domolibrary 0.1.107

No project description provided

domolibrary: a powerful pydomo alternative

Formatting

use_snake_case for everything except class names is_bool - prefix any boolean parameters with is_bool

Patterns to recycle

UPSERT

domolibrary.classes.DomoUser upsert_user

search entity

user_routes.SearchUser_NoResults()

What is it?

domolibrary is a Python package that provides a OOP (class-based) and a functional approach to interacting with Domo’s API framework.

All accessed APIs are documented under DataCrew’s Domo Documentation page.

This library was created by DataCrew contributor Jae Wilson.

Install

The DataCrew team is hard at work expanding the list of available classes and routes. We have a ton of work completed, it’s just a matter of migrating and documenting the code into this library.

pip install domolibrary

How to use

Authentication

For each task, consider the appropriate DomoAuth mechanism. In most cases DomoFullAuth or DomoTokenAuth will be appropriate as this library predominately accesses private APIs.

Any Public routes or methods will be labeled appropriately in which case you should use DomoDeveloperAuth. Public routes are APIs enumerated and documented under Developer.Domo.com.

Typically each project will begin with configuring an auth object. If you are accessing multiple Domo instances, you’ll probably need multiple auth objects.

# configure an auth method
import os
import domolibrary.client.DomoAuth as dmda

token_auth = dmda.DomoTokenAuth( domo_instance = 'domo-community', domo_access_token = os.environ['DOMO_DOJO_ACCESS_TOKEN'])

Option 1: class based programming

In this project domo entities, DomoActivityLog, DomoDataset are all prefixed ‘Domo’ and can be found in the classes folder. Each class method will call one or more routes. Each route will interact with one and only one API.

Although most methods will be standard methods that will be called after creating an instance of the class, some methods will be classmethods which return an instance of the class.

In the example below, DomoDataset.get_from_id is a classmethod.

Note: DomoLibrary uses the asynchronous aiohttp requests library to offer users the ability to write concurrently executing code.

# import domolibrary.classes.DomoDataset as dmds

# # this is a class method
# domo_ds = await dmds.DomoDataset.get_from_id(auth=token_auth, dataset_id=os.environ['DOJO_DATASET_ID'])
# domo_ds

Once instantiated, you can call methods to interact with that object. You typically won’t have to pass auth creds again because they are saved to the object.

In the example below we are retrieving the DomoDataset_Schema which consists of subclass DomoDataset_Schema_Column using the DomoDataset_Schema.get method.

We take the approach of where possible converting dictionaries from Domo APIs into classes because it provides greater predictability when users are creating integrations between platforms (ex. Domo to Trello).

# await domo_ds.schema.get()

Typically all information about an entity is saved in the object

# domo_ds.__dict__

Option 2 functional programming

Although classes add a pretty wrapper for interacting with Domo APIs, users can opt to interact directly with APIs by way of routes.

All route functions will exclusively call one API and will always return a ResponseGetData object OR raise an Exception if appropriate.

For example we can implement similar functionality as the Option 1 example by calling the get_dataset_by_id function.

import domolibrary.routes.dataset as dataset_routes

ds_res = await dataset_routes.get_dataset_by_id( auth = token_auth, dataset_id = os.environ['DOJO_DATASET_ID'])
ds_res

ResponseGetData(status=200, response={'id': '04c1574e-c8be-4721-9846-c6ffa491144b', 'displayType': 'domo-jupyterdata', 'dataProviderType': 'domo-jupyterdata', 'type': 'Jupyter', 'name': 'domo_kbs', 'owner': {'id': '1893952720', 'name': 'Jae Wilson', 'type': 'USER', 'group': False}, 'status': 'SUCCESS', 'created': 1668379680000, 'lastTouched': 1668385822000, 'lastUpdated': 1668385822045, 'rowCount': 1185, 'columnCount': 7, 'cardInfo': {'cardCount': 0, 'cardViewCount': 0}, 'properties': {'formulas': {'formulas': {}}}, 'state': 'SUCCESS', 'validConfiguration': True, 'validAccount': True, 'streamId': 825, 'transportType': 'API', 'adc': False, 'adcExternal': False, 'cloudId': 'domo', 'cloudName': 'Domo', 'permissions': 'READ_WRITE_DELETE_SHARE_ADMIN', 'hidden': False, 'tags': '["developer_documentation","hackercore"]', 'scheduleActive': True, 'cryoStatus': 'ADRENALINE'}, is_success=True)

ResponseGetData will always include a boolean is_success, the API status, and raw API response.

Typically the route methods will not alter the response unless the API does not include a descriptive response (ex, routes.dataset.set_dataset_tags does not return a response so we artificially alter the response in the function.)

[(prop, type(getattr(ds_res , prop))) for prop in dir(ds_res) if not prop.startswith('_')]

[('auth', domolibrary.client.DomoAuth.DomoTokenAuth),
 ('is_success', bool),
 ('response', dict),
 ('set_response', method),
 ('status', int)]

ds_res.response

{'id': '04c1574e-c8be-4721-9846-c6ffa491144b',
 'displayType': 'domo-jupyterdata',
 'dataProviderType': 'domo-jupyterdata',
 'type': 'Jupyter',
 'name': 'domo_kbs',
 'owner': {'id': '1893952720',
  'name': 'Jae Wilson',
  'type': 'USER',
  'group': False},
 'status': 'SUCCESS',
 'created': 1668379680000,
 'lastTouched': 1668385822000,
 'lastUpdated': 1668385822045,
 'rowCount': 1185,
 'columnCount': 7,
 'cardInfo': {'cardCount': 0, 'cardViewCount': 0},
 'properties': {'formulas': {'formulas': {}}},
 'state': 'SUCCESS',
 'validConfiguration': True,
 'validAccount': True,
 'streamId': 825,
 'transportType': 'API',
 'adc': False,
 'adcExternal': False,
 'cloudId': 'domo',
 'cloudName': 'Domo',
 'permissions': 'READ_WRITE_DELETE_SHARE_ADMIN',
 'hidden': False,
 'tags': '["developer_documentation","hackercore"]',
 'scheduleActive': True,
 'cryoStatus': 'ADRENALINE'}

Access Paginated APIs using the Looper

A hidden advantage of using the DomoLibrary is that paginated API requests are baked into the route’s definition.

Consider query_dataset_private from the routes.dataset.

Inside this function we are using looper from client.get_data to paginate over the API response.