Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
LZW COMPRESSION
8 views
Skip to first unread message
robin
Mar 3, 2010, 7:07:57 AM3/3/10
to
Hi,
I thought I'd have a look at this topic, and thus read what Wikipedia
had on the subject.
I implemented a basic version of the compress and decompress algorithms,
namely, to deal with compression at the character level, as described on Wikipedia.
The explanation about repeating codes "aSaSa" was not borne out in practice.
The explanation about using more bits (when changing from
an index of 31 to an index of 32) seemed to be inadequate,
as the decompress algorithm has no idea when this change could occur.
I thought I'd have a look at this topic, and thus read what Wikipedia
had on the subject.
I implemented a basic version of the compress and decompress algorithms,
namely, to deal with compression at the character level, as described on Wikipedia.
The explanation about repeating codes "aSaSa" was not borne out in practice.
The explanation about using more bits (when changing from
an index of 31 to an index of 32) seemed to be inadequate,
as the decompress algorithm has no idea when this change could occur.
Just wondered whether anyone else had explored this topic
on Wikipedia, and if so, what he/she thought of it.
Peter Flass
Mar 3, 2010, 4:32:05 PM3/3/10
to
Somewhere in my library I have a book on compression algorithms (that
I've never read). I'll see if it talks about LZW, and possibly compare
to wikipedia.
robin
Mar 3, 2010, 6:48:08 PM3/3/10
to
robin wrote in message ...
>Hi,
> I thought I'd have a look at this topic, and thus read what Wikipedia
>had on the subject.
> I implemented a basic version of the compress and decompress algorithms,
>namely, to deal with compression at the character level, as described on Wikipedia.
> The explanation about repeating codes "aSaSa" was not borne out in practice.
> The explanation about using more bits (when changing from
>an index of 31 to an index of 32) seemed to be inadequate,
>as the decompress algorithm has no idea when this change could occur.
>Hi,
> I thought I'd have a look at this topic, and thus read what Wikipedia
>had on the subject.
> I implemented a basic version of the compress and decompress algorithms,
>namely, to deal with compression at the character level, as described on Wikipedia.
> The explanation about repeating codes "aSaSa" was not borne out in practice.
> The explanation about using more bits (when changing from
>an index of 31 to an index of 32) seemed to be inadequate,
>as the decompress algorithm has no idea when this change could occur.
I looked again (at the number of bits) from a different angle, and have
changed my mind. That seeks to be OK.
robin
Mar 6, 2010, 9:47:50 AM3/6/10
to
"Peter Flass" <Peter...@Yahoo.com> wrote in message news:hmmkgl$smo$1...@news.eternal-september.org...
| Somewhere in my library I have a book on compression algorithms (that
| I've never read). I'll see if it talks about LZW, and possibly compare
| to wikipedia.
Thanks. I look forward to any developments.
0 new messages