diff options
| author | Chayim I. Kirshen <c@kirshen.com> | 2022-11-29 11:09:33 +0200 |
|---|---|---|
| committer | Chayim I. Kirshen <c@kirshen.com> | 2022-11-29 11:09:33 +0200 |
| commit | 596244fbc4828ad4848df470ab0e16f4663b903f (patch) | |
| tree | 864ca5dbb46303ce8f77ebc109582079cc8df18b /redis/parsers | |
| parent | f492f85af723b277fa1ec99f01e6612e5d567b3f (diff) | |
| download | redis-py-596244fbc4828ad4848df470ab0e16f4663b903f.tar.gz | |
Reorganizing the code, in preparation for RESP3
Diffstat (limited to 'redis/parsers')
| -rw-r--r-- | redis/parsers/__init__.py | 16 | ||||
| -rw-r--r-- | redis/parsers/commands.py | 259 | ||||
| -rw-r--r-- | redis/parsers/encoders.py | 44 | ||||
| -rw-r--r-- | redis/parsers/hiredis.py | 217 | ||||
| -rw-r--r-- | redis/parsers/protocol.py | 107 | ||||
| -rw-r--r-- | redis/parsers/resp2.py | 203 | ||||
| -rw-r--r-- | redis/parsers/socket.py | 140 |
7 files changed, 986 insertions, 0 deletions
diff --git a/redis/parsers/__init__.py b/redis/parsers/__init__.py new file mode 100644 index 0000000..c6d950d --- /dev/null +++ b/redis/parsers/__init__.py @@ -0,0 +1,16 @@ +from .commands import AsyncCommandsParser, CommandsParser +from .encoders import Encoder +from .hiredis import AsyncHiredisParser, HiredisParser +from .protocol import BaseParser +from .resp2 import AsyncRESP2Parser, RESP2Parser + +__all__ = [ + "AsyncCommandsParser", + "AsyncHiredisParser", + "AsyncRESP2Parser", + "CommandsParser", + "Encoder", + "BaseParser", + "HiredisParser", + "RESP2Parser", +] diff --git a/redis/parsers/commands.py b/redis/parsers/commands.py new file mode 100644 index 0000000..7981a52 --- /dev/null +++ b/redis/parsers/commands.py @@ -0,0 +1,259 @@ +from typing import TYPE_CHECKING, Any, Dict, Optional, Tuple, Union + +from redis.exceptions import RedisError, ResponseError +from redis.utils import str_if_bytes + +if TYPE_CHECKING: + from redis.asyncio.cluster import ClusterNode + + +class AsyncCommandsParser: + """ + Parses Redis commands to get command keys. + + COMMAND output is used to determine key locations. + Commands that do not have a predefined key location are flagged with 'movablekeys', + and these commands' keys are determined by the command 'COMMAND GETKEYS'. + + NOTE: Due to a bug in redis<7.0, this does not work properly + for EVAL or EVALSHA when the `numkeys` arg is 0. + - issue: https://github.com/redis/redis/issues/9493 + - fix: https://github.com/redis/redis/pull/9733 + + So, don't use this with EVAL or EVALSHA. + """ + + __slots__ = ("commands", "node") + + def __init__(self) -> None: + self.commands: Dict[str, Union[int, Dict[str, Any]]] = {} + + async def initialize(self, node: Optional["ClusterNode"] = None) -> None: + if node: + self.node = node + + commands = await self.node.execute_command("COMMAND") + for cmd, command in commands.items(): + if "movablekeys" in command["flags"]: + commands[cmd] = -1 + elif command["first_key_pos"] == 0 and command["last_key_pos"] == 0: + commands[cmd] = 0 + elif command["first_key_pos"] == 1 and command["last_key_pos"] == 1: + commands[cmd] = 1 + self.commands = {cmd.upper(): command for cmd, command in commands.items()} + + # As soon as this PR is merged into Redis, we should reimplement + # our logic to use COMMAND INFO changes to determine the key positions + # https://github.com/redis/redis/pull/8324 + async def get_keys(self, *args: Any) -> Optional[Tuple[str, ...]]: + if len(args) < 2: + # The command has no keys in it + return None + + try: + command = self.commands[args[0]] + except KeyError: + # try to split the command name and to take only the main command + # e.g. 