jupyterlite-simple-cors-proxy

Simple CORS proxy wrapper for making http requests from JupyterLite. Uses https://corsproxy.io/

Installation

pip install jupyterlite-simple-cors-proxy

Usage

from jupyterlite_simple_cors_proxy.proxy import cors_proxy_get, robust_get_request, furl, xurl

# Set up
url = "https://api.example.com/data"
# Optional params
params = {"key": "value"}

# Get a cross-origin proxied url
cross_origin_url = xurl(url) # xurl(url, params)

# Get a file like object
# (Make the request, then create a file like object
# from the response)
file_ob = furl(url) # furl(url, params)

# Make a request
response = cors_proxy_get(url, params)

# Use like requests
print(response.text)
data = response.json()
raw = response.content

The robust_get_request() will first try a simple request, then a proxied request: robust_get_request(url, params)

Features

• Simple CORS proxy wrapper
• Requests response object
• Support for URL parameters

Using an API key with a CORS proxy

Some proxies, such as corsproxy.io, support (or require) an API key, passed as ?key=abcd1234&url=.... Pass it via the api_key argument on any of the helper functions:

from jupyterlite_simple_cors_proxy.proxy import xurl, cors_proxy_get

cross_origin_url = xurl(url, api_key="abcd1234")
response = cors_proxy_get(url, params, api_key="abcd1234")

Or set a default api_key once when using the CorsProxy class:

from jupyterlite_simple_cors_proxy.cacheproxy import CorsProxy

proxy = CorsProxy(api_key="abcd1234")
response = proxy.cors_proxy_get(url, params)  # api_key applied automatically

fastf1 cors proxy

A monkey patch for fastf1 is provided as:

import fast f1
from jupyterlite_simple_cors_proxy.fastf1_proxy import enable_cors_proxy

enable_cors_proxy(
#    domains=["api.formula1.com", "livetiming.formula1.com"],
#    debug=True,
#    proxy_url="https://corsproxy.io/",
)

CorsProxy with cache facility

Via claude.ai, the package is now further enriched.

Note that pyodide sqlite can't write to /drive so the cache path dir needs to be something like /tmp or a dir created on /.

I'm not convinced the following works in pyodide and xeus-python yet - requests-cache dependency issues etc. requests-cache has requirements attrs, cattrs,platformdirs, url-normalize.

from simple_cors_proxy.proxy import CorsProxy

# Create a cached proxy instance
proxy = CorsProxy(use_cache=True, expire_after=3600)  # Cache for 1 hour

# Optionally pass an api_key (and/or proxy) to use by default
proxy = CorsProxy(use_cache=True, expire_after=3600, api_key="abcd1234")

# Use furl directly from your proxy instance
file_like = proxy.furl('https://example.com/somefile.csv')

#----
import pandas as pd
from simple_cors_proxy.cacheproxy import CorsProxy

proxy = CorsProxy(use_cache=True)
file_like = proxy.furl('https://example.com/data.csv')
df = pd.read_csv(file_like)

#----

from simple_cors_proxy.proxy import create_cached_proxy

proxy = create_cached_proxy(cache_name='my_cache', expire_after=86400)  # Cache for 1 day
file_like = proxy.furl('https://example.com/somefile.csv')