Python Key Value Data Abstraction Layer

PyKvDal lets you easily do object-to-key-value-database-abstraction. It supports memcached out of the box but other backends can be used without much effort.

Getting started

PyKvDal connects to memcached by default:

pykvdal.dal_connect(['127.0.0.1:11211'])

Or repurpose dal_set(), dal_get() and dal_delete() to use other backends. Any backend that can support key-value storing, retreiving and deletion is supported: MySQL, PostgreSQL, memcached, mongodb, redis, couchdb ... and many more.

All models must subclass BaseModel, that's all:

class Person(pykvdal.Model):
  pass

Now you can do stuff like:

p = Person()
p.name = 'foo'
p.age = 13
p.save()  # stores it in the database, assigns p.id

Retreiving the saved instance:

p_identical = Person.load(p.id)
print p_identical.name  # >>> 'foo'

Simple querying is supported, first define an Index:

class Person(pykvdal.Model):
  get_by_age = pykvdal.Index('age')

Now to query the above created Person:

res = Person.get_by_age(13)
for p in res:
  print p.name

Limited field validation is implemented using PyContracts:

class Person(pykvdal.Model):
  age = pykvdal.Field('age', default=0, contract='int')
  name = pykvdal.Field('name', contract='str')

Why use PyKvDal

Ever started a project when you didn't know what datastore to use? SQL or NOSQL? PostgreSQL or Mongodb? The requirements were unclear, specially the non-functional requirements so you chose MySQL since its reasonable fast and supports transactions? PyKvDal makes you not worry about these things at the start of the project. PyKvDal exists of only one module and it is designed to be replaced by a proper ORM or document store mapper when the type of database has been decided. PyKvDal just lets you begin without worries while you know you can easily transfer to another solution or just connect PyKvDal to the correct database.

Manual

TODO: Write proper manual. (For now check the source of pykvdal.py ;-)

Feature list

• Popuplar backends implemented
• Proper subclasses for validated fields e.b. StringField, IntField, etc.