'memory' for 'memory usage' + args = args[0].split() + list(args[1:]) + cmd_name = args[0].upper() + if cmd_name not in self.commands: + # We'll try to reinitialize the commands cache, if the engine + # version has changed, the commands may not be current + await self.initialize() + if cmd_name not in self.commands: + raise RedisError( + f"{cmd_name} command doesn't exist in Redis commands" + ) + + command = self.commands[cmd_name] + + if command == 1: + return (args[1],) + if command == 0: + return None + if command == -1: + return await self._get_moveable_keys(*args) + + last_key_pos = command["last_key_pos"] + if last_key_pos < 0: + last_key_pos = len(args) + last_key_pos + return args[command["first_key_pos"] : last_key_pos + 1 : command["step_count"]] + + async def _get_moveable_keys(self, *args: Any) -> Optional[Tuple[str, ...]]: + try: + keys = await self.node.execute_command("COMMAND GETKEYS", *args) + except ResponseError as e: + message = e.__str__() + if ( + "Invalid arguments" in message + or "The command has no key arguments" in message + ): + return None + else: + raise e + return keys + + +class CommandsParser: + """ + Parses Redis commands to get command keys. + COMMAND output is used to determine key locations. + Commands that do not have a predefined key location are flagged with + 'movablekeys', and these commands' keys are determined by the command + 'COMMAND GETKEYS'. + """ + + def __init__(self, redis_connection): + self.commands = {} + self.initialize(redis_connection) + + def initialize(self, r): + commands = r.execute_command("COMMAND") + uppercase_commands = [] + for cmd in commands: + if any(x.isupper() for x in cmd): + uppercase_commands.append(cmd) + for cmd in uppercase_commands: + commands[cmd.lower()] = commands.pop(cmd) + self.commands = commands + + def parse_subcommand(self, command, **options): + cmd_dict = {} + cmd_name = str_if_bytes(command[0]) + cmd_dict["name"] = cmd_name + cmd_dict["arity"] = int(command[1]) + cmd_dict["flags"] = [str_if_bytes(flag) for flag in command[2]] + cmd_dict["first_key_pos"] = command[3] + cmd_dict["last_key_pos"] = command[4] + cmd_dict["step_count"] = command[5] + if len(command) > 7: + cmd_dict["tips"] = command[7] + cmd_dict["key_specifications"] = command[8] + cmd_dict["subcommands"] = command[9] + return cmd_dict + + # As soon as this PR is merged into Redis, we should reimplement + # our logic to use COMMAND INFO changes to determine the key positions + # https://github.com/redis/redis/pull/8324 + def get_keys(self, redis_conn, *args): + """ + Get the keys from the passed command. + + NOTE: Due to a bug in redis<7.0, this function does not work properly + for EVAL or EVALSHA when the `numkeys` arg is 0. + - issue: https://github.com/redis/redis/issues/9493 + - fix: https://github.com/redis/redis/pull/9733 + + So, don't use this function with EVAL or EVALSHA. + """ + if len(args) < 2: + # The command has no keys in it + return None + + cmd_name = args[0].lower() + if cmd_name not in self.commands: + # try to split the command name and to take only the main command, + # e.g. 'memory' for 'memory usage' + cmd_name_split = cmd_name.split() + cmd_name = cmd_name_split[0] + if cmd_name in self.commands: + # save the splitted command to args + args = cmd_name_split + list(args[1:]) + else: + # We'll try to reinitialize the commands cache, if the engine + # version has changed, the commands may not be current + self.initialize(redis_conn) + if cmd_name not in self.commands: + raise RedisError( + f"{cmd_name.upper()} command doesn't exist in Redis commands" + ) + + command = self.commands.get(cmd_name) + if "movablekeys" in command["flags"]: + keys = self._get_moveable_keys(redis_conn, *args) + elif "pubsub" in command["flags"] or command["name"] == "pubsub": + keys = self._get_pubsub_keys(*args) + else: + if ( + command["step_count"] == 0 + and command["first_key_pos"] == 0 + and command["last_key_pos"] == 0 + ): + is_subcmd = False + if "subcommands" in command: + subcmd_name = f"{cmd_name}|{args[1].lower()}" + for subcmd in command["subcommands"]: + if str_if_bytes(subcmd[0]) == subcmd_name: + command = self.parse_subcommand(subcmd) + is_subcmd = True + + # The command doesn't have keys in it + if not is_subcmd: + return None + last_key_pos = command["last_key_pos"] + if last_key_pos < 0: + last_key_pos = len(args) - abs(last_key_pos) + keys_pos = list( + range(command["first_key_pos"], last_key_pos + 1, command["step_count"]) + ) + keys = [args[pos] for pos in keys_pos] + + return keys + + def _get_moveable_keys(self, redis_conn, *args): + """ + NOTE: Due to a bug in redis<7.0, this function does not work properly + for EVAL or EVALSHA when the `numkeys` arg is 0. + - issue: https://github.com/redis/redis/issues/9493 + - fix: https://github.com/redis/redis/pull/9733 + + So, don't use this function with EVAL or EVALSHA. + """ + pieces = [] + cmd_name = args[0] + # The command name should be splitted into separate arguments, + # e.g. 'MEMORY USAGE' will be splitted into ['MEMORY', 'USAGE'] + pieces = pieces + cmd_name.split() + pieces = pieces + list(args[1:]) + try: + keys = redis_conn.execute_command("COMMAND GETKEYS", *pieces) + except ResponseError as e: + message = e.__str__() + if ( + "Invalid arguments" in message + or "The command has no key arguments" in message + ): + return None + else: + raise e + return keys + + def _get_pubsub_keys(self, *args): + """ + Get the keys from pubsub command. + Although PubSub commands have predetermined key locations, they are not + supported in the 'COMMAND's output, so the key positions are hardcoded + in this method + """ + if len(args) < 2: + # The command has no keys in it + return None + args = [str_if_bytes(arg) for arg in args] + command = args[0].upper() + keys = None + if command == "PUBSUB": + # the second argument is a part of the command name, e.g. + # ['PUBSUB', 'NUMSUB', 'foo']. + pubsub_type = args[1].upper() + if pubsub_type in ["CHANNELS", "NUMSUB"]: + keys = args[2:] + elif command in ["SUBSCRIBE", "PSUBSCRIBE", "UNSUBSCRIBE", "PUNSUBSCRIBE"]: + # format example: + # SUBSCRIBE channel [channel ...] + keys = list(args[1:]) + elif command == "PUBLISH": + # format example: + # PUBLISH channel message + keys = [args[1]] + return keys diff --git a/redis/parsers/encoders.py b/redis/parsers/encoders.py new file mode 100644 index 0000000..85fb292 --- /dev/null +++ b/redis/parsers/encoders.py @@ -0,0 +1,44 @@ +from redis.typing import EncodableT, EncodedT + +from ..exceptions import DataError + + +class Encoder: + "Encode strings to bytes-like and decode bytes-like to strings" + + def __init__(self, encoding, encoding_errors, decode_responses): + self.encoding = encoding + self.encoding_errors = encoding_errors + self.decode_responses = decode_responses + + def encode(self, value): + "Return a bytestring or bytes-like representation of the value" + if isinstance(value, (bytes, memoryview)): + return value + elif isinstance(value, (int, float)): + value = repr(value).encode() + elif isinstance(value, str): + value = value.encode(self.encoding, self.encoding_errors) + elif isinstance(value, bool): + # special case bool since it is a subclass of int + raise DataError( + "Invalid input of type: 'bool'. Convert to a " + "bytes, string, int or float first." + ) + elif not isinstance(value, str): + # a value we don't know how to deal with. throw an error + typename = type(value).__name__ + raise DataError( + f"Invalid input of type: '{typename}'. " + f"Convert to a bytes, string, int or float first." + ) + return value + + def decode(self, value, force=False): + "Return a unicode string from the bytes-like representation" + if self.decode_responses or force: + if isinstance(value, memoryview): + value = value.tobytes() + if isinstance(value, bytes): + value = value.decode(self.encoding, self.encoding_errors) + return value diff --git a/redis/parsers/hiredis.py b/redis/parsers/hiredis.py new file mode 100644 index 0000000..44d5b81 --- /dev/null +++ b/redis/parsers/hiredis.py @@ -0,0 +1,217 @@ +import asyncio +import socket +from typing import Callable, List, Optional, TypedDict, Union + +import async_timeout + +from ..exceptions import ( + AuthenticationError, + ConnectionError, + InvalidResponse, + RedisError, +) +from ..typing import EncodableT +from ..utils import HIREDIS_AVAILABLE +from .protocol import AsyncBaseParser, BaseParser +from .socket import ( + NONBLOCKING_EXCEPTION_ERROR_NUMBERS, + NONBLOCKING_EXCEPTIONS, + SENTINEL, + SERVER_CLOSED_CONNECTION_ERROR, +) + + +class _HiredisReaderArgs(TypedDict, total=False): + protocolError: Callable[[str], Exception] + replyError: Callable[[str], Exception] + encoding: Optional[str] + errors: Optional[str] + + +class HiredisParser(BaseParser): + "Parser class for connections using Hiredis" + + def __init__(self, socket_read_size): + if not HIREDIS_AVAILABLE: + raise RedisError("Hiredis is not installed") + self.socket_read_size = socket_read_size + self._buffer = bytearray(socket_read_size) + + def __del__(self): + try: + self.on_disconnect() + except Exception: + pass + + def on_connect(self, connection, **kwargs): + import hiredis + + self._sock = connection._sock + self._socket_timeout = connection.socket_timeout + kwargs = { + "protocolError": InvalidResponse, + "replyError": self.parse_error, + "errors": connection.encoder.encoding_errors, + } + + if connection.encoder.decode_responses: + kwargs["encoding"] = connection.encoder.encoding + self._reader = hiredis.Reader(**kwargs) + self._next_response = False + + def on_disconnect(self): + self._sock = None + self._reader = None + self._next_response = False + + def can_read(self, timeout): + if not self._reader: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + + if self._next_response is False: + self._next_response = self._reader.gets() + if self._next_response is False: + return self.read_from_socket(timeout=timeout, raise_on_timeout=False) + return True + + def read_from_socket(self, timeout=SENTINEL, raise_on_timeout=True): + sock = self._sock + custom_timeout = timeout is not SENTINEL + try: + if custom_timeout: + sock.settimeout(timeout) + bufflen = self._sock.recv_into(self._buffer) + if bufflen == 0: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + self._reader.feed(self._buffer, 0, bufflen) + # data was read from the socket and added to the buffer. + # return True to indicate that data was read. + return True + except socket.timeout: + if raise_on_timeout: + raise TimeoutError("Timeout reading from socket") + return False + except NONBLOCKING_EXCEPTIONS as ex: + # if we're in nonblocking mode and the recv raises a + # blocking error, simply return False indicating that + # there's no data to be read. otherwise raise the + # original exception. + allowed = NONBLOCKING_EXCEPTION_ERROR_NUMBERS.get(ex.__class__, -1) + if not raise_on_timeout and ex.errno == allowed: + return False + raise ConnectionError(f"Error while reading from socket: {ex.args}") + finally: + if custom_timeout: + sock.settimeout(self._socket_timeout) + + def read_response(self, disable_decoding=False): + if not self._reader: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + + # _next_response might be cached from a can_read() call + if self._next_response is not False: + response = self._next_response + self._next_response = False + return response + + if disable_decoding: + response = self._reader.gets(False) + else: + response = self._reader.gets() + + while response is False: + self.read_from_socket() + if disable_decoding: + response = self._reader.gets(False) + else: + response = self._reader.gets() + # if the response is a ConnectionError or the response is a list and + # the first item is a ConnectionError, raise it as something bad + # happened + if isinstance(response, (AuthenticationError, ConnectionError)): + raise response + elif ( + isinstance(response, list) + and response + and isinstance(response[0], (AuthenticationError, ConnectionError)) + ): + raise response[0] + return response + + +class AsyncHiredisParser(AsyncBaseParser): + """PAsync implementation of parser class for connections using Hiredis""" + + __slots__ = AsyncBaseParser.