:orphan: 1.x Changelog ============= .. changelog:: 1.8.2 :date: 2025-12-12 .. change:: add `db` group alias :type: feature :pr: 622 Add a `db` shorthand alias to the `database` group. This allows `litestar|alchemy db` or `litestar|alchemy database` to work interchangeably. .. change:: import error while generating migrations :type: bugfix :pr: 630 Fixes `passlib` and `pwdlib` import errors while creating migrations Cause: We added the `sa.PasslibHasher = PasslibHasher` and `sa.PwdlibHasher = PwdlibHasher` types in `script.py.mako`. As a result, when a user installs only Advanced Alchemy and creates a migration, these files are imported. Since they reference types from `passlib` and `pwdlib`, which are not installed by default, the import fails and triggers this error. .. change:: add missing type parameter to AsyncServiceT_co and SyncServiceT_… :type: bugfix :pr: 612 Discovered a runtime issue with an inconsistent type declaration when upgrading a litestar project to use version 1.8.0 introduced .. changelog:: 1.8.1 :date: 2025-12-06 .. change:: pin default installed python to 3.10 :type: bugfix :pr: 601 Update the installation process to set the default Python version to 3.10 instead of 3.9. There are some testing & docs packages we use that are causing issues. We can pin 3.10 until 3.9 support is removed. There is still a CI tests for 3.9 .. change:: adding string representation to PasswordHash and EncryptedString :type: bugfix :pr: 598 :issue: 596 Add string representation while generating migrations for models with `PasswordHash` or `EncryptedString` columns. .. change:: error message handling and isolation in repositories :type: bugfix :pr: 605 :issue: 597 Correct error message retrieval and ensure that error message overrides are isolated for different repository instances. This improves the clarity and reliability of error messages across the application. .. change:: correct race condition in `with_for_update` :type: bugfix :pr: 607 This corrects an issue in the `with_for_update` behavior: - Before the change, passing `with_for_update` to service.update() or repository.update() only affected the post-flush session.refresh() call. The row that gets copied and mutated was always retrieved with a plain SELECT, so two concurrent writers could both read the same version - Now the `with_for_update` flag is honored when the row is first fetched (both in the service’s item_id branch and inside SQLAlchemyAsyncRepository.get()). When you call service.update(..., with_for_update=True) (or pass the richer dict form/ForUpdateArg), the initial SELECT ... FOR UPDATE runs, so the session holds the expected lock before any field copying or merges occur. .. changelog:: 1.8.0 :date: 2025-10-28 .. change:: ensure `has_dict_attribute` checks for `__dict__` attribute :type: bugfix :pr: 579 The previous implementation used `isinstance` against the `DictProtocol` type returned `True` for any object. The `isinstance` call is replaced with `hasattr` (which should also be faster). With this change, I believe the `DictProtocol` class could also be removed, but I kept the changes to a minimum. .. change:: adding string representation to StoredObject :type: bugfix :pr: 582 Issues running migration created for `advanced_alchemy.types.FileObject` column on a model. .. change:: surface FileObject session errors; align commit/rollback semantics :type: bugfix :pr: 580 :issue: 543 - Previously save/delete failures were only logged, so callers believed commits succeeded when storage ops had already failed. - Sync commit: processes saves sequentially, then deletes. Logs failures with tracebacks and re-raises. Ignores FileNotFoundError on delete. Stops on first save error. Clears internal state only on full success. - Async commit: runs saves and deletes concurrently via asyncio.gather. Logs each failure with exc_info. Raises the first real Exception, and lets BaseException (e.g., CancelledError) bubble. Attempts deletes even if a save fails. Clears state only on full success. - Rollback (sync/async): deletes only files saved during the current transaction. Ignores FileNotFoundError. Logs and re-raises other errors. Clears state after processing. - Tracking: records successful saves in _saved_in_transaction for targeted rollback cleanup; state is retained on error to allow inspection/retry. .. change:: add `sort_order` to mixin columns for consistent table layout :type: feature :pr: 581 - Add sort_order=-100 to primary key