4.6.7 fix (#752)

projects/app/package.json

@@ -42,7 +42,6 @@
     "next": "13.5.2",
     "next-i18next": "^13.3.0",
     "nprogress": "^0.2.0",
-    "papaparse": "^5.4.1",
     "react": "18.2.0",
     "react-day-picker": "^8.7.1",
     "react-dom": "18.2.0",
@@ -66,7 +65,6 @@
     "@types/jsonwebtoken": "^9.0.3",
     "@types/lodash": "^4.14.191",
     "@types/node": "^20.8.5",
-    "@types/papaparse": "^5.3.7",
     "@types/react": "18.2.0",
     "@types/react-dom": "18.2.0",
     "@types/react-syntax-highlighter": "^15.5.6",

projects/app/public/locales/zh/common.json

@@ -226,7 +226,7 @@
     "Chat test": "测试对话",
     "Max Token": "单条数据上限",
     "Start chat": "立即对话",
-    "Total chars": "总字符数: {{total}}",
+    "Total chars": "总字数: {{total}}",
     "Total tokens": "总 Tokens: {{total}}",
     "ai": {
       "Model": "AI 模型",
@@ -541,8 +541,7 @@
             "success": "开始同步"
           }
         },
-        "training": {
-        }
+        "training": {}
       },
       "data": {
         "Auxiliary Data": "辅助数据",

projects/app/src/components/core/module/Flow/components/modules/ButtonEdge.tsx

@@ -17,7 +17,7 @@ const ButtonEdge = (props: EdgeProps) => {
     style = {}
   } = props;
 
-  const [labelX, labelY] = getBezierPath({
+  const [, labelX, labelY] = getBezierPath({
     sourceX,
     sourceY,
     sourcePosition,

projects/app/src/global/core/api/datasetRes.d.ts

@@ -8,6 +8,3 @@ import { DatasetCollectionSchemaType } from '@fastgpt/global/core/dataset/type';
 /* ======= collection =========== */
 
 /* ==== data ===== */
-export type PushDataResponse = {
-  insertLen: number;
-};

projects/app/src/global/core/dataset/api.d.ts

@@ -27,13 +27,7 @@ export type CreateDatasetParams = {
 export type InsertOneDatasetDataProps = PushDatasetDataChunkProps & {
   collectionId: string;
 };
-export type PushDatasetDataProps = {
-  collectionId: string;
-  data: PushDatasetDataChunkProps[];
-  trainingMode: `${TrainingModeEnum}`;
-  prompt?: string;
-  billId?: string;
-};
+
 export type UpdateDatasetDataProps = {
   id: string;
   q?: string; // embedding content

projects/app/src/pages/api/core/dataset/collection/create/text.ts

@@ -16,11 +16,15 @@ import { checkDatasetLimit } from '@fastgpt/service/support/permission/limit/dat
 import { predictDataLimitLength } from '@fastgpt/global/core/dataset/utils';
 import { pushDataToTrainingQueue } from '@/service/core/dataset/data/controller';
 import { hashStr } from '@fastgpt/global/common/string/tools';
+import { createTrainingBill } from '@fastgpt/service/support/wallet/bill/controller';
+import { BillSourceEnum } from '@fastgpt/global/support/wallet/bill/constants';
+import { getQAModel, getVectorModel } from '@/service/core/ai/model';
 
 export default async function handler(req: NextApiRequest, res: NextApiResponse<any>) {
   try {
     await connectToDatabase();
     const {
+      name,
       text,
       trainingType = TrainingModeEnum.chunk,
       chunkSize = 512,
@@ -29,7 +33,7 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse<
       ...body
     } = req.body as TextCreateDatasetCollectionParams;
 
-    const { teamId, tmbId } = await authDataset({
+    const { teamId, tmbId, dataset } = await authDataset({
       req,
       authToken: true,
       authApiKey: true,
@@ -52,21 +56,32 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse<
       insertLen: predictDataLimitLength(trainingType, chunks)
     });
 
