pymdict 0.0.4

PyMDict

Advanced Python Mongo Dict. A Python dictionary based on a MongoDB. It allows to treat a collection as a dictionary in Python, with extensive capabilities, like allowing basic queries, bulk operations and versioning (forks).

INSTALLATION

Currently it is only supported Python3.4 onwards. It can be installed through pip:

$ pip3 install pymdict

USAGE

In order to run, a MongoDB server is required.

A Mongo-based dictionary can be instantiated as follows:

>>> from pymdict.mongo_dict import MongoDict
>>>
>>> m = MongoDict("custom_id", mongo_host="localhost", mongo_port=27017)

Once m is instantiated, it can be used as a normal dictionary.

>>> m["key"] = "value"
>>> m[44] = "value2"
>>> m["key2"] = "value3"
>>> m["number"] = 44

>>> print(list(m.keys()))
["key", 44, "key2, "number"]

>>> for key, value in m.items():
...     print("{}: {}".format(key, value))
key: value
44: value2
key2: value3
number: 44

In addition, there are advanced functionalities like queries:

>>> for key, value, _ in m('key % ey'):
...     print("{}: {}".format(key, value))
key: value
key2: value3

>>> for key, value, _ in m('key % ey or value = 44'):
...     print("{}: {}".format(key, value))
key: value
key2: value3
number: 44

Queries also support to query with sub-dict elements:

>>> m["first"] = {"example": 44}
>>> m["second"] = {"example": 45}
>>> m["third"] = {"example": 46}

>>> for key, value, _ in m('value.example > 44 and value.example < 46'):
...     print("{}: {}".format(key, value))
second: 45

(TODO: Check the wiki page for more information about the query syntax)

Note that all the stores and removals are stored within a MongoDB. This means for each addition,edit and removal there is at least one connection to the MongoDB backend. In order to optimize it, a bulk operation can be used to wrap such amount of operations in a single connection:

>>> with m.bulk(buffer_size=100) as m:
...     for x in range(2000):
...         m["key{}".format(x)] = {"example": x}

Also, a mongo dict can be forked without the need to copy its content. This is specially useful if the target dict is extremely big and a copy is wanted. Note that a fork is an immediate process, and it allows to override or remove elements without modifying an original dictionary. It is achieved by applying a versioning technique with the dictionaries and it is still in an experimental state.

(TODO: More information about forking and versioning in the wiki page)

>>> m['foo'] = "bar"
>>> fork = m.fork()
>>> print(fork['foo'])
bar
>>> fork['foo'] = "foo"
>>> print(fork['foo'], m['foo'])
foo bar