__slots__ + ("_reader",) + + def __init__(self, socket_read_size: int): + if not HIREDIS_AVAILABLE: + raise RedisError("Hiredis is not available.") + + import hiredis + + super().__init__(socket_read_size=socket_read_size) + self._reader: Optional[hiredis.Reader] = None + + def on_connect(self, connection: "Connection"): + import hiredis + + self._stream = connection._reader + kwargs: _HiredisReaderArgs = { + "protocolError": InvalidResponse, + "replyError": self.parse_error, + } + if connection.encoder.decode_responses: + kwargs["encoding"] = connection.encoder.encoding + kwargs["errors"] = connection.encoder.encoding_errors + + self._reader = hiredis.Reader(**kwargs) + + def on_disconnect(self): + self._stream = None + self._reader = None + + async def can_read_destructive(self): + if not self._stream or not self._reader: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + if self._reader.gets(): + return True + try: + async with async_timeout.timeout(0): + return await self.read_from_socket() + except asyncio.TimeoutError: + return False + + async def read_from_socket(self): + buffer = await self._stream.read(self._read_size) + if not buffer or not isinstance(buffer, bytes): + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) from None + self._reader.feed(buffer) + # data was read from the socket and added to the buffer. + # return True to indicate that data was read. + return True + + async def read_response( + self, disable_decoding: bool = False + ) -> Union[EncodableT, List[EncodableT]]: + if not self._stream or not self._reader: + self.on_disconnect() + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) from None + + response = self._reader.gets() + while response is False: + await self.read_from_socket() + response = self._reader.gets() + + # if the response is a ConnectionError or the response is a list and + # the first item is a ConnectionError, raise it as something bad + # happened + if isinstance(response, (AuthenticationError, ConnectionError)): + raise response + elif ( + isinstance(response, list) + and response + and isinstance(response[0], (AuthenticationError, ConnectionError)) + ): + raise response[0] + return response diff --git a/redis/parsers/protocol.py b/redis/parsers/protocol.py new file mode 100644 index 0000000..6faaddd --- /dev/null +++ b/redis/parsers/protocol.py @@ -0,0 +1,107 @@ +from abc import ABC +from asyncio import StreamReader +from typing import List, Optional, Union + +from redis.typing import EncodableT + +from ..exceptions import ( + AuthenticationError, + AuthenticationWrongNumberOfArgsError, + BusyLoadingError, + ConnectionError, + ExecAbortError, + ModuleError, + NoPermissionError, + NoScriptError, + ReadOnlyError, + ResponseError, +) + +MODULE_LOAD_ERROR = "Error loading the extension. " "Please check the server logs." +NO_SUCH_MODULE_ERROR = "Error unloading module: no such module with that name" +MODULE_UNLOAD_NOT_POSSIBLE_ERROR = "Error unloading module: operation not " "possible." +MODULE_EXPORTS_DATA_TYPES_ERROR = ( + "Error unloading module: the module " + "exports one or more module-side data " + "types, can't unload" +) +# user send an AUTH cmd to a server without authorization configured +NO_AUTH_SET_ERROR = { + # Redis >= 6.0 + "AUTH <password> called without any password " + "configured for the default user. Are you sure " + "your configuration is correct?": AuthenticationError, + # Redis < 6.0 + "Client sent AUTH, but no password is set": AuthenticationError, +} + + +class BaseParser(ABC): + + EXCEPTION_CLASSES = { + "ERR": { + "max number of clients reached": ConnectionError, + "invalid password": AuthenticationError, + # some Redis server versions report invalid command syntax + # in lowercase + "wrong number of arguments " + "for 'auth' command": AuthenticationWrongNumberOfArgsError, + # some Redis server versions report invalid command syntax + # in uppercase + "wrong number of arguments " + "for 