件表示的是一個文件的全部內容.大部分時候,可以把它當作 Tag 物件,是一個特殊的 Tag,我們可以分別獲取它的型別,名稱,以及屬性來感受一下

 

1

2

3

4

5

6

print type(soup.name)

#<type 'unicode'>

print soup.name

# [document]

print soup.attrs

#{} 空字典

 

(4)Comment

Comment 物件是一個特殊型別的 NavigableString 物件,其實輸出的內容仍然不包括註釋符號,但是如果不好好處理它,可能會對我們的文字處理造成意想不到的麻煩。

我們找一個帶註釋的標籤

 

1

2

3

print soup.a

print soup.a.string

print type(soup.a.string)

 

執行結果如下

 

 

1

2

3

<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>

Elsie

<class 'bs4.element.Comment'>

a 標籤裡的內容實際上是註釋,但是如果我們利用 .string 來輸出它的內容,我們發現它已經把註釋符號去掉了,所以這可能會給我們帶來不必要的麻煩。

另外我們列印輸出下它的型別,發現它是一個 Comment 型別,所以,我們在使用前最好做一下判斷,判斷程式碼如下

 

1

2

if type(soup.a.string)==bs4.element.Comment:

    print soup.a.string

上面的程式碼中,我們首先判斷了它的型別,是否為 Comment 型別,然後再進行其他操作,如列印輸出。

6. 遍歷文件樹

(1)直接子節點

要點:.contents  .children  屬性

.contents

tag 的 .content 屬性可以將tag的子節點以列表的方式輸出

 

1

2

print soup.head.contents 

#[<title>The Dormouse's story</title>]

輸出方式為列表,我們可以用列表索引來獲取它的某一個元素

 

1

2

print soup.head.contents[0]

#<title>The Dormouse's story</title>

.children

它返回的不是一個 list,不過我們可以通過遍歷獲取所有子節點。

我們列印輸出 .children 看一下,可以發現它是一個 list 生成器物件

 

1

2

print soup.head.children

#<listiterator object at 0x7f71457f5710>

我們怎樣獲得裡面的內容呢?很簡單,遍歷一下就好了,程式碼及結果如下

 

1

2

for child in  soup.body.children:

    print child

 

 

1

2

3

4

5

<p class="title" name="dromouse"><b>The Dormouse's story</b></p>

 

<p class="story">Once upon a time there were three little sisters; and their names were

<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>,

<a class="sister" href="http://example.com/lacie" id="link2">Lacie</a> and

 

 (2)所有子孫節點

知識點:.descendants 屬性

.descendants

.contents 和 .children 屬性僅包含tag的直接子節點,.descendants 屬性可以對所有tag的子孫節點進行遞迴迴圈,和 children類似,我們也需要遍歷獲取其中的內容。

 

1

2

for child in soup.descendants:

    print child

執行結果如下,可以發現,所有的節點都被打印出來了,先生最外層的 HTML標籤,其次從 head 標籤一個個剝離,以此類推。

 

1

2

3

4

5

<html><head><title>The Dormouse's story</title></head>

<body>

<p class="title" name="dromouse"><b>The Dormouse's story</b></p>

<p class="story">Once upon a time there were three little sisters; and their names were

<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>,

 

 (3)節點內容

知識點:.string 屬性

如果tag只有一個 NavigableString 型別子節點,那麼這個tag可以使用 .string 得到子節點。如果一個tag僅有一個子節點,那麼這個tag也可以使用 .string 方法,輸出結果與當前唯一子節點的 .string 結果相同。

通俗點說就是:如果一個標籤裡面沒有標籤了,那麼 .string 就會返回標籤裡面的內容。如果標籤裡面只有唯一的一個標籤了,那麼 .string 也會返回最裡面的內容。例如

 

1

2

3

4

print soup.head.string

#The Dormouse's story

print soup.title.string

#The Dormouse's story

如果tag包含了多個子節點,tag就無法確定,string 方法應該呼叫哪個子節點的內容, .string 的輸出結果是 None

 

1

2

print soup.html.string

# None

 

(4)多個內容

知識點: .strings  .stripped_strings 屬性

.strings

獲取多個內容,不過需要遍歷獲取,比如下面的例子

 

