srijanmukherjee / huffman-compression

A huffman coding based text compression utilility tool.

huffman-compression

A simple cli tool to compress/decompress text (for now) files using huffman coding algorithm.

Getting started

# Clone the repository (SSH)
git clone --recurse-submodules [email protected]:srijanmukherjee/huffman-compression.git

# Clone the repository (HTTPS)
git clone --recurse-submodules https://github.com/srijanmukherjee/huffman-compression.git

# initialize build directory
make init

# to build the project
make

# to compress a text file
./build/huffman filepath

# to decompress
./build/huffman -d filepath

Usage

Usage: huffman [--help] [--version] [--verbose] [--decompress] [--output VAR] file

Positional arguments:
  file              path to file 

Optional arguments:
  -h, --help        shows help message and exits 
  -v, --version     prints version information and exits 
  --verbose         enable verbose output 
  -d, --decompress  decompress file 
  -o, --output      output file path [default: ""]

Result

• examples/sample0.txt

1009K sample0.txt.bin
1.8M  examples/sample0.txt

Dependencies

• p-ranav/argparse for cli argument parsing

License

This project is available under the MIT license

huffman-compression's Issues

Compressed file size greater than original file size

If the compressed size (table + compressed content) > original file size, we will not put the compressed content in the file instead do something like set some sort of flag and put the original content.

The flag will denote that the content of the compressed file contains the original content, so removing the flag will give the uncompressed file.

Example:

• Original file (3 bytes)

abc

• Table size = 3 * 4 + 3 bytes
• Content = 4 bits

Validate compressed file while before decompressing

Before decompressing, check if the file is actually a compressed file, and contains the huffman table.

Possible solutions:

• embedding a hash into the compressed file in the first few bytes