nebius.aio.token.exchangeable.Receiver

class documentation

class Receiver(ParentReceiver): (source)

Constructor: Receiver(requester, service, max_retries)

Receiver that performs a token exchange over gRPC.

The receiver constructs an exchange request using the provided nebius.base.service_account.service_account.TokenRequester and calls the TokenExchangeServiceClient.exchange RPC.

Parameters
requester	Object that can construct the exchange request.
service	Either a ready-made `TokenExchangeServiceClient` or an awaitable resolving to one (useful when the channel is created asynchronously).
max_retries	Maximum retry attempts for this receiver.

Method	`__init__`	Initialize the receiver.
Method	`can_retry`	Decide whether the receiver should attempt another retry.
Async Method	`_fetch`	Perform the exchange RPC and return a `Token`.
Method	`_raise_request_error`	Convert a gRPC AioRpcError into a RequestError with diagnostics.
Instance Variable	`_max_retries`	Undocumented
Instance Variable	`_requester`	Undocumented
Instance Variable	`_status`	Undocumented
Instance Variable	`_svc`	Undocumented
Instance Variable	`_trial`	Undocumented

Inherited from Receiver:

Async Method	`fetch`	Fetch a token and record it as the latest value.
Property	`latest`	Return the latest fetched token or `None`.
Instance Variable	`_latest`	The most recently fetched token, or `None` if no token has been fetched yet. Implementations may update this to support inspection or caching by callers.

def __init__(self, requester: TokenRequester, service: TokenExchangeServiceClient | Awaitable[TokenExchangeServiceClient], max_retries: int = 2): (source) ¶

Initialize the receiver.

def can_retry(self, err: Exception, options: dict[str, str] | None = None) -> bool: (source) ¶

overrides nebius.aio.token.token.Receiver.can_retry

Decide whether the receiver should attempt another retry.

The method honours the per-request OPTION_MAX_RETRIES override if present and parses it as an integer. Invalid values are logged and ignored.

async def _fetch(self, timeout: float | None = None, options: dict[str, str] | None = None) -> Token: (source) ¶

overrides nebius.aio.token.token.Receiver._fetch

Perform the exchange RPC and return a Token.

Parameters
timeout:`float \| None`	Optional RPC timeout in seconds forwarded to the underlying gRPC stub.
options:`dict[str, str] \| None`	Optional map of request-specific options (currently unused).
Returns
`Token`	A `Token` representing the exchanged access token.
Raises
`nebius.aio.service_error.RequestError`	when the RPC returns a service error.
`UnsupportedResponseError`	when the RPC returns an unexpected payload type.
`UnsupportedTokenTypeError`	when the received token type is not `"Bearer"`.

def _raise_request_error(self, err: AioRpcError): (source) ¶

Convert a gRPC AioRpcError into a RequestError with diagnostics.

The function extracts request and trace ids from the initial metadata and converts the gRPC status into the SDK's nebius.aio.service_error.RequestError type.

Parameters
err:`AioRpcError`	The original gRPC AioRpcError.
Raises
`RequestError`	always raised with enriched diagnostics.

_max_retries = (source) ¶

Undocumented

_requester = (source) ¶

Undocumented

_status = (source) ¶

Undocumented

_svc = (source) ¶

Undocumented

_trial: int = (source) ¶

Undocumented