1 # Revision history for Perl extension Encode.
3 # $Id: Changes,v 1.32 2002/04/09 20:06:15 dankogai Exp dankogai $
6 1.32 $Date: 2002/04/09 20:06:15 $
9 - ucm/macDevanaga.ucm Unicode Character Map
10 - ucm/macGujarati.ucm Unicode Character Map
11 - ucm/macGurmukhi.ucm Unicode Character Map
12 A utility to check integrity of .ucm files. t/bogus.ucm is a
13 ucm that is deliberately bogus. unused Indic mappings are removed
16 resolve_alias() added as suggested by jhi. Same as
17 find_encoding("alias")->name. For convenience. This one is
18 defined in Encode.pm instead of Alias.pm.
19 Message-Id: <20020409215846.H17022@alpha.hut.fi>
21 Memory Allocate but detected during the devel of ucmlint -- fixed.
22 Message-Id: <C0DDCE16-4BE7-11D6-9204-00039301D480@dan.co.jp>
23 ! lib/Encode/Unicode.pm
24 valid_ucs2(0) is false but must be true.
25 3 patches from NI-S as follows. This also has fixed the incident
28 find_alias() recursion prevention
30 Checks for the patch above
32 An extra "F" that causes valid_ucs2() return a bogus value fixed
33 Message-Id: <20020409133927.17803.1@bactrian.elixent.com>
34 Message-Id: <Pine.SOL.4.10.10204091338220.10390-100000@maxwell.phys.lafayette.edu>
35 2 Small Patches from jhi as follows:
37 Encode->encodings() lists in case-insensitve order (as it was)
39 -l option prints avaiable encodings to STDOUT instead of STDERR
40 ! lib/Encode/Aliases.pm
41 s/defintion/definition/
42 Message-Id: <200204082306.CAA21033@alpha.hut.fi>
44 ! lib/Encode/Supported.pod
45 ! lib/Encode/Unicode.pm
46 POD revise by Philip Newton. This adds Philip to AUTHORS list.
47 Thank you for the exact quote of Douglas Adams :)
48 Message-Id: <22s3bu4gpvhhsses64nj3afuu0lo927rv3@4ax.com>
50 1.31 2002/04/08 18:08:07
51 ! lib/Encode/Encoder.pm
53 Encode::Encoder, once just a placeholder of an idea, is now much more
54 practical. See t/Encode.t to find how practical it can be.
55 + lib/Encode/Config.pm
57 my false laziness at Encode.pm is fixed. Now %ExtModules are set
58 in Encode::Config and they are all literally, not programatically
59 set. My false laziness was resulting many encodings missing from
61 ! lib/Encode/Unicode.pm
63 BOM for 32LE was bogus as noted by Anton. t/Unicode.t is fixed
64 so that it does not rely Encode::Unicode for BOM values
65 Message-Id: <FFEC33E9-4AFB-11D6-B415-00039301D480@dan.co.jp>
67 1.30 2002/04/08 02:34:51
68 + lib/Encode/Encoder.pm
69 Object Oriented Encoder. I reckon something like this is in need.
72 ! lib/Encode/Supported.pod
73 * autoloading bug that prevented upper-case canonicals such as UTF-16
74 is fixed. Now even UTF/UCS are autoloaded!
75 * encodings() is now more intuitive.
76 * t/Unicode.t fixed to explicitly use Unicode.pm -- BOM values are
78 * Obligatory fixes to the POD.
79 ! lib/Encode/Supported.pod
80 Patch from Anton applied.
81 Message-Id: <66641479.20020408033300@motor.ru>
83 ! lib/Encode/Unicode.pm
84 Cosmetic changes: "bless $obj, $class" => "bless $obj => class"
86 1.28 2002/04/07 18:58:42
90 Just a MANIFEST for those missing files.
92 1.26 Date: 2002/04/07 15:22:04
95 Schwarn's patches against Makefile.PL has zapped jis*.ucm. Restored.
96 And t/Aliases.t fixed to make sure they all exist.
98 1.25 2002/04/07 15:01:25 (Unreleased)
100 ! lib/Encode/Unicode.pm
103 - lib/Encode/UTF_EBCDIC.pm
104 - lib/Encode/Internal.pm
106 Integrated into Encode.pm as closures. That way "one package, one file"
107 rule is preserved yet less files to require.
