NoSQL databases like MongoDB are a hot tool these days. The ease of usage and the speed of development make them the right choice of tool for most of application use cases. The fact that you can dump any JSON (read BSON) data to databases is the real advantage over tight physical schemas. But the lack of data integrity and validations could mean that all the validations and datatype checks have to re-implemented at the application layer.
Mongorm is a lightweight ORM available to perform basic data type checks and constraint validations. It can be imported to create a base class used to define models that represent a collection in the database. ORM also provides a bunch of default datatypes that can be used alongside any custom datatypes that you may want to provde.
In a regular use case you would create a new collection and store stuff as:
collection = pymongo.MongoClient().sample_database.sample_coll
A regular update operation would look like
collection.insert({"data": "hello", "data_list": [1]})
collection.update({"data": "hello"}, {"$set": {"data_set": 2}})
If you notice, the data type of data_list has changed from a list to a string withour any errors. Ideally you would like an error to be raised but pyMongo would not do that. Mongorm solves this for you.
A quick re-implementation of the above requirement will looks like this.
from mongorm.base import ModelBase
from mongorm.datatypes import Unichar, List
class SampleModel(ModelBase):
__tablename__ = "sample_coll"
data = Unichar(nullable=False)
data_list = List(default=[], nullable=False)
@classmethod
def using(cls):
return pymongo.MongoClient().sample_database
sm = SampleModel.insert({"data_list": [1]})
Raises Error: mongorm.base.ORMException: ['data is a required field']
OR
SampleModel.update({"data": "hello"}, {"$set": {"data_list": 2}})
Raises Error: mongorm.base.ORMException: ["Field: datalist, Error: Expected <type 'list'>. Found 2"]
OR
SampleModel.insert({"data": 1, "data_list": 2}})
mongorm.base.ORMException: ['Field: data_list, Error: Expected a List but found 12', 'Field: data, Error: A field in this form requires a text string but you entered this instead: 1']
How wonderful !!