feat: add knowledge management API and permissions

Author: liuruibin

apps/common/constants/permission_constants.py

@@ -153,8 +153,11 @@ class PermissionConstants(Enum):
     KNOWLEDGE_MODULE_EDIT = Permission(group=Group.KNOWLEDGE, operate=Operate.EDIT, role_list=[RoleConstants.ADMIN,
                                                                                                RoleConstants.USER])
     KNOWLEDGE_MODULE_DELETE = Permission(group=Group.KNOWLEDGE, operate=Operate.DELETE, role_list=[RoleConstants.ADMIN,
-                                                                                                   RoleConstants.USER])
-
+                                                                                         RoleConstants.USER])
+    KNOWLEDGE_READ = Permission(group=Group.KNOWLEDGE, operate=Operate.READ, role_list=[RoleConstants.ADMIN,
+                                                                                         RoleConstants.USER])
+    KNOWLEDGE_CREATE = Permission(group=Group.KNOWLEDGE, operate=Operate.CREATE, role_list=[RoleConstants.ADMIN,
+                                                                                         RoleConstants.USER])
     def get_workspace_application_permission(self):
         return lambda r, kwargs: Permission(group=self.value.group, operate=self.value.operate,
                                             resource_path=

apps/knowledge/api/__init__.py

@@ -0,0 +1,54 @@
+from drf_spectacular.types import OpenApiTypes
+from drf_spectacular.utils import OpenApiParameter
+
+from common.mixins.api_mixin import APIMixin
+from common.result import ResultSerializer
+from knowledge.serializers.knowledge import KnowledgeBaseCreateRequest, KnowledgeModelSerializer
+
+
+class KnowledgeCreateResponse(ResultSerializer):
+    def get_data(self):
+        return KnowledgeModelSerializer()
+
+
+class KnowledgeCreateAPI(APIMixin):
+    @staticmethod
+    def get_parameters():
+        return [
+            OpenApiParameter(
+                name="workspace_id",
+                description="工作空间id",
+                type=OpenApiTypes.STR,
+                location='path',
+                required=True,
+            )
+        ]
+
+    @staticmethod
+    def get_request():
+        return KnowledgeBaseCreateRequest
+
+    @staticmethod
+    def get_response():
+        return KnowledgeCreateResponse
+
+
+class KnowledgeTreeReadAPI(APIMixin):
+    @staticmethod
+    def get_parameters():
+        return [
+            OpenApiParameter(
+                name="workspace_id",
+                description="工作空间id",
+                type=OpenApiTypes.STR,
+                location='path',
+                required=True,
+            ),
+            OpenApiParameter(
+                name="module_id",
+                description="模块id",
+                type=OpenApiTypes.STR,
+                location='query',
+                required=False,
+            )
+        ]

apps/knowledge/api/knowledge.py

@@ -14,9 +14,7 @@ def insert_default_data(apps, schema_editor):
     KnowledgeModule.objects.create(id='root', name='根目录', user_id='f0dd8f71-e4ee-11ee-8c84-a8a1595801ab')
 
