fix: file size limit to 15M

.github/workflows/stale.yml

@@ -27,4 +27,4 @@ jobs:
         stale-pr-message: "Close due to it's no longer active, if you have any questions, you can reopen it."
         stale-issue-label: 'no-issue-activity'
         stale-pr-label: 'no-pr-activity'
-        only-labels: 'duplicate,question,invalid,wontfix,no-issue-activity,no-pr-activity,enhancement'
+        any-of-labels: 'duplicate,question,invalid,wontfix,no-issue-activity,no-pr-activity,enhancement'

api/core/index/readers/xlsx_parser.py

@@ -27,5 +27,7 @@ class XLSXParser(BaseParser):
                     if keys == []:
                         keys = list(map(str, row))
                     else:
-                        data.append(json.dumps(dict(zip(keys, list(map(str, row)))), ensure_ascii=False))
+                        row_dict = dict(zip(keys, row))
+                        row_dict = {k: v for k, v in row_dict.items() if v}
+                        data.append(json.dumps(row_dict, ensure_ascii=False))
         return '\n\n'.join(data)

web/app/(commonLayout)/datasets/(datasetDetailLayout)/[datasetId]/layout.tsx

@@ -124,7 +124,7 @@ const DatasetDetailLayout: FC<IAppDetailLayoutProps> = (props) => {
         ? (
           <>
             <div className={s.subTitle}>{relatedApps?.total || '--'} {t('common.datasetMenus.relatedApp')}</div>
-            {relatedApps?.data?.map(item => (<LikedItem detail={item} />))}
+            {relatedApps?.data?.map((item, index) => (<LikedItem key={index} detail={item} />))}
           </>
         )
         : (

web/app/components/datasets/create/file-uploader/index.tsx

@@ -30,7 +30,7 @@ const ACCEPTS = [
   '.csv',
 ]
 
-const MAX_SIZE = 10 * 1024 * 1024
+const MAX_SIZE = 15 * 1024 * 1024
 const BATCH_COUNT = 5
 
 const FileUploader = ({

web/app/components/datasets/create/step-one/index.module.css

@@ -10,7 +10,9 @@
 }
 
 .form {
+  position: relative;
   padding: 12px 64px;
+  background-color: #fff;
 }
 
 .dataSourceTypeList {

web/app/components/datasets/hit-testing/index.tsx

 'use client'
-import React, { useState, FC, useMemo } from 'react'
+import type { FC } from 'react'
+import React, { useMemo, useState } from 'react'
 import { useTranslation } from 'react-i18next'
 import useSWR from 'swr'
-import { fetchTestingRecords } from '@/service/datasets'
 import { omit } from 'lodash-es'
-import Pagination from '@/app/components/base/pagination'
-import Modal from '@/app/components/base/modal'
-import Loading from '@/app/components/base/loading'
-import type { HitTestingResponse, HitTesting } from '@/models/datasets'
 import cn from 'classnames'
 import dayjs from 'dayjs'
 import SegmentCard from '../documents/detail/completed/SegmentCard'
@@ -15,8 +11,14 @@ import docStyle from '../documents/detail/completed/style.module.css'
 import Textarea from './textarea'
 import s from './style.module.css'
 import HitDetail from './hit-detail'
+import type { HitTestingResponse } from '@/models/datasets'
+import { HitTesting } from '@/models/datasets'
+import Loading from '@/app/components/base/loading'
+import Modal from '@/app/components/base/modal'
+import Pagination from '@/app/components/base/pagination'
+import { fetchTestingRecords } from '@/service/datasets'
 
-const limit = 10;
+const limit = 10
 
 type Props = {
   datasetId: string
@@ -32,23 +34,24 @@ const RecordsEmpty: FC = () => {
   </div>
 }
 
