mirror of
https://github.com/Abdulazizzn/n8n-enterprise-unlocked.git
synced 2025-12-17 18:12:04 +00:00
feat(Read PDF Node): Replace pdf-parse with pdfjs, and add support for streaming and encrypted PDFs (#6640)
This commit is contained in:
कारतोफ्फेलस्क्रिप्ट™
committed by
GitHub
GitHub
parent
acda7f269f
commit
0a31b8e2b4
@@ -0,0 +1,87 @@
|
||||
{
|
||||
"nodes": [
|
||||
{
|
||||
"name": "When clicking \"Execute Workflow\"",
|
||||
"type": "n8n-nodes-base.manualTrigger",
|
||||
"typeVersion": 1,
|
||||
"parameters": {},
|
||||
"position": [660, 580]
|
||||
},
|
||||
{
|
||||
"name": "Read sample-encrypted.pdf",
|
||||
"type": "n8n-nodes-base.readBinaryFile",
|
||||
"typeVersion": 1,
|
||||
"parameters": {
|
||||
"filePath": "C:\\Test\\sample-encrypted.pdf"
|
||||
},
|
||||
"position": [880, 780]
|
||||
},
|
||||
{
|
||||
"name": "Read PDF (encrypted)",
|
||||
"type": "n8n-nodes-base.readPDF",
|
||||
"typeVersion": 1,
|
||||
"parameters": {
|
||||
"encrypted": true,
|
||||
"password": "ReaderPassword"
|
||||
},
|
||||
"position": [1100, 780]
|
||||
}
|
||||
],
|
||||
"pinData": {
|
||||
"Read PDF (encrypted)": [
|
||||
{
|
||||
"binary": {
|
||||
"data": {
|
||||
"fileExtension": "pdf",
|
||||
"fileName": "sample-encrypted.pdf",
|
||||
"fileSize": "18.9 kB",
|
||||
"mimeType": "application/pdf"
|
||||
}
|
||||
},
|
||||
"json": {
|
||||
"numpages": 1,
|
||||
"numrender": 1,
|
||||
"info": {
|
||||
"PDFFormatVersion": "1.7",
|
||||
"Language": null,
|
||||
"EncryptFilterName": "Standard",
|
||||
"IsLinearized": false,
|
||||
"IsAcroFormPresent": false,
|
||||
"IsXFAPresent": false,
|
||||
"IsCollectionPresent": false,
|
||||
"IsSignaturesPresent": false,
|
||||
"ModDate": "D:20230210122750Z",
|
||||
"Producer": "iLovePDF",
|
||||
"Title": "sample"
|
||||
},
|
||||
"text": "N8N\nSample PDF\nLorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor\ninvidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et\njusto duo dolores et ea rebum.",
|
||||
"version": "2.16.105"
|
||||
}
|
||||
}
|
||||
]
|
||||
},
|
||||
"connections": {
|
||||
"When clicking \"Execute Workflow\"": {
|
||||
"main": [
|
||||
[
|
||||
{
|
||||
"node": "Read sample-encrypted.pdf",
|
||||
"type": "main",
|
||||
"index": 0
|
||||
}
|
||||
]
|
||||
]
|
||||
},
|
||||
"Read sample-encrypted.pdf": {
|
||||
"main": [
|
||||
[
|
||||
{
|
||||
"node": "Read PDF (encrypted)",
|
||||
"type": "main",
|
||||
"index": 0
|
||||
}
|
||||
]
|
||||
]
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -1,60 +1,11 @@
|
||||
/* eslint-disable @typescript-eslint/no-loop-func */
|
||||
import * as Helpers from '@test/nodes/Helpers';
|
||||
import type { WorkflowTestData } from '@test/nodes/types';
|
||||
import { executeWorkflow } from '@test/nodes/ExecuteWorkflow';
|
||||
import path from 'path';
|
||||
import { getWorkflowFilenames, initBinaryDataManager, testWorkflows } from '@test/nodes/Helpers';
|
||||
|
||||
describe('Test Read PDF Node', () => {
|
||||
beforeEach(async () => {
|
||||
await Helpers.initBinaryDataManager();
|
||||
const workflows = getWorkflowFilenames(__dirname);
|
||||
|
||||
beforeAll(async () => {
|
||||
await initBinaryDataManager();
|
||||
});
|
||||
|
||||
const workflow = Helpers.readJsonFileSync('nodes/ReadPdf/test/ReadPDF.workflow.json');
|
||||
const node = workflow.nodes.find((n: any) => n.name === 'Read Binary File');
|
||||
node.parameters.filePath = path.join(__dirname, 'sample.pdf');
|
||||
|
||||
const testData: WorkflowTestData = {
|
||||
description: 'nodes/ReadPdf/test/ReadPDF.workflow.json',
|
||||
input: {
|
||||
workflowData: workflow,
|
||||
},
|
||||
output: {
|
||||
nodeData: {
|
||||
'Read PDF': [
|
||||
[
|
||||
{
|
||||
json: {
|
||||
numpages: 1,
|
||||
numrender: 1,
|
||||
info: {
|
||||
PDFFormatVersion: '1.4',
|
||||
IsAcroFormPresent: false,
|
||||
IsXFAPresent: false,
|
||||
Title: 'sample',
|
||||
Producer: 'iLovePDF',
|
||||
ModDate: 'D:20230210122750Z',
|
||||
},
|
||||
