feat(file): Add file ancestor directory lookup feature by go (#14037)

### What problem does this PR solve? feat(file): Add file ancestor directory lookup feature by go ### Type of change - [x] New Feature (non-breaking change which adds functionality)
2026-08-01 13:33:48 +08:00 · 2026-04-14 15:22:03 +08:00
parent 6aec8058bb
commit 1031aebc8f
7 changed files with 224 additions and 133 deletions
--- a/web/src/pages/agent/form/parser-form/common-form-fields.tsx
+++ b/web/src/pages/agent/form/parser-form/common-form-fields.tsx
@@ -93,9 +93,10 @@ export function TwoColumnCheckFormField({ prefix }: CommonProps) {
  return (
    <RAGFlowFormItem
      name={buildFieldNameWithPrefix(`enable_multi_column`, prefix)}
-      label={t('flow.enableMultiColumn', 'Enable multi column')}
+      label={t('flow.enableMultiColumn')}
      horizontal={true}
-      labelClassName="w-[200px]"
+      labelClassName="w-full"
+      tooltip={t('flow.enableMultiColumnTip')}
    >
      {(field) => (
        <Checkbox
@@ -114,9 +115,10 @@ export function RmdirFormField({ prefix }: CommonProps) {
  return (
    <RAGFlowFormItem
      name={buildFieldNameWithPrefix(`remove_toc`, prefix)}
-      label={t('flow.remove_toc', 'Remove TOC')}
+      label={t('flow.removeToc')}
      horizontal={true}
-      labelClassName="w-[200px]"
+      tooltip={t('flow.removeTocTip')}
+      labelClassName="w-full"
    >
      {(field) => (
        <Checkbox
--- a/web/src/pages/agent/form/parser-form/index.tsx
+++ b/web/src/pages/agent/form/parser-form/index.tsx
@@ -5,14 +5,13 @@ import {
 import { RAGFlowFormItem } from '@/components/ragflow-form';
 import { BlockButton, Button } from '@/components/ui/button';
 import { Form } from '@/components/ui/form';
-import { MultiSelect } from '@/components/ui/multi-select';
 import { Separator } from '@/components/ui/separator';
 import { cn } from '@/lib/utils';
 import { buildOptions } from '@/utils/form';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { useHover } from 'ahooks';
 import { Trash2 } from 'lucide-react';
-import { memo, useCallback, useEffect, useMemo, useRef } from 'react';
+import { memo, useCallback, useMemo, useRef } from 'react';
 import {
  useFieldArray,
  UseFieldArrayRemove,
@@ -25,8 +24,6 @@ import {
  FileType,
  InitialOutputFormatMap,
  initialParserValues,
-  MAIN_CONTENT_PREPROCESS_VALUE,
-  PreprocessValue,
 } from '../../constant/pipeline';
 import { useFormValues } from '../../hooks/use-form-values';
 import { useWatchFormChange } from '../../hooks/use-watch-form-change';
@@ -49,82 +46,82 @@ import { WordFormFields } from './word-form-fields';

 const outputList = buildOutputList(initialParserValues.outputs);

-type PreprocessOptionConfig = {
-  value: PreprocessValue;
-  required?: boolean;
-};
+// type PreprocessOptionConfig = {
+//   value: PreprocessValue;
+//   required?: boolean;
+// };

-const DefaultPreprocessOptionConfigs: PreprocessOptionConfig[] = [
-  { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-];
+// const DefaultPreprocessOptionConfigs: PreprocessOptionConfig[] = [
+//   { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+// ];

