Feature/Multer to s3 (#2408)

* add ability to store files from multer to s3

* add check to bypass doc loader

packages/server/src/services/documentstore/index.ts

@@ -2,7 +2,7 @@ import { getRunningExpressApp } from '../../utils/getRunningExpressApp'
 import { DocumentStore } from '../../database/entities/DocumentStore'
 // @ts-ignore
 import {
-    addFileToStorage,
+    addSingleFileToStorage,
     getFileFromStorage,
     ICommonObject,
     IDocument,
@@ -343,7 +343,7 @@ const _saveFileToStorage = async (fileBase64: string, entity: DocumentStore) =>
     if (mimePrefix) {
         mime = mimePrefix.split(';')[0].split(':')[1]
     }
-    await addFileToStorage(mime, bf, filename, DOCUMENT_STORE_BASE_FOLDER, entity.id)
+    await addSingleFileToStorage(mime, bf, filename, DOCUMENT_STORE_BASE_FOLDER, entity.id)
     return {
         id: uuidv4(),
         name: filename,

packages/server/src/utils/buildChatflow.ts

@@ -1,5 +1,5 @@
 import { Request } from 'express'
-import { IFileUpload, convertSpeechToText, ICommonObject, addFileToStorage } from 'flowise-components'
+import { IFileUpload, convertSpeechToText, ICommonObject, addSingleFileToStorage, addArrayFilesToStorage } from 'flowise-components'
 import { StatusCodes } from 'http-status-codes'
 import { IncomingInput, IMessage, INodeData, IReactFlowObject, IReactFlowNode, IDepthQueue, chatType, IChatMessage } from '../Interface'
 import { InternalFlowiseError } from '../errors/internalFlowiseError'
@@ -71,7 +71,7 @@ export const utilBuildChatflow = async (req: Request, socketIO?: Server, isInter
                     const splitDataURI = upload.data.split(',')
                     const bf = Buffer.from(splitDataURI.pop() || '', 'base64')
                     const mime = splitDataURI[0].split(':')[1].split(';')[0]
-                    await addFileToStorage(mime, bf, filename, chatflowid, chatId)
+                    await addSingleFileToStorage(mime, bf, filename, chatflowid, chatId)
                     upload.type = 'stored-file'
                     // Omit upload.data since we don't store the content in database
                     fileUploads[i] = omit(upload, ['data'])
@@ -111,20 +111,21 @@ export const utilBuildChatflow = async (req: Request, socketIO?: Server, isInter
 
         let isStreamValid = false
 
-        const files = (req.files as any[]) || []
+        const files = (req.files as Express.Multer.File[]) || []
 
         if (files.length) {
             const overrideConfig: ICommonObject = { ...req.body }
+            const fileNames: string[] = []
             for (const file of files) {
-                const fileData = fs.readFileSync(file.path, { encoding: 'base64' })
-                const dataBase64String = `data:${file.mimetype};base64,${fileData},filename:${file.filename}`
+                const fileBuffer = fs.readFileSync(file.path)
+
+                const storagePath = await addArrayFilesToStorage(file.mimetype, fileBuffer, file.originalname, fileNames, chatflowid)
 
                 const fileInputField = mapMimeTypeToInputField(file.mimetype)
-                if (overrideConfig[fileInputField]) {
-                    overrideConfig[fileInputField] = JSON.stringify([...JSON.parse(overrideConfig[fileInputField]), dataBase64String])
-                } else {
-                    overrideConfig[fileInputField] = JSON.stringify([dataBase64String])
-                }
+
+                overrideConfig[fileInputField] = storagePath
+
+                fs.unlinkSync(file.path)
             }
             incomingInput = {
                 question: req.body.question ?? 'hello',

packages/server/src/utils/index.ts

@@ -374,6 +374,44 @@ export const saveUpsertFlowData = (nodeData: INodeData, upsertHistory: Record<st
     return existingUpsertFlowData
 }
 
