Files
deb-python-cassandra-driver/changelog
2014-06-09 14:34:15 -07:00

200 lines
6.4 KiB
Plaintext

CHANGELOG
0.14.0
* fix for setting map to empty (Lifto)
* report when creating models with attributes that conflict with cqlengine (maedhroz)
* use stable version of sure package (maedhroz)
* performance improvements
0.13.0
* adding support for post batch callbacks (thanks Daniel Dotsenko github.com/dvdotsenko)
* fixing sync table for tables with multiple keys (thanks Daniel Dotsenko github.com/dvdotsenko)
* fixing bug in Bytes column (thanks Netanel Cohen-Tzemach github.com/natict)
* fixing bug with timestamps and DST (thanks Netanel Cohen-Tzemach github.com/natict)
0.12.0
* Normalize and unquote boolean values. (Thanks Kevin Deldycke github.com/kdeldycke)
* Fix race condition in connection manager (Thanks Roey Berman github.com/bergundy)
* allow access to instance columns as if it is a dict (Thanks Kevin Deldycke github.com/kdeldycke)
* Added support for blind partial updates to queryset (Thanks Danny Cosson github.com/dcosson)
* Model instance equality check bugfix (Thanks to Michael Hall, github.com/mahall)
* Fixed bug syncing tables with camel cased names (Thanks to Netanel Cohen-Tzemach, github.com/natict)
* Fixed bug dealing with eggs (Thanks Kevin Deldycke github.com/kdeldycke)
0.11.0
* support for USING TIMESTAMP <microseconds from epoch> via a .timestamp(timedelta(seconds=30)) syntax
- allows for long, timedelta, and datetime
* fixed use of USING TIMESTAMP in batches
* clear TTL and timestamp off models after persisting to DB
* allows UUID without dashes - (Thanks to Michael Hall, github.com/mahall)
* fixes regarding syncing schema settings (thanks Kai Lautaportti github.com/dokai)
0.10.0
* support for execute_on_exception within batches
0.9.2
* fixing create keyspace with network topology strategy
* fixing regression with query expressions
0.9
* adding update method
* adding support for ttls
* adding support for per-query consistency
* adding BigInt column (thanks @Lifto)
* adding support for timezone aware time uuid functions (thanks @dokai)
* only saving collection fields on insert if they've been modified
* adding model method that returns a list of modified columns
0.8.5
* adding support for timeouts
0.8.4
* changing value manager previous value copying to deepcopy
0.8.3
* better logging for operational errors
0.8.2
* fix for connection failover
0.8.1
* fix for models not exactly matching schema
0.8.0
* support for table polymorphism
* var int type
0.7.1
* refactoring query class to make defining custom model instantiation logic easier
0.7.0
* added counter columns
* added support for compaction settings at the model level
* deprecated delete_table in favor of drop_table
* deprecated create_table in favor of sync_table
* added support for custom QuerySets
0.6.0
* added table sync
0.5.2
* adding hex conversion to Bytes column
0.5.1
* improving connection pooling
* fixing bug with clustering order columns not being quoted
0.5
* eagerly loading results into the query result cache, the cql driver does this anyway,
and pulling them from the cursor was causing some problems with gevented queries,
this will cause some breaking changes for users calling execute directly
0.4.10
* changing query parameter placeholders from uuid1 to uuid4
0.4.7
* adding support for passing None into query batch methods to clear any batch objects
0.4.6
* fixing the way models are compared
0.4.5
* fixed bug where container columns would not call their child to_python method, this only really affected columns with special to_python logic
0.4.4
* adding query logging back
* fixed bug updating an empty list column
0.4.3
* fixed bug with Text column validation
0.4.2
* added support for instantiating container columns with column instances
0.4.1
* fixed bug in TimeUUID from datetime method
0.4.0
* removed default values from all column types
* explicit primary key is required (automatic id removed)
* added validation on keyname types on .create()
* changed table_name to __table_name__, read_repair_chance to __read_repair_chance__, keyspace to __keyspace__
* modified table name auto generator to ignore module name
* changed internal implementation of model value get/set
* added TimeUUID.from_datetime(), used for generating UUID1's for a specific
time
0.3.3
* added abstract base class models
0.3.2
* comprehesive rewrite of connection management (thanks @rustyrazorblade)
0.3
* added support for Token function (thanks @mrk-its)
* added support for compound partition key (thanks @mrk-its)s
* added support for defining clustering key ordering (thanks @mrk-its)
* added values_list to Query class, bypassing object creation if desired (thanks @mrk-its)
* fixed bug with Model.objects caching values (thanks @mrk-its)
* fixed Cassandra 1.2.5 compatibility bug
* updated model exception inheritance
0.2.1
* adding support for datetimes with tzinfo (thanks @gdoermann)
* fixing bug in saving map updates (thanks @pandu-rao)
0.2
* expanding internal save function to use update where appropriate
* adding set, list, and map collection types
* adding support for allow filtering flag
* updating management functions to work with cassandra 1.2
* fixed a bug querying datetimes
* modifying datetime serialization to preserve millisecond accuracy
* adding cql function call generators MaxTimeUUID and MinTimeUUID
0.1.1
* fixed a bug occurring when creating tables from the REPL
0.1
* added min_length and max_length validators to the Text column
* added == and != equality operators to model class
* fixed bug with default values that would evaluate to False (ie: 0, '')
0.0.9
* fixed column inheritance bug
* manually defined table names are no longer inherited by subclasses
0.0.8
* added configurable read repair chance to model definitions
* added configurable keyspace strategy class and replication factor to keyspace creator
0.0.7
* fixed manual table name bug
* changed model level db_name field to table_name
0.0.6
* added TimeUUID column
0.0.5
* added connection pooling
* adding a few convenience query classmethods to the model class
0.0.4-ALPHA
* added Sphinx docs
* changing create_column_family management function to create_table
* changing delete_column_family management function to delete_table
* added partition key validation to QuerySet delete method
* added .get() method to QuerySet
* added create method shortcut to the model class
0.0.3-ALPHA
* added queryset result caching
* added queryset array index access and slicing
* updating table name generation (more readable)