tortoise-orm

• Add more index types.
• Add force_index, use_index to queryset.
• Fix F in update error with update_fields.
• Make delete query work with limit and order_by. (#697)
• Filter backward FK fields with IS NULL and NOT IS NULL filters (#700)
• Add select_for_update in update_or_create. (#702)
• Add Model.select_for_update.
• Add __search full text search to queryset.

• Fix type for modules.
• Fix select_related when related model specified more than once. (#679)
• Add __iter__ to model, now can just return model/models in fastapi response.
• Fix in_transaction bug caused by router. (#677) (#678)

• Add date part extract filtering.
• Add Manager support.
• Add db router support.
• Add nowait, skip_locked, of parameters to queryset.select_for_update.
• Add field name to validation exceptions.
• Compatible with asyncmy.
• Replace pypika to pypika-tortoise.

• Fixed validating JSON before decoding. (#623)
• Add model method update_or_create.
• Add batch_size parameter for bulk_create method.
• Fix save with F expression and field with source_field.

• Add model field validators.
• Allow function results in group by. (#608)

• Replace set TZ environment variable to TIMEZONE to avoid affecting global timezone.
• Allow passing module objects to models_paths param of Tortoise.init_models(). (#561)
• Implement PydanticMeta.backward_relations. (#536)
• Allow overriding PydanticMeta in PydanticModelCreator. (#536)
• Fixed make_native typo to make_naive in timezone module

• Support custom function in update. (#537)
• Add Model.refresh_from_db. (#549)
• Add timezone support, be careful to upgrade to this version, see docs for details. (#335)
• Remove aerich in case of cyclic dependency. (#558)

• Add on_delete in ManyToManyField. (#508)
• Support F expression in annotate. (#475)
• Fix QuerySet.select_related in case of join same table twice. (#525)
• Integrate Aerich into the install. (#530)

• Fixed inconsistency in integrity error exception of FastAPI
• add OSError to _get_comments except block

• Make DateField accept valid date str.
• Add QuerySet.select_for_update().
• check default for not None on pydantic model creation
• propagate default to pydantic model
• Add QuerySet.select_related().
• Add custom attribute name for Prefetch instruction.
• Add db_constraint for RelationalField family.

• We now do CI runs on a Windows VM as well, to try and prevent Windows specific regressions.
• Make F expression work with QuerySet.filter().
• Include py.typed in source distribution.
• Added datetime parsing from int for fields.DatetimeField.
• get_or_create passes the using_db= on if provided.
• Allow custom loop and connection_class parameters to be passed on to asyncpg.

• Default install of tortoise-orm now installs with no C-dependencies, if you want to use the C accelerators, please do a pip install tortoise-orm[accel] instead.
• Added <instance>.clone() method that will create a cloned instance in memory. To persist it you still need to call .save()
• .clone() will raise a ParamsError if tortoise can't generate a primary key. In that case do a .clone(pk=<newval>)
• If manually setting the primary key value to None and the primary key can be automatically generated, this will create a new record. We however still recommend the .clone() method instead.
• .save() can be forced to do a create by setting force_create=True
• .save() can be forced to do an update by setting force_update=True
• Setting update_fields for a .save() operation will strongly prefer to do an update if possible

• Make Field.default effect on db level when generate table
• Add converters instead of importing from pymysql
• Fix PostgreSQL BooleanField default value convertion
• Fix JSONField typed in pydantic_model_creator
• Add .sql() method on QuerySet

• fix: sqlite://:memory: in Windows thrown OSError: [WinError 123]
• Support bulk_create() insertion of records with overridden primary key when the primary key is DB-generated
• Add queryset.exists() and Model.exists().
• Add model subscription lookup, Model[<pkval>] that will return the object or raise KeyError

• Fix bad import of basestring
• Better handling of NULL characters in strings. Fixes SQLite, raises better error for PostgreSQL.
• Support .group_by() with join now

• Support F expression in .save() now
• IntEnumField accept valid int value and CharEnumField accept valid str value
• Pydantic models get created with globally unique identifier
• Leaf-detection to minimize duplicate Pydantic model creation
• Pydantic models with a Primary Key that is also a raw field of a relation is now not hidden when exclude_raw_fields=True as it is a critically important field
• Raise an informative error when a field is set as nullable and primary key at the same time
• Foreign key id's are now described to have the positive-integer range of the field it is related to
• Fixed prefetching over OneToOne relations
• Fixed __contains for non-text fields (e.g. JSONB)

• Allow Q expression to function with _filter parameter on aggregations
• Add manual .group_by() support
• Fixed regression where GROUP BY class is missing for an aggregate with a specified order.

• Fixed regression where GROUP BY class is missing for an aggregate with a specified order.

• Added preliminary support for Python 3.9
• TruncationTestCase now properly quotes table names when it clears them out.
• Add model signals support
• Added app_label to test initializer(...) and TORTOISE_TEST_APP as test environment variable.

This is a security fix release. We strongly recommend people upgrade.

Security fixes:

• Fixed SQL injection issue in MySQL
• Fixed SQL injection issues in MySQL when using contains, starts_with or ends_with filters (and their case-insensitive counterparts)
• Fixed malformed SQL for PostgreSQL and SQLite when using contains, starts_with or ends_with filters (and their case-insensitive counterparts)

Other changes:

• Added support for partial models:

  To create a partial model, one can do a .only(<fieldnames-as-strings>) as part of the QuerySet.
  This will create model instances that only have those values fetched.

  Persisting changes on the model is allowed only when:

  • All the fields you want to update is specified in <model>.save(update_fields=[...])
  • You included the Model primary key in the `.only(...)``

  To protect against common mistakes we ensure that errors get raised:

  • If you access a field that is not specified, you will get an AttributeError.
  • If you do a <model>.save() a IncompleteInstanceError will be raised as the model is, as requested, incomplete.
  • If you do a <model>.save(update_fields=[...]) and you didn't include the primary key in the .only(...),
    then IncompleteInstanceError will be raised indicating that updates can't be done without the primary key being known.
  • If you do a <model>.save(update_fields=[...]) and one of the fields in update_fields was not in the .only(...),
    then IncompleteInstanceError as that field is not available to be updated.

• Fixed bad SQL generation when doing a .values() query over a Foreign Key
• Added <model>.update_from_dict({...}) that will mass update values safely from a dictionary
• Fixed processing URL encoded password in connection string