feat：完成上传文档时调整嵌入参数模块初版
This commit is contained in:
PingChuan
@@ -121,7 +121,7 @@ export async function toggleDocumentStatus(
|
||||
}
|
||||
|
||||
/**
|
||||
* 上传文件到知识库
|
||||
* 上传文件到知识库（使用默认配置）
|
||||
*
|
||||
* @param datasetId - 知识库 ID
|
||||
* @param file - 文件对象
|
||||
@@ -132,19 +132,44 @@ export async function uploadDocument(
|
||||
datasetId: string,
|
||||
file: File,
|
||||
onProgress?: (percent: number) => void
|
||||
): Promise<any> {
|
||||
): Promise<UploadDocumentResponse> {
|
||||
return uploadDocumentWithConfig(
|
||||
datasetId,
|
||||
file,
|
||||
{
|
||||
indexing_technique: 'high_quality',
|
||||
process_rule: { mode: 'automatic' },
|
||||
},
|
||||
onProgress
|
||||
);
|
||||
}
|
||||
|
||||
/**
|
||||
* 上传文件到知识库（使用自定义配置）
|
||||
*
|
||||
* @param datasetId - 知识库 ID
|
||||
* @param file - 文件对象
|
||||
* @param config - 上传配置（索引模式、分段规则等）
|
||||
* @param onProgress - 上传进度回调
|
||||
* @returns 创建的文档信息，包含 batch 用于查询索引进度
|
||||
*/
|
||||
export async function uploadDocumentWithConfig(
|
||||
datasetId: string,
|
||||
file: File,
|
||||
config: UploadDocumentConfig,
|
||||
onProgress?: (percent: number) => void
|
||||
): Promise<UploadDocumentResponse> {
|
||||
const formData = new FormData();
|
||||
formData.append('file', file);
|
||||
formData.append('data', JSON.stringify({
|
||||
indexing_technique: 'high_quality',
|
||||
process_rule: {
|
||||
mode: 'automatic',
|
||||
},
|
||||
}));
|
||||
formData.append('data', JSON.stringify(config));
|
||||
|
||||
console.log('[Dataset Client] 上传文档:', { datasetId, fileName: file.name });
|
||||
console.log('[Dataset Client] 上传文档:', {
|
||||
datasetId,
|
||||
fileName: file.name,
|
||||
config,
|
||||
});
|
||||
|
||||
const response = await axios.post(
|
||||
const response = await axios.post<UploadDocumentResponse>(
|
||||
`${API_URL}/datasets/${datasetId}/documents`,
|
||||
formData,
|
||||
{
|
||||
@@ -200,21 +225,41 @@ export async function fetchUploadFileInfo(
|
||||
return response.data;
|
||||
}
|
||||
|
||||
/**
|
||||
* 预处理规则 ID
|
||||
*/
|
||||
export type PreProcessingRuleId = 'remove_extra_spaces' | 'remove_urls_emails';
|
||||
|
||||
/**
|
||||
* 预处理规则配置
|
||||
*/
|
||||
export interface PreProcessingRule {
|
||||
id: PreProcessingRuleId;
|
||||
enabled: boolean;
|
||||
}
|
||||
|
||||
/**
|
||||
* 分段配置
|
||||
*/
|
||||
export interface SegmentationConfig {
|
||||
separator: string;
|
||||
max_tokens: number;
|
||||
}
|
||||
|
||||
/**
|
||||
* 自定义处理规则
|
||||
*/
|
||||
export interface CustomRules {
|
||||
pre_processing_rules?: PreProcessingRule[];
|
||||
segmentation?: SegmentationConfig;
|
||||
}
|
||||
|
||||
/**
|
||||
* 文档处理规则配置
|
||||
*/
|
||||
export interface ProcessRule {
|
||||
mode: 'automatic' | 'custom';
|
||||
rules?: {
|
||||
pre_processing_rules?: Array<{
|
||||
id: 'remove_extra_spaces' | 'remove_urls_emails';
|
||||
enabled: boolean;
|
||||
}>;
|
||||
segmentation?: {
|
||||
separator: string;
|
||||
max_tokens: number;
|
||||
};
|
||||
};
|
||||
rules?: CustomRules;
|
||||
}
|
||||
|
||||
/**
|
||||
@@ -225,6 +270,45 @@ export interface UpdateDocumentSettings {
|
||||
process_rule?: ProcessRule;
|
||||
}
|
||||
|
||||
/**
|
||||
* 上传文档配置参数
|
||||
*/
|
||||
export interface UploadDocumentConfig {
|
||||
indexing_technique: 'high_quality' | 'economy';
|
||||
process_rule: ProcessRule;
|
||||
}
|
||||
|
||||
/**
|
||||
* 上传文档响应
|
||||
*/
|
||||
export interface UploadDocumentResponse {
|
||||
document: {
|
||||
id: string;
|
||||
position: number;
|
||||
data_source_type: string;
|
||||
data_source_info: {
|
||||
upload_file_id: string;
|
||||
};
|
||||
dataset_process_rule_id: string;
|
||||
name: string;
|
||||
created_from: string;
|
||||
created_by: string;
|
||||
created_at: number;
|
||||
tokens: number;
|
||||
indexing_status: string;
|
||||
error: string | null;
|
||||
enabled: boolean;
|
||||
disabled_at: number | null;
|
||||
disabled_by: string | null;
|
||||
archived: boolean;
|
||||
display_status: string;
|
||||
word_count: number;
|
||||
hit_count: number;
|
||||
doc_form: string;
|
||||
};
|
||||
batch: string;
|
||||
}
|
||||
|
||||
/**
|
||||
* 更新文档设置并重新处理
|
||||
* 注意：Dify API 不直接支持修改已有文档的分段设置
|
||||
@@ -252,3 +336,50 @@ export async function updateDocumentWithSettings(
|
||||
);
|
||||
return response.data;
|
||||
}
|
||||
|
||||
/**
|
||||
* 通过文件更新文档
|
||||
* Dify API: POST /datasets/{dataset_id}/documents/{document_id}/update-by-file
|
||||
*
|
||||
* 用于在用户修改分段参数后，使用同一文件重新处理文档
|
||||
*
|
||||
* @param datasetId - 知识库 ID
|
||||
* @param documentId - 文档 ID
|
||||
* @param file - 文件对象（使用原上传的文件）
|
||||
* @param config - 新的分段配置
|
||||
* @param onProgress - 上传进度回调
|
||||
* @returns 更新后的文档信息，包含新的 batch 用于查询索引进度
|
||||
*/
|
||||
export async function updateDocumentByFile(
|
||||
datasetId: string,
|
||||
documentId: string,
|
||||
file: File,
|
||||
config: UploadDocumentConfig,
|
||||
onProgress?: (percent: number) => void
|
||||
): Promise<UploadDocumentResponse> {
|
||||
const formData = new FormData();
|
||||
formData.append('file', file);
|
||||
formData.append('data', JSON.stringify(config));
|
||||
|
||||
console.log('[Dataset Client] 通过文件更新文档:', {
|
||||
datasetId,
|
||||
documentId,
|
||||
fileName: file.name,
|
||||
config,
|
||||
});
|
||||
|
||||
const response = await axios.post<UploadDocumentResponse>(
|
||||
`${API_URL}/datasets/${datasetId}/documents/${documentId}/update-by-file`,
|
||||
formData,
|
||||
{
|
||||
withCredentials: true,
|
||||
onUploadProgress: (progressEvent) => {
|
||||
if (progressEvent.total && onProgress) {
|
||||
const percent = Math.round((progressEvent.loaded * 100) / progressEvent.total);
|
||||
onProgress(percent);
|
||||
}
|
||||
},
|
||||
}
|
||||
);
|
||||
return response.data;
|
||||
}
|
||||
|
||||
@@ -18,10 +18,19 @@ export {
|
||||
deleteDocument,
|
||||
toggleDocumentStatus,
|
||||
uploadDocument,
|
||||
uploadDocumentWithConfig,
|
||||
updateDocumentByFile,
|
||||
fetchIndexingStatus,
|
||||
fetchUploadFileInfo,
|
||||
} from './documentApi';
|
||||
|
||||
// 文档 API 类型
|
||||
export type {
|
||||
ProcessRule,
|
||||
UploadDocumentConfig,
|
||||
UploadDocumentResponse,
|
||||
} from './documentApi';
|
||||
|
||||
// 分段、子分段、检索 API
|
||||
export {
|
||||
fetchSegments,
|
||||
|
||||
@@ -11,7 +11,6 @@ import {
|
||||
message,
|
||||
Empty,
|
||||
Spin,
|
||||
Upload,
|
||||
} from 'antd';
|
||||
import {
|
||||
SearchOutlined,
|
||||
@@ -28,7 +27,8 @@ import {
|
||||
} from '@ant-design/icons';
|
||||
import type { ColumnsType } from 'antd/es/table';
|
||||
import type { Document, IndexingStatus } from '~/api/dify-dataset/type/documentTypes';
|
||||
import { deleteDocument, toggleDocumentStatus, uploadDocument } from '~/api/dify-dataset/api/documentApi';
|
||||
import { deleteDocument, toggleDocumentStatus } from '~/api/dify-dataset/api/documentApi';
|
||||
import DocumentUpload from './document-upload';
|
||||
import '../../styles/components/dify-dataset-manager/index.css';
|
||||
|
||||
interface DocumentListProps {
|
||||
@@ -63,9 +63,11 @@ export default function DocumentList({
|
||||
onViewDocument,
|
||||
}: DocumentListProps) {
|
||||
const [searchValue, setSearchValue] = useState('');
|
||||
const [uploading, setUploading] = useState(false);
|
||||
const [deletingId, setDeletingId] = useState<string | null>(null);
|
||||
|
||||
// 显示上传页面的状态
|
||||
const [showUploadPage, setShowUploadPage] = useState(false);
|
||||
|
||||
/**
|
||||
* 获取状态标签配置
|
||||
*/
|
||||
@@ -141,29 +143,29 @@ export default function DocumentList({
|
||||
};
|
||||
|
||||
/**
|
||||
* 处理文件上传
|
||||
* 点击上传按钮，显示上传页面
|
||||
*/
|
||||
const handleUpload = async (file: File) => {
|
||||
const handleUploadClick = () => {
|
||||
if (!datasetId) {
|
||||
message.error('请先选择知识库');
|
||||
return false;
|
||||
return;
|
||||
}
|
||||
setShowUploadPage(true);
|
||||
};
|
||||
|
||||
setUploading(true);
|
||||
try {
|
||||
await uploadDocument(datasetId, file, (percent) => {
|
||||
console.log('上传进度:', percent);
|
||||
});
|
||||
message.success('上传成功，正在处理...');
|
||||
onRefresh();
|
||||
} catch (err: any) {
|
||||
console.error('上传文件失败:', err);
|
||||
message.error(err.message || '上传失败');
|
||||
} finally {
|
||||
setUploading(false);
|
||||
}
|
||||
/**
|
||||
* 关闭上传页面
|
||||
*/
|
||||
const handleUploadClose = () => {
|
||||
setShowUploadPage(false);
|
||||
};
|
||||
|
||||
return false;
|
||||
/**
|
||||
* 上传成功回调
|
||||
*/
|
||||
const handleUploadSuccess = () => {
|
||||
setShowUploadPage(false);
|
||||
onRefresh();
|
||||
};
|
||||
|
||||
// 过滤文档
|
||||
@@ -271,115 +273,114 @@ export default function DocumentList({
|
||||
];
|
||||
|
||||
return (
|
||||
<div className="document-list-page">
|
||||
{/* 页面头部 */}
|
||||
<div className="page-header">
|
||||
<div className="header-left">
|
||||
<h1>文档</h1>
|
||||
{/* <p className="page-description">
|
||||
知识库的所有文件都在这里显示，整个知识库都可以被接到 Dify 引用或通过 Chat 插件进行索引。
|
||||
</p> */}
|
||||
</div>
|
||||
<div className="header-actions">
|
||||
<Tooltip title="刷新">
|
||||
<Button
|
||||
icon={<ReloadOutlined />}
|
||||
onClick={onRefresh}
|
||||
loading={loading}
|
||||
/>
|
||||
</Tooltip>
|
||||
<Upload
|
||||
beforeUpload={handleUpload}
|
||||
showUploadList={false}
|
||||
accept=".txt,.md,.pdf,.docx,.doc,.csv,.xlsx,.xls"
|
||||
disabled={!datasetId}
|
||||
>
|
||||
<Button
|
||||
type="primary"
|
||||
icon={<CloudUploadOutlined />}
|
||||
loading={uploading}
|
||||
disabled={!datasetId}
|
||||
>
|
||||
添加文件
|
||||
</Button>
|
||||
</Upload>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
{/* 搜索栏 */}
|
||||
<div className="document-search-bar">
|
||||
<Input
|
||||
placeholder="搜索文档..."
