4.6.4-alpha (#582)

files/deploy/fastgpt/docker-compose.yml

@@ -1,4 +1,5 @@
 # 非 host 版本, 不使用本机代理
+# (不懂 Docker 的，只需要关心 OPENAI_BASE_URL 和 CHAT_API_KEY 即可！)
 version: '3.3'
 services:
   pg:
@@ -47,7 +48,7 @@ services:
     environment:
       # root 密码，用户名为: root
       - DEFAULT_ROOT_PSW=1234
-      # 中转地址，如果是用官方号，不需要管
+      # 中转地址，如果是用官方号，不需要管。务必加 /v1
       - OPENAI_BASE_URL=https://api.openai.com/v1
       - CHAT_API_KEY=sk-xxxx
       - DB_MAX_LINK=5 # database max link

packages/global/common/error/code/common.ts

@@ -0,0 +1,24 @@
+import { ErrType } from '../errorCode';
+
+/* dataset: 507000 */
+const startCode = 507000;
+export enum CommonErrEnum {
+  fileNotFound = 'fileNotFound'
+}
+const datasetErr = [
+  {
+    statusText: CommonErrEnum.fileNotFound,
+    message: 'error.fileNotFound'
+  }
+];
+export default datasetErr.reduce((acc, cur, index) => {
+  return {
+    ...acc,
+    [cur.statusText]: {
+      code: startCode + index,
+      statusText: cur.statusText,
+      message: cur.message,
+      data: null
+    }
+  };
+}, {} as ErrType<`${CommonErrEnum}`>);

packages/global/common/error/code/dataset.ts

@@ -13,23 +13,23 @@ export enum DatasetErrEnum {
 const datasetErr = [
   {
     statusText: DatasetErrEnum.unAuthDataset,
-    message: '无权操作该知识库'
+    message: 'core.dataset.error.unAuthDataset'
   },
   {
     statusText: DatasetErrEnum.unAuthDatasetCollection,
-    message: '无权操作该数据集'
+    message: 'core.dataset.error.unAuthDatasetCollection'
   },
   {
     statusText: DatasetErrEnum.unAuthDatasetData,
-    message: '无权操作该数据'
+    message: 'core.dataset.error.unAuthDatasetData'
   },
   {
     statusText: DatasetErrEnum.unAuthDatasetFile,
-    message: '无权操作该文件'
+    message: 'core.dataset.error.unAuthDatasetFile'
   },
   {
     statusText: DatasetErrEnum.unCreateCollection,
-    message: '无权创建数据集'
+    message: 'core.dataset.error.unCreateCollection'
   },
   {
     statusText: DatasetErrEnum.unLinkCollection,

packages/global/common/error/errorCode.ts

@@ -6,6 +6,7 @@ import pluginErr from './code/plugin';
 import outLinkErr from './code/outLink';
 import teamErr from './code/team';
 import userErr from './code/user';
+import commonErr from './code/common';
 
 export const ERROR_CODE: { [key: number]: string } = {
   400: '请求失败',
@@ -96,5 +97,6 @@ export const ERROR_RESPONSE: Record<
   ...outLinkErr,
   ...teamErr,
   ...userErr,
-  ...pluginErr
+  ...pluginErr,
+  ...commonErr
 };

packages/global/common/file/api.d.ts

@@ -1,3 +1,10 @@
+export type UploadImgProps = {
+  base64Img: string;
+  expiredTime?: Date;
+  metadata?: Record<string, any>;
+  shareId?: string;
+};
+
 export type UrlFetchParams = {
   urlList: string[];
   selector?: string;

packages/global/common/file/tools.ts

@@ -49,7 +49,14 @@ export const cheerioToHtml = ({
     }
   });
 
-  return $(selector || 'body').html();
+  const html = $(selector || 'body')
+    .map((item, dom) => {
+      return $(dom).html();
+    })
+    .get()
+    .join('\n');
+
+  return html;
 };
 export const urlsFetch = async ({
   urlList,

packages/global/common/string/markdown.ts

@@ -26,10 +26,14 @@ export const simpleMarkdownText = (rawText: string) => {
   rawText = rawText.replace(/\\\\n/g, '\\n');
 
