source: trunk/essentials/dev-lang/perl/ext/Encode/Changes@ 3314

# Revision history for Perl extension Encode.
#
# $Id: Changes,v 2.12 2005/09/08 14:17:17 dankogai Exp dankogai $
#
$Revision: 2.12 $ $Date: 2005/09/08 14:17:17 $
! Encode.xs Encode.pm t/fallback.t
  Now accepts coderef for CHECK!
! ucm/8859-7.ucm
  Updated to newer version at unicode.org
  http://rt.cpan.org/NoAuth/Bug.html?id=14222
! lib/Encode/Supported.pod
  More POD typo fixed.
  <[email protected]>
! encoding.pm
  More POD typo leftover fixed.
  Message-Id: <[email protected]>

2.11  2005/08/05 10:58:25
! AUTHORS CHANGES
  To reflect changes below
! Encode.pm encoding.pm 
  lib/Encode/Alias.pm lib/Encode/PerlIO.pod lib/Encode/Supported.pod
  Typo fixed by Piotr Fusik in Change 25261 & 25266
  Message-ID: <001401c595bd$dccb5d80$0bd34dd5@piec>
! Encode.xs
  Addresses "BUG REPORT: panic in Encode.xs".
  Message-Id: <[email protected]>
+ lib/Encode/MIME/Header/ISO_2022_JP.pm mime_header_iso2022jp.t
! lib/Encode/MIME/Header.pm lib/Encode/Config.pm
  Encoding 'MIME-Header-ISO_2022_JP' is introduced by Makamaka
  Message-Id: <[email protected]>
! Encode/encode.h Encode.pm Encode.xs
  PerlIO's "encoding(utf-8-strict)" got a problem w/ partial character.
  Found and addressed by KONNO Hiroharu <[email protected]>
  See also ext/PerlIO/encoding/encoding.pm   
  Message-Id: <[email protected]>

2.10 2005/05/16 18:46:36
! Encode.pm
  fixed decode_utf8() accordingly to RT#8872
  http://rt.cpan.org/NoAuth/Bug.html?id=8872
! Encode.xs AUTHORS
  s/SvIVX/SvIV_set/ by Steve Peters.
  Message-Id: <[email protected]>
! AUTHORS
  GAAS was missing!
! Encode.pm
  New Pod section: "UTF-8 vs utf8"; explains utf-8-strict
+ t/utf8strict.t
  Tests utf-8-strict, accordingly to 
  UTF-8 decoder capability and stress test" by Markus Kuhn
  http://smontagu.damowmow.com/utf8test.html
  Note that malformed and overlong sequences are not test here
  because perl already does that for you, utf-8-strict or not.
! Encode.pm Encode/encode.h t/fallback.t
  Addressed "encode(..., Encode::LEAVE_SRC) does not work".
  Now FB_(PERLQQ|HTMLCREF|XMLCREF) implies LEAVE_SRC so
  you can (en|de)code constant strings with these fallbacks.
  http://rt.cpan.org/NoAuth/Bug.html?id=8736  
! Encode.pm Encode.xs lib/Encode/Alias.pm t/Aliases.t
  Make Encode.pm support the real UTF-8, by GAAS
  Message-Id: <[email protected]>
  Message-Id: <[email protected]>
! Encode.pm Encode.xs
  post-2.09 comment patches from GAAS applied.
  Message-Id: <[email protected]>
  Message-Id: <[email protected]>

2.09 2004/12/03 19:16:53 
! Encode.pm Encode.xs
  Addressed " :encoding(utf8) broken in perl-5.8.6".
  Message-Id: <[email protected]>
! Encode.pm
  Addressed "(de|en)code($valid_encoding, undef) does not warn".
  http://rt.cpan.org/NoAuth/Bug.html?id=8723
! Encode.pm t/Encode.t
  Addressed "Can't encode URI".  When a reference is fed to (en|de)code,
  Encode now stringifies instead of returning undef.
  http://rt.cpan.org/NoAuth/Bug.html?id=8725
! Encode.xs t/fallback.t
  Addressed "FB_HTMLCREF and FB_XMLCREF for the UTF-8 decoder".
  http://rt.cpan.org/NoAuth/Bug.html?id=8694
! Encode.pm
  Addressed "s/digit/number/".
  http://rt.cpan.org/NoAuth/Bug.html?id=8695
! Encode.pm
  Addressed "while (defined(read )) { ... } is an infinite loop".
  http://rt.cpan.org/NoAuth/Bug.html?id=8696
! Encode.pm
  Addressed "What the heck is UCM?".
  Document fixed so that it no longer contains "UCM-Based Encodings".
  http://rt.cpan.org/NoAuth/Bug.html?id=8697

