LANGUAGE » PYTHON » PACKAGE

Requests

Last updated:
Reference:requests.readthedocs.io
Package:pypi

Usage ​

python
import requests

Request structure:

python
url = 'https://site.com/api'
params = {  # => '?greet=true'
    'greet': 'true'
}
headers = {
    'x-api-user': '12345678-90ab-416b-cdef-1234567890ab',
    'x-api-key': '12345678-90ab-416b-cdef-1234567890ab'
}
data = {
    'field1': 'data1',
    'field2': 'data2'
}

requests.get(url, params=params, headers=headers)
requests.post(url, headers=headers, data=data, json={})
requests.put(url, headers=headers, data=data, json={})
requests.delete(url, headers=headers)

Check the response content:

python
response.status_code
response.text
response.json()

Retry ​

Stackoverflow

Create a requests session instance, then mount it with a HTTP adapter using the desired retry configuration.

By creating a session, the underlying TCP connection is reused (reference), therefore the retries have better performance.

python
from urllib3.util.retry import Retry

with requests.Session() as session:
    session.mount('https://', requests.adapters.HTTPAdapter(
        max_retries=Retry(total=5, backoff_factor=2)
    ))
    response = session.get('https://example.com/')

Dealing with huge files ​

Reference

Use the file pointer instead of loading all its contents to avoid memory issues.

python
import requests

with open('file_50GB.bin', 'rb') as f:
    requests.put('http://myserv.com', data=f)           # OK (stream upload)
    # requests.put('http://myserv.com', data=f.read())  # NG