ads2_2022/code/python/models/README.md

52 lines
1.6 KiB
Markdown
Raw Normal View History

# Models #
- In this folder the configuration files for the models of data classes (as `*.yaml`-files) are stored.
- These are interpreted by `openapi` to generate the classes for the source code during the `build` phase of the programme.
- Once the models are generated, the app configurations are read at run time and interpreted as these models.
- Generating instead of manually encoding the classes is safer/more stable as it allows for validation.
## Folder structure ##
As per the 3rd point, the `yaml`-file/s for the models (schemata)
and the `yaml`-file/s for actual configuration values for the app
are to be kept separately.
All models are to be stored in [./models](../models/),
whereas configuration files are to be stored in [./assets](../assets/).
Note that the generated python files are not stored in the repository.
When deployed on a server, these are generated as part of the `build`-process.
## Developer notes ##
### Prerequisites ###
For the python source code, we currently use:
- Python: `v3.10.*`
- Modules:
- `datamodel-code-generator==0.12.0`
(This is all taken care of during the `build` process.)
### Building the models ###
To build the models before run time, use the following command + options:
```bash
datamodel-codegen
--input-file-type openapi
--encoding "UTF-8"
--disable-timestamp
--use-schema-description
--snake-case-field
--strict-nullable
--input <path/to/file.yml>
--output <path/to/file.py>
```
(cf. https://pydantic-docs.helpmanual.io/datamodel_code_generator/).
Alternatively, call:
```
just build
```