eth-pydantic-types 0.2.6

Pydantic Types for Ethereum

eth-pydantic-types

The types in this package are pydantic types for Ethereum inspired from eth-typing.

HexStr

When your model involves a string serializes to a hex-str, HexStr is the type to use. Examples of HexStr might be a hash.

Use HexStr in your models:

from pydantic import BaseModel
from eth_pydantic_types import HexStr, HexStr32

class TransactionData(BaseModel):
    hash_any_size: HexStr
    sized_hash: HexStr32

data = TransactionData(hash_any_size="0x123", sized_hash="0x000123")
assert isinstance(data.nonce, str)
assert isinstance(data.gas, str)

HexBytes

When your model involves bytes that serialize to a hex-str, HexBytes is the type to use. Examples of HexBytes might be a hash.

Use HexBytes in your models:

from pydantic import BaseModel
from eth_pydantic_types import HexBytes, HexBytes32

class TransactionData(BaseModel):
    hash_any_size: HexBytes
    sized_hash: HexBytes32

data = TransactionData(hash_any_size="0x123", sized_hash="0x000123")
assert isinstance(data.nonce, str)
assert isinstance(data.gas, str)

HexInt

When your model involves an integer that serializes to a hex-str, HexInt is the type to use. Examples of HexInt are transaction-type, nonce, and gas values.

Use HexInt in your models:

from pydantic import BaseModel
from eth_pydantic_types import HexInt

class TransactionData(BaseModel):
    nonce: HexInt
    gas: HexInt

data = TransactionData(nonce="0x123", gas="0x000123")
assert isinstance(data.nonce, int)
assert isinstance(data.gas, int)

Address

Use the Address class for working with checksummed-addresses. Addresses get validated and checksummed in model construction. Addresses serialize to str in the Pydantic core schema and string in the JSON schema with a binary format.

from pydantic import BaseModel
from eth_pydantic_types import Address

class Account(BaseModel):
    address: Address

# NOTE: The address ends up checksummed
#   ("0x0837207e343277CBd6c114a45EC0e9Ec56a1AD84")
account = Account(address="0x837207e343277cbd6c114a45ec0e9ec56a1ad84")

HexStr

Use hex str when you only care about un-sized hex strings. The HexStr type serializes to str in the Pydantic core schema and a string in the JSON schema with a binary format.

from eth_pydantic_types import HexStr
from pydantic import BaseModel

class Tx(BaseModel):
    data: HexStr

tx = Tx(data="0x0123")

Bip122Uri

Use BIP-122 URIs in your models by annotating with the Bip122Uri type. This type serializes to a str in the Pydantic core schema as well as a string in the JSON schema, however the individual hashes are validated.

from eth_pydantic_types import Bip122Uri
from pydantic import BaseModel

class Message(BaseModel):
    path: Bip122Uri

message = Message(
    path=(
        "blockchain://d4e56740f876aef8c010b86a40d5f56745a118d0906a34e69aec8c0db1cb8fa3"
        "/block/752820c0ad7abc1200f9ad42c4adc6fbb4bd44b5bed4667990e64565102c1ba6"
    )
)

Custom Types

You can build your own hex types by subclassing and passing kwargs to the validators. For example, here is a 32-byte hash type without the 0x prefix:

from pydantic import BaseModel
from pydantic_core.core_schema import with_info_before_validator_function, str_schema

from eth_pydantic_types import HexStr32


class MyAddress(HexStr32):
    @classmethod
    def __get_pydantic_core_schema__(cls, value, handler=None):
        str_size = cls.size * 2
        return with_info_before_validator_function(
            cls.__eth_pydantic_validate__,
            str_schema(max_length=str_size, min_length=str_size),
        )

    @classmethod
    def __eth_pydantic_validate__(cls, value, info=None, **kwargs):
        return super().__eth_pydantic_validate__(value, info=info, prefixed=False, **kwargs)


class MyModel(BaseModel):
    address: MyAddress

model = MyModel(address="0x" + "ab" * 32)

Padding

For types like HexStr or HexBytes, you can control the padding by using @field_validator().

from pydantic import BaseModel, field_validator
from eth_pydantic_types import HexStr20, HexBytes20
from eth_pydantic_types.utils import Pad

class MyModel(BaseModel):
    my_str: HexStr20
    my_bytes: HexBytes20

    @field_validator("my_str", "my_bytes", mode="before")
    @classmethod
    def validate_value(cls, value, info):
        field_type = cls.model_fields[info.field_name].annotation
        return field_type.__eth_pydantic_validate__(value, pad=Pad.RIGHT)

Else, by default, if you validate integer values, it will pad left. Other inputs pad right. This mirrors Solidity types, like bytes32, that automatically pad-right when given smaller values. Integer and address types automatically pad-left.