farcasclaudiu/Flowise
1
0
Fork 0
mirror of https://github.com/farcasclaudiu/Flowise.git synced 2026-06-28 21:00:58 +03:00
Code Issues Packages Projects Releases Wiki Activity

add fixes to multiple documents

Browse Source
This commit is contained in:
Henry
2023-05-15 20:06:22 +01:00
parent 9f53111aac
commit 324bd3a430
6 changed files with 128 additions and 78 deletions
Download Patch File Download Diff File Expand all files Collapse all files
packages/components/nodes/documentloaders/Pdf/Pdf.ts
+36 -18
View File
@@ -1,7 +1,6 @@
import { INode, INodeData, INodeParams } from '../../../src/Interface'
import { TextSplitter } from 'langchain/text_splitter'
import { PDFLoader } from 'langchain/document_loaders/fs/pdf'
import { getBlob } from '../../../src/utils'
class Pdf_DocumentLoaders implements INode {
label: string
@@ -66,30 +65,49 @@ class Pdf_DocumentLoaders implements INode {
const usage = nodeData.inputs?.usage as string
const metadata = nodeData.inputs?.metadata
const blob = new Blob(getBlob(pdfFileBase64))
let docs = []
if (usage === 'perFile') {
// @ts-ignore
const loader = new PDFLoader(blob, { splitPages: false, pdfjs: () => import('pdf-parse/lib/pdf.js/v1.10.100/build/pdf.js') })
if (textSplitter) {
docs = await loader.loadAndSplit(textSplitter)
} else {
docs = await loader.load()
}
let alldocs = []
let files: string[] = []
if (pdfFileBase64.startsWith('[') && pdfFileBase64.endsWith(']')) {
files = JSON.parse(pdfFileBase64)
} else {
// @ts-ignore
const loader = new PDFLoader(blob, { pdfjs: () => import('pdf-parse/lib/pdf.js/v1.10.100/build/pdf.js') })
if (textSplitter) {
docs = await loader.loadAndSplit(textSplitter)
files = [pdfFileBase64]
}
for (const file of files) {
const splitDataURI = file.split(',')
splitDataURI.pop()
const bf = Buffer.from(splitDataURI.pop() || '', 'base64')
if (usage === 'perFile') {
const loader = new PDFLoader(new Blob([bf]), {
splitPages: false,
// @ts-ignore
pdfjs: () => import('pdf-parse/lib/pdf.js/v1.10.100/build/pdf.js')
})
if (textSplitter) {
const docs = await loader.loadAndSplit(textSplitter)
alldocs.push(...docs)
} else {
const docs = await loader.load()
alldocs.push(...docs)
}
} else {
docs = await loader.load()
// @ts-ignore
const loader = new PDFLoader(new Blob([bf]), { pdfjs: () => import('pdf-parse/lib/pdf.js/v1.10.100/build/pdf.js') })
if (textSplitter) {
const docs = await loader.loadAndSplit(textSplitter)
alldocs.push(...docs)
} else {
const docs = await loader.load()
alldocs.push(...docs)
}
}
}
if (metadata) {
const parsedMetadata = typeof metadata === 'object' ? metadata : JSON.parse(metadata)
let finaldocs = []
for (const doc of docs) {
for (const doc of alldocs) {
const newdoc = {
...doc,
metadata: {
@@ -102,7 +120,7 @@ class Pdf_DocumentLoaders implements INode {
return finaldocs
}
return docs
return alldocs
}
}