109 commented out binmode(STDERR ...
118 ! Encode/Makefile_PL.e2x
119 Schwarn's MM-compliance patch merged
120 Message-Id: <20020406082609.GA28758@blackrider>
122 ! lib/Encode/Unicode.pm
123 + lib/Encode/UTF_EBCDIC.pm
125 - lib/Encode/10646_1.pm
126 - lib/Encode/ucs2_le.pm
127 (UCS-2|UTF-(16|32))(LE|BE)? implementation and cleanups. Instead of
128 per-module based (en|de)code, I saved a number of .pm by
129 reorganizing it as per-object base (Well, this is what Encode::XS
130 does under the hood). See Encode::Unicode for details.
131 The original Unicode.pm is now correctly renamed to UTF_EBCDIC.pm.
132 This module is used only on EBCDIC environments.
134 1.21 2002/04/05 14:46:34 (Not Released)
140 Are back to make Perl/Tk happy Smile, NI-S.
142 ! lib/Encode/Alias.pm
143 ! lib/Encode/Supported.pm
144 ! lib/Encode/10646_1.pm
145 ! lib/Encode/ucs2_le.pm
146 UCS-16BE is now canonical for UCS-2/ISO-10646-1.
147 Leftover implicit aliases in ucs2_le.pm removed. Tests and documents
148 updated to reflect changes.
149 essage-Id: <20020405114024.1290.17@bactrian.ni-s.u-net.com>
150 ! lib/Encode/Alias.pm
151 ! lib/Encode/Supported.pm
152 Anton's revision commited. Added Dan's own fixes as well.
153 Message-Id: <159103166906.20020405161134@motor.ru>
154 ! lib/Encode/Alias.pm
156 < qr/^UCS2-le$/i => '"UCS-2"', );
158 > qr/^UCS2-LE$/i => '"UTF-16LE"');
159 Sigh. Thank you, Anton.
160 Message-Id: <14567692196.20020405062020@motor.ru>
161 Message-Id: <69FEC0B4-483E-11D6-A045-00039301D480@dan.co.jp>
163 1.20 2002/04/04 19:50:52
165 the last minute addtion. Just give it a try. Docs remains to be done.
166 Not installed by default.
167 ! lib/Encode/Supported.pod
170 ! lib/Encode/Alias.pm
172 ! lib/Encode/10464_1.pm
173 ! lib/Encode/ucs2_le.pm
174 Canonical name for 'UCS-2le" is now "UTF-16LE". UCS-2 left
175 unchanged but UTF-16BE is added as an alias. Implicit aliases
176 move to Encode::Alias so init_alias() works more as expected.
177 Also, 'utf8' is now canonical with 'UTF-8' being an alias.
178 Though pedantically wrong, This should make perl mongers happier.
179 t/Alias.t is enhanced to test all these.
180 Message-Id: <9C39BD58-47AF-11D6-9D82-00039301D480@dan.co.jp>
182 Now all .ucm are stacked in byte_t; They all share ascii part so 50%
183 of the codepoints are common. CJKT left as is because the saving is
189 ! Encode/Makefile_PL.e2x
197 All occurance of _def.h replaced with .exh so djgpp works happily
198 ever after! To credit this amazing discovery, Laszlo is now in
200 Message-Id: <20020403181424.GA8778@freemail.hu>
201 Message-Id: <B5BF0C6F-4732-11D6-B13D-00039301D480@dan.co.jp>
204 ! Encode/Makefile_PL.skel
206 No more @INC fiddling! Uses $ENV{PERL_CORE} instead
207 Message-Id: <20020401222744.GX2000@blackrider>, et al.
209 Two more tests by added jhi
210 Message-Id: <200204020000.DAA25121@alpha.hut.fi>
213 The showstopper fixed -- Memory reallocation bug was causing
214 Encode::XS to fall into infinite loop on certain conditions.
216 Message-Id: <9572CAC4-463C-11D6-ABA5-00039301D480@dan.co.jp>, et al
221 ! lib/Encode/Supported.pod
222 Vendor encodings rebuilt out of original map files at unicode.org.
223 Indic languages such as MacDevanagali remain unspported do to the
224 shortcoming of encengine capabilities (they need algorithmical
225 conversion and I have no knowledge on that!). Pods fixed for added
227 Oh, macJapan.ucm renamed to macJapanese.ucm.
228 macROMnn is macRomanian and macRUMnn is macRumanian.
