source: trunk/docs/contents/changelog.rst @ 901

Last change on this file since 901 was 901, checked in by cito, 3 years ago

Improve creation of named tuples in Python 2.6 and 3.0

  • Property svn:keywords set to Author Date Id Revision
File size: 26.9 KB
4Version 5.0.4 (to be released)
6- query_formatted() can now be used without parameters.
7- The automatic renaming of columns that are invalid as field names of
8  named tuples now works more accurately in Python 2.6 and 3.0.
10Version 5.0.3 (2016-12-10)
12- It is now possible to use a custom array cast function by changing
13  the type caster for the 'anyarray' type.  For instance, by calling
14  set_typecast('anyarray', lambda v, c: v) you can have arrays returned
15  as strings instead of lists.  Note that in the pg module, you can also
16  call set_array(False) in order to return arrays as strings.
17- The namedtuple classes used for the rows of query results are now cached
18  and reused internally, since creating namedtuples classes in Python is a
19  somewhat expensive operation.  By default the cache has a size of 1024
20  entries, but this can be changed with the set_row_factory_size() function.
21  In certain cases this change can notably improve the performance.
22- The namedresult() method in the classic API now also tries to rename
23  columns that would result in invalid field names.
25Version 5.0.2 (2016-09-13)
27- Fixed an infinite recursion problem in the DB wrapper class of the classic
28  module that could occur when the underlying connection could not be properly
29  opened (bug report by Justin Pryzby).
31Version 5.0.1 (2016-08-18)
33- The update() and delete() methods of the DB wrapper now use the OID instead
34  of the primary key if both are provided. This restores backward compatibility
35  with PyGreSQL 4.x and allows updating the primary key itself if an OID exists.
36- The connect() function of the DB API 2.0 module now accepts additional keyword
37  parameters such as "application_name" which will be passed on to PostgreSQL.
38- PyGreSQL now adapts some queries to be able to access older PostgreSQL 8.x
39  databases (as suggested on the mailing list by Andres Mejia). However, these
40  old versions of PostgreSQL are not officially supported and tested any more.
41- Fixed an issue with Postgres types that have an OID >= 0x80000000 (reported
42  on the mailing list by Justin Pryzby).
43- Allow extra values that are not used in the command in the parameter dict
44  passed to the query_formatted() method (as suggested by Justin Pryzby).
45- Improved handling of empty arrays in the classic module.
46- Unused classic connections were not properly garbage collected which could
47  cause memory leaks (reported by Justin Pryzby).
48- Made C extension compatible with MSVC 9 again (this was needed to compile for
49  Python 2 on Windows).
51Version 5.0 (2016-03-20)
53- This version now runs on both Python 2 and Python 3.
54- The supported versions are Python 2.6 to 2.7, and 3.3 to 3.5.
55- PostgreSQL is supported in all versions from 9.0 to 9.5.
56- Changes in the classic PyGreSQL module (pg):
57    - The classic interface got two new methods get_as_list() and get_as_dict()
58      returning a database table as a Python list or dict. The amount of data
59      returned can be controlled with various parameters.
60    - A method upsert() has been added to the DB wrapper class that utilitses
61      the "upsert" feature that is new in PostgreSQL 9.5. The new method nicely
62      complements the existing get/insert/update/delete() methods.
63    - When using insert/update/upsert(), you can now pass PostgreSQL arrays as
64      lists and PostgreSQL records as tuples in the classic module.
65    - Conversely, when the query method returns a PostgreSQL array, it is passed
66      to Python as a list. PostgreSQL records are converted to named tuples as
67      well, but only if you use one of the get/insert/update/delete() methods.
68      PyGreSQL uses a new fast built-in parser to achieve this.  The automatic
69      conversion of arrays to lists can be disabled with set_array(False).
70    - The pkey() method of the classic interface now returns tuples instead
71      of frozenset. The order of the tuples is like in the primary key index.
72    - Like the DB-API 2 module, the classic module now also returns bool values
73      from the database as Python bool objects instead of strings.  You can
74      still restore the old behavior by calling set_bool(False).