'AUTH' command": AuthenticationWrongNumberOfArgsError, + MODULE_LOAD_ERROR: ModuleError, + MODULE_EXPORTS_DATA_TYPES_ERROR: ModuleError, + NO_SUCH_MODULE_ERROR: ModuleError, + MODULE_UNLOAD_NOT_POSSIBLE_ERROR: ModuleError, + **NO_AUTH_SET_ERROR, + }, + "WRONGPASS": AuthenticationError, + "EXECABORT": ExecAbortError, + "LOADING": BusyLoadingError, + "NOSCRIPT": NoScriptError, + "READONLY": ReadOnlyError, + "NOAUTH": AuthenticationError, + "NOPERM": NoPermissionError, + } + + def parse_error(self, response): + "Parse an error response" + error_code = response.split(" ")[0] + if error_code in self.EXCEPTION_CLASSES: + response = response[len(error_code) + 1 :] + exception_class = self.EXCEPTION_CLASSES[error_code] + if isinstance(exception_class, dict): + exception_class = exception_class.get(response, ResponseError) + return exception_class(response) + return ResponseError(response) + + +class AsyncBaseParser(BaseParser): + """Base parsing class for the python-backed async parser""" + + __slots__ = "_stream", "_read_size" + + def __init__(self, socket_read_size: int): + self._stream: Optional[StreamReader] = None + self._read_size = socket_read_size + + def __del__(self): + try: + self.on_disconnect() + except Exception: + pass + + def on_disconnect(self): + raise NotImplementedError() + + def on_connect(self, connection: "Connection"): + raise NotImplementedError() + + async def can_read_destructive(self) -> bool: + raise NotImplementedError() + + async def read_response( + self, disable_decoding: bool = False + ) -> Union[EncodableT, ResponseError, None, List[EncodableT]]: + raise NotImplementedError() diff --git a/redis/parsers/resp2.py b/redis/parsers/resp2.py new file mode 100644 index 0000000..32872ea --- /dev/null +++ b/redis/parsers/resp2.py @@ -0,0 +1,203 @@ +from asyncio import IncompleteReadError, TimeoutError +from typing import Any, Optional, Union + +import async_timeout + +from redis.typing import EncodableT + +from ..exceptions import ConnectionError, InvalidResponse, RedisError, ResponseError +from .encoders import Encoder +from .protocol import AsyncBaseParser, BaseParser +from .socket import SERVER_CLOSED_CONNECTION_ERROR, SocketBuffer + + +class RESP2Parser(BaseParser): + "Plain Python parsing class" + + def __init__(self, socket_read_size): + self.socket_read_size = socket_read_size + self.encoder = None + self._sock = None + self._buffer = None + + def __del__(self): + try: + self.on_disconnect() + except Exception: + pass + + def on_connect(self, connection): + "Called when the socket connects" + self._sock = connection._sock + self._buffer = SocketBuffer( + self._sock, self.socket_read_size, connection.socket_timeout + ) + self.encoder = connection.encoder + + def on_disconnect(self): + "Called when the socket disconnects" + self._sock = None + if self._buffer is not None: + self._buffer.close() + self._buffer = None + self.encoder = None + + def can_read(self, timeout): + return self._buffer and self._buffer.can_read(timeout) + + def read_response(self, disable_decoding=False): + raw = self._buffer.readline() + if not raw: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + + byte, response = raw[:1], raw[1:] + + if byte not in (b"-", b"+", b":", b"$", b"*"): + raise InvalidResponse(f"Protocol Error: {raw!r}") + + # server returned an error + if byte == b"-": + response = response.decode("utf-8", errors="replace") + error = self.parse_error(response) + # if the error is a ConnectionError, raise immediately so the user + # is notified + if isinstance(error, ConnectionError): + raise error + # otherwise, we're dealing with a ResponseError that might belong + # inside a pipeline response. the connection's read_response() + # and/or the pipeline's execute() will raise this error if + # necessary, so just return the exception instance here. + return error + # single value + elif byte == b"+": + pass + # int value + elif byte == b":": + response = int(response) + # bulk response + elif byte == b"$": + length = int(response) + if length == -1: + return None + response = self._buffer.read(length) + # multi-bulk response + elif byte == b"*": + length = int(response) + if length == -1: + return None + response = [ + self.read_response(disable_decoding=disable_decoding) + for i in range(length) + ] + if isinstance(response, bytes) and disable_decoding is False: + response = self.encoder.decode(response) + return response + + +class AsyncRESP2Parser(AsyncBaseParser): + """Parsing class for the RESP2 protocol""" + + __slots__ = AsyncBaseParser.