|
||||
prefix={<SearchOutlined />}
|
||||
value={searchValue}
|
||||
onChange={(e) => setSearchValue(e.target.value)}
|
||||
allowClear
|
||||
style={{ width: 280 }}
|
||||
<>
|
||||
{/* 上传页面 */}
|
||||
{showUploadPage ? (
|
||||
<DocumentUpload
|
||||
datasetId={datasetId}
|
||||
onClose={handleUploadClose}
|
||||
onSuccess={handleUploadSuccess}
|
||||
/>
|
||||
</div>
|
||||
|
||||
{/* 文档表格 */}
|
||||
<div className="document-table-wrapper">
|
||||
{loading && documents.length === 0 ? (
|
||||
<div className="loading-state">
|
||||
<Spin size="large" />
|
||||
<div className="loading-text">加载中...</div>
|
||||
) : (
|
||||
<div className="document-list-page">
|
||||
{/* 页面头部 */}
|
||||
<div className="page-header">
|
||||
<div className="header-left">
|
||||
<h1>文档</h1>
|
||||
</div>
|
||||
<div className="header-actions">
|
||||
<Tooltip title="刷新">
|
||||
<Button
|
||||
icon={<ReloadOutlined />}
|
||||
onClick={onRefresh}
|
||||
loading={loading}
|
||||
/>
|
||||
</Tooltip>
|
||||
<Button
|
||||
type="primary"
|
||||
icon={<CloudUploadOutlined />}
|
||||
onClick={handleUploadClick}
|
||||
disabled={!datasetId}
|
||||
>
|
||||
添加文件
|
||||
</Button>
|
||||
</div>
|
||||
</div>
|
||||
) : filteredDocuments.length === 0 ? (
|
||||
<div className="empty-state">
|
||||
<Empty description={searchValue ? '未找到匹配的文档' : '暂无文档'}>
|
||||
{!searchValue && (
|
||||
<Upload
|
||||
beforeUpload={handleUpload}
|
||||
showUploadList={false}
|
||||
accept=".txt,.md,.pdf,.docx,.doc,.csv,.xlsx,.xls"
|
||||
|
||||
{/* 搜索栏 */}
|
||||
<div className="document-search-bar">
|
||||
<Input
|
||||
placeholder="搜索文档..."
|
||||
prefix={<SearchOutlined />}
|
||||
value={searchValue}
|
||||
onChange={(e) => setSearchValue(e.target.value)}
|
||||
allowClear
|
||||
style={{ width: 280 }}
|
||||
/>
|
||||
</div>
|
||||
|
||||
{/* 文档表格 */}
|
||||
<div className="document-table-wrapper">
|
||||
{loading && documents.length === 0 ? (
|
||||
<div className="loading-state">
|
||||
<Spin size="large" />
|
||||
<div className="loading-text">加载中...</div>
|
||||
</div>
|
||||
) : filteredDocuments.length === 0 ? (
|
||||
<div className="empty-state">
|
||||
<Empty description={searchValue ? '未找到匹配的文档' : '暂无文档'}>
|
||||
{!searchValue && (
|
||||
<Button
|
||||
type="primary"
|
||||
icon={<CloudUploadOutlined />}
|
||||
onClick={handleUploadClick}
|
||||
>
|
||||
上传第一个文档
|
||||
</Button>
|
||||
)}
|
||||
</Empty>
|
||||
</div>
|
||||
) : (
|
||||
<Table
|
||||
className="document-table"
|
||||
columns={columns}
|
||||
dataSource={filteredDocuments}
|
||||
rowKey="id"
|
||||
loading={loading}
|
||||
pagination={false}
|
||||
size="small"
|
||||
scroll={{ x: 'max-content' }}
|
||||
/>
|
||||
)}
|
||||
</div>
|
||||
|
||||
{/* 底部分页器 */}
|
||||
{filteredDocuments.length > 0 && (
|
||||
<div className="document-pagination">
|
||||
<span className="pagination-total">共 {total} 条</span>
|
||||
<div className="pagination-controls">
|
||||
<Button
|
||||
size="small"
|
||||
disabled={page <= 1}
|
||||
onClick={() => onPageChange(page - 1)}
|
||||
>
|
||||
<Button type="primary" icon={<CloudUploadOutlined />}>
|
||||
上传第一个文档
|
||||
</Button>
|
||||
</Upload>
|
||||
)}
|
||||
</Empty>
|
||||
</div>
|
||||
) : (
|
||||
<Table
|
||||
className="document-table"
|
||||
columns={columns}
|
||||
dataSource={filteredDocuments}
|
||||
rowKey="id"
|
||||
loading={loading}
|
||||
pagination={false}
|
||||
size="small"
|
||||
scroll={{ x: 'max-content' }}
|
||||
/>
|
||||
)}
|
||||
</div>
|
||||
|
||||
{/* 底部分页器 */}
|
||||
{filteredDocuments.length > 0 && (
|
||||
<div className="document-pagination">
|
||||
<span className="pagination-total">共 {total} 条</span>
|
||||
<div className="pagination-controls">
|
||||
<Button
|
||||
size="small"
|
||||
disabled={page <= 1}
|
||||
onClick={() => onPageChange(page - 1)}
|
||||
>
|
||||
上一页
|
||||
</Button>
|
||||
<span className="pagination-info">
|
||||
第 {page} 页 / 共 {Math.ceil(total / pageSize)} 页
|
||||
</span>
|
||||
<Button
|
||||
size="small"
|
||||
disabled={page >= Math.ceil(total / pageSize)}
|
||||
onClick={() => onPageChange(page + 1)}
|
||||
>
|
||||
下一页
|
||||
</Button>
|
||||
</div>
|
||||
上一页
|
||||
</Button>
|
||||
<span className="pagination-info">
|
||||
第 {page} 页 / 共 {Math.ceil(total / pageSize)} 页
|
||||
</span>
|
||||
<Button
|
||||
size="small"
|
||||
disabled={page >= Math.ceil(total / pageSize)}
|
||||
onClick={() => onPageChange(page + 1)}
|
||||
>
|
||||
下一页
|
||||
</Button>
|
||||
</div>
|
||||
</div>
|
||||
)}
|
||||
</div>
|
||||
)}
|
||||
</div>
|
||||
</>
|
||||
);
|
||||
}
|
||||
|
||||
@@ -0,0 +1,914 @@
|
||||
import { useState, useEffect, useRef, useCallback } from 'react';
|
||||
import {
|
||||
Input,
|
||||
Button,
|
||||
InputNumber,
|
||||
Checkbox,
|
||||
Select,
|
||||
Card,
|
||||
Empty,
|
||||
Spin,
|
||||
message,
|
||||
Divider,
|
||||
Tooltip,
|
||||
Progress,
|
||||
Upload,
|
||||
} from 'antd';
|
||||
import type { UploadFile, UploadProps } from 'antd';
|
||||
import {
|
||||
QuestionCircleOutlined,
|
||||
FileTextOutlined,
|
||||
CheckCircleOutlined,
|
||||
LoadingOutlined,
|
||||
ExclamationCircleOutlined,
|
||||
InboxOutlined,
|
||||
DeleteOutlined,
|
||||
ArrowLeftOutlined,
|
||||
} from '@ant-design/icons';
|
||||
import type { IndexingStatus } from '~/api/dify-dataset/type/documentTypes';
|
||||
import type { Segment } from '~/api/dify-dataset/type';
|
||||
import {
|
||||
uploadDocumentWithConfig,
|
||||
updateDocumentByFile,
|
||||
fetchIndexingStatus,
|
||||
} from '~/api/dify-dataset/api/documentApi';
|
||||
import { fetchSegments } from '~/api/dify-dataset/api/segmentApi';
|
||||
|
||||
const { Dragger } = Upload;
|
||||
|
||||
interface DocumentUploadProps {
|
||||
datasetId: string;
|
||||
onClose: () => void;
|
||||
onSuccess: () => void;
|
||||
}
|
||||
|
||||
/**
|
||||
* 分段设置配置
|
||||
*/
|
||||
interface SegmentationSettings {
|
||||
separator: string;
|
||||
maxTokens: number;
|
||||
chunkOverlap: number;
|
||||
removeExtraSpaces: boolean;
|
||||
removeUrlsEmails: boolean;
|
||||
indexingTechnique: 'high_quality' | 'economy';
|
||||
}
|
||||
|
||||
/**
|
||||
* 默认分段设置
|
||||
*/
|
||||
const DEFAULT_SETTINGS: SegmentationSettings = {
|
||||
separator: '\\n\\n',
|
||||
maxTokens: 1024,
|
||||
chunkOverlap: 50,
|
||||
removeExtraSpaces: true,
|
||||
removeUrlsEmails: false,
|
||||
indexingTechnique: 'high_quality',
|
||||
};
|
||||
|
||||
/**
|
||||
* 单个文档的上传状态
|
||||
*/
|
||||
type DocumentStage = 'pending' | 'uploading' | 'indexing' | 'completed' | 'error';
|
||||
|
||||
/**
|
||||
* 上传的文档信息（支持多文件）
|
||||
*/
|
||||
interface UploadedDocument {
|
||||
file: File;
|
||||
documentId: string;
|
||||
batch: string;
|
||||
stage: DocumentStage;
|
||||
indexingStatus: IndexingStatus;
|
||||
uploadProgress: number;
|
||||
error?: string;
|
||||
settings: SegmentationSettings;
|
||||
segments: Segment[];
|
||||
}
|
||||
|
||||
/**
|
||||
* 索引状态配置
|
||||
*/
|
||||
const INDEXING_STATUS_CONFIG: Record<IndexingStatus, { text: string; percent: number }> = {
|
||||
waiting: { text: '等待处理...', percent: 10 },
|
||||
parsing: { text: '解析文档...', percent: 30 },
|
||||
cleaning: { text: '清洗文本...', percent: 50 },
|
||||
splitting: { text: '分段处理...', percent: 70 },
|
||||
indexing: { text: '建立索引...', percent: 85 },
|
||||
completed: { text: '处理完成', percent: 100 },
|
||||
paused: { text: '已暂停', percent: 0 },
|
||||
error: { text: '处理失败', percent: 0 },
|
||||
};
|
||||
|
||||
/**
|
||||
* 支持的文件格式
|
||||
*/
|
||||
const SUPPORTED_FORMATS = 'TXT, MARKDOWN, MDX, PDF, HTML, XLSX, XLS, DOCX, CSV, VTT, PROPERTIES, MD, HTM';
|
||||
|
||||
/**
|
||||
* 文档上传组件
|
||||
* 支持多文件上传，两步流程：选择文件 → 上传并配置分段
|
||||
*/
|
||||
export default function DocumentUpload({
|
||||
datasetId,
|
||||
onClose,
|
||||
onSuccess,
|
||||
}: DocumentUploadProps) {
|
||||
// 步骤控制
|
||||
const [step, setStep] = useState<1 | 2>(1);
|
||||
|
||||
// 文件相关
|
||||
const [selectedFiles, setSelectedFiles] = useState<File[]>([]);
|
||||
const [fileList, setFileList] = useState<UploadFile[]>([]);
|
||||
|
||||
// 多文档状态管理
|
||||
const [uploadedDocuments, setUploadedDocuments] = useState<UploadedDocument[]>([]);
|
||||
// 当前选中查看的文档索引
|
||||
const [currentDocIndex, setCurrentDocIndex] = useState(0);
|
||||
|
||||
// 当前显示的分段设置（来自当前选中的文档）
|
||||
const [currentSettings, setCurrentSettings] = useState<SegmentationSettings>(DEFAULT_SETTINGS);
|
||||
|
||||
// 预览相关
|
||||
const [previewLoading, setPreviewLoading] = useState(false);
|
||||
|
||||
// 轮询定时器（支持多个文档）
|
||||
const pollingTimersRef = useRef<Map<string, NodeJS.Timeout>>(new Map());
|
||||
|
||||
// 清理所有轮询定时器
|
||||
useEffect(() => {
|
||||
return () => {
|
||||
pollingTimersRef.current.forEach(timer => clearInterval(timer));
|
||||
pollingTimersRef.current.clear();
|
||||
};
|
||||
}, []);
|
||||
|
||||
/**
|
||||
* 停止指定文档的轮询
|
||||
*/
|
||||
const stopPolling = useCallback((documentId: string) => {
|
||||
const timer = pollingTimersRef.current.get(documentId);
|
||||
if (timer) {
|
||||
clearInterval(timer);
|
||||
pollingTimersRef.current.delete(documentId);
|
||||
}
|
||||
}, []);
|
||||
|
||||
/**
|
||||
* 停止所有轮询
|
||||
*/
|
||||
const stopAllPolling = useCallback(() => {
|
||||
pollingTimersRef.current.forEach(timer => clearInterval(timer));
|
||||
pollingTimersRef.current.clear();
|
||||
}, []);
|
||||
|
||||
/**
|
||||
* 加载分段预览
|
||||
*/
|
||||
const loadSegmentsPreview = useCallback(async (documentId: string, docIndex: number) => {
|
||||
setPreviewLoading(true);
|
||||
try {
|
||||
const response = await fetchSegments(datasetId, documentId, 1, 50);
|
||||
const segments = response.data || [];
|
||||
// 更新对应文档的分段
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === docIndex ? { ...doc, segments } : doc
|
||||
));
|
||||
} catch (err: any) {
|
||||
console.error('加载分段预览失败:', err);
|
||||
message.error('加载分段预览失败');
|
||||
} finally {
|
||||
setPreviewLoading(false);
|
||||
}
|
||||
}, [datasetId]);
|
||||
|
||||
/**
|
||||
* 轮询索引状态
|
||||
*/
|
||||
const pollIndexingStatus = useCallback(async (batch: string, documentId: string, docIndex: number) => {
|
||||
try {
|
||||
const response = await fetchIndexingStatus(datasetId, batch);
|
||||
const documentStatus = response.data?.[0];
|
||||
|
||||
if (documentStatus) {
|
||||
const status = documentStatus.indexing_status as IndexingStatus;
|
||||
|
||||
// 更新文档状态
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) => {
|
||||
if (idx !== docIndex) return doc;
|
||||
return { ...doc, indexingStatus: status };
|
||||
}));
|
||||
|
||||
if (status === 'completed') {
|
||||
stopPolling(documentId);
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === docIndex ? { ...doc, stage: 'completed' } : doc
|
||||
));
|
||||
// message.success(`文档 "${uploadedDocuments[docIndex]?.file.name}" 处理完成！`);
|
||||
// 自动加载分段预览
|
||||
loadSegmentsPreview(documentId, docIndex);
|
||||
} else if (status === 'error') {
|
||||
stopPolling(documentId);
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === docIndex ? { ...doc, stage: 'error', error: documentStatus.error || '处理失败' } : doc
|
||||
));
|
||||
}
|
||||
}
|
||||
} catch (err) {
|
||||
console.error('获取索引状态失败:', err);
|
||||
}
|
||||
