Version 5.2.3 (2022-01-30)

  • This version officially supports the new Python 3.10 and PostgreSQL 14.

  • Some improvements and fixes in the inserttable() method of the pg module: - Sync with PQendcopy() when there was an error (#60) - Allow specifying a schema in the table name (#61) - Improved check for internal result (#62) - Catch buffer overflows when building the copy command - Data can now be passed as an iterable, not just list or tuple (#66)

  • Some more fixes in the pg module: - Fix upsert with limited number of columns (#58). - Fix argument handling of is/set_non_blocking(). - Add missing get/set_typecasts in list of exports.

  • Fixed a reference counting issue when casting JSON columns (#57).

Version 5.2.2 (2020-12-09)

  • Added a missing adapter method for UUIDs in the classic pg module.

  • Performance optimizations for fetchmany() in the pgdb module (#51).

  • Fixed a reference counting issue in the cast_array/record methods (#52).

  • Ignore incompatible libpq.dll in Windows PATH for Python >= 3.8 (#53).

Version 5.2.1 (2020-09-25)

  • This version officially supports the new Python 3.9 and PostgreSQL 13.

  • The copy_to() and copy_from() methods in the pgdb module now also work with table names containing schema qualifiers (#47).

Version 5.2 (2020-06-21)

  • We now require Python version 2.7 or 3.5 and newer.

  • All Python code is now tested with flake8 and made PEP8 compliant.

  • Changes to the classic PyGreSQL module (pg):
    • New module level function get_pqlib_version() that gets the version of the pqlib used by PyGreSQL (needs PostgreSQL >= 9.1 on the client).

    • New query method memsize() that gets the memory size allocated by the query (needs PostgreSQL >= 12 on the client).

    • New query method fieldinfo() that gets name and type information for one or all field(s) of the query. Contributed by Justin Pryzby (#39).

    • Experimental support for asynchronous command processing. Additional connection parameter nowait, and connection methods send_query(), poll(), set_non_blocking(), is_non_blocking(). Generously contributed by Patrick TJ McPhee (#19).

    • The types parameter of format_query can now be passed as a string that will be split on whitespace when values are passed as a sequence, and the types can now also be specified using actual Python types instead of type names. Suggested by Justin Pryzby (#38).

    • The inserttable() method now accepts an optional column list that will be passed on to the COPY command. Contributed by Justin Pryzby (#24).

    • The DBTypes class now also includes the typlen attribute with information about the size of the type (contributed by Justin Pryzby).

    • Large objects on the server are not closed any more when they are deallocated as Python objects, since this could cause several problems. Bug report and analysis by Justin Pryzby (#30).

  • Changes to the DB-API 2 module (pgdb):
    • When using Python 2, errors are now derived from StandardError instead of Exception, as required by the DB-API 2 compliance test.

    • Connection arguments containing single quotes caused problems (reported and fixed by Tyler Ramer and Jamie McAtamney).

Version 5.1.2 (2020-04-19)

  • Improved handling of build_ext options for disabling certain features.

  • Avoid compiler warnings with proper casts. This should solve problems when building PyGreSQL on MaCOS.

  • Export only the public API on wildcard imports

Version 5.1.1 (2020-03-05)

  • This version officially supports the new Python 3.8 and PostgreSQL 12.

  • This version changes internal queries so that they cannot be exploited using a PostgreSQL security vulnerability described as CVE-2018-1058.

  • Removed NO_PQSOCKET switch which is not needed any longer.

  • Fixed documentation for other compilation options which had been renamed.

  • Started using GitHub as development platform.

Version 5.1 (2019-05-17)

  • Changes to the classic PyGreSQL module (pg):
    • Support for prepared statements (following a suggestion and first implementation by Justin Pryzby on the mailing list).

    • DB wrapper objects based on existing connections can now be closed and reopened properly (but the underlying connection will not be affected).

    • The query object can now be used as an iterator similar to query.getresult() and will then yield the rows as tuples. Thanks to Justin Pryzby for the proposal and most of the implementation.