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

for string in soup.strings:

    print(repr(string))

    # u"The Dormouse's story"

    # u'\n\n'

    # u"The Dormouse's story"

    # u'\n\n'

    # u'Once upon a time there were three little sisters; and their names were\n'

    # u'Elsie'

    # u',\n'

    # u'Lacie'

    # u' and\n'

    # u'Tillie'

    # u';\nand they lived at the bottom of a well.'

    # u'\n\n'

    # u'...'

    # u'\n'

.stripped_strings 

輸出的字串中可能包含了很多空格或空行,使用 .stripped_strings 可以去除多餘空白內容

 

1

2

3

4

5

6

7

8

9

10

11

12

for string in soup.stripped_strings:

    print(repr(string))

    # u"The Dormouse's story"

    # u"The Dormouse's story"

    # u'Once upon a time there were three little sisters; and their names were'

    # u'Elsie'

    # u','

    # u'Lacie'

    # u'and'

    # u'Tillie'

    # u';\nand they lived at the bottom of a well.'

    # u'...'

 

(5)父節點

 知識點: .parent 屬性

 

 

1

2

3

p = soup.p

print p.parent.name

#body

 

 

1

2

3

content = soup.head.title.string

print content.parent.name

#title

 

(6)全部父節點

知識點:.parents 屬性

通過元素的 .parents 屬性可以遞迴得到元素的所有父輩節點,例如

 

1

2

3

content = soup.head.title.string

for parent in  content.parents:

    print parent.name

 

 

1

2

3

4

title

head

html

[document]

 

 (7)兄弟節點

知識點:.next_sibling  .previous_sibling 屬性

兄弟節點可以理解為和本節點處在統一級的節點,.next_sibling 屬性獲取了該節點的下一個兄弟節點,.previous_sibling 則與之相反,如果節點不存在,則返回 None

注意:實際文件中的tag的 .next_sibling 和 .previous_sibling 屬性通常是字串或空白,因為空白或者換行也可以被視作一個節點,所以得到的結果可能是空白或者換行

 

1

2

3

4

5

6

7

8

9

10

11

print soup.p.next_sibling

#       實際該處為空白

print soup.p.prev_sibling

#None   沒有前一個兄弟節點,返回 None

print soup.p.next_sibling.next_sibling

#<p class="story">Once upon a time there were three little sisters; and their names were

#<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>,

#<a class="sister" href="http://example.com/lacie" id="link2">Lacie</a> and

#<a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>;

#and they lived at the bottom of a well.</p>

#下一個節點的下一個兄弟節點是我們可以看到的節點

 

(8)全部兄弟節點

知識點:.next_siblings  .previous_siblings 屬性

通過 .next_siblings 和 .previous_siblings 屬性可以對當前節點的兄弟節點迭代輸出

 

1

2

3

4

5

6

7

8

for sibling in soup.a.next_siblings:

    print(repr(sibling))

    # u',\n'

    # <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>

    # u' and\n'

    # <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>

    # u'; and they lived at the bottom of a well.'

    # None

 

(9)前後節點

知識點:.next_element  .previous_element 屬性

與 .next_sibling  .previous_sibling 不同,它並不是針對於兄弟節點,而是在所有節點,不分層次

比如 head 節點為

 

1

<head><title>The Dormouse's story</title></head>

那麼它的下一個節點便是 title,它是不分層次關係的

 

1

2

print soup.head.next_element

#<title>The Dormouse's story</title>

 

(10)所有前後節點

知識點:.next_elements  .previous_elements 屬性

通過 .next_elements 和 .previous_elements 的迭代器就可以向前或向後訪問文件的解析內容,就好像文件正在被解析一樣

 

1

2

3

4

5

6

7

8

9

for element in last_a_tag.next_elements:

    print(repr(element))

# u'Tillie'

# u';\nand they lived at the bottom of a well.'

# u'\n\n'

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

# u'...'

# u'\n'

# None

以上是遍歷文件樹的基本用法。

7.搜尋文件樹

(1)find_all( name , attrs , recursive , text , **kwargs )

find_all() 方法搜尋當前tag的所有tag子節點,並判斷是否符合過濾器的條件

1)name 引數

name 引數可以查詢所有名字為 name 的tag,字串物件會被自動忽略掉

A.傳字串

最簡單的過濾器是字串.在搜尋方法中傳入一個字串引數,Beautiful Soup會查詢與字串完整匹配的內容,下面的例子用於查詢文件中所有的<b>標籤

 