}, [datasetId, stopPolling, loadSegmentsPreview, uploadedDocuments]);
|
||||
|
||||
/**
|
||||
* 开始轮询
|
||||
*/
|
||||
const startPolling = useCallback((batch: string, documentId: string, docIndex: number) => {
|
||||
// 先停止之前的轮询
|
||||
stopPolling(documentId);
|
||||
|
||||
// 开始新的轮询
|
||||
const timer = setInterval(() => {
|
||||
pollIndexingStatus(batch, documentId, docIndex);
|
||||
}, 2000);
|
||||
pollingTimersRef.current.set(documentId, timer);
|
||||
|
||||
// 立即执行一次
|
||||
pollIndexingStatus(batch, documentId, docIndex);
|
||||
}, [stopPolling, pollIndexingStatus]);
|
||||
|
||||
/**
|
||||
* 构建上传配置
|
||||
*/
|
||||
const buildConfig = (s: SegmentationSettings) => ({
|
||||
indexing_technique: s.indexingTechnique,
|
||||
process_rule: {
|
||||
mode: 'custom' as const,
|
||||
rules: {
|
||||
pre_processing_rules: [
|
||||
{ id: 'remove_extra_spaces' as const, enabled: s.removeExtraSpaces },
|
||||
{ id: 'remove_urls_emails' as const, enabled: s.removeUrlsEmails },
|
||||
],
|
||||
segmentation: {
|
||||
separator: s.separator.replace(/\\n/g, '\n'),
|
||||
max_tokens: s.maxTokens,
|
||||
},
|
||||
},
|
||||
},
|
||||
});
|
||||
|
||||
/**
|
||||
* 更新当前文档的设置
|
||||
*/
|
||||
const updateCurrentSettings = (key: keyof SegmentationSettings, value: any) => {
|
||||
const newSettings = { ...currentSettings, [key]: value };
|
||||
setCurrentSettings(newSettings);
|
||||
// 同步更新到文档列表
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === currentDocIndex ? { ...doc, settings: newSettings } : doc
|
||||
));
|
||||
};
|
||||
|
||||
/**
|
||||
* 处理文件选择变化
|
||||
*/
|
||||
const handleFileChange: UploadProps['onChange'] = ({ fileList: newFileList }) => {
|
||||
setFileList(newFileList);
|
||||
// 提取实际文件对象
|
||||
const files = newFileList
|
||||
.filter(f => f.originFileObj)
|
||||
.map(f => f.originFileObj as File);
|
||||
setSelectedFiles(files);
|
||||
};
|
||||
|
||||
/**
|
||||
* 移除文件
|
||||
*/
|
||||
const handleRemoveFile = (file: UploadFile) => {
|
||||
const newFileList = fileList.filter(f => f.uid !== file.uid);
|
||||
setFileList(newFileList);
|
||||
const files = newFileList
|
||||
.filter(f => f.originFileObj)
|
||||
.map(f => f.originFileObj as File);
|
||||
setSelectedFiles(files);
|
||||
};
|
||||
|
||||
/**
|
||||
* 上传单个文件
|
||||
*/
|
||||
const uploadSingleFile = async (file: File, index: number): Promise<void> => {
|
||||
try {
|
||||
// 更新状态为上传中
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === index ? { ...doc, stage: 'uploading' as DocumentStage } : doc
|
||||
));
|
||||
|
||||
const config = buildConfig(DEFAULT_SETTINGS);
|
||||
const result = await uploadDocumentWithConfig(
|
||||
datasetId,
|
||||
file,
|
||||
config,
|
||||
(percent) => {
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === index ? { ...doc, uploadProgress: percent } : doc
|
||||
));
|
||||
}
|
||||
);
|
||||
|
||||
// 更新文档信息
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === index ? {
|
||||
...doc,
|
||||
documentId: result.document.id,
|
||||
batch: result.batch,
|
||||
stage: 'indexing' as DocumentStage,
|
||||
indexingStatus: 'waiting' as IndexingStatus,
|
||||
} : doc
|
||||
));
|
||||
|
||||
// 开始轮询索引状态
|
||||
startPolling(result.batch, result.document.id, index);
|
||||
} catch (err: any) {
|
||||
console.error(`上传文档 ${file.name} 失败:`, err);
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === index ? {
|
||||
...doc,
|
||||
stage: 'error' as DocumentStage,
|
||||
error: err.message || '上传失败',
|
||||
} : doc
|
||||
));
|
||||
}
|
||||
};
|
||||
|
||||
/**
|
||||
* 点击"下一步"：立即上传所有文件
|
||||
*/
|
||||
const handleNextStep = async () => {
|
||||
if (selectedFiles.length === 0) {
|
||||
message.warning('请先选择文件');
|
||||
return;
|
||||
}
|
||||
|
||||
// 初始化所有文档状态
|
||||
const docs: UploadedDocument[] = selectedFiles.map(file => ({
|
||||
file,
|
||||
documentId: '',
|
||||
batch: '',
|
||||
stage: 'pending' as DocumentStage,
|
||||
indexingStatus: 'waiting' as IndexingStatus,
|
||||
uploadProgress: 0,
|
||||
settings: { ...DEFAULT_SETTINGS },
|
||||
segments: [],
|
||||
}));
|
||||
setUploadedDocuments(docs);
|
||||
setCurrentDocIndex(0);
|
||||
setCurrentSettings({ ...DEFAULT_SETTINGS });
|
||||
setStep(2);
|
||||
|
||||
// 依次上传所有文件
|
||||
for (let i = 0; i < selectedFiles.length; i++) {
|
||||
await uploadSingleFile(selectedFiles[i], i);
|
||||
}
|
||||
};
|
||||
|
||||
/**
|
||||
* 切换查看的文档
|
||||
*/
|
||||
const handleDocumentChange = (docId: string) => {
|
||||
const index = uploadedDocuments.findIndex(doc => doc.documentId === docId || doc.file.name === docId);
|
||||
if (index !== -1) {
|
||||
setCurrentDocIndex(index);
|
||||
const doc = uploadedDocuments[index];
|
||||
setCurrentSettings(doc.settings);
|
||||
}
|
||||
};
|
||||
|
||||
/**
|
||||
* 修改参数后重新处理当前文档
|
||||
*/
|
||||
const handleReprocess = async () => {
|
||||
const currentDoc = uploadedDocuments[currentDocIndex];
|
||||
if (!currentDoc || !currentDoc.documentId) return;
|
||||
|
||||
// 更新状态
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === currentDocIndex ? {
|
||||
...doc,
|
||||
stage: 'uploading' as DocumentStage,
|
||||
uploadProgress: 0,
|
||||
segments: [],
|
||||
} : doc
|
||||
));
|
||||
|
||||
try {
|
||||
const config = buildConfig(currentSettings);
|
||||
const result = await updateDocumentByFile(
|
||||
datasetId,
|
||||
currentDoc.documentId,
|
||||
currentDoc.file,
|
||||
config,
|
||||
(percent) => {
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === currentDocIndex ? { ...doc, uploadProgress: percent } : doc
|
||||
));
|
||||
}
|
||||
);
|
||||
|
||||
// 更新 batch
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === currentDocIndex ? {
|
||||
...doc,
|
||||
batch: result.batch,
|
||||
stage: 'indexing' as DocumentStage,
|
||||
indexingStatus: 'waiting' as IndexingStatus,
|
||||
} : doc
|
||||
));
|
||||
|
||||
startPolling(result.batch, currentDoc.documentId, currentDocIndex);
|
||||
} catch (err: any) {
|
||||
console.error('重新处理失败:', err);
|
||||
setUploadedDocuments(prev => prev.map((doc, idx) =>
|
||||
idx === currentDocIndex ? {
|
||||
...doc,
|
||||
stage: 'error' as DocumentStage,
|
||||
error: err.message || '重新处理失败',
|
||||
} : doc
|
||||
));
|
||||
message.error(err.message || '重新处理失败');
|
||||
}
|
||||
};
|
||||
|
||||
/**
|
||||
* 返回上一步
|
||||
*/
|
||||
const handlePrevStep = () => {
|
||||
// 检查是否有文档正在处理
|
||||
const hasProcessing = uploadedDocuments.some(doc =>
|
||||
doc.stage === 'uploading' || doc.stage === 'indexing'
|
||||
);
|
||||
if (hasProcessing) {
|
||||
message.warning('还有文档正在处理中，请等待完成');
|
||||
return;
|
||||
}
|
||||
stopAllPolling();
|
||||
setStep(1);
|
||||
setUploadedDocuments([]);
|
||||
setCurrentDocIndex(0);
|
||||
setCurrentSettings(DEFAULT_SETTINGS);
|
||||
};
|
||||
|
||||
/**
|
||||
* 返回文档列表
|
||||
*/
|
||||
const handleGoToDocuments = () => {
|
||||
stopAllPolling();
|
||||
const hasCompleted = uploadedDocuments.some(doc => doc.stage === 'completed');
|
||||
if (hasCompleted) {
|
||||
onSuccess();
|
||||
}
|
||||
onClose();
|
||||
};
|
||||
|
||||
/**
|
||||
* 获取当前文档
|
||||
*/
|
||||
const getCurrentDocument = (): UploadedDocument | null => {
|
||||
return uploadedDocuments[currentDocIndex] || null;
|
||||
};
|
||||
|
||||
/**
|
||||
* 获取当前文档的进度
|
||||
*/
|
||||
const getCurrentProgress = () => {
|
||||
const doc = getCurrentDocument();
|
||||
if (!doc) return 0;
|
||||
if (doc.stage === 'uploading') {
|
||||
return doc.uploadProgress;
|
||||
}
|
||||
if (doc.stage === 'indexing' || doc.stage === 'completed') {
|
||||
return INDEXING_STATUS_CONFIG[doc.indexingStatus]?.percent || 0;
|
||||
}
|
||||
return 0;
|
||||
};
|
||||
|
||||
/**
|
||||
* 获取当前文档的状态文本
|
||||
*/
|
||||
const getStatusText = () => {
|
||||
const doc = getCurrentDocument();
|
||||
if (!doc) return '';
|
||||
if (doc.stage === 'uploading') {
|
||||
return `正在上传... ${doc.uploadProgress}%`;
|
||||
}
|
||||
if (doc.stage === 'indexing') {
|
||||
return INDEXING_STATUS_CONFIG[doc.indexingStatus]?.text || '处理中...';
|
||||
}
|
||||
if (doc.stage === 'completed') {
|
||||
return '处理完成';
|
||||
}
|
||||
if (doc.stage === 'error') {
|
||||
return doc.error || '处理失败';
|
||||
}
|
||||
return '';
|
||||
};
|
||||
|
||||
/**
|
||||
* 判断当前文档是否正在处理
|
||||
*/
|
||||
const isCurrentDocProcessing = () => {
|
||||
const doc = getCurrentDocument();
|
||||
return doc?.stage === 'uploading' || doc?.stage === 'indexing';
|
||||
};
|
||||
|
||||
/**
|
||||
* 获取所有文档的完成状态统计
|
||||
*/
|
||||
const getCompletionStats = () => {
|
||||
const completed = uploadedDocuments.filter(doc => doc.stage === 'completed').length;
|
||||
const total = uploadedDocuments.length;
|
||||
return { completed, total };
|
||||
};
|
||||
|
||||
/**
|
||||
* 渲染步骤指示器（两步流程）
|
||||
*/
|
||||
const renderSteps = () => (
|
||||
<div className="upload-steps">
|
||||
<div className={`step-item ${step === 1 ? 'active' : ''} ${step > 1 ? 'completed' : ''}`}>
|
||||
<span className="step-number">1</span>
|
||||
<span className="step-title">选择数据源</span>
|
||||
</div>
|
||||
<div className={`step-divider ${step > 1 ? 'completed' : ''}`}></div>
|
||||
<div className={`step-item ${step === 2 ? 'active' : ''}`}>
|
||||
<span className="step-number">2</span>
|
||||
<span className="step-title">文本分段与清洗</span>
|
||||
</div>
|
||||
</div>
|
||||
);
|
||||
|
||||
/**
|
||||
* 渲染第一步：选择文件（支持多文件）
|
||||
*/
|
||||
const renderStep1 = () => (
|
||||
<div className="upload-step-content step1">
|
||||
<h2 className="step-heading">上传文本文件</h2>
|
||||
<p className="step-description">
|
||||
文档需上传至知识智能理解法治知识库，广东烟草智能理解将按照于知识库，你可以在聊后指数文档所据案中检索它
|
||||
</p>
|
||||
|
||||
<div className="file-drop-zone">
|
||||
<Dragger
|
||||
fileList={fileList}
|
||||
onChange={handleFileChange}
|
||||
beforeUpload={() => false}
|
||||
multiple={true}
|
||||
accept=".txt,.md,.mdx,.pdf,.html,.htm,.xlsx,.xls,.docx,.csv,.vtt,.properties"
|
||||
showUploadList={false}
|
||||
>
|
||||
<p className="ant-upload-drag-icon">
|
||||
<InboxOutlined />
|
||||
</p>
|
||||
<p className="ant-upload-text">拖拽文件或文件夹至此，或者 <span className="upload-link">选择文件</span></p>
|
||||
<p className="ant-upload-hint">
|
||||
已支持 {SUPPORTED_FORMATS}，每个文件不超过 15MB。支持批量上传多个文件。
|
||||
</p>
|
||||
</Dragger>
|
||||
</div>
|
||||
|
||||
{/* 已选文件列表 */}
|
||||
{selectedFiles.length > 0 && (
|
||||
<div className="selected-files-section">
|
||||
<h3 className="section-subtitle">嵌入已就绪 ({selectedFiles.length} 个文件)</h3>
|
||||
<div className="selected-files-list">
|
||||
{fileList.map((file) => (
|
||||
<div key={file.uid} className="selected-file-item">
|
||||
<FileTextOutlined className="file-icon" />
|
||||
<div className="file-info">
|
||||
<span className="file-name">{file.name}</span>
|
||||
<span className="file-size">
|
||||
{file.originFileObj
|
||||
? `${file.originFileObj.type?.split('/')[1]?.toUpperCase() || 'FILE'}，${(file.originFileObj.size / 1024 / 1024).toFixed(2)}MB`
|
||||
: ''}
|
||||
</span>
|
||||
</div>
|
||||
<Button
|
||||
type="text"
|
||||
icon={<DeleteOutlined />}
|
||||
onClick={() => handleRemoveFile(file)}
|
||||
className="remove-file-btn"
|
||||
/>
|
||||
</div>
|
||||
))}
|
||||
</div>
|
||||
</div>
|
||||
)}
|
||||
|
||||
<div className="step-actions">
