benzene-pydantic

Validate a Benzene handler's request with a pydantic model — the Python ecosystem's standard for parsing and validating request data — for the Benzene Python port.

Depends on benzene-core and pydantic. The core stays pydantic-free; this optional adapter is the one place the dependency lives.

pip install benzene-pydantic

from benzene.core import message
from benzene.pydantic import validated
from benzene.results import Result
from pydantic import BaseModel

class PlaceOrder(BaseModel):
    sku: str
    quantity: int = 1

@message("orders:place")          # no request_type — the raw body flows in and @validated checks it
@validated(PlaceOrder)
async def place(order: PlaceOrder) -> Result:
    return Result.created(order)  # a pydantic model returned as payload serializes on the wire

@validated(Model) validates the decoded body into Model before your handler runs. A pydantic.ValidationError becomes a validation-error Result carrying one structured error per bad field, so a malformed request never reaches your handler and never crashes the pipeline.

pydantic already knows, per failure, the message, the field it came from and the rule that rejected it, and all three cross into the RFC 9457 problem document unchanged — loc becomes field and pydantic's type becomes code:

// POST {"quantity": "not-an-int"}  ->  status "validation-error"
{
  "type": "https://benzene.app/problems/validation-error",
  "title": "Validation failed",
  "detail": "Field required, Input should be a valid integer, unable to parse string as an integer",
  "benzeneStatus": "validation-error",
  "errors": [
    { "message": "Field required", "field": "sku", "code": "missing" },
    { "message": "Input should be a valid integer, unable to parse string as an integer",
      "field": "quantity", "code": "int_parsing" }
  ]
}

That is the difference between an error a UI can attach to the right input and one it can only print. detail remains the messages joined, for a caller that logs a single line.

benzene.pydantic.validation_errors(exc) exposes the same mapping directly if you validate somewhere other than a handler boundary; format_validation_errors(exc) still returns the flat "field: message" strings.

A pydantic model returned as a success payload is serialized by benzene.core's wire mapper (model_dump(by_alias=True)). Unlike a dataclass — whose fields the mapper auto-camelCases — a pydantic model is dumped under its own field names, so give the model a camelCase alias_generator to cross the wire in the Benzene naming policy (then it matches a dataclass response).

Contributes the benzene.pydantic subpackage to the shared benzene namespace.