Skip to content

Latest commit

 

History

History
32 lines (26 loc) · 1.8 KB

README.md

File metadata and controls

32 lines (26 loc) · 1.8 KB

irs-990-parse: finding value in 3.5M Non-profit tax returns

This repository contains guidance and Python scripts for parsing IRS 990 releases in Amazon S3

To begin, you'll need to download the index.json file from the S3 bucket, it's 1Gb so don't try it on a mobile tether 👎

The easiest method is to install the AWS CLI and run the following command from your desired destination folder: aws s3 cp s3://irs-form-990/index.json index.json

the CLI should download the file in parts for you, and you'll end up with a big, big JSON file.

The contents of this file are a list of objects contained within a top-level AllFilings tag, each similar to:

{
	"EIN": "270678774",
	"SubmittedOn": "2016-02-09",
	"TaxPeriod": "201509",
	"DLN": "93492308002265",
	"LastUpdated": "2016-03-21T17:23:53",
	"URL": "https://s3.amazonaws.com/irs-form-990/201513089349200226_public.xml",
	"FormType": "990EZ",
	"ObjectId": "201513089349200226",
	"OrganizationName": "KIWANIS CLUB OF GLENDORA PROJECTS FUND INC",
	"IsElectronic": true,
	"IsAvailable": true
}

Not all attributes are available in all objects, for example if IsAvailable is false, URL will not be present.

Parsing through the file in-memory is generally not possible, so a streaming approach can be used through the module ijson. To acquire ijson, run pip install ijson from your Python environment.

Some useful filtering options available for this large file are to search by OrganizationName and by IsAvailable. An example is provided in filter_vt_records.py. This will parse the entire file and pull out any record containing the word "VERMONT", exporting it to a more manageable CSV: Sample Image