75    - Like the DB-API 2 module, the classic module now also returns bytea
76      data fetched from the database as byte strings, so you don't need to
77      call unescape_bytea() any more.  This has been made configurable though,
78      and you can restore the old behavior by calling set_bytea_escaped(True).
79    - A method set_jsondecode() has been added for changing or removing the
80      function that automatically decodes JSON data coming from the database.
81      By default, decoding JSON is now enabled and uses the decoder function
82      in the standard library with its default parameters.
83    - The table name that is affixed to the name of the OID column returned
84      by the get() method of the classic interface will not automatically
85      be fully qualified any more. This reduces overhead from the interface,
86      but it means you must always write the table name in the same way when
87      you call the methods using it and you are using tables with OIDs.
88      Also, OIDs are now only used when access via primary key is not possible.
89      Note that OIDs are considered deprecated anyway, and they are not created
90      by default any more in PostgreSQL 8.1 and later.
91    - The internal caching and automatic quoting of class names in the classic
92      interface has been simplified and improved, it should now perform better
93      and use less memory. Also, overhead for quoting values in the DB wrapper
94      methods has been reduced and security has been improved by passing the
95      values to libpq separately as parameters instead of inline.
96    - It is now possible to use regular type names instead of the simpler
97      type names that are used by default in PyGreSQL, without breaking any
98      of the mechanisms for quoting and typecasting, which rely on the type
99      information. This is achieved while maintaining simplicity and backward
100      compatibility by augmenting the type name string objects with all the
101      necessary information under the cover. To switch regular type names on
102      or off (this is the default), call the DB wrapper method use_regtypes().
103    - A new method query_formatted() has been added to the DB wrapper class
104      that allows using the format specifications from Python.  A flag "inline"
105      can be set to specify whether parameters should be sent to the database
106      separately or formatted into the SQL.
107    - A new type helper Bytea() has been added.
108- Changes in the DB-API 2 module (pgdb):
109    - The DB-API 2 module now always returns result rows as named tuples
110      instead of simply lists as before. The documentation explains how
111      you can restore the old behavior or use custom row objects instead.
112    - The names of the various classes used by the classic and DB-API 2
113      modules have been renamed to become simpler, more intuitive and in
114      line with the names used in the DB-API 2 documentation.
115      Since the API provides only objects of these types through constructor
116      functions, this should not cause any incompatibilities.
117    - The DB-API 2 module now supports the callproc() cursor method. Note
118      that output parameters are currently not replaced in the return value.
119    - The DB-API 2 module now supports copy operations between data streams
120      on the client and database tables via the COPY command of PostgreSQL.
121      The cursor method copy_from() can be used to copy data from the database
122      to the client, and the cursor method copy_to() can be used to copy data
123      from the client to the database.
124    - The 7-tuples returned by the description attribute of a pgdb cursor
125      are now named tuples, i.e. their elements can be also accessed by name.
126      The column names and types can now also be requested through the
127      colnames and coltypes attributes, which are not part of DB-API 2 though.
128      The type_code provided by the description attribute is still equal to
129      the PostgreSQL internal type name, but now carries some more information
130      in additional attributes. The size, precision and scale information that
131      is part of the description is now properly set for numeric types.
132    - If you pass a Python list as one of the parameters to a DB-API 2 cursor,
133      it is now automatically bound using an ARRAY constructor. If you pass a
134      Python tuple, it is bound using a ROW constructor. This is useful for
135      passing records as well as making use of the IN syntax.
136    - Inversely, when a fetch method of a DB-API 2 cursor returns a PostgreSQL
137      array, it is passed to Python as a list, and when it returns a PostgreSQL
138      composite type, it is passed to Python as a named tuple. PyGreSQL uses
139      a new fast built-in parser to achieve this. Anonymous composite types are
140      also supported, but yield only an ordinary tuple containing text strings.
141    - New type helpers Interval() and Uuid() have been added.
142    - The connection has a new attribute "closed" that can be used to check
143      whether the connection is closed or broken.
144    - SQL commands are always handled as if they include parameters, i.e.
