Skip to content

Python library to scrape social media data via the EnsembleData API.

License

Notifications You must be signed in to change notification settings

EnsembleData/ensembledata-python

Repository files navigation

EnsembleData Python API

pypi pypi

Documentation

Check out the API docs to see which endpoints are available for each social media and for detailed descriptions of their parameters and functionality.

Installation

Install the package with pip as you would any other python package.

pip install ensembledata

Requirements

  • The package only supports Python 3.7 and above.

Usage

Register to get your free API token.

from ensembledata.api import EDClient


client = EDClient("API-TOKEN")
result = client.tiktok.user_info_from_username(username="daviddobrik")

print("Data: ", result.data)
print("Units charged:", result.units_charged)

# Other Examples:
# result = client.instagram.user_info(username="daviddobrik")
# result = client.youtube.channel_subscribers(channel_id="UCnQghMm3Z164JFhScQYFTBw")

TikTok Guide

Get started with the TikTok API Guide. Find out which endpoints are available, and how you can use them with examples using the ensembledata python package.

Instagram Guide

Get started with the Instagram API Guide. Find out which endpoints are available, and how you can use them with examples using the ensembledata python package.



Missing Endpoints / Parameters

If you find that one of the endpoints from our API docs is not yet available in this package, you can use the EDClient.request method to specify the endpoint manually in the meantime.

from ensembledata.api import EDClient

client = EDClient("API-TOKEN")
result = client.request("/instagram/[example]", params={"foo": "...", "bar": "..."})

If you find that one the parameters to an existing endpoint is missing, you can still send this parameter via the extra_params dictionary which is available on all endpoint methods. See the example below:

from ensembledata.api import EDClient

client = EDClient("API-TOKEN")
result = client.instagram.user_info(username="...", extra_params={"baz": "..."})


Handling Errors

In the API docs, each endpoint lists a handful of possible errors the API may respond with. You can handle these errors by catching the EDError exception.

from ensembledata.api import EDClient, EDError, errors


client = EDClient("API-TOKEN")
try:
    result = client.tiktok.user_info_from_username(username="daviddobrik")
except EDError as e:

    # Rate limit exceeded...
    if e.status_code == errors.STATUS_429_RATE_LIMIT_EXCEEDED:
        print(e.detail)

    # Subscription expired...
    if e.status_code == errors.STATUS_493_SUBSCRIPTION_EXPIRED:
        print(e.detail)

except Exception as e:
    # Some other error occurred, unrelated to the EnsembleData API
    # E.g. httpx.RequestError, json.JSONDecodeError
    pass

    


Async

This package provides an asynchronous client, EDAsyncClient, which will give you access to async versions of all the same methods that can be found on the EDClient.

import asyncio

from ensembledata.api import EDAsyncClient


async def main():
    client = EDAsyncClient("API-TOKEN")
    result = await client.tiktok.user_info_from_username(username="daviddobrik")

if __name__ == "__main__":
    asyncio.run(main())

Network Retries

By default the EDClient will perform 3 retries when it encounters network issues. If you would like to customise this behaviour, you can pass in the max_network_retries param as show below:

Note: if the request times out, it will not be retried.

from ensembledata.api import EDClient

client = EDClient("API-TOKEN", max_network_retries=5)

Configure Timeout

If you would like control over the read timeout, you can configure this either for all request by setting timeout when creating the EDClient, or you can specify the timeout per request, on any of the individual methods as shown below:

Note: the timeout is specified in seconds.

from ensembledata.api import EDClient

client = EDClient("API-TOKEN", timeout=120)
result = client.tiktok.user_info_from_username(username="daviddobrik", timeout=10)

Types

The package uses type hints, and is type checked with the latest version of mypy. If you experience any type checking related issues with the package, please let us know by creating an issue.