Build cmd in rust

## Motivation / Description
The second cpu-intensive part of the request processing is building
the cmd.

Also instead of building dicts of flags we can use a single
flags object, which also simplifies the API of the lower
commands.

I chose to still built a single flags object, but we could
explore building one flags object per meta-command, as the
flags that they support differ, and it could lead to a more
type-safe low-level implementation.

## Performance:
* Initial:
multithreaded: Overall: 110779.55 RPS / 9.03 us/req
singlethreaded: Overall: 111545.63 RPS / 8.96 us/req

* Rust only for response parsing
multithreaded: Overall: 245898.34 RPS / 4.07 us/req
singlethreaded: Overall: 246165.19 RPS / 4.06 us/req

* Now (rust also for build_cmd)
multithreaded: Overall: 319587.03 RPS / 3.13 us/req
singlethreaded: Overall: 323101.77 RPS / 3.10 us/req

Author: bisho

poetry.lock

@@ -14,7 +14,7 @@ packages = [{include = "meta_memcache", from="src"}]
 python = "^3.8"
 uhashring = "^2.1"
 marisa-trie = "^1.0.0"
-meta-memcache-socket = "^0.1.0"
+meta-memcache-socket = "0.1.1"
 
 [tool.poetry.group.extras.dependencies]
 prometheus-client = "^0.17.1"

pyproject.toml

@@ -33,17 +33,15 @@
 )
 from meta_memcache.protocol import (
     Conflict,
-    Flag,
-    IntFlag,
     Key,
     MetaCommand,
     Miss,
     NotStored,
     ServerVersion,
     ResponseFlags,
+    RequestFlags,
     SetMode,
     Success,
-    TokenFlag,
     Value,
 )
 from meta_memcache.routers.default import DefaultRouter

src/meta_memcache/__init__.py

@@ -1,4 +1,4 @@
-from typing import Callable, Iterable, Optional, Tuple
+from typing import Callable, Iterable, Optional
 
 from meta_memcache.base.base_cache_client import BaseCacheClient
 from meta_memcache.commands.high_level_commands import HighLevelCommandsMixin
@@ -25,7 +25,7 @@ def cache_client_from_servers(
         servers: Iterable[ServerAddress],
         connection_pool_factory_fn: Callable[[ServerAddress], ConnectionPool],
         serializer: Optional[BaseSerializer] = None,
-        key_encoder_fn: Callable[[Key], Tuple[bytes, bool]] = default_key_encoder,
+        key_encoder_fn: Callable[[Key], bytes] = default_key_encoder,
         raise_on_server_error: bool = True,
     ) -> CacheApi:
         executor = DefaultExecutor(
@@ -48,7 +48,7 @@ def cache_client_with_gutter_from_servers(
         gutter_ttl: int,
         connection_pool_factory_fn: Callable[[ServerAddress], ConnectionPool],
         serializer: Optional[BaseSerializer] = None,
-        key_encoder_fn: Callable[[Key], Tuple[bytes, bool]] = default_key_encoder,
+        key_encoder_fn: Callable[[Key], bytes] = default_key_encoder,
         raise_on_server_error: bool = True,
     ) -> CacheApi:
         executor = DefaultExecutor(
@@ -76,7 +76,7 @@ def ephemeral_cache_client_from_servers(
         max_ttl: int,
         connection_pool_factory_fn: Callable[[ServerAddress], ConnectionPool],
         serializer: Optional[BaseSerializer] = None,
-        key_encoder_fn: Callable[[Key], Tuple[bytes, bool]] = default_key_encoder,
+        key_encoder_fn: Callable[[Key], bytes] = default_key_encoder,
         raise_on_server_error: bool = True,
     ) -> CacheApi:
         executor = DefaultExecutor(