+// eslint-disable-next-line @typescript-eslint/no-redeclare
 const HitTesting: FC<Props> = ({ datasetId }: Props) => {
   const { t } = useTranslation()
-  const [hitResult, setHitResult] = useState<HitTestingResponse | undefined>(); // 初始化记录为空数组
-  const [submitLoading, setSubmitLoading] = useState(false);
+  const [hitResult, setHitResult] = useState<HitTestingResponse | undefined>() // 初始化记录为空数组
+  const [submitLoading, setSubmitLoading] = useState(false)
   const [currParagraph, setCurrParagraph] = useState<{ paraInfo?: HitTesting; showModal: boolean }>({ showModal: false })
-  const [text, setText] = useState('');
+  const [text, setText] = useState('')
 
   const [currPage, setCurrPage] = React.useState<number>(0)
   const { data: recordsRes, error, mutate: recordsMutate } = useSWR({
     action: 'fetchTestingRecords',
     datasetId,
-    params: { limit, page: currPage + 1, }
+    params: { limit, page: currPage + 1 },
   }, apiParams => fetchTestingRecords(omit(apiParams, 'action')))
 
   const total = recordsRes?.total || 0
 
-  const points = useMemo(() => (hitResult?.records.map((v) => [v.tsne_position.x, v.tsne_position.y]) || []), [hitResult?.records])
+  const points = useMemo(() => (hitResult?.records.map(v => [v.tsne_position.x, v.tsne_position.y]) || []), [hitResult?.records])
 
   const onClickCard = (detail: HitTesting) => {
     setCurrParagraph({ paraInfo: detail, showModal: true })
@@ -71,50 +74,54 @@ const HitTesting: FC<Props> = ({ datasetId }: Props) => {
           text={text}
         />
         <div className={cn(s.title, 'mt-8 mb-2')}>{t('datasetHitTesting.recents')}</div>
-        {!recordsRes && !error ? (
-          <div className='flex-1'><Loading type='app' /></div>
-        ) : recordsRes?.data?.length ? (
-          <>
-            <table className={`w-full border-collapse border-0 mt-3 ${s.table}`}>
-              <thead className="h-8 leading-8 border-b border-gray-200 text-gray-500 font-bold">
-                <tr>
-                  <td className='w-28'>{t('datasetHitTesting.table.header.source')}</td>
-                  <td>{t('datasetHitTesting.table.header.text')}</td>
-                  <td className='w-48'>{t('datasetHitTesting.table.header.time')}</td>
-                </tr>
-              </thead>
-              <tbody className="text-gray-500">
-                {recordsRes?.data?.map((record) => {
-                  return <tr
-                    key={record.id}
-                    className='group border-b border-gray-200 h-8 hover:bg-gray-50 cursor-pointer'
-                    onClick={() => setText(record.content)}
-                  >
-                    <td className='w-24'>
-                      <div className='flex items-center'>
-                        <div className={cn(s[`${record.source}_icon`], s.commonIcon, 'mr-1')} />
-                        <span className='capitalize'>{record.source.replace('_', ' ')}</span>
-                      </div>
-                    </td>
-                    <td className='max-w-xs group-hover:text-primary-600'>{record.content}</td>
-                    <td className='w-36'>
-                      {dayjs.unix(record.created_at).format(t('datasetHitTesting.dateTimeFormat') as string)}
-                    </td>
-                  </tr>
-                })}
-              </tbody>
-            </table>
-            {(total && total > limit)
-              ? <Pagination current={currPage} onChange={setCurrPage} total={total} limit={limit} />
-              : null}
-          </>
-        ) : (
-          <RecordsEmpty />
-        )}
+        {(!recordsRes && !error)
+          ? (
+            <div className='flex-1'><Loading type='app' /></div>
+          )
+          : recordsRes?.data?.length
+            ? (
+              <>
+                <table className={`w-full border-collapse border-0 mt-3 ${s.table}`}>
+                  <thead className="h-8 leading-8 border-b border-gray-200 text-gray-500 font-bold">
+                    <tr>
+                      <td className='w-28'>{t('datasetHitTesting.table.header.source')}</td>
+                      <td>{t('datasetHitTesting.table.header.text')}</td>
+                      <td className='w-48'>{t('datasetHitTesting.table.header.time')}</td>
+                    </tr>
+                  </thead>
+                  <tbody className="text-gray-500">
+                    {recordsRes?.data?.map((record) => {
+                      return <tr
+                        key={record.id}
+                        className='group border-b border-gray-200 h-8 hover:bg-gray-50 cursor-pointer'
+                        onClick={() => setText(record.content)}
+                      >
