| Name | zninit JSON |
| Version |
0.1.11
JSON |
| download |
| home_page | |
| Summary | Descriptor based dataclass implementation |
| upload_time | 2023-09-21 15:50:57 |
| maintainer | |
| docs_url | None |
| author | zincwarecode |
| requires_python | >=3.8,<4.0 |
| license | Apache-2.0 |
| keywords |
dataclass
descriptor
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
[](https://coveralls.io/github/zincware/ZnInit?branch=main)
[](https://badge.fury.io/py/zninit)
[](https://github.com/psf/black/)
[](https://mybinder.org/v2/gh/zincware/ZnInit/HEAD)
[](https://github.com/zincware)
# ZnInit - Automatic Generation of `__init__` based on Descriptors
This package provides a base class for `dataclass` like structures with the
addition of using
[Descriptors](https://docs.python.org/3/howto/descriptor.html). The main
functionality is the automatic generation of an keyword-only`__init__` based on
selected descriptors. The descriptors can e.g. overwrite `__set__` or `__get__`
or have custom metadata associated with them. The `ZnInit` package is used by
[ZnTrack](https://github.com/zincware/ZnTrack) to enable lazy loading data from
files as well as distinguishing between different types of descriptors such as
`zn.params` or `zn.outputs`. An example can be found in the `examples`
directory.
# Example
The most simple use case is a replication of a dataclass like structure.
```python
from zninit import ZnInit, Descriptor
class Human(ZnInit):
name: str = Descriptor()
language: str = Descriptor("EN")
# This will generate the following init:
def __init__(self, *, name, language="EN"):
self.name = name
self.language = language
fabian = Human(name="Fabian")
# or
fabian = Human(name="Fabian", language="DE")
```
The benefit of using `ZnInit` comes with using descriptors. You can subclass the
`zninit.Descriptor` class and only add certain kwargs to the `__init__` defined
in `init_descriptors: list`. Furthermore, a `post_init` method is available to
run code immediately after initializing the class.
```python
from zninit import ZnInit, Descriptor
class Input(Descriptor):
"""A Parameter"""
class Metric(Descriptor):
"""An Output"""
class Human(ZnInit):
_init_descriptors_ = [Input] # only add Input descriptors to the __init__
name: str = Input()
language: str = Input("DE")
date: str = Metric() # will not appear in the __init__
def _post_init_(self):
self.date = "2022-09-16"
julian = Human(name="Julian")
print(julian) # Human(language='DE', name='Julian')
print(julian.date) # 2022-09-16
print(Input.get_dict(julian)) # {"name": "Julian", "language": "DE"}
```
One benefit of `ZnInit` is that it also allows for inheritance.
```python
from zninit import ZnInit, Descriptor
class Animal(ZnInit):
age: int = Descriptor()
class Cat(Animal):
name: str = Descriptor()
billy = Cat(age=4, name="Billy")
```
Raw data
{
"_id": null,
"home_page": "",
"name": "zninit",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8,<4.0",
"maintainer_email": "",
"keywords": "dataclass,descriptor",
"author": "zincwarecode",
"author_email": "zincwarecode@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/06/34/5a2b779e62ddb7fc42a25d7056f41bb95a0a2241268f6cb301ff367b58a6/zninit-0.1.11.tar.gz",
"platform": null,
"description": "[](https://coveralls.io/github/zincware/ZnInit?branch=main)\n\n[](https://badge.fury.io/py/zninit)\n[](https://github.com/psf/black/)\n[](https://mybinder.org/v2/gh/zincware/ZnInit/HEAD)\n[](https://github.com/zincware)\n\n# ZnInit - Automatic Generation of `__init__` based on Descriptors\n\nThis package provides a base class for `dataclass` like structures with the\naddition of using\n[Descriptors](https://docs.python.org/3/howto/descriptor.html). The main\nfunctionality is the automatic generation of an keyword-only`__init__` based on\nselected descriptors. The descriptors can e.g. overwrite `__set__` or `__get__`\nor have custom metadata associated with them. The `ZnInit` package is used by\n[ZnTrack](https://github.com/zincware/ZnTrack) to enable lazy loading data from\nfiles as well as distinguishing between different types of descriptors such as\n`zn.params` or `zn.outputs`. An example can be found in the `examples`\ndirectory.