145      literal percent signs must always be doubled. This consistent behavior
146      is necessary for using pgdb with wrappers like SQLAlchemy.
147    - PyGreSQL 5.0 will be supported as a database driver by SQLAlchemy 1.1.
148- Changes concerning both modules:
149    - PyGreSQL now tries to raise more specific and appropriate subclasses of
150      DatabaseError than just ProgrammingError. Particularly, when database
151      constraints are violated, it raises an IntegrityError now.
152    - The modules now provide get_typecast() and set_typecast() methods
153      allowing to control the typecasting on the global level.  The connection
154      objects have got type caches with the same methods which give control
155      over the typecasting on the level of the current connection.
156      See the documentation on details about the type cache and the typecast
157      mechanisms provided by PyGreSQL.
158    - Dates, times, timestamps and time intervals are now returned as the
159      corresponding Python objects from the datetime module of the standard
160      library.  In earlier versions of PyGreSQL they had been returned as
161      strings.  You can restore the old behavior by deactivating the respective
162      typecast functions, e.g. set_typecast('date', str).
163    - PyGreSQL now support the "uuid" data type, converting such columns
164      automatically to and from Python uuid.UUID objects.
165    - PyGreSQL now supports the "hstore" data type, converting such columns
166      automatically to and from Python dictionaries.  If you want to insert
167      Python objects as JSON data using DB-API 2, you should wrap them in the
168      new HStore() type constructor as a hint to PyGreSQL.
169    - PyGreSQL now supports the "json" and "jsonb" data types, converting such
170      columns automatically to and from Python objects. If you want to insert
171      Python objects as JSON data using DB-API 2, you should wrap them in the
172      new Json() type constructor as a hint to PyGreSQL.
173    - A new type helper Literal() for inserting parameters literally as SQL
174      has been added.  This is useful for table names, for instance.
175    - Fast parsers cast_array(), cast_record() and cast_hstore for the input
176      and output syntax for PostgreSQL arrays, composite types and the hstore
177      type have been added to the C extension module. The array parser also
178      allows using multi-dimensional arrays with PyGreSQL.
179    - The tty parameter and attribute of database connections has been
180      removed since it is not supported any more since PostgreSQL 7.4.
182Version 4.2.2 (2016-03-18)
184- The get_relations() and get_tables() methods now also return system views
185  and tables if you set the optional "system" parameter to True.
186- Fixed a regression when using temporary tables with DB wrapper methods
187  (thanks to Patrick TJ McPhee for reporting).
189Version 4.2.1 (2016-02-18)
191- Fixed a small bug when setting the notice receiver.
192- Some more minor fixes and re-packaging with proper permissions.
194Version 4.2 (2016-01-21)
196- The supported Python versions are 2.4 to 2.7.
197- PostgreSQL is supported in all versions from 8.3 to 9.5.
198- Set a better default for the user option "escaping-funcs".
199- Force build to compile with no errors.
200- New methods get_parameters() and set_parameters() in the classic interface
201  which can be used to get or set run-time parameters.
202- New method truncate() in the classic interface that can be used to quickly
203  empty a table or a set of tables.
204- Fix decimal point handling.
205- Add option to return boolean values as bool objects.
206- Add option to return money values as string.
207- get_tables() does not list information schema tables any more.
208- Fix notification handler (Thanks Patrick TJ McPhee).
209- Fix a small issue with large objects.
210- Minor improvements of the NotificationHandler.
211- Converted documentation to Sphinx and added many missing parts.
212- The tutorial files have become a chapter in the documentation.
213- Greatly improved unit testing, tests run with Python 2.4 to 2.7 again.
215Version 4.1.1 (2013-01-08)
217- Add NotificationHandler class and method.  Replaces need for pgnotify.
218- Sharpen test for inserting current_timestamp.
219- Add more quote tests.  False and 0 should evaluate to NULL.
220- More tests - Any number other than 0 is True.
221- Do not use positional parameters internally.
222  This restores backward compatibility with version 4.0.
223- Add methods for changing the decimal point.
225Version 4.1 (2013-01-01)
227- Dropped support for Python below 2.5 and PostgreSQL below 8.3.
