LfApi: a Python SDK

Installation

The latest version of the library can be installed directly from PyPI:

pip install lfapi

Usage

Getting Started

The lfapi.Client class is the main interface with the ListenFirst API:

from lfapi import Client

It requires at minimum an API key as well as a set of client credentials for access token generation. Token generation is handled by the lfapi.Auth class:

from lfapi import Auth
auth = Auth(<CLIENT_ID>, <CLIENT_SECRET>)
token = auth.access_token

The Client class is instantiated with an API key and an instance of the Auth class:

client = Client(<API_KEY>, auth)

Alternatively, all three credentials can be loaded directly to the Client class from a JSON file:

client = Client.load(<JSON_FILE_NAME_OR_FILE_OBJECT>)

# The JSON file should follow this schema:
# {
#   "api_key": <api_key>,
#   "client_id": <client_id>
#   "client_secret": <client_secret>
# }

These credentials can be retrieved from the platform's API settings page.

Accessing the API

Once instantiated, the Client object can be used to make customized HTTP requests to various API endpoints. The lowest-level request mechanism is built around two methods, secure_get() and secure_post(). Each takes a positional endpoint argument, as well as a params argument as in the requests library. secure_post() additionally takes a json argument, again mirroring the requests library. Both return requests.Response objects upon successful completion:

• client.secure_get(endpoint, params=None)
  Make a secure GET request to the ListenFirst API.

• client.secure_post(endpoint, json=None, params=None)
  Make a secure POST request to the ListenFirst API.

Commonly used endpoints have dedicated instance methods:

• client.fetch(json)
POST request to /analytics/fetch to perform a synchronous query.

• client.create_fetch_job(json)
POST request to /analytics/fetch_job to create an asynchronous query.

• client.show_fetch_job(job_id)
GET request to /analytics/fetch_job/{id} to view a summary of an existing asynchronous query.

• client.latest_fetch_job(params=None)
GET request to /analytics/fetch_job/latest to view a summary of the most recent asynchronous query.

• client.list_fetch_jobs(params=None)
GET request to /analytics/fetch_job to view an abridged summary for all asynchronous queries.

• client.create_schedule_config(json)
POST request to /analytics/schedule_config to create an schedule configuration.

• client.show_schedule_config(schedule_config_id)
GET request to /analytics/schedule_config/{id} to view a summary of an existing schedule configuration.

• client.list_schedule_configs(params=None)
GET request to /analytics/schedule_config to view an abridged summary for all schedule configurations.

• client.get_brand(brand_id, params=None)
GET request to /brand_views/{id} to view a summary of a brand view.

• client.list_brands(params=None)
GET request to /brand_views to view a summary for all brand views.

• client.get_brand_set(brand_set_id)
GET request to /brand_view_sets/{id} to view a summary of a brand view set.

• client.list_brand_sets(params=None)
GET request to /brand_view_sets to view a summary for all brand view sets.

• client.get_dataset(dataset_id)
GET request to /dictionary/datasets/{id} to view a summary of a dataset.

• client.list_datasets()
GET request to /dictionary/datasets to view an abridged summary for all datasets.

• client.get_field_values(params)
GET request to /dictionary/field_values to view a list of values for a given field.

With the exception of get_field_values(), these methods wrap the API responses in instances of lfapi.Model subclasses. These wrapper classes offer some convenient extended functionality, such as JSON and CSV conversion.

In addition, the Client object implements a number of convenience methods around the /analytics endpoints for managing data queries:

• client.poll_fetch_job(job_id)
  Pull fetch job summary until state is one of 'completed', 'failed'.

• client.sync_analytic_query(fetch_params, per_page=None, max_pages=inf)
  Run multiple pages of synchronous analytic queries.

• client.async_analytic_query(fetch_params, client_context=None, max_rows=None, emails=None)
  Construct and poll an async analytic query, and download page URLs upon completion.

For code examples, see our examples wiki.