Releases: collerek/ormar
Releases · collerek/ormar
Fix saving string pks
Fix issue #46 , also added refresh after save() if any field has server_default set and field not set (so it will be populated by server default
Add upsert, save_related, allow dict in fields and exclude_fields, optimisation and cleanup etc.
Add upsert, save_related, allow dict in fields and exclude_fields, optimisation and cleanup etc.
Pre-release
Pre-release
- Added save status -> you can check if model is saved with
ModelInstance.saved
property- Model is saved after
save/update/load/upsert
method on model - Model is saved after
create/get/first/all/get_or_create/update_or_create
method - Model is saved when passed to
bulk_update
andbulk_create
- Model is saved after adding/removing
ManyToMany
related objects (through model instance auto saved/deleted) - Model is not saved after change of any own field (including pk as
Model.pk
alias) - Model is not saved after adding/removing
ForeignKey
related object (fk column not saved) - Model is not saved after instantation with
__init__
(w/oQuerySet.create
or before callingsave
)
- Model is saved after
- Added
Model.upsert(**kwargs)
that performssave()
if pk not set otherwiseupdate(**kwargs)
- Added
Model.save_related(follow=False)
that iterates all related objects in all relations and checks if they are saved. If not it callsupsert()
on each of them. - Breaking: added raising exceptions if
add
-ing/remove
-ing not saved (pk is None) models toManyToMany
relation - Allow passing dictionaries and sets to fields and exclude_fields
- Auto translate str and lists to dicts for fields and exclude_fields
- Breaking: passing nested models to fields and exclude_fields is now by related ForeignKey name and not by target model name
- Performance optimizations - in modelproxy, newbasemodel - > less queries, some properties are cached on models
- Cleanup of unused relations code
- Optional performance dependency orjson added (strongly recommended)
- Updated docs
Add exclude_fields, fix order_by
- add exclude_fields() method to exclude fields from sql
- refactor column names setting (aliases)
- fix ordering by for column with aliases
- additional tests for fields and exclude_fields
- update docs
Include property methods as fields in dict() and json()
Merge pull request #34 from collerek/include_properties_in_dict_and_json Include properties in dict and json
Allow returning related models with only pk set through fastapi
Merge pull request #32 from collerek/allow_pk_models_fastapi allow in fastapi to return related models with only pk populated
Add order_by to QuerySet
Add order_by to QuerySet
Update docs
Provide support for proper type hints - change of model definition (non breaking)
Pre-release
- Changed notation in Model definition -> now use name = ormar.Field() not name: ormar.Field()
- Note that old notation is still supported but deprecated and will not play nice with static checkers like mypy and pydantic pycharm plugin
- Type hint docs and test
- Use mypy for tests also not, only ormar package
- Fix scale and precision translation with max_digits and decimal_places pydantic Decimal field
- Update docs - add best practices for dependencies
- Refactor metaclass and model_fields to play nice with type hints
- Add mypy and pydantic plugin to docs
- Expand the docs on ManyToMany relation
Fix server_default field parameter and update docs
See details in #26
Fix postgresql check to avoid exceptions with drivers not installed
Fix issue #23
Fix json schema generation
- Fix json schema generation as of #19
- Fix for not initialized ManyToMany relations in fastapi copies of ormar.Models
- Update docs in regard of fastapi use
- Add tests to verify fastapi/docs proper generation
- Modify schema so that ManyToMany fields show list of nested models in fastapi schema