mirror of
https://github.com/Frooodle/Stirling-PDF.git
synced 2024-12-21 19:08:24 +01:00
233 lines
7.8 KiB
JavaScript
233 lines
7.8 KiB
JavaScript
import express from 'express';
|
|
import crypto from 'crypto';
|
|
import stream from "stream";
|
|
import Archiver from 'archiver';
|
|
|
|
import * as Functions from "../../functions.js";
|
|
import { traverseOperations } from "../../../shared-operations/traverseOperations.js";
|
|
|
|
const activeWorkflows = {};
|
|
|
|
const router = express.Router();
|
|
|
|
router.post("/:workflowUuid?", [
|
|
async (req, res) => {
|
|
if(req.files == null) {
|
|
res.status(400).json({"error": "No files were uploaded."});
|
|
return;
|
|
}
|
|
|
|
if(Array.isArray(req.files.files)) {
|
|
req.files = req.files.files;
|
|
}
|
|
else {
|
|
req.files = [req.files.files];
|
|
}
|
|
|
|
const workflow = JSON.parse(req.body.workflow);
|
|
// TODO: Validate input further (json may fail or not be a valid workflow)
|
|
|
|
const inputs = await Promise.all(req.files.map(async file => {
|
|
return {
|
|
originalFileName: file.name.replace(/\.[^/.]+$/, ""),
|
|
fileName: file.name.replace(/\.[^/.]+$/, ""),
|
|
buffer: new Uint8Array(await file.data)
|
|
}
|
|
}));
|
|
|
|
// Allow option to do it synchronously and just make a long request
|
|
if(req.body.async === "false") {
|
|
console.log("Don't do async");
|
|
|
|
const traverse = traverseOperations(workflow.operations, inputs, Functions);
|
|
|
|
let pdfResults;
|
|
let iteration;
|
|
while (true) {
|
|
iteration = await traverse.next();
|
|
if (iteration.done) {
|
|
pdfResults = iteration.value;
|
|
console.log("Done");
|
|
break;
|
|
}
|
|
console.log(iteration.value);
|
|
}
|
|
|
|
console.log("Download");
|
|
downloadHandler(res, pdfResults);
|
|
}
|
|
else {
|
|
console.log("Start Aync Workflow");
|
|
// TODO: UUID collision checks
|
|
let workflowID = req.params.workflowUuid
|
|
if(!workflowID)
|
|
workflowID = generateWorkflowID();
|
|
|
|
activeWorkflows[workflowID] = {
|
|
createdAt: Date.now(),
|
|
finished: false,
|
|
eventStream: null,
|
|
result: null,
|
|
// TODO: When auth is implemented: owner
|
|
}
|
|
const activeWorkflow = activeWorkflows[workflowID];
|
|
|
|
res.status(200).json({
|
|
"workflowID": workflowID,
|
|
"data-recieved": {
|
|
"fileCount": req.files.length,
|
|
"workflow": workflow
|
|
}
|
|
});
|
|
|
|
const traverse = traverseOperations(workflow.operations, inputs, Functions);
|
|
|
|
let pdfResults;
|
|
let iteration;
|
|
while (true) {
|
|
iteration = await traverse.next();
|
|
if (iteration.done) {
|
|
pdfResults = iteration.value;
|
|
if(activeWorkflow.eventStream) {
|
|
activeWorkflow.eventStream.write(`data: processing done\n\n`);
|
|
activeWorkflow.eventStream.end();
|
|
}
|
|
break;
|
|
}
|
|
if(activeWorkflow.eventStream)
|
|
activeWorkflow.eventStream.write(`data: ${iteration.value}\n\n`);
|
|
}
|
|
|
|
activeWorkflow.result = pdfResults;
|
|
activeWorkflow.finished = true;
|
|
}
|
|
}
|
|
]);
|
|
|
|
router.get("/progress/:workflowUuid", (req, res, nex) => {
|
|
if(!req.params.workflowUuid) {
|
|
res.status(400).json({"error": "No workflowUuid weres provided."});
|
|
return;
|
|
}
|
|
if(!activeWorkflows.hasOwnProperty(req.params.workflowUuid)) {
|
|
res.status(400).json({"error": `No workflow with workflowUuid "${req.params.workflowUuid}" was found.`});
|
|
return;
|
|
}
|
|
|
|
// Return current progress
|
|
const workflow = activeWorkflows[req.params.workflowUuid];
|
|
res.status(200).json({ createdAt: workflow.createdAt, finished: workflow.finished });
|
|
});
|
|
|
|
router.get("/progress-stream/:workflowUuid", (req, res, nex) => {
|
|
if(!req.params.workflowUuid) {
|
|
res.status(400).json({"error": "No workflowUuid weres provided."});
|
|
return;
|
|
}
|
|
if(!activeWorkflows.hasOwnProperty(req.params.workflowUuid)) {
|
|
res.status(400).json({"error": `No workflow with workflowUuid "${req.params.workflowUuid}" was found.`});
|
|
return;
|
|
}
|
|
|
|
// TODO: Check if already done
|
|
|
|
// Send realtime updates
|
|
res.setHeader('Cache-Control', 'no-cache');
|
|
res.setHeader('Content-Type', 'text/event-stream');
|
|
res.setHeader('Access-Control-Allow-Origin', '*');
|
|
res.setHeader('Connection', 'keep-alive');
|
|
res.flushHeaders(); // flush the headers to establish SSE with client
|
|
|
|
const workflow = activeWorkflows[req.params.workflowUuid];
|
|
workflow.eventStream = res;
|
|
|
|
res.on('close', () => {
|
|
res.end();
|
|
// TODO: Abort if not already done?
