isage-middleware 0.1.4__py3-none-any.whl → 0.1.5.1__py3-none-any.whl

sage/__init__.py

@@ -1,37 +1,4 @@
-"""
-SAGE - Streaming-Augmented Generative Execution
-"""
+# This is a namespace package
+# 下面这段话可以让 pylance 检查到 middleware，删除该文件将会导致 pylance 无法提示补全
 
-# 动态版本加载
-def _load_version():
-    """从项目根目录动态加载版本信息"""
-    from pathlib import Path
-    
-    # 获取项目根目录
-    current_file = Path(__file__).resolve()
-    root_dir = current_file.parent.parent.parent.parent
-    version_file = root_dir / "_version.py"
-    
-    # 加载版本信息
-    if version_file.exists():
-        version_globals = {}
-        with open(version_file, 'r', encoding='utf-8') as f:
-            exec(f.read(), version_globals)
-        return {
-            'version': version_globals.get('__version__', '0.1.4'),
-            'author': version_globals.get('__author__', 'SAGE Team'),
-            'email': version_globals.get('__email__', 'shuhao_zhang@hust.edu.cn')
-        }
-    
-    # 默认值
-    return {
-        'version': '0.1.4',
-        'author': 'SAGE Team', 
-        'email': 'shuhao_zhang@hust.edu.cn'
-    }
-
-# 加载信息
-_info = _load_version()
-__version__ = _info['version']
-__author__ = _info['author']
-__email__ = _info['email']
+__path__ = __import__("pkgutil").extend_path(__path__, __name__)

sage/middleware/__init__.py

@@ -2,36 +2,11 @@
 SAGE - Streaming-Augmented Generative Execution
 """
 
-# 动态版本加载
-def _load_version():
-    """从项目根目录动态加载版本信息"""
-    from pathlib import Path
-    
-    # 获取项目根目录
-    current_file = Path(__file__).resolve()
-    root_dir = current_file.parent.parent.parent.parent.parent
-    version_file = root_dir / "_version.py"
-    
-    # 加载版本信息
-    if version_file.exists():
-        version_globals = {}
-        with open(version_file, 'r', encoding='utf-8') as f:
-            exec(f.read(), version_globals)
-        return {
-            'version': version_globals.get('__version__', '0.1.4'),
-            'author': version_globals.get('__author__', 'SAGE Team'),
-            'email': version_globals.get('__email__', 'shuhao_zhang@hust.edu.cn')
-        }
-    
-    # 默认值
-    return {
-        'version': '0.1.4',
-        'author': 'SAGE Team', 
-        'email': 'shuhao_zhang@hust.edu.cn'
-    }
-
-# 加载信息
-_info = _load_version()
-__version__ = _info['version']
-__author__ = _info['author']
-__email__ = _info['email']
+# 直接从本包的_version模块加载版本信息
+try:
+    from sage.middleware._version import __author__, __email__, __version__
+except ImportError:
+    # 备用硬编码版本
+    __version__ = "0.1.4"
+    __author__ = "IntelliStream Team"
+    __email__ = "shuhao_zhang@hust.edu.cn"

sage/middleware/_version.py

@@ -0,0 +1,6 @@
+"""Version information for sage-middleware package."""
+
+# 独立硬编码版本
+__version__ = "0.1.5.1"
+__author__ = "IntelliStream Team"
+__email__ = "shuhao_zhang@hust.edu.cn"

sage/middleware/components/cmake/pybind11_dependency.cmake

@@ -0,0 +1,34 @@
+# Shared pybind11 dependency resolver for SAGE C++ extensions
+#
+# Usage:
+#   include("${CMAKE_CURRENT_SOURCE_DIR}/../cmake/pybind11_dependency.cmake")
+#   # afterwards the function `pybind11_add_module` and targets `pybind11::module`
+#   # will be available.
+#
+# Behaviour:
+#   1. Prefer an existing pybind11 installation discoverable via find_package.
+#   2. Fallback to FetchContent on a pinned, vetted version (v2.13.0).
+#   3. Guarded so the FetchContent path only runs once per configure step.
+
+if(NOT TARGET pybind11::module)
+    find_package(pybind11 CONFIG QUIET)
+    if(pybind11_FOUND)
+        message(STATUS "Using system pybind11 ${pybind11_VERSION} from ${pybind11_CONFIG}")
+    else()
+        include(FetchContent)
+        if(NOT DEFINED pybind11_POPULATED)
+            set(_sage_pybind11_git_tag "v2.13.0" CACHE STRING "Pinned pybind11 version for SAGE extensions")
+            FetchContent_Declare(
+                pybind11
+                GIT_REPOSITORY https://github.com/pybind/pybind11.git
+                GIT_TAG ${_sage_pybind11_git_tag}
+            )
+        endif()
+        FetchContent_MakeAvailable(pybind11)
+        message(STATUS "Fetched pybind11 ${_sage_pybind11_git_tag} for SAGE extensions")
+    endif()
+endif()
+
+if(NOT COMMAND pybind11_add_module)
+    message(FATAL_ERROR "pybind11_add_module is unavailable even after resolving pybind11 dependency")
+endif()

sage/middleware/components/cmake/sage_shared_dependencies.cmake

@@ -0,0 +1,64 @@
+# Centralized dependency setup for SAGE C++ extensions when built via the Python superbuild.
+# This file is optional: individual C++ projects can still build stand-alone by
+# falling back to their local dependency scripts when SAGE_COMMON_DEPS_FILE isn't set.
+
+cmake_minimum_required(VERSION 3.20)
+
+# --- Configuration knobs ---------------------------------------------------
+set(SAGE_PYBIND11_VERSION "2.13.0" CACHE STRING "Pinned pybind11 version for all SAGE extensions")
+option(SAGE_ENABLE_GPERFTOOLS "Build extensions with gperftools/tcmalloc" OFF)
+# Allow callers to provide a pre-installed gperftools root
+set(SAGE_GPERFTOOLS_ROOT "" CACHE PATH "Optional root path for gperftools installation")
+
+# --- pybind11 ---------------------------------------------------------------
+find_package(pybind11 ${SAGE_PYBIND11_VERSION} EXACT CONFIG QUIET)
+if(NOT pybind11_FOUND)
+    include(FetchContent)
+    if(NOT DEFINED pybind11_POPULATED)
+        FetchContent_Declare(
+            pybind11
+            GIT_REPOSITORY https://github.com/pybind/pybind11.git
+            GIT_TAG v${SAGE_PYBIND11_VERSION}
+        )
+    endif()
+    FetchContent_MakeAvailable(pybind11)
+endif()
+
+# --- gperftools -------------------------------------------------------------
+set(SAGE_GPERFTOOLS_LIBS "")
+set(SAGE_GPERFTOOLS_INCLUDE "")
+if(SAGE_ENABLE_GPERFTOOLS)
+    if(SAGE_GPERFTOOLS_ROOT)
+        list(APPEND CMAKE_PREFIX_PATH ${SAGE_GPERFTOOLS_ROOT})
+    endif()
+    find_package(gperftools QUIET)
+    if(gperftools_FOUND)
+        set(SAGE_GPERFTOOLS_LIBS gperftools::profiler gperftools::tcmalloc)
+        if(TARGET gperftools::profiler)
+            get_target_property(_tmp_include gperftools::profiler INTERFACE_INCLUDE_DIRECTORIES)
+            set(SAGE_GPERFTOOLS_INCLUDE ${_tmp_include})
+        endif()
+    else()
+        include(FetchContent)
+        if(NOT TARGET gperftools::profiler)
+            FetchContent_Declare(
+                gperftools
+                GIT_REPOSITORY https://github.com/gperftools/gperftools.git
+                GIT_TAG gperftools-2.15
+            )
+            FetchContent_MakeAvailable(gperftools)
+        endif()
+        set(SAGE_GPERFTOOLS_LIBS gperftools::profiler gperftools::tcmalloc)
+    endif()
+endif()
+
+# --- Common compile flags / definitions ------------------------------------
+set(SAGE_COMMON_COMPILE_DEFINITIONS
+    PYBIND11_INTERNALS_ID="sage_pybind11_shared"
+    _GLIBCXX_USE_CXX11_ABI=1
+    CACHE INTERNAL "")
+
+set(SAGE_COMMON_COMPILE_OPTIONS
+    -fvisibility=hidden
+    -fvisibility-inlines-hidden
+    CACHE INTERNAL "")

sage/middleware/components/extensions_compat.py

@@ -0,0 +1,102 @@
+"""
+SAGE Middleware Components - 运行时兼容性检测
+"""
+
+import warnings
+
+# 尝试导入C++扩展，失败时使用纯Python实现
+_SAGE_DB_AVAILABLE = False
+_SAGE_FLOW_AVAILABLE = False
+
+try:
+    from sage.middleware.components.sage_db.python import _sage_db
+
+    _SAGE_DB_AVAILABLE = True
+except ImportError as e:
+    _sage_db = None
+    warnings.warn(
+        f"SAGE DB C++扩展不可用，某些高性能功能将受限。错误: {e}\n"
+        "安装完整版本：pip install --force-reinstall isage-middleware",
+        UserWarning,
+    )
+
+try:
+    from sage.middleware.components.sage_flow.python import _sage_flow
+    from sage.middleware.components.sage_flow.python.sage_flow import *  # noqa: F401, F403
+
+    _SAGE_FLOW_AVAILABLE = True
+except ImportError as e:
+    _sage_flow = None
+    warnings.warn(
+        f"SAGE Flow C++扩展不可用，流处理功能将受限。错误: {e}\n"
+        "安装完整版本：pip install --force-reinstall isage-middleware",
+        UserWarning,
+    )
+
+
+def is_sage_db_available() -> bool:
+    """检查SAGE DB扩展是否可用"""
+    return _SAGE_DB_AVAILABLE
+
+
+def is_sage_flow_available() -> bool:
+    """检查SAGE Flow扩展是否可用"""
+    return _SAGE_FLOW_AVAILABLE
+
+
+def get_extension_status() -> dict:
+    """获取所有扩展的状态"""
+    return {
+        "sage_db": _SAGE_DB_AVAILABLE,
+        "sage_flow": _SAGE_FLOW_AVAILABLE,
+        "total_available": sum([_SAGE_DB_AVAILABLE, _SAGE_FLOW_AVAILABLE]),
+        "total_extensions": 2,
+    }
+
+
+def check_extensions_availability() -> dict:
+    """检查扩展可用性，返回兼容格式用于CI"""
+    return {
+        "sage_db": _SAGE_DB_AVAILABLE,
+        "sage_flow": _SAGE_FLOW_AVAILABLE,
+    }
+
+
+def require_sage_db():
+    """要求SAGE DB扩展可用，否则抛出异常"""
+    if not _SAGE_DB_AVAILABLE:
+        raise ImportError(
+            "此功能需要SAGE DB C++扩展。请安装完整版本：\n"
+            "pip install --force-reinstall isage-middleware\n"
+            "或安装构建依赖后重新安装：\n"
+            "Ubuntu/Debian: sudo apt-get install build-essential cmake\n"
+            "macOS: brew install cmake"
+        )
+    return _sage_db
+
+
+def require_sage_flow():
+    """要求SAGE Flow扩展可用，否则抛出异常"""
+    if not _SAGE_FLOW_AVAILABLE:
+        raise ImportError(
+            "此功能需要SAGE Flow C++扩展。请安装完整版本：\n"
+            "pip install --force-reinstall isage-middleware\n"
+            "或安装构建依赖后重新安装：\n"
+            "Ubuntu/Debian: sudo apt-get install build-essential cmake\n"
+            "macOS: brew install cmake"
+        )
+    return _sage_flow
+
+
+# 在模块导入时显示状态
+if __name__ != "__main__":
+    status = get_extension_status()
+    if status["total_available"] < status["total_extensions"]:
+        print(
+            f"ℹ️  SAGE扩展状态: {status['total_available']}/{status['total_extensions']} 可用"
+        )
+        if not _SAGE_DB_AVAILABLE:
+            print("  ❌ SAGE DB: C++扩展不可用")
+        if not _SAGE_FLOW_AVAILABLE:
+            print("  ❌ SAGE Flow: C++扩展不可用")
+        print("  💡 提示: 安装构建依赖后重新安装可启用完整功能")

sage/middleware/components/neuromem/__init__.py

@@ -2,36 +2,16 @@
 SAGE - Streaming-Augmented Generative Execution
 """
 
-# 动态版本加载
-def _load_version():
-    """从项目根目录动态加载版本信息"""
-    from pathlib import Path
-    
-    # 获取项目根目录
-    current_file = Path(__file__).resolve()
-    root_dir = current_file.parent.parent.parent.parent.parent.parent.parent
-    version_file = root_dir / "_version.py"
-    
-    # 加载版本信息
-    if version_file.exists():
-        version_globals = {}
-        with open(version_file, 'r', encoding='utf-8') as f:
-            exec(f.read(), version_globals)
-        return {
-            'version': version_globals.get('__version__', '0.1.4'),
-            'author': version_globals.get('__author__', 'SAGE Team'),
-            'email': version_globals.get('__email__', 'shuhao_zhang@hust.edu.cn')
-        }
-    
-    # 默认值
-    return {
-        'version': '0.1.4',
-        'author': 'SAGE Team', 
-        'email': 'shuhao_zhang@hust.edu.cn'
-    }
+# 直接从本包的_version模块加载版本信息
+try:
+    from sage.middleware._version import __author__, __email__, __version__
+except ImportError:
+    # 备用硬编码版本
+    __version__ = "0.1.4"
+    __author__ = "IntelliStream Team"
+    __email__ = "shuhao_zhang@hust.edu.cn"
 
-# 加载信息
-_info = _load_version()
-__version__ = _info['version']
-__author__ = _info['author']
-__email__ = _info['email']
+# 导入主要组件
+from sage.middleware.components.neuromem.micro_service.neuromem_vdb import NeuroMemVDB
+
+__all__ = ["NeuroMemVDB"]

sage/middleware/components/neuromem/memory_collection/__init__.py

@@ -2,36 +2,11 @@
 SAGE - Streaming-Augmented Generative Execution
 """
 
-# 动态版本加载
-def _load_version():
-    """从项目根目录动态加载版本信息"""
-    from pathlib import Path
-    
-    # 获取项目根目录
-    current_file = Path(__file__).resolve()
-    root_dir = current_file.parent.parent.parent.parent.parent.parent.parent.parent
-    version_file = root_dir / "_version.py"
-    
-    # 加载版本信息
-    if version_file.exists():
-        version_globals = {}
-        with open(version_file, 'r', encoding='utf-8') as f:
-            exec(f.read(), version_globals)
-        return {
-            'version': version_globals.get('__version__', '0.1.4'),
-            'author': version_globals.get('__author__', 'SAGE Team'),
-            'email': version_globals.get('__email__', 'shuhao_zhang@hust.edu.cn')
-        }
-    
-    # 默认值
-    return {
-        'version': '0.1.4',
-        'author': 'SAGE Team', 
-        'email': 'shuhao_zhang@hust.edu.cn'
-    }
-
-# 加载信息
-_info = _load_version()
-__version__ = _info['version']
-__author__ = _info['author']
-__email__ = _info['email']
+# 直接从本包的_version模块加载版本信息
+try:
+    from sage.middleware._version import __author__, __email__, __version__
+except ImportError:
+    # 备用硬编码版本
+    __version__ = "0.1.4"
+    __author__ = "IntelliStream Team"
+    __email__ = "shuhao_zhang@hust.edu.cn"

sage/middleware/components/neuromem/memory_collection/base_collection.py

@@ -1,24 +1,23 @@
 # file: sage.middleware.services.neuromem./memory_collection/base_collection.py
 # python -m sage.core.sage.middleware.services.neuromem.memory_collection.base_collection
 
-import os
 import hashlib
-from dotenv import load_dotenv
-from typing import Dict, Optional, Callable, Any, List
-from sage.middleware.components.neuromem.utils.path_utils import get_default_data_dir
-from sage.middleware.components.neuromem.storage_engine.metadata_storage import MetadataStorage
+from typing import Any, Callable, Dict, List, Optional
+
+from sage.middleware.components.neuromem.storage_engine.metadata_storage import (
+    MetadataStorage,
+)
 from sage.middleware.components.neuromem.storage_engine.text_storage import TextStorage
 
 # from sage.middleware.services.neuromem..storage_engine.text_storage import TextStorage
 # from sage.middleware.services.neuromem..storage_engine.metadata_storage import MetadataStorage
 
 
-
-
 # 加载工程根目录下 sage/.env 配置
 # Load configuration from .env file under the sage directory
 # load_dotenv(dotenv_path=os.path.join(os.path.dirname(__file__), '../../../.env'))
 
+
 class BaseMemoryCollection:
     """
     Base memory collection with support for raw text and metadata management.
@@ -41,7 +40,7 @@ class BaseMemoryCollection:
         self,
         ids: List[str],
         metadata_filter_func: Optional[Callable[[Dict[str, Any]], bool]] = None,
-        **metadata_conditions
+        **metadata_conditions,
     ) -> List[str]:
         """
         Filter given IDs based on metadata filter rag or exact match conditions.
@@ -89,29 +88,37 @@ class BaseMemoryCollection:
             for field_name in metadata.keys():
                 if not self.metadata_storage.has_field(field_name):
                     self.metadata_storage.add_field(field_name)
-            
+
             self.metadata_storage.store(stable_id, metadata)
 
         return stable_id
-    
+
     def retrieve(
         self,
         with_metadata: bool = False,
         metadata_filter_func: Optional[Callable[[Dict[str, Any]], bool]] = None,
-        **metadata_conditions
+        **metadata_conditions,
     ):
         """
         Retrieve raw texts optionally filtered by metadata.
         根据元数据（条件或函数）检索原始文本。
         """
         all_ids = self.get_all_ids()
-        matched_ids = self.filter_ids(all_ids, metadata_filter_func, **metadata_conditions)
+        matched_ids = self.filter_ids(
+            all_ids, metadata_filter_func, **metadata_conditions
+        )
         # return [self.text_storage.get(i) for i in matched_ids]
         if with_metadata:
-            return [{"text": self.text_storage.get(i), "metadata": self.metadata_storage.get(i)} for i in matched_ids]
+            return [
+                {
+                    "text": self.text_storage.get(i),
+                    "metadata": self.metadata_storage.get(i),
+                }
+                for i in matched_ids
+            ]
         else:
             return [self.text_storage.get(i) for i in matched_ids]
-        
+
     def clear(self):
         """
         Clear all stored text and metadata.
@@ -119,12 +126,14 @@ class BaseMemoryCollection:
         """
         self.text_storage.clear()
         self.metadata_storage.clear()
-    
+
+
 if __name__ == "__main__":
 
     def basetest():
         import time
         from datetime import datetime
+
         col = BaseMemoryCollection("demo")
         col.add_metadata_field("source")
         col.add_metadata_field("lang")
@@ -132,9 +141,18 @@ if __name__ == "__main__":
 
