feat:完成dify知识库文档基础CRUD模块

2025-11-30 21:28:49 +08:00
parent d85010bada
commit 754ec2c7b5
21 changed files with 1142 additions and 706 deletions
@@ -1,13 +1,13 @@
 /**
- * Dify 服务端基础 API 模块
+ * Dify Chat 服务端 API 模块
 *
 * 提供 Node.js 服务端调用 FastAPI 后端的基础功能
- * 包括配置管理和基础请求函数
+ * Dify 的 API_KEY 和 APP_ID 由 FastAPI 后端管理，前端只负责转发请求
 *
 * 调用链路：
- * Remix Server → FastAPI /dify/* → Dify
+ * Remix Server → FastAPI /dify_chat/* → Dify
 *
- * @module api/dify/client.server
+ * @module api/dify-chat/client.server
 */

 import { API_BASE_URL } from '~/config/api-config';
@@ -17,45 +17,25 @@ import { API_BASE_URL } from '~/config/api-config';
 // ============================================================================

 /**
- * 获取环境变量的服务端函数
+ * Dify Chat API 代理地址
+ * 通过 FastAPI 后端的 /dify_chat 路由代理访问 Dify
+ * Dify 的认证（API_KEY）由 FastAPI 后端处理
 */
-const getServerEnvVar = (name: string, defaultValue: string = ''): string => {
-    return process.env[name] || defaultValue;
-};
-
-/**
- * Dify API 客户端配置
- * 通过 FastAPI 后端的 /dify 路由代理访问 Dify
- */
-const DIFY_CONFIG = {
-    API_URL: `${API_BASE_URL}/dify_chat`,
-    API_KEY: getServerEnvVar('NEXT_PUBLIC_APP_KEY', ''),
-    APP_ID: (() => {
-        const rawAppId = getServerEnvVar('NEXT_PUBLIC_APP_ID', '');
-        const match = rawAppId.match(/\/app\/([a-f0-9-]{36})/);
-        return match ? match[1] : rawAppId;
-    })(),
-};
-
-console.log('[Dify Server] 配置:', {
-    apiUrl: DIFY_CONFIG.API_URL,
-    apiBaseUrl: API_BASE_URL,
-    appId: DIFY_CONFIG.APP_ID,
-    configComplete: !!(DIFY_CONFIG.API_URL && DIFY_CONFIG.APP_ID)
-});
+const DIFY_CHAT_API_URL = `${API_BASE_URL}/dify_chat`;

 // ============================================================================
 // 基础请求函数
 // ============================================================================

 /**
- * Dify API 基础请求函数
+ * Dify Chat API 基础请求函数
 *
- * 使用 JWT 认证通过 FastAPI 代理访问 Dify
+ * 使用用户 JWT 认证通过 FastAPI 代理访问 Dify
+ * FastAPI 后端会验证 JWT 并添加 Dify API_KEY
 *
 * @param endpoint - API 端点路径
 * @param options - fetch 请求选项
- * @param jwt - JWT 认证令牌
+ * @param jwt - 用户 JWT 认证令牌
 * @returns Response 对象
 */
 export async function difyFetch(
@@ -63,7 +43,7 @@ export async function difyFetch(
    options: RequestInit = {},
    jwt?: string
 ): Promise<Response> {
-    const url = `${DIFY_CONFIG.API_URL}/${endpoint.replace(/^\//, '')}`;
+    const url = `${DIFY_CHAT_API_URL}/${endpoint.replace(/^\//, '')}`;

    const headers: HeadersInit = {
        'Content-Type': 'application/json',
@@ -73,7 +53,7 @@ export async function difyFetch(
    if (jwt) {
        (headers as Record<string, string>)['Authorization'] = `Bearer ${jwt}`;
    } else {
-        console.warn('[Dify Server] 没有提供 JWT，转发fastapi请求可能失败');
+        console.warn('[Dify Chat] 没有提供 JWT，FastAPI 请求可能失败');
    }

    const response = await fetch(url, {
@@ -83,7 +63,7 @@ export async function difyFetch(

    if (!response.ok) {
        const errorText = await response.text();
-        console.error('[Dify Server] API 转发错误:', {
+        console.error('[Dify Chat] API 转发错误:', {
            status: response.status,
            statusText: response.statusText,
            error: errorText
@@ -99,19 +79,5 @@ export async function difyFetch(
    return response;
 }