2.08 2004/10/24 13:00:29
! Encode.xs lib/Encode/Encoding.pm  Unicode/Unicode.{pm,xs}
  Resolved the issue that was raised by 2.07 --  Encode::utf8 fallbacks
  that was introduce messed up  PerlIO::encoding.
  * To do so, ->renew() is renewed and ->renewed() was introduced to
  tell whether the caller is PerlIO or not.
  Message-Id: <[email protected]>

2.07 2004/10/22 19:35:52
! lib/Encode/Encoding.pm
  "Remove Carp from warnings.pm" that influences Encode, by Tels.
  Message-Id: <[email protected]>
! Encode.xs AUTHORS t/fallback.t
  Now Encode::utf8's fallbacks are compliant to Encode standard.
  Thank Bjoern Hoehrmann for persistently convincing me.
  Message-Id: <[email protected]>
! Encode.pm
  POD further revised.

2.06 2004/10/22 06:23:11
! ucm/mac*
  RT #8083 reports that MacThai mapping was obsolete
  Updated all mac* encodings accordingly to the URI below.
  One remaining mystery is that MacRomanian vs. MacRumanian.
  MacRumanian is not found in unicode.org...
  http://www.unicode.org/Public/MAPPINGS/VENDORS/APPLE/
! Encode.pm t/Encode.t
  Fixed RT #8081: "decode(..., bless{},'x') segfault"
  Two more tests added to test that.
  http://rt.cpan.org/NoAuth/Bug.html?id=8081
! Encode.pm
  POD revised accordingly to RT #7966
  http://rt.cpan.org/NoAuth/Bug.html?id=7966
! Unicode/Unicode.pm
  POD updated explaining why Encode::Unicode always croaks on error
  rather than giving users choices.
  http://rt.cpan.org/NoAuth/Bug.html?id=7892

2.05 2004/10/19 04:55:01
! encoding.pm
  "unnuke" jhi's patch in bleedperl, with minor correction by dankogai.
  Message-ID: <[email protected]>
  Message-ID: <20041018233442.7418113f@r2d2>
  Message-Id: <[email protected]>

2.04 2004/10/16 21:22:44
! Makefle.PL 
  From:   [email protected]
  Subject: [PATCH ext/Encode/Makefile.PL] make Encode.c dependency explicit
  Message-Id: <[email protected]>

2.03 2004/10/06 05:07:20
! lib/Encode/Alias.pm
    Resolved some alias case sensitivity glitches reported via RT.
    http://rt.cpan.org/NoAuth/Bug.html?id=7835
! bin/piconv
    Resolved Win32 glitches reported via RT. 
    (Fixed by dankogai and tested by Steve Hay)
    http://rt.cpan.org/Ticket/Display.html?id=7831
! JP/JP.pm lib/Encode/Alias.pm lib/Encode/Supported.pod AUTHORS
    /\bwindows-31j$/i is now an alias of CP932, by Steve Hay.
    http://rt.cpan.org/NoAuth/Bug.html?id=6695

2.02 2004/08/31 10:55:34
! ucm/big5-hkscs.ucm AUTHORS t/big5-hkscs.enc t/big5-hkscs.utf
  New map submitted by Deng Liu and Autrijus.  Test data needed
  to be upgrade as well, done by dankogai
  Message-Id: <[email protected]>
! bin/ucmsort
  Now works for characters U+10000 and above.  This fix was needed
  to "tidy" the original map that was submitted.
! bin/enc2xs
  "ucmsort" now mentioned in pod

2.01 2004/05/25 16:27:14
! bin/enc2xs AUTHORS
  From:   [email protected]
  Subject: [PATCH] Correct statistics from enc2xs
  <[email protected]>
  <[email protected]>
! lib/Encode/Alias.pm
  Addressed "False [] range "\s-" in regex;" in Encode::Alias.pm
  <[email protected]>

2.01 2004/05/25 16:27:14
! lib/Encode/CN/HZ.pm lib/Encode/Unicode/UTF7.pm
  "If someone thinks utf8::upgrade($1) should be croaked like 
  chom?p($1),please try the following patch for Encode.pm."
  -- sadahiro-san
  <[email protected]>

2.0 2004/05/16 20:55:15
* version updated to 2.00
   -- sorry, no big feature change.  I just hate version 1.100 :)
! lib/Encode/Guess.pm
  Unicode/Unicode.pm
  addressed  UTF-(8|32LE) + BOM misguessing
  https://rt.cpan.org/Ticket/Display.html?id=6279
! Encode.pm
  s/is_utif8/is_utf8/ in POD
! Encode/lib/Encode/CN/HZ.pm 
  Fixes "make test" failure after the patch to pp_hot.c
  by Sadahiro-san
  Message-Id: <[email protected]>
! bin/piconv
  From:   [email protected]
  Subject: [PATCH] "piconv -C 512" badly broken
  Message-Id: <1072870210.769.5.camel@localhost>

1.99 2003/12/29 02:47:16
! Unicode/Unicode.xs
  find_encoding("UTF-16BE")->encode("abc") now null terminates
  http://www.mail-archive.com/[email protected]/msg69766.html
! Encode.pm
  prototype bug in decode_utf8() fixed
  Message-Id: <[email protected]>
! Encode.pm /MANIFEST encoding.pm lib/Encode/Supported.pod 
  t/at-cn.t t/at-tw.t  t/gsm0338.t ucm/gsm0338.ucm
+ t/gsm0338.t
  Merged from maintperl@21987

1.98 2003/08/20 11:15:31
! lib/Encode/MIME/Header.pm AUTHORS t/mime-header.t
  Dave Evans has found and corrected a bug in Encode::MIME::Header.
  Test suite added by Dan Kogai.
  Message-Id: <[email protected]>
! encoding.pm
  Typo fixes rolled back in from bleedperl
! t/at-cn.t t/at-tw.t
  v-strings, now depreciated in perl 5.8.1, is replaced by sadahiro
  Message-Id: <[email protected]>
! bin/enc2xs
  argv case nit for VMS by Craig
  Message-ID: <[email protected]>
! t/enc_eucjp.t t/enc_utf8.t AUTHORS
  Encode test fixes for VMS by Peter Prymmer
  Message-ID: <OFBD4A7559.D7CF9517-ON85256D6B.00534853-85256D6B.00538131@factset.com>
! lib/Encode/Alias.pm t/Aliases.t
  koi-8 aliases bug detected and patched by sadahiro.
  Further fix and test suite by dankogai
  Message-Id: <[email protected]>

1.97 2003/07/08 21:52:14
! encoding.pm lib/Encode/Guess.pm lib/Encode/Alias.pm 
  lib/Encode/JP/JIS7.pm lib/Encode/Encoder.pm Encode.pm
  $DEBUG replaced with DEBUG() so perl optimizes better,
   by Rafael with further fixes by dankogai
  Message-Id: <[email protected]>
! lib/Encode/Aliases.pm
  Was: define_alias( qr/\bGB[-_ ]?2312(?:\D.*$|$)/i => '"euc-cn"' );
  Now: define_alias( qr/\bGB[-_ ]?2312(?!-?raw)/i => '"euc-cn"' );
  So new hash seeding introduced in bleedperl works.
  Message-Id: <[email protected]>
! lib/Encode/Guess.pm
  $Encode::Guess::NoUTFAutoGuess is added so you can turn off
  automatic  utf(8|16|32) guessing -- originally by Autrijus
  Message-Id: <[email protected]>
! Encode.pm
  Addressed the following;
  Subject: [perl #22835] FB_QUIET doesn't work with Encode::encode 
  Message-Id: <[email protected]>

