Python -使用pandas和enumerate进行web爬行

Question

所以我最近在网上找到了这段代码，它是用python编写的，在pandas中使用了enumerate表达式。

import pandas as pd
url = 'http://myurl.com/mypage/'

for i, df in enumerate(pd.read_html(url)):
    df.to_csv('myfile_%s.csv' % i)

有没有办法重写它，这样它就可以遍历网页列表而不是单个url，并将每个页面表格中的所有信息放入一个.csv文件中？我的主要猜测类似于for循环。

url_base = 'http://myurl.com/mypage/'
count = 1
for i in range(1,5):

    url = '%s%s' %(url_base,count)

    for i, df in enumerate(pd.read_html(url)):
        df.to_csv('myfile_%s.csv' % i)
    count = count + 1  

Answer 1

如果您的所有csvs都具有相同的列，则可以执行以下操作

pd.concat([pd.read_html(url) for url in urls], ignore_index=True)

如果您的urls与示例中的urls具有相同的基数，您将执行以下操作

url_base = 'http://myurl.com/mypage/{}'
df = pd.concat([pd.read_html(base.format(i)) for i in range(num)], ignore_index=True)
df.to_csv('alldata.csv')

Answer 2

这个怎么样？

import pandas as pd
from concurrent import futures

urls = [your list of urls]

def read_html(url):
    return pd.read_html(url)

with futures.ThreadPoolExecutor(max_workers=6) as executor:
    fetched_urls = dict((executor.submit(read_html, url), url)
                         for url in urls)

nums = range(1, len(fetched_url)+1)

for future, num in zip(futures.as_completed(fetched_urls), nums):
    if future.result():
        future.result().to_csv('myfile_{}.csv'.format(num), index=False)
    elif future.exception():
        print '{} yielded no results'.format(fetched_urls[future])
    else:
        pass