After setting up history for a model using django-simple-history
, I wanted to run populate_history
to populate the history table based on the existing contents of the table. However, other users have already made a number of changes, causing the history table to be partially populated. Running populate_history --auto
simply results in message Existing history found, skipping model
.
I wish to retain the existing history, but populate history for all records not currently stored in the history. Is there a way to do this?
I ended up writing a modified script based on populate_history
. It identified all objects with no historical records, and added them in to the history table. A simple version (no batching) is below.
from django.apps import apps
from simple_history.utils import get_history_manager_for_model, get_history_model_for_model
def populate_model_history(model):
history = get_history_model_for_model(model)
history_manager = get_history_manager_for_model(model)
# Insert historical records for objects without existing history
# NOTE: A better approach would be to do this in batches, as in populate_history.py
for instance in model.objects.exclude(pk__in=history.objects.values_list(model._meta.pk.name)):
history_manager.bulk_history_create([instance], batch_size=1)
model = apps.get_model('app', 'my_model')
populate_model_history(model)