precis-cli 0.1.3__py3-none-any.whl

app/shared/services/llm/cache/__init__.py

@@ -0,0 +1,18 @@
+# SPDX-License-Identifier: Apache-2.0
+#
+# Copyright 2026 Precis Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+@fileoverview LLM 响应缓存模块入口
+"""

app/shared/services/llm/cache/response_cache.py

@@ -0,0 +1,80 @@
+# SPDX-License-Identifier: Apache-2.0
+#
+# Copyright 2026 Precis Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""LLM 响应缓存：LRU + TTL 的线程安全缓存与缓存键生成。"""
+
+from __future__ import annotations
+
+import hashlib
+import json
+import time
+from collections import OrderedDict
+from threading import Lock
+from typing import Any
+
+from app.shared.services.llm.providers.base import ChatMessage, ChatRequest
+
+
+class ResponseCache:
+    """LLM 响应缓存（LRU + TTL，线程安全）。"""
+
+    def __init__(self, max_entries: int = 100, ttl_seconds: float = 3600):
+        self._store: OrderedDict[str, tuple[Any, float]] = OrderedDict()
+        self._max = max_entries
+        self._ttl = ttl_seconds
+        self._lock = Lock()
+
+    def get(self, key: str) -> Any | None:
+        with self._lock:
+            entry = self._store.get(key)
+            if entry is None:
+                return None
+            value, expire_at = entry
+            if time.time() > expire_at:
+                self._store.pop(key, None)
+                return None
+            self._store.move_to_end(key)
+            return value
+
+    def put(self, key: str, value: Any) -> None:
+        with self._lock:
+            self._store[key] = (value, time.time() + self._ttl)
+            self._store.move_to_end(key)
+            while len(self._store) > self._max:
+                self._store.popitem(last=False)
+
+
+def cache_key(req: ChatRequest) -> str:
+    """规范化哈希请求，作为缓存键。"""
+    payload = {
+        "model": req.model,
+        "temperature": req.temperature,
+        "messages": [_normalize_message(m) for m in req.messages],
+        "tools": req.tools,
+        "tool_choice": req.tool_choice,
+    }
+    return hashlib.sha256(json.dumps(payload, sort_keys=True, ensure_ascii=False, default=str).encode()).hexdigest()
+
+
+def _normalize_message(m: ChatMessage) -> dict:
+    """将 ChatMessage 转为可哈希 dict，去除 None 字段。"""
+    d: dict[str, Any] = {"role": m.role}
+    if m.content is not None:
+        d["content"] = m.content
+    if m.tool_calls is not None:
+        d["tool_calls"] = m.tool_calls
+    if m.tool_call_id is not None:
+        d["tool_call_id"] = m.tool_call_id
+    return d

app/shared/services/llm/chat/__init__.py

@@ -0,0 +1,35 @@
+# SPDX-License-Identifier: Apache-2.0
+#
+# Copyright 2026 Precis Team
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""
+@fileoverview LLM 对话与响应解析子包
+
+功能概述:
+- build_system_prompt: 系统提示词构建
+- ActionParser: LLM JSON 响应解析器
+
+架构设计:
+- 直接调用 Provider 的 async 接口（chat/chat_stream），不再有同步包装层
+  （历史上的 ChatLLMService + _run_async 已移除，因为它在 async 上下文里
+  会开子线程跑 asyncio.run，导致 httpx 连接池清理任务泄漏 event loop）
+"""
+
+from app.shared.services.llm.chat.chat_system_prompt import build_system_prompt
+from app.shared.services.llm.chat.response_parser import ActionParser
+
+__all__ = [
+    "ActionParser",
+    "build_system_prompt",
+]