__slots__ + ("encoder",) + + def __init__(self, socket_read_size: int): + super().__init__(socket_read_size) + self.encoder: Optional[Encoder] = None + + def on_connect(self, connection: "Connection"): + """Called when the stream connects""" + self._stream = connection._reader + if self._stream is None: + raise RedisError("Buffer is closed.") + + self.encoder = connection.encoder + + def on_disconnect(self): + """Called when the stream disconnects""" + if self._stream is not None: + self._stream = None + self.encoder = None + + async def can_read_destructive(self) -> bool: + if self._stream is None: + raise RedisError("Buffer is closed.") + try: + async with async_timeout.timeout(0): + return await self._stream.read(1) + except TimeoutError: + return False + + async def read_response( + self, disable_decoding: bool = False + ) -> Union[EncodableT, ResponseError, None]: + if not self._stream or not self.encoder: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + raw = await self._readline() + if not raw: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + response: Any + byte, response = raw[:1], raw[1:] + + if byte not in (b"-", b"+", b":", b"$", b"*"): + raise InvalidResponse(f"Protocol Error: {raw!r}") + + # server returned an error + if byte == b"-": + response = response.decode("utf-8", errors="replace") + error = self.parse_error(response) + # if the error is a ConnectionError, raise immediately so the user + # is notified + if isinstance(error, ConnectionError): + raise error + # otherwise, we're dealing with a ResponseError that might belong + # inside a pipeline response. the connection's read_response() + # and/or the pipeline's execute() will raise this error if + # necessary, so just return the exception instance here. + return error + # single value + elif byte == b"+": + pass + # int value + elif byte == b":": + response = int(response) + # bulk response + elif byte == b"$": + length = int(response) + if length == -1: + return None + response = await self._read(length) + # multi-bulk response + elif byte == b"*": + length = int(response) + if length == -1: + return None + response = [ + (await self.read_response(disable_decoding)) for _ in range(length) + ] + if isinstance(response, bytes) and disable_decoding is False: + response = self.encoder.decode(response) + return response + + async def _read(self, length: int) -> bytes: + """ + Read `length` bytes of data. These are assumed to be followed + by a '\r\n' terminator which is subsequently discarded. + """ + if self._stream is None: + raise RedisError("Buffer is closed.") + try: + data = await self._stream.readexactly(length + 2) + except IncompleteReadError as error: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) from error + return data[:-2] + + async def _readline(self) -> bytes: + """ + read an unknown number of bytes up to the next '\r\n' + line separator, which is discarded. + """ + if self._stream is None: + raise RedisError("Buffer is closed.") + data = await self._stream.readline() + if not data.endswith(b"\r\n"): + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + return data[:-2] diff --git a/redis/parsers/socket.py b/redis/parsers/socket.py new file mode 100644 index 0000000..b021051 --- /dev/null +++ b/redis/parsers/socket.py @@ -0,0 +1,140 @@ +import errno +import io +import socket + +from ..exceptions import