|
||||
<Button
|
||||
type="primary"
|
||||
onClick={handleNextStep}
|
||||
disabled={selectedFiles.length === 0}
|
||||
className="next-btn"
|
||||
>
|
||||
下一步 →
|
||||
</Button>
|
||||
</div>
|
||||
</div>
|
||||
);
|
||||
|
||||
/**
|
||||
* 渲染第二步：分段配置与预览
|
||||
* 左侧始终显示配置面板，右侧预览框内显示进度或分段内容
|
||||
*/
|
||||
const renderStep2 = () => {
|
||||
const currentDoc = getCurrentDocument();
|
||||
const isProcessing = isCurrentDocProcessing();
|
||||
const stats = getCompletionStats();
|
||||
|
||||
return (
|
||||
<div className="upload-step-content step2">
|
||||
{/* 分段配置与预览 */}
|
||||
<div className="document-detail-content">
|
||||
{/* 左侧设置区域 */}
|
||||
<div className="settings-panel">
|
||||
<div className="settings-section">
|
||||
<h3 className="section-title">分段设置</h3>
|
||||
|
||||
{/* 分段标识符 */}
|
||||
<div className="setting-item">
|
||||
<label className="setting-label">
|
||||
分段标识符
|
||||
<Tooltip title="系统会在遇到指定分隔符时自动分段，默认值为 \n\n（按段落分段）">
|
||||
<QuestionCircleOutlined className="help-icon" />
|
||||
</Tooltip>
|
||||
</label>
|
||||
<Input
|
||||
value={currentSettings.separator}
|
||||
onChange={(e) => updateCurrentSettings('separator', e.target.value)}
|
||||
placeholder="\n\n"
|
||||
className="setting-input"
|
||||
disabled={isProcessing}
|
||||
/>
|
||||
</div>
|
||||
|
||||
{/* 分段最大长度 */}
|
||||
<div className="setting-item">
|
||||
<label className="setting-label">
|
||||
分段最大长度
|
||||
<Tooltip title="指定每个分段允许的最大字符数（100-4000），超过此限制系统会强制分段">
|
||||
<QuestionCircleOutlined className="help-icon" />
|
||||
</Tooltip>
|
||||
</label>
|
||||
<div className="setting-input-with-suffix">
|
||||
<InputNumber
|
||||
value={currentSettings.maxTokens}
|
||||
onChange={(value) => updateCurrentSettings('maxTokens', value || 1024)}
|
||||
min={100}
|
||||
max={4000}
|
||||
className="setting-input-number"
|
||||
disabled={isProcessing}
|
||||
/>
|
||||
<span className="input-suffix">characters</span>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
{/* 分段重叠长度 */}
|
||||
<div className="setting-item">
|
||||
<label className="setting-label">
|
||||
分段重叠长度
|
||||
<Tooltip title="相邻分段之间重叠的字符数，有助于保持上下文连贯性">
|
||||
<QuestionCircleOutlined className="help-icon" />
|
||||
</Tooltip>
|
||||
</label>
|
||||
<div className="setting-input-with-suffix">
|
||||
<InputNumber
|
||||
value={currentSettings.chunkOverlap}
|
||||
onChange={(value) => updateCurrentSettings('chunkOverlap', value || 50)}
|
||||
min={0}
|
||||
max={500}
|
||||
className="setting-input-number"
|
||||
disabled={isProcessing}
|
||||
/>
|
||||
<span className="input-suffix">characters</span>
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<Divider />
|
||||
|
||||
{/* 文本预处理规则 */}
|
||||
<div className="settings-section">
|
||||
<h3 className="section-title">文本预处理规则</h3>
|
||||
<div className="checkbox-group">
|
||||
<Checkbox
|
||||
checked={currentSettings.removeExtraSpaces}
|
||||
onChange={(e) => updateCurrentSettings('removeExtraSpaces', e.target.checked)}
|
||||
disabled={isProcessing}
|
||||
>
|
||||
替换掉连续的空格、换行符和制表符
|
||||
</Checkbox>
|
||||
<Checkbox
|
||||
checked={currentSettings.removeUrlsEmails}
|
||||
onChange={(e) => updateCurrentSettings('removeUrlsEmails', e.target.checked)}
|
||||
disabled={isProcessing}
|
||||
>
|
||||
删除所有 URL 和电子邮件地址
|
||||
</Checkbox>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<Divider />
|
||||
|
||||
{/* 索引方式 */}
|
||||
<div className="settings-section">
|
||||
<h3 className="section-title">索引方式</h3>
|
||||
<div className="index-options">
|
||||
<div
|
||||
className={`index-option ${currentSettings.indexingTechnique === 'high_quality' ? 'active' : ''} ${isProcessing ? 'disabled' : ''}`}
|
||||
onClick={() => !isProcessing && updateCurrentSettings('indexingTechnique', 'high_quality')}
|
||||
>
|
||||
<span className="option-radio"></span>
|
||||
<span className="option-label">高质量</span>
|
||||
<span className="option-badge recommended">推荐</span>
|
||||
</div>
|
||||
<div
|
||||
className={`index-option ${currentSettings.indexingTechnique === 'economy' ? 'active' : ''} ${isProcessing ? 'disabled' : ''}`}
|
||||
onClick={() => !isProcessing && updateCurrentSettings('indexingTechnique', 'economy')}
|
||||
>
|
||||
<span className="option-radio"></span>
|
||||
<span className="option-label">经济</span>
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
{/* 操作按钮 */}
|
||||
<div className="settings-actions">
|
||||
<Button onClick={handlePrevStep} disabled={isProcessing}>
|
||||
<ArrowLeftOutlined /> 上一步
|
||||
</Button>
|
||||
<Button
|
||||
type="primary"
|
||||
onClick={handleReprocess}
|
||||
loading={isProcessing}
|
||||
disabled={isProcessing || !currentDoc?.documentId}
|
||||
>
|
||||
更新嵌入配置
|
||||
</Button>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
{/* 右侧预览区域 */}
|
||||
<div className="preview-panel">
|
||||
<Card
|
||||
title={
|
||||
<div className="preview-header">
|
||||
<span>预览</span>
|
||||
{uploadedDocuments.length > 0 && (
|
||||
<>
|
||||
<Select
|
||||
value={currentDoc?.documentId || currentDoc?.file.name}
|
||||
style={{ width: 500 }}
|
||||
onChange={handleDocumentChange}
|
||||
options={uploadedDocuments.map((doc, idx) => ({
|
||||
value: doc.documentId || doc.file.name,
|
||||
label: (
|
||||
<span className="file-select-option">
|
||||
{doc.stage === 'completed' && <CheckCircleOutlined style={{ color: '#52c41a', marginRight: 4 }} />}
|
||||
{(doc.stage === 'uploading' || doc.stage === 'indexing') && <LoadingOutlined style={{ color: '#00684a', marginRight: 4 }} />}
|
||||
{doc.stage === 'error' && <ExclamationCircleOutlined style={{ color: '#ff4d4f', marginRight: 4 }} />}
|
||||
{doc.file.name}
|
||||
</span>
|
||||
),
|
||||
}))}
|
||||
/>
|
||||
{!isProcessing && currentDoc?.segments && (
|
||||
<span className="segment-count">
|
||||
{currentDoc.segments.length} 段块
|
||||
</span>
|
||||
)}
|
||||
</>
|
||||
)}
|
||||
</div>
|
||||
}
|
||||
className="preview-card"
|
||||
>
|
||||
{/* 处理进度（在预览框内显示） */}
|
||||
{isProcessing ? (
|
||||
<div className="preview-processing">
|
||||
<div className="processing-file">
|
||||
<FileTextOutlined className="file-icon" />
|
||||
<span className="file-name">{currentDoc?.file.name}</span>
|
||||
<LoadingOutlined className="status-icon loading" />
|
||||
</div>
|
||||
<Progress
|
||||
percent={getCurrentProgress()}
|
||||
status="active"
|
||||
strokeColor={{
|
||||
'0%': '#00684a',
|
||||
'100%': '#52c41a',
|
||||
}}
|
||||
/>
|
||||
<div className="status-text">{getStatusText()}</div>
|
||||
|
||||
{/* 索引阶段详情 */}
|
||||
{currentDoc?.stage === 'indexing' && (
|
||||
<div className="indexing-stages">
|
||||
<div className={`stage-item ${['waiting', 'parsing', 'cleaning', 'splitting', 'indexing', 'completed'].includes(currentDoc.indexingStatus) ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>等待处理</span>
|
||||
</div>
|
||||
<div className={`stage-item ${['parsing', 'cleaning', 'splitting', 'indexing', 'completed'].includes(currentDoc.indexingStatus) ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>解析文档</span>
|
||||
</div>
|
||||
<div className={`stage-item ${['cleaning', 'splitting', 'indexing', 'completed'].includes(currentDoc.indexingStatus) ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>清洗文本</span>
|
||||
</div>
|
||||
<div className={`stage-item ${['splitting', 'indexing', 'completed'].includes(currentDoc.indexingStatus) ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>分段处理</span>
|
||||
</div>
|
||||
<div className={`stage-item ${['indexing', 'completed'].includes(currentDoc.indexingStatus) ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>建立索引</span>
|
||||
</div>
|
||||
<div className={`stage-item ${currentDoc.indexingStatus === 'completed' ? 'active' : ''}`}>
|
||||
<span className="stage-dot"></span>
|
||||
<span>完成</span>
|
||||
</div>
|
||||
</div>
|
||||
)}
|
||||
</div>
|
||||
) : currentDoc?.stage === 'error' ? (
|
||||
<div className="preview-error">
|
||||
<ExclamationCircleOutlined className="error-icon" />
|
||||
<div className="error-text">{currentDoc.error || '处理失败'}</div>
|
||||
</div>
|
||||
) : previewLoading ? (
|
||||
<div className="preview-loading">
|
||||
<Spin size="large" />
|
||||
<div className="loading-text">加载中...</div>
|
||||
</div>
|
||||
) : (currentDoc?.segments?.length ?? 0) === 0 ? (
|
||||
<div className="preview-empty">
|
||||
<Empty description="等待处理完成后显示分段预览" />
|
||||
</div>
|
||||
) : (
|
||||
<div className="preview-segments">
|
||||
{currentDoc?.segments.map((segment, index) => (
|
||||
<div key={segment.id} className="segment-item">
|
||||
<div className="segment-header">
|
||||
<span className="segment-index">#{index + 1}</span>
|
||||
<span className="segment-chars">
|
||||
{segment.word_count} 字符
|
||||
</span>
|
||||
</div>
|
||||
<div className="segment-content">
|
||||
{segment.content}
|
||||
</div>
|
||||
</div>
|
||||
))}
|
||||
</div>
|
||||
)}
|
||||
</Card>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
{/* 完成状态底部操作 */}
|
||||
{stats.completed > 0 && (
|
||||
<div className="completion-actions">
|
||||
<span className="completion-stats">
|
||||
{stats.completed}/{stats.total} 个文档处理完成
|
||||
</span>
|
||||
<Button type="primary" onClick={handleGoToDocuments}>
|
||||
前往文档
|
||||
</Button>
|
||||
</div>
|
||||
)}
|
||||
</div>
|
||||
);
|
||||
};
|
||||
|
||||
return (
|
||||
<div className="document-upload-page">
|
||||
{/* 页面头部 */}
|
||||
<div className="upload-header">
|
||||
<Button
|
||||
type="text"
|
||||
icon={<ArrowLeftOutlined />}
|
||||
onClick={onClose}
|
||||
className="back-btn"
|
||||
>
|
||||
知识库
|
||||
</Button>
|
||||
{renderSteps()}
|
||||
</div>
|
||||
|
||||
{/* 内容区域 */}
|
||||
<div className="upload-content">
|
||||
{step === 1 && renderStep1()}
|
||||
{step === 2 && renderStep2()}
|
||||
</div>
|
||||
</div>
|
||||
);
|
||||
}
|
||||
@@ -0,0 +1,59 @@
|
||||
import { type ActionFunctionArgs } from '@remix-run/node';
|
||||
import { API_BASE_URL } from '~/config/api-config';
|
||||
|
||||
/**
|
||||
* POST /api/dataset/datasets/:datasetId/documents/:documentId/update-by-file
|
||||
* 通过文件更新文档（重新处理）
|
||||
*/
|
||||
export async function action({ request, params }: ActionFunctionArgs) {
|
||||
try {
|
||||
// 获取用户会话信息和 JWT
|
||||
const { getUserSession } = await import("~/api/login/auth.server");
|
||||
const { frontendJWT } = await getUserSession(request);
|
||||
|
||||
if (!frontendJWT) {
|
||||
return new Response(
|
||||
JSON.stringify({ error: 'JWT认证失败，请重新登录' }),
|
||||
{ status: 401, headers: { 'Content-Type': 'application/json' } }
|
||||
);
|
||||
}
|
||||
|
||||
const { datasetId, documentId } = params;
|
||||
if (!datasetId || !documentId) {
|
||||
return new Response(
|
||||
JSON.stringify({ error: '缺少 datasetId 或 documentId 参数' }),
|
||||
{ status: 400, headers: { 'Content-Type': 'application/json' } }
|
||||
);
|
||||
}
|
||||
|
||||
// 获取表单数据
|
||||
const formData = await request.formData();
|
||||
|
||||
console.log('[API] Update Document By File:', { datasetId, documentId });
|
||||
|
||||
// 转发请求到 FastAPI
|
||||
// Dify API: POST /datasets/{dataset_id}/documents/{document_id}/update-by-file
|
||||