1

2

soup.find_all('b')

# [<b>The Dormouse's story</b>]

 

 

1

2

print soup.find_all('a')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>, <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>, <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

B.傳正則表示式

如果傳入正則表示式作為引數,Beautiful Soup會通過正則表示式的 match() 來匹配內容.下面例子中找出所有以b開頭的標籤,這表示<body>和<b>標籤都應該被找到

 

1

2

3

4

5

import re

for tag in soup.find_all(re.compile("^b")):

    print(tag.name)

# body

# b

C.傳列表

如果傳入列表引數,Beautiful Soup會將與列表中任一元素匹配的內容返回.下面程式碼找到文件中所有<a>標籤和<b>標籤

 

1

2

3

4

5

soup.find_all(["a", "b"])

# [<b>The Dormouse's story</b>,

#  <a class="sister" href="http://example.com/elsie" id="link1">Elsie</a>,

#  <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>,

#  <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

D.傳 True

True 可以匹配任何值,下面程式碼查詢到所有的tag,但是不會返回字串節點

 

1

2

3

4

5

6

7

8

9

10

11

for tag in soup.find_all(True):

    print(tag.name)

# html

# head

# title

# body

# p

# b

# p

# a

# a

E.傳方法

如果沒有合適過濾器,那麼還可以定義一個方法,方法只接受一個元素引數 [4] ,如果這個方法返回 True 表示當前元素匹配並且被找到,如果不是則反回 False

下面方法校驗了當前元素,如果包含 class 屬性卻不包含 id 屬性,那麼將返回 True:

 

1

2

def has_class_but_no_id(tag):

    return tag.has_attr('class') and not tag.has_attr('id')

將這個方法作為引數傳入 find_all() 方法,將得到所有<p>標籤:

 

1

2

3

4

soup.find_all(has_class_but_no_id)

# [<p class="title"><b>The Dormouse's story</b></p>,

#  <p class="story">Once upon a time there were...</p>,

#  <p class="story">...</p>]

2)keyword 引數

注意:如果一個指定名字的引數不是搜尋內建的引數名,搜尋時會把該引數當作指定名字tag的屬性來搜尋,如果包含一個名字為 id 的引數,Beautiful Soup會搜尋每個tag的”id”屬性

 

1

2

soup.find_all(id='link2')

# [<a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>]

如果傳入 href 引數,Beautiful Soup會搜尋每個tag的”href”屬性

 

1

2

soup.find_all(href=re.compile("elsie"))

# [<a class="sister" href="http://example.com/elsie" id="link1">Elsie</a>]

使用多個指定名字的引數可以同時過濾tag的多個屬性

 

1

2

soup.find_all(href=re.compile("elsie"), id='link1')

# [<a class="sister" href="http://example.com/elsie" id="link1">three</a>]

在這裡我們想用 class 過濾,不過 class 是 python 的關鍵詞,這怎麼辦?加個下劃線就可以

 

1

2

3

4

soup.find_all("a", class_="sister")

# [<a class="sister" href="http://example.com/elsie" id="link1">Elsie</a>,

#  <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>,

#  <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

有些tag屬性在搜尋不能使用,比如HTML5中的 data-* 屬性

 

1

2

3

data_soup = BeautifulSoup('<div data-foo="value">foo!</div>')

data_soup.find_all(data-foo="value")

# SyntaxError: keyword can't be an expression

但是可以通過 find_all() 方法的 attrs 引數定義一個字典引數來搜尋包含特殊屬性的tag

 

1

2

data_soup.find_all(attrs={"data-foo": "value"})

# [<div data-foo="value">foo!</div>]

3)text 引數

通過 text 引數可以搜搜文件中的字串內容.與 name 引數的可選值一樣, text 引數接受 字串 , 正則表示式 , 列表, True

 

1

2

3

4

5

6

7

8

soup.find_all(text="Elsie")

# [u'Elsie']

 

soup.find_all(text=["Tillie", "Elsie", "Lacie"])

# [u'Elsie', u'Lacie', u'Tillie']

 

soup.find_all(text=re.compile("Dormouse"))

[u"The Dormouse's story", u"The Dormouse's story"]

4)limit 引數

find_all() 方法返回全部的搜尋結構,如果文件樹很大那麼搜尋會很慢.如果我們不需要全部結果,可以使用 limit 引數限制返回結果的數量.效果與SQL中的limit關鍵字類似,當搜尋到的結果數量達到 limit 的限制時,就停止搜尋返回結果.

