[Xml-bin] Status of projects?

[Alaric Snell]

Quoting "Anders W. Tell" <opensource@toolsmiths.se>:

> Yes I manage to avoid that with a 2 pass algorithm which collects all 
> chars in one single array and then put it through a GZIP Stream. The 
> problem are all those small references to the char array from the markup
> 
> area.
> 
> Im working on another solution but the block features of gzip algorithm
> 
> creates great difficulties when designing a 1 pass algorithm.

You really do want to just gzip the lot, IMHO. It'd be nice if gzip had an
interface for context-switching, to transparently embed multiple "streams" into
the output and have each block being read come with the name of the stream it's
from (ideal for embedding any kind of escaped data, to be honest)

I did some research into this when I was a youngster. The entropy encoding stage
of an algorithm like deflate takes values in the 0..255 range and encodes them
as variable length bit strings; however, there's nothing to stop it from taking
values in the 0..256 range, and using the special non-valid-byte 256 as a marker
for metadata or data block boundaries. But I digress...

> Cheers
> Anders

ABS

-- 
                               Alaric B. Snell
 http://www.alaric-snell.com/  http://RFC.net/  http://www.warhead.org.uk/
   Any sufficiently advanced technology can be emulated in software