const apiUrl = `${API_BASE_URL}/dify_dataset/datasets/${datasetId}/documents/${documentId}/update-by-file`;
|
||||
const response = await fetch(apiUrl, {
|
||||
method: 'POST',
|
||||
headers: {
|
||||
'Authorization': `Bearer ${frontendJWT}`,
|
||||
},
|
||||
body: formData,
|
||||
});
|
||||
|
||||
const data = await response.json();
|
||||
|
||||
return new Response(JSON.stringify(data), {
|
||||
status: response.status,
|
||||
headers: { 'Content-Type': 'application/json' },
|
||||
});
|
||||
|
||||
} catch (error: any) {
|
||||
console.error('[API] Update Document By File - Error:', error.message);
|
||||
return new Response(
|
||||
JSON.stringify({ error: error.message || 'Failed to update document by file' }),
|
||||
{ status: 500, headers: { 'Content-Type': 'application/json' } }
|
||||
);
|
||||
}
|
||||
}
|
||||
@@ -1425,3 +1425,919 @@
|
||||
.segment-manager-page .anticon-spin {
|
||||
color: rgb(0 104 74) !important;
|
||||
}
|
||||
|
||||
/* ============================================================================
|
||||
* 上传配置对话框样式
|
||||
* ============================================================================ */
|
||||
|
||||
.upload-config-modal .ant-modal-header {
|
||||
border-bottom: 1px solid #f0f0f0;
|
||||
padding: 16px 24px;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-modal-title {
|
||||
font-size: 16px;
|
||||
font-weight: 600;
|
||||
color: #1a1a1a;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-modal-body {
|
||||
padding: 24px;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-modal-footer {
|
||||
border-top: 1px solid #f0f0f0;
|
||||
padding: 12px 24px;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-btn-primary {
|
||||
background-color: #00684a;
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-btn-primary:hover {
|
||||
background-color: #005a3f;
|
||||
border-color: #005a3f;
|
||||
}
|
||||
|
||||
/* 文件信息区域 */
|
||||
.upload-config-form .file-info-section,
|
||||
.upload-progress-container .file-info-section {
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.upload-config-form .file-info,
|
||||
.upload-progress-container .file-info {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 12px;
|
||||
padding: 12px 16px;
|
||||
background: #f5f7f9;
|
||||
border-radius: 8px;
|
||||
border: 1px solid #e5e5e5;
|
||||
}
|
||||
|
||||
.upload-config-form .file-icon,
|
||||
.upload-progress-container .file-icon {
|
||||
font-size: 24px;
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.upload-config-form .file-details,
|
||||
.upload-progress-container .file-details {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
gap: 2px;
|
||||
flex: 1;
|
||||
min-width: 0;
|
||||
}
|
||||
|
||||
.upload-config-form .file-name,
|
||||
.upload-progress-container .file-name {
|
||||
font-size: 14px;
|
||||
font-weight: 500;
|
||||
color: #1a1a1a;
|
||||
white-space: nowrap;
|
||||
overflow: hidden;
|
||||
text-overflow: ellipsis;
|
||||
}
|
||||
|
||||
.upload-config-form .file-size,
|
||||
.upload-progress-container .file-size {
|
||||
font-size: 12px;
|
||||
color: #666;
|
||||
}
|
||||
|
||||
/* 设置项样式 */
|
||||
.upload-config-form .setting-item {
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.upload-config-form .setting-label {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 6px;
|
||||
font-size: 13px;
|
||||
font-weight: 500;
|
||||
color: #333;
|
||||
margin-bottom: 8px;
|
||||
}
|
||||
|
||||
.upload-config-form .help-icon {
|
||||
color: #999;
|
||||
font-size: 12px;
|
||||
cursor: help;
|
||||
}
|
||||
|
||||
.upload-config-form .checkbox-group {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
gap: 10px;
|
||||
}
|
||||
|
||||
.upload-config-form .checkbox-group .ant-checkbox-wrapper {
|
||||
color: #333;
|
||||
font-size: 13px;
|
||||
}
|
||||
|
||||
.upload-config-form .checkbox-group .ant-checkbox-checked .ant-checkbox-inner {
|
||||
background-color: #00684a;
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
/* 进度容器 */
|
||||
.upload-progress-container {
|
||||
min-height: 200px;
|
||||
}
|
||||
|
||||
.upload-progress-container .progress-section {
|
||||
padding: 24px 0;
|
||||
}
|
||||
|
||||
.upload-progress-container .status-text {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
gap: 8px;
|
||||
margin-top: 16px;
|
||||
font-size: 14px;
|
||||
color: #666;
|
||||
}
|
||||
|
||||
.upload-progress-container .status-icon {
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
/* 成功状态 */
|
||||
.upload-progress-container .success-state {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 24px;
|
||||
}
|
||||
|
||||
.upload-progress-container .success-icon {
|
||||
font-size: 48px;
|
||||
color: #52c41a;
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.upload-progress-container .success-text {
|
||||
font-size: 16px;
|
||||
font-weight: 500;
|
||||
color: #1a1a1a;
|
||||
}
|
||||
|
||||
/* 错误状态 */
|
||||
.upload-progress-container .error-state {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 24px;
|
||||
}
|
||||
|
||||
.upload-progress-container .error-icon {
|
||||
font-size: 48px;
|
||||
color: #ff4d4f;
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.upload-progress-container .error-text {
|
||||
font-size: 14px;
|
||||
color: #ff4d4f;
|
||||
text-align: center;
|
||||
}
|
||||
|
||||
/* 索引阶段指示器 */
|
||||
.upload-progress-container .indexing-stages {
|
||||
display: flex;
|
||||
flex-wrap: wrap;
|
||||
gap: 8px;
|
||||
margin-top: 24px;
|
||||
padding: 16px;
|
||||
background: #f5f7f9;
|
||||
border-radius: 8px;
|
||||
}
|
||||
|
||||
.upload-progress-container .stage-item {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 6px;
|
||||
font-size: 12px;
|
||||
color: #999;
|
||||
padding: 4px 8px;
|
||||
background: #fff;
|
||||
border-radius: 4px;
|
||||
border: 1px solid #e5e5e5;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.upload-progress-container .stage-item.active {
|
||||
color: #00684a;
|
||||
border-color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.05);
|
||||
}
|
||||
|
||||
.upload-progress-container .stage-dot {
|
||||
width: 6px;
|
||||
height: 6px;
|
||||
border-radius: 50%;
|
||||
background: #d9d9d9;
|
||||
transition: background 0.2s;
|
||||
}
|
||||
|
||||
.upload-progress-container .stage-item.active .stage-dot {
|
||||
background: #00684a;
|
||||
}
|
||||
|
||||
/* 进度条主题色 */
|
||||
.upload-config-modal .ant-progress-bg {
|
||||
background-color: #00684a;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-progress-status-active .ant-progress-bg::before {
|
||||
background: #fff;
|
||||
}
|
||||
|
||||
/* Divider 样式 */
|
||||
.upload-config-form .ant-divider {
|
||||
margin: 16px 0;
|
||||
border-color: #e5e5e5;
|
||||
}
|
||||
|
||||
.upload-config-form .ant-divider-inner-text {
|
||||
font-size: 13px;
|
||||
font-weight: 500;
|
||||
color: #666;
|
||||
}
|
||||
|
||||
/* Select 和 Input 主题色 */
|
||||
.upload-config-modal .ant-select:hover .ant-select-selector {
|
||||
border-color: #00684a !important;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-select-focused .ant-select-selector {
|
||||
border-color: #00684a !important;
|
||||
box-shadow: 0 0 0 2px rgba(0, 104, 74, 0.1) !important;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-input:hover,
|
||||
.upload-config-modal .ant-input-number:hover {
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.upload-config-modal .ant-input:focus,
|
||||
.upload-config-modal .ant-input-focused,
|
||||
.upload-config-modal .ant-input-number:focus,
|
||||
.upload-config-modal .ant-input-number-focused {
|
||||
border-color: #00684a;
|
||||
box-shadow: 0 0 0 2px rgba(0, 104, 74, 0.1);
|
||||
}
|
||||
|
||||
/* Spin 主题色 */
|
||||
.upload-config-modal .ant-spin .ant-spin-dot-item {
|
||||
background-color: #00684a !important;
|
||||
}
|
||||
|
||||
/* ============================================================================
|
||||
* 文档上传页面样式
|
||||
* ============================================================================ */
|
||||
|
||||
.document-upload-page {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
height: 100%;
|
||||
background: #fff;
|
||||
overflow: hidden;
|
||||
}
|
||||
|
||||
/* 上传页面头部 */
|
||||
.upload-header {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 16px 24px;
|
||||
border-bottom: 1px solid #f0f0f0;
|
||||
flex-shrink: 0;
|
||||
position: relative;
|
||||
}
|
||||
|
||||
.upload-header .back-btn {
|
||||
position: absolute;
|
||||
left: 24px;
|
||||
color: #666;
|
||||
font-size: 14px;
|
||||
padding: 4px 12px;
|
||||
height: auto;
|
||||
}
|
||||
|
||||
.upload-header .back-btn:hover {
|
||||
color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.08);
|
||||
}
|
||||
|
||||
/* 步骤指示器 */
|
||||
.upload-steps {
|
||||
display: inline-flex;
|
||||
align-items: center;
|
||||
gap: 0;
|
||||
}
|
||||
|
||||
.step-item {
|
||||
display: inline-flex;
|
||||
align-items: center;
|
||||
gap: 8px;
|
||||
color: #999;
|
||||
font-size: 13px;
|
||||
white-space: nowrap;
|
||||
}
|
||||
|
||||
.step-item.active {
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.step-item.completed {
|
||||
color: #52c41a;
|
||||
}
|
||||
|
||||
.step-number {
|
||||
width: 24px;
|
||||
height: 24px;
|
||||
border-radius: 50%;
|
||||
background: #f0f0f0;
|
||||
display: inline-flex;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
font-size: 12px;
|
||||
font-weight: 600;
|
||||
transition: all 0.2s;
|
||||
flex-shrink: 0;
|
||||
}
|
||||
|
||||
.step-item.active .step-number {
|
||||
background: #00684a;
|
||||
color: #fff;
|
||||
}
|
||||
|
||||
.step-item.completed .step-number {
|
||||
background: #52c41a;
|
||||
color: #fff;
|
||||
}
|
||||
|
||||
.step-title {
|
||||
font-weight: 500;
|
||||
}
|
||||
|
||||
.step-divider {
|
||||
width: 40px;
|
||||
height: 1px;
|
||||
background: #e5e5e5;
|
||||
margin: 0 12px;
|
||||
flex-shrink: 0;
|
||||
}
|
||||
|
||||
.step-divider.completed {
|
||||
background: #52c41a;
|
||||
}
|
||||
|
||||
/* 上传内容区域 */
|
||||
.upload-content {
|
||||
flex: 1;
|
||||
overflow-y: auto;
|
||||
padding: 24px;
|
||||
}
|
||||
|
||||
/* 第一步：选择文件 */
|
||||
.upload-step-content.step1 {
|
||||
max-width: 800px;
|
||||
margin: 0 auto;
|
||||
}
|
||||
|
||||
.step-heading {
|
||||
font-size: 24px;
|
||||
font-weight: 600;
|
||||
color: #1a1a1a;
|
||||
margin: 0 0 8px 0;
|
||||
}
|
||||
|
||||
.step-description {
|
||||
font-size: 14px;
|
||||
color: #666;
|
||||
margin: 0 0 24px 0;
|
||||
line-height: 1.6;
|
||||
}
|
||||
|
||||
/* 文件拖拽区域 */
|
||||
.file-drop-zone {
|
||||
margin-bottom: 24px;
|
||||
}
|
||||
|
||||
.file-drop-zone .ant-upload-drag {
|
||||
border: 2px dashed #d9d9d9;
|
||||
border-radius: 12px;
|
||||
background: #fafafa;
|
||||
padding: 40px 20px;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.file-drop-zone .ant-upload-drag:hover {
|
||||
border-color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.02);
|
||||
}
|
||||
|
||||
.file-drop-zone .ant-upload-drag-icon {
|
||||
color: #999;
|
||||
font-size: 48px !important;
|
||||
}
|
||||
|
||||
.file-drop-zone .ant-upload-text {
|
||||
font-size: 14px;
|
||||
color: #333;
|
||||
}
|
||||
|
||||
.file-drop-zone .upload-link {
|
||||
color: #00684a;
|
||||
font-weight: 500;