文件樹中有3個tag符合搜尋條件,但結果只返回了2個,因為我們限制了返回數量

 

1

2

3

soup.find_all("a", limit=2)

# [<a class="sister" href="http://example.com/elsie" id="link1">Elsie</a>,

#  <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>]

5)recursive 引數

呼叫tag的 find_all() 方法時,Beautiful Soup會檢索當前tag的所有子孫節點,如果只想搜尋tag的直接子節點,可以使用引數 recursive=False .

一段簡單的文件:

 

1

2

3

4

5

6

7

<html>

<head>

  <title>

   The Dormouse's story

  </title>

</head>

...

是否使用 recursive 引數的搜尋結果:

 

1

2

3

4

5

soup.html.find_all("title")

# [<title>The Dormouse's story</title>]

 

soup.html.find_all("title", recursive=False)

# []

(2)find( name , attrs , recursive , text , **kwargs )

它與 find_all() 方法唯一的區別是 find_all() 方法的返回結果是值包含一個元素的列表,而 find() 方法直接返回結果

(3)find_parents()  find_parent()

find_all() 和 find() 只搜尋當前節點的所有子節點,孫子節點等. find_parents() 和 find_parent() 用來搜尋當前節點的父輩節點,搜尋方法與普通tag的搜尋方法相同,搜尋文件搜尋文件包含的內容

(4)find_next_siblings()  find_next_sibling()

這2個方法通過 .next_siblings 屬性對當 tag 的所有後面解析的兄弟 tag 節點進行迭代, find_next_siblings() 方法返回所有符合條件的後面的兄弟節點,find_next_sibling() 只返回符合條件的後面的第一個tag節點

(5)find_previous_siblings()  find_previous_sibling()

這2個方法通過 .previous_siblings 屬性對當前 tag 的前面解析的兄弟 tag 節點進行迭代, find_previous_siblings()方法返回所有符合條件的前面的兄弟節點, find_previous_sibling() 方法返回第一個符合條件的前面的兄弟節點

(6)find_all_next()  find_next()

這2個方法通過 .next_elements 屬性對當前 tag 的之後的 tag 和字串進行迭代, find_all_next() 方法返回所有符合條件的節點, find_next() 方法返回第一個符合條件的節點

(7)find_all_previous() 和 find_previous()

這2個方法通過 .previous_elements 屬性對當前節點前面的 tag 和字串進行迭代, find_all_previous() 方法返回所有符合條件的節點, find_previous()方法返回第一個符合條件的節點

注:以上(2)(3)(4)(5)(6)(7)方法引數用法與 find_all() 完全相同,原理均類似,在此不再贅述。

8.CSS選擇器

我們在寫 CSS 時,標籤名不加任何修飾,類名前加點,id名前加 #,在這裡我們也可以利用類似的方法來篩選元素,用到的方法是 soup.select(),返回型別是 list

(1)通過標籤名查詢

 

 

1

2

print soup.select('title') 

#[<title>The Dormouse's story</title>]

 

 

1

2

print soup.select('a')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>, <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>, <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

 

 

1

2

print soup.select('b')

#[<b>The Dormouse's story</b>]

 

(2)通過類名查詢

 

 

1

2

print soup.select('.sister')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>, <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>, <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

 

(3)通過 id 名查詢

 

 

1

2

print soup.select('#link1')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>]

 

(4)組合查詢

組合查詢即和寫 class 檔案時,標籤名與類名、id名進行的組合原理是一樣的,例如查詢 p 標籤中,id 等於 link1的內容,二者需要用空格分開

 