ConnectionError, TimeoutError +from ..utils import SSL_AVAILABLE + +NONBLOCKING_EXCEPTION_ERROR_NUMBERS = {BlockingIOError: errno.EWOULDBLOCK} + +if SSL_AVAILABLE: + import ssl + + if hasattr(ssl, "SSLWantReadError"): + NONBLOCKING_EXCEPTION_ERROR_NUMBERS[ssl.SSLWantReadError] = 2 + NONBLOCKING_EXCEPTION_ERROR_NUMBERS[ssl.SSLWantWriteError] = 2 + else: + NONBLOCKING_EXCEPTION_ERROR_NUMBERS[ssl.SSLError] = 2 + +NONBLOCKING_EXCEPTIONS = tuple(NONBLOCKING_EXCEPTION_ERROR_NUMBERS.keys()) + +SERVER_CLOSED_CONNECTION_ERROR = "Connection closed by server." +SENTINEL = object() + + +class SocketBuffer: + def __init__(self, socket, socket_read_size, socket_timeout): + self._sock = socket + self.socket_read_size = socket_read_size + self.socket_timeout = socket_timeout + self._buffer = io.BytesIO() + # number of bytes written to the buffer from the socket + self.bytes_written = 0 + # number of bytes read from the buffer + self.bytes_read = 0 + + @property + def length(self): + return self.bytes_written - self.bytes_read + + def _read_from_socket(self, length=None, timeout=SENTINEL, raise_on_timeout=True): + sock = self._sock + socket_read_size = self.socket_read_size + buf = self._buffer + buf.seek(self.bytes_written) + marker = 0 + custom_timeout = timeout is not SENTINEL + + try: + if custom_timeout: + sock.settimeout(timeout) + while True: + data = self._sock.recv(socket_read_size) + # an empty string indicates the server shutdown the socket + if isinstance(data, bytes) and len(data) == 0: + raise ConnectionError(SERVER_CLOSED_CONNECTION_ERROR) + buf.write(data) + data_length = len(data) + self.bytes_written += data_length + marker += data_length + + if length is not None and length > marker: + continue + return True + except socket.timeout: + if raise_on_timeout: + raise TimeoutError("Timeout reading from socket") + return False + except NONBLOCKING_EXCEPTIONS as ex: + # if we're in nonblocking mode and the recv raises a + # blocking error, simply return False indicating that + # there's no data to be read. otherwise raise the + # original exception. + allowed = NONBLOCKING_EXCEPTION_ERROR_NUMBERS.get(ex.__class__, -1) + if not raise_on_timeout and ex.errno == allowed: + return False + raise ConnectionError(f"Error while reading from socket: {ex.args}") + finally: + if custom_timeout: + sock.settimeout(self.socket_timeout) + + def can_read(self, timeout): + return bool(self.length) or self._read_from_socket( + timeout=timeout, raise_on_timeout=False + ) + + def read(self, length): + length = length + 2 # make sure to read the \r\n terminator + # make sure we've read enough data from the socket + if length > self.length: + self._read_from_socket(length - self.length) + + self._buffer.seek(self.bytes_read) + data = self._buffer.read(length) + self.bytes_read += len(data) + + # purge the buffer when we've consumed it all so it doesn't + # grow forever + if self.bytes_read == self.bytes_written: + self.purge() + + return data[:-2] + + def readline(self): + buf = self._buffer + buf.seek(self.bytes_read) + data = buf.readline() + while not data.endswith(b"\r\n"): + # there's more data in the socket that we need + self._read_from_socket() + buf.seek(self.bytes_read) + data = buf.readline() + + self.bytes_read += len(data) + + # purge the buffer when we've consumed it all so it doesn't + # grow forever + if self.bytes_read == self.bytes_written: + self.purge() + + return data[:-2] + + def purge(self): + self._buffer.seek(0) + self._buffer.truncate() + self.bytes_written = 0 + self.bytes_read = 0 + + def close(self): + try: + self.purge() + self._buffer.close() + except Exception: + # issue #633 suggests the purge/close somehow raised a + # BadFileDescriptor error. Perhaps the client ran out of + # memory or something else? It's probably OK to ignore + # any error being raised from purge/close since we're + # removing the reference to the instance below. + pass + self._buffer = None + self._sock = None |