229 txt2ucm is a crude script that is used to convert them.
231 Unicode Compound Characters (used extensively on Mac) supported
233 Typo fixes and improvements by jhi
234 Message-Id: <200204010201.FAA03564@alpha.hut.fi>, et al.
236 1.11 $Date: 2002/04/09 20:06:15 $
240 Missing files from the MANIFEST fixed.
241 Message-Id: <20020401010156.H10509@alpha.hut.fi>
242 Version incremented just to make CPAN happy.
244 1.10 2002/03/31 21:32:42
247 INSTALL_UCM option added to Makefile.PL so you can install *.ucm
248 if you want. This should make Autrijus happy. Also, piconv
249 is added to default install.
252 Here-documented files that enc2xs generates are now exported
253 to *.e2x. Much cleaner and easier to debug.
255 encoding enhances so you can make it act more like such
256 (now prehistoric ) "localized" variations of perl like Jperl.
258 Further test for encoding.pm. Written in euc-jp
262 Message-Id: <20020330174618.B10154@alpha.hut.fi>
267 *.ucm relocated to ucm/ so MakeMaker will not install'em by default.
274 ! Encode/macIceland.ucm
275 ! lib/Encode/Alias.pm
276 ! lib/Encode/Supported.pod
277 MacIceland fixes and Pod Typo fixes. This adds Andreas to AUTHORS.
278 Message-Id: <m3lmcavhjt.fsf@anima.de>
280 1.01 2002/03/29 20:59:39
283 s/USE_SCRIPTS/MORE_SCRIPTS/
285 installs enc2xs by default for external Encode:: modules in CPAN,
286 such as Encode::HanExtra
288 More sensible perl core detection via $ENV{PERL_CORE}
290 Message-Id: <200203291007.FAA07329@Orb.Nashua.NH.US>
292 Perl core ditection via $^X =~ m/\bminiperl$/o
293 Message-Id: <A5C7B0CA-42F1-11D6-B5AD-00039301D480@dan.co.jp>
297 The version of all files is updated to 1.00 via "ci -f -l1.00",
298 commemorating version 1.00. All files, including *.ucm are now
299 under version control.
302 encode.h moved to Encode/ so it will be installed for the later
305 h2xs-like feature added via "h2xs -M Name *.(enc|ucm)"
310 compile renamed to enc2xs.
311 Affected Makefle.PL updated
313 "Punt it. HanExtra can take care of that later." -- Autrijus
314 Message-Id: <20020328154338.GA7351@not.autrijus.org>
318 ! lib/Encode/CJKConstants.pm
319 ! lib/Encode/KR/2022_KR.pm
320 Table patches for Euro Signs, 2022-KR fixups by Jungshik
321 Message-Id: <Pine.LNX.4.44.0203280616190.2259-200000@www.ykga.org>
325 bin/ added for example scripts. They are not installed by default.
326 to install them, "perl Makefile.PL USE_SCRIPTS".
327 piconv is iconv reinvented in perl. in addition to all features
328 of iconv, it also adds perlish features. See L<piconv/1> for more
330 ! lib/Encode/Alias.pm
331 qr/^ replaced with qr/\b so it directly matches locale names
332 such as en_US.US-ASCII
335 Patch by MJD to fix the following problem applied.
336 Subject: [PATCH 5.7.3 Encode]
337 Aliases.t not properly skipped when Encode extension not built
338 Message-Id: <20020328091850.18677.qmail@plover.com>
339 ! lib/Encode/KR/2022_KR.pm
340 ! lib/Encode/CJKConstants.pm
341 Another patch from Jungshik to make iso-2022-kr actually work
342 Message-Id: <Pine.LNX.4.44.0203271745210.30462-200000@www.ykga.org>
343 ! Encode/Encode/euc-kr.ucm
344 + Encode/Encode/johab.ucm
345 ! Encode/Encode/ksc5601.ucm
347 ! Encode/KR/Makefile.PL
348 ! Encode/lib/Encode/Alias.pm
350 Johab support and complete revision of Korean Encoding by Jungshik
351 Message-Id: <Pine.LNX.4.44.0203271105060.30462-200000@www.ykga.org>
353 Revised to make up with now-dropped Encode::Details.
354 - lib/Encode/Details.pod
355 Dropped. Besides being obsolete, the topics are now covered in
361 Korean aliases fixed thanks to Jungshik Shin
362 /ks[-_ ]?c[-_ ]?5601-1987$/i => cp936
363 Message-Id: <Pine.LNX.4.44.0203262102250.1237-100000@www.ykga.org>
365 =head1 NAME added to all modules to make buildtoc happy
366 Message-Id: <20020327041151.A10618@alpha.hut.fi>
367 - lib/Encode/CJKguide.pod
368 Too controversial and dropped from the dist. Will be available
369 separately on the web.