   // Remove headings and code blocks front spaces
-  ['####', '###', '##', '#', '```', '~~~'].forEach((item) => {
+  ['####', '###', '##', '#', '```', '~~~'].forEach((item, i) => {
+    const isMarkdown = i <= 3;
     const reg = new RegExp(`\\n\\s*${item}`, 'g');
     if (reg.test(rawText)) {
-      rawText = rawText.replace(new RegExp(`\\n\\s*(${item})`, 'g'), '\n$1');
+      rawText = rawText.replace(
+        new RegExp(`(\\n)\\s*(${item})`, 'g'),
+        isMarkdown ? '\n$1$2' : '$1$2'
+      );
     }
   });
 

packages/global/common/string/textSplitter.ts

@@ -12,12 +12,13 @@ export const splitText2Chunks = (props: {
   text: string;
   chunkLen: number;
   overlapRatio?: number;
+  customReg?: string[];
 }): {
   chunks: string[];
   tokens: number;
   overlapRatio?: number;
 } => {
-  let { text = '', chunkLen, overlapRatio = 0.2 } = props;
+  let { text = '', chunkLen, overlapRatio = 0.2, customReg = [] } = props;
   const splitMarker = 'SPLIT_HERE_SPLIT_HERE';
   const codeBlockMarker = 'CODE_BLOCK_LINE_MARKER';
   const overlapLen = Math.round(chunkLen * overlapRatio);
@@ -29,22 +30,29 @@ export const splitText2Chunks = (props: {
 
   // The larger maxLen is, the next sentence is less likely to trigger splitting
   const stepReges: { reg: RegExp; maxLen: number }[] = [
-    { reg: /^(#\s[^\n]+)\n/gm, maxLen: chunkLen * 1.4 },
-    { reg: /^(##\s[^\n]+)\n/gm, maxLen: chunkLen * 1.4 },
-    { reg: /^(###\s[^\n]+)\n/gm, maxLen: chunkLen * 1.4 },
-    { reg: /^(####\s[^\n]+)\n/gm, maxLen: chunkLen * 1.4 },
-
-    { reg: /([\n](`))/g, maxLen: chunkLen * 4 }, // code block
-    { reg: /([\n](?![\*\-|>0-9]))/g, maxLen: chunkLen * 1.8 }, // (?![\*\-|>`0-9]): markdown special char
-    { reg: /([\n])/g, maxLen: chunkLen * 1.4 },
-
-    { reg: /([。]|([a-zA-Z])\.\s)/g, maxLen: chunkLen * 1.4 },
-    { reg: /([！]|!\s)/g, maxLen: chunkLen * 1.4 },
-    { reg: /([？]|\?\s)/g, maxLen: chunkLen * 1.6 },
-    { reg: /([；]|;\s)/g, maxLen: chunkLen * 1.8 },
+    ...customReg.map((text) => ({ reg: new RegExp(`([${text}])`, 'g'), maxLen: chunkLen * 1.4 })),
+    { reg: /^(#\s[^\n]+)\n/gm, maxLen: chunkLen * 1.2 },
+    { reg: /^(##\s[^\n]+)\n/gm, maxLen: chunkLen * 1.2 },
+    { reg: /^(###\s[^\n]+)\n/gm, maxLen: chunkLen * 1.2 },
+    { reg: /^(####\s[^\n]+)\n/gm, maxLen: chunkLen * 1.2 },
+
+    { reg: /([\n]([`~]))/g, maxLen: chunkLen * 4 }, // code block
+    { reg: /([\n](?!\s*[\*\-|>0-9]))/g, maxLen: chunkLen * 2 }, // (?![\*\-|>`0-9]): markdown special char
+    { reg: /([\n])/g, maxLen: chunkLen * 1.2 },
+
+    { reg: /([。]|([a-zA-Z])\.\s)/g, maxLen: chunkLen * 1.2 },
+    { reg: /([！]|!\s)/g, maxLen: chunkLen * 1.2 },
+    { reg: /([？]|\?\s)/g, maxLen: chunkLen * 1.4 },
+    { reg: /([；]|;\s)/g, maxLen: chunkLen * 1.6 },
     { reg: /([，]|,\s)/g, maxLen: chunkLen * 2 }
   ];
 
