python-kasa/kasa/smartprotocol.py
Steven B 7fd5c213e6
Defer module updates for less volatile modules (#1052)
Addresses stability issues on older hw device versions

 - Handles module timeout errors better by querying modules individually on errors and disabling problematic modules like Firmware that go out to the internet to get updates.
- Addresses an issue with the Led module on P100 hardware version 1.0 which appears to have a memory leak and will cause the device to crash after approximately 500 calls.
- Delays updates of modules that do not have regular changes like LightPreset and LightEffect and enables them to be updated on the next update cycle only if required values have changed.
2024-07-11 16:21:59 +01:00

481 lines
17 KiB
Python

"""Implementation of the TP-Link AES Protocol.
Based on the work of https://github.com/petretiandrea/plugp100
under compatible GNU GPL3 license.
"""
from __future__ import annotations
import asyncio
import base64
import logging
import time
import uuid
from pprint import pformat as pf
from typing import Any
from .exceptions import (
SMART_AUTHENTICATION_ERRORS,
SMART_RETRYABLE_ERRORS,
AuthenticationError,
DeviceError,
KasaException,
SmartErrorCode,
TimeoutError,
_ConnectionError,
_RetryableError,
)
from .json import dumps as json_dumps
from .protocol import BaseProtocol, BaseTransport, md5
_LOGGER = logging.getLogger(__name__)
class SmartProtocol(BaseProtocol):
"""Class for the new TPLink SMART protocol."""
BACKOFF_SECONDS_AFTER_TIMEOUT = 1
DEFAULT_MULTI_REQUEST_BATCH_SIZE = 5
def __init__(
self,
*,
transport: BaseTransport,
) -> None:
"""Create a protocol object."""
super().__init__(transport=transport)
self._terminal_uuid: str = base64.b64encode(md5(uuid.uuid4().bytes)).decode()
self._request_id_generator = SnowflakeId(1, 1)
self._query_lock = asyncio.Lock()
self._multi_request_batch_size = (
self._transport._config.batch_size or self.DEFAULT_MULTI_REQUEST_BATCH_SIZE
)
def get_smart_request(self, method, params=None) -> str:
"""Get a request message as a string."""
request = {
"method": method,
"params": params,
"requestID": self._request_id_generator.generate_id(),
"request_time_milis": round(time.time() * 1000),
"terminal_uuid": self._terminal_uuid,
}
return json_dumps(request)
async def query(self, request: str | dict, retry_count: int = 3) -> dict:
"""Query the device retrying for retry_count on failure."""
async with self._query_lock:
return await self._query(request, retry_count)
async def _query(self, request: str | dict, retry_count: int = 3) -> dict:
for retry in range(retry_count + 1):
try:
return await self._execute_query(
request, retry_count=retry, iterate_list_pages=True
)
except _ConnectionError as ex:
if retry == 0:
_LOGGER.debug(
"Device %s got a connection error, will retry %s times: %s",
self._host,
retry_count,
ex,
)
if retry >= retry_count:
_LOGGER.debug("Giving up on %s after %s retries", self._host, retry)
raise ex
continue
except AuthenticationError as ex:
await self._transport.reset()
_LOGGER.debug(
"Unable to authenticate with %s, not retrying: %s", self._host, ex
)
raise ex
except _RetryableError as ex:
if retry == 0:
_LOGGER.debug(
"Device %s got a retryable error, will retry %s times: %s",
self._host,
retry_count,
ex,
)
await self._transport.reset()
if retry >= retry_count:
_LOGGER.debug("Giving up on %s after %s retries", self._host, retry)
raise ex
await asyncio.sleep(self.BACKOFF_SECONDS_AFTER_TIMEOUT)
continue
except TimeoutError as ex:
if retry == 0:
_LOGGER.debug(
"Device %s got a timeout error, will retry %s times: %s",
self._host,
retry_count,
ex,
)
await self._transport.reset()
if retry >= retry_count:
_LOGGER.debug("Giving up on %s after %s retries", self._host, retry)
raise ex
await asyncio.sleep(self.BACKOFF_SECONDS_AFTER_TIMEOUT)
continue
except KasaException as ex:
await self._transport.reset()
_LOGGER.debug(
"Unable to query the device: %s, not retrying: %s",
self._host,
ex,
)