columns (id) across all mixins - Add sort_order=3001 to sentinel column - Add sort_order=3002 to created_at audit column - Add sort_order=3003 to updated_at audit column This ensures consistent column ordering in generated tables: primary keys first, user columns in middle, audit/sentinel columns last. .. change:: use `property` in `SQLAlchemyDTO` with `MappedAsDataclass` :type: feature :pr: 447 Allow better compatibility with `MappedAsDataclass` and the `SQLALchemyDTO` .. change:: add support for SQLAlchemy func expressions in filter classes :type: feature :pr: 585 :issue: 519 Adds support for SQLAlchemy func() expressions in filter classes to eliminate type checker errors when using database functions like `func.random()` or `func.lower()` **Changes** - Updated `OrderBy`, `BeforeAfter`, `OnBeforeAfter`, `CollectionFilter`, `NotInCollectionFilter`, `ComparisonFilter` - Enhanced `_get_instrumented_attr()` to handle new types - Mock repositories extract field names from `InstrumentedAttribute`, raise helpful error for func expressions (can't execute SQL in-memory) - Added integration tests for func expressions Note: Different databases use different function names (PostgreSQL: `func.random()`, MySQL: `func.rand()`, SQL Server: `func.newid()`) .. changelog:: 1.7.0 :date: 2025-10-13 .. change:: handle and compare `numpy` arrays :type: feature :pr: 550 Direct equality comparisons (`!=`) with numpy arrays in repository update methods raised `ValueError: The truth value of an array with more than one element is ambiguous` Adds a safe comparison utility that handle numpy arrays gracefully .. change:: enhance visibility of syntax blocks in docs :type: feature :pr: 570 Migrates custom documentation styling from hardcoded colors to Shibuya theme's native CSS variable system. This ensures consistent theming across light/dark modes and leverages the configured `accent_color: "amber"` setting. .. change:: align CLI commands with complete Alembic API :type: feature :pr: 569 :issue: 566, 568 Implements complete API parity with Alembic 1.16.5 CLI by adding 9 missing commands and completing the stamp command with all options. .. change:: lazy attributes getting accessed on repository update method :type: bugfix :pr: 553 :issue: 552 This fixes #552 by moving the guard condition up. .. change:: prevent update() from overwriting unset fields with None (#560) :type: bugfix :pr: 563 :issue: 560 Added `was_attribute_set()` helper function that uses SQLAlchemy's instance state inspection to check which attributes were actually modified/set on the input instance. The update() method now only copies attributes that have been explicitly set by the user. .. change:: updated_at not being updated :type: bugfix :pr: 551 `updated_at` was not correctly updated on some base models. .. change:: ensure to_model called with update operation for all data types :type: bugfix :pr: 575 :issue: 555 This PR fixes a bug in the Service layer's `update()` method where dict, Pydantic, msgspec, and attrs data bypassed the `to_model()` operation map, preventing custom `to_model()` implementations from being invoked during update operations. .. change:: handle deleted objects gracefully in auto_expunge :type: bugfix :pr: 574 :issue: 514 Resolves InvalidRequestError when deleting objects with `auto_expunge=True` and `auto_commit=True` enabled. .. change:: punctuation symbols not visible in dark mode syntax highlighting :type: bugfix :pr: 576 Add explicit color styling for `.highlight .p` (punctuation tokens) in both light and dark modes using Shibuya theme's `--gray-12` variable. This ensures brackets, parentheses, commas, and other punctuation symbols are properly visible when viewing code examples in dark mode. .. change:: correct closure variable capture in filter provider loops :type: bugfix :pr: 573 :issue: 507 Fixes a closure bug where multiple fields in `in_fields` and `not_in_fields` arrays resulted in only the last field working correctly. This was caused by loop variables being captured by reference rather than by value in nested function definitions. .. change:: agent workflow and knowledge repository :type: misc :pr: 567 Introduce a structured agent workflow and comprehensive development guides for the AI agent based development. This includes detailed instructions for agents, directory structures for requirements, and updates to existing documentation. .. changelog:: 1.6.3 :date: 2025-09-19 .. change:: additional update and update_many