|
||||
cursor: pointer;
|
||||
}
|
||||
|
||||
.file-drop-zone .ant-upload-hint {
|
||||
font-size: 12px;
|
||||
color: #999;
|
||||
margin-top: 8px;
|
||||
}
|
||||
|
||||
/* 已选文件区域 */
|
||||
.selected-files-section {
|
||||
margin-bottom: 24px;
|
||||
}
|
||||
|
||||
.section-subtitle {
|
||||
font-size: 14px;
|
||||
font-weight: 600;
|
||||
color: #1a1a1a;
|
||||
margin: 0 0 12px 0;
|
||||
}
|
||||
|
||||
.selected-files-list {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
gap: 8px;
|
||||
}
|
||||
|
||||
.selected-file-item {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 12px;
|
||||
padding: 12px 16px;
|
||||
background: #f5f7f9;
|
||||
border-radius: 8px;
|
||||
border: 1px solid #e5e5e5;
|
||||
}
|
||||
|
||||
.selected-file-item .file-icon {
|
||||
font-size: 24px;
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.selected-file-item .file-info {
|
||||
flex: 1;
|
||||
min-width: 0;
|
||||
}
|
||||
|
||||
.selected-file-item .file-name {
|
||||
font-size: 14px;
|
||||
font-weight: 500;
|
||||
color: #1a1a1a;
|
||||
display: block;
|
||||
white-space: nowrap;
|
||||
overflow: hidden;
|
||||
text-overflow: ellipsis;
|
||||
}
|
||||
|
||||
.selected-file-item .file-size {
|
||||
font-size: 12px;
|
||||
color: #666;
|
||||
}
|
||||
|
||||
.selected-file-item .remove-file-btn {
|
||||
color: #999;
|
||||
}
|
||||
|
||||
.selected-file-item .remove-file-btn:hover {
|
||||
color: #ff4d4f;
|
||||
}
|
||||
|
||||
/* 步骤操作按钮 */
|
||||
.step-actions {
|
||||
display: flex;
|
||||
justify-content: flex-end;
|
||||
margin-top: 24px;
|
||||
}
|
||||
|
||||
.step-actions .next-btn {
|
||||
min-width: 120px;
|
||||
height: 40px;
|
||||
font-size: 14px;
|
||||
background: #00684a;
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.step-actions .next-btn:hover {
|
||||
background: #005a3f;
|
||||
border-color: #005a3f;
|
||||
}
|
||||
|
||||
.step-actions .next-btn:disabled {
|
||||
background: #d9d9d9;
|
||||
border-color: #d9d9d9;
|
||||
}
|
||||
|
||||
/* 第二步：分段配置与预览 */
|
||||
.upload-step-content.step2 {
|
||||
height: 100%;
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
}
|
||||
|
||||
/* 处理进度区 */
|
||||
.processing-status {
|
||||
background: #f5f7f9;
|
||||
border-radius: 12px;
|
||||
padding: 24px;
|
||||
margin-bottom: 24px;
|
||||
}
|
||||
|
||||
.processing-file {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 12px;
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.processing-file .file-icon {
|
||||
font-size: 24px;
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.processing-file .file-name {
|
||||
flex: 1;
|
||||
font-size: 14px;
|
||||
font-weight: 500;
|
||||
color: #1a1a1a;
|
||||
}
|
||||
|
||||
.processing-file .status-icon {
|
||||
font-size: 18px;
|
||||
}
|
||||
|
||||
.processing-file .status-icon.success {
|
||||
color: #52c41a;
|
||||
}
|
||||
|
||||
.processing-file .status-icon.error {
|
||||
color: #ff4d4f;
|
||||
}
|
||||
|
||||
.processing-file .status-icon.loading {
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.processing-status .status-text {
|
||||
text-align: center;
|
||||
font-size: 13px;
|
||||
color: #666;
|
||||
margin-top: 12px;
|
||||
}
|
||||
|
||||
/* 索引阶段指示器（Step2） */
|
||||
.processing-status .indexing-stages {
|
||||
display: flex;
|
||||
flex-wrap: wrap;
|
||||
gap: 8px;
|
||||
margin-top: 16px;
|
||||
justify-content: center;
|
||||
}
|
||||
|
||||
.processing-status .stage-item {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 6px;
|
||||
font-size: 12px;
|
||||
color: #999;
|
||||
padding: 4px 10px;
|
||||
background: #fff;
|
||||
border-radius: 4px;
|
||||
border: 1px solid #e5e5e5;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.processing-status .stage-item.active {
|
||||
color: #00684a;
|
||||
border-color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.05);
|
||||
}
|
||||
|
||||
.processing-status .stage-dot {
|
||||
width: 6px;
|
||||
height: 6px;
|
||||
border-radius: 50%;
|
||||
background: #d9d9d9;
|
||||
transition: background 0.2s;
|
||||
}
|
||||
|
||||
.processing-status .stage-item.active .stage-dot {
|
||||
background: #00684a;
|
||||
}
|
||||
|
||||
/* 错误状态 */
|
||||
.error-state {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 48px;
|
||||
text-align: center;
|
||||
}
|
||||
|
||||
.error-state .error-icon {
|
||||
font-size: 48px;
|
||||
color: #ff4d4f;
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.error-state .error-text {
|
||||
font-size: 14px;
|
||||
color: #666;
|
||||
margin-bottom: 24px;
|
||||
}
|
||||
|
||||
/* 索引方式选项 */
|
||||
.index-options {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
gap: 8px;
|
||||
}
|
||||
|
||||
.index-option {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 10px;
|
||||
padding: 12px 16px;
|
||||
background: #fff;
|
||||
border: 1px solid #e5e5e5;
|
||||
border-radius: 8px;
|
||||
cursor: pointer;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.index-option:hover {
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.index-option.active {
|
||||
border-color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.05);
|
||||
}
|
||||
|
||||
.index-option .option-radio {
|
||||
width: 16px;
|
||||
height: 16px;
|
||||
border-radius: 50%;
|
||||
border: 2px solid #d9d9d9;
|
||||
position: relative;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.index-option.active .option-radio {
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.index-option.active .option-radio::after {
|
||||
content: '';
|
||||
position: absolute;
|
||||
top: 50%;
|
||||
left: 50%;
|
||||
transform: translate(-50%, -50%);
|
||||
width: 8px;
|
||||
height: 8px;
|
||||
border-radius: 50%;
|
||||
background: #00684a;
|
||||
}
|
||||
|
||||
.index-option .option-label {
|
||||
font-size: 14px;
|
||||
color: #333;
|
||||
}
|
||||
|
||||
.index-option .option-badge {
|
||||
font-size: 11px;
|
||||
padding: 2px 6px;
|
||||
border-radius: 4px;
|
||||
margin-left: auto;
|
||||
}
|
||||
|
||||
.index-option .option-badge.recommended {
|
||||
background: rgba(0, 104, 74, 0.1);
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
/* 完成状态底部操作 */
|
||||
.completion-actions {
|
||||
display: flex;
|
||||
justify-content: center;
|
||||
padding: 24px;
|
||||
border-top: 1px solid #f0f0f0;
|
||||
margin-top: auto;
|
||||
}
|
||||
|
||||
.completion-actions .ant-btn-primary {
|
||||
min-width: 140px;
|
||||
height: 40px;
|
||||
font-size: 14px;
|
||||
background: #00684a;
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.completion-actions .ant-btn-primary:hover {
|
||||
background: #005a3f;
|
||||
border-color: #005a3f;
|
||||
}
|
||||
|
||||
/* 上传页面的分段内容布局（复用 document-detail-content） */
|
||||
.upload-step-content .document-detail-content {
|
||||
flex: 1;
|
||||
min-height: 0;
|
||||
}
|
||||
|
||||
/* 上传页面的设置操作按钮 */
|
||||
.upload-step-content .settings-actions {
|
||||
margin-top: 24px;
|
||||
display: flex;
|
||||
gap: 12px;
|
||||
}
|
||||
|
||||
.upload-step-content .settings-actions .ant-btn-primary {
|
||||
background: #00684a;
|
||||
border-color: #00684a;
|
||||
}
|
||||
|
||||
.upload-step-content .settings-actions .ant-btn-primary:hover {
|
||||
background: #005a3f;
|
||||
border-color: #005a3f;
|
||||
}
|
||||
|
||||
/* 进度条主题色 */
|
||||
.document-upload-page .ant-progress-bg {
|
||||
background-color: #00684a;
|
||||
}
|
||||
|
||||
.document-upload-page .ant-progress-status-active .ant-progress-bg::before {
|
||||
background: #fff;
|
||||
}
|
||||
|
||||
/* Spin 主题色 */
|
||||
.document-upload-page .ant-spin .ant-spin-dot-item {
|
||||
background-color: #00684a !important;
|
||||
}
|
||||
|
||||
/* 预览框内的进度显示 */
|
||||
.preview-processing {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 40px 24px;
|
||||
min-height: 300px;
|
||||
}
|
||||
|
||||
.preview-processing .processing-file {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 12px;
|
||||
margin-bottom: 24px;
|
||||
width: 100%;
|
||||
max-width: 400px;
|
||||
}
|
||||
|
||||
.preview-processing .file-icon {
|
||||
font-size: 24px;
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.preview-processing .file-name {
|
||||
flex: 1;
|
||||
font-size: 14px;
|
||||
font-weight: 500;
|
||||
color: #1a1a1a;
|
||||
white-space: nowrap;
|
||||
overflow: hidden;
|
||||
text-overflow: ellipsis;
|
||||
}
|
||||
|
||||
.preview-processing .status-icon {
|
||||
font-size: 18px;
|
||||
}
|
||||
|
||||
.preview-processing .status-icon.loading {
|
||||
color: #00684a;
|
||||
}
|
||||
|
||||
.preview-processing .ant-progress {
|
||||
width: 100%;
|
||||
max-width: 400px;
|
||||
}
|
||||
|
||||
.preview-processing .status-text {
|
||||
text-align: center;
|
||||
font-size: 13px;
|
||||
color: #666;
|
||||
margin-top: 12px;
|
||||
}
|
||||
|
||||
.preview-processing .indexing-stages {
|
||||
display: flex;
|
||||
flex-wrap: wrap;
|
||||
gap: 8px;
|
||||
margin-top: 24px;
|
||||
justify-content: center;
|
||||
max-width: 500px;
|
||||
}
|
||||
|
||||
.preview-processing .stage-item {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
gap: 6px;
|
||||
font-size: 12px;
|
||||
color: #999;
|
||||
padding: 4px 10px;
|
||||
background: #f5f7f9;
|
||||
border-radius: 4px;
|
||||
border: 1px solid #e5e5e5;
|
||||
transition: all 0.2s;
|
||||
}
|
||||
|
||||
.preview-processing .stage-item.active {
|
||||
color: #00684a;
|
||||
border-color: #00684a;
|
||||
background: rgba(0, 104, 74, 0.05);
|
||||
}
|
||||
|
||||
.preview-processing .stage-dot {
|
||||
width: 6px;
|
||||
height: 6px;
|
||||
border-radius: 50%;
|
||||
background: #d9d9d9;
|
||||
transition: background 0.2s;
|
||||
}
|
||||
|
||||
.preview-processing .stage-item.active .stage-dot {
|
||||
background: #00684a;
|
||||
}
|
||||
|
||||
/* 索引选项禁用状态 */
|
||||
.index-option.disabled {
|
||||
opacity: 0.6;
|
||||
cursor: not-allowed;
|
||||
}
|
||||
|
||||
.index-option.disabled:hover {
|
||||
border-color: #e5e5e5;
|
||||
}
|
||||
|
||||
/* 预览框内错误状态 */
|
||||
.preview-error {
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
justify-content: center;
|
||||
padding: 60px 24px;
|
||||
min-height: 300px;
|
||||
}
|
||||
|
||||
.preview-error .error-icon {
|
||||
font-size: 48px;
|
||||
color: #ff4d4f;
|
||||
margin-bottom: 16px;
|
||||
}
|
||||
|
||||
.preview-error .error-text {
|
||||
color: #666;
|
||||
font-size: 14px;
|
||||
text-align: center;
|
||||
}
|
||||
|
||||
/* 完成统计样式 */
|
||||
.completion-stats {
|
||||
color: #666;
|
||||
font-size: 14px;
|
||||
margin-right: 16px;
|
||||
}
|
||||
|
||||
/* 文件选择下拉选项 */
|
||||
.file-select-option {
|
||||
display: flex;
|
||||
align-items: center;
|
||||
}
|
||||
|
||||
+462
@@ -0,0 +1,462 @@
|
||||
# Dify 知识库 API 功能实现清单
|
||||
|
||||
## 概述
|
||||
|
||||
本文档对比 Dify 官方知识库 API 与当前项目已实现的功能。
|
||||
|
||||
- ✅ 已实现
|
||||
- ❌ 未实现
|
||||
- 🚫 不开放
|
||||
|
||||
---
|
||||
|
||||
## 系统架构
|
||||
|
||||
### 完整数据流向
|
||||
|
||||
```
|
||||
┌─────────────────────────────────────────────────────────────────────┐
|
||||
│ 第一层：React 组件 (浏览器) │
|
||||
│ 位置：app/components/dify-dataset-manager/ │
|
||||
│ 调用：客户端 API 函数 │
|
||||
└────────────────────────────┬────────────────────────────────────────┘
|
||||
│ 使用 axios 发送 HTTP 请求
|
||||
│ URL: /api/dataset/...
