Real time data from web page with python

[Anonymous]

I am trying to create a real time report from the following website. 

The website updates data every hour by generating a new .zip file at the bottom of the list.

The difficulty is twofold:

1. the .csv data is zipped;
2. the name of the desired file changes every hour. Therefore, the url changes as well;

I have written the following python script to download the latest report which works when I run it in the Command Prompt:

import requests
import os
import pandas as pd
from bs4 import BeautifulSoup
from urllib.request import urlopen
from io import BytesIO
from zipfile import ZipFile

# important addresses
parent_dir = "http://www.nemweb.com.au"
URL = "http://www.nemweb.com.au/REPORTS/CURRENT/Short_Term_PASA_Reports/"
dest_dir = "D:\Python\DELWP"

# web scraping
page = requests.get(URL)
soup = BeautifulSoup(page.content, "html.parser")
zip_reports = soup.find_all("a")

report_link = parent_dir + zip_reports[-1].get("href")
resp = urlopen(report_link)

# extracting zip file and converting it to dataframe
zip_file = ZipFile(BytesIO(resp.read()))
fname = zip_file.namelist()[0]
df = pd.read_csv(zip_file.open(fname), skiprows=3, dtype=object)
zip_file.close()

print(df)

However, it doesn't work when I execute it from Python Script in PowerBI.

I get the following error:

python error

I have set the Python Home directory to the same address that I get when I type where python in the command prompt.

home directory

execution and location

Can someone help, please?

PS: The solutions proposed by the responses in this community message don't work because they treat the url address as permanent.

[Anonymous · Accepted solution]

OK. I solved the problem. It was a matter of reinstalling a few packages.

I did the following:

1. pip uninstall numpy; then pip install numpy;
2. pip uninstall matplotlib; pip install matplotlib;
3. pip install --upgrade Pillow

Cheers