imgcmprs 0.1.1

Fast, safe CLI to compress JPEG & PNG, lossless/lossy, batch or single images.

img

A fast, easy-to-use Python CLI tool for compressing JPEG and PNG images with both lossless and lossy modes.

---

Features

• Compresses images individually or in bulk (folders)
• Supports JPEG and PNG
• Lossless mode (-l): Optimizes files without visible quality loss
• Lossy mode (-q): Reduce file size by lowering image quality
• Recursive directory support
• Custom output folder
• Prompts to delete originals after compression
• Skips replacement if compressed file is larger
• Cross-platform, requires Python 3.7+

---

Installation

1. Install Python 3.7+
2. Clone this repo and install:

   pip install -e .
   # Or for user-local install:
   pip install --user .
   

---

Usage

Compress a single image (lossy, 60% quality):

img -i myphoto.jpg -q 60

Compress a folder (lossless, best for PNG):

img -i images/ -o optimized/ -l -r

Options (Single-letter flags)

Flag	Meaning	Example
-i	Input file or folder (required)	-i mypic.jpg
-o	Output file or folder (optional)	-o compressed/
-q	JPEG quality (default 60; ignored in lossless mode)	-q 80
-l	Use lossless compression	-l
-r	Recursively process folders	-r

• All flags are single-letter for speed and ease: -i, -o, -q, -l, -r.
• After compressing, you will be prompted to delete the original files.
• If the compressed file is larger, the original is kept and a warning is printed.

---

Requirements

• Python 3.7+
• Pillow (pip install Pillow)

---

Notes

• Lossless for PNG is truly lossless; for JPEG, uses quality=100 with optimizations (minor effect but no further visual loss).
• Output defaults to _compressed directory if not specified for folders.
• Re-run with -l to optimize previously compressed images further (if possible).

---

Publishing to PyPI

1. Ensure you have an account on PyPI.
2. Install required tools:

   pip install build twine
   

3. Build your package:

   python -m build
   

   This creates a dist/ folder with your distributable files (tar.gz and .whl).
4. Upload to PyPI:

   python -m twine upload dist/*
   

5. Enter your PyPI credentials when prompted.

---

License

MIT