This repository contains the source code for Dwolla's Python-based SDK, which allows developers to interact with Dwolla's server-side API via a Python API. Any action that can be performed via an HTTP request can be made using this SDK when executed within a server-side environment.
To begin using this SDK, you will first need to download it to your machine. We use PyPi to distribute this package from where you can automagically download it via pip.
$ pip install dwollav2
Before any API requests can be made, you must first determine which environment you will be using, as well as fetch the application key and secret. To fetch your application key and secret, please visit one of the following links:
- Production: https://dashboard.dwolla.com/applications
- Sandbox: https://dashboard-sandbox.dwolla.com/applications
Finally, you can create an instance of Client
with key
and secret
replaced with the application key and secret that you fetched from one of the aforementioned links, respectively.
client = dwollav2.Client(
key = os.environ['DWOLLA_APP_KEY'],
secret = os.environ['DWOLLA_APP_SECRET'],
environment = 'sandbox', # defaults to 'production'
requests = {'timeout': 0.001}
)
An on_grant
callback is useful for storing new tokens when they are granted. The on_grant
callback is called with the Token
that was just granted by the server.
client = dwollav2.Client(
key = os.environ['DWOLLA_APP_KEY'],
secret = os.environ['DWOLLA_APP_SECRET'],
on_grant = lambda t: save(t)
)
It is highly recommended that you encrypt any token data you store.
Application access tokens are used to authenticate against the API on behalf of an application. Application tokens can be used to access resources in the API that either belong to the application itself (webhooks
, events
, webhook-subscriptions
) or the Dwolla Account that owns the application (accounts
, customers
, funding-sources
, etc.). Application tokens are obtained by using the client_credentials
OAuth grant type:
application_token = client.Auth.client()
Application access tokens are short-lived: 1 hour. They do not include a refresh_token
. When it expires, generate a new one using client.Auth.client()
.
The Dwolla Sandbox Dashboard allows you to generate tokens for your application. A Token
can be initialized with the following attributes:
client.Token(access_token = '...',
expires_in = 123)
Once you've created a Token
, currently, you can make low-level HTTP requests.
To make low-level HTTP requests, you can use the get()
, post()
, and delete()
methods. These methods will return a Response
object.
# GET api.dwolla.com/resource?foo=bar
token.get('resource', foo = 'bar')
# GET requests can also use objects as parameters
# GET api.dwolla.com/resource?foo=bar
token.get('resource', {'foo' = 'bar', 'baz' = 'foo'})
# POST api.dwolla.com/resource {"foo":"bar"}
token.post('resource', foo = 'bar')
# POST api.dwolla.com/resource multipart/form-data foo=...
token.post('resource', foo = ('mclovin.jpg', open('mclovin.jpg', 'rb'), 'image/jpeg'))
# DELETE api.dwolla.com/resource
token.delete('resource')
To set additional headers on a request you can pass a dict
of headers as the 3rd argument.
For example:
token.post('customers', { 'firstName': 'John', 'lastName': 'Doe', 'email': '[email protected]' },
{ 'Idempotency-Key': 'a52fcf63-0730-41c3-96e8-7147b5d1fb01' })
The following snippets demonstrate successful and errored responses from the Dwolla API.
An errored response is returned when Dwolla's servers respond with a status code that is greater than or equal to 400, whereas a successful response is when Dwolla's servers respond with a 200-level status code.
res = token.get('/')
res.status
# => 200
res.headers
# => {'server'=>'cloudflare-nginx', 'date'=>'Mon, 28 Mar 2016 15:30:23 GMT', 'content-type'=>'application/vnd.dwolla.v1.hal+json; charset=UTF-8', 'content-length'=>'150', 'connection'=>'close', 'set-cookie'=>'__cfduid=d9dcd0f586c166d36cbd45b992bdaa11b1459179023; expires=Tue, 28-Mar-17 15:30:23 GMT; path=/; domain=.dwolla.com; HttpOnly', 'x-request-id'=>'69a4e612-5dae-4c52-a6a0-2f921e34a88a', 'cf-ray'=>'28ac1f81875941e3-MSP'}
res.body['_links']['events']['href']
# => 'https://api-sandbox.dwolla.com/events'
If the server returns an error, a dwollav2.Error
(or one of its subclasses) will be raised.
dwollav2.Error
s are similar to Response
s.