1.96 2003/06/18 09:29:02
! lib/Encode/JP/JP.pm t/guess.t
  m/(...)/ in void context then $1 is considered a Bad Thing
  Message-Id: <[email protected]>
! Encode.pm
  Mentions in POD that as of perl 5.8.1 utf8::is_utf8() is
  also available.
! encengine.c
  More typecast from maintperl@19739
  Message-Id: <[email protected]>
! t/perlio.t
  Tests 37 & 38 failed on Win32 -- yet another CRLF issue
  Message-Id: <[email protected]>
! t/Encode.t
  Now skips for EBCDIC platform.
  Message-Id: <[email protected]>
! t/perlio.t
  Craig's patch applied that addresses "Many systems (DOS, VMS) cannot
  have more than one C<.> in their filenames." -- perlport.
  Message-Id: <[email protected]>
! bin/piconv
  Found and fixed the back that -p,--perlqq does not work.
  Induced by the change from Getopt::Std to Getopt::Long.
! encoding.pm
  Addressed [cpan #2629] Wrong assumption in numeric comparison
  Message-Id: <[email protected]>
! Encode.pm Encode.xs Unicode/Unicode.pm Unicode/Unicode.xs
 lib/Encode/Encoding.pm t/perlio.t
 ! API Change: ->new_sequence() => ->renew()
 + Encode::Unicode makes use of it so it can handle BOM on PerlIO
 + Encode::XS and Encode::utf8 now supports ->renew()
 + Encode::Encoding now documents this with examples
 - Non-XS (en|de)code stripped out of Encode::Unicode
 Message-Id: <[email protected]>

1.95 2003/05/21 08:41:11
! ucm/8859-*.ucm
  Since bogus entries were found in iso-8859-6, all entries are
  re-generated once again out of
  http://www.unicode.org/Public/MAPPINGS/ISO8859/8859-*.TXT
  Thank David Graff <[email protected]> for the discovery
  Message-Id: <[email protected]>
+ lib/Encode/Unicode/UTF7.pm
! lib/Encode/Config.pm lib/Encode/Alias.pm Unicode/Unicode.pm t/Unicode.t
  lib/Encode/Supported.pod
  UTF-7 support is now added.  With this Encode now has all transcoding 
  methods in Unicode::String.

1.94 2003/05/10 18:13:59
! lib/Encode/MIME/Header.pm
  A more sophisticated solution for double-encoding by dankogai
! lib/Encode/MIME/Header.pm AUTHORS
  Two bugs fixed by Bjoern Jacke
  * "Double Encoding" was not possible
    i.e. encode("MIME-B" => "=?UTF-8?B?w4RwZmVs?=")
  * encode("MIME-Q") had UTF-8 flag on
  Message-Id: <[email protected]>
! lib/Encode/MIME/Header.pm AUTHORS
  Two occurances of "croak ()" fixed as "croak qq()".
  Simon Cozens is added to AUTHORS as a result.
  Message-Id: <[email protected]>
! bin/piconv
  POD fixes that reflect enhancements by jhi
! bin/piconv
  Two enhancements by jhi.
  + Now uses Getopt::Long so it accepts long name options 
   (--from for -f, for example)
  + New option: -r,--resolve
  Message-Id: <[email protected]>
! MANIFEST META.yml
  META.yml added upon request of Schwern
  Message-Id: <[email protected]>
! AUTHORS
  Enache Adrian removed upon request -- to live longer than Encode
  and/or FreeBSD (toy-)?thread :)
  Message-Id: <[email protected]>
! t/enc_module.t
  "close STDOUT unless $^O eq 'freebsd';" once again relocated
  to keep VMS happy in which case "$^O eq 'freebsd'" is required
  to keep FreeBSD+thread happy.  Sigh.
  Message-Id: <[email protected]>

1.93 2003/04/24 17:43:16
! t/enc_eucjp.t
  added "no warnings 'pack'" in for loop to keep bleedperl from
  complaining "Character in 'C' format wrapped in pack".
! Makefile.PL
  More elegant perl core detection inspired by Ilya Zakharevich
  (but further elaborated for general cases).
! lib/Encode/Encoding.pm lib/Encode/PerlIO.pod
  POD fixes.
! t/euc-jp.ucm
  like cp9??, \x80-\x9F (control + 0x80) are zapped so they
  are less likely to be confused w/ ISO-8859-*
! t/CJKT.t
  RT tests added (vendor encodings are exemplified) 
  -- that successfully found a flaw on iso-2022-kr before the patch.
! lib/Encode/CJKConstants.pm lib/Encode/KR/2022_KR.pm
  decode("ISO-2022-KR") has been buggy but no one ever sited
  that since no one seems to be using it.  Bugs discovered by
  SADAHIRO-san
  Message-Id: <[email protected]>
! lib/Encode/CN/HZ.pm t/perlio.t
  HZ is now perlio_ok, thanks to SADAHIRO-san. perlio.t modified
  so it adds test for HZ.
  Message-Id: <[email protected]>
! lib/Encode/Guess.pm
  Now guesses UTF-(16|32)(BE|LE) when the string contains \x00.
  So long as the string contains \x{00}-\x{ff} it does not fail.
  See perldoc for details.
  Message-Id: <[email protected]>