raise ex
# make mypy happy, this should never be reached..
raise KasaException("Query reached somehow to unreachable")
async def _execute_multiple_query(self, requests: dict, retry_count: int) -> dict:
debug_enabled = _LOGGER.isEnabledFor(logging.DEBUG)
multi_result: dict[str, Any] = {}
smart_method = "multipleRequest"
multi_requests = [
{"method": method, "params": params} for method, params in requests.items()
]
end = len(multi_requests)
# Break the requests down as there can be a size limit
step = self._multi_request_batch_size
if step == 1:
# If step is 1 do not send request batches
for request in multi_requests:
method = request["method"]
req = self.get_smart_request(method, request["params"])
resp = await self._transport.send(req)
self._handle_response_error_code(resp, method, raise_on_error=False)
multi_result[method] = resp["result"]
return multi_result
for batch_num, i in enumerate(range(0, end, step)):
requests_step = multi_requests[i : i + step]
smart_params = {"requests": requests_step}
smart_request = self.get_smart_request(smart_method, smart_params)
batch_name = f"multi-request-batch-{batch_num+1}-of-{int(end/step)+1}"
if debug_enabled:
_LOGGER.debug(
"%s %s >> %s",
self._host,
batch_name,
pf(smart_request),
)
response_step = await self._transport.send(smart_request)
if debug_enabled:
_LOGGER.debug(
"%s %s << %s",
self._host,
batch_name,
pf(response_step),
)
try:
self._handle_response_error_code(response_step, batch_name)
except DeviceError as ex:
# P100 sometimes raises JSON_DECODE_FAIL_ERROR or INTERNAL_UNKNOWN_ERROR
# on batched request so disable batching
if (
ex.error_code
in {
SmartErrorCode.JSON_DECODE_FAIL_ERROR,
SmartErrorCode.INTERNAL_UNKNOWN_ERROR,
}
and self._multi_request_batch_size != 1
):
self._multi_request_batch_size = 1
raise _RetryableError(
"JSON Decode failure, multi requests disabled"
) from ex
raise ex
responses = response_step["result"]["responses"]
for response in responses:
method = response["method"]
self._handle_response_error_code(response, method, raise_on_error=False)
result = response.get("result", None)
await self._handle_response_lists(
result, method, retry_count=retry_count
)
multi_result[method] = result
# Multi requests don't continue after errors so requery any missing
for method, params in requests.items():
if method not in multi_result:
resp = await self._transport.send(
self.get_smart_request(method, params)
)
self._handle_response_error_code(resp, method, raise_on_error=False)
multi_result[method] = resp.get("result")
return multi_result
async def _execute_query(
self, request: str | dict, *, retry_count: int, iterate_list_pages: bool = True
) -> dict:
debug_enabled = _LOGGER.isEnabledFor(logging.DEBUG)
if isinstance(request, dict):
if len(request) == 1:
smart_method = next(iter(request))
smart_params = request[smart_method]
else:
return await self._execute_multiple_query(request, retry_count)
else:
smart_method = request
smart_params = None
smart_request = self.get_smart_request(smart_method, smart_params)
if debug_enabled:
_LOGGER.debug(
"%s >> %s",
self._host,
pf(smart_request),
)
response_data = await self._transport.send(smart_request)
if debug_enabled:
_LOGGER.debug(
"%s << %s",
self._host,
pf(response_data),
)
self._handle_response_error_code(response_data, smart_method)
# Single set_ requests do not return a result
result = response_data.get("result")
if iterate_list_pages and result:
await self._handle_response_lists(
result, smart_method, retry_count=retry_count
)
return {smart_method: result}
async def _handle_response_lists(
self, response_result: dict[str, Any], method, retry_count
):
if (
response_result is None
or isinstance(response_result, SmartErrorCode)
or "start_index" not in response_result
or (list_sum := response_result.get("sum")) is None
):
return
response_list_name = next(
iter(
[
key
for key in response_result
if isinstance(response_result[key], list)
]
)
)
while (list_length := len(response_result[response_list_name])) < list_sum:
response = await self._execute_query(
{method: {"start_index": list_length}},
retry_count=retry_count,
iterate_list_pages=False,
)
next_batch = response[method]
# In case the device returns empty lists avoid infinite looping
if not next_batch[response_list_name]:
_LOGGER.error(
f"Device {self._host} returned empty "
+ f"results list for method {method}"
)
break
response_result[response_list_name].extend(next_batch[response_list_name])
def _handle_response_error_code(self, resp_dict: dict, method, raise_on_error=True):
error_code_raw = resp_dict.get("error_code")
try:
error_code = SmartErrorCode.from_int(error_code_raw)
except ValueError:
_LOGGER.warning(
"Device %s received unknown error code: %s", self._host, error_code_raw
)
error_code = SmartErrorCode.INTERNAL_UNKNOWN_ERROR
if error_code is SmartErrorCode.SUCCESS:
return
if not raise_on_error:
resp_dict["result"] = error_code
return
msg = (
f"Error querying device: {self._host}: "
+ f"{error_code.name}({error_code.value})"
+ f" for method: {method}"
)
if error_code in SMART_RETRYABLE_ERRORS:
raise _RetryableError(msg, error_code=error_code)
if error_code in SMART_AUTHENTICATION_ERRORS:
raise AuthenticationError(msg, error_code=error_code)
raise DeviceError(msg, error_code=error_code)
async def close(self) -> None:
"""Close the underlying transport."""
await self._transport.close()
class SnowflakeId:
"""Class for generating snowflake ids."""