-// ============================================================================
-// 工具函数
-// ============================================================================
-
-/**
- * Dify 工具函数
- */
-export const difyUtils = {
-    /**
-     * 获取 Dify 配置
-     */
-    getConfig: () => DIFY_CONFIG,
-};
-
 // 重新导出 chat 模块的 difyClient
 export { difyClient } from './chat';
@@ -0,0 +1,84 @@
+/**
+ * Dify Dataset 服务端 API 模块
+ *
+ * 提供 Node.js 服务端调用 FastAPI 后端的基础功能
+ * Dify 的 DATASET_API_KEY 由 FastAPI 后端管理，前端只负责转发请求
+ *
+ * 调用链路：
+ * Remix Server → FastAPI /dify_dataset/* → Dify Knowledge API
+ *
+ * @module api/dify-dataset/client.server
+ */
+
+import { API_BASE_URL } from '~/config/api-config';
+
+// ============================================================================
+// 配置
+// ============================================================================
+
+/**
+ * Dify Dataset API 代理地址
+ * 通过 FastAPI 后端的 /dify_dataset 路由代理访问 Dify Knowledge API
+ * Dify 的认证（DATASET_API_KEY）由 FastAPI 后端处理
+ */
+const DIFY_DATASET_API_URL = `${API_BASE_URL}/dify_dataset`;
+
+// ============================================================================
+// 基础请求函数
+// ============================================================================
+
+/**
+ * Dify Dataset API 基础请求函数
+ *
+ * 使用用户 JWT 认证通过 FastAPI 代理访问 Dify Knowledge API
+ * FastAPI 后端会验证 JWT 并添加 Dify DATASET_API_KEY
+ *
+ * @param endpoint - API 端点路径
+ * @param options - fetch 请求选项
+ * @param jwt - 用户 JWT 认证令牌
+ * @returns Response 对象
+ */
+export async function difyDatasetFetch(
+    endpoint: string,
+    options: RequestInit = {},
+    jwt?: string
+): Promise<Response> {
+    const url = `${DIFY_DATASET_API_URL}/${endpoint.replace(/^\//, '')}`;
+
+    const headers: HeadersInit = {
+        ...options.headers,
+    };
+
+    // 如果不是 FormData，设置 Content-Type 为 JSON
+    if (!(options.body instanceof FormData)) {
+        (headers as Record<string, string>)['Content-Type'] = 'application/json';
+    }
+
+    if (jwt) {
+        (headers as Record<string, string>)['Authorization'] = `Bearer ${jwt}`;
+    } else {
+        console.warn('[Dify Dataset] 没有提供 JWT，FastAPI 请求可能失败');
+    }
+
+    const response = await fetch(url, {
+        ...options,
+        headers,
+    });
+
+    if (!response.ok) {
+        const errorText = await response.text();
+        console.error('[Dify Dataset] API 转发错误:', {
+            status: response.status,
+            statusText: response.statusText,
+            error: errorText
+        });
+
+        if (response.status === 401) {
+            throw new Error('JWT认证失败，请重新登录');
+        }
+
+        throw new Error(`Dify Dataset API Error: ${response.status} ${response.statusText}`);
+    }
+
+    return response;
+}
@@ -9,11 +9,15 @@

 import axios from 'axios';
 import type {
+    Dataset,
    DatasetsResponse,
    DocumentsResponse,
    SegmentsResponse,
    Document,
    OperationResult,
+    IndexingStatusResponse,
+    UploadFileInfo,
+    UpdateDatasetRequest,
 } from './types';

