V4.9.4 feature (#4470)

* Training status (#4424) * dataset data training state (#4311) * dataset data training state * fix * fix ts * fix * fix api format * fix * fix * perf: count training * format * fix: dataset training state (#4417) * fix * add test * fix * fix * fix test * fix test * perf: training count * count * loading status --------- Co-authored-by: heheer <heheer@sealos.io> * doc * website sync feature (#4429) * perf: introduce BullMQ for website sync (#4403) * perf: introduce BullMQ for website sync * feat: new redis module * fix: remove graceful shutdown * perf: improve UI in dataset detail - Updated the "change" icon SVG file. - Modified i18n strings. - Added new i18n string "immediate_sync". - Improved UI in dataset detail page, including button icons and background colors. * refactor: Add chunkSettings to DatasetSchema * perf: website sync ux * env template * fix: clean up website dataset when updating chunk settings (#4420) * perf: check setting updated * perf: worker currency * feat: init script for website sync refactor (#4425) * website feature doc --------- Co-authored-by: a.e. <49438478+I-Info@users.noreply.github.com> * pro migration (#4388) (#4433) * pro migration * reuse customPdfParseType Co-authored-by: gggaaallleee <91131304+gggaaallleee@users.noreply.github.com> * perf: remove loading ui * feat: config chat file expired time * Redis cache (#4436) * perf: add Redis cache for vector counting (#4432) * feat: cache * perf: get cache key --------- Co-authored-by: a.e. <49438478+I-Info@users.noreply.github.com> * perf: mobile voice input (#4437) * update:Mobile voice interaction (#4362) * Add files via upload * Add files via upload * Update ollama.md * Update ollama.md * Add files via upload * Update useSpeech.ts * Update ChatInput.tsx * Update useSpeech.ts * Update ChatInput.tsx * Update useSpeech.ts * Update constants.ts * Add files via upload * Update ChatInput.tsx * Update useSpeech.ts * Update useSpeech.ts * Update useSpeech.ts * Update ChatInput.tsx * Add files via upload * Update common.json * Update VoiceInput.tsx * Update ChatInput.tsx * Update VoiceInput.tsx * Update useSpeech.ts * Update useSpeech.ts * Update common.json * Update common.json * Update common.json * Update VoiceInput.tsx * Update VoiceInput.tsx * Update ChatInput.tsx * Update VoiceInput.tsx * Update ChatInput.tsx * Update VoiceInput.tsx * Update ChatInput.tsx * Update useSpeech.ts * Update common.json * Update chat.json * Update common.json * Update chat.json * Update common.json * Update chat.json * Update VoiceInput.tsx * Update ChatInput.tsx * Update useSpeech.ts * Update VoiceInput.tsx * speech ui * 优化语音输入组件，调整输入框显示逻辑，修复语音输入遮罩层样式，更新画布背景透明度，增强用户交互体验。 (#4435) * perf: mobil voice input --------- Co-authored-by: dreamer6680 <1468683855@qq.com> * Test completion v2 (#4438) * add v2 completions (#4364) * add v2 completions * completion config * config version * fix * frontend * doc * fix * fix: completions v2 api --------- Co-authored-by: heheer <heheer@sealos.io> * package * Test mongo log (#4443) * feat: mongodb-log (#4426) * perf: mongo log * feat: completions stop reasoner * mongo db log --------- Co-authored-by: Finley Ge <32237950+FinleyGe@users.noreply.github.com> * update doc * Update doc * fix external var ui (#4444) * action * fix: ts (#4458) * preview doc action add docs preview permission update preview action udpate action * update doc (#4460) * update preview action * update doc * remove * update * schema * update mq export;perf: redis cache (#4465) * perf: redis cache * update mq export * perf: website sync error tip * add error worker * website sync ui (#4466) * Updated the dynamic display of the voice input pop-up (#4469) * Update VoiceInput.tsx * Update VoiceInput.tsx * Update VoiceInput.tsx * fix: voice input --------- Co-authored-by: heheer <heheer@sealos.io> Co-authored-by: a.e. <49438478+I-Info@users.noreply.github.com> Co-authored-by: gggaaallleee <91131304+gggaaallleee@users.noreply.github.com> Co-authored-by: dreamer6680 <1468683855@qq.com> Co-authored-by: Finley Ge <32237950+FinleyGe@users.noreply.github.com>
2025-04-08 12:05:04 +08:00
parent 5839325f77
commit f642c9603b
151 changed files with 5434 additions and 1354 deletions
--- a/packages/service/core/dataset/collection/controller.ts
+++ b/packages/service/core/dataset/collection/controller.ts
@@ -1,6 +1,7 @@
 import {
  DatasetCollectionTypeEnum,
-  DatasetCollectionDataProcessModeEnum
+  DatasetCollectionDataProcessModeEnum,
+  DatasetTypeEnum
 } from '@fastgpt/global/core/dataset/constants';
 import type { CreateDatasetCollectionParams } from '@fastgpt/global/core/dataset/api.d';
 import { MongoDatasetCollection } from './schema';
@@ -104,7 +105,8 @@ export const createCollectionAndInsertData = async ({
      hashRawText: hashStr(rawText),
      rawTextLength: rawText.length,
      nextSyncTime: (() => {
-        if (!dataset.autoSync) return undefined;
+        // ignore auto collections sync for website datasets
+        if (!dataset.autoSync && dataset.type === DatasetTypeEnum.websiteDataset) return undefined;
        if (
          [DatasetCollectionTypeEnum.link, DatasetCollectionTypeEnum.apiFile].includes(
            createCollectionParams.type
--- a/packages/service/core/dataset/collection/schema.ts
+++ b/packages/service/core/dataset/collection/schema.ts
@@ -1,13 +1,8 @@
 import { connectionMongo, getMongoModel } from '../../../common/mongo';
-const { Schema, model, models } = connectionMongo;
+const { Schema } = connectionMongo;
 import { DatasetCollectionSchemaType } from '@fastgpt/global/core/dataset/type.d';
-import {
-  DatasetCollectionTypeMap,
-  DatasetCollectionDataProcessModeEnum,
-  ChunkSettingModeEnum,
-  DataChunkSplitModeEnum
-} from '@fastgpt/global/core/dataset/constants';
-import { DatasetCollectionName } from '../schema';
+import { DatasetCollectionTypeMap } from '@fastgpt/global/core/dataset/constants';
+import { ChunkSettings, DatasetCollectionName } from '../schema';
 import {
  TeamCollectionName,
  TeamMemberCollectionName
@@ -90,25 +85,7 @@ const DatasetCollectionSchema = new Schema({
  customPdfParse: Boolean,