371 RCS tags added so table debugging gets easier (should that be
372 needed! I hope they all stay 1.00!)
373 + lib/Encode/CJKguide.pod
374 A detailed guide to mainly, but not limited to, CJK multibyte
377 + Encode/hp-roman8.ucm
379 ! Encode/Supported.pod
380 All occurance of "roman8" replaced with "hp-roman8" to avoid
382 ! Encode/Supported.pod
385 Mac Encodings now comply the Inside Macintosh
387 Test for '-raw' conventions added.
389 aliased gb2312 -> euc-cn, ksc5601 -> euc-kr
393 "-raw" appended to canonical names.
394 File mames stay unchanged thanks to UCM format.
395 ! lib/Encode/CN/HZ.pm
396 Patch from Autrijus to fix gb2312 -> gb2312-raw + code linting
397 Message-Id: <20020326035210.GA2091@not.autrijus.org>
400 - lib/Encode/JP/Const.pm
401 + lib/Encode/CJKConstants.pm
402 + lib/Encode/CN/2022_CN.pm
403 + lib/Encode/KR/2022_KR.pm
412 * Support for ISO-2022-KR and ISO-2022-CN added.
414 * more t/*.{euc,ref} added, which was autogenerated from ucm2table
415 * ucm2table autogenerates character table out of UCM files.
418 - lib/Encode/Supports.pod
419 + lib/Encode/Supported.pod
420 Names reverted due to popular demand.
421 8.3 rule applies only when there is a conflict.
422 Message-Id: <20020325095924.GD44120@not.autrijus.org>
427 - lib/Encode/Format/Enc.pod
429 * Character tables is now 100% ucm.
430 * All files under Encode/ is now 8.3-compliant
431 * some of missing encodings added (i.e. gsm0338 and nextstep)
