dbm — Unix Key-Value Databases¶
|Purpose:||dbm provides a generic dictionary-like interface to DBM-style, string-keyed databases|
dbm is a front-end for DBM-style databases that use simple
string values as keys to access records containing strings. It uses
whichdb() to identify databases, then opens them with the
appropriate module. It is used as a back-end for
stores objects in a DBM database using
Python comes with several modules for accessing DBM-style databases. The default implementation selected depends on the libraries available on the current system and the options used when Python was compiled. Separate interfaces to the specific implementations allow Python programs to exchange data with programs in other languages that do not automatically switch between available formats, or to write portable data files that will work on multiple platforms.
dbm.gnu is an interface to the version of the
library from the GNU project. It works the same as the other DBM
implementations described here, with a few changes to the
Besides the standard
'f'to open the database in fast mode. In fast mode, writes to the database are not synchronized.
's'to open the database in synchronized mode. Changes to the database are written to the file as they are made, rather than being delayed until the database is closed or synced explicitly.
'u'to open the database unlocked.
dbm.ndbm module provides an interface to the Unix ndbm
implementations of the dbm format, depending on how the module was
configured during compilation. The module attribute
identifies the name of the library
configure was able to find when
the extension module was compiled.
dbm.dumb module is a portable fallback implementation of
the DBM API when no other implementations are available. No external
dependencies are required to use
dbm.dumb, but it is slower
than most other implementations.
Creating a New Database¶
The storage format for new databases is selected by looking for usable versions of each of the sub-modules in order.
open() function takes
flags to control how the database
file is managed. To create a new database when necessary, use
'n' always creates a new database, overwriting an
import dbm with dbm.open('/tmp/example.db', 'n') as db: db['key'] = 'value' db['today'] = 'Sunday' db['author'] = 'Doug'
In this example, the file is always re-initialized.
$ python3 dbm_new.py
whichdb() reports the type of database that was created.
import dbm print(dbm.whichdb('/tmp/example.db'))
Output from the example program will vary, depending on which modules are installed on the system.
$ python3 dbm_whichdb.py dbm.ndbm
Opening an Existing Database¶
To open an existing database, use
flags of either
'w' (for read-write). Existing databases are
automatically given to
whichdb() to identify, so it as long as a
file can be identified, the appropriate module is used to open it.
import dbm with dbm.open('/tmp/example.db', 'r') as db: print('keys():', db.keys()) for k in db.keys(): print('iterating:', k, db[k]) print('db["author"] =', db['author'])
db is a dictionary-like object. New keys are always
converted to byte strings when added to the database, and returned as
$ python3 dbm_existing.py keys(): [b'key', b'today', b'author'] iterating: b'key' b'value' iterating: b'today' b'Sunday' iterating: b'author' b'Doug' db["author"] = b'Doug'
The keys of the database need to be strings.
import dbm with dbm.open('/tmp/example.db', 'w') as db: try: db = 'one' except TypeError as err: print(err)
Passing another type results in a
$ python3 dbm_intkeys.py dbm mappings have bytes or string keys only
Values must be strings or
import dbm with dbm.open('/tmp/example.db', 'w') as db: try: db['one'] = 1 except TypeError as err: print(err)
TypeError is raised if a value is not a string.
$ python3 dbm_intvalue.py dbm mappings have byte or string elements only
- Standard library documentation for dbm
- Python 2 to 3 porting notes for anydbm
- Python 2 to 3 porting notes for whichdb
shelve– Examples for the
shelvemodule, which uses
dbmto store data.