@ecfjs/upload — Enterprise File Ingestion Pipeline Platform

@ecfjs/upload is a robust file ingestion pipeline platform for the ECF (Enterprise Core Framework) ecosystem.

Features

• 🕵️ Binary Magic Byte Sniffing: Header signature sniffing for JPEG, PNG, GIF, WEBP, PDF, ZIP, and MP4 without external binary dependencies.
• 📐 Image Dimension Parser: Header inspection for width and height extraction.
• 🛡️ Executable Neutralization: Prevents path traversal and neutralizes executable extensions (.exe, .php, .sh).
• ☣️ Virus Scanning & Quarantine: Pluggable IVirusScanner with QuarantineManager routing.
• 🧩 Chunked & Resumable Uploads: Tus-style chunk session management (initiate, appendChunk, assemble, cleanupAbandoned).
• ✍️ Signed Direct-to-Storage Uploads: Pre-signed upload URLs for direct S3 / R2 / MinIO client uploads.
• 🧪 Upload Testing Fake: Upload.fake() and UploadedFile.fake('avatar.jpg').

Quick Start

1. Basic Ingestion Pipeline

import { Application } from "@ecfjs/core";
import { StorageServiceProvider } from "@ecfjs/storage";
import { UploadServiceProvider, Upload, UploadedFile } from "@ecfjs/upload";

const app = new Application();
app.register(StorageServiceProvider);
app.register(UploadServiceProvider);
app.boot();

// Ingest uploaded file through 'avatar' profile
const rawFile = UploadedFile.fake("profile.jpg", { size: 2048, mime: "image/jpeg" });
const { file, manifest } = await Upload.process(rawFile, "avatar");

// Persist to storage disk
await file.store("avatars", "local");

License

MIT