 
-
 class Migration(migrations.Migration):
-
     initial = True
 
     dependencies = [
@@ -30,8 +28,12 @@ class Migration(migrations.Migration):
             fields=[
                 ('create_time', models.DateTimeField(auto_now_add=True, verbose_name='创建时间')),
                 ('update_time', models.DateTimeField(auto_now=True, verbose_name='修改时间')),
-                ('id', models.UUIDField(default=uuid_utils.compat.uuid7, editable=False, primary_key=True, serialize=False, verbose_name='主键id')),
+                ('id',
+                 models.UUIDField(default=uuid_utils.compat.uuid7, editable=False, primary_key=True, serialize=False,
+                                  verbose_name='主键id')),
                 ('file_name', models.CharField(default='', max_length=256, verbose_name='文件名称')),
+                ('workspace_id',
+                 models.CharField(db_index=True, default='default', max_length=64, verbose_name='工作空间id')),
                 ('loid', models.IntegerField(verbose_name='loid')),
                 ('meta', models.JSONField(default=dict, verbose_name='文件关联数据')),
             ],
@@ -40,39 +42,60 @@ class Migration(migrations.Migration):
             },
         ),
         migrations.CreateModel(
-            name='Knowledge',
+            name='KnowledgeModule',
             fields=[
                 ('create_time', models.DateTimeField(auto_now_add=True, verbose_name='创建时间')),
                 ('update_time', models.DateTimeField(auto_now=True, verbose_name='修改时间')),
-                ('id', models.UUIDField(default=uuid_utils.compat.uuid7, editable=False, primary_key=True, serialize=False, verbose_name='主键id')),
-                ('name', models.CharField(max_length=150, verbose_name='知识库名称')),
-                ('desc', models.CharField(max_length=256, verbose_name='描述')),
-                ('type', models.IntegerField(choices=[(0, '通用类型'), (1, 'web站点类型'), (2, '飞书类型'), (3, '语雀类型')], default=0, verbose_name='类型')),
-                ('meta', models.JSONField(default=dict, verbose_name='元数据')),
-                ('embedding_mode', models.ForeignKey(default=knowledge.models.knowledge.default_model, on_delete=django.db.models.deletion.DO_NOTHING, to='models_provider.model', verbose_name='向量模型')),
-                ('user', models.ForeignKey(on_delete=django.db.models.deletion.DO_NOTHING, to='users.user', verbose_name='所属用户')),
+                ('id', models.CharField(editable=False, max_length=64, primary_key=True, serialize=False,
+                                        verbose_name='主键id')),
+                ('name', models.CharField(max_length=64, verbose_name='文件夹名称')),
+                ('workspace_id',
+                 models.CharField(db_index=True, default='default', max_length=64, verbose_name='工作空间id')),
+                ('lft', models.PositiveIntegerField(editable=False)),
+                ('rght', models.PositiveIntegerField(editable=False)),
+                ('tree_id', models.PositiveIntegerField(db_index=True, editable=False)),
+                ('level', models.PositiveIntegerField(editable=False)),
+                ('parent',
+                 mptt.fields.TreeForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.CASCADE,
+                                            related_name='children', to='knowledge.knowledgemodule')),
+                ('user', models.ForeignKey(on_delete=django.db.models.deletion.DO_NOTHING, to='users.user',
+                                           verbose_name='用户id')),
             ],
             options={
-                'db_table': 'knowledge',
+                'db_table': 'knowledge_module',
             },
         ),
         migrations.CreateModel(
-            name='KnowledgeModule',
+            name='Knowledge',
             fields=[
                 ('create_time', models.DateTimeField(auto_now_add=True, verbose_name='创建时间')),
                 ('update_time', models.DateTimeField(auto_now=True, verbose_name='修改时间')),
-                ('id', models.CharField(editable=False, max_length=64, primary_key=True, serialize=False, verbose_name='主键id')),
-                ('name', models.CharField(max_length=64, verbose_name='文件夹名称')),
-                ('workspace_id', models.CharField(db_index=True, default='default', max_length=64, verbose_name='工作空间id')),
-                ('lft', models.PositiveIntegerField(editable=False)),
-                ('rght', models.PositiveIntegerField(editable=False)),
-                ('tree_id', models.PositiveIntegerField(db_index=True, editable=False)),
-                ('level', models.PositiveIntegerField(editable=False)),
-                ('parent', mptt.fields.TreeForeignKey(blank=True, null=True, on_delete=django.db.models.deletion.CASCADE, related_name='children', to='knowledge.knowledgemodule')),
-                ('user', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to='users.user', verbose_name='用户id')),
+                ('id',
+                 models.UUIDField(default=uuid_utils.compat.uuid7, editable=False, primary_key=True, serialize=False,
+                                  verbose_name='主键id')),
+                ('name', models.CharField(max_length=150, verbose_name='知识库名称')),
+                ('workspace_id',
+                 models.CharField(db_index=True, default='default', max_length=64, verbose_name='工作空间id')),
+                ('desc', models.CharField(max_length=256, verbose_name='描述')),
+                ('type',
+                 models.IntegerField(choices=[(0, '通用类型'), (1, 'web站点类型'), (2, '飞书类型'), (3, '语雀类型')],
+                                     default=0, verbose_name='类型')),
+                ('meta', models.JSONField(default=dict, verbose_name='元数据')),
+                ('scope',
+                 models.CharField(choices=[('SHARED', '共享'), ('WORKSPACE', '工作空间可用')], default='WORKSPACE',
+                                  max_length=20, verbose_name='可用范围')),
+                ('module',
+                 models.ForeignKey(default='root', on_delete=django.db.models.deletion.CASCADE,
+                                   to='knowledge.knowledgemodule',
+                                   verbose_name='模块id')),
+                ('embedding_model', models.ForeignKey(default=knowledge.models.knowledge.default_model,
+                                                     on_delete=django.db.models.deletion.DO_NOTHING,
+                                                     to='models_provider.model', verbose_name='向量模型')),
+                ('user', models.ForeignKey(on_delete=django.db.models.deletion.DO_NOTHING, to='users.user',
+                                           verbose_name='所属用户')),
             ],
             options={
-                'db_table': 'knowledge_module',
+                'db_table': 'knowledge',
             },
         ),
         migrations.RunPython(insert_default_data),

apps/knowledge/migrations/0001_initial.py

@@ -12,10 +12,15 @@
 
 
 class KnowledgeType(models.IntegerChoices):
-    base = 0, '通用类型'
-    web = 1, 'web站点类型'
-    lark = 2, '飞书类型'
-    yuque = 3, '语雀类型'
+    BASE = 0, '通用类型'
+    WEB = 1, 'web站点类型'
+    LARK = 2, '飞书类型'
+    YUQUE = 3, '语雀类型'
+
+
+class KnowledgeScope(models.TextChoices):
+    SHARED = "SHARED", '共享'
+    WORKSPACE = "WORKSPACE", "工作空间可用"
 
 
 def default_model():
@@ -26,7 +31,7 @@ def default_model():
 class KnowledgeModule(MPTTModel, AppModelMixin):
     id = models.CharField(primary_key=True, max_length=64, editable=False, verbose_name="主键id")
     name = models.CharField(max_length=64, verbose_name="文件夹名称")
-    user = models.ForeignKey(User, on_delete=models.CASCADE, verbose_name="用户id")
+    user = models.ForeignKey(User, on_delete=models.DO_NOTHING, verbose_name="用户id")
     workspace_id = models.CharField(max_length=64, verbose_name="工作空间id", default="default", db_index=True)
     parent = TreeForeignKey('self', on_delete=models.CASCADE, null=True, blank=True, related_name='children')
 
@@ -37,13 +42,17 @@ class MPTTMeta:
         order_insertion_by = ['name']
 
 
+
 class Knowledge(AppModelMixin):
     id = models.UUIDField(primary_key=True, max_length=128, default=uuid.uuid7, editable=False, verbose_name="主键id")
     name = models.CharField(max_length=150, verbose_name="知识库名称")
+    workspace_id = models.CharField(max_length=64, verbose_name="工作空间id", default="default", db_index=True)
     desc = models.CharField(max_length=256, verbose_name="描述")
     user = models.ForeignKey(User, on_delete=models.DO_NOTHING, verbose_name="所属用户")
-    type = models.IntegerField(verbose_name='类型', choices=KnowledgeType.choices, default=KnowledgeType.base)
-    embedding_mode = models.ForeignKey(Model, on_delete=models.DO_NOTHING, verbose_name="向量模型",
+    type = models.IntegerField(verbose_name='类型', choices=KnowledgeType.choices, default=KnowledgeType.BASE)
+    scope = models.CharField(max_length=20, verbose_name='可用范围', choices=KnowledgeScope.choices, default=KnowledgeScope.WORKSPACE)
+    module = models.ForeignKey(KnowledgeModule, on_delete=models.CASCADE, verbose_name="模块id", default='root')
+    embedding_model = models.ForeignKey(Model, on_delete=models.DO_NOTHING, verbose_name="向量模型",
                                        default=default_model)
     meta = models.JSONField(verbose_name="元数据", default=dict)
 
@@ -54,6 +63,7 @@ class Meta:
 class File(AppModelMixin):
     id = models.UUIDField(primary_key=True, max_length=128, default=uuid.uuid7, editable=False, verbose_name="主键id")
     file_name = models.CharField(max_length=256, verbose_name="文件名称", default="")
+    workspace_id = models.CharField(max_length=64, verbose_name="工作空间id", default="default", db_index=True)
     loid = models.IntegerField(verbose_name="loid")
     meta = models.JSONField(verbose_name="文件关联数据", default=dict)
 