228- Added support for Python up to 2.7 and PostgreSQL up to 9.2.
229- Particularly, support PQescapeLiteral() and PQescapeIdentifier().
230- The query method of the classic API now supports positional parameters.
231  This an effective way to pass arbitrary or unknown data without worrying
232  about SQL injection or syntax errors (contribution by Patrick TJ McPhee).
233- The classic API now supports a method namedresult() in addition to
234  getresult() and dictresult(), which returns the rows of the result
235  as named tuples if these are supported (Python 2.6 or higher).
236- The classic API has got the new methods begin(), commit(), rollback(),
237  savepoint() and release() for handling transactions.
238- Both classic and DBAPI 2 connections can now be used as context
239  managers for encapsulating transactions.
240- The execute() and executemany() methods now return the cursor object,
241  so you can now write statements like "for row in cursor.execute(...)"
242  (as suggested by Adam Frederick).
243- Binary objects are now automatically escaped and unescaped.
244- Bug in money quoting fixed.  Amounts of $0.00 handled correctly.
245- Proper handling of date and time objects as input.
246- Proper handling of floats with 'nan' or 'inf' values as input.
247- Fixed the set_decimal() function.
248- All DatabaseError instances now have a sqlstate attribute.
249- The getnotify() method can now also return payload strings (#15).
250- Better support for notice processing with the new methods
251  set_notice_receiver() and get_notice_receiver()
252  (as suggested by Michael Filonenko, see #37).
253- Open transactions are rolled back when pgdb connections are closed
254  (as suggested by Peter Harris, see #46).
255- Connections and cursors can now be used with the "with" statement
256  (as suggested by Peter Harris, see #46).
257- New method use_regtypes() that can be called to let getattnames()
258  return regular type names instead of the simplified classic types (#44).
260Version 4.0 (2009-01-01)
262- Dropped support for Python below 2.3 and PostgreSQL below 7.4.
263- Improved performance of fetchall() for large result sets
264  by speeding up the type casts (as suggested by Peter Schuller).
265- Exposed exceptions as attributes of the connection object.
266- Exposed connection as attribute of the cursor object.
267- Cursors now support the iteration protocol.
268- Added new method to get parameter settings.
269- Added customizable row_factory as suggested by Simon Pamies.
270- Separated between mandatory and additional type objects.
271- Added keyword args to insert, update and delete methods.
272- Added exception handling for direct copy.
273- Start transactions only when necessary, not after every commit().
274- Release the GIL while making a connection
275  (as suggested by Peter Schuller).
276- If available, use decimal.Decimal for numeric types.
277- Allow DB wrapper to be used with DB-API 2 connections
278  (as suggested by Chris Hilton).
279- Made private attributes of DB wrapper accessible.
280- Dropped dependence on mx.DateTime module.
281- Support for PQescapeStringConn() and PQescapeByteaConn();
282  these are now also used by the internal _quote() functions.
283- Added 'int8' to INTEGER types. New SMALLINT type.
284- Added a way to find the number of rows affected by a query()
285  with the classic pg module by returning it as a string.
286  For single inserts, query() still returns the oid as an integer.
287  The pgdb module already provides the "rowcount" cursor attribute
288  for the same purpose.
289- Improved getnotify() by calling PQconsumeInput() instead of
290  submitting an empty command.
291- Removed compatibility code for old OID munging style.
292- The insert() and update() methods now use the "returning" clause
293  if possible to get all changed values, and they also check in advance
294  whether a subsequent select is possible, so that ongoing transactions
295  won't break if there is no select privilege.
296- Added "protocol_version" and "server_version" attributes.
297- Revived the "user" attribute.
298- The pg module now works correctly with composite primary keys;
299  these are represented as frozensets.
300- Removed the undocumented and actually unnecessary "view" parameter
301  from the get() method.
302- get() raises a nicer ProgrammingError instead of a KeyError
303  if no primary key was found.
304- delete() now also works based on the primary key if no oid available
305  and returns whether the row existed or not.
307Version 3.8.1 (2006-06-05)
309- Use string methods instead of deprecated string functions.
