爬虫批量获取中文文献引用[selenium+知网]

应用场景：毕业论文中文献往往高达几十篇，调整引用格式（如标点后需加“空格”，作者超过三人，第三人以后需改为“，等”），虽然有各种论文管理软件，但能利用自己所学技能批量获取文献引用也挺有趣。

核心：selenium+知网。selenium是python写爬虫常用的库，不知道的小伙伴可以先在CSDN搜索下，做好selenium的前置准备。本文selenium使用的浏览器为Google浏览器。

具体逻辑：首先将文献下载保存至一个文件夹，文件命名格式为 " XXXXXX_作者.pdf "（知网下载pdf默认格式），然后通过python提取文件名，利用selenium访问知网实现搜索、引用的功能，然后抓取该文献引用格式。

详细代码如下：

其中file_address即为保存有文献的文件夹路径，修改即可。

import time
from selenium import webdriver
from selenium.webdriver import ActionChains
from selenium.webdriver.common.by import By
import os
import random
import re

time_s=random.random()*2+2
#保存文献文件夹
#file_address=input("文献文件夹：")
file_address=r"C:\Users\86156\Desktop\研二下\开题报告\文献\开挖变形\监测"
#读取文件名
#储存引用容器
cite_box=[]
#default
default=[]
# 遍历当前路径下所有文件
file = os.listdir(file_address)
for f in file:
    #论文名容器
    thesis=[]
    #论文名
    try:
        name=re.search(r".*_",f).group()
        name=name.replace("_","")
    except:
        default.append(f)
        continue
    thesis.append(name)
    #论文作者
    e=f.split("_")
    peo=e[len(e)-1].replace(".pdf","")
    thesis.append(peo)
    #print(thesis)

    driver = webdriver.Chrome()
    driver.implicitly_wait(5)
    #print("读取js")
    with open('JS/stealth.min.js') as f:
        js = f.read()
    driver.execute_cdp_cmd('Page.addScriptToEvaluateOnNewDocument', {'source': js})
    #print("读取完成，开始打开网页")
#driver.maximize_window()#最大化窗口
    #打开知网
    driver.get('https://www.cnki.net/')
    time.sleep(0.25)
    #打开输入栏，键入文章标题
    search=driver.find_element(By.XPATH, "/html/body/div[2]/div[2]/div/div[1]/input[1]")
    search_click=driver.find_element(By.XPATH, "/html/body/div[2]/div[2]/div/div[1]/input[2]")
    ActionChains(driver) \
        .send_keys_to_element(search, name)  \
        .click(search_click) \
        .perform()
    time.sleep(0.5)
    #搜索结果
    number=driver.find_element(By.XPATH,"/html/body/div[3]/div[2]/div[2]/div[2]/form/div/div[1]/div[1]/span/em").text
    number=int(number)
    #作者检查
    fla=True
    i=1
    while fla:
        try:
            us=driver.find_element(By.XPATH,"/html/body/div[3]/div[2]/div[2]/div[2]/form/div/table/tbody/tr["+str(i)+"]/td[3]/a[1]").text
        except:
            default.append(thesis[0])
            fla=False
        if us==thesis[1]:
            #print(number,i)
            # 两个中，第二个#gridTable > table > tbody > tr:nth-child(2) > td.operat > a.icon-quote
            # 两个中，第一个#gridTable > table > tbody > tr.odd > td.operat > a.icon-quote
            # 3~多个中，第i个#gridTable > table > tbody > tr:nth-child(i) > td.operat > a.icon-quote
            if number == 1 and i == 1:
                cite = driver.find_element(By.CSS_SELECTOR,"#gridTable > table > tbody > tr > td.operat > a.icon-quote")
                ActionChains(driver) \
                    .click(cite) \
                    .perform()
                time.sleep(0.5)
                cite_info = driver.find_element(By.CSS_SELECTOR,"#layui-layer1 > div.layui-layer-content > table > tbody > tr:nth-child(1) > td.quote-r").text
                time.sleep(0.5)
            elif number == 2 and i == 1:
                cite = driver.find_element(By.CSS_SELECTOR,"#gridTable > table > tbody > tr.odd > td.operat > a.icon-quote")
                ActionChains(driver) \
                    .click(cite) \
                    .perform()
                time.sleep(0.5)
                cite_info = driver.find_element(By.CSS_SELECTOR,"#layui-layer1 > div.layui-layer-content > table > tbody > tr:nth-child(1) > td.quote-r").text
                time.sleep(0.5)
            elif number == 2 and i == 2:
                cite = driver.find_element(By.CSS_SELECTOR,"#gridTable > table > tbody > tr:nth-child("+str(i)+") > td.operat > a.icon-quote")
                ActionChains(driver) \
                    .click(cite) \
                    .perform()
                time.sleep(0.5)
                cite_info = driver.find_element(By.CSS_SELECTOR,"#layui-layer1 > div.layui-layer-content > table > tbody > tr:nth-child(1) > td.quote-r").text
                time.sleep(0.5)
            elif number > 2:
                cite = driver.find_element(By.CSS_SELECTOR, "#gridTable > table > tbody > tr:nth-child(" + str(i) + ") > td.operat > a.icon-quote")
                ActionChains(driver) \
                    .click(cite) \
                    .perform()
                time.sleep(0.5)
                cite_info = driver.find_element(By.CSS_SELECTOR,"#layui-layer1 > div.layui-layer-content > table > tbody > tr:nth-child(1) > td.quote-r").text
                time.sleep(0.5)
            #引用信息
            cite_box.append(cite_info)
            #跳出
            fla=False
        i=i+1
print("获取引用文献 ",len(cite_box)," 篇。")
for ci in cite_box:
    # 姓名省略
    names = re.search(r".*?\.", ci).group()
    # print("前段:",names)
    name_li = names.split(",")
    be_name = names
    if len(name_li) > 3:
        be_name = name_li[0] + "," + name_li[1] + "," + name_li[2] + ",等."
    # 后段
    af = ci.split(".")
    af_content = af[1] + af[2] + "."
    # 拼接
    cite_info_per = be_name + af_content
    # print("后段:",af_content)
    # 打空格
    ci = cite_info_per.replace(",", ", ").replace(".", ". ").replace(":", ": ")
    print(ci)
if len(default)!=0:
    print("============失败文件:===========")
    for i in  default:
        print(i)

例：

文件夹内容如下：

获取结果如下：

 缺点：缺点很多，包括英文文章没有考虑，文献命名格式固定等等，不过作为一个小demo玩玩还是可以的。