feat: add knowledge_ids field to ApplicationVersion and populate from ResourceMapping

apps/application/migrations/0014_applicationversion_knowledge_ids.py

@@ -0,0 +1,61 @@
+# Generated by Django 5.2.15 on 2026-07-29 02:42
+
+from django.db import migrations, models
+
+
+def forwards(apps, schema_editor):
+    Application = apps.get_model("application", "Application")
+    ResourceMapping = apps.get_model("system_manage", "ResourceMapping")
+    ApplicationVersion = apps.get_model("application", "ApplicationVersion")
+
+    APPLICATION = "APPLICATION"
+    KNOWLEDGE = "KNOWLEDGE"
+    SIMPLE = "SIMPLE"
+    db_alias = schema_editor.connection.alias
+    simple_application_ids = {
+        str(app_id)
+        for app_id in Application.objects.using(db_alias)
+        .filter(type=SIMPLE)
+        .values_list("id", flat=True)
+    }
+    mapping = {}
+    qs = (
+        ResourceMapping.objects.using(db_alias)
+        .filter(source_type=APPLICATION, target_type=KNOWLEDGE)
+        .values_list("source_id", "target_id")
+    )
+    for source_id, target_id in qs.iterator():
+        if source_id in simple_application_ids:
+            mapping.setdefault(source_id, []).append(target_id)
+    mapping = {k: list(dict.fromkeys(v)) for k, v in mapping.items()}
+
+    updates = []
+    for obj in ApplicationVersion.objects.using(db_alias).iterator():
+        app_id = str(obj.application_id)
+        if app_id not in simple_application_ids:
+            continue
+        knowledge_ids = mapping.get(app_id)
+        if knowledge_ids:
+            obj.knowledge_ids = knowledge_ids
+            updates.append(obj)
+    if updates:
+        ApplicationVersion.objects.using(db_alias).bulk_update(
+            updates, ["knowledge_ids"], batch_size=500
+        )
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('application', '0013_application_long_term_enable_and_more'),
+        ('system_manage', '0005_resourcemapping'),
+    ]
+
+    operations = [
+        migrations.AddField(
+            model_name='applicationversion',
+            name='knowledge_ids',
+            field=models.JSONField(default=list, verbose_name='数据集id列表'),
+        ),
+        migrations.RunPython(forwards, migrations.RunPython.noop),
+    ]

apps/application/models/application.py

@@ -192,6 +192,7 @@ class ApplicationVersion(AppModelMixin):
     long_term_model_params_setting = models.JSONField(verbose_name="长期记忆模型参数相关设置", default=dict)
     long_term_trigger_type = models.CharField(verbose_name='长期记忆触发类型', default='ROUND')
     long_term_trigger_setting = models.JSONField(verbose_name='长期记忆触发配置', default=dict)
+    knowledge_ids = models.JSONField(verbose_name="数据集id列表", default=list)
 
     class Meta:
         db_table = "application_version"

apps/application/serializers/application.py

@@ -86,7 +86,7 @@ def get_bound_tool_ids(instance: Dict) -> List[str]:
     """
     tool_ids = set()
     for key in ("tool_ids", "skill_tool_ids", "mcp_tool_ids"):
-        for tool_id in (instance.get(key) or []):
+        for tool_id in instance.get(key) or []:
             tool_ids.add(str(tool_id))
     if instance.get("mcp_tool_id"):
         tool_ids.add(str(instance.get("mcp_tool_id")))
@@ -96,7 +96,7 @@ def collect(node_data):
             if node_data.get(key):
                 tool_ids.add(str(node_data.get(key)))
         for key in ("mcp_tool_ids", "tool_ids", "skill_tool_ids"):
-            for tool_id in (node_data.get(key) or []):
+            for tool_id in node_data.get(key) or []:
                 tool_ids.add(str(tool_id))
 
     _walk_workflow_nodes(instance.get("work_flow"), collect)
@@ -109,11 +109,11 @@ def get_bound_application_ids(instance: Dict) -> List[str]:
     ai-chat-node 的 node_data 包含 application_ids 列表。
     """
     application_ids = set()
-    for app_id in (instance.get("application_ids") or []):
+    for app_id in instance.get("application_ids") or []:
         application_ids.add(str(app_id))
 
     def collect(node_data):
-        for app_id in (node_data.get("application_ids") or []):
+        for app_id in node_data.get("application_ids") or []:
             application_ids.add(str(app_id))
 