    • Deprecated query.ntuples() in the classic API, since len(query) can now be used and returns the same number.

    • The i-th row of the result can now be accessed as query[i].

    • New method query.scalarresult() that gets only the first field of each row as a list of scalar values.

    • New methods, query.onenamed(), query.onedict() and query.onescalar() that fetch only one row from the result or None if there are no more rows, similar to the cursor.fetchone() method in DB-API 2.

    • New methods query.single(), query.singlenamed(), query.singledict() and query.singlescalar() that fetch only one row from the result, and raise an error if the result does not have exactly one row.

    • New methods query.dictiter(), query.namediter() and query.scalariter() returning the same values as query.dictresult(), query.namedresult() and query.scalarresult(), but as iterables instead of lists. This avoids creating a Python list of all results and can be slightly more efficient.

    • Removed pg.get/set_namedresult. You can configure the named tuples factory with the pg.set_row_factory_size() function and change the implementation with pg.set_query_helpers(), but this is not recommended and this function is not part of the official API.

    • Added new connection attributes socket, backend_pid, ssl_in_use and ssl_attributes (the latter need PostgreSQL >= 9.5 on the client).

  • Changes to the DB-API 2 module (pgdb):
    • Connections now have an autocommit attribute which is set to False by default but can be set to True to switch to autocommit mode where no transactions are started and calling commit() is not required. Note that this is not part of the DB-API 2 standard.

Version 5.0.7 (2019-05-17)

  • This version officially supports the new PostgreSQL 11.

  • Fixed a bug in parsing array subscript ranges (reported by Justin Pryzby).

  • Fixed an issue when deleting a DB wrapper object with the underlying connection already closed (bug report by Jacob Champion).

Version 5.0.6 (2018-07-29)

  • This version officially supports the new Python 3.7.

  • Correct trove classifier for the PostgreSQL License.

Version 5.0.5 (2018-04-25)

  • This version officially supports the new PostgreSQL 10.

  • The memory for the string with the number of rows affected by a classic pg module query() was already freed (bug report and fix by Peifeng Qiu).

Version 5.0.4 (2017-07-23)

  • This version officially supports the new Python 3.6 and PostgreSQL 9.6.

  • query_formatted() can now be used without parameters.

  • The automatic renaming of columns that are invalid as field names of named tuples now works more accurately in Python 2.6 and 3.0.

  • Fixed error checks for unlink() and export() methods of large objects (bug report by Justin Pryzby).

  • Fixed a compilation issue under OS X (bug report by Josh Johnston).

Version 5.0.3 (2016-12-10)

  • It is now possible to use a custom array cast function by changing the type caster for the ‘anyarray’ type. For instance, by calling set_typecast(‘anyarray’, lambda v, c: v) you can have arrays returned as strings instead of lists. Note that in the pg module, you can also call set_array(False) in order to return arrays as strings.

  • The namedtuple classes used for the rows of query results are now cached and reused internally, since creating namedtuples classes in Python is a somewhat expensive operation. By default the cache has a size of 1024 entries, but this can be changed with the set_row_factory_size() function. In certain cases this change can notably improve the performance.

  • The namedresult() method in the classic API now also tries to rename columns that would result in invalid field names.

Version 5.0.2 (2016-09-13)

  • Fixed an infinite recursion problem in the DB wrapper class of the classic module that could occur when the underlying connection could not be properly opened (bug report by Justin Pryzby).

Version 5.0.1 (2016-08-18)

  • The update() and delete() methods of the DB wrapper now use the OID instead of the primary key if both are provided. This restores backward compatibility with PyGreSQL 4.x and allows updating the primary key itself if an OID exists.

  • The connect() function of the DB API 2.0 module now accepts additional keyword parameters such as “application_name” which will be passed on to PostgreSQL.

  • PyGreSQL now adapts some queries to be able to access older PostgreSQL 8.x databases (as suggested on the mailing list by Andres Mejia). However, these old versions of PostgreSQL are not officially supported and tested any more.

