Fix: Fixed an issue where parser configurations could be added infinitely #9869 (#10464)

### What problem does this PR solve?

Fix: Fixed an issue where parser configurations could be added
infinitely #9869

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

web/src/components/home-card.tsx

@@ -39,9 +39,9 @@ export function HomeCard({
           />
         </div>
         <div className="flex flex-col justify-between gap-1 flex-1 h-full w-[calc(100%-50px)]">
-          <section className="flex justify-between w-full">
-            <section className="flex gap-1 items-center w-full">
-              <div className="text-base font-bold w-80% text-ellipsis overflow-hidden leading-snug">
+          <section className="flex justify-between">
+            <section className="flex flex-1 min-w-0 gap-1 items-center">
+              <div className="text-base font-bold leading-snug truncate">
                 {data.name}
               </div>
               {icon}

web/src/layouts/next-header.tsx

@@ -106,7 +106,7 @@ export function Header() {
   }, [navigate]);
 
   return (
-    <section className="p-5 pr-14 flex justify-between items-center ">
+    <section className="py-5 px-10 flex justify-between items-center ">
       <div className="flex items-center gap-4">
         <img
           src={'/logo.svg'}

web/src/locales/en.ts

@@ -1729,13 +1729,15 @@ This delimiter is used to split the input text into several text pieces echo of
       regularExpressions: 'Regular Expressions',
       overlappedPercent: 'Overlapped percent',
       searchMethod: 'Search method',
+      searchMethodTip: `Defines how the content can be searched — by full-text, embedding, or both.
+The Tokenizer will store the content in the corresponding data structures for the selected methods.`,
       begin: 'File',
       parserMethod: 'Parsing method',
       systemPrompt: 'System Prompt',
       systemPromptPlaceholder:
         'Enter system prompt for image analysis, if empty the system default value will be used',
       exportJson: 'Export JSON',
-      viewResult: 'View Result',
+      viewResult: 'View result',
       running: 'Running',
       summary: 'Augmented Context',
       keywords: 'Keywords',

web/src/locales/zh.ts

@@ -1634,6 +1634,8 @@ General：实体和关系提取提示来自 GitHub - microsoft/graphrag：基于
       regularExpressions: '正则表达式',
       overlappedPercent: '重叠百分比',
       searchMethod: '搜索方法',
+      searchMethodTip: `决定该数据集启用的搜索方式，可选择全文、向量，或两者兼有。
+Tokenizer 会根据所选方式将内容存储为对应的数据结构。`,
       filenameEmbdWeight: '文件名嵌入权重',
       begin: '文件',
       parserMethod: '解析方法',

web/src/pages/data-flow/form/parser-form/common-form-fields.tsx

@@ -9,18 +9,29 @@ import {
   SelectWithSearchFlagOptionType,
 } from '@/components/originui/select-with-search';
 import { RAGFlowFormItem } from '@/components/ragflow-form';
-import { upperFirst } from 'lodash';
+import { upperCase, upperFirst } from 'lodash';
 import { useTranslation } from 'react-i18next';
-import { FileType, OutputFormatMap, PdfOutputFormat } from '../../constant';
+import {
+  FileType,
+  OutputFormatMap,
+  SpreadsheetOutputFormat,
+} from '../../constant';
 import { CommonProps } from './interface';
 import { buildFieldNameWithPrefix } from './utils';
 
+const UppercaseFields = [
+  SpreadsheetOutputFormat.Html,
+  SpreadsheetOutputFormat.Json,
+];
+
 function buildOutputOptionsFormatMap() {
   return Object.entries(OutputFormatMap).reduce<
     Record<string, SelectWithSearchFlagOptionType[]>
   >((pre, [key, value]) => {
     pre[key] = Object.values(value).map((v) => ({
-      label: v === PdfOutputFormat.Json ? 'JSON' : upperFirst(v),
+      label: UppercaseFields.some((x) => x === v)
+        ? upperCase(v)
+        : upperFirst(v),
       value: v,
     }));
     return pre;

