调整了pipeline执行顺序。现在按domain顺序读取数据源数据，防止前面domain的更新影响后面domain.

将target_project_ids替换为更通用的data_id_filter

docs/datasource_guide.md

@@ -38,7 +38,7 @@ JSONL 数据源对应实现：
 - 从目录中按 `{node_type}_N.jsonl` 规则发现并加载数据
 - 批量 create/update/delete 时，先改内存缓存，再统一刷盘
 - 支持 `read_only`，在只读模式下禁止写回
-- 支持按项目过滤（`target_project_ids`）在 load 阶段预过滤
+- 支持按项目过滤（`handler_configs.project.data_id_filter`）在 load 阶段预过滤
 
 ### 2.2 行为特点
 
@@ -53,7 +53,7 @@ JSONL 数据源对应实现：
 - `type: jsonl`
 - `jsonl_dir`: 数据目录
 - `read_only`: 是否只读
-- `target_project_ids`: 项目过滤列表
+- `handler_configs.project.data_id_filter`: 项目过滤列表
 - `handler_configs`: 按节点类型的 handler 参数
 
 ---
@@ -159,8 +159,9 @@ local_datasource:
   type: jsonl
   jsonl_dir: ${PROJECT_ROOT}/filtered_datasource/datasource/filtered
   read_only: false
-  target_project_ids: []
+   handler_configs:
-  handler_configs: {}
+      project:
+         data_id_filter: []
 ```
 
 ### 4.2 API 数据源配置示例（含限流）
@@ -181,9 +182,10 @@ remote_datasource:
   api_rate_limit_max_requests: 10
   api_rate_limit_window_seconds: 10.0
 
-  target_project_ids:
+   handler_configs:
-    - "project-id-1"
+      project:
-  handler_configs: {}
+         data_id_filter:
+            - "project-id-1"
 ```
 
 ### 4.3 配置模型位置

docs/factory_registry_refactor_plan.md

@@ -93,7 +93,6 @@
 - endpoint 创建 handler
 - endpoint 统一注入：
   - `handler_config`
-  - `target_project_ids`
   - datasource-specific runtime object（如 api client）
 
 最终让 `factory` 只负责：

docs/library_integration_guide.md

@@ -239,7 +239,7 @@ backend/
 
 - 不要在 `sync_state_machine` 包里 import backend 的 model/repository
 - 不要让 `DomainRegistry` 注册逻辑散落在 backend 业务层之外
-- 不要在 datasource 之外做额外的 project 过滤；`target_project_ids` 只应由 datasource 自己解释
+- 不要在 datasource 之外做额外的 project 过滤；项目范围应通过 `handler_configs.project.data_id_filter` 下沉到具体 handler 解释
 
 ---
 
@@ -320,7 +320,9 @@ async def run_demo_sync() -> None:
             "api_base_url": "https://example.com",
             "api_uid": "xxx",
             "api_secret": "xxx",
-            "target_project_ids": ["demo-project"],
+          "handler_configs": {
+            "project": {"data_id_filter": ["demo-project"]}
+          },
         },
     }
 

docs/runtime_config_guide.md

@@ -18,9 +18,6 @@
 - `node_types`
   - 本轮参与同步的业务类型列表。
   - 顺序会影响依赖链执行时机（当前按 node_type 串行执行 bind/create/update）。
-- `target_project_ids`
-  - 兼容字段。作为两侧 datasource 的项目白名单兜底值。
-  - 若 datasource 自身配置了 `target_project_ids`，则优先使用 datasource 级配置。
 - `local_datasource` / `remote_datasource`
   - 分别定义两侧数据来源与执行端。
 - `strategies`
@@ -37,8 +34,8 @@
 - `type=jsonl`：使用本地 JSONL 文件作为数据源。
 - `jsonl_dir`：JSONL 目录路径。
 - `read_only=false`：允许该侧执行 create/update 回写；若为 true，通常用于只读加载。
-- `target_project_ids=[]`：该侧项目白名单；为空时不进行项目筛选。
+- `handler_configs.project.data_id_filter=[]`：项目白名单；为空时不进行项目筛选。
-- `handler_configs`：预留给各业务 handler 的扩展参数。
+- `handler_configs`：按节点类型下发给各业务 handler 的扩展参数。
 
 ### 3.2 API 数据源
 
@@ -46,7 +43,7 @@
 - `api_base_url` / `api_uid` / `api_secret`：接口访问参数。
 - `api_debug`：开启后输出更多调试信息。
 - `poll_max_retries` / `poll_interval`：异步任务轮询次数与间隔。
-- `target_project_ids`：**必填**，该侧项目白名单；至少需要指定一个项目ID，以避免加载过多远程数据。
+- `handler_configs.project.data_id_filter`：建议显式配置的项目白名单；至少需要指定一个项目ID，以避免加载过多远程数据。
 - `handler_configs`：业务 handler 扩展参数。
 
 ## 4. 策略字段（Strategy）

docs/sync_flow.md

@@ -2,10 +2,12 @@
 
 ## 1. 全流程主线
 1. pipeline 装配组件并加载持久化。
-2. pipeline 调用 `strategy.bind()`。
+2. pipeline 为 datasource 绑定 collection，但不在 bootstrap 阶段全量拉取业务数据。
-3. pipeline 调用 `strategy.create()` / `strategy.update()` 生成动作节点。
+3. pipeline 在每个 `node_type` 开始执行前，单独加载该 `node_type` 的最新数据。
-4. pipeline 调用 `datasource.sync_all()` 执行动作并回写执行状态。
+4. pipeline 调用 `strategy.bind()`。
-5. pipeline 输出统计并持久化。
+5. pipeline 调用 `strategy.create()` / `strategy.update()` 生成动作节点。
+6. pipeline 调用 `datasource.sync_all()` 执行动作并回写执行状态。
+7. pipeline 输出统计并持久化。
 
 ## 2. reset 时序
 

run_profiles/preset/jsonl_to_api.default.yaml

@@ -22,10 +22,11 @@ local_datasource:
   type: jsonl
   jsonl_dir: ${PROJECT_ROOT}/filtered_datasource/datasource/filtered
   read_only: false
-  target_project_ids:
+  handler_configs:
-    - "0f9c3e22-f4bb-4803-825a-4f238eeb06d5"
+    project:
-    - "f3e02e84-e81c-4aa6-88d4-f5aa108c8227"
+      data_id_filter:
-  handler_configs: {}
+        - "0f9c3e22-f4bb-4803-825a-4f238eeb06d5"
+        - "f3e02e84-e81c-4aa6-88d4-f5aa108c8227"
 remote_datasource:
   type: api
   api_base_url: http://localhost:8000/api/third/v2
@@ -36,11 +37,12 @@ remote_datasource:
   api_rate_limit_window_seconds: 10.0
   poll_max_retries: 1
   poll_interval: 0.5
-  target_project_ids:
-  # 请替换为实际的项目ID，至少需要一个
-    - "0f9c3e22-f4bb-4803-825a-4f238eeb06d5"
-    - "f3e02e84-e81c-4aa6-88d4-f5aa108c8227"
   handler_configs:
+    project:
+      data_id_filter:
+      # 请替换为实际的项目ID，至少需要一个
+        - "0f9c3e22-f4bb-4803-825a-4f238eeb06d5"
+        - "f3e02e84-e81c-4aa6-88d4-f5aa108c8227"
     supplier: {}
 persist:
   backend: sqlite

run_profiles/preset/jsonl_to_jsonl.default.yaml

@@ -23,13 +23,11 @@ local_datasource:
   type: jsonl
   jsonl_dir: ${PROJECT_ROOT}/filtered_datasource/datasource/filtered
   read_only: false
-  target_project_ids: []
   handler_configs: {}
 remote_datasource:
   type: jsonl
   jsonl_dir: ${PROJECT_ROOT}/remote_datasource/datasource
   read_only: false
-  target_project_ids: []
   handler_configs: {}
 persist:
   backend: sqlite

scripts/auto_test_domains/base.py

@@ -292,8 +292,8 @@ def build_temp_run_profile_for_project_auto_bind(
     project_detail_push_keys: set[str] | None = None,
 ) -> tuple[str, list[str]]:
     profile = load_yaml_config(config["pipeline"]["run_profile"])
-    profile.setdefault("local_datasource", {})["target_project_ids"] = [local_project_id]
+    profile.setdefault("local_datasource", {}).setdefault("handler_configs", {}).setdefault("project", {})["data_id_filter"] = [local_project_id]
-    profile.setdefault("remote_datasource", {})["target_project_ids"] = [remote_project_id]
+    profile.setdefault("remote_datasource", {}).setdefault("handler_configs", {}).setdefault("project", {})["data_id_filter"] = [remote_project_id]
 
     project_strategy = profile.setdefault("strategies", {}).setdefault("project", {}).setdefault("config", {})
     project_strategy["auto_bind"] = False

sync_state_machine/common/collection.py

@@ -514,9 +514,9 @@ class DataCollection:
 
         await self.persistence.save_nodes_bulk(self.collection_id, list(self._nodes.values()))
 
-    async def filter_by_project_ids(self, target_project_ids: List[str]) -> int:
+    async def filter_by_project_ids(self, data_id_filter: List[str]) -> int:
-        """按目标项目集合过滤当前 collection，返回删除数量。"""
+        """按目标项目 data_id 集合过滤当前 collection，返回删除数量。"""
-        target_set = set(target_project_ids or [])
+        target_set = set(data_id_filter or [])
         if not target_set:
             return 0
 

sync_state_machine/config/datasource_config.py

@@ -11,7 +11,6 @@ class BaseDataSourceConfig(BaseModel):
     model_config = ConfigDict(extra="forbid", validate_assignment=True)
 
     type: str
-    target_project_ids: List[str] = Field(default_factory=list)
     handler_configs: Dict[str, Dict[str, Any]] = Field(default_factory=dict)
 
 
@@ -33,7 +32,6 @@ class ApiDataSourceConfig(BaseDataSourceConfig):
     poll_interval: float = 0.5
     api_rate_limit_max_requests: int = 10
     api_rate_limit_window_seconds: float = 10.0
-    target_project_ids: List[str] = Field(default_factory=list)
 
 
 DataSourceConfig = Annotated[

sync_state_machine/datasource/api/handler.py

@@ -69,9 +69,13 @@ class BaseApiHandler(NodeHandler[T]):
     def set_handler_config(self, config: Dict[str, Any]) -> None:
         self.handler_config = dict(config)
 
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
+    def get_data_id_filter(self) -> List[str]:
-        """默认忽略项目过滤列表，具体子类按需覆盖。"""
+        value = self.handler_config.get("data_id_filter", [])
-        return
+        if isinstance(value, str):
+            return [value] if value else []
+        if isinstance(value, list):
+            return [str(item) for item in value if str(item)]
+        return []
     
     @property
     def node_type(self) -> str:

sync_state_machine/datasource/datasource.py

@@ -74,7 +74,6 @@ class BaseDataSource(ABC):
     def __init__(self, poll_max_retries: int = 1, poll_interval: float = 0.5):
         self._handlers: Dict[str, "NodeHandler"] = {}
         self._collection: Optional["DataCollection"] = None
-        self._target_project_ids: List[str] = []
         
         # 按节点类型统计信息
         # {"company": {"loaded": 10, "synced": 10, "success": 10, "failed": 0}, ...}
@@ -87,13 +86,6 @@ class BaseDataSource(ABC):
         self._poll_interval = max(0, poll_interval)
         self._sm_runtime: Optional[StateMachineRuntime] = None
 
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        self._target_project_ids = [str(item) for item in target_project_ids if str(item)]
-
-    @property
-    def target_project_ids(self) -> List[str]:
-        return list(self._target_project_ids)
-
     def _ensure_sm_runtime(self) -> StateMachineRuntime:
         if self._sm_runtime is None:
             cfg_path = Path(__file__).resolve().parents[1] / "config" / "node_state_machine.yaml"
@@ -379,17 +371,6 @@ class BaseDataSource(ABC):
             datasource.register_handler(ContractNodeHandler(api_client))
         """
         self._handlers[handler.node_type] = handler
-        self._apply_target_project_ids_to_handler(handler)
-
-    def set_target_project_ids(self, target_project_ids: Optional[List[str]]) -> None:
-        """设置项目过滤列表，并在注册阶段下发给所有 handler。"""
-        normalized = [str(pid) for pid in (target_project_ids or []) if str(pid)]
-        self._target_project_ids = normalized
-        for handler in self._handlers.values():
-            self._apply_target_project_ids_to_handler(handler)
-
-    def _apply_target_project_ids_to_handler(self, handler: "NodeHandler") -> None:
-        handler.set_target_project_ids(self._target_project_ids)
     
     def get_handler(self, node_type: str) -> "NodeHandler":
         """获取 Handler"""

sync_state_machine/datasource/handler.py

@@ -74,10 +74,6 @@ class NodeHandler(Protocol[T]):
         """注入 handler 级别配置。默认可忽略。"""
         ...
 
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        """注入项目过滤列表。默认可忽略。"""
-        ...
-
     def set_api_client(self, client: "ApiClient") -> None:
         """注入 API 客户端。默认可忽略。"""
         ...
@@ -360,14 +356,18 @@ class BaseNodeHandler(ABC, Generic[T]):
         """
         self._collection = collection
 
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        """默认忽略项目过滤能力。"""
-        return
-
     def set_handler_config(self, config: Dict[str, Any]) -> None:
         """默认保存 handler 级别配置，供子类按需读取。"""
         self.handler_config = dict(config)
 
+    def get_data_id_filter(self) -> List[str]:
+        value = self.handler_config.get("data_id_filter", [])
+        if isinstance(value, str):
+            return [value] if value else []
+        if isinstance(value, list):
+            return [str(item) for item in value if str(item)]
+        return []
+
     def set_api_client(self, client: "ApiClient") -> None:
         """默认忽略 API 客户端注入。"""
         return
@@ -479,6 +479,59 @@ class BaseNodeHandler(ABC, Generic[T]):
         return self.create_node(data)
 
 
+class CompatibleNodeHandler(BaseNodeHandler[T]):
+    def __init__(self, handler: BaseNodeHandler[T]):
+        super().__init__(handler.node_type, handler.node_class, handler.schema)
+        self._handler = handler
+
+    def set_collection(self, collection: "DataCollection") -> None:
+        super().set_collection(collection)
+        self._handler.set_collection(collection)
+
+    def set_handler_config(self, config: Dict[str, Any]) -> None:
+        self._handler.set_handler_config(config)
+        self.handler_config = dict(self._handler.handler_config)
+
+    def set_api_client(self, client: "ApiClient") -> None:
+        self._handler.set_api_client(client)
+
+    def set_poll_mode(self, mode: str) -> None:
+        self._handler.set_poll_mode(mode)
+
+    def get_update_fields(self, *, exclude: frozenset[str] = frozenset({"id"})) -> List[str]:
+        return self._handler.get_update_fields(exclude=exclude)
+
+    async def load(self) -> List['SyncNode']:
+        return await self._handler.load()
+
+    async def create_all(
+        self,
+        nodes: List["SyncNode"]
+    ) -> List[TaskResult]:
+        return await self._handler.create_all(nodes)
+
+    async def update_all(
+        self,
+        nodes: List["SyncNode"]
+    ) -> List[TaskResult]:
+        return await self._handler.update_all(nodes)
+
+    async def delete_all(
+        self,
+        nodes: List["SyncNode"]
+    ) -> List[TaskResult]:
+        return await self._handler.delete_all(nodes)
+
+    async def poll_tasks(self, task_ids: List[str]) -> Dict[str, TaskResult]:
+        return await self._handler.poll_tasks(task_ids)
+
+    async def validate(self, data: Dict[str, Any]) -> ValidationResult:
+        return await self._handler.validate(data)
+
+    def extract_id(self, data: Dict[str, Any]) -> str:
+        return self._handler.extract_id(data)
+
+
 class NodeHandlerRegistry:
     """节点处理器注册表"""
     

sync_state_machine/datasource/jsonl/handler.py

@@ -50,22 +50,17 @@ class BaseJsonlHandler(BaseNodeHandler):
     ):
         super().__init__(node_type, node_class, schema)
         self.datasource = datasource
-        self._target_project_ids: List[str] = []
-
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        """注入项目过滤列表的本地副本。真正的过滤来源仍然是 datasource。"""
-        self._target_project_ids = [str(pid) for pid in target_project_ids if str(pid)]
 
     def _should_skip_item_by_project_filter(self, item: Dict[str, Any], item_id: str) -> bool:
-        target_project_ids = set(self.datasource.target_project_ids)
+        data_id_filter = set(self.get_data_id_filter())
-        if not target_project_ids:
+        if not data_id_filter:
             return False
         if self.node_type == "project":
-            return item_id not in target_project_ids
+            return item_id not in data_id_filter
 
         project_id = item.get("project_id")
         if project_id:
-            return str(project_id) not in target_project_ids
+            return str(project_id) not in data_id_filter
         return False
     
     async def load(self) -> List['SyncNode']:

sync_state_machine/domain/preparation/api_handler.py

@@ -27,17 +27,8 @@ class PreparationApiHandler(BaseApiHandler):
         self._node_type = "preparation"
         self._node_class = PreparationSyncNode
         self._schema = PreparationDetail
-        self._filter_project_id = None
         self.update_schemas = [PostPreparation]
 
-    def set_filter_project_id(self, project_id: str | List[str]):
-        """设置要加载的项目ID过滤列表"""
-        self._filter_project_id = project_id
-
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        """设置目标项目ID列表"""
-        self.set_filter_project_id(target_project_ids)
-
     async def load(self) -> List[SyncNode]:
         """从 API 加载里程碑数据（需要项目上下文）"""
         from ..project.api_handler import ProjectApiHandler

sync_state_machine/domain/project/api_handler.py

@@ -71,26 +71,11 @@ class ProjectApiHandler(BaseApiHandler):
     # 类级别缓存：存储 all_info 数据
     _all_info_cache: Dict[str, Dict[str, Any]] = {}
     
-    def __init__(self, filter_project_id: str | List[str] | None = None):
+    def __init__(self):
-        """
-        初始化 Project Handler
-        
-        Args:
-            filter_project_id: 指定只加载的项目ID列表（可选，用于多项目测试）
-        """
         super().__init__()
         self._node_type = "project"
         self._node_class = ProjectSyncNode
         self._schema = ProjectResponseBase
-        if filter_project_id is None:
-            self._filter_project_id = None
-        elif isinstance(filter_project_id, str):
-            self._filter_project_id = [filter_project_id]
-        else:
-            self._filter_project_id = filter_project_id  # 保存过滤条件
-
-    def set_target_project_ids(self, target_project_ids: List[str]) -> None:
-        self._filter_project_id = list(target_project_ids) if target_project_ids else None
     
     @classmethod
     async def get_all_info(cls, client, project_id: str) -> Dict[str, Any]:
@@ -127,15 +112,12 @@ class ProjectApiHandler(BaseApiHandler):
             
             nodes = []
             
-            # 确定要加载的项目ID列表：参数 > 构造函数配置
+            project_id_filter = self.get_data_id_filter()
-            target_project_ids = self._filter_project_id or []
-            if isinstance(target_project_ids, str):
-                target_project_ids = [target_project_ids]
             
-            if target_project_ids:
+            if project_id_filter:
                 # 只加载指定项目列表
                 projects : List[ProjectResponseBase]= []
-                for project_id in target_project_ids:
+                for project_id in project_id_filter:
                     project_response = await api_get_project(self.api_client, project_id)
                     projects.append(project_response)
             else:

sync_state_machine/pipeline/factory.py

@@ -277,9 +277,6 @@ async def create_pipeline_from_config(
             project_root=_project_root(),
         )
         await _initialize_datasource(remote_ds, endpoint_name="remote")
-        local_ds.set_target_project_ids([str(pid) for pid in config.local_datasource.target_project_ids if str(pid)])
-        remote_ds.set_target_project_ids([str(pid) for pid in config.remote_datasource.target_project_ids if str(pid)])
-
         _register_handlers(config, local_ds, remote_ds, all_types)
 
         local_collection = DataCollection("local", persistence, auto_persist=False)
@@ -434,8 +431,6 @@ async def create_jsonl_to_api_pipeline(
     api_secret: str,
     # 节点类型
     node_types: List[str],
-    # 项目过滤
-    target_project_ids: Optional[List[str]] = None,
     strategy_overrides: Optional[Dict[str, Dict[str, Any]]] = None,
     # Pipeline 行为
     enable_persistence: bool = True,
@@ -446,8 +441,8 @@ async def create_jsonl_to_api_pipeline(
         poll_interval: float = 0.5,
     api_rate_limit_max_requests: int = 30,
     api_rate_limit_window_seconds: float = 10.0,
-    # Handler 自定义配置（可选，作用于 API 端）
+    local_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
-    handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
+    remote_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
     # Logger 配置
     initialize_logger: bool = True,
     logger_file_path: Optional[str] = None,
@@ -459,7 +454,7 @@ async def create_jsonl_to_api_pipeline(
             type="jsonl",
             jsonl_dir=local_jsonl_dir,
             read_only=False,
-            target_project_ids=target_project_ids or [],
+            handler_configs=local_handler_configs or {},
         ),
         remote_datasource=ApiDataSourceConfig(
             type="api",
@@ -471,8 +466,7 @@ async def create_jsonl_to_api_pipeline(
             poll_interval=poll_interval,
             api_rate_limit_max_requests=api_rate_limit_max_requests,
             api_rate_limit_window_seconds=api_rate_limit_window_seconds,
-            target_project_ids=target_project_ids or [],
+            handler_configs=remote_handler_configs or {},
-            handler_configs=handler_configs or {},
         ),
         strategies=[],
         persist=PersistConfig(
@@ -502,8 +496,6 @@ async def create_api_to_jsonl_pipeline(
     persistence_db: str,
     # 节点类型
     node_types: List[str],
-    # 项目过滤
-    target_project_ids: Optional[List[str]] = None,
     strategy_overrides: Optional[Dict[str, Dict[str, Any]]] = None,
     # Pipeline 行为
     enable_persistence: bool = True,
@@ -514,8 +506,8 @@ async def create_api_to_jsonl_pipeline(
         poll_interval: float = 0.5,
     api_rate_limit_max_requests: int = 30,
     api_rate_limit_window_seconds: float = 10.0,
-    # Handler 自定义配置（可选，作用于 API 端）
+    local_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
-    handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
+    remote_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
     # Logger 配置
     initialize_logger: bool = True,
     logger_file_path: Optional[str] = None,
@@ -533,14 +525,13 @@ async def create_api_to_jsonl_pipeline(
             poll_interval=poll_interval,
             api_rate_limit_max_requests=api_rate_limit_max_requests,
             api_rate_limit_window_seconds=api_rate_limit_window_seconds,
-            target_project_ids=target_project_ids or [],
+            handler_configs=local_handler_configs or {},
-            handler_configs=handler_configs or {},
         ),
         remote_datasource=JsonlDataSourceConfig(
             type="jsonl",
             jsonl_dir=remote_jsonl_dir,
             read_only=False,
-            target_project_ids=target_project_ids or [],
+            handler_configs=remote_handler_configs or {},
         ),
         strategies=[],
         persist=PersistConfig(
@@ -565,10 +556,11 @@ async def create_jsonl_to_jsonl_pipeline(
     remote_jsonl_dir: str,
     persistence_db: str,
     node_types: List[str],
-    target_project_ids: Optional[List[str]] = None,
     enable_persistence: bool = True,
     wipe_persistence_on_start: bool = False,
     strategy_overrides: Optional[Dict[str, Dict[str, Any]]] = None,
+    local_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
+    remote_handler_configs: Optional[Dict[str, Dict[str, Any]]] = None,
     initialize_logger: bool = True,
     logger_file_path: Optional[str] = None,
     logger_level: int = logging.INFO,
@@ -579,13 +571,13 @@ async def create_jsonl_to_jsonl_pipeline(
             type="jsonl",
             jsonl_dir=local_jsonl_dir,
             read_only=False,
-            target_project_ids=target_project_ids or [],
+            handler_configs=local_handler_configs or {},
         ),
         remote_datasource=JsonlDataSourceConfig(
             type="jsonl",
             jsonl_dir=remote_jsonl_dir,
             read_only=False,
-            target_project_ids=target_project_ids or [],
+            handler_configs=remote_handler_configs or {},
         ),
         strategies=[],
         persist=PersistConfig(

sync_state_machine/pipeline/full_sync_pipeline.py

@@ -90,6 +90,7 @@ class FullSyncPipeline:
         self._consistency_by_type: Dict[str, Dict[str, Any]] = {}
         self._logger = logger
         self._closed = False
+        self._loaded_node_types: set[str] = set()
 
     def _resolve_pipeline_runtime(self) -> StateMachineRuntime:
         if self.strategies:
@@ -119,6 +120,7 @@ class FullSyncPipeline:
             await self.close()
 
     async def phase_sync_by_node_type(self) -> None:
+        self._prepare_datasources()
         strategy_map = {s.node_type: s for s in self.strategies}
         for node_type in self.sync_order:
             strategy = strategy_map.get(node_type)
@@ -130,6 +132,8 @@ class FullSyncPipeline:
                 self._logger.info(f"🧩 Strategy start: {node_type}")
                 self._logger.info("-" * section_width)
 
+                await self._load_node_type(node_type, reason="pre-strategy refresh")
+
                 await strategy.bind()
                 self._logger.info(f"✅ Strategy bind: {node_type}")
 
@@ -169,10 +173,11 @@ class FullSyncPipeline:
         await self._load_persistence_state()
         self._logger.info("✅ Bootstrap: persistence loaded")
 
-        await self._load_from_datasource()
+        self._prepare_datasources()
-        self._logger.info("✅ Bootstrap: datasource fetched")
+        self._logger.info("✅ Bootstrap: datasource prepared")
 
     async def phase_bind(self) -> None:
+        self._prepare_datasources()
         strategy_map = {s.node_type: s for s in self.strategies}
         for node_type in self.sync_order:
             strategy = strategy_map.get(node_type)
@@ -186,6 +191,7 @@ class FullSyncPipeline:
                 continue
 
     async def phase_create(self) -> None:
+        self._prepare_datasources()
         strategy_map = {s.node_type: s for s in self.strategies}
         for node_type in self.sync_order:
             strategy = strategy_map.get(node_type)
@@ -207,6 +213,7 @@ class FullSyncPipeline:
                 continue
 
     async def phase_update(self) -> None:
+        self._prepare_datasources()
         strategy_map = {s.node_type: s for s in self.strategies}
         for node_type in self.sync_order:
             strategy = strategy_map.get(node_type)
@@ -271,26 +278,30 @@ class FullSyncPipeline:
             self._logger.info(f"🧹 Reset cleanup: {total_cleaned} CREATE-failed zombies cleaned")
         self._logger.info("🔄 Post-load reset cleanup completed")
 
-    async def _load_from_datasource(self) -> None:
+    def _prepare_datasources(self) -> None:
-        """从数据源加载数据"""
+        """为后续按 node_type 加载准备 collection 绑定。"""
         self.local_datasource.set_collection(self.local_collection)
-        await self.local_datasource.load_all(order=self.load_order)
-        
         self.remote_datasource.set_collection(self.remote_collection)
-        await self.remote_datasource.load_all(order=self.load_order)
 
-        for node_type in self.node_types:
+    async def _load_node_type(self, node_type: str, *, reason: str) -> None:
-            self.stats["loaded"] += len(self.local_collection.filter(node_type=node_type))
+        self._logger.info(f"🔄 Load node_type={node_type}, reason={reason}")
-            self.stats["loaded"] += len(self.remote_collection.filter(node_type=node_type))
+        await self.local_datasource.load_all(order=[node_type])
+        await self.remote_datasource.load_all(order=[node_type])
+
+        if node_type in self._loaded_node_types:
+            return
+
+        self._loaded_node_types.add(node_type)
+        self.stats["loaded"] += len(self.local_collection.filter(node_type=node_type))
+        self.stats["loaded"] += len(self.remote_collection.filter(node_type=node_type))
+
     def _get_action_success_count(self, datasource: "BaseDataSource", node_type: str, action_key: str) -> int:
         summary = datasource.get_action_summary().get(node_type, {})
         action_summary = summary.get(action_key, {}) if isinstance(summary, dict) else {}
         return int(action_summary.get("success", 0)) if isinstance(action_summary, dict) else 0
 
     async def _reload_node_type(self, node_type: str, *, reason: str) -> None:
-        self._logger.info(f"🔄 Reload after write: node_type={node_type}, reason={reason}")
+        await self._load_node_type(node_type, reason=reason)
-        await self.local_datasource.load_all(order=[node_type])
-        await self.remote_datasource.load_all(order=[node_type])
 
     async def _evaluate_consistency_for_node_type(self, node_type: str) -> Dict[str, Any]:
         strategy = next((item for item in self.strategies if item.node_type == node_type), None)

tests/unit/test_auto_test_project_profile.py

@@ -27,8 +27,8 @@ def test_build_temp_run_profile_uses_explicit_project_prebind(tmp_path: Path) ->
     source_profile.write_text(
         yaml.safe_dump(
             {
-                "local_datasource": {"target_project_ids": ["old-local"]},
+                "local_datasource": {"handler_configs": {"project": {"data_id_filter": ["old-local"]}}},
-                "remote_datasource": {"target_project_ids": ["old-remote"]},
+                "remote_datasource": {"handler_configs": {"project": {"data_id_filter": ["old-remote"]}}},
                 "strategies": {"project": {"config": {}}},
             },
             allow_unicode=True,
@@ -55,8 +55,8 @@ def test_build_temp_run_profile_uses_explicit_project_prebind(tmp_path: Path) ->
     profile = yaml.safe_load(Path(profile_path).read_text(encoding="utf-8"))
     project_config = profile["strategies"]["project"]["config"]
 
-    assert profile["local_datasource"]["target_project_ids"] == ["local-project"]
+    assert profile["local_datasource"]["handler_configs"]["project"]["data_id_filter"] == ["local-project"]
-    assert profile["remote_datasource"]["target_project_ids"] == ["remote-project"]
+    assert profile["remote_datasource"]["handler_configs"]["project"]["data_id_filter"] == ["remote-project"]
     assert project_config["auto_bind"] is False
     assert project_config["auto_bind_fields"] == []
     assert project_config["pre_bind_data_id"] == [

tests/unit/test_datasource_target_project_ids_resolution.py

@@ -21,12 +21,12 @@ def _build_config(
         local_datasource=JsonlDataSourceConfig(
             type="jsonl",
             jsonl_dir="./filtered_datasource/datasource/filtered",
-            target_project_ids=local_ids or [],
+            handler_configs={"project": {"data_id_filter": local_ids or []}},
         ),
         remote_datasource=JsonlDataSourceConfig(
             type="jsonl",
             jsonl_dir="./remote_datasource/datasource",
-            target_project_ids=remote_ids or [],
+            handler_configs={"project": {"data_id_filter": remote_ids or []}},
         ),
         strategies=[],
         persist=PersistConfig(db_path="./test_results/sync_state_machine/test.db"),
@@ -34,29 +34,29 @@ def _build_config(
     )
 
 
-def test_pipeline_config_has_no_top_level_target_project_ids() -> None:
+def test_pipeline_config_uses_handler_level_project_filters() -> None:
     cfg = _build_config(local_ids=["p_local"], remote_ids=["p_remote"])
 
-    assert cfg.local_datasource.target_project_ids == ["p_local"]
+    assert cfg.local_datasource.handler_configs["project"]["data_id_filter"] == ["p_local"]
-    assert cfg.remote_datasource.target_project_ids == ["p_remote"]
+    assert cfg.remote_datasource.handler_configs["project"]["data_id_filter"] == ["p_remote"]
-    with pytest.raises(AttributeError):
+    legacy_field_name = "_".join(("target", "project", "ids"))
-        _ = cfg.target_project_ids
+    assert legacy_field_name not in type(cfg).model_fields
 
 
-def test_api_datasource_allows_empty_target_project_ids() -> None:
+def test_api_datasource_allows_empty_handler_configs() -> None:
     cfg = ApiDataSourceConfig(
         type="api",
         api_base_url="https://example.com",
     )
 
-    assert cfg.target_project_ids == []
+    assert cfg.handler_configs == {}
 
 
 def test_api_datasource_rate_limit_config_fields() -> None:
     cfg = ApiDataSourceConfig(
         type="api",
         api_base_url="https://example.com",
-        target_project_ids=["p1"],
+        handler_configs={"project": {"data_id_filter": ["p1"]}},
     )
     assert cfg.api_rate_limit_max_requests == 10
     assert cfg.api_rate_limit_window_seconds == 10.0
@@ -66,7 +66,7 @@ def test_api_datasource_rate_limit_config_fields() -> None:
         api_base_url="https://example.com",
         api_rate_limit_max_requests=5,
         api_rate_limit_window_seconds=2.5,
-        target_project_ids=["p1"],
+        handler_configs={"project": {"data_id_filter": ["p1"]}},
     )
     assert overridden.api_rate_limit_max_requests == 5
     assert overridden.api_rate_limit_window_seconds == 2.5

tests/unit/test_factory_db_handler_registration.py

@@ -98,7 +98,6 @@ def test_register_handlers_uses_db_handler_for_db_datasource(tmp_path: Path) ->
         remote_datasource=JsonlDataSourceConfig(
             type="jsonl",
             jsonl_dir=str(tmp_path),
-            target_project_ids=[],
         ),
         strategies=[],
         persist=PersistConfig(db_path=str(tmp_path / "pipeline.db")),

tests/unit/test_factory_endpoint_initialization.py

@@ -103,12 +103,10 @@ async def test_factory_initializes_same_datasource_type_with_distinct_handler_co
         local_datasource=EndpointInitConfig(
             type="endpoint_init_memory",
             handler_configs={"factory_endpoint_demo": {"side": "local", "batch_size": 10}},
-            target_project_ids=["project_local"],
         ),
         remote_datasource=EndpointInitConfig(
             type="endpoint_init_memory",
             handler_configs={"factory_endpoint_demo": {"side": "remote", "batch_size": 20}},
-            target_project_ids=["project_remote"],
         ),
         strategies=[],
         persist=PersistConfig(db_path=str(tmp_path / "factory-init.db")),

tests/unit/test_full_sync_pipeline_downstream_reload.py

@@ -0,0 +1,96 @@
+from __future__ import annotations
+
+from types import SimpleNamespace
+
+import pytest
+
+from sync_state_machine.common.binding import BindingManager
+from sync_state_machine.common.collection import DataCollection
+from sync_state_machine.pipeline.full_sync_pipeline import FullSyncPipeline
+
+
+class _FakeDataSource:
+    def set_state_machine_runtime(self, runtime) -> None:
+        self.runtime = runtime
+
+    def set_collection(self, collection) -> None:
+        self.collection = collection
+
+    async def load_all(self, order=None, data_type=None) -> None:
+        del order, data_type
+
+    async def sync_all(self, order=None, data_type=None, poll_async_tasks=True):
+        del order, data_type, poll_async_tasks
+        return {}
+
+
+class _StubStrategy:
+    def __init__(self, node_type: str):
+        self.node_type = node_type
+        self.skip_sync = False
+        self.config = SimpleNamespace()
+        self.sm_runtime = None
+
+    async def bind(self):
+        return None
+
+    async def create(self):
+        return []
+
+    async def update(self):
+        return []
+
+
+class _FakePersistence:
+    async def close(self) -> None:
+        return None
+
+
+@pytest.mark.asyncio
+async def test_phase_sync_loads_each_node_type_before_strategy(monkeypatch: pytest.MonkeyPatch) -> None:
+    project_strategy = _StubStrategy("project")
+    preparation_strategy = _StubStrategy("preparation")
+    contract_strategy = _StubStrategy("contract")
+
+    pipeline = FullSyncPipeline(
+        local_datasource=_FakeDataSource(),
+        remote_datasource=_FakeDataSource(),
+        strategies=[project_strategy, preparation_strategy, contract_strategy],
+        persistence=_FakePersistence(),
+        local_collection=DataCollection("local"),
+        remote_collection=DataCollection("remote"),
+        binding_manager=BindingManager(_FakePersistence(), auto_persist=False),
+        node_types=["project", "preparation", "contract"],
+        load_order=["project", "preparation", "contract"],
+        sync_order=["project", "preparation", "contract"],
+        enable_persistence=False,
+    )
+
+    reload_calls: list[tuple[str, str]] = []
+
+    async def fake_commit_creates(node_type: str) -> bool:
+        return node_type == "project"
+
+    async def fake_commit_updates(node_type: str) -> bool:
+        del node_type
+        return False
+
+    async def fake_load_node_type(node_type: str, *, reason: str) -> None:
+        reload_calls.append((node_type, reason))
+
+    async def fake_normalize(node_type: str) -> None:
+        del node_type
+
+    monkeypatch.setattr(pipeline, "commit_creates", fake_commit_creates)
+    monkeypatch.setattr(pipeline, "commit_updates", fake_commit_updates)
+    monkeypatch.setattr(pipeline, "_load_node_type", fake_load_node_type)
+    monkeypatch.setattr(pipeline, "normalize_create_success_to_update_entry", fake_normalize)
+
+    await pipeline.phase_sync_by_node_type()
+
+    assert reload_calls == [
+        ("project", "pre-strategy refresh"),
+        ("project", "create wrote data"),
+        ("preparation", "pre-strategy refresh"),
+        ("contract", "pre-strategy refresh"),
+    ]

tests/unit/test_pipeline_strategy_continue_on_error.py

@@ -79,6 +79,11 @@ async def test_phase_sync_by_node_type_continues_after_strategy_error(tmp_path:
         enable_persistence=False,
     )
 
+    async def _noop_load_node_type(node_type: str, reason: str = "") -> None:
+        return None
+
+    pipeline._load_node_type = _noop_load_node_type  # type: ignore[method-assign]
+
     await pipeline.phase_sync_by_node_type()
 
     assert ok_strategy.bound is True

tests/unit/test_project_api_handler_updates.py

@@ -38,7 +38,8 @@ class _MockApiClient:
 