try:
token.get('/not-found')
except dwollav2.NotFoundError as e:
e.status
# => 404
e.headers
# => {"server"=>"cloudflare-nginx", "date"=>"Mon, 28 Mar 2016 15:35:32 GMT", "content-type"=>"application/vnd.dwolla.v1.hal+json; profile=\"http://nocarrier.co.uk/profiles/vnd.error/\"; charset=UTF-8", "content-length"=>"69", "connection"=>"close", "set-cookie"=>"__cfduid=da1478bfdf3e56275cd8a6a741866ccce1459179332; expires=Tue, 28-Mar-17 15:35:32 GMT; path=/; domain=.dwolla.com; HttpOnly", "access-control-allow-origin"=>"*", "x-request-id"=>"667fca74-b53d-43db-bddd-50426a011881", "cf-ray"=>"28ac270abca64207-MSP"}
e.body.code
# => "NotFound"
except dwollav2.Error:
# ...
See https://developers.dwolla.com/api-reference#errors for more info.
dwollav2.AccessDeniedError
dwollav2.InvalidCredentialsError
dwollav2.NotFoundError
dwollav2.BadRequestError
dwollav2.InvalidGrantError
dwollav2.RequestTimeoutError
dwollav2.ExpiredAccessTokenError
dwollav2.InvalidRequestError
dwollav2.ServerError
dwollav2.ForbiddenError
dwollav2.InvalidResourceStateError
dwollav2.TemporarilyUnavailableError
dwollav2.InvalidAccessTokenError
dwollav2.InvalidScopeError
dwollav2.UnauthorizedClientError
dwollav2.InvalidAccountStatusError
dwollav2.InvalidScopesError
dwollav2.UnsupportedGrantTypeError
dwollav2.InvalidApplicationStatusError
dwollav2.InvalidVersionError
dwollav2.UnsupportedResponseTypeError
dwollav2.InvalidClientError
dwollav2.MethodNotAllowedError
dwollav2.ValidationError
dwollav2.TooManyRequestsError
dwollav2.ConflictError
Take a look at the
Sample Application for examples
on how to use this SDK to call the Dwolla API. Before you can begin using the app, however,
you will need to specify a DWOLLA_APP_KEY
and DWOLLA_APP_SECRET
environment variable.
- 2.2.1
- Add extra check in URL's to ensure they are clean. #36.
- 2.2.0
- Update JSON request bodies to serialize via
simplejson
so datatypes likeDecimal
still serialize like they did pre2.0.0
- Update JSON request bodies to serialize via
- 2.1.0
- Do not share
requests.session()
across instances ofdwollav2.Client
- Do not share
- 2.0.0
- JSON request bodies now contain sorted keys to ensure the same request body for a given set of
arguments, no matter the order they are passed to
dwolla.post
. This ensures theIdempotency-Key
header will work as intended without additional effort by developers. - NOTE: Because this change alters the formatting of JSON request bodies, we are releasing it
as a major new version. The request body of a request made with
1.6.0
will not match the request body of the same request made in2.0.0
. This will nullify the effect of theIdempotency-Key
header when upgrading, so please take this into account. If you have any questions please reach out to us! There are no other changes since1.6.0
.
- JSON request bodies now contain sorted keys to ensure the same request body for a given set of
arguments, no matter the order they are passed to
- 1.6.0 Allow configuration of
requests
options ondwollav2.Client
. - 1.5.0 Add integrations auth functionality
- 1.4.0
Pass kwargs from(Removed in v1.6)get
,post
, anddelete
methods to underlying requests methods. - 1.3.0 Change token URLs, update dependencies.
- 1.2.4 Create a new session for each Token.
- 1.2.3 Check if IOBase when checking to see if something is a file.
- 1.2.2 Strip domain from URLs provided to token.* methods.
- 1.2.1 Update sandbox URLs from uat => sandbox.
- 1.2.0 Refer to Client id as key.
- 1.1.8 Support
verified_account
anddwolla_landing
auth flags - 1.1.7 Use session over connections for performance improvement (#8 - Thanks @bfeeser!
- 1.1.5 Fix file upload bug when using with Python 2 (#6)
- 1.1.2 Add
TooManyRequestsError
andConflictError
- 1.1.1 Add MANIFEST.in
- 1.1.0 Support per-request headers
- If you have any feedback, please reach out to us on our forums or by creating a GitHub issue.
- If you would like to contribute to this library, bug reports and pull requests are always appreciated!
- After checking out the repo, run
pip install -r requirements.txt
to install dependencies. Then, runpython setup.py
test to run the tests. - To install this gem onto your local machine,
run pip install -e .
.
- After checking out the repo, run
If you prefer to use Docker to run dwolla-v2-python locally, a Dockerfile is included at the root directory. Follow these instructions from Docker's website to create a Docker image from the Dockerfile, and run it.
To learn more about Dwolla and how to integrate our product with your application, please consider visiting the following resources and becoming a member of our community!