olihawkins / feedstream

A Python package for downloading data from Feedly.

feedstream

A Python package for downloading data from the Feedly API. This is not a wrapper for the full Feedly API but a Python library that is designed to support a particular workflow, in which articles are classified in Feedly by saving them to boards and the data on the articles is downloaded for analysis.

Usage

Simply set the API token and timezone in config.json and download data on all articles that have been saved as entries to boards in Feedly with:

import feedstream as fs
download = fs.download_entries()

This returns a dictionary with three keys:

• timestamp contains a Feedly timestamp of the time this data was downloaded from Feedly
• fieldnames is a list of the fieldnames used as keys for each item in the items list
• entries is a list of all entries saved to boards, along with the id and name of their board

Entries can be downloaded directly to a csv with fs.download_entries_csv(), or to a pandas dataframe with timestamp, df = fs.download_entries_df().

You can run the package as a program directly from the command line with python -m feedstream, which downloads the data to a csv in the application data directory. You can set feedstream to only download articles that have been added to boards since the last time data was saved by setting download_new to True in config.json.

Tests

Run python -m unittest -v to run the unit tests.