OSDev.org

hi all, Im trying to write a simple jpeg decoder, Im stuck at the stage of huffman table format in jpegs, some encoders use standard huffman codewords some create their own huffman tables (and codewords), my problem is that codewords are not directly given in huffman table of jpeg so my question is: do we have to create huffman tree -from huffman table given in jpeg file- in order to get codewords or is there a simpler method to get codewords from a jpeg file's huffman table? I dont want to get into tree creation, im looking for a more easier way. Thanx.

hi, I have done it in the hard way and generated the codewords from jpeg's huffman table, but im open to other ideas since this thing is complex and might have bugs, I could not test it with enough encoders, but it creates 150 codewords for default ac list correctly , I hope it works ok...