home .. forth .. colorforth mail list archive ..

Re: [colorforth] Unused character encodings available in Colorforth

On Sunday 22 February 2004 11:46 am, Mark Slicker wrote:
> What Chuck is doing might be described as hand coded huffman. He does
> assign the more frequently occuring characters to shorter codes,
> however it apears to me the codes are asigned in such a way to make
> characters simpler to decode. I'm speculating here, because I have not

Yes, it's optimized for table-lookups based on a set series of bits, 
versus performing a bit-level traversal of a decoding tree.

I thought it was Shannon encoding because I've seen Shannon codes which 
follow a similar 'ease of decoding' sequence, where the first 'n' bits 
are used to decode one of several different tables.  True, each table 
did have a set probability.

--
Samuel A. Falvo II


---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com