Python SDK

---

This package is a server-side SDK for applications written in Python for the Tailslide feature flag framework.

Visit the https://github.com/tailslide-io repository or see Tailslide’s case study page for more information.

Installation

---

Install the Tailslide npm package with pip install tailslide-sdk

Basic Usage

---

Instantiating and Initializing FlagManager

The FlagManagerclass is the entry point of this SDK. It is responsible for retrieving all the flag rulesets for a given app with its app_id and creating new Toggler instances to handle toggling of feature flags within that app. To instantiate a FlagManager object, a user must provide a configuration object:

import asyncio
from tailslide import FlagManager

config = {
    "nats_server": "nats://localhost:4222",
    "nats_stream": "flags_ruleset",
    "app_id": 1,
    "user_context": "375d39e6-9c3f-4f58-80bd-e5960b710295",
    "sdk_key": "myToken",
    "redis_host": "http://localhost",
    "redis_port": 6379,
}

async def main():
    manager = FlagManager(**config)
    await manager.initialize()

asyncio.run(main())

• nats_server is the NATS JetStream server address:port
• nats_stream is the NATS JetStream’s stream name that stores all the apps and their flag rulesets
• app_id is the ID number of the app the user wants to retrieve its flag ruleset from
• user_context is the UUID string that identifies the current user
• sdk_key is the SDK key for the Tailslide, it is used as a password for NATS JetStream token authentication
• redis_host is the address to the Redis database
• redis_port is the port number that the Redis database runs on

After instantiating a FlagManager, invoke the initialize method. This method connects the FlagManager instance to both NATS JetStream and Redis Timeseries, and asynchronously retrieves the latest and any new flag ruleset data.

---

Using Feature Flag with Toggler

Once the FlagManager is initialized, it can create a Toggler, with the new_toggler method, for each feature flag that the developer wants to wrap the new and old features in. A Toggler’s is_flag_active method checks whether the flag with its flag_name is active or not based on the flag ruleset. A Toggler’s is_flag_active method returns a boolean value, which can be used to evaluate whether a new feature should be used or not.

flag_config = {
    "flag_name": 'App 1 Flag 1',
}

flag_toggler = manager.new_toggler(flag_config)

if flag_toggler.is_flag_active():
    # call new feature here
else:
    # call old feature here

---

Emitting Success or Failture

To use a Toggler instance to record successful or failed operations, call its emit_success or emit_failure methods:

if successCondition:
    flag_toggler.emit_success()
else:
    flag_toggler.emit_failure()

Documentation

---

FlagManager

The FlagManager class is the entry point of the SDK. A new FlagManager object will need to be created for each app.

FlagManager Constructor

Parameters:

• An object with the following keys
  • nats_server is the NATS JetStream server address:port
  • nats_stream is the NATS JetStream’s stream name that stores all the apps and their flag rulesets
  • app_id a number representing the application the microservice belongs to
  • sdk_key a string generated via the Tower front-end for NATS JetStream authentication
  • user_context a string representing the user’s UUID
  • redis_host a string that represents the url of the Redis server
  • redis_port a number that represents the port number of the Redis server

---

Instance Methods

flagmanager.initialize()

Asynchronously initialize flagmanager connections to NATS JetStream and Redis database

Parameters:

• None

Return Value:

• None

---

flagmanager.set_user_context(new_user_context)

Set the current user's context for the flagmanager

Parameters:

• new_user_context: A UUID string that represents the current active user

Return Value:

• None

---

flagmanager.get_user_context()

Returns the current user context

Parameters:

• None

Return Value:

• The UUID string that represents the current active user

---

flagmanager.new_toggler(options)

Creates a new toggler to check for a feature flag's status from the current app's flag ruleset by the flag's name.

Parameters:

• options: An object with key of flag_name and a string value representing the name of the feature flag for the new toggler to check whether the new feature is enabled

Return Value:

• A Toggler object

---

flagmanager.disconnect()

Asynchronously disconnects the FlagManager instance from NATS JetStream and Redis database

Parameters:

• None

Return Value:

• None

---

Toggler

The Toggler class provides methods that determine whether or not new feature code is run and handles success/failure emissions. Each toggler handles one feature flag, and is created by flagmanager.new_toggler().

---

Instance Methods

toggler.is_flag_active()

Checks for flag status, whitelisted users, and rollout percentage in that order to determine whether the new feature is enabled.

• If the flag's active status is false, the function returns False
• If current user's UUID is in the whitelist of users, the function returns True
• If current user's UUID hashes to a value within user rollout percentage, the function returns True
• If current user's UUID hashes to a value outside user rollout percentage, the function returns False

Parameters:

• None

Return Value

• True or False depending on whether the feature flag is active

---

toggler.emit_success()

Records a successful operation to the Redis Timeseries database, with key flagId:success and value of current timestamp

Parameters:

• None

Return Value

• None

---

toggler.emit_failure()

Records a failure operation to the Redis Timeseries database, with key flagId:success and value of current timestamp

Parameters:

• None

Return Value

• None