perl-unicode

Re: 0th MODULAR Encode module is now available

2002-01-28 15:01:45
Dan Kogai <dankogai(_at_)dan(_dot_)co(_dot_)jp> writes:
jhi, Nick, Sadahiro and folks @ perl-unicode,

  I have just finished 'modularizing' Encode module.  

I appologise in advance for sounding rather negative about the idea.
I hate to turn down offers of help, but really wish you could do the 
bits I don't understand rather than re-working the bits I am quite 
good at.

Modularizing I
mean that Encode module now compiles AFTER you have installed perl5.7.2.

Err, it used to do that once - I have installed it in 5.6.0 for instance.
(Where it collapsed in a heap due to lack of enough UTF-8-ness.)
[Encode was once an add-on to assist Tk's "support" for Unicode ...]  


    + Based upon perl(_at_)14471
    + Files like Changes, MANIFEST are added
    + Encoding tables are all relocated from Encode/ to tables/

Any reason why? The files were under Encode to that an @INC search 
for Encode/... could be done allowing folk to install encodings 
in the "normal" manner.

    + One package, One File

What do you mean by that? 
iso-2022 is to be one file?! - we replicate 0-128 in all the ASCII-oid 
tables?!

    + Tab stop = 4 (Sorry, Nick!)

An indent of 4 and a tab-stop of 8 is _the_ perl (and most open source)
standard. I have just about beaten my editor into doing that now...


  As you see real work is yet to be done but at least now we can develop
Encode ASYNCHRONOUSLY from perl core.

Well I am far from sure that is a good idea until it works better than it 
does now - given that various parts of the core assume its existance.
We have various CPAN modules that do things "like" Encode - the rationale
for building it (as a framework) into the core was to try and provide 
an interface that people could rely on existing.

It was already "modular" in intent - in that some CPAN module could 
provide new encodings (tables) and encoders (algorithms).
It may have rough edges (patches welcome) and I know the multi-byte
and escape encodings are not finished but ...

  You can get one via

http://www.dan.co.jp/~dankogai/Encode-0.30.tar.gz

How big is it? - will take a look.


  and upon the approval of jhi, I will upload this to CPAN. jhi, please
have a look and tell me what you think.  FYI, Version 0.30 is simply how
RCS does 0.02++.

Well it's wrong ;-)


CAVEAT

* In terms of functionality (and the very lack thereof), it makes no
difference from the version that comes with perl(_at_)14471
* Perl 5.7.2 or later needed so far
* CJK decode is still kaput (Actually worsened since 5.7.2 release)

Dan the Encoder of Yours
-- 
Nick Ing-Simmons
http://www.ni-s.u-net.com/