Installation

Install with pip:

pip install sofart

Example

Easy use:

>>> from sofart import Database

Create Database and a collection:

>>> db = Database('/tmp/so.fart')
>>> db.create_collection('test_collection')
>>> db.collection_names()
['test_collection']

Play with collection:

>>> c = db.test_collection
>>> post = {"artist": "Jambon",
...         "track": "I love my jambon"}
>>> c.save(post)
>>> for i in c.find_one():
...     print(i)
...
{'track': 'I love my jambon', '_id': 'b2d6bf60-6c11-4e26-9357-efb28056e60d', 'artist': 'Jambon'}
>>> c.remove('b2d6bf60-6c11-4e26-9357-efb28056e60d')
>>> c.find_one()

Some filter:

      >>> c.find_one({"artist": "Jambon"})
      {'track': 'I love my jambon', '_id': 'b2d6bf60-6c11-4e26-9357-efb28056e60d', 'artist': 'Jambon'}
      >>> c.find_one({"artist":"Bieber"})
>>>

Tests

You can run test under tests/test_sofart.py. And there is a populate script into tests/populate.py.

Performances

Single(in-memory) is highly faster than multi cause it’s mainly work in RAM and just down data when sync method is called. In otherwise multi down data at each request.

But you can have a pretty data control with sync method which down data in file when you call it.

Serializers

Pypy and msgpack-pure are not supported.

Misc.

You can easily write your own serializer, have a look at serializers/_msgpack.py or _json.py file.

Remember that if someone write Ruby or other language driver for sofart, maybe using json or msgpack could be a good idea.

Docs

Database(str(path), str(mode), str(serializer)) : Path is database file path
                                                : Mode is single or multi (Default: `single`)
                                                : Serializer like msgpack, json or pickle (Default: `json`)

path: Return database path

create_collection(str(name))    : Create new collection
drop_collection(str(name))      : Drop collection
rename(str(name), str(new_name)): Rename collection `name` to `new_name`
collection_names(): Return database collections list (same as `collections` attribut)
count()           : Return total database entries
sync()            : Save every changes in database file

c = db.my_collection
c = db['my_collection']

name: Return collection name

drop()           : Drop collection
count()          : Return total collection entries
save(dict(enreg)): Save entry into collection
remove(str(_id)) : Remove entry from collection
sync()           : Save every changes in database file
rename(str(name)): Rename collection to `name`
find_one(spec_or_id)              : Return first founded result
find(dict(spec_or_id), int(limit)): Iterator which return `limit` result founded (limit=0 return all)

Query

At this moment just following operands are available:

• <

• <=

• >

• >=

• all

• exists

• mod

• ne

• in

• nin

This is an example:

>>> c.save({"value": 2})
>>> c.find({"value": {"$exists": True}})
[{'_id': '47e53aea-85b4-434b-8961-40e89c877b41', 'value': 2}]
>>> c.find({"value": {"$in": [2, 3, 67]}})
[{'_id': '47e53aea-85b4-434b-8961-40e89c877b41', 'value': 2}]
>>> c.find({"value" : { "$gt": 1 }})
[{'_id': '42567296-7d78-43b7-a4e0-50447b80eca8', 'value': 2}]

And another:

>>> c.find({"value" : { "$gte": 2 }})
[{'_id': '42567296-7d78-43b7-a4e0-50447b80eca8', 'value': 2}]
>>> c.find({"value" : { "$gte": 2, "$lt" : 1 }})
[]
>>> c.find({"value": {"$mod": [2, 0]}})
[{'_id': '47e53aea-85b4-434b-8961-40e89c877b41', 'value': 2}]

More informations here.

See LICENSE.