Home

Welcome to the Compression wiki!

• Using std::map<Char, Freq> as frequency table.
• Note : move to frequency map , huffman tree not tested yet.
• class Encoder : read a file and build a vector to hold content, for now just limited characters.
• class BitString : a sequential container to store data in bits, push back only.
• protocol :

FrequencyTable|CompressedPart|Remainder

-->

FrequencyTable|CompressedPart|Remainder|RemainderSize

So the last bits that less than 8 should be append to Remainder part without compressing.

• 2 4 (8) 12 16

---

• Design for Version II

1. Read file char by char and save all data as an large BitString.
2. Split the BitString object into an unsigned[], based on the size specified, like 1, 2, 3..16.
3. Construct a FrqMap<unsigned int, FreqType> using the unsigned[].
4. Construct a HuffmanTree using the FrqMap.
5. Construct a CodeMap using HuffmanTree.
6. Traverse the unsigned[] with looking up CodeMap and save the resulted into an output BitString.
7. Split the output BitString into char[].
8. Output the char[] as the compressed file.