Master Python Web Scraping & Data Extraction with Requests, lxml, pandas

1. Introduction

Earlier in a Python community a user asked about a Python web‑scraping problem; the following solution is shared.

The original code uses requests, lxml, and csv to crawl GDP data from a website, extract rows via XPath, and write them to a CSV file.

import requests
from lxml import etree
import csv
import time
import pandas as pd 

def gdpData(page):
    url = f'https://www.hongheiku.com/category/gdjsgdp/page/{page}'
    headers ={'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36'}
    resp = requests.get(url,headers = headers)
    # print(resp.text)
    data(resp.text)
file = open('data.csv',mode='a',encoding='utf-8',newline='')
csv_write=csv.DictWriter(file,fieldnames=['排名','地区','GDP','年份'])
csv_write.writeheader()

def data(text):
    e = etree.HTML(text)
    lst = e.xpath('//*[@id="tablepress-48"]/tbody/tr[@class="even"]')
    for l in lst:
        no = l.xpath('./td[1]/center/span/text()')
        name = l.xpath('./td[2]/a/center/text()')
        team = l.xpath('./td[3]/center/text()')
        year = l.xpath('./td[4]/center/text()')
        data_dict ={
            '排名':no,
            '地区':name,
            'GDP':team,
            '年份':year
        }
        print(f'排名：{no} 地区:{name} GDP:{team} 年份:{year} ')
        csv_write.writerow(data_dict)
file.close()
url = 'https://www.hongheiku.com/category/gdjsgdp'
headers ={'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36'}
resp = requests.get(url,headers = headers)
# print(resp.text)
data(resp.text)
e = etree.HTML(resp.text)
#//*[@id="tablepress-48"]/tbody/tr[192]/td[3]/center
count = e.xpath('//div[@class="pagination pagination-multi"][last()]/ul/li[last()]/span/text()')[0].split(' ')[1]
for index in range(int(count) - 1):
    gdpData(index + 2)

2. Implementation

The original script works but sometimes fails to retrieve data because the XPath rules are not robust.

A revised solution leverages pandas to read HTML tables directly, simplifying extraction and eliminating missing data.

import requests
import pandas as pd 
url = 'https://www.hongheiku.com/category/gdjsgdp'
headers ={'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/109.0.0.0 Safari/537.36'}
resp = requests.get(url,headers = headers)
df = pd.read_html(resp.text)[0].dropna()
df.to_excel('1.xlsx',index=None)
df

The pandas approach produces a complete dataset without gaps.

Another participant also wrote a working script that yields the expected results.

3. Conclusion

This article presented a Python web‑scraping issue, provided detailed analysis and code implementations, and helped the user obtain the desired data.