j2labs / tobin

Schematics, Data, and Databases.

Tobin

Schematics, Data, and Databases.

Tobin is a data persistence layer for Schematics. We store data in denormalized form and perform joins in software. Good habits early make smoother sailing later.

This project is named after Amon Tobin, a musician I am regularly inspired by. If you're not familiar with his work, try Hey Mr. Tree first.

Quick Example

First, let's create a song model.

from schematics.models import Model
from schematics.types import StringType, URLType
from schematics.types.serializable import serializable


class Song(Model):
    name = StringType()
    artist = StringType()
    url = URLType()
    @serializable
    def id(self):
        return '%s/%s' % (self.artist, self.name)

Create a couple song instances.

>>> s1 = Song()
>>> s1.name = 'Hey Mr. Tree'
>>> s1.artist = 'Amon Tobin'
>>> s1.url = 'http://www.youtube.com/watch?v=1JCZCYB0df0'
>>>
>>> s2 = Song()
>>> s2.artist = 'Shipyards'
>>> s2.name = 'Headfirst Dive'
>>> s2.url = 'https://soundcloud.com/shipyards/headfirst-dive'

Tobin aims for key-value'ish behavior, but we haven't set an id on the model. Instead, we will use a function to generate the key from field values. In this case, we join the artist name and song name for a composite key.

Instantiate the query set and then save.

>>> from tobin.base import DictQuerySet >>> dqs = DictQuerySet() >>> dqs.create(Song, [s1, s2])

Looping across data looks like this:

>>> for (status, datum) in dqs.read(Song): ... print datum ... {'url': u'http://www.youtube.com/watch?v=1JCZCYB0df0', 'id': 'Amon Tobin/Hey Mr. Tree', 'name': u'Hey Mr. Tree', 'artist': u'Amon Tobin'} {'url': u'https://soundcloud.com/shipyards/headfirst-dive', 'id': 'Shipyards/Headfirst Dive', 'name': u'Headfirst Dive', 'artist': u'Shipyards'}

We could instantiate these as models too:

>>> songs = list() >>> for (status, datum) in dqs.read(Song): ... song = Song(datum) ... songs.append(song) ... >>> songs [<Song: Song object>, <Song: Song object>]