English

简体中文

English

简体中文

Appearance

Data Models in Airalogy ​

In Airalogy, the term “Model” refers to a data model—a Python class that inherits from pydantic.BaseModel and defines a specific data structure.

All public models can be imported through a single entry point, for example:

python
from airalogy.models import CheckValue

Built-in Models ​

CheckValue ​

A small schema used for Airalogy Checkpoints.

python
from pydantic import BaseModel

class CheckValue(BaseModel):
    checked: bool
    annotation: str
python
from airalogy.models import CheckValue   # standard import path

StepValue ​

Similar to CheckValue, but the checked flag can be None, allowing a step to remain “unchecked.”

python
from pydantic import BaseModel

class StepValue(BaseModel):
    checked: bool | None
    annotation: str
python
from airalogy.models import StepValue

How Models Differ from Built-in Types ​

Airalogy also ships a set of built-in field types—for example UserName, CurrentTime, AiralogyMarkdown, and others. These types control UI behaviour on the platform (auto-filling, special widgets, etc.) and are therefore grouped under the dedicated module airalogy.built_in_types:

python
from airalogy.built_in_types import UserName

You can freely combine built-in types with ordinary Pydantic models:

python
# model.py
from pydantic import BaseModel
from airalogy.built_in_types import UserName

class VarModel(BaseModel):
    user_name: UserName

By contrast, regular models such as CheckValue or StepValue exist purely to validate and manipulate data within Airalogy’s Python runtime—they do not influence how the recording interface is rendered.

Example: Using a Model Inside an Assigner ​

python
# assigner.py
from airalogy.assigner import AssignerResult, assigner
from airalogy.models import CheckValue

@assigner(
    assigned_fields=["a_gt_b"],
    dependent_fields=["a", "b"],   # both floats
    mode="auto",
)
def check_a_gt_b(dependent_fields: dict) -> AssignerResult:
    a = dependent_fields["a"]
    b = dependent_fields["b"]
    result = a > b
    return AssignerResult(
        assigned_fields={
            "a_gt_b": CheckValue(
                checked=result,
                annotation="a > b"
            )
        }
    )

In this snippet, CheckValue enforces a consistent structure for the data that the Assigner returns, while any UI-related fields (e.g. UserName) would be handled elsewhere in the protocol.