Talk:LZW compression: Difference between revisions
m
→Details of LZW: -- belatedly sign section
(Nitpicky details) |
m (→Details of LZW: -- belatedly sign section) |
||
(2 intermediate revisions by 2 users not shown) | |||
Line 3:
The task currently does not actually specify the LZW algorithm to be implemented. Instead, it defers to whatever is currently present in the Wikipedia presentation of the algorithm. This has left some compatibility details unspecified. (As near as I can tell, none of the implementations here include support for these issues.) But the task page is also back-linked from wikipedia, and someone should take responsibility for mentioning these things:
(1) One consequence is that the implementations do not include the LZW bitpacking algorithm. Instead, they represent the numbers which would be fed to that algorithm. The bitpacking algorithm shifts indices into position to take advantage of all guaranteed zeros in the representation of the previous index. (
(2) Another consequence is that the implementations tend to not include the LZW clear code
(3) Another issue is that the LZW file format begins with a three byte sequence (0x1F 0x9d
(At the time I am writing this, this is an
--[[User:Rdm|Rdm]] ([[User talk:Rdm|talk]]) 21:39, 10 October 2020 (UTC)
==Task and example==
Line 146 ⟶ 148:
</lang>
--[[User:Mick P.|Mick P.]] ([[User talk:Mick P.|talk]]) 11:08, 27 October 2015 (UTC)
: It took me a small while to find where the '''GOTO CC''' er, ... goes. Funny place to nearly hide a label. -- [[User:Gerard Schildberger|Gerard Schildberger]] ([[User talk:Gerard Schildberger|talk]]) 21:19, 9 October 2020 (UTC)
|