1

2

print soup.select('p #link1')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>]

直接子標籤查詢

 

1

2

print soup.select("head > title")

#[<title>The Dormouse's story</title>]

 

(5)屬性查詢

查詢時還可以加入屬性元素,屬性需要用中括號括起來,注意屬性和標籤屬於同一節點,所以中間不能加空格,否則會無法匹配到。

 

1

2

print soup.select('a[class="sister"]')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>, <a class="sister" href="http://example.com/lacie" id="link2">Lacie</a>, <a class="sister" href="http://example.com/tillie" id="link3">Tillie</a>]

 

 

1

2

print soup.select('a[href="http://example.com/elsie"]')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>]

同樣,屬性仍然可以與上述查詢方式組合,不在同一節點的空格隔開,同一節點的不加空格

 

1

2

print soup.select('p a[href="http://example.com/elsie"]')

#[<a class="sister" href="http://example.com/elsie" id="link1"><!-- Elsie --></a>]

以上的 select 方法返回的結果都是列表形式,可以遍歷形式輸出,然後用 get_text() 方法來獲取它的內容。

 

1

2

3

4

5

6

soup = BeautifulSoup(html, 'lxml')

print type(soup.select('title'))

print soup.select('title')[0].get_text()

 

for title in soup.select('title'):

    print title.get_text()

 

簡述 request 模組

一開始要匯入Requests模組:

>>> import requests

然後,嘗試獲取某個網頁。本例子中,我們來獲取Github的公共時間線

>>> r = requests.get('https://github.com/timeline.json')

現在,我們有一個名為 r 的 Response 物件。可以從這個物件中獲取所有我們想要的資訊。

Requests簡便的API意味著所有HTTP請求型別都是顯而易見的。例如,你可以這樣傳送一個HTTP POST請求:

>>> r = requests.post("http://httpbin.org/post")

漂亮,對吧?那麼其他HTTP請求型別:PUT, DELETE, HEAD以及OPTIONS又是如何的呢?都是一樣的簡單:

>>> r = requests.put("http://httpbin.org/put")
>>> r = requests.delete("http://httpbin.org/delete")
>>> r = requests.head("http://httpbin.org/get")
>>> r = requests.options("http://httpbin.org/get")

都很不錯吧,但這也僅是Requests的冰山一角呢。

為URL傳遞引數

你也許經常想為URL的查詢字串(query string)傳遞某種資料。如果你是手工構建URL,那麼資料會以鍵/值 對的形式置於URL中,跟在一個問號的後面。例如,httpbin.org/get?key=val 。 Requests允許你使用 params 關鍵字引數,以一個字典來提供這些引數。舉例來說,如果你想傳遞 key1=value1 和 key2=value2 到 httpbin.org/get ,那麼你可以使用如下程式碼:

>>> payload = {'key1': 'value1', 'key2': 'value2'}
>>> r = requests.get("http://httpbin.org/get", params=payload)

通過列印輸出該URL,你能看到URL已被正確編碼:

>>> print r.url
u'http://httpbin.org/get?key2=value2&key1=value1'

響應內容

我們能讀取伺服器響應的內容。再次以Github時間線為例:

>>> import requests
>>> r = requests.get('https://github.com/timeline.json')
>>> r.text
'[{"repository":{"open_issues":0,"url":"https://github.com/...

Requests會自動解碼來自伺服器的內容。大多數unicode字符集都能被無縫地解碼。

請求發出後,Requests會基於HTTP頭部對響應的編碼作出有根據的推測。當你訪問r.text 之時,Requests會使用其推測的文字編碼。你可以找出Requests使用了什麼編碼,並且能夠使用 r.encoding 屬性來改變它:

>>> r.encoding
'utf-8'
>>> r.encoding = 'ISO-8859-1'

如果你改變了編碼,每當你訪問 r.text ,Request都將會使用 r.encoding 的新值。

在你需要的情況下,Requests也可以使用定製的編碼。如果你建立了自己的編碼,並使用codecs 模組進行註冊,你就可以輕鬆地使用這個解碼器名稱作為 r.encoding 的值, 然後由Requests來為你處理編碼。

二進位制響應內容

你也能以位元組的方式訪問請求響應體,對於非文字請求:

>>> r.content
b'[{"repository":{"open_issues":0,"url":"https://github.com/...

