使用带有循环的漂亮汤在 Python 中抓取交互式图表

Question

The below code provide information from all the numeric tags in the page. Can I use a filter to extract once for each region

For example : https://opensignal.com/reports/2019/04/uk/mobile-network-experience , I am interested in numbers only under the regional analysis tab and for all regions.

import requests
from bs4 import BeautifulSoup

html=requests.get("https://opensignal.com/reports/2019/04/uk/mobile-network-experience").text
soup=BeautifulSoup(html,'html.parser')
items=soup.find_all('div',class_='c-ru-graph__rect')


for item in items:
    provider=item.find('span', class_='c-ru-graph__label').text
    prodvalue=item.find_next_sibling('span').find('span', class_='c-ru-graph__number').text
    print(provider + " : " + prodvalue)

I want a table or df as below Easter Region

                       o2      Vodaphone   3    EE
4G Availability        82      76.9        73.0   89.2
Upload Speed Experience 5.6    5.9         6.8    9.5

Any pointers that can help in getting the result ?

Answer 1

Assuming fixed the order of companies (it is, indeed), you can simply reduce the content to examine to only those div's containing the information you need.

import requests
from bs4 import BeautifulSoup

html = requests.get("https://opensignal.com/reports/2019/04/uk/mobile-network-experience").text
soup = BeautifulSoup(html,'html.parser')

res = soup.find_all('div', {'id':'eastern'})

aval = res[0].find_all('div', {'data-chart-name':'4g-availability'})
avalname = aval[0].find('span', {'class':'js-metric-name'}).text

upload = res[0].find_all('div', {'data-chart-name':'upload-speed'})
uploadname = upload[0].find('span', {'class':'js-metric-name'}).text

companies = [i.text for i in aval[0].find_all('span', class_='c-ru-graph__label')]

row1 = [i.text for i in aval[0].find_all('span', class_='c-ru-graph__number')]
row2 = [i.text for i in upload[0].find_all('span', class_='c-ru-graph__number')]

import pandas as pd

df = pd.DataFrame({avalname:row1,
                   uploadname:row2})


df.index = companies

df = df.T

output

                          O2    Vodafone      3      EE
4G Availability         82.0        76.9   73.0    89.2
Upload Speed Experience  5.6         5.9    6.8     9.5

Answer 2

Here is how I would do it for all regions. Requires bs4 4.7.1. AFAICS you have to assume consistent ordering of companies.

import requests
from bs4 import BeautifulSoup
import pandas as pd

r = requests.get("https://opensignal.com/reports/2019/04/uk/mobile-network-experience")
soup = BeautifulSoup(r.content,'lxml') #'html.parser' if lxml not installed
metrics = ['4g-availability', 'video-experience', 'download-speed' , 'upload-speed', 'latency']
headers = ['02', 'Vodaphone', '3', 'EE']
results = []

for region in soup.select('.s-regional-analysis__region'):
    for metric in metrics:
        providers = [item.text for item in region.select('.c-ru-chart:has([data-metric="' + metric + '"]) .c-ru-graph__number')]
        row = {headers[i] : providers[i] for i in range(len(providers))}
        row['data-metric'] = metric
        row['region'] = region['id'] 
        results.append(row)

df = pd.DataFrame(results, columns = ['region', 'data-metric', '02','Vodaphone', '3', 'EE'] )
print(df)

---

Sample output: