Device registry port (#49)

* Update README and MQTTClient for installation instructions and code improvements * feat: 支持local_config启动 add: 增加对crt path的说明，为传入config.py的相对路径 move: web component * add: registry description * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * feat: node_info_update srv fix: OTDeck cant create * close #12 feat: slave node registry * feat: show machine name fix: host node registry not uploaded * feat: add hplc registry * feat: add hplc registry * fix: hplc status typo * fix: devices/ * 完成启动OT并联动rviz * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * 完成启动OT并联动rviz * fix: device.class possible null * fix: HPLC additions with online service * fix: slave mode spin not working * fix: slave mode spin not working * 修复rviz位置问题, 修复rviz位置问题, 在无tf变动时减缓发送频率在backend中添加物料跟随方法 * feat: 多ProtocolNode 允许子设备ID相同 feat: 上报发现的ActionClient feat: Host重启动，通过discover机制要求slaveNode重新注册，实现信息及时上报 * feat: 支持env设置config * fix: running logic * fix: running logic * fix: missing ot * 在main中直接初始化republisher和物料的mesh节点 * 将joint_republisher和resource_mesh_manager添加进 main_slave_run.py中 * Device visualization (#14) * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * 完成启动OT并联动rviz * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * 完成启动OT并联动rviz * 修复rviz位置问题, 修复rviz位置问题, 在无tf变动时减缓发送频率在backend中添加物料跟随方法 * fix: running logic * fix: running logic * fix: missing ot * 在main中直接初始化republisher和物料的mesh节点 * 将joint_republisher和resource_mesh_manager添加进 main_slave_run.py中 --------- Co-authored-by: zhangshixiang <@zhangshixiang> Co-authored-by: wznln <18435084+Xuwznln@users.noreply.github.com> * fix: missing hostname in devices_names fix: upload_file for model file * fix: missing paho-mqtt package bump version to 0.9.0 * fix startup add ResourceCreateFromOuter.action * fix type hint * update actions * update actions * host node add_resource_from_outer fix cmake list * pass device config to device class * add: bind_parent_ids to resource create action fix: message convert string * fix: host node should not be re_discovered * feat: resource tracker support dict * feat: add more necessary params * feat: fix boolean null in registry action data * feat: add outer resource * 编写mesh添加action * feat: append resource * add action * feat: vis 2d for plr * fix * fix: browser on rviz * fix: cloud bridge error fallback to local * fix: salve auto run rviz * 初始化两个plate * Device visualization (#22) * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * 完成启动OT并联动rviz * add 3d visualization * 完成在main中启动设备可视化完成在main中启动设备可视化，并输出物料ID：mesh的对应关系resource_model 添加物料模型管理类，遍历物料与resource_model，完成TF数据收集 * 完成TF发布 * 修改模型方向，在yaml中添加变换属性 * 添加物料tf变化时，发送topic到前端另外修改了物料初始化的方法，防止在tf还未发布时提前建立物料模型与发布话题 * 添加关节发布节点与物料可视化节点进入unilab * 使用json启动plr与3D模型仿真 * 完成启动OT并联动rviz * 修复rviz位置问题, 修复rviz位置问题, 在无tf变动时减缓发送频率在backend中添加物料跟随方法 * fix: running logic * fix: running logic * fix: missing ot * 在main中直接初始化republisher和物料的mesh节点 * 将joint_republisher和resource_mesh_manager添加进 main_slave_run.py中 * 编写mesh添加action * add action * fix * fix: browser on rviz * fix: cloud bridge error fallback to local * fix: salve auto run rviz * 初始化两个plate --------- Co-authored-by: zhangshixiang <@zhangshixiang> Co-authored-by: wznln <18435084+Xuwznln@users.noreply.github.com> * fix: multi channel * fix: aspirate * fix: aspirate * fix: aspirate * fix: aspirate * 提交 * fix: jobadd * fix: jobadd * fix: msg converter * tijiao * add resource creat easy action * identify debug msg * mq client id * unify liquid_handler definition * Update virtual_device.yaml * 更正了stir和heater的连接方式 * 区分了虚拟仪器中的八通阀和电磁阀，添加了两个阀门的驱动 * 修改了add protocol * 修复了阀门更新版的bug * 修复了添加protocol前缀导致的不能启动的bug * Fix handles * bump version to 0.9.6 * add resource edge upload * update container registry and handles * add virtual_separator virtual_rotavap fix transfer_pump * fix container value add parent_name to edge device id * 大图的问题都修复好了，添加了gassource和vacuum pump的驱动以及注册表 * default resource upload mode is false * 添加了icon的文件名在注册表里面 * 修改了json图中link的格式 * fix resource and edge upload * fix device ports * Fix edge id * 移除device的父节点关联 * separate registry sync and resource_add * 默认不进行注册表报送，通过命令unilabos-register或者增加启动参数 * 完善tip * protocol node不再嵌套显示 * bump version to 0.9.7 新增一个测试PumpTransferProtocol的teststation，亲测可以运行，将八通阀们和转移泵与pump_protocol适配 * protocol node 执行action不应携带自身device id * 添加了一套简易双八通阀工作站JSON，亲测能跑 * 修复了很多protocol，亲测能跑 * 添加了run column和filter through的protocol，亲测能跑 * fix mock_reactor * 修改了大图和小图的json，但是在前端上没看到改变 --------- Co-authored-by: Harvey Que <Q-Query@outlook.com> Co-authored-by: wznln <18435084+Xuwznln@users.noreply.github.com> Co-authored-by: zhangshixiang <@zhangshixiang> Co-authored-by: q434343 <73513873+q434343@users.noreply.github.com> Co-authored-by: Junhan Chang <changjh@pku.edu.cn>
2026-02-06 14:34:58 +00:00 · 2025-06-22 12:59:59 +08:00
parent f7db8d17c5
commit 46cec82a51
85 changed files with 12934 additions and 1573 deletions
--- a/unilabos/ros/main_slave_run.py
+++ b/unilabos/ros/main_slave_run.py
@@ -45,6 +45,7 @@ def exit() -> None:
 def main(
    devices_config: Dict[str, Any] = {},
    resources_config: list=[],
+    resources_edge_config: list=[],
    graph: Optional[Dict[str, Any]] = None,
    controllers_config: Dict[str, Any] = {},
    bridges: List[Any] = [],
@@ -62,6 +63,7 @@ def main(
        "host_node",
        devices_config,
        resources_config,
+        resources_edge_config,
        graph,
        controllers_config,
        bridges,
@@ -97,6 +99,7 @@ def main(
 def slave(
    devices_config: Dict[str, Any] = {},
    resources_config=[],
+    resources_edge_config=[],
    graph: Optional[Dict[str, Any]] = None,
    controllers_config: Dict[str, Any] = {},
    bridges: List[Any] = [],
--- a/unilabos/ros/msgs/message_converter.py
+++ b/unilabos/ros/msgs/message_converter.py
@@ -100,7 +100,7 @@ _action_mapping: Dict[Type, Dict[str, Any]] = {

