fengzhyuan / lzw

Introduction

A C++ STL based implementation of LZW (Lempel–Ziv–Welch) compression algorithm; inspired by the following repos:

• https://github.com/pieroxy/lz-string (author website: http://pieroxy.net/blog/pages/lz-string/index.html)
• https://github.com/jawa-the-hutt/lz-string-csharp

APIs

Implemented the following encoder/decoders:

• basic implementation Compress and Decompress
• UTF16 CompressToUTF16 and DecompressFromUTF16
• Base64 CompressToBase64 and DecompressFromBase64

Not yet implemented:

• encode to/decode from EncodeURIComponent (mainly for JavaScript)

Util functions:

• ToWStr: convert std::string to std::wstring
• ToStr: truncted version of string convertion; convert std::wstring to std::string based on MSDN: If wcstombs_s encounters a wide character it cannot convert to a multibyte character, it puts 0 in *pReturnValue, sets the destination buffer to an empty string, sets errno to EILSEQ, and returns EILSEQ. (wcstombs and wsctombs_s always return error code 42, any one knows how to do it correctly?)

Each method provides 2 APIs for different parameters std::string and std::wstring, respectively.

Test case (non-standard / messy version):

The file data.json is used for test purpose; below is the output of local machine (seconds):

[Test Default algorithm]
string length:1161516 compress ratio:23
time elapsed (s): 0.486
checking correctness: [string compressed == decompressed]:1

[Test UTF16]
string length:1161516 compress ratio:21
time elapsed (s): 0.457
checking correctness: [string compressed == decompressed]:1

[Test Base64]
string length:1161516 compress ratio:8
time elapsed (s): 0.468
checking correctness: [string compressed == decompressed]:1