  // Chunk settings
-  imageIndex: Boolean,
-  autoIndexes: Boolean,
-  trainingType: {
-    type: String,
-    enum: Object.values(DatasetCollectionDataProcessModeEnum)
-  },
-  chunkSettingMode: {
-    type: String,
-    enum: Object.values(ChunkSettingModeEnum)
-  },
-  chunkSplitMode: {
-    type: String,
-    enum: Object.values(DataChunkSplitModeEnum)
-  },
-  chunkSize: Number,
-  chunkSplitter: String,
-
-  indexSize: Number,
-  qaPrompt: String
+  ...ChunkSettings
 });

 DatasetCollectionSchema.virtual('dataset', {
--- a/packages/service/core/dataset/controller.ts
+++ b/packages/service/core/dataset/controller.ts
@@ -9,6 +9,8 @@ import { deleteDatasetDataVector } from '../../common/vectorStore/controller';
 import { MongoDatasetDataText } from './data/dataTextSchema';
 import { DatasetErrEnum } from '@fastgpt/global/common/error/code/dataset';
 import { retryFn } from '@fastgpt/global/common/system/utils';
+import { removeWebsiteSyncJobScheduler } from './websiteSync';
+import { DatasetTypeEnum } from '@fastgpt/global/core/dataset/constants';

 /* ============= dataset ========== */
 /* find all datasetId by top datasetId */
--- a/packages/service/core/dataset/schema.ts
+++ b/packages/service/core/dataset/schema.ts
@@ -1,7 +1,8 @@
 import { getMongoModel, Schema } from '../../common/mongo';
 import {
-  DatasetStatusEnum,
-  DatasetStatusMap,
+  ChunkSettingModeEnum,
+  DataChunkSplitModeEnum,
+  DatasetCollectionDataProcessModeEnum,
  DatasetTypeEnum,
  DatasetTypeMap
 } from '@fastgpt/global/core/dataset/constants';
@@ -13,6 +14,28 @@ import type { DatasetSchemaType } from '@fastgpt/global/core/dataset/type.d';

 export const DatasetCollectionName = 'datasets';

+export const ChunkSettings = {
+  imageIndex: Boolean,
+  autoIndexes: Boolean,
+  trainingType: {
+    type: String,
+    enum: Object.values(DatasetCollectionDataProcessModeEnum)
+  },
+  chunkSettingMode: {
+    type: String,
+    enum: Object.values(ChunkSettingModeEnum)
+  },
+  chunkSplitMode: {
+    type: String,
+    enum: Object.values(DataChunkSplitModeEnum)
+  },
+  chunkSize: Number,
+  chunkSplitter: String,
+
+  indexSize: Number,
+  qaPrompt: String
+};
+
 const DatasetSchema = new Schema({
  parentId: {
    type: Schema.Types.ObjectId,
@@ -40,11 +63,6 @@ const DatasetSchema = new Schema({
    required: true,
    default: DatasetTypeEnum.dataset
  },
-  status: {
-    type: String,
-    enum: Object.keys(DatasetStatusMap),
-    default: DatasetStatusEnum.active
-  },
  avatar: {
    type: String,
    default: '/icon/logo.svg'
@@ -84,6 +102,9 @@ const DatasetSchema = new Schema({
      }
    }
  },
+  chunkSettings: {
+    type: ChunkSettings
+  },
  inheritPermission: {
    type: Boolean,
    default: true
@@ -98,9 +119,8 @@ const DatasetSchema = new Schema({
    type: Object
  },