Requests會自動為你解碼 gzip 和 deflate 傳輸編碼的響應資料。

例如,以請求返回的二進位制資料建立一張圖片,你可以使用如下程式碼:

>>> from PIL import Image
>>> from StringIO import StringIO
>>> i = Image.open(StringIO(r.content))

JSON響應內容

Requests中也有一個內建的JSON解碼器,助你處理JSON資料:

>>> import requests
>>> r = requests.get('https://github.com/timeline.json')
>>> r.json()
[{u'repository': {u'open_issues': 0, u'url': 'https://github.com/...

如果JSON解碼失敗, r.json 就會丟擲一個異常。

原始響應內容

在罕見的情況下你可能想獲取來自伺服器的原始套接字響應,那麼你可以訪問 r.raw 。 如果你確實想這麼幹,那請你確保在初始請求中設定了 stream=True 。具體的你可以這麼做:

>>> r = requests.get('https://github.com/timeline.json', stream=True)
>>> r.raw
<requests.packages.urllib3.response.HTTPResponse object at 0x101194810>
>>> r.raw.read(10)
'\x1f\x8b\x08\x00\x00\x00\x00\x00\x00\x03'

定製請求頭

如果你想為請求新增HTTP頭部,只要簡單地傳遞一個 dict 給 headers 引數就可以了。

例如,在前一個示例中我們沒有指定content-type:

>>> import json
>>> url = 'https://api.github.com/some/endpoint'
>>> payload = {'some': 'data'}
>>> headers = {'content-type': 'application/json'}

>>> r = requests.post(url, data=json.dumps(payload), headers=headers)

更加複雜的POST請求

通常,你想要傳送一些編碼為表單形式的資料—非常像一個HTML表單。 要實現這個,只需簡單地傳遞一個字典給 data 引數。你的資料字典 在發出請求時會自動編碼為表單形式:

>>> payload = {'key1': 'value1', 'key2': 'value2'}
>>> r = requests.post("http://httpbin.org/post", data=payload)
>>> print r.text
{
  ...
  "form": {
    "key2": "value2",
    "key1": "value1"
  },
  ...
}

很多時候你想要傳送的資料並非編碼為表單形式的。如果你傳遞一個 string 而不是一個dict ,那麼資料會被直接釋出出去。

例如,Github API v3接受編碼為JSON的POST/PATCH資料:

>>> import json
>>> url = 'https://api.github.com/some/endpoint'
>>> payload = {'some': 'data'}

>>> r = requests.post(url, data=json.dumps(payload))

POST一個多部分編碼(Multipart-Encoded)的檔案

Requests使得上傳多部分編碼檔案變得很簡單:

>>> url = 'http://httpbin.org/post'
>>> files = {'file': open('report.xls', 'rb')}

>>> r = requests.post(url, files=files)
>>> r.text
{
  ...
  "files": {
    "file": "<censored...binary...data>"
  },
  ...
}

你可以顯式地設定檔名:

>>> url = 'http://httpbin.org/post'
>>> files = {'file': ('report.xls', open('report.xls', 'rb'))}

>>> r = requests.post(url, files=files)
>>> r.text
{
  ...
  "files": {
    "file": "<censored...binary...data>"
  },
  ...
}

如果你想,你也可以傳送作為檔案來接收的字串:

>>> url = 'http://httpbin.org/post'
>>> files = {'file': ('report.csv', 'some,data,to,send\nanother,row,to,send\n')}

>>> r = requests.post(url, files=files)
>>> r.text
{
  ...
  "files": {
    "file": "some,data,to,send\\nanother,row,to,send\\n"
  },
  ...
}

響應狀態碼

我們可以檢測響應狀態碼:

>>> r = requests.get('http://httpbin.org/get')
>>> r.status_code
200

為方便引用,Requests還附帶了一個內建的狀態碼查詢物件:

>>> r.status_code == requests.codes.ok
True

如果傳送了一個失敗請求(非200響應),我們可以通過 Response.raise_for_status() 來丟擲異常:

>>> bad_r = requests.get('http://httpbin.org/status/4