web/src/pages/data-flow/form/parser-form/index.tsx

@@ -1,4 +1,7 @@
-import { SelectWithSearch } from '@/components/originui/select-with-search';
+import {
+  SelectWithSearch,
+  SelectWithSearchFlagOptionType,
+} from '@/components/originui/select-with-search';
 import { RAGFlowFormItem } from '@/components/ragflow-form';
 import { BlockButton, Button } from '@/components/ui/button';
 import { Form } from '@/components/ui/form';
@@ -49,6 +52,7 @@ type ParserItemProps = {
   index: number;
   fieldLength: number;
   remove: UseFieldArrayRemove;
+  fileFormatOptions: SelectWithSearchFlagOptionType[];
 };
 
 export const FormSchema = z.object({
@@ -67,7 +71,13 @@ export const FormSchema = z.object({
 
 export type ParserFormSchemaType = z.infer<typeof FormSchema>;
 
-function ParserItem({ name, index, fieldLength, remove }: ParserItemProps) {
+function ParserItem({
+  name,
+  index,
+  fieldLength,
+  remove,
+  fileFormatOptions,
+}: ParserItemProps) {
   const { t } = useTranslation();
   const form = useFormContext<ParserFormSchemaType>();
   const ref = useRef(null);
@@ -79,23 +89,15 @@ function ParserItem({ name, index, fieldLength, remove }: ParserItemProps) {
   const values = form.getValues();
   const parserList = values.setups.slice(); // Adding, deleting, or modifying the parser array will not change the reference.
 
-  const FileFormatOptions = buildOptions(
-    FileType,
-    t,
-    'dataflow.fileFormatOptions',
-  ).filter(
-    (x) => x.value !== FileType.Video, // Temporarily hide the video option
-  );
-
   const filteredFileFormatOptions = useMemo(() => {
     const otherFileFormatList = parserList
       .filter((_, idx) => idx !== index)
       .map((x) => x.fileFormat);
 
-    return FileFormatOptions.filter((x) => {
+    return fileFormatOptions.filter((x) => {
       return !otherFileFormatList.includes(x.value);
     });
-  }, [FileFormatOptions, index, parserList]);
+  }, [fileFormatOptions, index, parserList]);
 
   const Widget =
     typeof fileFormat === 'string' && fileFormat in FileFormatWidgetMap
@@ -158,6 +160,14 @@ const ParserForm = ({ node }: INextOperatorForm) => {
   const { t } = useTranslation();
   const defaultValues = useFormValues(initialParserValues, node);
 
+  const FileFormatOptions = buildOptions(
+    FileType,
+    t,
+    'dataflow.fileFormatOptions',
+  ).filter(
+    (x) => x.value !== FileType.Video, // Temporarily hide the video option
+  );
+
   const form = useForm<z.infer<typeof FormSchema>>({
     defaultValues,
     resolver: zodResolver(FormSchema),
@@ -194,12 +204,15 @@ const ParserForm = ({ node }: INextOperatorForm) => {
               index={index}
               fieldLength={fields.length}
               remove={remove}
+              fileFormatOptions={FileFormatOptions}
             ></ParserItem>
           );
         })}
-        <BlockButton onClick={add} type="button" className="mt-2.5">
-          {t('dataflow.addParser')}
-        </BlockButton>
+        {fields.length < FileFormatOptions.length && (
+          <BlockButton onClick={add} type="button" className="mt-2.5">
+            {t('dataflow.addParser')}
+          </BlockButton>
+        )}
       </form>
       <div className="p-5">
         <Output list={outputList}></Output>

web/src/pages/data-flow/form/tokenizer-form/index.tsx

@@ -58,6 +58,7 @@ const TokenizerForm = ({ node }: INextOperatorForm) => {
         <RAGFlowFormItem
           name="search_method"
           label={t('dataflow.searchMethod')}
+          tooltip={t('dataflow.searchMethodTip')}
         >
           {(field) => (
             <MultiSelect

web/src/pages/data-flow/index.tsx

@@ -162,9 +162,7 @@ export default function DataFlow() {
             onClick={handleRunAgent}
             loading={running}
           >
-            <CirclePlay
-              className={isParsing || isLogEmpty ? 'animate-spin' : ''}
-            />
+            <CirclePlay className={isParsing ? 'animate-spin' : ''} />
 