+  const customRegLen = customReg.length;
+  const checkIsCustomStep = (step: number) => step < customRegLen;
+  const checkIsMarkdownSplit = (step: number) => step >= customRegLen && step <= 3 + customRegLen;
+  const checkIndependentChunk = (step: number) => step >= customRegLen && step <= 4 + customRegLen;
+  const checkForbidOverlap = (step: number) => step <= 6 + customRegLen;
+
   // if use markdown title split, Separate record title title
   const getSplitTexts = ({ text, step }: { text: string; step: number }) => {
     if (step >= stepReges.length) {
@@ -55,11 +63,13 @@ export const splitText2Chunks = (props: {
         }
       ];
     }
-    const isMarkdownSplit = step <= 3;
+    const isMarkdownSplit = checkIsMarkdownSplit(step);
+    const independentChunk = checkIndependentChunk(step);
+
     const { reg } = stepReges[step];
 
     const splitTexts = text
-      .replace(reg, isMarkdownSplit ? `${splitMarker}$1` : `$1${splitMarker}`)
+      .replace(reg, independentChunk ? `${splitMarker}$1` : `$1${splitMarker}`)
       .split(`${splitMarker}`)
       .filter((part) => part.trim());
 
@@ -76,7 +86,7 @@ export const splitText2Chunks = (props: {
   };
 
   const getOneTextOverlapText = ({ text, step }: { text: string; step: number }): string => {
-    const forbidOverlap = step <= 6;
+    const forbidOverlap = checkForbidOverlap(step);
     const maxOverlapLen = chunkLen * 0.4;
 
     // step >= stepReges.length: Do not overlap incomplete sentences
@@ -114,7 +124,8 @@ export const splitText2Chunks = (props: {
     lastText: string;
     mdTitle: string;
   }): string[] => {
-    const isMarkdownSplit = step <= 3;
+    const independentChunk = checkIndependentChunk(step);
+    const isCustomStep = checkIsCustomStep(step);
 
     // mini text
     if (text.length <= chunkLen) {
@@ -134,12 +145,13 @@ export const splitText2Chunks = (props: {
       return chunks;
     }
 
-    const { maxLen } = stepReges[step];
-    const minChunkLen = chunkLen * 0.7;
-
     // split text by special char
     const splitTexts = getSplitTexts({ text, step });
 
+    const maxLen = splitTexts.length > 1 ? stepReges[step].maxLen : chunkLen;
+    const minChunkLen = chunkLen * 0.7;
+    const miniChunkLen = 30;
+
     const chunks: string[] = [];
     for (let i = 0; i < splitTexts.length; i++) {
       const item = splitTexts[i];
@@ -170,8 +182,8 @@ export const splitText2Chunks = (props: {
           mdTitle: currentTitle
         });
         const lastChunk = innerChunks[innerChunks.length - 1];
-        // last chunk is too small, concat it to lastText
-        if (!isMarkdownSplit && lastChunk.length < minChunkLen) {
+        // last chunk is too small, concat it to lastText(next chunk start)
+        if (!independentChunk && lastChunk.length < minChunkLen) {
           chunks.push(...innerChunks.slice(0, -1));
           lastText = lastChunk;
         } else {
@@ -189,10 +201,14 @@ export const splitText2Chunks = (props: {
       lastText = newText;
 
       // markdown paragraph block: Direct addition; If the chunk size reaches, add a chunk
-      if (isMarkdownSplit || newTextLen >= chunkLen) {
+      if (
+        isCustomStep ||
+        (independentChunk && newTextLen > miniChunkLen) ||
+        newTextLen >= chunkLen
+      ) {
         chunks.push(`${currentTitle}${lastText}`);
 
-        lastText = isMarkdownSplit ? '' : getOneTextOverlapText({ text: lastText, step });
+        lastText = getOneTextOverlapText({ text: lastText, step });
       }
     }
 

packages/global/core/app/utils.ts

@@ -24,7 +24,7 @@ export const getDefaultAppForm = (templateId = 'fastgpt-universal'): AppSimpleEd
     dataset: {
       datasets: [],
       similarity: 0.4,
-      limit: 5,
+      limit: 1500,
       searchEmptyText: '',
       searchMode: DatasetSearchModeEnum.embedding
     },

packages/global/core/chat/constants.ts

@@ -55,3 +55,5 @@ export const LOGO_ICON = `/icon/logo.svg`;
 
 export const IMG_BLOCK_KEY = 'img-block';
 export const FILE_BLOCK_KEY = 'file-block';
+
+export const MARKDOWN_QUOTE_SIGN = 'QUOTE SIGN';

packages/global/core/module/template/system/datasetSearch.ts

@@ -54,17 +54,10 @@ export const DatasetSearchModule: FlowModuleTemplateType = {
     {
       key: ModuleInputKeyEnum.datasetLimit,
       type: FlowNodeInputTypeEnum.hidden,
-      label: '单次搜索上限',
-      description: '最多取 n 条记录作为本次问题引用',
-      value: 5,
+      label: '引用上限',
+      description: '单次搜索最大的 Tokens 数量，中文约1字=1.7Tokens，英文约1字=1Tokens',
+      value: 1500,
       valueType: ModuleDataTypeEnum.number,
-      min: 1,
-      max: 20,
-      step: 1,
-      markList: [
-        { label: '1', value: 1 },
-        { label: '20', value: 20 }
-      ],
       showTargetInApp: false,
       showTargetInPlugin: false
     },

packages/service/common/file/gridfs/controller.ts

@@ -3,6 +3,7 @@ import { BucketNameEnum } from '@fastgpt/global/common/file/constants';
 import fsp from 'fs/promises';
 import fs from 'fs';
 import { DatasetFileSchema } from '@fastgpt/global/core/dataset/type';
+import { delImgByFileIdList } from '../image/controller';
 
 export function getGFSCollection(bucket: `${BucketNameEnum}`) {
   return connectionMongo.connection.db.collection(`${bucket}.files`);
@@ -69,24 +70,65 @@ export async function getFileById({
     _id: new Types.ObjectId(fileId)
   });
 
-  if (!file) {
-    return Promise.reject('File not found');
-  }
+  // if (!file) {
+  //   return Promise.reject('File not found');
+  // }
 
-  return file;
+  return file || undefined;
 }
 
-export async function delFileById({
+export async function delFileByFileIdList({
   bucketName,
-  fileId
+  fileIdList,
+  retry = 3
 }: {
   bucketName: `${BucketNameEnum}`;
-  fileId: string;
+  fileIdList: string[];
+  retry?: number;
+}): Promise<any> {
+  try {
+    const bucket = getGridBucket(bucketName);
+
+    await Promise.all(fileIdList.map((id) => bucket.delete(new Types.ObjectId(id))));
+  } catch (error) {
+    if (retry > 0) {
+      return delFileByFileIdList({ bucketName, fileIdList, retry: retry - 1 });
+    }
+  }
+}
+// delete file by metadata(datasetId)
+export async function delFileByMetadata({
+  bucketName,
+  datasetId
+}: {
+  bucketName: `${BucketNameEnum}`;
+  datasetId?: string;
 }) {
   const bucket = getGridBucket(bucketName);
 
-  await bucket.delete(new Types.ObjectId(fileId));
-  return true;
+  const files = await bucket
+    .find(
+      {
+        ...(datasetId && { 'metadata.datasetId': datasetId })
+      },
+      {
+        projection: {
+          _id: 1
+        }
+      }
+    )
+    .toArray();
+
+  const idList = files.map((item) => String(item._id));
+
+  // delete img
+  await delImgByFileIdList(idList);
+
+  // delete file
+  await delFileByFileIdList({
+    bucketName,
+    fileIdList: idList
+  });
 }
 
 export async function getDownloadStream({