-  autoSync: Boolean,
-
  // abandoned
+  autoSync: Boolean,
  externalReadUrl: {
    type: String
  },
--- a/packages/service/core/dataset/training/schema.ts
+++ b/packages/service/core/dataset/training/schema.ts
@@ -98,7 +98,9 @@ const TrainingDataSchema = new Schema({
      }
    ],
    default: []
-  }
+  },
+
+  errorMsg: String
 });

 TrainingDataSchema.virtual('dataset', {
--- a/packages/service/core/dataset/websiteSync/index.ts
+++ b/packages/service/core/dataset/websiteSync/index.ts
@@ -0,0 +1,101 @@
+import { Processor } from 'bullmq';
+import { getQueue, getWorker, QueueNames } from '../../../common/bullmq';
+import { DatasetStatusEnum } from '@fastgpt/global/core/dataset/constants';
+
+export type WebsiteSyncJobData = {
+  datasetId: string;
+};
+
+export const websiteSyncQueue = getQueue<WebsiteSyncJobData>(QueueNames.websiteSync, {
+  defaultJobOptions: {
+    attempts: 3, // retry 3 times
+    backoff: {
+      type: 'exponential',
+      delay: 1000 // delay 1 second between retries
+    }
+  }
+});
+export const getWebsiteSyncWorker = (processor: Processor<WebsiteSyncJobData>) => {
+  return getWorker<WebsiteSyncJobData>(QueueNames.websiteSync, processor, {
+    removeOnFail: {
+      age: 15 * 24 * 60 * 60, // Keep up to 15 days
+      count: 1000 // Keep up to 1000 jobs
+    },
+    concurrency: 1 // Set worker to process only 1 job at a time
+  });
+};
+
+export const addWebsiteSyncJob = (data: WebsiteSyncJobData) => {
+  const datasetId = String(data.datasetId);
+  // deduplication: make sure only 1 job
+  return websiteSyncQueue.add(datasetId, data, { deduplication: { id: datasetId } });
+};
+
+export const getWebsiteSyncDatasetStatus = async (datasetId: string) => {
+  const jobId = await websiteSyncQueue.getDeduplicationJobId(datasetId);
+  if (!jobId) {
+    return {
+      status: DatasetStatusEnum.active,
+      errorMsg: undefined
+    };
+  }
+  const job = await websiteSyncQueue.getJob(jobId);
+  if (!job) {
+    return {
+      status: DatasetStatusEnum.active,
+      errorMsg: undefined
+    };
+  }
+
+  const jobState = await job.getState();
+
+  if (jobState === 'failed' || jobState === 'unknown') {
+    return {
+      status: DatasetStatusEnum.error,
+      errorMsg: job.failedReason
+    };
+  }
+  if (['waiting-children', 'waiting'].includes(jobState)) {
+    return {
+      status: DatasetStatusEnum.waiting,
+      errorMsg: undefined
+    };
+  }
+  if (jobState === 'active') {
+    return {
+      status: DatasetStatusEnum.syncing,
+      errorMsg: undefined
+    };
+  }
+
+  return {
+    status: DatasetStatusEnum.active,
+    errorMsg: undefined
+  };
+};
+
+// Scheduler setting
+const repeatDuration = 24 * 60 * 60 * 1000; // every day
+export const upsertWebsiteSyncJobScheduler = (data: WebsiteSyncJobData, startDate?: number) => {
+  const datasetId = String(data.datasetId);
+
+  return websiteSyncQueue.upsertJobScheduler(
+    datasetId,
+    {
+      every: repeatDuration,
+      startDate: startDate || new Date().getTime() + repeatDuration // First run tomorrow
+    },
+    {
+      name: datasetId,
+      data
+    }
+  );
+};
+
+export const getWebsiteSyncJobScheduler = (datasetId: string) => {
+  return websiteSyncQueue.getJobScheduler(String(datasetId));
+};
+
+export const removeWebsiteSyncJobScheduler = (datasetId: string) => {
+  return websiteSyncQueue.removeJobScheduler(String(datasetId));
+};