corrections. :type: bugfix :pr: 537 :issue: 464, 535 Updates `update` and `update_many` to properly handle relationships and returning support. .. change:: TypeError when initializing SQLAlchemyAsyncQueryRepository :type: bugfix :pr: 538 :issue: 534 Corrects a TypeError reported from the init method of `SQLAlchemyAsyncQueryRepository` .. change:: property correction :type: bugfix :pr: 539 Property correction for updates .. change:: improve `sync_tools` thread handling and update dependencies :type: bugfix :pr: 545 Enhance thread management in `sync_tools` to improve performance and reliability. .. changelog:: 1.6.2 :date: 2025-08-29 .. change:: enable loading config from working directory :type: feature :pr: 527 :issue: 491 Correctly allow loading configuration from the current directory .. change:: correctly handle lazy attributes on update :type: bugfix :pr: 533 Correctly handle `viewownly` and `lazy` loaded relationships during update. .. change:: prevent AttributeError when schema_dump receives None values :type: bugfix :pr: 530 - `advanced_alchemy/service/typing.py`: Fixed `has_dict_attribute()` function - `tests/unit/test_attrs_integration.py`: Added test case for `None` value handling .. change:: add warning message when using fallback value for a primary key :type: bugfix :pr: 529 - Add warning message when using `uuid` instead of `nanoid` - Add warning message when using `uuid4` instead of `uuid6` or `uuid7` .. change:: litestar fileobject example :type: bugfix :pr: 531 Fix litestar fileobject example .. change:: pass Content-Type and metadata to backend storage :type: bugfix :pr: 528 - Modified `save_object` and `save_object_async` methods to collect attributes from `FileObject` - Pass `content_type` as `"Content-Type"` in the `attributes` parameter - Include any custom metadata from `FileObject.metadata` - Added proper typing for the attributes dictionary .. changelog:: 1.6.1 :date: 2025-08-26 .. change:: `to_schema` and `attrs` type hint correction :type: bugfix :pr: 516 Corrects an issue where the type hint was incorrect when `attrs` or `cattrs` were not installed. .. change:: suppress `passlib` caused pytest warnings and other session warnings :type: bugfix :pr: 518 Suppress warnings caused by `passlib` during testing. .. change:: `IdentityPrimaryKey` correctly generates an `IDENTITY` DDL :type: bugfix :pr: 523 Ensure that the `IdentityPrimaryKey` correctly generates `IDENTITY` DDL across multiple database dialects, including PostgreSQL, Oracle, and SQL Server. Update dependencies and add tests to verify the functionality. .. changelog:: 1.6.0 :date: 2025-08-18 .. change:: server side session backend :type: feature :pr: 429 Implements a server side session backend using SQLAlchemy. Works with an Async or Sync configuration. .. change:: handle relationship data in model_from_dict for service.create() :type: bugfix :pr: 512 Fixed regression where service.create() method stopped handling relationship data correctly when passed SQLAlchemy model instances. Changed model_from_dict() in _util.py to use `__mapper__.attrs.keys()` instead of `__mapper__.columns.keys()` to include relationship attributes alongside column attributes. - Use `attrs.keys()` to include both columns and relationships - Add comprehensive tests for relationship handling in model_from_dict - Verify unknown attributes are still ignored .. changelog:: 1.5.0 :date: 2025-08-13 .. change:: correct typing issue in `litestar` example :type: bugfix :pr: 498 Fixed typing issue in `Litestar` usage documentation .. change:: correctly handle `id_attribute` with `update` :type: bugfix :pr: 502 Correctly merge attributes onto existing instance when using `id_attribute` and `update` .. change:: gzip and zipped fixture support :type: feature :pr: 500 Contains support for automatically extracting and loading data from zipped fixture files .. change:: match against complex types :type: feature :pr: 501 Correctly handle complex data types for matching fields .. change:: `attrs` integration :type: feature :pr: 503 Adds `attrs` support into the `ResultConverter` mixin. This enables `to_schema` and `schema_dump` to natively understand `attrs`. .. changelog:: 1.4.5 :date: 2025-06-28 .. change:: add the DefaultBase class to __all__ :type: feature :pr: 482 :issue: 481 Adds [`DefaultBase`](https://github.com/litestar-org/advanced-alchemy/blob/6cc26ef8d53bc04f89a070337f8b0ab07a1bac46/advanced_alchemy/base.py#L517) class to `__all__` to match other