-    // 3. create collection
-    const collectionId = await createOneCollection({
-      ...body,
-      teamId,
-      tmbId,
-      type: DatasetCollectionTypeEnum.virtual,
+    // 3. create collection and training bill
+    const [collectionId, { billId }] = await Promise.all([
+      createOneCollection({
+        ...body,
+        teamId,
+        tmbId,
+        type: DatasetCollectionTypeEnum.virtual,
 
-      trainingType,
-      chunkSize,
-      chunkSplitter,
-      qaPrompt,
+        name,
+        trainingType,
+        chunkSize,
+        chunkSplitter,
+        qaPrompt,
 
-      hashRawText: hashStr(text),
-      rawTextLength: text.length
-    });
+        hashRawText: hashStr(text),
+        rawTextLength: text.length
+      }),
+      createTrainingBill({
+        teamId,
+        tmbId,
+        appName: name,
+        billSource: BillSourceEnum.training,
+        vectorModel: getVectorModel(dataset.vectorModel)?.name,
+        agentModel: getQAModel(dataset.agentModel)?.name
+      })
+    ]);
 
     // 4. push chunks to training queue
     const insertResults = await pushDataToTrainingQueue({
@@ -74,6 +89,8 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse<
       tmbId,
       collectionId,
       trainingMode: trainingType,
+      prompt: qaPrompt,
+      billId,
       data: chunks.map((text, index) => ({
         q: text,
         chunkIndex: index
@@ -90,3 +107,11 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse<
     });
   }
 }
+
+export const config = {
+  api: {
+    bodyParser: {
+      sizeLimit: '10mb'
+    }
+  }
+};

projects/app/src/pages/api/core/dataset/data/pushData.ts

@@ -3,8 +3,10 @@ import type { NextApiRequest, NextApiResponse } from 'next';
 import { jsonRes } from '@fastgpt/service/common/response';
 import { connectToDatabase } from '@/service/mongo';
 import { withNextCors } from '@fastgpt/service/common/middle/cors';
-import type { PushDataResponse } from '@/global/core/api/datasetRes.d';
-import type { PushDatasetDataProps } from '@/global/core/dataset/api.d';
+import type {
+  PushDatasetDataProps,
+  PushDatasetDataResponse
+} from '@fastgpt/global/core/dataset/api.d';
 import { authDatasetCollection } from '@fastgpt/service/support/permission/auth/dataset';
 import { checkDatasetLimit } from '@fastgpt/service/support/permission/limit/dataset';
 import { predictDataLimitLength } from '@fastgpt/global/core/dataset/utils';
@@ -39,7 +41,7 @@ export default withNextCors(async function handler(req: NextApiRequest, res: Nex
       insertLen: predictDataLimitLength(collection.trainingType, data)
     });
 
-    jsonRes<PushDataResponse>(res, {
+    jsonRes<PushDatasetDataResponse>(res, {
       data: await pushDataToTrainingQueue({
         ...req.body,
         teamId,

projects/app/src/pages/api/plusApi/[...path].ts

@@ -12,16 +12,13 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse)
     const method = (req.method || 'POST') as Method;
     const { path = [], ...query } = req.query as any;
 
-    const url = `/${path?.join('/')}`;
+    const url = `/${path?.join('/')}?${new URLSearchParams(query).toString()}`;
 
     if (!url) {
       throw new Error('url is empty');
     }
 
-    const data = {
-      ...req.body,
-      ...query
-    };
+    const data = req.body || query;
 
     const repose = await request(
       url,
@@ -56,3 +53,12 @@ export default async function handler(req: NextApiRequest, res: NextApiResponse)
     });
   }
 }
+
+export const config = {
+  api: {
+    bodyParser: {
+      sizeLimit: '10mb'
+    },
+    responseLimit: '10mb'
+  }
+};

