perl-unicode

Encode-0.90 is now available

2002-03-19 07:06:02
From: Dan Kogai <dankogai(_at_)dan(_dot_)co(_dot_)jp>
Date: Tue Mar 19, 2002  11:04:28  Asia/Tokyo
To: Dan Kogai <dankogai(_at_)dan(_dot_)co(_dot_)jp>
Subject: Encode-0.90 now available

Encode Hackers,

  I have made Encode-0.90 available at

http://www.dan.co.jp/~dankogai/Encode-0.90.tar.gz

Changes since perl(_at_)15300
------------------------

+ Pod bugs in Encode::XX fixed
+ t/*.t now works independently of perl core so Encode is CPAN-uploadable
  (all "@INC = '../lib';" replaced with "unshift @INC, '../lib', for
   instance)
+ Module versions all set to 0.90 with RCS stub
  hopefully this will be 1.00 when 5.8.0 is ripe and uploaded to CPAN
+ partial support for jisx0212-1990.  See CAVEAT
+ Encode/euc-jp+0212.ucm added for the functionality above
! t/table.* renamed to jisx0208.*

CAVEAT
------

jisx0212 -> utf8 (decode) does not work while utf8 -> jisx0212 does. I've tried to work it out on my own but gave up for the time being because I want Nick to check it out. Nick, please try

perl5.7.3 -Mblib t/JP 1

to see what I mean. the last '1' enables jisx0212 tests. I will keep working on this problem but I need your help.

Well, jisx0212 is not round-trippable; it contains TILDE and others that are already mapped in ASCII and/or jisx0208. Still, it immediately croaks with a message like

euc-jp "\x8F" does not map to Unicode (3) at t/JP.t line 50, <$src> line 288.

Once multibyte chars with more than 3bytes/char works, we can even support such biggies as GB18030 (with the biggest char 4 byte big; icu distribution does contain this).

Dan the Yet Another Encode Hacker

<Prev in Thread] Current Thread [Next in Thread>