|
||||
│ 自动携带 cookies (JWT)
|
||||
↓
|
||||
┌─────────────────────────────────────────────────────────────────────┐
|
||||
│ 第二层：客户端 API 层 (浏览器侧) │
|
||||
│ 位置：app/api/dify-dataset/api/*.ts │
|
||||
│ 作用：封装 axios 请求，提供类型安全的函数接口 │
|
||||
│ 请求：axios.get('/api/dataset/datasets', { withCredentials: true })│
|
||||
└────────────────────────────┬────────────────────────────────────────┘
|
||||
│ HTTP 请求 (浏览器 → Remix 服务器)
|
||||
↓
|
||||
┌─────────────────────────────────────────────────────────────────────┐
|
||||
│ 第三层：Remix 路由层 (Node.js 服务端) │
|
||||
│ 位置：app/routes/api.dataset.*.tsx │
|
||||
│ 作用：接收浏览器请求，验证 JWT，转发到 FastAPI │
|
||||
│ 请求：fetch(`${API_BASE_URL}/dify_dataset/...`, { headers: JWT }) │
|
||||
└────────────────────────────┬────────────────────────────────────────┘
|
||||
│ HTTP 请求 (Remix → FastAPI)
|
||||
↓
|
||||
┌─────────────────────────────────────────────────────────────────────┐
|
||||
│ 第四层：FastAPI 后端代理 (Python) │
|
||||
│ 位置：docauditai/routers/dify_dataset.py │
|
||||
│ 作用：验证用户 JWT，添加 Dify DATASET_API_KEY，转发请求 │
|
||||
└────────────────────────────┬────────────────────────────────────────┘
|
||||
│ HTTP 请求 (FastAPI → Dify)
|
||||
↓
|
||||
┌─────────────────────────────────────────────────────────────────────┐
|
||||
│ 第五层：Dify 官方知识库 API │
|
||||
│ URL：https://api.dify.ai/v1/datasets/... │
|
||||
│ 鉴权：Authorization: Bearer {DATASET_API_KEY} │
|
||||
└─────────────────────────────────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
### 为什么有两层 API？
|
||||
|
||||
| 层级 | 位置 | 执行环境 | HTTP 库 | 作用 |
|
||||
|------|------|----------|---------|------|
|
||||
| **客户端 API** | `app/api/dify-dataset/api/*.ts` | 浏览器 | axios | 供 React 组件调用，类型安全 |
|
||||
| **Remix 路由** | `app/routes/api.dataset.*.tsx` | Node.js | fetch | 接收浏览器请求，转发到 FastAPI |
|
||||
|
||||
**调用链路**：
|
||||
```
|
||||
React 组件 → 客户端 API (axios) → Remix 路由 (fetch) → FastAPI → Dify API
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 一、知识库管理
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 路由文件 | 客户端函数 |
|
||||
|------|----------|------|------|----------|-----------|
|
||||
| 获取知识库列表 | /datasets | GET | ✅ | `api.dataset.datasets.tsx` | `fetchDatasets()` |
|
||||
| 查看知识库详情 | /datasets/{dataset_id} | GET | ✅ | `api.dataset.datasets.$datasetId.tsx` | `fetchDataset()` |
|
||||
| 创建空知识库 | /datasets | POST | ❌ | - | - |
|
||||
| 修改知识库名称 | /datasets/{dataset_id} | PATCH | ✅ | `api.dataset.datasets.$datasetId.tsx` | `updateDatasetName()` |
|
||||
| 删除知识库 | /datasets/{dataset_id} | DELETE | 🚫 | - | - |
|
||||
|
||||
**说明**：
|
||||
- 修改知识库：仅允许修改 `name` 字段，其他字段不开放
|
||||
- 删除知识库：出于安全考虑不对用户开放
|
||||
|
||||
---
|
||||
|
||||
## 二、文档管理
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 路由文件 | 客户端函数 |
|
||||
|------|----------|------|------|----------|-----------|
|
||||
| 获取文档列表 | /datasets/{id}/documents | GET | ✅ | `api.dataset.datasets.$datasetId.documents.tsx` | `fetchDocuments()` |
|
||||
| 获取文档详情 | /datasets/{id}/documents/{docId} | GET | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.tsx` | `fetchDocument()` |
|
||||
| 通过文件创建文档 | /datasets/{id}/document/create-by-file | POST | ✅ | `api.dataset.datasets.$datasetId.documents.tsx` | `uploadDocument()` |
|
||||
| 通过文本创建文档 | /datasets/{id}/document/create-by-text | POST | ❌ | - | - |
|
||||
| 通过文件更新文档 | /datasets/{id}/documents/{docId}/update-by-file | POST | ❌ | - | - |
|
||||
| 通过文本更新文档 | /datasets/{id}/documents/{docId}/update-by-text | POST | ❌ | - | - |
|
||||
| 删除文档 | /datasets/{id}/documents/{docId} | DELETE | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.tsx` | `deleteDocument()` |
|
||||
| 更新文档状态 | /datasets/{id}/documents/status/{action} | PATCH | ✅ | `api.dataset.datasets.$datasetId.documents.status.$action.tsx` | `toggleDocumentStatus()` |
|
||||
| 获取文档嵌入状态 | /datasets/{id}/documents/{batch}/indexing-status | GET | ✅ | `api.dataset.datasets.$datasetId.documents.$batch.indexing-status.tsx` | `fetchIndexingStatus()` |
|
||||
| 获取上传文件信息 | /datasets/{id}/documents/{docId}/upload-file | GET | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.upload-file.tsx` | `fetchUploadFileInfo()` |
|
||||
| 索引预估（预览分段） | /datasets/{id}/indexing-estimate | POST | ✅ | `api.dataset.datasets.$datasetId.indexing-estimate.tsx` | `fetchIndexingEstimate()` |
|
||||
| 重新处理文档 | /datasets/{id}/documents/reprocess | POST | ✅ | `api.dataset.datasets.$datasetId.documents.reprocess.tsx` | `reprocessDocument()` |
|
||||
|
||||
**说明**：
|
||||
- 上传文档：支持 multipart/form-data 格式
|
||||
- 文档状态：action 可选值为 `enable` / `disable` / `archive` / `un_archive`
|
||||
- 索引预估：用于预览分段效果，不会实际修改文档
|
||||
- 重新处理文档：通过 `original_document_id` 参数使用新的分段设置重新处理已有文档
|
||||
|
||||
---
|
||||
|
||||
## 三、分段管理
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 路由文件 | 客户端函数 |
|
||||
|------|----------|------|------|----------|-----------|
|
||||
| 获取分段列表 | /datasets/{id}/documents/{docId}/segments | GET | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.segments.tsx` | `fetchSegments()` |
|
||||
| 获取分段详情 | /datasets/{id}/documents/{docId}/segments/{segId} | GET | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.tsx` | `fetchSegment()` |
|
||||
| 新增分段 | /datasets/{id}/documents/{docId}/segments | POST | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.segments.tsx` | `createSegments()` |
|
||||
| 更新分段 | /datasets/{id}/documents/{docId}/segments/{segId} | POST | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.tsx` | `updateSegment()` |
|
||||
| 删除分段 | /datasets/{id}/documents/{docId}/segments/{segId} | DELETE | ✅ | `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.tsx` | `deleteSegment()` |
|
||||
|
||||
**说明**：
|
||||
- 新增分段：支持批量新增 `{ segments: [...] }`
|
||||
- 更新分段：可更新 `content`, `answer`, `keywords`, `enabled`
|
||||
|
||||
---
|
||||
|
||||
## 四、子分段管理（父子模式）
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 路由文件 | 客户端函数 |
|
||||
|------|----------|------|------|----------|-----------|
|
||||
| 查询子分段 | .../segments/{segId}/child_chunks | GET | ✅ | `...segments.$segmentId.child_chunks.tsx` | `fetchChildChunks()` |
|
||||
| 新增子分段 | .../segments/{segId}/child_chunks | POST | ✅ | `...segments.$segmentId.child_chunks.tsx` | `createChildChunk()` |
|
||||
| 更新子分段 | .../segments/{segId}/child_chunks/{chunkId} | PATCH | ✅ | `...segments.$segmentId.child_chunks.$childChunkId.tsx` | `updateChildChunk()` |
|
||||
| 删除子分段 | .../segments/{segId}/child_chunks/{chunkId} | DELETE | ✅ | `...segments.$segmentId.child_chunks.$childChunkId.tsx` | `deleteChildChunk()` |
|
||||
|
||||
**说明**：子分段用于 Dify 的父子模式分段策略
|
||||
|
||||
---
|
||||
|
||||
## 五、检索功能
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 路由文件 | 客户端函数 |
|
||||
|------|----------|------|------|----------|-----------|
|
||||
| 检索知识库 | /datasets/{id}/retrieve | POST | ✅ | `api.dataset.datasets.$datasetId.retrieve.tsx` | `retrieveDataset()` |
|
||||
|
||||
**检索参数详解**：
|
||||
```typescript
|
||||
{
|
||||
query: string; // 检索关键词
|
||||
retrieval_model: {
|
||||
search_method: 'keyword_search' | 'semantic_search' | 'full_text_search' | 'hybrid_search';
|
||||
reranking_enable: boolean; // 是否开启 rerank
|
||||
reranking_model?: object; // Rerank 模型配置
|
||||
top_k: number; // 返回结果数量
|
||||
score_threshold_enabled: boolean;
|
||||
score_threshold: number; // 分数阈值 (0-1)
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 六、元数据管理
|
||||
|
||||
**当前状态：❌ 全部未实现**
|
||||
|
||||
| 功能 | API 端点 | 方法 | 说明 |
|
||||
|------|----------|------|------|
|
||||
| 新增元数据 | /datasets/{id}/metadata | POST | type, name |
|
||||
| 更新元数据 | /datasets/{id}/metadata/{metaId} | PATCH | name |
|
||||
| 删除元数据 | /datasets/{id}/metadata/{metaId} | DELETE | |
|
||||
| 查询元数据列表 | /datasets/{id}/metadata | GET | |
|
||||
| 启用/禁用内置元数据 | /datasets/{id}/metadata/built-in/{action} | POST | |
|
||||
| 更新文档元数据 | /datasets/{id}/documents/metadata | POST | 批量更新 |
|
||||
|
||||
---
|
||||
|
||||
## 七、模型查询
|
||||
|
||||
| 功能 | API 端点 | 方法 | 状态 | 说明 |
|
||||
|------|----------|------|------|------|
|
||||
| 获取嵌入模型列表 | /workspaces/current/models/model-types/text-embedding | GET | ❌ | 创建知识库时需要 |
|
||||
|
||||
---
|
||||
|
||||
## 八、标签管理
|
||||
|
||||
**当前状态：❌ 全部未实现**
|
||||
|
||||
| 功能 | API 端点 | 方法 | 说明 |
|
||||
|------|----------|------|------|
|
||||
| 新增标签 | /datasets/tags | POST | name (最大50字符) |
|
||||
| 获取标签列表 | /datasets/tags | GET | |
|
||||
| 修改标签名称 | /datasets/tags | PATCH | name, tag_id |
|
||||
| 删除标签 | /datasets/tags | DELETE | tag_id |
|
||||
| 绑定知识库到标签 | /datasets/tags/binding | POST | tag_ids, target_id |
|
||||
| 解绑知识库和标签 | /datasets/tags/unbinding | POST | tag_id, target_id |
|
||||
| 查询知识库已绑定的标签 | /datasets/{id}/tags | POST | |
|
||||
|
||||
---
|
||||
|
||||
## 功能统计
|
||||
|
||||
| 类别 | 已实现 | 未实现 | 不开放 | 完成度 |
|
||||
|------|--------|--------|--------|--------|
|
||||
| 知识库管理 | 3 | 1 | 1 | 75% |
|
||||
| 文档管理 | 9 | 3 | 0 | 75% |
|
||||
| 分段管理 | 5 | 0 | 0 | 100% |
|
||||
| 子分段管理 | 4 | 0 | 0 | 100% |
|
||||
| 检索功能 | 1 | 0 | 0 | 100% |
|
||||
| 元数据管理 | 0 | 6 | 0 | 0% |
|
||||
| 模型查询 | 0 | 1 | 0 | 0% |
|
||||
| 标签管理 | 0 | 7 | 0 | 0% |