 // ============================================================================
@@ -59,14 +63,38 @@ export async function fetchDatasets(
 * @param datasetId - 知识库 ID
 * @returns 知识库详情
 */
-export async function fetchDataset(datasetId: string): Promise<any> {
-    const response = await axios.get(
+export async function fetchDataset(datasetId: string): Promise<Dataset> {
+    const response = await axios.get<Dataset>(
        `${API_URL}/datasets/${datasetId}`,
        { withCredentials: true }
    );
    return response.data;
 }

+/**
+ * 更新知识库详情
+ *
+ * @param datasetId - 知识库 ID
+ * @param data - 更新数据
+ * @returns 更新后的知识库详情
+ */
+export async function updateDataset(
+    datasetId: string,
+    data: UpdateDatasetRequest
+): Promise<Dataset> {
+    console.log('[Dataset Client] 更新知识库:', { datasetId, data });
+
+    const response = await axios.patch<Dataset>(
+        `${API_URL}/datasets/${datasetId}`,
+        data,
+        {
+            headers: { 'Content-Type': 'application/json' },
+            withCredentials: true,
+        }
+    );
+    return response.data;
+}
+
 // ============================================================================
 // 文档 API
 // ============================================================================
@@ -144,6 +172,8 @@ export async function deleteDocument(

 /**
 * 启用/禁用文档
+ * Dify API: PATCH /datasets/{dataset_id}/documents/status/{action}
+ * action: enable / disable / archive / un_archive
 *
 * @param datasetId - 知识库 ID
 * @param documentId - 文档 ID
@@ -155,11 +185,12 @@ export async function toggleDocumentStatus(
    documentId: string,
    enabled: boolean
 ): Promise<OperationResult> {
-    console.log('[Dataset Client] 切换文档状态:', { datasetId, documentId, enabled });
+    const action = enabled ? 'enable' : 'disable';
+    console.log('[Dataset Client] 切换文档状态:', { datasetId, documentId, action });

    const response = await axios.patch<OperationResult>(
-        `${API_URL}/datasets/${datasetId}/documents/${documentId}/status`,
-        { enabled },
+        `${API_URL}/datasets/${datasetId}/documents/status/${action}`,
+        { document_ids: [documentId] },
        {
            headers: { 'Content-Type': 'application/json' },
            withCredentials: true,
@@ -231,6 +262,8 @@ export async function deleteSegment(

 /**
 * 启用/禁用分段
+ * Dify API: POST /datasets/{dataset_id}/documents/{document_id}/segments/{segment_id}
+ * 通过更新分段的方式来切换状态
 *
 * @param datasetId - 知识库 ID
 * @param documentId - 文档 ID
@@ -244,9 +277,11 @@ export async function toggleSegmentStatus(
    segmentId: string,
    enabled: boolean
 ): Promise<OperationResult> {
-    const response = await axios.patch<OperationResult>(
-        `${API_URL}/datasets/${datasetId}/documents/${documentId}/segments/${segmentId}/status`,
-        { enabled },
+    console.log('[Dataset Client] 切换分段状态:', { datasetId, documentId, segmentId, enabled });
+
+    const response = await axios.post<OperationResult>(
+        `${API_URL}/datasets/${datasetId}/documents/${documentId}/segments/${segmentId}`,
+        { segment: { enabled } },
        {
            headers: { 'Content-Type': 'application/json' },
            withCredentials: true,
@@ -284,7 +319,7 @@ export async function uploadDocument(
    console.log('[Dataset Client] 上传文档:', { datasetId, fileName: file.name });

