Skip to main content

development version of persistentdicts

Project description


Build Status Coverage Status License Supported Python versions

dict-like interfaces for various databases

Key differences with competing projects

If you search on pypi, you will find a few competing projects: pdicts, durabledicts, etc. The key differences between persistentdicts and those are:

  • Implementations of persistentdicts do not keep a local cache dictionary: changes are immediately written to database. Iterators are also proceeding directly on the database. It allows to interact with datasets that would not fit in RAM otherwise.

  • The test suite is significantly bigger than in all the other implementation I have seen. When relevant, tests have been backported from the CPython test suite of dict.

  • Serialization is done in json rather than using pickle.


You can’t modify a value of the dictionary in place. For example:

>>> import persistentdicts

>>> d = persistentdicts.sqlitedict.SqliteDict()
>>> d["a"] = []
>>> d["a"].append(1)

>>> d["a"] # with a normal dict, you would get [1]

That is because d["a"] returns a copy of the database entry for the key "a", and not a reference to a python object. Modifying this copy (with append) does not affect the database itself.

To circumvent this, you should do:

>>> import persistentdicts

>>> d = persistentdicts.sqlitedict.SqliteDict()
>>> d["a"] = []
>>> d["a"] = d["a"] + [1]

>>> d["a"]

Similarly, setdefault will not work as expected since it does not return a reference to the stored value, but a copy of this value.

>>> import persistentdicts

>>> d = persistentdicts.sqlitedict.SqliteDict()
>>> d.setdefault("a", []).append(1)

>>> d["a"]

Databases supported so far


  • kyotocabinet

  • sqlite

  • cassandra


  • leveldb

  • redis

  • memcachedb

  • lightcloud

You can request new formats on the bug tracker.

Short documentation


persistentdicts.sqlitedict.SqliteDict(path=":memory:", table="dict", isolation_level="DEFERRED", *args, **kwargs)
  • path is the path to the file where you wish to store the data

  • dict is the table to use in this file

  • isolation_level is the isolation level used for all transactions. See the sqlite documentation for more details.

  • the remaining arguments *args, **kwargs are used to fill the dictionary (like a normal dict)


persistentdicts.kyotocabinetdict.KyotoCabinetDict(path, *args, **kwargs)
  • path is the path to the file where you wish to store the data. The file extension matters and will determine which format is going to be used internally (must be one of .kch, .kct, .kcd, .kcf or .kcx). See the kyotocabinet documentation for more details.

  • the remaining arguments *args, **kwargs are used to fill the dictionary (like a normal dict)


persistentdicts.cassandradict.CassandraDict(contact_points=("",), port=9042, keyspace="dict", table="dict", *args, **kwargs)
  • contact_points is an initial list of ip addresses which are part of the Cassandra cluster. The Cassandra driver will automatically discover the rest of the cluster.

  • port is the port on which Cassandra runs.

  • keyspace is the keyspace used to store the data. This keyspace will be deleted if the method .delete() is called on the CassandraDict

  • table is the name of the table used to store the data.

  • the remaining arguments *args, **kwargs are used to fill the dictionary (like a normal dict)

Project details

Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

persistentdicts-dev-2.0.4.dev2.tar.gz (7.3 kB view hashes)

Uploaded source

Supported by

AWS AWS Cloud computing Datadog Datadog Monitoring Facebook / Instagram Facebook / Instagram PSF Sponsor Fastly Fastly CDN Google Google Object Storage and Download Analytics Huawei Huawei PSF Sponsor Microsoft Microsoft PSF Sponsor NVIDIA NVIDIA PSF Sponsor Pingdom Pingdom Monitoring Salesforce Salesforce PSF Sponsor Sentry Sentry Error logging StatusPage StatusPage Status page