+/**
+ * Check if doc loader should be bypassed, ONLY if doc loader is connected to a vector store
+ * Reason being we dont want to load the doc loader again whenever we are building the flow, because it was already done during upserting
+ * TODO: Remove this logic when we remove doc loader nodes from the canvas
+ * @param {IReactFlowNode} reactFlowNode
+ * @param {IReactFlowNode[]} reactFlowNodes
+ * @param {IReactFlowEdge[]} reactFlowEdges
+ * @returns {boolean}
+ */
+const checkIfDocLoaderShouldBeIgnored = (
+    reactFlowNode: IReactFlowNode,
+    reactFlowNodes: IReactFlowNode[],
+    reactFlowEdges: IReactFlowEdge[]
+): boolean => {
+    let outputId = ''
+
+    if (reactFlowNode.data.outputAnchors.length) {
+        if (Object.keys(reactFlowNode.data.outputs || {}).length) {
+            const output = reactFlowNode.data.outputs?.output
+            const node = reactFlowNode.data.outputAnchors[0].options?.find((anchor) => anchor.name === output)
+            if (node) outputId = (node as ICommonObject).id
+        } else {
+            outputId = (reactFlowNode.data.outputAnchors[0] as ICommonObject).id
+        }
+    }
+
+    const targetNodeId = reactFlowEdges.find((edge) => edge.sourceHandle === outputId)?.target
+
+    if (targetNodeId) {
+        const targetNodeCategory = reactFlowNodes.find((nd) => nd.id === targetNodeId)?.data.category || ''
+        if (targetNodeCategory === 'Vector Stores') {
+            return true
+        }
+    }
+
+    return false
+}
+
 /**
  * Build langchain from start to end
  * @param {string[]} startingNodeIds
@@ -446,7 +484,6 @@ export const buildFlow = async (
 
             const reactFlowNodeData: INodeData = resolveVariables(flowNodeData, flowNodes, question, chatHistory)
 
-            // TODO: Avoid processing Text Splitter + Doc Loader once Upsert & Load Existing Vector Nodes are deprecated
             if (isUpsert && stopNodeId && nodeId === stopNodeId) {
                 logger.debug(`[server]: Upserting ${reactFlowNode.data.label} (${reactFlowNode.data.id})`)
                 const indexResult = await newNodeInstance.vectorStoreMethods!['upsert']!.call(newNodeInstance, reactFlowNodeData, {
@@ -464,6 +501,12 @@ export const buildFlow = async (
                 if (indexResult) upsertHistory['result'] = indexResult
                 logger.debug(`[server]: Finished upserting ${reactFlowNode.data.label} (${reactFlowNode.data.id})`)
                 break
+            } else if (
+                !isUpsert &&
+                reactFlowNode.data.category === 'Document Loaders' &&
+                checkIfDocLoaderShouldBeIgnored(reactFlowNode, reactFlowNodes, reactFlowEdges)
+            ) {
+                initializedNodes.add(nodeId)
             } else {
                 logger.debug(`[server]: Initializing ${reactFlowNode.data.label} (${reactFlowNode.data.id})`)
                 let outputResult = await newNodeInstance.init(reactFlowNodeData, question, {
@@ -935,7 +978,7 @@ export const mapMimeTypeToInputField = (mimeType: string) => {
         case 'text/yaml':
             return 'yamlFile'
         default:
-            return ''
+            return 'txtFile'
     }
 }
 

packages/server/src/utils/upsertVector.ts

@@ -1,7 +1,7 @@
 import { Request } from 'express'
 import * as fs from 'fs'
 import { cloneDeep, omit } from 'lodash'
-import { ICommonObject, IMessage } from 'flowise-components'
+import { ICommonObject, IMessage, addArrayFilesToStorage } from 'flowise-components'
 import telemetryService from '../services/telemetry'
 import logger from '../utils/logger'
 import {
@@ -48,20 +48,21 @@ export const upsertVector = async (req: Request, isInternal: boolean = false) =>
             }
         }
 
-        const files = (req.files as any[]) || []
+        const files = (req.files as Express.Multer.File[]) || []
 
         if (files.length) {
             const overrideConfig: ICommonObject = { ...req.body }
+            const fileNames: string[] = []
             for (const file of files) {
-                const fileData = fs.readFileSync(file.path, { encoding: 'base64' })
-                const dataBase64String = `data:${file.mimetype};base64,${fileData},filename:${file.filename}`
+                const fileBuffer = fs.readFileSync(file.path)
+
+                const storagePath = await addArrayFilesToStorage(file.mimetype, fileBuffer, file.originalname, fileNames, chatflowid)
 
                 const fileInputField = mapMimeTypeToInputField(file.mimetype)
-                if (overrideConfig[fileInputField]) {
-                    overrideConfig[fileInputField] = JSON.stringify([...JSON.parse(overrideConfig[fileInputField]), dataBase64String])
-                } else {
-                    overrideConfig[fileInputField] = JSON.stringify([dataBase64String])
-                }
+
+                overrideConfig[fileInputField] = storagePath
+
+                fs.unlinkSync(file.path)
             }
             incomingInput = {
                 question: req.body.question ?? 'hello',