    const response = await axios.post(
-        `${API_URL}/datasets/${datasetId}/documents/create-by-file`,
+        `${API_URL}/datasets/${datasetId}/documents`,
        formData,
        {
            withCredentials: true,
@@ -298,3 +333,43 @@ export async function uploadDocument(
    );
    return response.data;
 }
+
+/**
+ * 获取文档嵌入状态（索引进度）
+ *
+ * @param datasetId - 知识库 ID
+ * @param batch - 上传文档的批次号
+ * @returns 索引状态列表
+ */
+export async function fetchIndexingStatus(
+    datasetId: string,
+    batch: string
+): Promise<IndexingStatusResponse> {
+    console.log('[Dataset Client] 获取索引状态:', { datasetId, batch });
+
+    const response = await axios.get<IndexingStatusResponse>(
+        `${API_URL}/datasets/${datasetId}/documents/${batch}/indexing-status`,
+        { withCredentials: true }
+    );
+    return response.data;
+}
+
+/**
+ * 获取文档上传文件信息
+ *
+ * @param datasetId - 知识库 ID
+ * @param documentId - 文档 ID
+ * @returns 上传文件信息
+ */
+export async function fetchUploadFileInfo(
+    datasetId: string,
+    documentId: string
+): Promise<UploadFileInfo> {
+    console.log('[Dataset Client] 获取上传文件信息:', { datasetId, documentId });
+
+    const response = await axios.get<UploadFileInfo>(
+        `${API_URL}/datasets/${datasetId}/documents/${documentId}/upload-file`,
+        { withCredentials: true }
+    );
+    return response.data;
+}
@@ -16,21 +16,32 @@ export type {
    OperationResult,
    CreateDocumentResponse,
    UploadProgress,
+    DocumentIndexingStatus,
+    IndexingStatusResponse,
+    UploadFileInfo,
+    RetrievalModel,
+    UpdateDatasetRequest,
 } from './types';

-// 客户端 API 导出
+// 客户端 API 导出（浏览器端使用 axios）
 export {
    // 知识库
    fetchDatasets,
    fetchDataset,
+    updateDataset,
    // 文档
    fetchDocuments,
    fetchDocument,
    deleteDocument,
    toggleDocumentStatus,
    uploadDocument,
+    fetchIndexingStatus,
+    fetchUploadFileInfo,
    // 分段
    fetchSegments,
    deleteSegment,
    toggleSegmentStatus,
 } from './client';
+
+// 服务端 API 请直接从 client.server.ts 导入
+// import { difyDatasetFetch } from '~/api/dify-dataset/client.server';
@@ -165,3 +165,86 @@ export interface UploadProgress {
    total: number;
    percent: number;
 }
+
+// ============================================================================
+// 索引状态类型
+// ============================================================================
+
+/**
+ * 单个文档的索引状态
+ */
+export interface DocumentIndexingStatus {
+    id: string;
+    indexing_status: IndexingStatus;
+    processing_started_at: number | null;
+    parsing_completed_at: number | null;
+    cleaning_completed_at: number | null;
+    splitting_completed_at: number | null;
+    completed_at: number | null;
+    paused_at: number | null;
+    error: string | null;
+    stopped_at: number | null;
+    completed_segments: number;
+    total_segments: number;
+}
+
+/**
+ * 批量文档索引状态响应
+ */
+export interface IndexingStatusResponse {
+    data: DocumentIndexingStatus[];
+}
+
+// ============================================================================
+// 上传文件信息类型
+// ============================================================================
+
+/**
+ * 上传文件信息
+ */
+export interface UploadFileInfo {
+    id: string;
+    name: string;
+    size: number;
+    extension: string;
+    url: string;
+    download_url: string;
+    mime_type: string;
+    created_by: string;
+    created_at: number;
+}
+
+// ============================================================================
+// 知识库更新类型
+// ============================================================================
+
+/**
+ * 检索模型配置
+ */
+export interface RetrievalModel {
+    search_method: 'keyword_search' | 'semantic_search' | 'full_text_search' | 'hybrid_search';
+    reranking_enable?: boolean;
+    reranking_mode?: string | null;
+    reranking_model?: {
+        reranking_provider_name: string;
+        reranking_model_name: string;
+    };
+    weights?: number | null;
+    top_k?: number;
+    score_threshold_enabled?: boolean;
+    score_threshold?: number | null;
+}
+
+/**
+ * 更新知识库请求参数
+ */
+export interface UpdateDatasetRequest {
+    name?: string;
+    description?: string;
+    indexing_technique?: 'high_quality' | 'economy';
+    permission?: 'only_me' | 'all_team_members' | 'partial_members';
+    embedding_model_provider?: string;
+    embedding_model?: string;
+    retrieval_model?: RetrievalModel;
+    partial_member_list?: string[];
+}