Persistence: dbm, shelve, pickle

dbm and shelve Interfaces

dbm is an access-by-key file system. Using a persistent object shelve is identical, except that “anydbm” is replaced by “shelve,” and “value” can be almost any kind of Python object (but keys are still strings). shelve serializes objects with pickle, and stores them by key on a dbm file.

import anydbm

Gets dbm, gbmd, . . . whatever is installed.

file =‘filename’)

Creates a new or open an existing dbm file.

file [‘key1’] = value

Store: create or change the entry for ‘key1’.

value = file [‘key2’]

Fetch: load the value for entry ‘key2’.

count = len(file)

Size: return the number entries stored.

index = file.keys( )

Index: fetch the stored keys list (can use in a for).

found = file.has_key(‘key3’)

Query: see if there’s an entry for ‘key3’.

del file [‘key4’]

Delete: remove the entry for ‘key4’.

file.close( )

Manual close (often is optional).


dbm files and shelves work like dictionaries that must be opened before use; all mapping operations, and some dictionary methods work.

For dbm files, can also pass “mode” (‘r’, ‘w’, etc.) and “protection” (access-mode) parameters to open if desired (some dbm flavors require extra arguments).

Pickle Interface

Converts (nearly) arbitrary Python objects to/from serialized byte streams. Also see: cPickle (coded in C for speed).

P = pickle.Pickler(fileobject)

Makes a new pickler (for saving to an output file object).


Writes an object onto the pickler’s file/stream.

pickle.dump(object, fileobject) ...

Get Python Pocket Reference now with the O’Reilly learning platform.

O’Reilly members experience live online training, plus books, videos, and digital content from nearly 200 publishers.