1.92 2003/03/31 03:27:27
! ucm/big5-eten.ucm ucm/big5-hkscs.ucm
  Extraneous single-byte chars in range \x80-\xA0 and \xFA-\xFF
  removed.  FYI, IBM's ICU has none of these for java-Big5-1.3_P.ucm
  but glibc-BIG5-2.1.2.ucm does.
  Message-Id: <[email protected]>
! ucm/cp932.ucm ucm/cp936.ucm ucm/cp949.ucm ucm/cp950.ucm
  Maps regenerated again but this time based upon
  http://oss.software.ibm.com/cvs/icu/charset/data/ucm/
  (But where is THE DOCUMENT by MICROSOFT?)
! t/enc_module.t AUTHORS
  failure with threaded Perl on FreeBSD addressed.
  Enache Adrian <[email protected]> is added to AUTHORS for this.
  Message-Id: <[email protected]>  
! lib/Encode/Guess.pm
  Some POD fixes.
! t/CJKT.t
  Change 18989: Make the :bytes conditional on PerlIO.
  further Modified by Dan Kogai
  <[email protected]>
! t/enc_module.t
  Chnage 18966: another fix for failing test on windows ("use encoding" 
  puts  STDIN in :raw mode, so chomp() wasn't stripping the CR), by gsar
  Message-Id: <[email protected]>
! t/CJKT.t
  Change 18970: Hopefully this works also in Win32, by jhi
  Message-Id: <[email protected]>
  Change 18965: fix CJKT.t failures on windows due to incorrect 
  binmode(), by gsar
  Message-Id: <[email protected]>

1.91 2003/03/09 20:07:37
! encoding.pm
  even more proofread by jhi.
  Message-Id: <[email protected]>
! t/enc_module.t
  -use lib 't';
  +use lib qw(t ext/Encode/t ../ext/Encode/t);
  Message-Id: <[email protected]>
! AUTHORS
  s/Hirohito/Hiroto/ig;  Sorry, Hiroto-san.
  Message-Id: <[email protected]>
! encoding.pm
  s/logner/longer/
  Message-Id: <[email protected]>

1.90 2003/03/09 17:32:43
! encoding.pm
+ t/enc_data.t
  Inaba-san has added a patch for perl 5.8.1 or later that makes
  encoding.pm work for <DATA> filehandle.  t/enc_data.t is to test
  that.  POD is further revised.
  Message-Id: <[email protected]>
! encoding.pm t/enc_module.t
  encoding vs. ${^UNICODE} resolved.  POD revised accordingly.
  Message-Id: <[email protected]>

1.89 2003/02/28
! Encode.xs
  signed vs. unsigned issue discovered by Craig on OpenVM
  Message-Id: <a05200f12ba81fe9d6298@[172.16.52.1]>
! encoding.pm AUTHORS
+ t/Mod_EUCJP.pm t/enc_module.enc t/enc_module.t
  Because  binmode() stacks layers instead of overwrite, you have to 
  ":raw :encoding()" in encoding.pm or your are in trouble when you 
  call encoding.pm multiple times.  There are several workarounds
  but Inaba-san's idea is in. SUGAWARA Hajime <[email protected]>, 
  who was the first to address this problem was added to AUTHORS.
  The test suites was added for this, which is a modified version
  of SUGAWARA-san's scripts
  Message-Id: <[email protected]>

