+---------------------------+ | | | | | .-. _______ | | {}``; |==|_______D | | / (' /|\ | Django Trackman helps with | ( / | / | \ | tracking and logging. | \(_)_]] / | \ | | | | | +---------------------------+
- Customizable models: You'll have base models and an admin class mixins that you can extend to define your own tracking models.
- Separate database: You can decide to use a separate database for your tracking data, as it could grow fast. A database router will help you implement the Django multiple database feature.
- Loose coupling: The tracking table is not related to you application table with foreign keys. This ensures that your tracking data is kept independent of your application data.
- Admin action tracking: You can enable tracking of Django admin actions.
- API Endpoint: You can define django-rest-framework views to create API endpoints. This can be used to track actions made through API calls.
pip install django-trackman
Add 'trackman' to your INSTALLED_APPS in Django's settings:
INSTALLED_APPS = (
...
'trackman',
...
)
Then enable trackman by setting the variable TRACKMAN_ENABLED to True.
Django Trackman can be used to create custom tracking models. Here's an example of a model class that extends Django Trackman's TrackingActionModel`:
from trackman.models import TrackingActionModel
from django.db import models
class TrackingAction(TrackingActionModel):
team = models.CharField("Team", max_length=256, blank=True)
def __str__(self):
return f"Actor: {self.actor}, Action: {self.action}, Time: {self.created}"
class Meta:
verbose_name = "Action Log"
verbose_name_plural = "Action Logs"
How Django Trackman would know that this model should be used by default for tracking your actions ? You'll need to instruct it by defining a default tracking model in your settings.
TRACKMAN_MODELS = {
"default": "tracking.TrackingAction",
"data-quality": "tracking.DataQualityTracking",
}
Here, you'll notice that we have also defined a additional tracking model for data quality tracking.
If your models are not action oriented, you can always extend the TrackingBaseModel that's an empty abstract model that only serves as a way to let Trackman know that your model should be consider as a tracking model and thus should be taken into account when during database routing.
For managing your tracking models in Django admin, Django Trackman provides the mix-in class TrackingModelAdminMixin.
from django.contrib import admin
from trackman.admin import TrackingModelAdminMixin
from .models import TrackingAction
class TrackingActionAdmin(TrackingModelAdminMixin, admin.ModelAdmin):
list_display = [
"id",
"actor",
"team",
"action",
"object",
"target",
"description",
"created",
]
search_fields = ["actor", "team"] + TrackingModelAdminMixin.action_log_search_fields
list_filter = ["team", "action"]
if settings.TRACKMAN_ENABLED:
admin.site.register(TrackingAction, TrackingActionAdmin)
When you want to isolate tracking data from your main application data for performance, maintenance or data integrity reasons, you could route tracking database operations to a separate database.
Django's multiple database feature allows you to use more than one database in your project. It provides the flexibility to specify which models use which database - that's defined by a router. Django Trackman uses that feature to isolate you tracking data from your application data.
You will first need to instruct in your project's settings, the database alias that should be used.
TRACKMAN_DATABASE_ALIAS = "tracking"
Then define the database accesses.
DATABASES = {
"default": {
'ENGINE': 'django.db.backends.<app-database>',
'NAME': <app-database-name>,
'USER': <-app-database-user>,
'PASSWORD': <-app-database-password>,
'HOST': <-app-database-host>,
'PORT': <-app-database-port>,
},
}
if TRACKMAN_ENABLED:
DATABASE_ROUTERS = ["trackman.db_routers.TrackmanDatabaseRouter"]
DATABASES[TRACKMAN_DATABASE_ALIAS] = {
'ENGINE': 'django.db.backends.<tracking-db>',
'NAME': <tracking-database-name>,
'USER': <tracking-database-user>,
'PASSWORD': <tracking-database-password>,
'HOST': <tracking-database-host>,
'PORT': <tracking-database-port>,
}
If you are using a tool like dj-database-url with environnement variables:
from django.urls import dj_database_url.parse
DATABASES = {"default": dj_database_url.parse(env("APP_DATABASE_URL"))}
if TRACKMAN_ENABLED:
DATABASE_ROUTERS = ["trackman.db_routers.TrackingDatabaseRouter"]
db_url = env("TRACKING_DATABASE_URL")
DATABASES[TRACKMAN_DATABASE_ALIAS] = dj_database_url.parse(db_url)
Django Trackman, allows you to track Django admin actions - the actions that the admin users are performing.
Setting Up Django Trackman The main file is trackman/signals.py, and it requires importing in some AppConfig's ready method.
Let's suppose that you have a backoffice app that you can use to setup signals.
from django.apps import AppConfig
import sys
class BackofficeConfig(AppConfig):
name = "backoffice"
verbose_name = "Backoffice"
def ready(self):
if "migrate" not in sys.argv:
import trackman.signals # noqa
As a consequence, all admin actions will be copied to you tracking table.
Django Trackman provides a mixin you can use with Django Rest Framework's ViewSet to create an API end-point for your application to track actions. This could be useful for tracking front-end actions.
Here's how you can use TrackingViewSetMixin in a Django REST ViewSet for action tracking:
from rest_framework import viewsets
from trackman.api import TrackingViewSetMixin
class ActionTrackingViewSet(TrackingViewSetMixin, viewsets.ViewSet):
model_alias = "default"
def clean_action_details(self, action_details):
# Do some clean-up here...
cleaned_data = action_details.copy()
return cleaned_data
The model_alias points out which Django Trackman model alias to be used for saving the tracking data.
You'll need to add this new ViewSet ActionTrackingViewSet to your url configuration to have it active.