metadata: null,
|
||||
text: '\n\nN8N\nSample PDF\nLorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor\ninvidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et\njusto duo dolores et ea rebum.',
|
||||
version: '1.10.100',
|
||||
},
|
||||
},
|
||||
],
|
||||
],
|
||||
},
|
||||
},
|
||||
};
|
||||
|
||||
const nodeTypes = Helpers.setup(testData);
|
||||
|
||||
test(testData.description, async () => {
|
||||
const { result } = await executeWorkflow(testData, nodeTypes);
|
||||
const resultNodeData = Helpers.getResultNodeData(result, testData);
|
||||
|
||||
// delete binary data because we test against json only
|
||||
delete resultNodeData[0].resultData[0]![0].binary;
|
||||
expect(resultNodeData[0].resultData).toEqual(testData.output.nodeData['Read PDF']);
|
||||
|
||||
expect(result.finished).toEqual(true);
|
||||
});
|
||||
testWorkflows(workflows);
|
||||
});
|
||||
|
||||
@@ -1,47 +1,75 @@
|
||||
{
|
||||
"name": "Read PDF node unit test",
|
||||
"nodes": [
|
||||
{
|
||||
"parameters": {},
|
||||
"id": "0c9db33c-dd15-4088-9d12-b9f3b8f1fa96",
|
||||
"name": "When clicking \"Execute Workflow\"",
|
||||
"type": "n8n-nodes-base.manualTrigger",
|
||||
"typeVersion": 1,
|
||||
"position": [960, 540]
|
||||
},
|
||||
{
|
||||
"parameters": {},
|
||||
"id": "86abdc3b-206d-4b67-a37f-6b67b6bd3bbc",
|
||||
"name": "Read PDF",
|
||||
"type": "n8n-nodes-base.readPDF",
|
||||
"typeVersion": 1,
|
||||
"position": [1400, 540]
|
||||
"position": [660, 580]
|
||||
},
|
||||
{
|
||||
"name": "Read sample.pdf",
|
||||
"type": "n8n-nodes-base.readBinaryFile",
|
||||
"typeVersion": 1,
|
||||
"parameters": {
|
||||
"filePath": "C:\\Test\\sample.pdf"
|
||||
},
|
||||
"id": "2f6d241e-44a4-4213-b49a-166201946a89",
|
||||
"name": "Read Binary File",
|
||||
"type": "n8n-nodes-base.readBinaryFile",
|
||||
"position": [880, 580]
|
||||
},
|
||||
{
|
||||
"name": "Read PDF",
|
||||
"type": "n8n-nodes-base.readPDF",
|
||||
"typeVersion": 1,
|
||||
"position": [1180, 540]
|
||||
"parameters": {},
|
||||
"position": [1100, 580]
|
||||
}
|
||||
],
|
||||
"pinData": {},
|
||||
"pinData": {
|
||||
"Read PDF": [
|
||||
{
|
||||
"binary": {
|
||||
"data": {
|
||||
"fileExtension": "pdf",
|
||||
"fileName": "sample.pdf",
|
||||
"fileSize": "17.8 kB",
|
||||
"mimeType": "application/pdf"
|
||||
}
|
||||
},
|
||||
"json": {
|
||||
"numpages": 1,
|
||||
"numrender": 1,
|
||||
"info": {
|
||||
"PDFFormatVersion": "1.4",
|
||||
"Language": null,
|
||||
"EncryptFilterName": null,
|
||||
"IsLinearized": false,
|
||||
"IsAcroFormPresent": false,
|
||||
"IsXFAPresent": false,
|
||||
"IsCollectionPresent": false,
|
||||
"IsSignaturesPresent": false,
|
||||
"Title": "sample",
|
||||
"Producer": "iLovePDF",
|
||||
"ModDate": "D:20230210122750Z"
|
||||
},
|
||||
"text": "N8N\nSample PDF\nLorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor\ninvidunt ut labore et dolore magna aliquyam erat, sed diam voluptua. At vero eos et accusam et\njusto duo dolores et ea rebum.",
|
||||
"version": "2.16.105"
|
||||
}
|
||||
}
|
||||
]
|
||||
},
|
||||
"connections": {
|
||||
"When clicking \"Execute Workflow\"": {
|
||||
"main": [
|
||||
[
|
||||
{
|
||||
"node": "Read Binary File",
|
||||
"node": "Read sample.pdf",
|
||||
"type": "main",
|
||||
"index": 0
|
||||
}
|
||||
]
|
||||
]
|
||||
},
|
||||
"Read Binary File": {
|
||||
"Read sample.pdf": {
|
||||
"main": [
|
||||
[
|
||||
{
|
||||
@@ -52,13 +80,5 @@
|
||||
]
|
||||
]
|
||||
}
|
||||
},
|
||||
"active": false,
|
||||
"settings": {},
|
||||
"versionId": "9802b48d-727a-40ef-ad87-d544a9a648a7",
|
||||
"id": "188",
|
||||
"meta": {
|
||||
"instanceId": "104a4d08d8897b8bdeb38aaca515021075e0bd8544c983c2bb8c86e6a8e6081c"
|
||||
},
|
||||
"tags": []
|
||||
}
|
||||
}
|
||||
|
||||
BIN
packages/nodes-base/nodes/ReadPdf/test/sample-encrypted.pdf
Normal file
BIN
packages/nodes-base/nodes/ReadPdf/test/sample-encrypted.pdf
Normal file
Binary file not shown.
Reference in New Issue
Block a user