-const PreprocessOptionConfigsMap: Partial<
-  Record<FileType, PreprocessOptionConfig[]>
-> = {
-  [FileType.PDF]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-    { value: PreprocessValue.abstract },
-    { value: PreprocessValue.author },
-    { value: PreprocessValue.section_title },
-  ],
-  [FileType.PowerPoint]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-  ],
-  [FileType.Spreadsheet]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-  ],
-  [FileType.TextMarkdown]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-    { value: PreprocessValue.section_title },
-  ],
-  [FileType.Code]: [{ value: MAIN_CONTENT_PREPROCESS_VALUE, required: true }],
-  [FileType.Html]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-    { value: PreprocessValue.section_title },
-  ],
-  [FileType.Doc]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-    { value: PreprocessValue.section_title },
-  ],
-  [FileType.Docx]: [
-    { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
-    { value: PreprocessValue.section_title },
-  ],
-};
+// const PreprocessOptionConfigsMap: Partial<
+//   Record<FileType, PreprocessOptionConfig[]>
+// > = {
+//   [FileType.PDF]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//     { value: PreprocessValue.abstract },
+//     { value: PreprocessValue.author },
+//     { value: PreprocessValue.section_title },
+//   ],
+//   [FileType.PowerPoint]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//   ],
+//   [FileType.Spreadsheet]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//   ],
+//   [FileType.TextMarkdown]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//     { value: PreprocessValue.section_title },
+//   ],
+//   [FileType.Code]: [{ value: MAIN_CONTENT_PREPROCESS_VALUE, required: true }],
+//   [FileType.Html]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//     { value: PreprocessValue.section_title },
+//   ],
+//   [FileType.Doc]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//     { value: PreprocessValue.section_title },
+//   ],
+//   [FileType.Docx]: [
+//     { value: MAIN_CONTENT_PREPROCESS_VALUE, required: true },
+//     { value: PreprocessValue.section_title },
+//   ],
+// };

-function getPreprocessOptionConfigs(fileType?: FileType) {
-  if (!fileType) {
-    return DefaultPreprocessOptionConfigs;
-  }
+// function getPreprocessOptionConfigs(fileType?: FileType) {
+//   if (!fileType) {
+//     return DefaultPreprocessOptionConfigs;
+//   }

-  return PreprocessOptionConfigsMap[fileType] ?? DefaultPreprocessOptionConfigs;
-}
+//   return PreprocessOptionConfigsMap[fileType] ?? DefaultPreprocessOptionConfigs;
+// }

-function normalizePreprocessValuesByFileType(
-  fileType: FileType | undefined,
-  values: string[] | undefined,
-) {
-  const optionConfigs = getPreprocessOptionConfigs(fileType);
-  const allowedValueSet = new Set(optionConfigs.map((x) => x.value));
-  const requiredValues = optionConfigs
-    .filter((x) => x.required)
-    .map((x) => x.value);
-  const normalizedOptionalValues = (Array.isArray(values) ? values : []).filter(
-    (value) => allowedValueSet.has(value as PreprocessValue),
-  ) as PreprocessValue[];
+// function normalizePreprocessValuesByFileType(
+//   fileType: FileType | undefined,
+//   values: string[] | undefined,
+// ) {
+//   const optionConfigs = getPreprocessOptionConfigs(fileType);
+//   const allowedValueSet = new Set(optionConfigs.map((x) => x.value));
+//   const requiredValues = optionConfigs
+//     .filter((x) => x.required)
+//     .map((x) => x.value);
+//   const normalizedOptionalValues = (Array.isArray(values) ? values : []).filter(
+//     (value) => allowedValueSet.has(value as PreprocessValue),
+//   ) as PreprocessValue[];

-  return Array.from(
-    new Set<PreprocessValue>([...requiredValues, ...normalizedOptionalValues]),
-  );
-}
+//   return Array.from(
+//     new Set<PreprocessValue>([...requiredValues, ...normalizedOptionalValues]),
+//   );
+// }

-function isSameStringArray(a: string[] | undefined, b: string[]) {
-  if (!a || a.length !== b.length) {
-    return false;
-  }
+// function isSameStringArray(a: string[] | undefined, b: string[]) {
+//   if (!a || a.length !== b.length) {
+//     return false;
+//   }

-  return a.every((item, idx) => item === b[idx]);
-}
+//   return a.every((item, idx) => item === b[idx]);
+// }

 const FileFormatWidgetMap = {
  [FileType.PDF]: PdfFormFields,
@@ -151,7 +148,7 @@ export const FormSchema = z.object({
  setups: z.array(
    z.object({
      fileFormat: z.string().nullish(),
-      preprocess: z.array(z.string()).optional(),
+      // preprocess: z.array(z.string()).optional(),
      output_format: z.string().optional(),
      parse_method: z.string().optional(),
      lang: z.string().optional(),
@@ -212,56 +209,56 @@ function ParserItem({
    [form, index],
  );

