https://github.com/esther-poniatowski/tessara
Define and manage parameters as structured objects with constraints, rules, and modular compositions.
https://github.com/esther-poniatowski/tessara
configuration modular-architecture parameters python schema type-checking validation
Last synced: 3 months ago
JSON representation
Define and manage parameters as structured objects with constraints, rules, and modular compositions.
- Host: GitHub
- URL: https://github.com/esther-poniatowski/tessara
- Owner: esther-poniatowski
- License: gpl-3.0
- Created: 2025-02-09T15:40:01.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-02-09T15:55:45.000Z (4 months ago)
- Last Synced: 2025-02-09T16:36:35.690Z (4 months ago)
- Topics: configuration, modular-architecture, parameters, python, schema, type-checking, validation
- Homepage:
- Size: 0 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Tessara
**Tessara** is a Python package for defining and managing parameters as structured objects with constraints, rules, and modular compositions. It provides a flexible framework for parameter validation, ensuring correctness, consistency, and composability in complex configurations.
## Key Features
- **Structured Parameters**: Define parameters as objects with validation constraints, through predefined and custom rules.
- **Parameter Sets**: Group parameters into modular, composable collections.
- **Sweeping Mechanism**: Define parameter sweeps for systematic exploration of configurations.
- **Hierarchical and MOdular Composition**: Merge and override parameter sets for flexibility.
- **Flexible Value Setting**: Define default values at the definition time or runtime values, possibly from configuration files.
- **Validation**: Enforce constraints on sigle parameters and across parameters, to ensure consistency and relations.
## Installation
```sh
pip install tessara
```
## Usage
### Defining Parameters
```python
from tessara.core.parameters import Param
from tessara.validation.rules import TypeRule, RangeRule
param = Param(
default=5,
rules=[
TypeRule(int),
RangeRule(gt=0, le=10)
]
)
print(param.get_value()) # Output: 5
param.set_value(7) # Valid assignment
```
### Defining Parameter Sweeps
```python
from tessara.core.parameters import SweepParam
sweep_param = SweepParam(sweep_values=[1, 2, 3, 4])
print(sweep_param.get_sweep_values()) # Output: [1, 2, 3, 4]
```
### Managing Parameter Sets
```python
from tessara.core.structures import ParameterSet
params = ParameterSet(
learning_rate=Param(default=0.01),
batch_size=Param(default=32)
)
print(params.get_value("learning_rate")) # Output: 0.01
```
### Applying Validation Rules
```python
from tessara.validation.validator import Validator
validator = Validator(params)
if validator.validate():
print("Validation passed.")
else:
print("Validation failed.")
```
## Key Components
### `parameters`
- **`Param`**: Defines an individual parameter with validation constraints.
- **`SweepParam`**: Defines a parameter representing a sweep over multiple values.
### `structures`
- **`ParameterSet`**: Manages a collection of parameters.
### `validation.rules`
- **`Rule`**: Base class for validation rules.
- **`TypeRule`**: Ensures a value matches a required type.
- **`RangeRule`**: Ensures a value is within a specified range.
- **`PatternRule`**: Checks if a value matches a regex pattern.
- **`OptionRule`**: Validates against a set of allowed options.
- **`CustomRule`**: Allows custom validation functions.
- **`MultiValueRule`**: Validates relations between multiple parameters.
### `validation.validator`
- **`Validator`**: Applies validation rules to parameters and parameter sets.
- **`Checker`**: Encapsulates a single validation process.
- **`ValidationRecorder`**: Tracks validation reports and errors.
## License
This project is licensed under the GNU License.