projects/app/src/pages/dataset/detail/components/Import/diffSource/FileLocal.tsx

@@ -27,7 +27,7 @@ const Upload = dynamic(() => import('../commonProgress/Upload'));
 const PreviewRawText = dynamic(() => import('../components/PreviewRawText'));
 
 type FileItemType = ImportSourceItemType & { file: File };
-const fileType = '.txt, .docx, .pdf, .md, .html';
+const fileType = '.txt, .docx, .csv, .pdf, .md, .html';
 const maxSelectFileCount = 1000;
 
 const FileLocal = ({ activeStep, goToNext }: ImportDataComponentProps) => {

projects/app/src/pages/dataset/detail/components/Import/diffSource/TableLocal.tsx

@@ -14,7 +14,8 @@ import { useImportStore } from '../Provider';
 import { feConfigs } from '@/web/common/system/staticData';
 
 import dynamic from 'next/dynamic';
-import { fileDownload, readCsvContent } from '@/web/common/file/utils';
+import { fileDownload } from '@/web/common/file/utils';
+import { readCsvContent } from '@fastgpt/web/common/file/read/csv';
 
 const PreviewData = dynamic(() => import('../commonProgress/PreviewData'));
 const Upload = dynamic(() => import('../commonProgress/Upload'));
@@ -56,7 +57,7 @@ const SelectFile = React.memo(function SelectFile({ goToNext }: { goToNext: () =
       {
         for await (const selectFile of files) {
           const { file, folderPath } = selectFile;
-          const { header, data } = await readCsvContent(file);
+          const { header, data } = await readCsvContent({ file });
 
           const filterData: FileItemType['chunks'] = data
             .filter((item) => item[0])

projects/app/src/pages/dataset/detail/components/InputDataModal.tsx

@@ -193,7 +193,10 @@ const InputDataModal = ({
       // not exactly same
       await putDatasetDataById({
         id: dataId,
-        ...e
+        ...e,
+        indexes: e.indexes.map((index) =>
+          index.defaultIndex ? getDefaultIndex({ q: e.q, a: e.a }) : index
+        )
       });
 
       return {

projects/app/src/pages/dataset/detail/components/Test.tsx

@@ -35,7 +35,8 @@ import dynamic from 'next/dynamic';
 import { useForm } from 'react-hook-form';
 import MySelect from '@/components/Select';
 import { useSelectFile } from '@/web/common/file/hooks/useSelectFile';
-import { fileDownload, readCsvContent } from '@/web/common/file/utils';
+import { fileDownload } from '@/web/common/file/utils';
+import { readCsvContent } from '@fastgpt/web/common/file/read/csv';
 import { delay } from '@fastgpt/global/common/system/utils';
 import QuoteItem from '@/components/core/dataset/QuoteItem';
 
@@ -125,7 +126,7 @@ const Test = ({ datasetId }: { datasetId: string }) => {
   const { mutate: onFileTest, isLoading: fileTestIsLoading } = useRequest({
     mutationFn: async ({ searchParams }: FormType) => {
       if (!selectFile) return Promise.reject('File is not selected');
-      const { data } = await readCsvContent(selectFile);
+      const { data } = await readCsvContent({ file: selectFile });
       const testList = data.slice(0, 100);
       const results: SearchTestResponse[] = [];
 

projects/app/src/service/common/system/cron.ts

@@ -3,6 +3,11 @@ import { generateQA } from '@/service/events/generateQA';
 import { generateVector } from '@/service/events/generateVector';
 import { setCron } from '@fastgpt/service/common/system/cron';
 