public classes in the module. .. change:: Update list and count :type: bugfix :pr: 487 Minor adjustment to the list and count method .. changelog:: 1.4.4 :date: 2025-05-26 .. change:: support for alembic 1.16 `toml_file` configuration :type: bugfix :pr: 479 Updates the AlembicCommand to use named arguments and support Alembic 1.16's new `toml_file` parameter. .. changelog:: 1.4.3 :date: 2025-05-12 .. change:: add __all__ exports for password hashing backends :type: feature :pr: 471 This update adds __all__ exports for the Argon2, Passlib, and Pwdlib hashing backends, improving module visibility and usability. .. change:: Add identity Mixin for Primary Keys :type: feature :pr: 473 :issue: 441 The sequences based BigInt key offers the most compatibility, but many would prefer to use the Identity column when the database supports it. This changes implements a basic Identity primary key mixin .. change:: `wrap_exceptions` is re-enabled :type: bugfix :pr: 475 :issue: 472 `wrap_exceptions` is now correctly passed into the exception handler context manager. Fixes #472 .. changelog:: 1.4.2 :date: 2025-05-04 .. change:: correct type hints for with_for_update to ForUpdateParameter :type: bugfix :pr: 465 This change fixes the type hint for the `with_for_update` parameter in the repositories. .. change:: BigIntPrimaryKey does not respect schema names :type: bugfix :pr: 469 :issue: 466 BigIntPrimaryKey will now respect schema names. Fixes #466 .. changelog:: 1.4.1 :date: 2025-04-28 .. change:: raise if filter operator is not in `operators_map` :type: bugfix :pr: 463 :issue: 453 Raise exception if filter operator does not exist in operators_map Fixes #453 .. change:: `uniquify` respects init method override :type: bugfix :pr: 462 Passing `uniquify` as an `__init__` argument now works as expected. .. changelog:: 1.4.0 :date: 2025-04-27 .. change:: PasswordHash field type :type: feature :pr: 452 Implements a PasswordHash field type with multiple supported backends. Includes built-in backends for: - `passlib` - `argon2` - `pwdlib` .. changelog:: 1.3.2 :date: 2025-04-25 .. change:: remove stringified type hint :type: bugfix :pr: 457 "De-stringifies" the Filter type hints to prevent runtime type resolutions in some cases .. change:: FileObject native Pydantic Core integration :type: bugfix :pr: 458 File object will now serialize properly in pydantic. More complete FastAPI examples added. .. changelog:: 1.3.1 :date: 2025-04-21 .. change:: updated example `litestar_service.py` model :type: bugfix :pr: 450 :issue: 449 ## fixes #449 relationship updated on models: - AuthorModel - BookModel .. change:: `create_service_provider` supports any configuration now :type: bugfix :pr: 451 The Litestar service provider now allows a user to specify the specific dependency key to use for the session. Previously the factory only worked with the `db_session` key. .. change:: update service provider to use dynamic session dependency key :type: bugfix :pr: 454 Update the Litestar service provider to use dynamic session dependency key .. change:: allows positional args for session :type: feature :pr: 455 This change allows for arguments to also be matched when generating a service provider closure. .. changelog:: 1.3.0 :date: 2025-04-18 .. change:: btn ui :type: bugfix :pr: 446 Corrects the button UI in the documentation under certain viewport sizes. .. change:: add dependency provider :type: feature :pr: 431 Add dependency factories for filters. .. changelog:: 1.2.0 :date: 2025-04-15 .. change:: migration generation produces duplicated unique constraints :type: bugfix :pr: 434 :issue: 427 Removes column re-ordering component was incorrectly causing incorrect constraints to be genreated. Fixes #427 .. change:: make `SentinelMixin` compatible with `MappedAsDataclass` :type: bugfix :pr: 442 `MappedAsDataclass` is a mixin introduced in SQLAlchemy 2.0. It introduces massive DX improvements to SQLAlchemy by introducing dataclass type validation to SQLAlchemy models. However, this mixin is incompatible with SQLAlchemy's recommended method of implementing a sentinel column as written in their [documentation](https://docs.sqlalchemy.org/en/20/core/connections.html#configuring-sentinel-columns). This PR fixes this incompatibility as suggested by the SQLAlchemy maintainer in this [discussion](https://github.com/sqlalchemy/sqlalchemy/discussions/12519#discussioncomment-12804658). .. change:: enable standard order by :type: feature :pr: 438 Enables