A MongoDB collection does not have a strict schema, the documents in one collection can have different fields. But in MongoEngine the schema is provided on an application level and there is a validation. So if you’ll remove a field from the Document and it’s still exists in the database, there will be a mongoengine.errors.FieldDoesNotExist exception. If there is no need in the schema validation, a dynamic schema can be used.

So when there is a strict schema, it’s good to have a tool to make an updates for the documents in the collection when something changes.

For SQL databases some apps exist to solve it, like Alembic or Django Migrations. I created a similar app Flask MongoEngine Migrations based on the idea from Django Data Migrations

An example of use

The app should be initialized with FlaskMigrations(app), then an empty migration can be created with flask migration_create name_of_migration_with_underscores

It will create a migration file 0001_name_of_migration_with_underscores.py in the migrations folder. You will need to modify it:

def up(db):
    db['client'].update_many({}, {'$unset': {'some_field': ''}})

def down(db):
    db['client'].update_many({}, {'$set': {'some_field': ''}})

After that it can be used with flask migration_up and flask migration_down 0001 commands.