 @pytest.mark.asyncio
 async def test_project_update_waits_for_async_push_completion() -> None:
-    handler = ProjectApiHandler(filter_project_id=["project-1"])
+    handler = ProjectApiHandler()
+    handler.set_handler_config({"data_id_filter": ["project-1"]})
     handler.api_client = _MockApiClient()
 
     node = _ProjectUpdateNode(

tests/unit/test_target_project_preload_filter.py

@@ -43,8 +43,9 @@ async def test_jsonl_project_load_applies_target_project_filter(tmp_path):
     )
 
     datasource = JsonlDataSource(tmp_path)
-    datasource.register_handler(BaseJsonlHandler(datasource, "project", _ProjectNode, _ProjectSchema))
+    handler = BaseJsonlHandler(datasource, "project", _ProjectNode, _ProjectSchema)
-    datasource.set_target_project_ids(["p1"])
+    handler.set_handler_config({"data_id_filter": ["p1"]})
+    datasource.register_handler(handler)
     datasource.set_collection(DataCollection("local"))
 
     await datasource.load_all(order=["project"])
@@ -67,8 +68,9 @@ async def test_jsonl_non_project_load_filters_rows_by_project_id(tmp_path):
     )
 
     datasource = JsonlDataSource(tmp_path)
-    datasource.register_handler(BaseJsonlHandler(datasource, "contract", _BizNode, _BizSchema))
+    handler = BaseJsonlHandler(datasource, "contract", _BizNode, _BizSchema)
-    datasource.set_target_project_ids(["p1"])
+    handler.set_handler_config({"data_id_filter": ["p1"]})
+    datasource.register_handler(handler)
     datasource.set_collection(DataCollection("local"))
 
     await datasource.load_all(order=["contract"])

ui/config_models.py

@@ -69,7 +69,6 @@ class PipelineUiConfig(BaseModel):
     )
     logging: LoggingConfig = Field(default_factory=LoggingConfig)
 
-    target_project_ids: List[str] = Field(default_factory=lambda: ["f3e02e84-e81c-4aa6-88d4-f5aa108c8227"])
     strategies: Dict[str, Dict[str, Any]] = Field(default_factory=dict)
 
 
@@ -102,7 +101,6 @@ def preset_ui_config(preset_name: str) -> PipelineUiConfig:
         remote_datasource=run_cfg.remote_datasource.model_copy(deep=True),
         persist=run_cfg.persist.model_copy(deep=True),
         logging=run_cfg.logging.model_copy(deep=True),
-        target_project_ids=list(run_cfg.target_project_ids),
         strategies=strategies,
     )
 
@@ -136,6 +134,5 @@ def file_ui_config(config_file: str) -> PipelineUiConfig:
         remote_datasource=run_cfg.remote_datasource.model_copy(deep=True),
         persist=run_cfg.persist.model_copy(deep=True),
         logging=run_cfg.logging.model_copy(deep=True),
-        target_project_ids=list(run_cfg.target_project_ids),
         strategies=strategies,
     )

ui/runner.py

@@ -177,7 +177,6 @@ class PipelineRunnerService:
 
         run_cfg = PipelineRunConfig(
             node_types=list(cfg.node_types),
-            target_project_ids=list(cfg.target_project_ids),
             local_datasource=local_ds,
             remote_datasource=remote_ds,
             strategies=[],
@@ -196,7 +195,6 @@ class PipelineRunnerService:
 
         run_cfg = PipelineRunConfig(
             node_types=list(cfg.node_types),
-            target_project_ids=list(cfg.target_project_ids),
             local_datasource=local_ds,
             remote_datasource=remote_ds,
             strategies=[],

ui/web/index.html

@@ -200,14 +200,30 @@
     function toCsv(list) { return (list || []).join(","); }
     function fromCsv(text) { return (text || "").split(",").map(s => s.trim()).filter(Boolean); }
 
+    function buildProjectFilterHandlerConfigs(baseConfigs, projectIds) {
+      const nextConfigs = { ...(baseConfigs || {}) };
+      const ids = Array.isArray(projectIds) ? projectIds : [];
+      if (!ids.length) return nextConfigs;
+      nextConfigs.project = {
+        ...(nextConfigs.project || {}),
+        data_id_filter: ids,
+      };
+      return nextConfigs;
+    }
+
+    function readProjectFilter(cfg) {
+      return cfg?.handler_configs?.project?.data_id_filter || [];
+    }
+
     function readForm() {
       const mode = el.mode.value;
-      const handlerConfigs = safeJsonParse(el.handlerConfigs.value, {});
+      const projectIds = fromCsv(el.targetProjectIds.value);
+      const handlerConfigs = buildProjectFilterHandlerConfigs(safeJsonParse(el.handlerConfigs.value, {}), projectIds);
       const localDs = {
         type: "jsonl",
         jsonl_dir: el.localJsonlDir.value,
         read_only: false,
-        handler_configs: {},
+        handler_configs: buildProjectFilterHandlerConfigs({}, projectIds),
       };
       const remoteDs = mode === "jsonl_to_api"
         ? {
@@ -244,7 +260,6 @@
           file_path: el.logFilePath.value || null,
           level: 20,
         },
-        target_project_ids: fromCsv(el.targetProjectIds.value),
         strategies: safeJsonParse(el.strategies.value, {}),
       };
     }
@@ -257,7 +272,11 @@
       el.persistBackend.value = cfg.persist?.backend || "sqlite";
       el.persistenceDb.value = cfg.persist?.db_path || "";
       el.backendOptions.value = JSON.stringify(cfg.persist?.backend_options || {}, null, 2);
-      el.targetProjectIds.value = toCsv(cfg.target_project_ids);
+      el.targetProjectIds.value = toCsv(
+        readProjectFilter(cfg.remote_datasource).length
+          ? readProjectFilter(cfg.remote_datasource)
+          : readProjectFilter(cfg.local_datasource)
+      );
       if (cfg.remote_datasource?.type === "api") {
         el.apiBaseUrl.value = cfg.remote_datasource.api_base_url || "";
         el.apiUid.value = cfg.remote_datasource.api_uid || "";