     _walk_workflow_nodes(instance.get("work_flow"), collect)
@@ -186,7 +186,9 @@ def validate_bound_tool_permissions(user_id: str, workspace_id: str, instance: D
     application_ids = get_bound_application_ids(instance)
     if application_ids:
         authorized_application_ids = set(get_authorized_application_ids(user_id, workspace_id, application_ids))
-        unauthorized_application_ids = [app_id for app_id in application_ids if app_id not in authorized_application_ids]
+        unauthorized_application_ids = [
+            app_id for app_id in application_ids if app_id not in authorized_application_ids
+        ]
         if unauthorized_application_ids:
             message = lazy_format(
                 _("No permission to use application(s): {application_ids}"),
@@ -1268,6 +1270,14 @@ def publish(self, instance, with_valid=True):
             workspace_id=workspace_id,
         )
         self.reset_application_version(work_flow_version, application)
+        # 如果是简易应用 需要存入 knowledge_ids
+        if application.type == ApplicationTypeChoices.SIMPLE:
+            work_flow_version.knowledge_ids = [
+                str(row.target_id)
+                for row in QuerySet(ResourceMapping).filter(
+                    source_id=str(application.id), source_type="APPLICATION", target_type="KNOWLEDGE"
+                )
+            ]
         work_flow_version.save()
         access_token = hashlib.md5(str(uuid.uuid7()).encode()).hexdigest()[8:24]
         application_access_token = QuerySet(ApplicationAccessToken).filter(application_id=application.id).first()
@@ -1781,9 +1791,7 @@ def batch_delete(self, instance: Dict, with_valid=True):
         id_list = instance.get("id_list")
         workspace_id = self.data.get("workspace_id")
         id_list = list(
-            QuerySet(Application)
-            .filter(id__in=id_list, workspace_id=workspace_id)
-            .values_list("id", flat=True)
+            QuerySet(Application).filter(id__in=id_list, workspace_id=workspace_id).values_list("id", flat=True)
         )
 
         QuerySet(ApplicationVersion).filter(application_id__in=id_list).delete()
@@ -1842,9 +1850,7 @@ def batch_clean_time(self, instance: Dict, with_valid=True):
 
 class BatchCleanTimeSerializer(BatchSerializer):
     clean_time = serializers.IntegerField(required=True, min_value=1, max_value=100000, label=_("Clean time"))
-    file_clean_time = serializers.IntegerField(
-        required=True, min_value=1, max_value=100000, label=_("File clean time")
-    )
+    file_clean_time = serializers.IntegerField(required=True, min_value=1, max_value=100000, label=_("File clean time"))
 
     def is_valid(self, *, model=None, raise_exception=False):
         super().is_valid(model=model, raise_exception=True)

apps/chat/serializers/chat.py

@@ -495,11 +495,16 @@ def re_open_chat(self, chat_id: str):
             return self.re_open_chat_work_flow(chat_id, application)
 
     def re_open_chat_simple(self, chat_id, application):
-        # 数据集id列表
-        knowledge_id_list = [str(row.target_id) for row in
-                             QuerySet(ResourceMapping).filter(source_id=str(application.id),
-                                                              source_type='APPLICATION',
-                                                              target_type='KNOWLEDGE')]
+        if self.data.get('debug'):
+            # 数据集id列表
+            knowledge_id_list = [str(row.target_id) for row in
+                                 QuerySet(ResourceMapping).filter(source_id=str(application.id),
+                                                                  source_type='APPLICATION',
+                                                                  target_type='KNOWLEDGE')]
+        else:
+            application_version = QuerySet(ApplicationVersion).filter(application_id=application.id).order_by(
+                '-create_time')[0:1].first()
+            knowledge_id_list = application_version.knowledge_ids
 
         # 需要排除的文档
         exclude_document_id_list = [str(document.id) for document in
@@ -584,10 +589,15 @@ def open_simple(self, application):
         ip_address = self.data.get("ip_address")
         source = self.data.get("source")
         debug = self.data.get("debug")
-        knowledge_id_list = [str(row.target_id) for row in
-                             QuerySet(ResourceMapping).filter(source_id=str(application_id),
-                                                              source_type='APPLICATION',
-                                                              target_type='KNOWLEDGE')]
+        if debug:
+            knowledge_id_list = [str(row.target_id) for row in
+                                 QuerySet(ResourceMapping).filter(source_id=str(application_id),
+                                                                  source_type='APPLICATION',
+                                                                  target_type='KNOWLEDGE')]
+        else:
+            application_version = QuerySet(ApplicationVersion).filter(application_id=application_id).order_by(
+                '-create_time')[0:1].first()
+            knowledge_id_list = application_version.knowledge_ids
 
         chat_id = str(uuid.uuid7())
         ChatInfo(chat_id, chat_user_id, chat_user_type, ip_address, source, knowledge_id_list,