310- Only use SQL-standard way of escaping quotes.
311- Added the functions escape_string() and escape/unescape_bytea()
312  (as suggested by Charlie Dyson and Kavous Bojnourdi a long time ago).
313- Reverted code in clear() method that set date to current.
314- Added code for backwards compatibility in OID munging code.
315- Reorder attnames tests so that "interval" is checked for before "int."
316- If caller supplies key dictionary, make sure that all has a namespace.
318Version 3.8 (2006-02-17)
320- Installed new favicon.ico from Matthew Sporleder <>
321- Replaced snprintf by PyOS_snprintf.
322- Removed NO_SNPRINTF switch which is not needed any longer
323- Clean up some variable names and namespace
324- Add get_relations() method to get any type of relation
325- Rewrite get_tables() to use get_relations()
326- Use new method in get_attnames method to get attributes of views as well
327- Add Binary type
328- Number of rows is now -1 after executing no-result statements
329- Fix some number handling
330- Non-simple types do not raise an error any more
331- Improvements to documentation framework
332- Take into account that nowadays not every table must have an oid column
333- Simplification and improvement of the inserttable() function
334- Fix up unit tests
335- The usual assortment of minor fixes and enhancements
337Version 3.7 (2005-09-07)
339Improvement of pgdb module:
341- Use Python standard `datetime` if `mxDateTime` is not available
343Major improvements and clean-up in classic pg module:
345- All members of the underlying connection directly available in `DB`
346- Fixes to quoting function
347- Add checks for valid database connection to methods
348- Improved namespace support, handle `search_path` correctly
349- Removed old dust and unnessesary imports, added docstrings
350- Internal sql statements as one-liners, smoothed out ugly code
352Version 3.6.2 (2005-02-23)
354- Further fixes to namespace handling
356Version 3.6.1 (2005-01-11)
358- Fixes to namespace handling
360Version 3.6 (2004-12-17)
362- Better DB-API 2.0 compliance
363- Exception hierarchy moved into C module and made available to both APIs
364- Fix error in update method that caused false exceptions
365- Moved to standard exception hierarchy in classic API
366- Added new method to get transaction state
367- Use proper Python constants where appropriate
368- Use Python versions of strtol, etc. Allows Win32 build.
369- Bug fixes and cleanups
371Version 3.5 (2004-08-29)
373Fixes and enhancements:
375- Add interval to list of data types
376- fix up method wrapping especially close()
377- retry pkeys once if table missing in case it was just added
378- wrap query method separately to handle debug better
379- use isinstance instead of type
380- fix free/PQfreemem issue - finally
381- miscellaneous cleanups and formatting
383Version 3.4 (2004-06-02)
385Some cleanups and fixes.
386This is the first version where PyGreSQL is moved back out of the
387PostgreSQL tree. A lot of the changes mentioned below were actually
388made while in the PostgreSQL tree since their last release.
390- Allow for larger integer returns
391- Return proper strings for true and false
392- Cleanup convenience method creation
393- Enhance debugging method
394- Add reopen method
395- Allow programs to preload field names for speedup
396- Move OID handling so that it returns long instead of int
397- Miscellaneous cleanups and formatting
399Version 3.3 (2001-12-03)
401A few cleanups.  Mostly there was some confusion about the latest version
402and so I am bumping the number to keep it straight.
404- Added NUMERICOID to list of returned types. This fixes a bug when
405  returning aggregates in the latest version of PostgreSQL.
407Version 3.2 (2001-06-20)
409Note that there are very few changes to PyGreSQL between 3.1 and 3.2.
410The main reason for the release is the move into the PostgreSQL
411development tree.  Even the WIN32 changes are pretty minor.
413- Add Win32 support (
414- Fix some DB-API quoting problems (
415- Moved development into PostgreSQL development tree.
417Version 3.1 (2000-11-06)
419- Fix some quoting functions.  In particular handle NULLs better.
420- Use a method to add primary key information rather than direct
421  manipulation of the class structures
422- Break decimal out in `_quote` (in and treat it as float
423- Treat timestamp like date for quoting purposes
424- Remove a redundant SELECT from the `get` method speeding it,
425  and `insert` (since it calls `get`) up a little.
