django-charitychecker 1.3

a small django app to verify information about nonprofits using their EINs

django-charitychecker

A django app for verifying U.S. nonprofits by their EIN number. For a general python module with similar functionality, check out charitycheck.

Visit the github.

Installation:

From pip:

pip install django-charitychecker

In your settings.py:

INSTALLED_APPS = (
    ...
    charitychecker
    ...
    )

Note that if you want django-charitychecker to work properly, you’ll need to periodically fetch data from the IRS to update your local database.

Useage:

django-charitychecker only checks U.S. based nonprofits, this means that it will not verify foreign nonprofits registered with the IRS (there are a few, but the data entry format for these organizations is inconsistent and so can’t be machine parsed easily).

django-charitychecker makes the following functions and models public:

Public Objects and Methods

Models

charitychecker.IRSNonprofitData

a model representing a nonprofit with the following fields:

Attributes

• IRSNonprofitData.ein: a string representing the nonprofit’s EIN number.

• IRSNonprofitData.name: a string representing the name of the Nonprofit as published in IRS Publication 78.

• IRSNonprofitData.city: a string representing the city in which the nonprofit is incorporated.

• IRSNonprofitData.state: a string representing the state in which the nonprofit is incorporated.

• IRSNonprofitData.country: a string representing the country in which the nonprofit is incorporated.

• IRSNonprofitData.deductability_code: the tax deductability code of the nonprofit organization. See the charitycheck README for more information on deductability codes.

Of course, you can retrieve a nonprofit’s info using their EIN by IRSNonprofitData.objects.get(ein="some ein string").

Methods

• IRSNonprofitData.verify_nonprofit(ein, name=None, city=None, state=None, country=None, deductability_code=None): return true if there is a nonprofit in the charitychecker database with information matching the information provided to the function as arguments, return false otherwise.

• IRSNonprofitData.get_deductability_code(ein, name=None, city=None, state=None, country=None): if a nonprofit is found in the charitychecker database with information matching the information provided as arguments to the function, then return that nonprofit’s deductability code, otherwise return the empty string.

Utilities

ignore_blank_space

A function which takes an iterator (it’s meant for file like objects but should take any iterator returning strings), and returns that iterator skipping all blank lines and stripping all white space from the end of any line.

open_zip_from_url

A context manager taking two arguments, zip_url and file_name. The context manager downloads the file from zip_url, attempts to unzip and then return the file at the path file_name from the downloaded zip archive. It does all of this in memory without writing to disk.

irs_nonprofit_data_context_manager

A context manager for dealing with the IRS Publication 78 data. It automatically downloads, unzips, and opens the file from the IRS website, closing the resources when finished. It skips any FORGN registered nonprofits, and returns all of the nonprofit data strings as a generator. Use is as you would any context manager:

with IRSNonprofitDataContextManager() as irs_data:
    # code using irs_data as a file object
    ....

The Publication 78 file is a series of lines of the form ein|name|city|state|country|deductability_code.

update_database_from_file

A function that takes file_manager, convert_line, pk_field, and model arguments, then uses a file to update in bulk your application’s database. Specifically, it takes:

• file_manager: a context manager which returns a generator yielding strings. Conceptually, we think of this iterator as a file like object, but any generator should work.

• convert_line: a function whose input is a line yielded by file_manager and that outputs a dictionary mapping keys which are field names of model and values which will be used to instantiate those field names.

• pk_field: The name field which is defined to be the primary key on model. This field should not be an AutoField for the following reasons: 1.) bulk updating commands in Django do not call the save method on the model, and thus do not set AutoField primary keys, 2.) if you are updating your database from some third-party source data, you want to be able to identify each line in the third-party data uniquely, thus some primary-key like value/unique identifier should already exist in your source data. Using an AutoField instead would mean that the function would have no way of distinguishing new data and old data that’s been updated.

• model: the model you want to update.

Note! This function will delete any data that is in your database and not present in the source file provided by file_manager.

update_charitychecker_data

A function that, when called, downloads a fresh copy of the IRS Publication 78 data, unzips it, and uses it to update the charitychecker database. It accepts no arguments.

Management Commands

update_charitychecker_data

A command that downloads a new copy of the IRS Publication 78 data, unzips it, and uses the data to update the charitychecker database.

This command takes no arguments, and only the default django command options. Run it by typing into the command prompt:

python manage.py update_charitychecker_data

Of course, you can only run the command after charitychecker is installed into your project’s settings.py file’s INSTALLED_APPS, and you’ve run python manage.py syncdb. This command could take a long time to finish, because it checks that your entire nonprofit database (800,000+ rows) is up to date.

Testing

Test the app as you would any other Django app. In a project with charitychecker installed run:

python manage.py test charitychecker

Note that since the test suite tests downloading and unzipping the data from the IRS, the test suite can take a minute or so, if you want to speed up the tests simply skip the tests using irs_nonprofit_data_context_manager and the test suite should run much faster.

Contributing

Pull requests are welcome. Also check out charitycheck for a general python equivalent of this package if you are interested in contributing.

Suggestions for contributing:

• python 3 support

• incorporate this module into a larger family of modules/packages dealing with nonprofit data (but of course, following the UNIX philosophy is really important here. Do one thing and do it well!)

• writing tests for the open_zip_from_url and update_database_from_file functions which currently are only tested indirectly by the tests on irs_nonprofit_data_context_manager and update_charitychecker_data, which are both wrappers for these more general functions.

• add tests for the newly added string methods on the IRSNonprofitData model.

Authors

Authored by Nicholas A. Lourie. You can contact him at [e-mail].