I want to compress an ascii text file, but would like the output to be

another ascii text file.

Is there an algorithm to do this that beats simply compressing and

uuencoding ?

Thanks, Maldwyn.

I want to compress an ascii text file, but would like the output to be

another ascii text file.

Is there an algorithm to do this that beats simply compressing and

uuencoding ?

Thanks, Maldwyn.

:I want to compress an ascii text file, but would like the output to be

:another ascii text file.

:Is there an algorithm to do this that beats simply compressing and

:uuencoding ?

Of course there is. uuencoding only uses a range of 64 characters per

position. Even just using an encoding algorithm that packed more densely

would be better.

There is no particular reason that a compression algorithm needs to

output into base 256 (a binary file), or indeed into a base that is

a power of 2. Arithmetic encoding in particular might be more

efficient with base 256, but works fine with other bases such as base

95 (the number of printable characters in ASCII.)

PGP with the "-c" option will compress your ascii file beforeQuote:>I want to compress an ascii text file, but would like the output to be

>another ascii text file.

>Is there an algorithm to do this that beats simply compressing and

>uuencoding ?

encrypting it with IDEA encryption. The encrypted ascii file will be

approximately 30% larger than an encrypted binary of the same source

file, but the compression more than compensates.

For example, for a text file containing the complete King James Bible:

uncompressed ascii = 4,844KB

PGP compressed (ascii) = 2,116KB

PGP compressed (binary) = 1,539KB

Zip compressed (binary) = 1,394KB

The result of PGP compression is of course unreadable (it's encrypted)

but it can be mailed through normal e-mail channels. MIME and UUENCODE

also compress files to ascii, though I don't know the compression

ratios.

Hope this helps.

_________________________________

>There is no particular reason that a compression algorithm needs to

>output into base 256 (a binary file), or indeed into a base that is

>a power of 2. Arithmetic encoding in particular might be more

>efficient with base 256, but works fine with other bases such as base

>95 (the number of printable characters in ASCII.)

and all sorts of ugly things crop up. I know --- I implemented AC

with residue number systems in '94, and when your base is not a power

of two, there is LOTS of trouble ;-)

Bernie

--

===========================================================================

"How does Windows work?"

Mike Battersby, BCS (Hon), System administrator at Deakin Uni, while trying

to handle Netscape/Windows during a meeting of the Linux Users of Victoria

> > ...

> > Is there an algorithm to do this that beats simply compressing and

> > uuencoding ?

> Of course there is...

> ...

( I had hoped my question kind of implied this, but there you go...)

Maldwyn.

>>There is no particular reason that a compression algorithm needs to

>>output into base 256 (a binary file), or indeed into a base that is

>>a power of 2. Arithmetic encoding in particular might be more

>>efficient with base 256, but works fine with other bases such as base

>>95 (the number of printable characters in ASCII.)

>You _really_ think so? In theory, of course, but try implementing it,

>and all sorts of ugly things crop up. I know --- I implemented AC

>with residue number systems in '94, and when your base is not a power

>of two, there is LOTS of trouble ;-)

with 95 characters, isn't it? The only ugly thing that crops up is

how to encode the end of the bitstream.

Leo

>>You _really_ think so? In theory, of course, but try implementing it,

>>and all sorts of ugly things crop up. I know --- I implemented AC

>>with residue number systems in '94, and when your base is not a power

>>of two, there is LOTS of trouble ;-)

>To encode into base 95 is to decode the bitstream using a uniform model

>with 95 characters, isn't it? The only ugly thing that crops up is

>how to encode the end of the bitstream.

anyway), and mixed up the internal workings of the AC and the output

alphabet. Of course, as long as they use the same number base, you are

fine --- but implementing a base-2 output AC with something other than

base-2 for the internal workings, _that's_ trouble.

Sorry for the confusion,

Bernie

--

===========================================================================

"How does Windows work?"

Mike Battersby, BCS (Hon), System administrator at Deakin Uni, while trying

to handle Netscape/Windows during a meeting of the Linux Users of Victoria

Hi,

Does anyone know of any Text to Text compression utilities or algorithm?

The scenario is that I have an XML file which has one of its elements

too huge[ about 4 MB ] and I don't want that to be sent over the

network causing severe delay. I'm planning to compress that element [

which has only 0-9 and A-F characters ] alone and then embed the

compressed one instead of the original element thereby reducing the

size of the XML file to be sent over the network. At the receiving side

I can write the inflation logic inside the Parser.

Does anyone know about any utility to do the compression or does anyone

know of better way of implementing this ?

Thanx,

Prem.

Sent via Deja.com http://www.deja.com/

Before you buy.

3. Text Compression, Image Compression and GEnetic Algorithms

4. Win2K backup

5. Which program to transform a scanned text-page in a text ?

6. Is it possible to make goodlooking document with mathematical symbols using matlab. How?

7. HP6200...Whats the smallest format to save text images(not using text for.)??

8. a2091 - Garbled ROMS - WTF?

9. Is Excel text not really text?

10. Getting text entry in one cell to return text entry in adjoining cell

11. 123 Formula (Text + Cell Reference + Text)

12. Convert XL or comma dlimited text to lineat text?

7 post • Page:**1** of **1**