426- Add test for BOOL type in typecast method to `pgdbTypeCache` class
427  (
428- Fix to send port as integer to lower level function
429  (
430- Change to speed up some operations
431- Allow updates on tables with no primary keys
433Version 3.0 (2000-05-30)
435- Remove strlen() call from pglarge_write() and get size from object
436  (
437- Add a little more error checking to the quote function in the wrapper
438- Add extra checking in `_quote` function
439- Wrap query in for debugging
440- Add DB-API 2.0 support to pgmodule.c (
441- Add DB-API 2.0 wrapper (
442- Correct keyword clash (temp) in tutorial
443- Clean up layout of tutorial
444- Return NULL values as None (
445  (WARNING: This will cause backwards compatibility issues)
446- Change None to NULL in insert and update
447- Change hash-bang lines to use /usr/bin/env
448- Clearing date should be blank (NULL) not TODAY
449- Quote backslashes in strings in `_quote` (brian@CSUA.Berkeley.EDU)
450- Expanded and clarified build instructions (
451- Make code thread safe (
452- Add README.distutils ( &
453- Many fixes and increased DB-API compliance by,
454, and others to get the final
455  version ready to release.
457Version 2.4 (1999-06-15)
459- Insert returns None if the user doesn't have select permissions
460  on the table.  It can (and does) happen that one has insert but
461  not select permissions on a table.
462- Added ntuples() method to query object (
463- Corrected a bug related to getresult() and the money type
464- Corrected a bug related to negative money amounts
465- Allow update based on primary key if munged oid not available and
466  table has a primary key
467- Add many __doc__ strings (
468- Get method works with views if key specified
470Version 2.3 (1999-04-17)
472- returns "localhost" when connected to Unix socket
473  (
474- Use `PyArg_ParseTupleAndKeywords` in connect() (
475- fixes and cleanups (
476- Fixed memory leak in dictresult() (
477- Deprecated - functionality now in
478- More cleanups to the tutorial
479- Added fileno() method - (Mikhail Terekhov)
480- added money type to quoting function
481- Compiles cleanly with more warnings turned on
482- Returns PostgreSQL error message on error
483- Init accepts keywords (Jarkko Torppa)
484- Convenience functions can be overridden (Jarkko Torppa)
485- added close() method
487Version 2.2 (1998-12-21)
489- Added user and password support thanks to Ng Pheng Siong (
490- Insert queries return the inserted oid
491- Add new `pg` wrapper (C module renamed to _pg)
492- Wrapped database connection in a class
493- Cleaned up some of the tutorial.  (More work needed.)
494- Added `version` and `__version__`.
495  Thanks to for the suggestion.
497Version 2.1 (1998-03-07)
499- return fields as proper Python objects for field type
500- Cleaned up
501- Added dictresult method
503Version 2.0  (1997-12-23)
505- Updated code for PostgreSQL 6.2.1 and Python 1.5
506- Reformatted code and converted to use full ANSI style prototypes
507- Changed name to PyGreSQL (from PyGres95)
508- Changed order of arguments to connect function
509- Created new type `pgqueryobject` and moved certain methods to it
510- Added a print function for pgqueryobject
511- Various code changes - mostly stylistic
513Version 1.0b (1995-11-04)
515- Keyword support for connect function moved from library file to C code
516  and taken away from library
517- Rewrote documentation
518- Bug fix in connect function
519- Enhancements in large objects interface methods
521Version 1.0a (1995-10-30)
523A limited release.
525- Module adapted to standard Python syntax
526- Keyword support for connect function in library file
527- Rewrote default parameters interface (internal use of strings)
528- Fixed minor bugs in module interface
529- Redefinition of error messages
531Version 0.9b (1995-10-10)
533The first public release.
535- Large objects implementation
536- Many bug fixes, enhancements, ...
538Version 0.1a (1995-10-07)
540- Basic libpq functions (SQL access)
Note: See TracBrowser for help on using the repository browser.