the standard `UnaryOperator` order by support in addition to the existing `OrderingPair` .. change:: additional filter configuration options :type: feature :pr: 444 Implements the following filters as configurable options: - NotInCollection - Collection Search now also accepts a set of strings in addition to a comma delimmited list. .. changelog:: 1.1.1 :date: 2025-04-07 .. change:: fsspec is not installed :type: bugfix :pr: 432 Corrects an import issue when `fsspec` and `obstore` are both missing. .. changelog:: 1.1.0 :date: 2025-04-06 .. change:: add stamp command :type: feature :pr: 428 Adds the Alembic `stamp` command to the CLI that will stamp the current database state into the migrations directory. .. change:: adds an `ExistsFilter` and `NotExists` filter :type: feature :pr: 336 :issue: 331 Implements new `Exists` and `NotExists` filters to more easily apply this type of logic to queries. Closes #331 .. change:: fully migrate to `pytest-databases` :type: feature :pr: 430 Migrates all database fixtures to `pytest-database` .. change:: file object data type :type: feature :pr: 291 :issue: 24 Implement a file data type that leverages `obstore` or `fsspec`. Supports any supported FSSpec or Obstore backend it including `sftp`, `gcs`, `s3`, `local`, and more. .. change:: Implements a `MultiFilter` type for complex searches :type: feature :pr: 311 This PR implements a "Multi-Filter" Filter type. It allows: - Create a collection of filters from an input - Allows filters to be groups with and/or logic .. changelog:: 1.0.2 :date: 2025-04-01 .. change:: prevent forward resolution issues :type: bugfix :pr: 423 Removes some stringified representations to help with the forward resolution of `datetime` and `Collection`. .. change:: correctly set `uniquify` from `new` :type: bugfix :pr: 424 Unquify is now correctly set when passed into the `new`/`init` methods. Introduced tests for `sync_tools` utilities, including `maybe_async_`, `maybe_async_context`, `SoonValue`, `TaskGroup`, and others. Improves coverage for async and sync function handling, context managers, and value management. .. change:: remove accidental litestar import :type: bugfix :pr: 426 Remove an incorrect import of `console` from `litestar.cli._utils` and replace it with a correct import from `rich`. This change ensures proper functionality without unnecessary dependencies. .. changelog:: 1.0.1 :date: 2025-03-19 .. change:: properly serialize `Relationship` type hints :type: bugfix :pr: 422 Adds `sqlalchemy.orm.Relationship` to the supported type hints for the `SQLAlchemyDTO` .. changelog:: 1.0.0 :date: 2025-03-18 .. change:: remove deprecated packages removed in `v1.0.0` :type: misc :pr: 419 Removes deprecated packages and prepares for 1.0 release. .. change:: logic correction for window function :type: bugfix :pr: 421 Corrects the logic for using a count with a window function. .. changelog:: 0.34.0 :date: 2025-03-10 .. change:: allow custom `not_found` error messages :type: feature :pr: 417 :issue: 391 Enhance the SQLAlchemy exception wrapper to handle NotFoundError with custom error messages and improved error handling. This includes: - Adding a 'not_found' key to ErrorMessages type - Extending wrap_sqlalchemy_exception to catch and handle NotFoundError - Updating default error message templates with a not_found message - Adding unit tests for custom NotFoundError handling .. change:: Refactor Sanic extension for multi-config support :type: feature :pr: 415 :issue: 375 This commit refactors the Sanic extension for Advanced Alchemy: - Refactored configuration handling with support for multiple database configurations - Added methods for retrieving async and sync sessions, engines, and configs - Improved dependency injection with new provider methods - Simplified extension initialization and registration - Updated example and test files to reflect new extension structure - Removed deprecated methods and simplified the extension interface .. changelog:: 0.33.2 :date: 2025-03-09 .. change:: simplify session type hints in service providers :type: bugfix :pr: 414 Remove unnecessary scoped session type hints from service provider functions. Prevents the following exception from being incorrectly raised: `TypeError: Type unions may not contain more than one custom type - type typing.Union[sqlalchemy.ext.asyncio.session.AsyncSession, sqlalchemy.ext.asyncio.scoping.async_scoped_session[sqlalchemy.ext.asyncio.session.AsyncSession], NoneType] is not supported.` .. changelog:: 0.33.1 :date: 2025-03-07 .. change:: add session to namespace signature :type: feature :pr: 412 The new filter providers expect that the sessions are in the signature namespace. This ensures there are no issues when configuring the plugin. .. changelog:: 0.33.0 :date: 2025-03-07 .. change:: Add dependency factory utilities :type: feature :pr: 405 Introduces a new module `advanced_alchemy.extensions.litestar.providers` with comprehensive dependency injection utilities for SQLAlchemy services in Litestar. The module provides: - Dynamic filter configuration generation - Dependency caching mechanism - Flexible filter and pagination support - Singleton metaclass for dependency management - Configurable filter and search dependencies .. changelog:: 0.32.2 :date: 2025-02-26 .. change:: Litestar extension: Use ``SerializationPlugin`` instead of ``SerializationPluginProtocol`` :type: misc :pr: 401 Use ``SerializationPlugin`` instead of ``SerializationPluginProtocol`` .. changelog:: 0.32.1 :date: 2025-02-26 .. change:: Litestar extension: Use ``CLIPlugin`` instead of ``CLIPluginProtocol`` :type: misc :pr: 399 Internal change migrating from using Litestar's ``CLIPluginProtocol`` to ``CLIPlugin``. .. changelog:: 0.32.0 :date: 2025-02-23 .. change:: remove `limit` and `offset` from count statement :type: bugfix :pr: 395 Remove `limit` and `offset` from count statement .. change:: rename `force_basic_query_mode` :type: misc :pr: 396 Renames `force_basic_query_mode` to `count_with_window_function`. This is also exposed as a class/init parameter for the service and repository. .. change:: add Enum to default type decoders :type: feature :pr: 397 Extends the default `msgspec` type decoders to handle Enum types by converting them to their underlying value during serialization .. changelog:: 0.31.0 :date: 2025-02-18 .. change:: Fix reference in `changelog.py` :type: bugfix :pr: 383 Should link to the AA repo, not litestar :) .. change:: Query repository list method for custom queries :type: bugfix :pr: 379 :issue: 338 Fix query repositories list method according to [documentation](https://docs.advanced-alchemy.litestar.dev/latest/usage/repositories.html#query-repository). Now its return a list of tuples with values instead of first column of the query. .. change:: remove 3.8 support :type: misc :pr: 386 Removes 3.8 support and removes future annotations in a few places for better compatibility .. change:: remove future annotations :type: feature :pr: 387 This removes the usage of future annotations. .. change:: add `uniquify` to service and repo :type: feature :pr: 389 Exposes the `uniquify` flag in all functions on the repository and add to the service .. change:: improved default serializer :type: feature :pr: 390 Improves the default serializer so that it handles various types a bit better .. changelog:: 0.30.3 :date: 2025-01-26 .. change:: add `wrap_exceptions` option to exception handler. :type: feature :pr: 363 :issue: 356 When `wrap_exceptions` is `False`, the original SQLAlchemy error message will be raised instead of the wrapped Repository error Fixes #356 (Bug: `wrap_sqlalchemy_exception` masks db errors) .. change:: simplify configuration hash :type: feature :pr: 366 The hashing method on the SQLAlchemy configs can be simplified. This should be enough to define a unique configuration. .. change:: use `lifespan` context manager in Starlette and FastAPI :type: bugfix :pr: 368 :issue: 367 Modifies the Starlette and FastAPI integrations to use the `lifespan` context manager instead of the `startup`\`shutdown` hooks. If the application already has a lifespan set, it is wrapped so that both execute. .. changelog:: 0.30.2 :date: 2025-01-21 .. change:: add hash to config classes :type: feature :pr: 358 :issue: 357 Adds hash function to `SQLAlchemySyncConfig` and `SQLAlchemyAsyncConfig` classes. .. changelog:: 0.30.1 :date: 2025-01-20 .. change:: Using init db CLI command creates migrations directory in unexpected place :type: bugfix :pr: 354 :issue: 351 When initializing migrations with the CLI, if no directory is specified, the directory from the configuration will be used. .. changelog:: 0.30.0 :date: 2025-01-19 .. change:: standardize on `autocommit_include_redirect` :type: bugfix :pr: 349 The flask plugin incorrectly used the term `autocommit_with_redirect` instead of the existing `autocommit_include_redirect`. This changes makes the name consistent before we bump to a `1.x` release .. change:: implement default schema serializer :type: bugfix :pr: 350 This corrects an issue that caused the Flask extension to use the incorrect serializer for encoding JSON .. change:: refactored integration with CLI support :type: feature :pr: 352 Refactored the Starlette and FastAPI integration to support multiple configurations and sessions. Additionally, FastAPI will now have the database commands automatically registered with the FastAPI CLI. .. change:: reorganize Sanic extension :type: feature :pr: 353 The Sanic integration now aligns with the structure and idioms used in the other integrations. .. changelog:: 0.29.1 :date: 2025-01-17 .. change:: add convenience hooks for `to_model` operations :type: feature :pr: 347 The service layer has always has a `to_model` function that accepts data and optionally an operation name. It would return a SQLAlchemy model no matter the input you gave it. It is possible to move business logic into this `to_model` layer for populating fields on insert. (i.e. slug fields or tags, etc.). When having logic for `insert`, `update`, `delete`, and `upsert`, that function can be a bit overwhelcoming. Now, there are helper functions that you can use that is specific to each DML hook: * `to_model_on_create` * `to_model_on_update` * `to_model_on_delete` * `to_model_on_upsert` .. changelog:: 0.29.0 :date: 2025-01-17 .. change:: fully qualify all `datetime` module references :type: bugfix :pr: 341 All date time references are now full qualified to prevent any forward resolution issues with `from datetime import datetime` and `import datetime` .. change:: disabled `timezone` in alembic.ini :type: bugfix :pr: 344 Disabled `timezone` in alembic.ini to fix `alembic.util.exc.CommandError: Can't locate timezone: UTC` error while applying migrations Reference: https://alembic.sqlalchemy.org/en/latest/tutorial.html#editing-the-ini-file .. change:: various typing improvements for services :type: feature :pr: 342 :issue: 261 Improved typing in the service layer and adds a additional type guards. .. change:: Auto extend Flask CLI and add session integration :type: feature :pr: 111 The Advanced Alchemy alembic CLI is now auto-extended to your Flask application. The Flask extension now also has a session handling middleware for handling auto-commits. Last, but not least, there's an experimental async portal that integrates a long running asyncio loop for running async operations in Flask. Using `foo = portal.call()` you can get the result of an asynchronous function from a sync context. .. changelog:: 0.28.0 :date: 2025-01-13 .. change:: add `bind-key` option to CLI :type: feature :pr: 339 Adds a `bind-key` option to the Advance Alchemy CLI groups. When present, the Alembic configs will be injected with the corresponding key. .. changelog:: 0.27.1 :date: 2025-01-11 .. change:: correction for `3.8` and `3.9` type hints :type: bugfix :pr: 330 Makes a few corrections to type hints in examples and tests to ensure 3.8 and 3.9 support .. changelog:: 0.27.0 :date: 2025-01-11 .. change:: add `error_messages` as class level configuration :type: feature :pr: 315 Exposes ``error_messages`` as a class level configuration in the repository and service classes. .. change:: implement reusable CLI :type: feature :pr: 320 Exposes a reusable CLI for creating and updating releases. This can be used to extend any existing Click or Typer CLI. .. change:: adds additional type guard helpers :type: feature :pr: 322 Addition typing utilities to help with type checking and validation. .. changelog:: 0.26.0 :date: 2025-01-11 .. change:: `AsyncAttrs` & remove `noload` default :type: feature :pr: 305 This PR adds the `AsyncAttrs` to the default declarative bases for convenience. It also changes the `inherit_lazy_relationships == False` behavior to use `lazyload`. SQLAlchemy will be deprecating `noload` in version 2.1 .. change:: `litestar` DTO enhancements :type: feature :pr: 310 :issue: 306 The Litestar DTO has been enhanced with: - The SQLAlchemyDTOConfig's `exclude`, `include`, and `rename_fields` fields will now accept string or `InstrumentedAttributes` - DTO supports `WriteOnlyMapped` and `DynamicMapped` .. change:: add default exception handler for `litestar` integration :type: feature :pr: 308 :issue: 275 This adds a configuration option to automatically enable an exception handler for Repository errors. This will update the exception handler if you do not have one already configured for the