apps/knowledge/models/knowledge.py

@@ -0,0 +1 @@
+# coding=utf-8

apps/knowledge/serializers/__init__.py

@@ -0,0 +1,52 @@
+import uuid_utils as uuid
+from django.utils.translation import gettext_lazy as _
+from rest_framework import serializers
+
+from knowledge.models import Knowledge, KnowledgeScope, KnowledgeType
+
+
+class KnowledgeModelSerializer(serializers.ModelSerializer):
+    class Meta:
+        model = Knowledge
+        fields = ['id', 'name', 'desc', 'meta', 'module_id', 'type', 'workspace_id', 'create_time', 'update_time']
+
+
+class KnowledgeBaseCreateRequest(serializers.Serializer):
+    name = serializers.CharField(required=True, label=_('knowledge name'))
+    desc = serializers.CharField(required=False, allow_null=True, allow_blank=True, label=_('knowledge description'))
+    embedding = serializers.CharField(required=True, label=_('knowledge embedding'))
+
+class KnowledgeWebCreateRequest(serializers.Serializer):
+    name = serializers.CharField(required=True, label=_('knowledge name'))
+    desc = serializers.CharField(required=False, allow_null=True, allow_blank=True, label=_('knowledge description'))
+    embedding = serializers.CharField(required=True, label=_('knowledge embedding'))
+
+
+class KnowledgeSerializer(serializers.Serializer):
+    class Create(serializers.Serializer):
+        user_id = serializers.UUIDField(required=True, label=_('user id'))
+        workspace_id = serializers.CharField(required=True, label=_('workspace id'))
+
+        def insert(self, instance, with_valid=True):
+            if with_valid:
+                self.is_valid(raise_exception=True)
+                KnowledgeBaseCreateRequest(data=instance).is_valid(raise_exception=True)
+            knowledge = Knowledge(
+                id=uuid.uuid7(),
+                name=instance.get('name'),
+                workspace_id=self.data.get('workspace_id'),
+                desc=instance.get('desc'),
+                type=instance.get('type', KnowledgeType.BASE),
+                user_id=self.data.get('user_id'),
+                scope=KnowledgeScope.WORKSPACE,
+                module_id=instance.get('module_id', 'root'),
+                embedding_model_id=instance.get('embedding'),
+                meta=instance.get('meta', {}),
+            )
+            knowledge.save()
+            return KnowledgeModelSerializer(knowledge).data
+
+
+class KnowledgeTreeSerializer(serializers.Serializer):
+    def get_knowledge_list(self, param):
+        pass

apps/knowledge/serializers/knowledge.py

@@ -1,3 +1,8 @@
-from django.shortcuts import render
+from django.urls import path
 
-# Create your views here.
+from . import views
+
+app_name = "knowledge"
+urlpatterns = [
+    path('workspace/<str:workspace_id>/knowledge', views.KnowledgeView.as_view()),
+]

apps/knowledge/urls.py

@@ -0,0 +1 @@
+from .knowledge import *

apps/knowledge/views/__init__.py

@@ -0,0 +1,44 @@
+from django.utils.translation import gettext_lazy as _
+from drf_spectacular.utils import extend_schema
+from rest_framework.request import Request
+from rest_framework.views import APIView
+
+from common.auth import TokenAuth
+from common.auth.authentication import has_permissions
+from common.constants.permission_constants import PermissionConstants
+from common.result import result
+from knowledge.api.knowledge import KnowledgeCreateAPI, KnowledgeTreeReadAPI
+from knowledge.serializers.knowledge import KnowledgeSerializer, KnowledgeTreeSerializer
+
+
+class KnowledgeView(APIView):
+    authentication_classes = [TokenAuth]
+
+    @extend_schema(
+        methods=['POST'],
+        description=_('Create knowledge'),
+        operation_id=_('Create knowledge'),
+        parameters=KnowledgeCreateAPI.get_parameters(),
+        request=KnowledgeCreateAPI.get_request(),
+        responses=KnowledgeCreateAPI.get_response(),
+        tags=[_('Knowledge Base')]
+    )
+    @has_permissions(PermissionConstants.KNOWLEDGE_CREATE.get_workspace_permission())
+    def post(self, request: Request, workspace_id: str):
+        return result.success(KnowledgeSerializer.Create(
+            data={'user_id': request.user.id, 'workspace_id': workspace_id}
+        ).insert(request.data))
+
+    @extend_schema(
+        methods=['GET'],
+        description=_('Get knowledge by module'),
+        operation_id=_('Get knowledge by module'),
+        parameters=KnowledgeTreeReadAPI.get_parameters(),
+        responses=KnowledgeTreeReadAPI.get_response(),
+        tags=[_('Knowledge Base')]
+    )
+    @has_permissions(PermissionConstants.KNOWLEDGE_READ.get_workspace_permission())
+    def get(self, request: Request, workspace_id: str):
+        return result.success(KnowledgeTreeSerializer(
+            data={'workspace_id': workspace_id}
+        ).get_knowledge_list(request.query_params.get('module_id')))