Pandas read_csv from URL and include request header

Question

As of Pandas 0.19.2, the function read_csv() can be passed a URL. See, for example, from this answer:

import pandas as pd

url="https://raw.githubusercontent.com/cs109/2014_data/master/countries.csv"
c=pd.read_csv(url)

The URL I'd like to use is: https://moz.com/top500/domains/csv

With the above code, this URL returns an error:

urllib2.HTTPError: HTTP Error 403: Forbidden

based on this post, I can get a valid response by passing a request header:

import urllib2,cookielib

site= "https://moz.com/top500/domains/csv"
hdr = {'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.11 (KHTML, like Gecko) Chrome/23.0.1271.64 Safari/537.11',
       'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
       'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.3',
       'Accept-Encoding': 'none',
       'Accept-Language': 'en-US,en;q=0.8',
       'Connection': 'keep-alive'}

req = urllib2.Request(site, headers=hdr)

try:
    page = urllib2.urlopen(req)
except urllib2.HTTPError, e:
    print (e.fp.read())

content = page.read()
print (content)

Is there any way to use the web URL functionality of Pandas read_csv(), but also pass a request header to make the request go through?

score 14 · Answer 1 · answered Apr 16 '19 at 15:12

14

I would recommend you using the requests and the io library for your task. The following code should do the job:

import pandas as pd
import requests
from io import StringIO

url = "https://moz.com:443/top500/domains/csv"
headers = {"User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:66.0) Gecko/20100101 Firefox/66.0"}
req = requests.get(url, headers=headers)
data = StringIO(req.text)

df = pd.read_csv(data)
print(df)

(If you want to add a custom header just modify the headers variable)

Hope this helps

answered Apr 16 '19 at 15:12

Nazim Kerimbekov

4,712
8
34
58

Thanks - I was not aware of the IO package previously. If possible, could you explain what the advantage of putting `req.text` into StringIO vs. reading the url directly with pandas like `df = pd.read_csv(url)` - actually I see you edited the question to reflect the new pandas version - do you believe that is the more efficient way? – thesimplevoodoo Nov 21 '19 at 22:43
2

@thesimplevoodoo Hey, the reason why I'm using StringIO here is that `pd.read_csv()` is expecting a filepath so giving it `url` or any other string including (`req.text`) would yield an error. By having `data = StringIO(req.text)` I can then use `data` as a file path (Do note that StringIO doesn't create any actual files but gives you the chance to read and write strings as files) – Nazim Kerimbekov Nov 21 '19 at 23:19
2

This is a nice solution, though it should probably not be an accepted answer. It does not answer the OP's question: "Is there any way to use the web URL functionality of Pandas read_csv(), but also pass a request header to make the request go through?" I'm personally much more interested in the question as it pertained to `read_csv` and potential header usage. – Jason R Stevens CFA Sep 09 '21 at 14:34

score 7 · Accepted Answer · answered Jul 30 '22 at 19:09

As of pandas 1.3.0, you can now pass custom HTTP(s) headers using storage_options argument:

url = "https://moz.com:443/top500/domains/csv"

hdr = {
    'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.11 (KHTML, like Gecko) Chrome/23.0.1271.64 Safari/537.11',
    'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
    'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.3',
    'Accept-Encoding': 'none',
    'Accept-Language': 'en-US,en;q=0.8',
    'Connection': 'keep-alive'
}

domains_df = pd.read_csv(url, storage_options=hdr)

Pandas read_csv from URL and include request header

2 Answers2

Linked