|
|
});
|
|
});
|
|
|
|
router.get("/result/:workflowUuid", (req, res, nex) => {
|
|
if(!req.params.workflowUuid) {
|
|
res.status(400).json({"error": "No workflowUuid weres provided."});
|
|
return;
|
|
}
|
|
if(!activeWorkflows.hasOwnProperty(req.params.workflowUuid)) {
|
|
res.status(400).json({"error": `No workflow with workflowUuid "${req.params.workflowUuid}" was found.`});
|
|
return;
|
|
}
|
|
|
|
/*
|
|
* If workflow isn't done return error
|
|
* Send file, TODO: if there are multiple outputs return as zip
|
|
* If download is done, delete results / allow deletion within the next 5-60 mins
|
|
*/
|
|
const workflow = activeWorkflows[req.params.workflowUuid];
|
|
if(!workflow.finished) {
|
|
res.status(202).json({ message: "Workflow hasn't finished yet. Check progress or connect to progress-steam to get notified when its done." });
|
|
return
|
|
}
|
|
|
|
downloadHandler(res, workflow.result);
|
|
// Delete workflow / results when done.
|
|
delete activeWorkflows[req.params.workflowUuid];
|
|
});
|
|
|
|
router.post("/abort/:workflowUuid", (req, res, nex) => {
|
|
if(!req.params.workflowUuid) {
|
|
res.status(400).json({"error": "No workflowUuid weres provided."});
|
|
return;
|
|
}
|
|
if(!activeWorkflows.hasOwnProperty(req.params.workflowUuid)) {
|
|
res.status(400).json({"error": `No workflow with workflowUuid "${req.params.workflowUuid}" was found.`});
|
|
return;
|
|
}
|
|
|
|
// TODO: Abort workflow
|
|
res.status(501).json({"warning": "Abortion has not been implemented yet."});
|
|
});
|
|
|
|
function generateWorkflowID() {
|
|
return crypto.randomUUID();
|
|
}
|
|
|
|
function downloadHandler(res, pdfResults) {
|
|
if(pdfResults.length == 0) {
|
|
res.status(500).json({"warning": "The workflow had no outputs."});
|
|
}
|
|
else if(pdfResults.length > 1) {
|
|
// TODO: Also allow the user to download multiple files without zip compressen, because this is kind of slow...
|
|
res.writeHead(200, {
|
|
'Content-Type': 'application/zip',
|
|
'Content-disposition': 'attachment; filename=workflow-results.zip'
|
|
});
|
|
|
|
var zip = Archiver('zip');
|
|
|
|
// Stream the file to the user.
|
|
zip.pipe(res);
|
|
|
|
console.log("Adding Files to ZIP...");
|
|
|
|
for (let i = 0; i < pdfResults.length; i++) {
|
|
// TODO: Implement other file types (mostly fro image & text extraction)
|
|
// TODO: Check for name collisions
|
|
zip.append(Buffer.from(pdfResults[i].buffer), { name: pdfResults[i].fileName + ".pdf" });
|
|
}
|
|
|
|
zip.finalize();
|
|
console.log("Sent");
|
|
}
|
|
else {
|
|
const readStream = new stream.PassThrough();
|
|
readStream.end(pdfResults[0].buffer);
|
|
|
|
// TODO: Implement other file types (mostly fro image & text extraction)
|
|
res.set("Content-disposition", 'attachment; filename=' + pdfResults[0].fileName + ".pdf");
|
|
res.set("Content-Type", "application/pdf");
|
|
|
|
readStream.pipe(res);
|
|
}
|
|
}
|
|
|
|
export default router; |