         # 添加带时间戳的数据
         current_time = time.time()
-        col.insert("hello world", {"source": "user", "lang": "en", "timestamp": current_time - 3600})  # 1小时前
-        col.insert("你好，世界", {"source": "user", "lang": "zh", "timestamp": current_time - 1800})  # 30分钟前
-        col.insert("bonjour le monde", {"source": "web", "lang": "fr", "timestamp": current_time})  # 现在
+        col.insert(
+            "hello world",
+            {"source": "user", "lang": "en", "timestamp": current_time - 3600},
+        )  # 1小时前
+        col.insert(
+            "你好，世界",
+            {"source": "user", "lang": "zh", "timestamp": current_time - 1800},
+        )  # 30分钟前
+        col.insert(
+            "bonjour le monde",
+            {"source": "web", "lang": "fr", "timestamp": current_time},
+        )  # 现在
 
         print("=== Filter by keyword ===")
         res1 = col.retrieve(source="user")
@@ -142,26 +160,39 @@ if __name__ == "__main__":
             print(r)
 
         print("\n=== Filter by custom rag (language) ===")
-        res2 = col.retrieve(metadata_filter_func=lambda m: m.get("lang") in {"zh", "fr"})
+        res2 = col.retrieve(
+            metadata_filter_func=lambda m: m.get("lang") in {"zh", "fr"}
+        )
         for r in res2:
             print(r)
 
         print(f"\nCurrent time: {datetime.fromtimestamp(current_time)}")
-        
+
         print("\n=== Filter by timestamp (last 45 minutes) ===")
         time_threshold = current_time - 2700  # 45分钟前
-        matched_ids = col.filter_ids(col.get_all_ids(), metadata_filter_func=lambda m: m.get("timestamp", 0) > time_threshold)
+        matched_ids = col.filter_ids(
+            col.get_all_ids(),
+            metadata_filter_func=lambda m: m.get("timestamp", 0) > time_threshold,
+        )
         for item_id in matched_ids:
             text = col.text_storage.get(item_id)
             metadata = col.metadata_storage.get(item_id)
-            print(f"{text} (timestamp: {datetime.fromtimestamp(metadata['timestamp']).strftime('%Y-%m-%d %H:%M:%S')})")
+            print(
+                f"{text} (timestamp: {datetime.fromtimestamp(metadata['timestamp']).strftime('%Y-%m-%d %H:%M:%S')})"
+            )
 
         print("\n=== Filter by timestamp range (30-60 minutes ago) ===")
         start_time = current_time - 3600  # 1小时前
-        end_time = current_time - 1800    # 30分钟前
-        matched_ids = col.filter_ids(col.get_all_ids(), metadata_filter_func=lambda m: start_time <= m.get("timestamp", 0) <= end_time)
+        end_time = current_time - 1800  # 30分钟前
+        matched_ids = col.filter_ids(
+            col.get_all_ids(),
+            metadata_filter_func=lambda m: start_time
+            <= m.get("timestamp", 0)
+            <= end_time,
+        )
         for item_id in matched_ids:
             text = col.text_storage.get(item_id)
             metadata = col.metadata_storage.get(item_id)
-            print(f"{text} (timestamp: {datetime.fromtimestamp(metadata['timestamp']).strftime('%Y-%m-%d %H:%M:%S')})")
-
+            print(
+                f"{text} (timestamp: {datetime.fromtimestamp(metadata['timestamp']).strftime('%Y-%m-%d %H:%M:%S')})"
+            )

sage/middleware/components/neuromem/memory_collection/graph_collection.py

@@ -1,11 +1,12 @@
 # file sage/core/sage.middleware.services.neuromem./memory_collection/graph_collection.py
-from sage.middleware.components.neuromem.memory_collection.base_collection import BaseMemoryCollection
-from sage.middleware.components.neuromem.utils.path_utils import get_default_data_dir
+from sage.middleware.components.neuromem.memory_collection.base_collection import (
+    BaseMemoryCollection,
+)
+
 
 class GraphMemoryCollection(BaseMemoryCollection):
     pass
 
     @classmethod
-    def load(cls,  a, b):
+    def load(cls, a, b):
         pass
-