4.6.7-alpha commit (#743)

Co-authored-by: Archer <545436317@qq.com>
Co-authored-by: heheer <71265218+newfish-cmyk@users.noreply.github.com>
This commit is contained in:
Archer
2024-01-19 11:17:28 +08:00
committed by GitHub
parent 8ee7407c4c
commit c031e6dcc9
324 changed files with 8509 additions and 4757 deletions

View File

@@ -3,9 +3,10 @@ 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';
import { MongoFileSchema } from './schema';
export function getGFSCollection(bucket: `${BucketNameEnum}`) {
MongoFileSchema;
return connectionMongo.connection.db.collection(`${bucket}.files`);
}
export function getGridBucket(bucket: `${BucketNameEnum}`) {
@@ -21,6 +22,7 @@ export async function uploadFile({
tmbId,
path,
filename,
contentType,
metadata = {}
}: {
bucketName: `${BucketNameEnum}`;
@@ -28,6 +30,7 @@ export async function uploadFile({
tmbId: string;
path: string;
filename: string;
contentType?: string;
metadata?: Record<string, any>;
}) {
if (!path) return Promise.reject(`filePath is empty`);
@@ -44,7 +47,7 @@ export async function uploadFile({
const stream = bucket.openUploadStream(filename, {
metadata,
contentType: metadata?.contentType
contentType
});
// save to gridfs
@@ -96,40 +99,6 @@ export async function delFileByFileIdList({
}
}
}
// delete file by metadata(datasetId)
export async function delFileByMetadata({
bucketName,
datasetId
}: {
bucketName: `${BucketNameEnum}`;
datasetId?: string;
}) {
const bucket = getGridBucket(bucketName);
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({
bucketName,

View File

@@ -0,0 +1,15 @@
import { connectionMongo, type Model } from '../../mongo';
const { Schema, model, models } = connectionMongo;
const FileSchema = new Schema({});
try {
FileSchema.index({ 'metadata.teamId': 1 });
FileSchema.index({ 'metadata.uploadDate': -1 });
} catch (error) {
console.log(error);
}
export const MongoFileSchema = models['dataset.files'] || model('dataset.files', FileSchema);
MongoFileSchema.syncIndexes();

View File

@@ -46,8 +46,8 @@ export async function readMongoImg({ id }: { id: string }) {
return data?.binary;
}
export async function delImgByFileIdList(fileIds: string[]) {
export async function delImgByRelatedId(relateIds: string[]) {
return MongoImage.deleteMany({
'metadata.fileId': { $in: fileIds.map((item) => String(item)) }
'metadata.relatedId': { $in: relateIds.map((id) => String(id)) }
});
}

View File

@@ -35,6 +35,8 @@ try {
ImageSchema.index({ expiredTime: 1 }, { expireAfterSeconds: 60 });
ImageSchema.index({ type: 1 });
ImageSchema.index({ teamId: 1 });
ImageSchema.index({ createTime: 1 });
ImageSchema.index({ 'metadata.relatedId': 1 });
} catch (error) {
console.log(error);
}

View File

@@ -1,68 +0,0 @@
import * as pdfjs from 'pdfjs-dist/legacy/build/pdf.mjs';
// @ts-ignore
import('pdfjs-dist/legacy/build/pdf.worker.min.mjs');
import { ReadFileParams } from './type';
type TokenType = {
str: string;
dir: string;
width: number;
height: number;
transform: number[];
fontName: string;
hasEOL: boolean;
};
export const readPdfFile = async ({ path }: ReadFileParams) => {
const readPDFPage = async (doc: any, pageNo: number) => {
const page = await doc.getPage(pageNo);
const tokenizedText = await page.getTextContent();
const viewport = page.getViewport({ scale: 1 });
const pageHeight = viewport.height;
const headerThreshold = pageHeight * 0.95;
const footerThreshold = pageHeight * 0.05;
const pageTexts: TokenType[] = tokenizedText.items.filter((token: TokenType) => {
return (
!token.transform ||
(token.transform[5] < headerThreshold && token.transform[5] > footerThreshold)
);
});
// concat empty string 'hasEOL'
for (let i = 0; i < pageTexts.length; i++) {
const item = pageTexts[i];
if (item.str === '' && pageTexts[i - 1]) {
pageTexts[i - 1].hasEOL = item.hasEOL;
pageTexts.splice(i, 1);
i--;
}
}
page.cleanup();
return pageTexts
.map((token) => {
const paragraphEnd = token.hasEOL && /([。?!.?!\n\r]|(\r\n))$/.test(token.str);
return paragraphEnd ? `${token.str}\n` : token.str;
})
.join('');
};
const loadingTask = pdfjs.getDocument(path);
const doc = await loadingTask.promise;
const pageTextPromises = [];
for (let pageNo = 1; pageNo <= doc.numPages; pageNo++) {
pageTextPromises.push(readPDFPage(doc, pageNo));
}
const pageTexts = await Promise.all(pageTextPromises);
loadingTask.destroy();
return {
rawText: pageTexts.join('')
};
};