432 * Vendor mappings aggregated with appropriate national std in
433 Makefile.PL, resulting smaller *.so especially for CJK.
434 Following is result on Dan's FreeBSD box.
436 ---------------------------------------------------------------
437 blib/arch/auto/Encode/Byte/Byte.so 157,279 171,042
438 blib/arch/auto/Encode/CN/CN.so 1,634,476 1,626,685
439 blib/arch/auto/Encode/EBCDIC/EBCDIC.so 18,476 18,476
440 blib/arch/auto/Encode/Encode.so 27,791 27,791
441 blib/arch/auto/Encode/JP/JP.so 1,408,056 1,832,811
442 blib/arch/auto/Encode/KR/KR.so 1,156,518 1,329,587
443 blib/arch/auto/Encode/Symbol/Symbol.so 23,940 20,990
444 blib/arch/auto/Encode/TW/TW.so* 948,761 1,316,437
445 ---------------------------------------------------------------
446 Total 5,375,297 6,343,819
448 * As a result of ucm-transition, Encode::Tcl dropped because
449 Encode::Tcl demands *.enc.
450 Encode::Tcl will be supplied in a separate tarball with *.enc.
451 Message-Id: <C024E294-3FC3-11D6-8347-00039301D480@dan.co.jp>
456 -lib/Encode/Supported.pod
457 +lib/Encode/Supports.pod
458 -lib/Encode/iso10646_1.pm
459 +lib/Encode/10646_1.pm
460 -lib/Encode/EncFormat.pod
461 +lib/Encode/Format/Enc.pod
462 Files renamed 8.3 filename compliance. Affected modules/scripts revised.
463 - lib/Encode/JP/Constants.pm
464 + lib/Encode/JP/Consts.pm
465 ! lib/Encode/JP/JIS.pm
466 ! lib/Encode/JP/H2Z.pm
467 Version nit problem and 8.3 rule fix.
468 > Package namespace installed latest in CPAN file
469 > Encode::JP::Constants 0.92 1.02 J/JH/JHI/perl-5.7.3.tar.gz
470 was noted by jhi then Dan discovers "Constants.pm" does not comply 8.3
471 rule. Contants.pm renamed to Consts.pm and affected modules are fixed
472 accordingly. In addition, legacy "use vars qw()..." are replaced with
474 Message-Id: <20020325011248.D1561@alpha.hut.fi>
475 Message-Id: <41023D51-3FB5-11D6-8347-00039301D480@dan.co.jp>
477 - lib/Encode/JP/ISO_2022_JP.pm
478 - lib/Encode/JP/ISO_2022_JP_1.pm
479 + lib/Encode/JP/2022_JP.pm
480 + lib/Encode/JP/2022_JP1.pm
482 8.3 naming conflict for vanilla fat addressed by jhi
483 Message-Id: <20020324201931.V22596@alpha.hut.fi>
486 Typecast fix addressed by jhi
487 Message-Id: <20020324185540.T22596@alpha.hut.fi>
490 ! lib/Encode/Supported.pod
492 + lib/Encode/JP/ISO_2022_JP_1.pm
493 ! lib/Encode/JP/ISO_2022_JP.pm
494 ! lib/Encode/JP/JIS.pm
496 Now Encode::JP is more strict on the difference between ISO-2022-JP
497 and ISO-2022-JP-1. See JP/JP.pm for details. I hope this move
498 makes Anton happier :) FYI the previous version implements
499 ISO-2022-JP as ISO-2022-JP-1 since it had X0212 support.
500 ! lib/Encode/Supported.pod
503 Avoid core-dump in Encode with PERLIO=mmap by NI-S
504 Message-Id: <20020324104139.1326.7@bactrian.ni-s.u-net.com>
509 ! lib/Encode/Suppoted.pod
510 pod fixes to replace F<http://...> to L<http://...>,
511 as suggested by Autrijius in:
512 Message-Id: <20020324083943.GA14901@not.autrijus.org>
513 ! lib/Encode/Suppoted.pod
514 fixes and enhancements by Anton
515 Message-Id: <10632060120.20020324103753@motor.ru>
516 ! lib/Encode/Alias.pm
517 > define_alias( qr/^GB[- ]?(\d+)$/i => '"gb$1"' );
518 added. Suggested by Anton then deobfuscated by Autrijius
519 Message-Id: <20020324064455.GA3667@not.autrijus.org>
521 Further fix by Nicholas Clark
522 Message-Id: <20020323145840.GD304@Bagpuss.unfortu.net>
523 - lib/EncodeFormat.pod
524 + lib/Encode/EncFormat.pod
526 File renamed as suggested by Autrijius
528 ! lib/Encode/Details.pod
529 ! lib/Encode/Supported.pod Sun Mar 24 13:29:35 2002
530 ! Encode.pm Sun Mar 24 13:43:47 2002
531 pod fixes by Autrijius.
532 Message-Id: <20020324062804.GA3595@not.autrijus.org>
533 Message-Id: <20020324075627.GB11986@not.autrijus.org>
535 ! lib/Encode/Alias.pm
537 now more EBCDIC conscious;
538 %ExtModules on EBCDIC system excludes CJK so that you don't
539 have to worry about the matched alias resulting cloaking.
540 t/Alias.t also revised to reflect changes. Verified by jhi
541 Message-Id: <20020324022929.D22596@alpha.hut.fi>
547 EBCDIC detection mechanism installed as in JP/JP.pm
548 Message-Id: <20020323211847.G19148@alpha.hut.fi>
556 Now all table files used by compile are postfixed '_t' to avoid
557 namespace collisions in case insensitive file systems once for all!
559 Message-ID: <58290227735.20020323195659@familiehaase.de>
561 Since the Encode::JP is unsupported under EBCDIC we
562 cannot run this test (aliases as such should work fine) -- jhi
563 Message-Id: <20020323202119.D19148@alpha.hut.fi>
565 duplicate occurance of ascii.ucm and 8859-1.ucm
566 causes MacOS X dlyd to cloak
572 < chdir 't' if -d 't';
574 > if (! -d 'blib' and -d 't'){ chdir 't' };
575 When you are "make test"-ing on Encode/ directory, you must not
576 change $ENV{PWD}. t/JP.t has been fixed before but others somehow
577 remain unchanced. Also the situation detection was made simpler
578 in t/JP.t, which was originally;
579 > chdir 't' if -d 't' and $ENV{PWD} !~ m,/Encode[^/]*$,o;
581 "Use of uninitialized value in string eq at Encode.pm line 96."
