Function register_custom_data_class 

pub fn register_custom_data_class(data_class: &Bound<'_, PyAny>) -> PyResult<()>

Registers a custom data type (class) with the catalog registry.

Use this when you prefer to pass the class instead of a sample instance. The class must have:

• type_name_static() class method or __name__ (used as type name in storage)
• from_json(data) class method
• decode_record_batch_py(metadata, batch) class method
• Instances must have ts_event, ts_init, and encode_record_batch_py(items).

Arguments

• data_class - The custom data class (e.g. MarketTickPython or module.MarketTickData)

Errors

Returns a PyErr if the class lacks required methods or the type is already registered.

Example

import json

from nautilus_trader.model import register_custom_data_class

class MarketTickPython:
    ts_event = 0
    ts_init = 0

    def to_json(self):
        return json.dumps(self.__dict__)

    @classmethod
    def from_json(cls, data):
        instance = cls()
        instance.__dict__.update(data)
        return instance

    def encode_record_batch_py(self, items):
        raise NotImplementedError("Arrow encoding is not configured")

    @classmethod
    def decode_record_batch_py(cls, metadata, batch):
        raise NotImplementedError("Arrow decoding is not configured")

register_custom_data_class(MarketTickPython)

The Arrow methods may raise for a message-bus-only class, but must be implemented before catalog persistence is used.