View File

@@ -1,18 +0,0 @@
export type ReadFileParams = {
preview: boolean;
teamId: string;
path: string;
metadata?: Record<string, any>;
};
export type ReadFileResponse = {
rawText: string;
};
export type ReadFileBufferItemType = ReadFileParams & {
rawText: string;
};
declare global {
var readFileBuffers: ReadFileBufferItemType[];
}

View File

@@ -1,50 +0,0 @@
import { readPdfFile } from './pdf';
import { readDocFle } from './word';
import { ReadFileBufferItemType, ReadFileParams } from './type';
global.readFileBuffers = global.readFileBuffers || [];
const bufferMaxSize = 200;
export const pushFileReadBuffer = (params: ReadFileBufferItemType) => {
global.readFileBuffers.push(params);
if (global.readFileBuffers.length > bufferMaxSize) {
global.readFileBuffers.shift();
}
};
export const getReadFileBuffer = ({ path, teamId }: ReadFileParams) =>
global.readFileBuffers.find((item) => item.path === path && item.teamId === teamId);
export const readFileContent = async (params: ReadFileParams) => {
const { path } = params;
const buffer = getReadFileBuffer(params);
if (buffer) {
return buffer;
}
const extension = path?.split('.')?.pop()?.toLowerCase() || '';
const { rawText } = await (async () => {
switch (extension) {
case 'pdf':
return readPdfFile(params);
case 'docx':
return readDocFle(params);
default:
return Promise.reject('Only support .pdf, .docx');
}
})();
pushFileReadBuffer({
...params,
rawText
});
return {
...params,
rawText
};
};

View File

@@ -1,22 +0,0 @@
import mammoth from 'mammoth';
import { htmlToMarkdown } from '../../string/markdown';
import { ReadFileParams } from './type';
/**
* read docx to markdown
*/
export const readDocFle = async ({ path, metadata = {} }: ReadFileParams) => {
try {
const { value: html } = await mammoth.convertToHtml({
path
});
const md = await htmlToMarkdown(html);
return {
rawText: md
};
} catch (error) {
console.log('error doc read:', error);
return Promise.reject('Can not read doc file, please convert to PDF');
}
};

View File

@@ -3,7 +3,6 @@ import multer from 'multer';
import path from 'path';
import { BucketNameEnum, bucketNameMap } from '@fastgpt/global/common/file/constants';
import { getNanoid } from '@fastgpt/global/common/string/tools';
import { tmpFileDirPath } from './constants';
type FileType = {
fieldname: string;
@@ -15,8 +14,6 @@ type FileType = {
size: number;
};
const expiredTime = 30 * 60 * 1000;
export const getUploadModel = ({ maxSize = 500 }: { maxSize?: number }) => {
maxSize *= 1024 * 1024;
class UploadModel {
@@ -31,15 +28,16 @@ export const getUploadModel = ({ maxSize = 500 }: { maxSize?: number }) => {
// },
filename: async (req, file, cb) => {
const { ext } = path.parse(decodeURIComponent(file.originalname));
cb(null, `${Date.now() + expiredTime}-${getNanoid(32)}${ext}`);
cb(null, `${getNanoid(32)}${ext}`);
}
})
}).any();
}).single('file');
async doUpload<T = Record<string, any>>(req: NextApiRequest, res: NextApiResponse) {
return new Promise<{
files: FileType[];
metadata: T;
file: FileType;
metadata: Record<string, any>;
data: T;
bucketName?: `${BucketNameEnum}`;
}>((resolve, reject) => {
// @ts-ignore
@@ -54,20 +52,28 @@ export const getUploadModel = ({ maxSize = 500 }: { maxSize?: number }) => {
return reject('BucketName is invalid');
}
// @ts-ignore
const file = req.file as FileType;
resolve({
...req.body,
files:
// @ts-ignore
req.files?.map((file) => ({
...file,
originalname: decodeURIComponent(file.originalname)
})) || [],
file: {
...file,
originalname: decodeURIComponent(file.originalname)
},
bucketName,
metadata: (() => {
if (!req.body?.metadata) return {};
try {
return JSON.parse(req.body.metadata);
} catch (error) {
console.log(error);
return {};
}
})(),
data: (() => {
if (!req.body?.data) return {};
try {
return JSON.parse(req.body.data);
} catch (error) {
return {};
}
})()

View File

@@ -1,5 +1,4 @@
import fs from 'fs';
import { tmpFileDirPath } from './constants';
export const removeFilesByPaths = (paths: string[]) => {
paths.forEach((path) => {
@@ -10,24 +9,3 @@ export const removeFilesByPaths = (paths: string[]) => {
});
});
};
/* cron job. check expired tmp files */
export const checkExpiredTmpFiles = () => {
// get all file name
const files = fs.readdirSync(tmpFileDirPath).map((name) => {
const timestampStr = name.split('-')[0];
const expiredTimestamp = timestampStr ? Number(timestampStr) : 0;
return {
filename: name,
expiredTimestamp,
path: `${tmpFileDirPath}/${name}`
};
});
// count expiredFiles
const expiredFiles = files.filter((item) => item.expiredTimestamp < Date.now());
// remove expiredFiles
removeFilesByPaths(expiredFiles.map((item) => item.path));
};