+export const startCron = () => {
+  setUpdateSystemConfigCron();
+  setTrainingQueueCron();
+};
+
 export const setUpdateSystemConfigCron = () => {
   setCron('*/5 * * * *', () => {
     initSystemConfig();
@@ -11,7 +16,7 @@ export const setUpdateSystemConfigCron = () => {
 };
 
 export const setTrainingQueueCron = () => {
-  setCron('*/3 * * * *', () => {
+  setCron('*/1 * * * *', () => {
     generateVector();
     generateQA();
   });

projects/app/src/service/core/dataset/data/controller.ts

@@ -9,13 +9,11 @@ import {
   recallFromVectorStore,
   updateDatasetDataVector
 } from '@fastgpt/service/common/vectorStore/controller';
-import { Types } from 'mongoose';
 import {
   DatasetDataIndexTypeEnum,
   DatasetSearchModeEnum,
   DatasetSearchModeMap,
-  SearchScoreTypeEnum,
-  TrainingModeEnum
+  SearchScoreTypeEnum
 } from '@fastgpt/global/core/dataset/constants';
 import { getDefaultIndex } from '@fastgpt/global/core/dataset/utils';
 import { jiebaSplit } from '@/service/common/string/jieba';
@@ -29,172 +27,26 @@ import {
 } from '@fastgpt/global/core/dataset/type';
 import { reRankRecall } from '../../ai/rerank';
 import { countPromptTokens } from '@fastgpt/global/common/string/tiktoken';
-import { hashStr, simpleText } from '@fastgpt/global/common/string/tools';
-import type { PushDatasetDataProps } from '@/global/core/dataset/api.d';
-import type { PushDataResponse } from '@/global/core/api/datasetRes';
-import { PushDatasetDataChunkProps } from '@fastgpt/global/core/dataset/api';
-import { MongoDatasetTraining } from '@fastgpt/service/core/dataset/training/schema';
-import { startQueue } from '@/service/utils/tools';
-import { getCollectionWithDataset } from '@fastgpt/service/core/dataset/controller';
-import { getQAModel, getVectorModel } from '../../ai/model';
-import { delay } from '@fastgpt/global/common/system/utils';
+import { hashStr } from '@fastgpt/global/common/string/tools';
+import type {
+  PushDatasetDataProps,
+  PushDatasetDataResponse
+} from '@fastgpt/global/core/dataset/api.d';
+import { pushDataListToTrainingQueue } from '@fastgpt/service/core/dataset/training/controller';
 
-export async function pushDataToTrainingQueue({
-  teamId,
-  tmbId,
-  collectionId,
-  data,
-  prompt,
-  billId,
-  trainingMode
-}: {
-  teamId: string;
-  tmbId: string;
-} & PushDatasetDataProps): Promise<PushDataResponse> {
-  const checkModelValid = async ({ collectionId }: { collectionId: string }) => {
-    const {
-      datasetId: { _id: datasetId, vectorModel, agentModel }
-    } = await getCollectionWithDataset(collectionId);
-
-    if (trainingMode === TrainingModeEnum.chunk) {
-      if (!collectionId) return Promise.reject(`CollectionId is empty`);
-      const vectorModelData = getVectorModel(vectorModel);
-      if (!vectorModelData) {
-        return Promise.reject(`Model ${vectorModel} is inValid`);
-      }
-
-      return {
-        datasetId,
-        maxToken: vectorModelData.maxToken * 1.5,
-        model: vectorModelData.model,
-        weight: vectorModelData.weight
-      };
-    }
-
-    if (trainingMode === TrainingModeEnum.qa) {
-      const qaModelData = getQAModel(agentModel);
-      if (!qaModelData) {
-        return Promise.reject(`Model ${agentModel} is inValid`);
-      }
-      return {
-        datasetId,
-        maxToken: qaModelData.maxContext * 0.8,
-        model: qaModelData.model,
-        weight: 0
-      };
-    }
-    return Promise.reject(`Mode ${trainingMode} is inValid`);
-  };
-
-  const { datasetId, model, maxToken, weight } = await checkModelValid({
-    collectionId
+export async function pushDataToTrainingQueue(
+  props: {
+    teamId: string;
+    tmbId: string;
+  } & PushDatasetDataProps
+): Promise<PushDatasetDataResponse> {
+  const result = await pushDataListToTrainingQueue({
+    ...props,
+    vectorModelList: global.vectorModels,
+    qaModelList: global.qaModels
   });
 