585 -- Problem on case insensitive file systems
586 "coexist of ebcdic.c <> EBCDIC.c on Cygwin not possible"
587 Message-ID: <88254111953.20020323095503@familiehaase.de>
590 "So I think it's a bug in gcc, not perl. But it still needs to be
592 Message-Id: <20020323145840.GD304@Bagpuss.unfortu.net>
593 Message-Id: <20020323170509.C96475@plum.flirble.org>
597 ! lib/Encode/Encoding.pm
598 ! lib/Encode/Alias.pm
599 ! lib/Encode/Supported.pod
601 Pod Fixes by Michael G Schwern <schwern@pobox.com> via jhi
602 Message-ID: <20020322073908.GB10539@blackrider>
605 "...I think we should include ISO 8859-1 as well." -- NI-S
606 Message-Id: <20020322120230.1332.8@bactrian.elixent.com>
611 ! lib/Encode/Alias.pm
612 alias definitions relocated to Encode::Alias so module autoloading
613 works for aliases also.
615 encodings() now accepts args to check ExtModules.
624 Latin and single byte encodings are reorganized so they are
625 demand-loaded like Encode::XX. Now only ascii is compiled into
627 ! lib/Encode/Alias.pm
628 for my $k (keys %hash){ delete $hash{$k}; }
629 is depreciated; fixed.
632 In this update, pod rewrites and alias fixes are the main issues
633 + lib/Encode/Supported.pod
634 Describes supported encodings
636 streamlined compiled-in encodings.
637 ! lib/Encode/Description.pod -> lib/Encode/Details.pod
639 + Encode/ibm-125?.ucm
640 Added from icu distibution with any occurance of
641 "IBM-125?" to "cp125?". Filenames remain unchanged to pay
642 some respect to icu staff, however.
643 + lib/Encode/Alias.pm
645 Alias difinitions in Encode.pm relocated.
648 packWARN patch from Paul Marquess via jhi
649 Message-Id: <20020321010101.O28978@alpha.hut.fi>
650 Paul added to AUTHORS as a result.
651 ! t/CJKalias.t -> t/Aliases.t
652 Renamed. Checks even more aliases and alias overloading
655 duplicate alias for ujis => euc-jp removed (Encode::JP has one)
656 gbk => cp936 relocated to CN.pm
658 Test::More with plans (by jhi)
661 + lib/Encode/Description.pod
662 ! lib/Encode/Encoding.pm
663 Now the pod in Encode.pm is abridged as programming references.
664 lib/Encode/Description.pod contains the original, detailed description
665 and Encode::Encoding explains how to write your own module to
666 add new encodings. So far, lib/Encode/Description.pod contains
667 the whole pod once in Encode.pm. This is intentional.
669 Pod revisions by Anton Tagunov
670 Message-Id: <517178431.20020320174824@motor.ru>
672 all occrance of Encode::Tcl::Extended removed including pod
674 test now checks $encoding->name only; $encoding->{name} are
675 no longer check to find the canonical name.
676 ! lib/Encode/JP/JIS.pm
677 ! lib/Encode/JP/ISO_2022_JP.pm
678 ->name() added to be more compliant with API
684 Patch by Autrijus to add aliases to TW and fixes to POD
685 Message-Id: <20020320090619.GA24774@not.autrijus.org>
687 SADAHIRO Tomoyuki added as should. My apologies.
690 * First release to be uploaded to CPAN. For prehistoric changes,
691 please see Changes file of perl distibution as well as
692 perl-unicode@perl.org archive, available at:
693 http://archive.develooper.com/perl-unicode@perl.org/
695 Changes Since 0.92 includes;
700 + Mention to perl-unicode@perl.org added
702 + Encoding aliases added so you can feed locale names
703 and MIME Charset="" directly.
704 - Mention to JISX0212 removed because it's fixed
707 + Encoding aliases added. Note TW is left untouched because
708 euc-tw is not implemented in TW but in Encode::HanExtra.
709 Autrijus, you may fix Encode::HanExtra.
711 + to test encode aliases added