1.88 2003/02/20 14:42:34
! Encode.xs
  one signedness nit for Encode by jhi
  <[email protected]>
! ucm/viscii.ucm
  VISCII map was incorrect; fixed by Sadahiro-san
  Message-Id: <[email protected]>
! t/enc_eucjp.t t/enc_utf8.t AUTHORS
  You can't unlink files that are opened in cygwin but the last
  file handle opened in t/enc_*.t left open.  Patch submitted
  by Yitzchak and he was added to AUTHORS.
  Message-Id: <[email protected]>
! t/CJKT.t
  now works with 'LC_ALL=en_US.UTF-8 PERL_UTF8_LOCALE=1'
  Message-Id: <[email protected]>  
! Unicode/Unicode.xs
  For 1.88: Unicode.xs =~ s/regog/recog/ -- jhi
  Message-Id: <[email protected]>

1.87 2003/02/06 01:52:11
! AUTHORS
  * Inaba "Sensei" Hirohito added (I thought I have done so a long 
  ago but apparently I did not).
  * SUZUKI Norio added for verious and useful bug reports.
! Byte/Byte.pm KR/KR.pm Unicode/Unicode.pm
  lib/Encode/Encoder.pm lib/Encode/CJKConstants.pm
  podchecked so all warnings are gone except for L<http://>.
! encoding.pm t/enc_eucjp.t
  * t/uni/tr_utf8.t now t ok on maintperl (sorry, jhi)
  * Filter option overhaul
  * POD revision           
! Encode.pm Encode.xs encengine.c Encode/encode.h 
  lib/Encode/Encoding.pm lib/Encode/JP/JIS7.pm
  Merged inaba-san's patch that fixes "use encoding 'shiftjis'"
  without filter.  podchecked by Dan Kogai.
  Message-Id: <[email protected]>
! lib/Encode/Alias.pm
  decode('alias', $1) went wild because of local $_ in find_alias()
  the evil local $_ is eradicated but that changes find_alias()
  format for coderef aliasing.  See Encode::Alias for details
  Message-Id: <[email protected]>

1.86 2003/01/22 03:29:07
! encoding.pm
  * Don't forget to canonize when you attempt an exact match!
  Message-Id: <[email protected]>
  * ${^ENCODING} exception is off for $] > 5.008
  Message-Id: <[email protected]>
! t/enc_utf8.t
  $] check commented out so it runs on 5.8.0
  
1.85 2003/01/21 22:19:14
! encoding.pm
  ${^ENCODING} exception is now explicit rather than handled by regex.
+ t/enc_eucjp.t t/enc_utf8.t
  Test suite for the better "encoding" pragma support for bleedperl.
  On 5.8.0, they will just be skipped.

1.84 2003/01/10 12:00:16
! encoding.pm
  ${^ENCODING} is no longer set for utf so encoding is no longer fun :)
  (That is to prevent duplicate encoding first by IO then ${^ENCODING})
  Message-Id: <[email protected]>
! Unicode/Unicode.xs
  %_ fixes saves the resulting .so .05% smaller, by NC
  Message-Id: <[email protected]>
! Encode.pm
  Silence Encode on undef, by Andreas
  Message-Id: <[email protected]>
  Message-Id: <[email protected]>
! Unicode/Unicode.xs
  s/regognised/recognised/ .  British spelling left intact to pay
  respect to two British Nicks :)
  Message-Id: <[email protected]>

1.83 2002/11/18 17:28:49
! Encode.xs lib/Encode/JIS7.pm
  Even more patches from Inaba-san has been applied.  With this
  patch t/uni/tr_7jis.t and t/uni/t_utf8.t of bleedperl will work.
  Message-Id: <[email protected]>

1.82 2002/11/14 23:06:12
! Encode.xs
  Encode::utf8 (XS Version) assertion botch first found in Cygwin, 
  later found in perls w/ -Dusemymalloc was fixed by NC.
  Message-Id: <[email protected]>

1.81 2002/11/08 18:29:27
! Encode.pm Encode.xs
  Non-XS version of Encode::utf8 is back (with XS being default).
  Encode::predefine_encodings(0) to turn off XS.
  This is primarily to cope w/ Cygwin smoke but Sadahiro-san has
  found that it was Test::More causing the problem, not Encode.
  But I have already made it configurable so it may be useful in
  some rare cases....
  Message-Id: <[email protected]>, et al.
! bin/enc2xs
  The ingenious patch by Nicholas Clark that reduces shlib sizes by
  50% with no penalty and backward compatibility preserved, is in.
  Message-Id: <[email protected]>

1.80 2002/10/21 20:39:09