  • Fixed an issue with Postgres types that have an OID >= 0x80000000 (reported on the mailing list by Justin Pryzby).

  • Allow extra values that are not used in the command in the parameter dict passed to the query_formatted() method (as suggested by Justin Pryzby).

  • Improved handling of empty arrays in the classic module.

  • Unused classic connections were not properly garbage collected which could cause memory leaks (reported by Justin Pryzby).

  • Made C extension compatible with MSVC 9 again (this was needed to compile for Python 2 on Windows).

Version 5.0 (2016-03-20)

  • This version now runs on both Python 2 and Python 3.

  • The supported versions are Python 2.6 to 2.7, and 3.3 to 3.5.

  • PostgreSQL is supported in all versions from 9.0 to 9.5.

  • Changes in the classic PyGreSQL module (pg):
    • The classic interface got two new methods get_as_list() and get_as_dict() returning a database table as a Python list or dict. The amount of data returned can be controlled with various parameters.

    • A method upsert() has been added to the DB wrapper class that utilizes the “upsert” feature that is new in PostgreSQL 9.5. The new method nicely complements the existing get/insert/update/delete() methods.

    • When using insert/update/upsert(), you can now pass PostgreSQL arrays as lists and PostgreSQL records as tuples in the classic module.

    • Conversely, when the query method returns a PostgreSQL array, it is passed to Python as a list. PostgreSQL records are converted to named tuples as well, but only if you use one of the get/insert/update/delete() methods. PyGreSQL uses a new fast built-in parser to achieve this. The automatic conversion of arrays to lists can be disabled with set_array(False).

    • The pkey() method of the classic interface now returns tuples instead of frozensets, with the same order of columns as the primary key index.

    • Like the DB-API 2 module, the classic module now also returns bool values from the database as Python bool objects instead of strings. You can still restore the old behavior by calling set_bool(False).

    • Like the DB-API 2 module, the classic module now also returns bytea data fetched from the database as byte strings, so you don’t need to call unescape_bytea() any more. This has been made configurable though, and you can restore the old behavior by calling set_bytea_escaped(True).

    • A method set_jsondecode() has been added for changing or removing the function that automatically decodes JSON data coming from the database. By default, decoding JSON is now enabled and uses the decoder function in the standard library with its default parameters.

    • The table name that is affixed to the name of the OID column returned by the get() method of the classic interface will not automatically be fully qualified any more. This reduces overhead from the interface, but it means you must always write the table name in the same way when you are using tables with OIDs and call methods that make use of these. Also, OIDs are now only used when access via primary key is not possible. Note that OIDs are considered deprecated anyway, and they are not created by default any more in PostgreSQL 8.1 and later.

    • The internal caching and automatic quoting of class names in the classic interface has been simplified and improved, it should now perform better and use less memory. Also, overhead for quoting values in the DB wrapper methods has been reduced and security has been improved by passing the values to libpq separately as parameters instead of inline.

    • It is now possible to use the registered type names instead of the more coarse-grained type names that are used by default in PyGreSQL, without breaking any of the mechanisms for quoting and typecasting, which rely on the type information. This is achieved while maintaining simplicity and backward compatibility by augmenting the type name string objects with all the necessary information under the cover. To switch registered type names on or off (this is the default), call the DB wrapper method use_regtypes().

    • A new method query_formatted() has been added to the DB wrapper class that allows using the format specifications from Python. A flag “inline” can be set to specify whether parameters should be sent to the database separately or formatted into the SQL.

    • A new type helper Bytea() has been added.

  • Changes in the DB-API 2 module (pgdb):
    • The DB-API 2 module now always returns result rows as named tuples instead of simply lists as before. The documentation explains how you can restore the old behavior or use custom row objects instead.

    • Various classes used by the classic and DB-API 2 modules have been renamed to become simpler, more intuitive and in line with the names used in the DB-API 2 documentation. Since the API provides objects of these types only through constructor functions, this should not cause any incompatibilities.