RepositoryException class .. changelog:: 0.25.0 :date: 2025-01-11 .. change:: add max length for encrypted string :type: feature :pr: 290 The EncryptedString field now has the ability to validate against a set length. .. change:: `AsyncAttrs` & remove `noload` default :type: feature :pr: 305 This PR adds the `AsyncAttrs` to the default declarative bases for convenience. It also changes the `inherit_lazy_relationships == False` behavior to use `lazyload`. SQLAlchemy will be deprecating `noload` in version 2.1 .. changelog:: 0.24.0 :date: 2025-01-11 .. change:: remove lambda statement usage :type: feature :pr: 288 :issue: 286, 287 Removes the use of lambda statements in the repository and service classes. This has no change on the end user API, however, it should remove strange queries errors seen. .. changelog:: 0.23.0 :date: 2025-01-11 .. change:: regression caused by conditional import Sequence for pagination.py :type: bugfix :pr: 274 :issue: 272 Import Sequence directly from collections.abc Remove conditional import using TYPE_CHECKING Add noqa comment to suppress potential linter warnings .. change:: make sure `anyio` is optional :type: bugfix :pr: 278 When running standalone or with a synchronous web framework, `anyio` is not required. This PR ensures that there are no module loading failures due to the missing import. .. change:: Improved typing of `ModelDictT` :type: feature :pr: 277 Fixes typing issues in service https://github.com/litestar-org/advanced-alchemy/issues/265 This still doesn't solve the problem of UnknownVariableType if the subtypes of ModelDictT are not installed (eg: Pydantic) But at least it solves the problem of incompatibilities when they are installed .. changelog:: 0.22.0 :date: 2025-01-11 .. change:: CLI argument adjustment :type: bugfix :pr: 270 Changes the argument name so that it matches the name given in `click.option`. .. changelog:: 0.21.0 :date: 2025-01-11 .. change:: bind session to session class instead of to the session maker :type: bugfix :pr: 268 :issue: 267 binds session into sanic extension as expected in the original code, session maker was defined and then the dependency for session overwrites it with a session maker as the type. this seems non-ideal -- you can't get the session maker and when you ask for the session maker you get a session object instead, this looks at the sessionmaker `class_` property for adding the sanic dependency .. change:: correct regex mappings for duplicate and foreign key errors :type: bugfix :pr: 266 :issue: 262 Swap the variable names for DUPLICATE_KEY_REGEXES and FOREIGN_KEY_REGEXES to correctly match their contents. This ensures that the error detection for duplicate keys and foreign key violations works as intended across different database backends. .. change:: Dump all tables as JSON :type: feature :pr: 259 Adds a new CLI command to export tables to JSON. Similar to a Django dumpdata command. .. changelog:: <=0.20.0 :date: 2025-01-11 .. change:: CollectionFilter returns all entries if values is empty :type: bugfix :pr: 52 :issue: 51 Fixes #51 Bug: CollectionFilter returns all entries if values is empty a simple `1=-1` is appended into the `where` clause when an empty list is passed into the `in` statement. .. change:: better handle empty collection filters :type: bugfix :pr: 62 Currently, [this](https://github.com/cofin/litestar-fullstack/blob/main/src/app/lib/dependencies.py#L169) is how you can inject these filters in your app. When using the `id_filter` dependency on it's own, you have to have an additional not-null check before passing it into the repository. This change handles that and allows you to pass in all filters into the repository function without checking their nullability. .. change:: service `exists` should use `exists` from repository :type: bugfix :pr: 68 The service should use the repository's implementation of `exists` instead of a new one with a `count`. .. change:: do not set `id` with `item_id` when `None` :type: bugfix :pr: 67 This PR prevents the primary key from being overrwitten with `None` when using the service without the `item_id` parameter. .. change:: sqlalchemy dto for models non `Column` fields :type: bugfix :pr: 75 Examples of such fields are `ColumnClause` and `Label`, these are generated when using `sqlalchemy.func` - Fix SQLAlchemy dto generation for litestar when using models that have fields that are not instances of `Column`. Such fields arise from using expressions such as `func`.