时间戳

首页 > 解决方案 > 使用名称从网站上抓取数据表

python - 使用名称从网站上抓取数据表

问题描述

在尝试抓取网站时，我遇到了一个独特的情况。我正在通过搜索栏搜索数百个名称，然后抓取表格。但是，与网站相比，我的列表中的某些名称是唯一的并且拼写不同。在这种情况下，我在网站上手动查找了几个名字，它仍然直接将我带到各个页面。其他时候，如果有多个名字相同或相似的人，它会进入名字列表（在这种情况下，我想要在 nba 打球的人。我已经考虑过了，但我认为有必要提及）。我该如何继续进入这些玩家的个人页面，而不必每次都运行脚本并点击错误以查看哪个玩家的拼写略有不同？再次，即使拼写略有不同或名称列表（需要 NBA 中的名称），数组中的名称也会直接将您带到单个页面。一些例子是 Georgios Papagiannis（在网站上列为 George Papagiannis）、Ognjen Kuzmic（列为 Ognen Kuzmic）、Nene（列为 Maybyner Nene，但会带您进入名单——https://basketball.realgm.com/search?q=nene）。这似乎很难，但我觉得这可能是可能的。此外，似乎不是将所有抓取的数据写入 csv，而是每次都被下一个玩家覆盖。万分感谢。

我得到的错误： AttributeError: 'NoneType' object has no attribute 'text'

import requests
from bs4 import BeautifulSoup
import pandas as pd


playernames=['Carlos Delfino', 'Nene', 'Yao Ming', 'Marcus Vinicius', 'Raul Neto', 'Timothe Luwawu-Cabarrot']

result = pd.DataFrame()
for name in playernames:

    fname=name.split(" ")[0]
    lname=name.split(" ")[1]
    url="https://basketball.realgm.com/search?q={}+{}".format(fname,lname)
    response = requests.get(url)
    soup = BeautifulSoup(response.content, 'html.parser')

    if soup.find('a',text=name).text==name:
        url="https://basketball.realgm.com"+soup.find('a',text=name)['href']
        print(url)
        response = requests.get(url)
        soup = BeautifulSoup(response.text, 'lxml')

    try:
        table1 = soup.find('h2',text='International Regular Season Stats - Per Game').findNext('table')
        table2 = soup.find('h2',text='International Regular Season Stats - Advanced Stats').findNext('table')

        df1 = pd.read_html(str(table1))[0]
        df2 = pd.read_html(str(table2))[0]

        commonCols = list(set(df1.columns) & set(df2.columns))
        df = df1.merge(df2, how='left', on=commonCols)
        df['Player'] = name
        print(df)
    except:
        print ('No international table for %s.' %name)
        df = pd.DataFrame([name], columns=['Player'])

result = result.append(df, sort=False).reset_index(drop=True)

cols = list(result.columns)
cols = [cols[-1]] + cols[:-1]
result = result[cols]
result.to_csv('international players.csv', index=False)

标签： pythonpandasweb-scrapingbeautifulsoup

解决方案

我对名字相似的 NBA 球员使用了循环。您可以在下面的 css 选择器中找到从搜索表中获取 NBA 球员的信息：

.tablesaw tr:has(a[href*="/nba/teams/"]) a[href*="/player/"]

CSS选择器含义：按tablesaw类查找table，查找table的children ，其trchildren包含text，然后查找其包含ahref/nba/teams/ahref/player/

我添加了Search Player Name和Real Player Name列，您可以查看如何找到玩家。使用此列作为第一列和第二列放置insert（请参阅代码中的注释）。

import requests
from bs4 import BeautifulSoup
import pandas as pd
from pandas import DataFrame

base_url = 'https://basketball.realgm.com'
player_names = ['Carlos Delfino', 'Nene', 'Yao Ming', 'Marcus Vinicius', 'Raul Neto', 'Timothe Luwawu-Cabarrot']

result = pd.DataFrame()


def def get_player_stats(search_name = None, real_name = None, player_soup = None):
    table_per_game = player_soup.find('h2', text='International Regular Season Stats - Per Game')
    table_advanced_stats = player_soup.find('h2', text='International Regular Season Stats - Advanced Stats')

    if table_per_game and table_advanced_stats:
        print('International table for %s.' % search_name)

        df1 = pd.read_html(str(table_per_game.findNext('table')))[0]
        df2 = pd.read_html(str(table_advanced_stats.findNext('table')))[0]

        common_cols = list(set(df1.columns) & set(df2.columns))
        df = df1.merge(df2, how='left', on=common_cols)

        # insert name columns for the first positions
        df.insert(0, 'Search Player Name', search_name)
        df.insert(1, 'Real Player Name', real_name)
    else:
        print('No international table for %s.' % search_name)
        df = pd.DataFrame([[search_name, real_name]], columns=['Search Player Name', 'Real Player Name'])

    return df


for name in player_names:
    url = f'{base_url}/search?q={name.replace(" ", "+")}'
    response = requests.get(url)
    soup = BeautifulSoup(response.content, 'html.parser')

    if url == response.url:
        # Get all NBA players
        for player in soup.select('.tablesaw tr:has(a[href*="/nba/teams/"]) a[href*="/player/"]'):
            response = requests.get(base_url + player['href'])
            player_soup = BeautifulSoup(response.content, 'lxml')
            player_data = get_player_stats(search_name=player.text, real_name=name, player_soup=player_soup)
            result = result.append(player_data, sort=False).reset_index(drop=True)
    else:
        player_data = get_player_stats(search_name=name, real_name=name, player_soup=soup)
        result = result.append(player_data, sort=False).reset_index(drop=True)

result.to_csv('international players.csv', index=False)
# Append to existing file
# result.to_csv('international players.csv', index=False, mode='a')

推荐阅读