-  const handlePreprocessChange = useCallback(
-    (value: PreprocessValue[]) => {
-      form.setValue(`setups.${index}.preprocess`, value, {
-        shouldDirty: true,
-        shouldValidate: true,
-        shouldTouch: true,
-      });
-    },
-    [form, index],
-  );
+  // const handlePreprocessChange = useCallback(
+  //   (value: PreprocessValue[]) => {
+  //     form.setValue(`setups.${index}.preprocess`, value, {
+  //       shouldDirty: true,
+  //       shouldValidate: true,
+  //       shouldTouch: true,
+  //     });
+  //   },
+  //   [form, index],
+  // );

-  const preprocessOptions = useMemo(() => {
-    const optionConfigs = getPreprocessOptionConfigs(fileFormat as FileType);
+  // const preprocessOptions = useMemo(() => {
+  //   const optionConfigs = getPreprocessOptionConfigs(fileFormat as FileType);

-    return optionConfigs.map((optionConfig) => {
-      const labelMap: Record<string, string> = {
-        [MAIN_CONTENT_PREPROCESS_VALUE]: t('flow.preprocess.mainContent'),
-        [PreprocessValue.section_title]: t('flow.preprocess.sectionTitle'),
-        [PreprocessValue.abstract]: t('flow.preprocess.abstract'),
-        [PreprocessValue.author]: t('flow.preprocess.author'),
-      };
+  //   return optionConfigs.map((optionConfig) => {
+  //     const labelMap: Record<string, string> = {
+  //       [MAIN_CONTENT_PREPROCESS_VALUE]: t('flow.preprocess.mainContent'),
+  //       [PreprocessValue.section_title]: t('flow.preprocess.sectionTitle'),
+  //       [PreprocessValue.abstract]: t('flow.preprocess.abstract'),
+  //       [PreprocessValue.author]: t('flow.preprocess.author'),
+  //     };

-      const label = labelMap[optionConfig.value] || optionConfig.value;
+  //     const label = labelMap[optionConfig.value] || optionConfig.value;

-      return {
-        value: optionConfig.value,
-        disabled: optionConfig.required,
-        label: label,
-      };
-    });
-  }, [fileFormat, t]);
+  //     return {
+  //       value: optionConfig.value,
+  //       disabled: optionConfig.required,
+  //       label: label,
+  //     };
+  //   });
+  // }, [fileFormat, t]);

-  useEffect(() => {
-    const currentPreprocessValues = form.getValues(
-      `setups.${index}.preprocess`,
-    ) as string[] | undefined;
-    const normalizedPreprocessValues = normalizePreprocessValuesByFileType(
-      fileFormat as FileType,
-      currentPreprocessValues,
-    );
+  // useEffect(() => {
+  //   const currentPreprocessValues = form.getValues(
+  //     `setups.${index}.preprocess`,
+  //   ) as string[] | undefined;
+  //   const normalizedPreprocessValues = normalizePreprocessValuesByFileType(
+  //     fileFormat as FileType,
+  //     currentPreprocessValues,
+  //   );

-    if (
-      !isSameStringArray(currentPreprocessValues, normalizedPreprocessValues)
-    ) {
-      form.setValue(`setups.${index}.preprocess`, normalizedPreprocessValues, {
-        shouldDirty: false,
-        shouldValidate: true,
-      });
-    }
-  }, [fileFormat, form, index]);
+  //   if (
+  //     !isSameStringArray(currentPreprocessValues, normalizedPreprocessValues)
+  //   ) {
+  //     form.setValue(`setups.${index}.preprocess`, normalizedPreprocessValues, {
+  //       shouldDirty: false,
+  //       shouldValidate: true,
+  //     });
+  //   }
+  // }, [fileFormat, form, index]);