|
||||
| **总计** | **22** | **18** | **1** | **55%** |
|
||||
|
||||
---
|
||||
|
||||
## 代码文件清单
|
||||
|
||||
### Remix 路由层 (服务端)
|
||||
|
||||
所有路由文件位于 `app/routes/` 目录：
|
||||
|
||||
| 文件名 | HTTP 方法 | 功能 |
|
||||
|--------|----------|------|
|
||||
| `api.dataset.datasets.tsx` | GET | 获取知识库列表 |
|
||||
| `api.dataset.datasets.$datasetId.tsx` | GET / PATCH | 知识库详情 / 修改名称 |
|
||||
| `api.dataset.datasets.$datasetId.documents.tsx` | GET / POST | 文档列表 / 上传文档 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.tsx` | GET / DELETE | 文档详情 / 删除文档 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.upload-file.tsx` | GET | 获取上传文件信息 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$batch.indexing-status.tsx` | GET | 获取嵌入状态 |
|
||||
| `api.dataset.datasets.$datasetId.documents.status.$action.tsx` | PATCH | 更新文档状态 |
|
||||
| `api.dataset.datasets.$datasetId.indexing-estimate.tsx` | POST | 索引预估（预览分段） |
|
||||
| `api.dataset.datasets.$datasetId.documents.reprocess.tsx` | POST | 重新处理文档 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.segments.tsx` | GET / POST | 分段列表 / 新增分段 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.tsx` | GET / POST / DELETE | 分段详情 / 更新 / 删除 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.child_chunks.tsx` | GET / POST | 子分段列表 / 新增 |
|
||||
| `api.dataset.datasets.$datasetId.documents.$documentId.segments.$segmentId.child_chunks.$childChunkId.tsx` | PATCH / DELETE | 子分段更新 / 删除 |
|
||||
| `api.dataset.datasets.$datasetId.retrieve.tsx` | POST | 检索知识库 |
|
||||
|
||||
### 客户端 API 层 (浏览器侧)
|
||||
|
||||
```
|
||||
app/api/dify-dataset/
|
||||
├── index.ts # 统一导出
|
||||
├── client.server.ts # 服务端基础请求函数（备用）
|
||||
├── type/ # 类型定义
|
||||
│ ├── index.ts # 类型统一导出
|
||||
│ ├── commonTypes.ts # 通用类型
|
||||
│ ├── datasetTypes.ts # 知识库类型
|
||||
│ ├── documentTypes.ts # 文档类型
|
||||
│ └── segmentTypes.ts # 分段/子分段/检索类型
|
||||
└── api/ # API 调用函数
|
||||
├── index.ts # 函数统一导出
|
||||
├── datasetApi.ts # 知识库 API
|
||||
├── documentApi.ts # 文档 API
|
||||
└── segmentApi.ts # 分段/子分段/检索 API
|
||||
```
|
||||
|
||||
### 客户端函数清单
|
||||
|
||||
**datasetApi.ts - 知识库管理**
|
||||
```typescript
|
||||
fetchDatasets(page, limit) // 获取知识库列表
|
||||
fetchDataset(datasetId) // 获取知识库详情
|
||||
updateDatasetName(datasetId, name) // 修改知识库名称
|
||||
```
|
||||
|
||||
**documentApi.ts - 文档管理**
|
||||
```typescript
|
||||
fetchDocuments(datasetId, page, limit, keyword) // 获取文档列表
|
||||
fetchDocument(datasetId, documentId) // 获取文档详情
|
||||
deleteDocument(datasetId, documentId) // 删除文档
|
||||
toggleDocumentStatus(datasetId, documentId, enabled) // 启用/禁用文档
|
||||
uploadDocument(datasetId, file, onProgress) // 上传文档
|
||||
fetchIndexingStatus(datasetId, batch) // 获取嵌入状态
|
||||
fetchUploadFileInfo(datasetId, documentId) // 获取上传文件信息
|
||||
fetchIndexingEstimate(datasetId, fileId, processRule, docForm, docLanguage) // 索引预估（预览分段效果）
|
||||
reprocessDocument(datasetId, originalDocumentId, processRule, docForm, docLanguage) // 重新处理文档
|
||||
```
|
||||
|
||||
**segmentApi.ts - 分段/子分段/检索**
|
||||
```typescript
|
||||
// 分段
|
||||
fetchSegments(datasetId, documentId, page, limit, keyword)
|
||||
fetchSegment(datasetId, documentId, segmentId)
|
||||
createSegments(datasetId, documentId, segments)
|
||||
updateSegment(datasetId, documentId, segmentId, segment)
|
||||
deleteSegment(datasetId, documentId, segmentId)
|
||||
toggleSegmentStatus(datasetId, documentId, segmentId, enabled)
|
||||
|
||||
// 子分段
|
||||
fetchChildChunks(datasetId, documentId, segmentId, page, limit, keyword)
|
||||
createChildChunk(datasetId, documentId, segmentId, content)
|
||||
updateChildChunk(datasetId, documentId, segmentId, childChunkId, content)
|
||||
deleteChildChunk(datasetId, documentId, segmentId, childChunkId)
|
||||
|
||||
// 检索
|
||||
retrieveDataset(datasetId, query, retrievalModel)
|
||||
```
|
||||
|
||||
### UI 组件
|
||||
|
||||
```
|
||||
app/components/dify-dataset-manager/
|
||||
├── index.tsx # 主容器组件 - 状态管理、标签页切换
|
||||
├── layout.tsx # 布局组件 - 左侧菜单栏 + 右侧内容区
|
||||
├── document-list.tsx # 文档列表 - 表格、搜索、上传、删除
|
||||
├── document-detail.tsx # 文档详情 - 分段设置、预览块
|
||||
├── retrieve-test.tsx # 召回测试 - 知识库检索测试
|
||||
└── dataset-settings.tsx # 知识库设置 - 名称、描述修改
|
||||
```
|
||||
|
||||
### 布局结构（仿 Dify 风格）
|
||||
|
||||
```
|
||||
┌─────────────────────────────────────────────────────────────┐
|
||||
│ dataset-layout │
|
||||
├──────────────────┬──────────────────────────────────────────┤
|
||||
│ dataset-sidebar │ dataset-main │
|
||||
│ │ │
|
||||
│ ┌─────────────┐ │ 根据 activeTab 渲染： │
|
||||
│ │ 知识库信息 │ │ - documents → DocumentList │
|
||||
│ │ (名称/数量) │ │ - documents + selectedDoc → DocumentDetail│
|
||||
│ └─────────────┘ │ - retrieve → RetrieveTest │
|
||||
│ │ - settings → DatasetSettings │
|
||||
│ ┌─────────────┐ │ │
|
||||
│ │ 文档 │ │ │
|
||||
│ │ 召回测试 │ │ │
|
||||
│ │ 设置 │ │ │
|
||||
│ └─────────────┘ │ │
|
||||
└──────────────────┴──────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 鉴权机制
|
||||
|
||||
### 三层认证流程
|
||||
|
||||
```
|
||||
┌────────────────────────────────────────────────────────────────┐
|
||||
│ 浏览器 → Remix 服务器 │
|
||||
│ 认证方式：Cookie (会话中的 JWT) │
|
||||
│ axios 配置：{ withCredentials: true } │
|
||||
└────────────────────────────────────────────────────────────────┘
|
||||
↓
|
||||
┌────────────────────────────────────────────────────────────────┐
|
||||
│ Remix 服务器 → FastAPI │
|
||||
│ 认证方式：Authorization: Bearer {frontendJWT} │
|
||||
│ JWT 来源：getUserSession(request) │
|
||||
└────────────────────────────────────────────────────────────────┘
|
||||
↓
|
||||
┌────────────────────────────────────────────────────────────────┐
|
||||
│ FastAPI → Dify API │
|
||||
│ 认证方式：Authorization: Bearer {DATASET_API_KEY} │
|
||||
│ API Key：服务端环境变量配置 │
|
||||
└────────────────────────────────────────────────────────────────┘
|
||||
```
|
||||
|
||||
### 关键配置
|
||||
|
||||
```typescript
|
||||
// app/config/api-config.ts
|
||||
export const API_BASE_URL = apiConfig.baseUrl; // 如：http://10.79.97.17:8000
|
||||
|
||||
// 根据端口自动选择配置
|
||||
const portConfigs = {
|
||||
'51703': { baseUrl: 'http://172.16.0.55:8073' }, // 梅州
|
||||
'51704': { baseUrl: 'http://10.79.97.17:8001' }, // 云浮
|
||||
'51707': { baseUrl: 'http://10.79.97.17:8004' }, // 省级
|
||||
// ...
|
||||
};
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 常见错误码
|
||||
|
||||
| code | status | message |
|
||||
|------|--------|---------|
|
||||
| no_file_uploaded | 400 | Please upload your file. |
|
||||
| too_many_files | 400 | Only one file is allowed. |
|
||||
| file_too_large | 413 | File size exceeded. |
|
||||
| unsupported_file_type | 415 | File type not allowed. |
|
||||
| high_quality_dataset_only | 400 | Current operation only supports 'high-quality' datasets. |
|
||||
| dataset_not_initialized | 400 | The dataset is still being initialized or indexing. |
|
||||
| archived_document_immutable | 403 | The archived document is not editable. |
|
||||
| dataset_name_duplicate | 409 | The dataset name already exists. |
|
||||
| invalid_action | 400 | Invalid action. |
|
||||
| document_already_finished | 400 | The document has been processed. |
|
||||
| document_indexing | 400 | The document is being processed and cannot be edited. |
|
||||
| invalid_metadata | 400 | The metadata content is incorrect. |
|
||||
|
||||
---
|
||||
|
||||
## 九、文档分段设置（上传时配置）
|
||||
|
||||
### API 支持的分段参数
|
||||
|
||||
在上传文档时，可以通过 `process_rule` 参数配置分段设置：
|
||||
|
||||
```typescript
|
||||
{
|
||||
indexing_technique: 'high_quality' | 'economy',
|
||||
process_rule: {
|
||||
mode: 'automatic' | 'custom',
|
||||
rules: {
|
||||
pre_processing_rules: [
|
||||
{ id: 'remove_extra_spaces', enabled: boolean }, // 替换连续空格
|
||||
{ id: 'remove_urls_emails', enabled: boolean } // 删除URL和邮件
|
||||
],
|
||||
segmentation: {
|
||||
separator: string, // 分段标识符，如 "\\n\\n"
|
||||
max_tokens: number // 分段最大长度，100-4000
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
### 功能支持情况
|
||||
|
||||
| 功能 | API 支持 | 参数 | 说明 |
|
||||
|------|----------|------|------|
|
||||
| 分段标识符 | ✅ | `separator` | 如 `\\n\\n`、`###` |
|
||||
| 分段最大长度 | ✅ | `max_tokens` | 100-4000 |
|
||||
| 替换连续空格 | ✅ | `remove_extra_spaces` | 预处理规则 |
|
||||
| 删除URL和邮件 | ✅ | `remove_urls_emails` | 预处理规则 |
|
||||
| 分段重叠长度 | ❌ | - | API 不支持 |
|
||||
| Q&A 分段 | ⚠️ | `doc_form: "qa_model"` | 需特殊配置 |
|
||||
|
||||
### 重要限制
|
||||
|
||||
⚠️ **已有文档无法直接修改分段设置**
|
||||
|
||||
Dify API 不支持修改已上传文档的分段规则。如需应用新设置，必须：
|
||||
1. 使用 `original_document_id` 参数重新上传文档
|
||||
2. 或删除文档后重新上传
|
||||
|
||||
---
|
||||
|
||||
## 优先级建议
|
||||
|
||||
### 高优先级（核心功能）
|
||||
1. ~~**检索知识库**~~ ✅ 已实现
|
||||
2. ~~获取文档嵌入状态~~ ✅ 已实现
|
||||
3. **创建空知识库** - 让用户能创建新的知识库
|
||||
4. **获取嵌入模型列表** - 创建知识库时需要选择模型
|
||||
|
||||
### 中优先级（完善功能）
|
||||
5. **通过文本创建文档** - 支持直接输入文本
|
||||
6. ~~**新增分段**~~ ✅ 已实现
|
||||
7. ~~修改知识库详情~~ ✅ 已实现
|
||||
8. ~~获取上传文件信息~~ ✅ 已实现
|
||||
|
||||
### 低优先级（扩展功能）
|
||||
9. ~~子分段管理系列~~ ✅ 已实现（API层）
|
||||
10. 元数据管理系列
|
||||
11. 标签管理系列
|
||||
12. 文档更新功能
|
||||
Reference in New Issue
Block a user