 # 添加Protocol action类型到映射
 for py_msgtype in imsg.__all__:
-    if py_msgtype not in _action_mapping and py_msgtype.endswith("Protocol"):
+    if py_msgtype not in _action_mapping and (py_msgtype.endswith("Protocol") or py_msgtype.startswith("Protocol")):
        try:
            protocol_class = msg_converter_manager.get_class(f"unilabos.messages.{py_msgtype}")
            action_name = py_msgtype.replace("Protocol", "")
@@ -117,6 +117,7 @@ for py_msgtype in imsg.__all__:
                    "result": {k: k for k in action_type.Result().get_fields_and_field_types().keys()},
                }
        except Exception:
+            traceback.print_exc()
            logger.debug(f"Failed to load Protocol class: {py_msgtype}")

 # Python到ROS消息转换器
--- a/unilabos/ros/nodes/presets/host_node.py
+++ b/unilabos/ros/nodes/presets/host_node.py
@@ -22,6 +22,7 @@ from unilabos_msgs.srv import (
 )  # type: ignore
 from unique_identifier_msgs.msg import UUID

+from unilabos.config.config import BasicConfig
 from unilabos.registry.registry import lab_registry
 from unilabos.resources.graphio import initialize_resource
 from unilabos.resources.registry import add_schema
@@ -58,6 +59,7 @@ class HostNode(BaseROS2DeviceNode):
        device_id: str,
        devices_config: Dict[str, Any],
        resources_config: list,
+        resources_edge_config: list[dict],
        physical_setup_graph: Optional[Dict[str, Any]] = None,
        controllers_config: Optional[Dict[str, Any]] = None,
        bridges: Optional[List[Any]] = None,
@@ -96,6 +98,7 @@ class HostNode(BaseROS2DeviceNode):
        self.server_latest_timestamp = 0.0  #
        self.devices_config = devices_config
        self.resources_config = resources_config
+        self.resources_edge_config = resources_edge_config
        self.physical_setup_graph = physical_setup_graph
        if controllers_config is None:
            controllers_config = {}
@@ -144,13 +147,15 @@ class HostNode(BaseROS2DeviceNode):

        self.device_status = {}  # 用来存储设备状态
        self.device_status_timestamps = {}  # 用来存储设备状态最后更新时间
+        if BasicConfig.upload_registry:
+            from unilabos.app.mq import mqtt_client