  return (
    <section
@@ -301,7 +298,7 @@ function ParserItem({
          fileType={fileFormat as FileType}
        />
      </div>
-      <RAGFlowFormItem
+      {/* <RAGFlowFormItem
        name={buildFieldNameWithPrefix(`preprocess`, prefix)}
        label={t('flow.preprocess.preprocess')}
      >
@@ -320,7 +317,7 @@ function ParserItem({
            options={preprocessOptions}
          ></MultiSelect>
        )}
-      </RAGFlowFormItem>
+      </RAGFlowFormItem> */}
      {index < fieldLength - 1 && <Separator />}
    </section>
  );
@@ -351,10 +348,10 @@ const ParserForm = ({ node }: INextOperatorForm) => {
      parse_method: '',
      lang: '',
      fields: [],
-      llm_id: '',
+      vlm: { llm_id: '' },
      table_result_type: '',
      markdown_image_response_type: '',
-      preprocess: [],
+      // preprocess: [],
    });
  }, [append]);

--- a/web/src/pages/agent/form/title-chunker-form/index.tsx
+++ b/web/src/pages/agent/form/title-chunker-form/index.tsx
@@ -8,7 +8,7 @@ import { Form } from '@/components/ui/form';
 import { Input } from '@/components/ui/input';
 import { zodResolver } from '@hookform/resolvers/zod';
 import { Trash2 } from 'lucide-react';
-import { memo, useEffect, useRef } from 'react';
+import { memo, useEffect, useRef, useState } from 'react';
 import { useFieldArray, useForm, useFormContext } from 'react-hook-form';
 import { useTranslation } from 'react-i18next';
 import { z } from 'zod';
@@ -197,6 +197,7 @@ const TitleChunkerForm = ({ node }: INextOperatorForm) => {
  });
  const isInitialized = useRef(false);
  const initialMode = useRef<string | undefined>(undefined);
+  const [showAllTip, setShowAllTip] = useState(true);

  const method = form.watch('method');
  const name = 'rules';
@@ -210,6 +211,7 @@ const TitleChunkerForm = ({ node }: INextOperatorForm) => {
    }

    if (method !== initialMode.current) {
+      setShowAllTip(true);
      const currentMode = initialMode.current;
      const hierarchyValue = form.getValues('hierarchy');
      const rulesValue = form.getValues('rules');
@@ -290,6 +292,34 @@ const TitleChunkerForm = ({ node }: INextOperatorForm) => {
            ],
          }}
        />
+        {/* <div className={cn("text-xs text-text-secondary w-full border p-1", showAllTip ? "block" : "")}>
+          {method === 'hierarchy' && t('flow.hierarchyTip')}
+          {method === 'group' && t('flow.groupTip')}
+        </div> */}
+        <div
+          className={`text-xs text-text-secondary w-full border rounded-sm p-2 cursor-pointer ${showAllTip ? 'block' : 'truncate'}`}
+          onClick={() => setShowAllTip(!showAllTip)}
+        >
+          <div className="flex flex-col justify-start items-center">
+            <span
+              className="flex self-start"
+              dangerouslySetInnerHTML={{
+                __html:
+                  method === 'hierarchy'
+                    ? t('flow.hierarchyTip')
+                    : method === 'group'
+                      ? t('flow.groupTip')
+                      : '',
+              }}
+            >
+              {/* {method === 'hierarchy' && t('flow.hierarchyTip')}
+              {method === 'group' && t('flow.groupTip')} */}
+            </span>
+            {/* <span className="flex ml-2 text-xs self-center">
+              {showAllTip ? '▲' : ''}
+            </span> */}
+          </div>
+        </div>
        <RAGFlowFormItem name={'hierarchy'} label={''}>
          <SelectWithSearch options={hierarchyOptions}></SelectWithSearch>
        </RAGFlowFormItem>
@@ -297,12 +327,9 @@ const TitleChunkerForm = ({ node }: INextOperatorForm) => {
          <RAGFlowFormItem
            name="include_heading_content"
            label={t('flow.includeHeadingContent', 'Include heading content')}
-            tooltip={t(
-              'flow.includeHeadingContentTip',
-              'When enabled, content directly under a heading is kept as its own chunk. Child chunks keep only the heading path.',
-            )}
+            tooltip={t('flow.includeHeadingContentTip')}
            horizontal={true}
-            labelClassName="w-[200px]"
+            labelClassName="w-full"
          >
            {(field) => (
              <Checkbox