    • The DB-API 2 module now supports the callproc() cursor method. Note that output parameters are currently not replaced in the return value.

    • The DB-API 2 module now supports copy operations between data streams on the client and database tables via the COPY command of PostgreSQL. The cursor method copy_from() can be used to copy data from the database to the client, and the cursor method copy_to() can be used to copy data from the client to the database.

    • The 7-tuples returned by the description attribute of a pgdb cursor are now named tuples, i.e. their elements can be also accessed by name. The column names and types can now also be requested through the colnames and coltypes attributes, which are not part of DB-API 2 though. The type_code provided by the description attribute is still equal to the PostgreSQL internal type name, but now carries some more information in additional attributes. The size, precision and scale information that is part of the description is now properly set for numeric types.

    • If you pass a Python list as one of the parameters to a DB-API 2 cursor, it is now automatically bound using an ARRAY constructor. If you pass a Python tuple, it is bound using a ROW constructor. This is useful for passing records as well as making use of the IN syntax.

    • Inversely, when a fetch method of a DB-API 2 cursor returns a PostgreSQL array, it is passed to Python as a list, and when it returns a PostgreSQL composite type, it is passed to Python as a named tuple. PyGreSQL uses a new fast built-in parser to achieve this. Anonymous composite types are also supported, but yield only an ordinary tuple containing text strings.

    • New type helpers Interval() and Uuid() have been added.

    • The connection has a new attribute “closed” that can be used to check whether the connection is closed or broken.

    • SQL commands are always handled as if they include parameters, i.e. literal percent signs must always be doubled. This consistent behavior is necessary for using pgdb with wrappers like SQLAlchemy.

    • PyGreSQL 5.0 will be supported as a database driver by SQLAlchemy 1.1.

  • Changes concerning both modules:
    • PyGreSQL now tries to raise more specific and appropriate subclasses of DatabaseError than just ProgrammingError. Particularly, when database constraints are violated, it raises an IntegrityError now.

    • The modules now provide get_typecast() and set_typecast() methods allowing to control the typecasting on the global level. The connection objects have type caches with the same methods which give control over the typecasting on the level of the current connection. See the documentation for details about the type cache and the typecast mechanisms provided by PyGreSQL.

    • Dates, times, timestamps and time intervals are now returned as the corresponding Python objects from the datetime module of the standard library. In earlier versions of PyGreSQL they had been returned as strings. You can restore the old behavior by deactivating the respective typecast functions, e.g. set_typecast(‘date’, str).

    • PyGreSQL now supports the “uuid” data type, converting such columns automatically to and from Python uuid.UUID objects.

    • PyGreSQL now supports the “hstore” data type, converting such columns automatically to and from Python dictionaries. If you want to insert Python objects as JSON data using DB-API 2, you should wrap them in the new HStore() type constructor as a hint to PyGreSQL.

    • PyGreSQL now supports the “json” and “jsonb” data types, converting such columns automatically to and from Python objects. If you want to insert Python objects as JSON data using DB-API 2, you should wrap them in the new Json() type constructor as a hint to PyGreSQL.

    • A new type helper Literal() for inserting parameters literally as SQL has been added. This is useful for table names, for instance.

    • Fast parsers cast_array(), cast_record() and cast_hstore for the input and output syntax for PostgreSQL arrays, composite types and the hstore type have been added to the C extension module. The array parser also allows using multi-dimensional arrays with PyGreSQL.

    • The tty parameter and attribute of database connections has been removed since it is not supported by PostgreSQL versions newer than 7.4.

Version 4.2.2 (2016-03-18)

  • The get_relations() and get_tables() methods now also return system views and tables if you set the optional “system” parameter to True.

  • Fixed a regression when using temporary tables with DB wrapper methods (thanks to Patrick TJ McPhee for reporting).

Version 4.2.1 (2016-02-18)

  • Fixed a small bug when setting the notice receiver.

  • Some more minor fixes and re-packaging with proper permissions.

Version 4.2 (2016-01-21)

  • The supported Python versions are 2.4 to 2.7.

  • PostgreSQL is supported in all versions from 8.3 to 9.5.

  • Set a better default for the user option “escaping-funcs”.

  • Force build to compile with no errors.

  • New methods get_parameters() and set_parameters() in the classic interface which can be used to get or set run-time parameters.

  • New method truncate() in the classic interface that can be used to quickly empty a table or a set of tables.

  • Fix decimal point handling.

  • Add option to return boolean values as bool objects.

  • Add option to return money values as string.

  • get_tables() does not list information schema tables any more.

  • Fix notification handler (Thanks Patrick TJ McPhee).

  • Fix a small issue with large objects.

  • Minor improvements of the NotificationHandler.

  • Converted documentation to Sphinx and added many missing parts.

  • The tutorial files have become a chapter in the documentation.

  • Greatly improved unit testing, tests run with Python 2.4 to 2.7 again.

Version 4.1.1 (2013-01-08)

  • Add NotificationHandler class and method. Replaces need for pgnotify.

  • Sharpen test for inserting current_timestamp.

  • Add more quote tests. False and 0 should evaluate to NULL.

  • More tests - Any number other than 0 is True.

  • Do not use positional parameters internally. This restores backward compatibility with version 4.0.

  • Add methods for changing the decimal point.

Version 4.1 (2013-01-01)

  • Dropped support for Python below 2.5 and PostgreSQL below 8.3.

  • Added support for Python up to 2.7 and PostgreSQL up to 9.2.

  • Particularly, support PQescapeLiteral() and PQescapeIdentifier().

  • The query method of the classic API now supports positional parameters. This an effective way to pass arbitrary or unknown data without worrying about SQL injection or syntax errors (contribution by Patrick TJ McPhee).

  • The classic API now supports a method namedresult() in addition to getresult() and dictresult(), which returns the rows of the result as named tuples if these are supported (Python 2.6 or higher).

  • The classic API has got the new methods begin(), commit(), rollback(), savepoint() and release() for handling transactions.

  • Both classic and DBAPI 2 connections can now be used as context managers for encapsulating transactions.

  • The execute() and executemany() methods now return the cursor object, so you can now write statements like “for row in cursor.execute(…)” (as suggested by Adam Frederick).

  • Binary objects are now automatically escaped and unescaped.

  • Bug in money quoting fixed. Amounts of $0.00 handled correctly.

  • Proper handling of date and time objects as input.

  • Proper handling of floats with ‘nan’ or ‘inf’ values as input.

  • Fixed the set_decimal() function.

  • All DatabaseError instances now have a sqlstate attribute.

  • The getnotify() method can now also return payload strings (#15).

  • Better support for notice processing with the new methods set_notice_receiver() and get_notice_receiver() (as suggested by Michael Filonenko, see #37).

  • Open transactions are rolled back when pgdb connections are closed (as suggested by Peter Harris, see #46).

  • Connections and cursors can now be used with the “with” statement (as suggested by Peter Harris, see #46).

  • New method use_regtypes() that can be called to let getattnames() return registered type names instead of the simplified classic types (#44).

Version 4.0 (2009-01-01)

  • Dropped support for Python below 2.3 and PostgreSQL below 7.4.

  • Improved performance of fetchall() for large result sets by speeding up the type casts (as suggested by Peter Schuller).

  • Exposed exceptions as attributes of the connection object.

  • Exposed connection as attribute of the cursor object.

  • Cursors now support the iteration protocol.

  • Added new method to get parameter settings.

  • Added customizable row_factory as suggested by Simon Pamies.

  • Separated between mandatory and additional type objects.

  • Added keyword args to insert, update and delete methods.

  • Added exception handling for direct copy.

  • Start transactions only when necessary, not after every commit().

  • Release the GIL while making a connection (as suggested by Peter Schuller).

  • If available, use decimal.Decimal for numeric types.

  • Allow DB wrapper to be used with DB-API 2 connections (as suggested by Chris Hilton).

  • Made private attributes of DB wrapper accessible.

  • Dropped dependence on mx.DateTime module.

