nqdb 0.1.0

Interface to Google Datastore, mimicking aspects of ndb, but allowing for different backends and caching

NQDB

Not quite the DB Google used to have, not quite the DB Google has now. Of course, better than both.

This is a mock of NDB, almost. It enables interfacing to Google Datastore, or to other DBs, by changing the client.

Usage

Models

from nqdb import *

client = DatastoreClient()

class Person(Model):
    _client = client
    first_name = StringProperty('first_name')
    last_name = StringProperty('last_name')

p = Person()
p.first_name = "Luca"
p.last_name = "de Alfaro"
p.put()
id = p.key_id
print("id:", id)
q = Person(id=id)
q.get()
print("First name:", q.first_name)
print("Last name:", q.last_name)

Keys

p = Person()
p.first_name = "Joe"
p.last_name = "Falchetto"
put(p)
id = p.key_id
k = Key(Person, id)
q = k.get()

Batch operations

from nqdb import put_multi

p = Person()
p.first_name = "Joe"
p.last_name = "Falchetto"
q = Person()
q.first_name = "Luca"
q.last_name = "de Alfaro"
put_multi([p, q])

The client used by put_multi is derived from the client used for p and q. You cannot mix in the same batch operations models with different clients.

Queries

q = Person.query()
q.filter(Person.first_name == 'Luca')
q.filter(Person.gender == 'Male')
for p in q.fetch():
    print(p.last_name)

You can add an ordering to a query by:

q.order(+Person.first_name)

You can use both +Person.first_name and -Person.first_name, but one of +, - should be present.

Caching

client = DatastoreClient(cache=MyCache())

Look at cache.Cache to see the (very few) methods you need to implement to use a new cache.

And as usual, look at the test files, because those do not lie.