🔨 Forge

A fast, native macOS app for batch file conversion — no dependencies, no bloat.

Drag in hundreds of images, videos, audio files, or PDFs. Pick a preset. Convert them all — in parallel, with RAM kept low by streaming. 100% Apple frameworks (Core Image, AVFoundation), zero external tools.

Why Forge

• Native & dependency-free — Core Image + AVFoundation do the work. No ImageMagick, FFmpeg, or LibreOffice to install. It just runs.
• Batch-first — drop hundreds of files, convert them in bounded parallel waves that respect your concurrency setting.
• Memory-conscious — streaming conversion keeps RAM flat even on large video.
• Watched folders — assign a preset to a folder; anything dropped in is converted automatically (FSEvents).
• Presets & history — reusable transformation pipelines, full processing history, JSON-backed (no Core Data).

Supported formats

Kind	Read	Write	Operations
Image	JPEG, PNG, TIFF, HEIC, WebP, BMP, GIF, TGA, ICO	JPEG, PNG, TIFF, HEIC…	convert · resize (4 modes) · quality · filters
Video	MP4, MOV, M4V, AVI, MKV, WebM	MP4, MOV, M4V (H.264)	convert · resolution · bitrate/quality
Audio	MP3, WAV, M4A, AAC, FLAC, ALAC, AIFF	MP3, M4A, WAV, AAC, FLAC	convert · bitrate · sample rate · channels
Documents	PDF, TXT, CSV, HTML, RTF	images (JPEG/PNG/TIFF), text	PDF → images · basic text conversions

Install

Download (recommended)

Grab the latest .dmg from the Releases page, open it, and drag Forge to Applications.

Unsigned build: on first launch, right-click the app → Open to bypass Gatekeeper.

Build from source

git clone https://github.com/thousandflowers/Forge.git
cd Forge
swift build -c release
.build/release/Forge

Requires macOS 13+ (Ventura) and Swift 5.9+.

Usage

Batch convert

1. Launch Forge, drag & drop files into the window.
2. Pick a preset (e.g. Instagram Post), choose destination (Overwrite / Copy to / Move to).
3. Click Start Processing.

Watched folder

1. Open the Folders tab, add a folder, assign it a preset.
2. Anything dropped in that folder is converted automatically.

Custom preset

1. Presets tab → +.
2. Set target format, resize, quality (1–100), optional filter. Save — it's now usable in batch and watched-folder modes.

Architecture

SwiftUI  ──▶  ProcessingCoordinator  ──▶  ProcessorRegistry  ──▶  Image / Video / Audio / Doc processors
                     │                                                    (Core Image · AVFoundation · PDFKit)
              bounded TaskGroup                             MonitoredFolderWatcher (FSEvents)
              (maxConcurrentNative)                         PersistenceManager (JSON in App Support)

• ProcessorRegistry picks the right processor for each file type.
• ProcessingCoordinator runs conversions, tracks active tasks for cancellation.
• RulePreset — serializable transformation pipelines.
• PersistenceManager — JSON storage for presets, history, monitored folders.

Development

swift test                 # run the unit suite
bash Scripts/generate_samples.sh   # generate sample test files

Roadmap

• Native image / video / audio / PDF conversion
• Bounded-concurrency batch processing with live progress
• Presets, processing history, JSON persistence
• Watched-folder UI polish + completion notifications
• Preset import/export
• CLI (forge) + Shortcuts integration

Contributing

PRs welcome. Fork → feature branch → tests green (swift test) → PR. Follow the Swift API Design Guidelines, keep memory bounded (stream, don't fully load), use async/await + actors.

License

MIT — see LICENSE.

Acknowledgments

Built with Core Image, AVFoundation, and SwiftUI. Inspired by the original Shift — this is a native, from-scratch rebuild.