             {isParsing || running ? t('dataflow.running') : t('flow.run')}
           </ButtonLoading>

web/src/pages/data-flow/log-sheet/dataflow-timeline.tsx

@@ -78,7 +78,7 @@ export function DataflowTimeline({ traceList }: DataflowTimelineProps) {
                       <div className="flex-1 flex items-center gap-5">
                         <Progress value={progress} className="h-1 flex-1" />
                         <span className="text-accent-primary text-xs">
-                          {progress}%
+                          {progress.toFixed(2)}%
                         </span>
                       </div>
                     </section>

web/src/pages/data-flow/log-sheet/index.tsx

@@ -55,10 +55,10 @@ export function LogSheet({
   return (
     <Sheet open onOpenChange={hideModal} modal={false}>
       <SheetContent
-        className={cn('top-20')}
+        className={cn('top-20 h-auto flex flex-col p-0 gap-0')}
         onInteractOutside={(e) => e.preventDefault()}
       >
-        <SheetHeader>
+        <SheetHeader className="p-5">
           <SheetTitle className="flex items-center gap-2.5">
             <Logs className="size-4" /> {t('flow.log')}
             {isCompleted && (
@@ -82,30 +82,32 @@ export function LogSheet({
             )}
           </SheetTitle>
         </SheetHeader>
-        <section className="max-h-[82vh] overflow-auto mt-6">
+        <section className="flex-1 overflow-auto px-5 pt-5">
           {isLogEmpty ? (
             <SkeletonCard className="mt-2" />
           ) : (
             <DataflowTimeline traceList={logs}></DataflowTimeline>
           )}
         </section>
-        {isParsing ? (
-          <Button
-            className="w-full mt-8 bg-state-error/10 text-state-error hover:bg-state-error hover:text-bg-base"
-            onClick={handleCancel}
-          >
-            <CirclePause /> {t('dataflow.cancel')}
-          </Button>
-        ) : (
-          <Button
-            onClick={handleDownloadJson}
-            disabled={isEndOutputEmpty(logs)}
-            className="w-full mt-8 bg-accent-primary-5 text-text-secondary hover:bg-accent-primary-5  hover:text-accent-primary hover:border-accent-primary hover:border"
-          >
-            <SquareArrowOutUpRight />
-            {t('dataflow.exportJson')}
-          </Button>
-        )}
+        <div className="px-5 pb-5">
+          {isParsing ? (
+            <Button
+              className="w-full mt-8 bg-state-error/10 text-state-error hover:bg-state-error hover:text-bg-base"
+              onClick={handleCancel}
+            >
+              <CirclePause /> {t('dataflow.cancel')}
+            </Button>
+          ) : (
+            <Button
+              onClick={handleDownloadJson}
+              disabled={isEndOutputEmpty(logs)}
+              className="w-full mt-8 bg-accent-primary-5 text-text-secondary hover:bg-accent-primary-5  hover:text-accent-primary hover:border-accent-primary hover:border"
+            >
+              <SquareArrowOutUpRight />
+              {t('dataflow.exportJson')}
+            </Button>
+          )}
+        </div>
       </SheetContent>
     </Sheet>
   );

web/src/pages/home/banner.tsx

@@ -42,7 +42,7 @@ export function Banner() {
 export function NextBanner() {
   const { t } = useTranslation();
   return (
-    <section className="text-5xl pt-10 pb-14 font-bold">
+    <section className="text-5xl pt-10 pb-14 font-bold px-10">
       <span className="text-text-primary">{t('header.welcome')}</span>
       <span className="pl-3 text-transparent bg-clip-text  bg-gradient-to-l from-[#40EBE3] to-[#4A51FF]">
         RAGFlow

web/src/pages/home/index.tsx

@@ -4,15 +4,13 @@ import { Datasets } from './datasets';
 
 const Home = () => {
   return (
-    <div className="mx-8">
-      <section>
-        <NextBanner></NextBanner>
-        <section className="h-[calc(100dvh-260px)] overflow-auto scrollbar-thin">
-          <Datasets></Datasets>
-          <Applications></Applications>
-        </section>
+    <section>
+      <NextBanner></NextBanner>
+      <section className="h-[calc(100dvh-260px)] overflow-auto px-10">
+        <Datasets></Datasets>
+        <Applications></Applications>
       </section>
-    </div>
+    </section>
   );
 };