-        from unilabos.app.mq import mqtt_client
-
-        for device_info in lab_registry.obtain_registry_device_info():
-            mqtt_client.publish_registry(device_info["id"], device_info)
-        for resource_info in lab_registry.obtain_registry_resource_info():
-            mqtt_client.publish_registry(resource_info["id"], resource_info)
+            for device_info in lab_registry.obtain_registry_device_info():
+                mqtt_client.publish_registry(device_info["id"], device_info)
+            for resource_info in lab_registry.obtain_registry_resource_info():
+                mqtt_client.publish_registry(resource_info["id"], resource_info)
+        else:
+            self.lab_logger().warning("本次启动注册表不报送云端，如果您需要联网调试，请使用unilab-register命令进行单独报送，或者在启动命令增加--upload_registry")
        time.sleep(1) # 等待MQTT连接稳定
        # 首次发现网络中的设备
        self._discover_devices()
@@ -191,24 +196,36 @@ class HostNode(BaseROS2DeviceNode):
        )
        resource_with_parent_name = []
        resource_ids_to_instance = {i["id"]: i for i in resources_config}
+        resource_name_to_with_parent_name = {}
        for res in resources_config:
-            if res.get("parent") and res.get("type") == "device" and res.get("class"):
-                parent_id = res.get("parent")
-                parent_res = resource_ids_to_instance[parent_id]
-                if parent_res.get("type") == "device" and parent_res.get("class"):
-                    resource_with_parent_name.append(copy.deepcopy(res))
-                    resource_with_parent_name[-1]["id"] = f"{parent_res['id']}/{res['id']}"
-                    continue
+            # if res.get("parent") and res.get("type") == "device" and res.get("class"):
+            #     parent_id = res.get("parent")
+            #     parent_res = resource_ids_to_instance[parent_id]
+            #     if parent_res.get("type") == "device" and parent_res.get("class"):
+            #         resource_with_parent_name.append(copy.deepcopy(res))
+            #         resource_name_to_with_parent_name[resource_with_parent_name[-1]["id"]] = f"{parent_res['id']}/{res['id']}"
+            #         resource_with_parent_name[-1]["id"] = f"{parent_res['id']}/{res['id']}"
+            #         continue
            resource_with_parent_name.append(copy.deepcopy(res))
+        # for edge in self.resources_edge_config:
+        #     edge["source"] = resource_name_to_with_parent_name.get(edge.get("source"), edge.get("source"))
+        #     edge["target"] = resource_name_to_with_parent_name.get(edge.get("target"), edge.get("target"))
        try:
            for bridge in self.bridges:
                if hasattr(bridge, "resource_add"):
+                    from unilabos.app.web.client import HTTPClient
+                    client: HTTPClient = bridge
                    resource_start_time = time.time()
-                    resource_add_res = bridge.resource_add(add_schema(resource_with_parent_name), True)
+                    resource_add_res = client.resource_add(add_schema(resource_with_parent_name), False)
                    resource_end_time = time.time()
                    self.lab_logger().info(
                        f"[Host Node-Resource] 物料上传 {round(resource_end_time - resource_start_time, 5) * 1000} ms"
                    )
+                    resource_add_res = client.resource_edge_add(self.resources_edge_config, False)
+                    resource_edge_end_time = time.time()
+                    self.lab_logger().info(
+                        f"[Host Node-Resource] 物料关系上传 {round(resource_edge_end_time - resource_end_time, 5) * 1000} ms"
+                    )
        except Exception as ex:
            self.lab_logger().error("[Host Node-Resource] 添加物料出错！")
            self.lab_logger().error(traceback.format_exc())
@@ -757,8 +774,10 @@ class HostNode(BaseROS2DeviceNode):
        self.lab_logger().info(f"[Host Node-Resource] Add request received: {len(resources)} resources")

        success = False
-        if len(self.bridges) > 0:
-            r = self.bridges[-1].resource_add(add_schema(resources))
+        if len(self.bridges) > 0:  # 边的提交待定
+            from unilabos.app.web.client import HTTPClient
+            client: HTTPClient = self.bridges[-1]
+            r = client.resource_add(add_schema(resources), False)
            success = bool(r)

        response.success = success
--- a/unilabos/ros/nodes/presets/protocol_node.py
+++ b/unilabos/ros/nodes/presets/protocol_node.py
@@ -110,7 +110,8 @@ class ROS2ProtocolNode(BaseROS2DeviceNode):

    def initialize_device(self, device_id, device_config):
        """初始化设备并创建相应的动作客户端"""
-        device_id_abs = f"{self.device_id}/{device_id}"
+        # device_id_abs = f"{self.device_id}/{device_id}"
+        device_id_abs = f"{device_id}"
        self.lab_logger().info(f"初始化子设备: {device_id_abs}")
        d = self.sub_devices[device_id] = initialize_device_from_dict(device_id_abs, device_config)

@@ -213,7 +214,7 @@ class ROS2ProtocolNode(BaseROS2DeviceNode):
        if device_id in ["", None, "self"]:
            action_id = f"/devices/{self.device_id}/{action_name}"
        else:
-            action_id = f"/devices/{self.device_id}/{device_id}/{action_name}"
+            action_id = f"/devices/{device_id}/{action_name}"  # 执行时取消了主节点信息 /{self.device_id}

        # 检查动作客户端是否存在
        if action_id not in self._action_clients: