From dd019e7ba1573ac1a1dd85d9112ad64f2e11de8b Mon Sep 17 00:00:00 2001
From: balibabu <cike8899@users.noreply.github.com>
Date: Mon, 23 Sep 2024 14:58:51 +0800
Subject: [PATCH] feat: Configurable for excel, html table or row based text
 #2516 (#2538)

### What problem does this PR solve?

feat: Configurable for excel, html table or row based text #2516

### Type of change

- [ ] Bug Fix (non-breaking change which fixes an issue)
- [x] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
---
 .../components/chunk-method-modal/index.tsx   |  5 +++++
 web/src/components/excel-to-html.tsx          | 19 +++++++++++++++++++
 web/src/locales/en.ts                         |  2 ++
 web/src/locales/zh-traditional.ts             |  2 ++
 web/src/locales/zh.ts                         |  2 ++
 .../knowledge-setting/configuration.tsx       |  2 ++
 6 files changed, 32 insertions(+)
 create mode 100644 web/src/components/excel-to-html.tsx

diff --git a/web/src/components/chunk-method-modal/index.tsx b/web/src/components/chunk-method-modal/index.tsx
index 9b20d3fab..876de5e0f 100644
--- a/web/src/components/chunk-method-modal/index.tsx
+++ b/web/src/components/chunk-method-modal/index.tsx
@@ -24,6 +24,7 @@ import { useFetchParserListOnMount } from './hooks';
 import { useTranslate } from '@/hooks/common-hooks';
 import Delimiter from '../delimiter';
 import EntityTypesItem from '../entity-types-item';
+import ExcelToHtml from '../excel-to-html';
 import LayoutRecognize from '../layout-recognize';
 import ParseConfiguration, {
   showRaptorParseConfiguration,
@@ -104,6 +105,9 @@ const ChunkMethodModal: React.FC<IProps> = ({
 
   const showEntityTypes = selectedTag === 'knowledge_graph';
 
+  const showExcelToHtml =
+    selectedTag === 'naive' && documentExtension === 'xlsx';
+
   const afterClose = () => {
     form.resetFields();
   };
@@ -279,6 +283,7 @@ const ChunkMethodModal: React.FC<IProps> = ({
             <Delimiter></Delimiter>
           </>
         )}
+        {showExcelToHtml && <ExcelToHtml></ExcelToHtml>}
         {showRaptorParseConfiguration(selectedTag) && (
           <ParseConfiguration></ParseConfiguration>
         )}
diff --git a/web/src/components/excel-to-html.tsx b/web/src/components/excel-to-html.tsx
new file mode 100644
index 000000000..e236a12f3
--- /dev/null
+++ b/web/src/components/excel-to-html.tsx
@@ -0,0 +1,19 @@
+import { useTranslate } from '@/hooks/common-hooks';
+import { Form, Switch } from 'antd';
+
+const ExcelToHtml = () => {
+  const { t } = useTranslate('knowledgeDetails');
+  return (
+    <Form.Item
+      name={['parser_config', 'html4excel']}
+      label={t('html4excel')}
+      initialValue={false}
+      valuePropName="checked"
+      tooltip={t('html4excelTip')}
+    >
+      <Switch />
+    </Form.Item>
+  );
+};
+
+export default ExcelToHtml;
diff --git a/web/src/locales/en.ts b/web/src/locales/en.ts
index ade4385aa..ed0586f88 100644
--- a/web/src/locales/en.ts
+++ b/web/src/locales/en.ts
@@ -150,6 +150,8 @@ export default {
       topK: 'Top-K',
       topKTip: `K chunks will be fed into rerank models.`,
       delimiter: `Delimiter`,
+      html4excel: 'Excel to HTML',
+      html4excelTip: `Excel will be parsed into HTML table or not. If it's FALSE, every row in Excel will be formed as a chunk.`,
     },
     knowledgeConfiguration: {
       titleDescription:
diff --git a/web/src/locales/zh-traditional.ts b/web/src/locales/zh-traditional.ts
index 2060907c3..04589bd6b 100644
--- a/web/src/locales/zh-traditional.ts
+++ b/web/src/locales/zh-traditional.ts
@@ -146,6 +146,8 @@ export default {
       topK: 'Top-K',
       topKTip: `K塊將被送入Rerank型號。`,
       delimiter: `分段標識符`,
+      html4excel: '表格轉HTML',
+      html4excelTip: `Excel 是否會被解析為 HTML 表格。如果為 FALSE，Excel 中的每一行都會形成一個區塊。`,
     },
     knowledgeConfiguration: {
       titleDescription: '在這裡更新您的知識庫詳細信息，尤其是解析方法。',
diff --git a/web/src/locales/zh.ts b/web/src/locales/zh.ts
index 038945325..796774d7b 100644
--- a/web/src/locales/zh.ts
+++ b/web/src/locales/zh.ts
@@ -147,6 +147,8 @@ export default {
       topK: 'Top-K',
       topKTip: `K块将被送入Rerank型号。`,
       delimiter: `分段标识符`,
+      html4excel: '表格转HTML',
+      html4excelTip: `Excel 是否将被解析为 HTML 表。如果为 FALSE，Excel 中的每一行都将形成一个块。`,
     },
     knowledgeConfiguration: {
       titleDescription: '在这里更新您的知识库详细信息，尤其是解析方法。',
diff --git a/web/src/pages/add-knowledge/components/knowledge-setting/configuration.tsx b/web/src/pages/add-knowledge/components/knowledge-setting/configuration.tsx
index 7e55f7608..64b1cbb82 100644
--- a/web/src/pages/add-knowledge/components/knowledge-setting/configuration.tsx
+++ b/web/src/pages/add-knowledge/components/knowledge-setting/configuration.tsx
@@ -1,5 +1,6 @@
 import Delimiter from '@/components/delimiter';
 import EntityTypesItem from '@/components/entity-types-item';
+import ExcelToHtml from '@/components/excel-to-html';
 import LayoutRecognize from '@/components/layout-recognize';
 import MaxTokenNumber from '@/components/max-token-number';
 import ParseConfiguration, {
@@ -124,6 +125,7 @@ const ConfigurationForm = ({ form }: { form: FormInstance }) => {
                   <MaxTokenNumber></MaxTokenNumber>
                   <Delimiter></Delimiter>
                   <LayoutRecognize></LayoutRecognize>
+                  <ExcelToHtml></ExcelToHtml>
                 </>
               )}
               {showRaptorParseConfiguration(parserId) && (