  • Support for PQescapeStringConn() and PQescapeByteaConn(); these are now also used by the internal _quote() functions.

  • Added ‘int8’ to INTEGER types. New SMALLINT type.

  • Added a way to find the number of rows affected by a query() with the classic pg module by returning it as a string. For single inserts, query() still returns the oid as an integer. The pgdb module already provides the “rowcount” cursor attribute for the same purpose.

  • Improved getnotify() by calling PQconsumeInput() instead of submitting an empty command.

  • Removed compatibility code for old OID munging style.

  • The insert() and update() methods now use the “returning” clause if possible to get all changed values, and they also check in advance whether a subsequent select is possible, so that ongoing transactions won’t break if there is no select privilege.

  • Added “protocol_version” and “server_version” attributes.

  • Revived the “user” attribute.

  • The pg module now works correctly with composite primary keys; these are represented as frozensets.

  • Removed the undocumented and actually unnecessary “view” parameter from the get() method.

  • get() raises a nicer ProgrammingError instead of a KeyError if no primary key was found.

  • delete() now also works based on the primary key if no oid available and returns whether the row existed or not.

Version 3.8.1 (2006-06-05)

  • Use string methods instead of deprecated string functions.

  • Only use SQL-standard way of escaping quotes.

  • Added the functions escape_string() and escape/unescape_bytea() (as suggested by Charlie Dyson and Kavous Bojnourdi a long time ago).

  • Reverted code in clear() method that set date to current.

  • Added code for backwards compatibility in OID munging code.

  • Reorder attnames tests so that “interval” is checked for before “int.”

  • If caller supplies key dictionary, make sure that all has a namespace.

Version 3.8 (2006-02-17)

  • Installed new favicon.ico from Matthew Sporleder <>

  • Replaced snprintf by PyOS_snprintf

  • Removed NO_SNPRINTF switch which is not needed any longer

  • Clean up some variable names and namespace

  • Add get_relations() method to get any type of relation

  • Rewrite get_tables() to use get_relations()

  • Use new method in get_attnames method to get attributes of views as well

  • Add Binary type

  • Number of rows is now -1 after executing no-result statements

  • Fix some number handling

  • Non-simple types do not raise an error any more

  • Improvements to documentation framework

  • Take into account that nowadays not every table must have an oid column

  • Simplification and improvement of the inserttable() function

  • Fix up unit tests

  • The usual assortment of minor fixes and enhancements

Version 3.7 (2005-09-07)

Improvement of pgdb module:

  • Use Python standard datetime if mxDateTime is not available

Major improvements and clean-up in classic pg module:

  • All members of the underlying connection directly available in DB

  • Fixes to quoting function

  • Add checks for valid database connection to methods

  • Improved namespace support, handle search_path correctly

  • Removed old dust and unnecessary imports, added docstrings

  • Internal sql statements as one-liners, smoothed out ugly code

Version 3.6.2 (2005-02-23)

  • Further fixes to namespace handling

Version 3.6.1 (2005-01-11)

  • Fixes to namespace handling

Version 3.6 (2004-12-17)

  • Better DB-API 2.0 compliance

  • Exception hierarchy moved into C module and made available to both APIs

  • Fix error in update method that caused false exceptions

  • Moved to standard exception hierarchy in classic API

  • Added new method to get transaction state

  • Use proper Python constants where appropriate

  • Use Python versions of strtol, etc. Allows Win32 build.

  • Bug fixes and cleanups

Version 3.5 (2004-08-29)

Fixes and enhancements:

  • Add interval to list of data types

  • fix up method wrapping especially close()

  • retry pkeys once if table missing in case it was just added

  • wrap query method separately to handle debug better

  • use isinstance instead of type

  • fix free/PQfreemem issue - finally

  • miscellaneous cleanups and formatting

Version 3.4 (2004-06-02)

Some cleanups and fixes. This is the first version where PyGreSQL is moved back out of the PostgreSQL tree. A lot of the changes mentioned below were actually made while in the PostgreSQL tree since their last release.