\n\n# Example\n\nThe most simple use case is a replication of a dataclass like structure.\n\n```python\nfrom zninit import ZnInit, Descriptor\n\n\nclass Human(ZnInit):\n name: str = Descriptor()\n language: str = Descriptor(\"EN\")\n\n\n# This will generate the following init:\ndef __init__(self, *, name, language=\"EN\"):\n self.name = name\n self.language = language\n\n\nfabian = Human(name=\"Fabian\")\n# or\nfabian = Human(name=\"Fabian\", language=\"DE\")\n```\n\nThe benefit of using `ZnInit` comes with using descriptors. You can subclass the\n`zninit.Descriptor` class and only add certain kwargs to the `__init__` defined\nin `init_descriptors: list`. Furthermore, a `post_init` method is available to\nrun code immediately after initializing the class.\n\n```python\nfrom zninit import ZnInit, Descriptor\n\n\nclass Input(Descriptor):\n \"\"\"A Parameter\"\"\"\n\n\nclass Metric(Descriptor):\n \"\"\"An Output\"\"\"\n\n\nclass Human(ZnInit):\n _init_descriptors_ = [Input] # only add Input descriptors to the __init__\n name: str = Input()\n language: str = Input(\"DE\")\n date: str = Metric() # will not appear in the __init__\n\n def _post_init_(self):\n self.date = \"2022-09-16\"\n\n\njulian = Human(name=\"Julian\")\nprint(julian) # Human(language='DE', name='Julian')\nprint(julian.date) # 2022-09-16\nprint(Input.get_dict(julian)) # {\"name\": \"Julian\", \"language\": \"DE\"}\n```\n\nOne benefit of `ZnInit` is that it also allows for inheritance.\n\n```python\nfrom zninit import ZnInit, Descriptor\n\nclass Animal(ZnInit):\n age: int = Descriptor()\n\nclass Cat(Animal):\n name: str = Descriptor()\n\nbilly = Cat(age=4, name=\"Billy\")\n```\n",
"bugtrack_url": null,
"license": "Apache-2.0",
"summary": "Descriptor based dataclass implementation",
"version": "0.1.11",
"project_urls": {
"repository": "https://github.com/zincware/ZnInit"
},
"split_keywords": [
"dataclass",
"descriptor"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "2e1a06bde8267e02d801d4fe04f4fb68540933ef051de4e471ee2d2b350ee65c",
"md5": "39735f00edbf91770f733ee1ab2c8da2",
"sha256": "78e5c7e1d0d50c131cf8efab58aad4a213446a1c38f38360731bf11b778e29e1"
},
"downloads": -1,
"filename": "zninit-0.1.11-py3-none-any.whl",
"has_sig": false,
"md5_digest": "39735f00edbf91770f733ee1ab2c8da2",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8,<4.0",
"size": 13668,
"upload_time": "2023-09-21T15:50:55",
"upload_time_iso_8601": "2023-09-21T15:50:55.441714Z",
"url": "https://files.pythonhosted.org/packages/2e/1a/06bde8267e02d801d4fe04f4fb68540933ef051de4e471ee2d2b350ee65c/zninit-0.1.11-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "06345a2b779e62ddb7fc42a25d7056f41bb95a0a2241268f6cb301ff367b58a6",
"md5": "901e0bc423b8494768d556f9fc114101",
"sha256": "fc14a55bd85a38f8f1411bd319cac541c2e6e16ab402f3f11a94e182b0681965"
},
"downloads": -1,
"filename": "zninit-0.1.11.tar.gz",
"has_sig": false,
"md5_digest": "901e0bc423b8494768d556f9fc114101",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8,<4.0",
"size": 13689,
"upload_time": "2023-09-21T15:50:57",
"upload_time_iso_8601": "2023-09-21T15:50:57.144828Z",
"url": "https://files.pythonhosted.org/packages/06/34/5a2b779e62ddb7fc42a25d7056f41bb95a0a2241268f6cb301ff367b58a6/zninit-0.1.11.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-09-21 15:50:57",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "zincware",
"github_project": "ZnInit",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "zninit"
}