EPOCH = 1420041600000 # Custom epoch (in milliseconds)
WORKER_ID_BITS = 5
DATA_CENTER_ID_BITS = 5
SEQUENCE_BITS = 12
MAX_WORKER_ID = (1 << WORKER_ID_BITS) - 1
MAX_DATA_CENTER_ID = (1 << DATA_CENTER_ID_BITS) - 1
SEQUENCE_MASK = (1 << SEQUENCE_BITS) - 1
def __init__(self, worker_id, data_center_id):
if worker_id > SnowflakeId.MAX_WORKER_ID or worker_id < 0:
raise ValueError(
"Worker ID can't be greater than "
+ str(SnowflakeId.MAX_WORKER_ID)
+ " or less than 0"
)
if data_center_id > SnowflakeId.MAX_DATA_CENTER_ID or data_center_id < 0:
raise ValueError(
"Data center ID can't be greater than "
+ str(SnowflakeId.MAX_DATA_CENTER_ID)
+ " or less than 0"
)
self.worker_id = worker_id
self.data_center_id = data_center_id
self.sequence = 0
self.last_timestamp = -1
def generate_id(self):
"""Generate a snowflake id."""
timestamp = self._current_millis()
if timestamp < self.last_timestamp:
raise ValueError("Clock moved backwards. Refusing to generate ID.")
if timestamp == self.last_timestamp:
# Within the same millisecond, increment the sequence number
self.sequence = (self.sequence + 1) & SnowflakeId.SEQUENCE_MASK
if self.sequence == 0:
# Sequence exceeds its bit range, wait until the next millisecond
timestamp = self._wait_next_millis(self.last_timestamp)
else:
# New millisecond, reset the sequence number
self.sequence = 0
# Update the last timestamp
self.last_timestamp = timestamp
# Generate and return the final ID
return (
(
(timestamp - SnowflakeId.EPOCH)
<< (
SnowflakeId.WORKER_ID_BITS
+ SnowflakeId.SEQUENCE_BITS
+ SnowflakeId.DATA_CENTER_ID_BITS
)
)
| (
self.data_center_id
<< (SnowflakeId.SEQUENCE_BITS + SnowflakeId.WORKER_ID_BITS)
)
| (self.worker_id << SnowflakeId.SEQUENCE_BITS)
| self.sequence
)
def _current_millis(self):
return round(time.time() * 1000)
def _wait_next_millis(self, last_timestamp):
timestamp = self._current_millis()
while timestamp <= last_timestamp:
timestamp = self._current_millis()
return timestamp
class _ChildProtocolWrapper(SmartProtocol):
"""Protocol wrapper for controlling child devices.
This is an internal class used to communicate with child devices,
and should not be used directly.
This class overrides query() method of the protocol to modify all
outgoing queries to use ``control_child`` command, and unwraps the
device responses before returning to the caller.
"""
def __init__(self, device_id: str, base_protocol: SmartProtocol):
self._device_id = device_id
self._protocol = base_protocol
self._transport = base_protocol._transport
def _get_method_and_params_for_request(self, request):
"""Return payload for wrapping.
TODO: this does not support batches and requires refactoring in the future.
"""
if isinstance(request, dict):
if len(request) == 1:
smart_method = next(iter(request))
smart_params = request[smart_method]
else:
smart_method = "multipleRequest"
requests = [
{"method": method, "params": params}
for method, params in request.items()
]
smart_params = {"requests": requests}
else:
smart_method = request
smart_params = None
return smart_method, smart_params
async def query(self, request: str | dict, retry_count: int = 3) -> dict:
"""Wrap request inside control_child envelope."""
return await self._query(request, retry_count)
async def _query(self, request: str | dict, retry_count: int = 3) -> dict:
"""Wrap request inside control_child envelope."""
method, params = self._get_method_and_params_for_request(request)
request_data = {
"method": method,
"params": params,
}
wrapped_payload = {
"control_child": {
"device_id": self._device_id,
"requestData": request_data,
}
}
response = await self._protocol.query(wrapped_payload, retry_count)
result = response.get("control_child")
# Unwrap responseData for control_child
if result and (response_data := result.get("responseData")):
result = response_data.get("result")
if result and (multi_responses := result.get("responses")):
ret_val = {}
for multi_response in multi_responses:
method = multi_response["method"]
self._handle_response_error_code(
multi_response, method, raise_on_error=False
)
ret_val[method] = multi_response.get("result")
return ret_val
self._handle_response_error_code(response_data, "control_child")
return {method: result}
async def close(self) -> None:
"""Do nothing as the parent owns the protocol."""