+                        <td className='w-24'>
+                          <div className='flex items-center'>
+                            <div className={cn(s[`${record.source}_icon`], s.commonIcon, 'mr-1')} />
+                            <span className='capitalize'>{record.source.replace('_', ' ')}</span>
+                          </div>
+                        </td>
+                        <td className='max-w-xs group-hover:text-primary-600'>{record.content}</td>
+                        <td className='w-36'>
+                          {dayjs.unix(record.created_at).format(t('datasetHitTesting.dateTimeFormat') as string)}
+                        </td>
+                      </tr>
+                    })}
+                  </tbody>
+                </table>
+                {(total && total > limit)
+                  ? <Pagination current={currPage} onChange={setCurrPage} total={total} limit={limit} />
+                  : null}
+              </>
+            )
+            : (
+              <RecordsEmpty />
+            )}
       </div>
       <div className={s.rightDiv}>
-        {submitLoading ?
-          <div className={s.cardWrapper}>
+        {submitLoading
+          ? <div className={s.cardWrapper}>
             <SegmentCard
               loading={true}
               scene='hitTesting'
@@ -125,32 +132,36 @@ const HitTesting: FC<Props> = ({ datasetId }: Props) => {
               scene='hitTesting'
               className='h-[216px]'
             />
-          </div> : !hitResult?.records.length ? (
-            <div className='h-full flex flex-col justify-center items-center'>
-              <div className={cn(docStyle.commonIcon, docStyle.targetIcon, '!bg-gray-200 !h-14 !w-14')} />
-              <div className='text-gray-300 text-[13px] mt-3'>
-                {t('datasetHitTesting.hit.emptyTip')}
-              </div>
-            </div>
-          ) : (
-            <>
-              <div className='text-gray-600 font-semibold mb-4'>{t('datasetHitTesting.hit.title')}</div>
-              <div className='overflow-auto flex-1'>
-                <div className={s.cardWrapper}>
-                  {hitResult?.records.map((record) => {
-                    return <SegmentCard
-                      loading={false}
-                      detail={record.segment as any}
-                      score={record.score}
-                      scene='hitTesting'
-                      className='h-[216px] mb-4'
-                      onClick={() => onClickCard(record as any)}
-                    />
-                  })}
+          </div>
+          : !hitResult?.records.length
+            ? (
+              <div className='h-full flex flex-col justify-center items-center'>
+                <div className={cn(docStyle.commonIcon, docStyle.targetIcon, '!bg-gray-200 !h-14 !w-14')} />
+                <div className='text-gray-300 text-[13px] mt-3'>
+                  {t('datasetHitTesting.hit.emptyTip')}
                 </div>
               </div>
-            </>
-          )
+            )
+            : (
+              <>
+                <div className='text-gray-600 font-semibold mb-4'>{t('datasetHitTesting.hit.title')}</div>
+                <div className='overflow-auto flex-1'>
+                  <div className={s.cardWrapper}>
+                    {hitResult?.records.map((record, idx) => {
+                      return <SegmentCard
+                        key={idx}
+                        loading={false}
+                        detail={record.segment as any}
+                        score={record.score}
+                        scene='hitTesting'
+                        className='h-[216px] mb-4'
+                        onClick={() => onClickCard(record as any)}
+                      />
+                    })}
+                  </div>
+                </div>
+              </>
+            )
         }
       </div>
       <Modal

web/i18n/lang/dataset-creation.en.ts

@@ -23,7 +23,7 @@ const translation = {
       title: 'Upload text file',
       button: 'Drag and drop file, or',
       browse: 'Browse',
-      tip: 'Supports txt, html, markdown, xlsx, and pdf. Max 10MB each.',
+      tip: 'Supports txt, html, markdown, xlsx, and pdf. Max 15MB each.',
       validation: {
         typeError: 'File type not supported',
         size: 'File too large. Maximum is 10MB',

web/i18n/lang/dataset-creation.zh.ts

@@ -23,7 +23,7 @@ const translation = {
       title: '上传文本文件',
       button: '拖拽文件至此，或者',
       browse: '选择文件',
-      tip: '已支持 TXT、 HTML、 Markdown、 PDF、 XLSX，每个文件不超过 10 MB。',
+      tip: '已支持 TXT、 HTML、 Markdown、 PDF、 XLSX，每个文件不超过 15 MB。',
       validation: {
         typeError: '文件类型不支持',
         size: '文件太大了，不能超过 10MB',