-  // format q and a, remove empty char
-  data.forEach((item) => {
-    item.q = simpleText(item.q);
-    item.a = simpleText(item.a);
-
-    item.indexes = item.indexes
-      ?.map((index) => {
-        return {
-          ...index,
-          text: simpleText(index.text)
-        };
-      })
-      .filter(Boolean);
-  });
-
-  // filter repeat or equal content
-  const set = new Set();
-  const filterResult: Record<string, PushDatasetDataChunkProps[]> = {
-    success: [],
-    overToken: [],
-    repeat: [],
-    error: []
-  };
-
-  data.forEach((item) => {
-    if (!item.q) {
-      filterResult.error.push(item);
-      return;
-    }
-
-    const text = item.q + item.a;
-
-    // count q token
-    const token = countPromptTokens(item.q);
-
-    if (token > maxToken) {
-      filterResult.overToken.push(item);
-      return;
-    }
-
-    if (set.has(text)) {
-      console.log('repeat', item);
-      filterResult.repeat.push(item);
-    } else {
-      filterResult.success.push(item);
-      set.add(text);
-    }
-  });
-
-  // 插入记录
-  const insertData = async (dataList: PushDatasetDataChunkProps[], retry = 3): Promise<number> => {
-    try {
-      const results = await MongoDatasetTraining.insertMany(
-        dataList.map((item, i) => ({
-          teamId,
-          tmbId,
-          datasetId,
-          collectionId,
-          billId,
-          mode: trainingMode,
-          prompt,
-          model,
-          q: item.q,
-          a: item.a,
-          chunkIndex: item.chunkIndex ?? i,
-          weight: weight ?? 0,
-          indexes: item.indexes
-        }))
-      );
-      await delay(500);
-      return results.length;
-    } catch (error) {
-      if (retry > 0) {
-        await delay(1000);
-        return insertData(dataList, retry - 1);
-      }
-      return Promise.reject(error);
-    }
-  };
-
-  let insertLen = 0;
-  const chunkSize = 50;
-  const chunkList = filterResult.success.reduce(
-    (acc, cur) => {
-      const lastChunk = acc[acc.length - 1];
-      if (lastChunk.length < chunkSize) {
-        lastChunk.push(cur);
-      } else {
-        acc.push([cur]);
-      }
-      return acc;
-    },
-    [[]] as PushDatasetDataChunkProps[][]
-  );
-  for await (const chunks of chunkList) {
-    insertLen += await insertData(chunks);
-  }
-
-  startQueue();
-  delete filterResult.success;
-
-  return {
-    insertLen,
-    ...filterResult
-  };
+  return result;
 }
 
 /* insert data.
@@ -341,6 +193,11 @@ export async function updateData2Dataset({
             text: qaStr
           }
         });
+      } else {
+        patchResult.push({
+          type: 'unChange',
+          index: item
+        });
       }
     } else {
       // not in database, create
@@ -379,6 +236,7 @@ export async function updateData2Dataset({
           model
         });
         item.index.dataId = result.insertId;
+
         return result;
       }
       if (item.type === 'delete' && item.index.dataId) {
@@ -397,13 +255,14 @@ export async function updateData2Dataset({
   );
 
   const charsLength = result.reduce((acc, cur) => acc + cur.charsLength, 0);
+  const newIndexes = patchResult.filter((item) => item.type !== 'delete').map((item) => item.index);
 