  • Allow for larger integer returns

  • Return proper strings for true and false

  • Cleanup convenience method creation

  • Enhance debugging method

  • Add reopen method

  • Allow programs to preload field names for speedup

  • Move OID handling so that it returns long instead of int

  • Miscellaneous cleanups and formatting

Version 3.3 (2001-12-03)

A few cleanups. Mostly there was some confusion about the latest version and so I am bumping the number to keep it straight.

  • Added NUMERICOID to list of returned types. This fixes a bug when returning aggregates in the latest version of PostgreSQL.

Version 3.2 (2001-06-20)

Note that there are very few changes to PyGreSQL between 3.1 and 3.2. The main reason for the release is the move into the PostgreSQL development tree. Even the WIN32 changes are pretty minor.

Version 3.1 (2000-11-06)

  • Fix some quoting functions. In particular handle NULLs better.

  • Use a method to add primary key information rather than direct manipulation of the class structures

  • Break decimal out in _quote (in and treat it as float

  • Treat timestamp like date for quoting purposes

  • Remove a redundant SELECT from the get method speeding it, and insert (since it calls get) up a little.

  • Add test for BOOL type in typecast method to pgdbTypeCache class (

  • Fix to send port as integer to lower level function (

  • Change to speed up some operations

  • Allow updates on tables with no primary keys

Version 3.0 (2000-05-30)

Version 2.4 (1999-06-15)

  • Insert returns None if the user doesn’t have select permissions on the table. It can (and does) happen that one has insert but not select permissions on a table.

  • Added ntuples() method to query object (

  • Corrected a bug related to getresult() and the money type

  • Corrected a bug related to negative money amounts

  • Allow update based on primary key if munged oid not available and table has a primary key

  • Add many __doc__ strings (

  • Get method works with views if key specified

Version 2.3 (1999-04-17)

  • returns “localhost” when connected to Unix socket (

  • Use PyArg_ParseTupleAndKeywords in connect() (

  • fixes and cleanups (

  • Fixed memory leak in dictresult() (

  • Deprecated - functionality now in

  • More cleanups to the tutorial

  • Added fileno() method - (Mikhail Terekhov)

  • added money type to quoting function

  • Compiles cleanly with more warnings turned on

  • Returns PostgreSQL error message on error

  • Init accepts keywords (Jarkko Torppa)

  • Convenience functions can be overridden (Jarkko Torppa)

  • added close() method

Version 2.2 (1998-12-21)

  • Added user and password support thanks to Ng Pheng Siong (

  • Insert queries return the inserted oid

  • Add new pg wrapper (C module renamed to _pg)

  • Wrapped database connection in a class

  • Cleaned up some of the tutorial. (More work needed.)

  • Added version and __version__. Thanks to for the suggestion.

Version 2.1 (1998-03-07)

  • return fields as proper Python objects for field type

  • Cleaned up

  • Added dictresult method

Version 2.0 (1997-12-23)

  • Updated code for PostgreSQL 6.2.1 and Python 1.5

  • Reformatted code and converted to use full ANSI style prototypes

  • Changed name to PyGreSQL (from PyGres95)

  • Changed order of arguments to connect function

  • Created new type pgqueryobject and moved certain methods to it

  • Added a print function for pgqueryobject

  • Various code changes - mostly stylistic

Version 1.0b (1995-11-04)

  • Keyword support for connect function moved from library file to C code and taken away from library

  • Rewrote documentation

  • Bug fix in connect function

  • Enhancements in large objects interface methods

Version 1.0a (1995-10-30)

A limited release.

  • Module adapted to standard Python syntax

  • Keyword support for connect function in library file

  • Rewrote default parameters interface (internal use of strings)

  • Fixed minor bugs in module interface

  • Redefinition of error messages

Version 0.9b (1995-10-10)

The first public release.

  • Large objects implementation

  • Many bug fixes, enhancements, …

Version 0.1a (1995-10-07)

  • Basic libpq functions (SQL access)