   // update mongo other data
   mongoData.q = q || mongoData.q;
   mongoData.a = a ?? mongoData.a;
   mongoData.fullTextToken = jiebaSplit({ text: mongoData.q + mongoData.a });
   // @ts-ignore
-  mongoData.indexes = indexes;
+  mongoData.indexes = newIndexes;
   await mongoData.save();
 
   return {

projects/app/src/service/mongo.ts

@@ -7,7 +7,7 @@ import { createDefaultTeam } from '@fastgpt/service/support/user/team/controller
 import { exit } from 'process';
 import { initVectorStore } from '@fastgpt/service/common/vectorStore/controller';
 import { getInitConfig } from '@/pages/api/common/system/getInitData';
-import { setUpdateSystemConfigCron, setTrainingQueueCron } from './common/system/cron';
+import { startCron } from './common/system/cron';
 
 /**
  * connect MongoDB and init data
@@ -23,8 +23,7 @@ export function connectToDatabase(): Promise<void> {
       getInitConfig();
 
       // cron
-      setUpdateSystemConfigCron();
-      setTrainingQueueCron();
+      startCron();
 
       initRootUser();
     }

projects/app/src/web/common/file/controller.ts

@@ -32,13 +32,24 @@ export const uploadFiles = ({
   });
 };
 
-export const getUploadBase64ImgController = (props: CompressImgProps & UploadImgProps) =>
-  compressBase64ImgAndUpload({
-    maxW: 4000,
-    maxH: 4000,
-    maxSize: 1024 * 1024 * 5,
-    ...props
-  });
+export const getUploadBase64ImgController = (
+  props: CompressImgProps & UploadImgProps,
+  retry = 3
+): Promise<string> => {
+  try {
+    return compressBase64ImgAndUpload({
+      maxW: 4000,
+      maxH: 4000,
+      maxSize: 1024 * 1024 * 5,
+      ...props
+    });
+  } catch (error) {
+    if (retry > 0) {
+      return getUploadBase64ImgController(props, retry - 1);
+    }
+    return Promise.reject(error);
+  }
+};
 
 /**
  * compress image. response base64

projects/app/src/web/common/file/utils.ts

@@ -1,29 +1,3 @@
-import Papa from 'papaparse';
-import { readFileRawText } from '@fastgpt/web/common/file/read/rawText';
-
-/**
- * read csv to json
- * @response {
- *  header: string[],
- *  data: string[][]
- * }
- */
-export const readCsvContent = async (file: File) => {
-  try {
-    const { rawText: textArr } = await readFileRawText(file);
-    const csvArr = Papa.parse(textArr).data as string[][];
-    if (csvArr.length === 0) {
-      throw new Error('csv 解析失败');
-    }
-    return {
-      header: csvArr.shift() as string[],
-      data: csvArr.map((item) => item)
-    };
-  } catch (error) {
-    return Promise.reject('解析 csv 文件失败');
-  }
-};
-
 /**
  * file download by text
  */

projects/app/src/web/core/dataset/api.ts

@@ -19,12 +19,14 @@ import type {
   SearchTestResponse
 } from '@/global/core/dataset/api.d';
 import type {
-  PushDatasetDataProps,
   UpdateDatasetDataProps,
   CreateDatasetParams,
   InsertOneDatasetDataProps
 } from '@/global/core/dataset/api.d';
-import type { PushDataResponse } from '@/global/core/api/datasetRes.d';
+import type {
+  PushDatasetDataProps,
+  PushDatasetDataResponse
+} from '@fastgpt/global/core/dataset/api.d';
 import type { DatasetCollectionItemType } from '@fastgpt/global/core/dataset/type';
 import {
   DatasetCollectionSyncResultEnum,
@@ -97,7 +99,7 @@ export const getDatasetDataItemById = (id: string) =>
  * push data to training queue
  */
 export const postChunks2Dataset = (data: PushDatasetDataProps) =>
-  POST<PushDataResponse>(`/core/dataset/data/pushData`, data);
+  POST<PushDatasetDataResponse>(`/core/dataset/data/pushData`, data);
 
 /**
  * insert one data to dataset (immediately insert)