kmx git

New in 1.18:
* Many more transliterations, in particular also of Emoji characters.
* The iconv_open function is now POSIX:2024 compliant: it recognizes a
  suffix //NON_IDENTICAL_DISCARD in the 'tocode' argument, with the effect
  that characters that cannot be represented in the target character set
  will be silently discarded. Whereas the suffix //IGNORE in the 'tocode'
  argument has the effect of discarding not only characters that cannot be
  represented in the target character set, but also invalid multibyte
  sequences in the input.
  Accordingly, the iconvctl function accepts requests
    ICONV_GET_DISCARD_INVALID, ICONV_SET_DISCARD_INVALID,
    ICONV_GET_DISCARD_NON_IDENTICAL, ICONV_SET_DISCARD_NON_IDENTICAL.
* The iconv_open function and the iconv program now support multiple suffixes,
  such as //TRANSLIT//IGNORE, not only one.
* GB18030 is now an alias for GB18030:2005. A new converter for GB18030:2022
  is added. Since this encoding merely cleans up a few private-use-area
  mappings, you can continue to use the GB18030 converter, for backward
  compatibility. Its Unicode to GB18030 conversion direction has been
  enhanced, to help transitioning away from PUA code points.
* When converting from/to an EBCDIC encoding, a non-standard way of
  converting newlines can be requested
    - at the C level, by calling iconvctl with argument ICONV_SET_FROM_SURFACE
      or ICONV_SET_TO_SURFACE, or
    - from the iconv program, by setting the environment variable
      ICONV_EBCDIC_ZOS_UNIX to a non-empty value.
* Special support for z/OS: The iconv program adds a charset metadata tag to
  its output file. (Contributed by Mike Fulton.)
* For conversions from UCS-2, UCS-4, UTF-16, UTF-32, invoking
  iconv(cd,NULL,NULL,...) now preserves the byte order state.

New in 1.17:
* The libiconv library is now licensed under the LGPL version 2.1, instead of
  the LGPL version 2.0. The iconv program continues to be licensed under GPL
  version 3.
* Added converters for many single-byte EBCDIC encodings:
  IBM-{037,273,277,278,280,282,284,285,297,423,424,425,500,838,870,871,875},
  IBM-{880,905,924,1025,1026,1047,1097,1112,1122,1123,1130,1132,1137,1140},
  IBM-{1141,1142,1143,1144,1145,1146,1147,1148,1149,1153,1154,1155,1156,1157},
  IBM-{1158,1160,1164,1165,1166,4971,12712,16804}.
  They are available through the configure option '--enable-extra-encodings'.

New in 1.16:
* The preloadable library has been removed.

New in 1.15:
* The UTF-8 converter now rejects surrogates and out-of-range code points.
* Added ISO-2022-JP-MS converter.
* Updated the CP1255 converter to map one more character.
* The functions now support strings longer than 2 GB.

New in 1.14:
* The 'iconv' program now produces its output as soon as it can. It no longer
  unnecessarily waits for more input.
* Updated the GB18030 converter to map 25 characters to code points that have
  been to Unicode since 2000, rather than to code points in the Private Use
  Area.
* Updated the BIG5-HKSCS converter. The old BIG5-HKSCS converter is renamed to
  BIG5-HKSCS:2004. A new converter BIG5-HKSCS:2008 is added. BIG5-HKSCS is now
  an alias for BIG5-HKSCS:2008.
* Fixed a bug in the conversion to wchar_t.
* Fixed a small bug in the CP1258 converter.

New in 1.13:
* The library and the iconv program now understand platform dependent aliases,
  for better compatibility with the platform's own iconv_open function.
  Examples: "646" on Solaris, "iso88591" on HP-UX, "IBM-1252" on AIX.
* For stateful encodings, when the input ends with a shift sequence followed
  by invalid input, the iconv function now increments the input pointer past
  the shift sequence before returning (size_t)(-1) with errno = EILSEQ. This
  is also like GNU libc's iconv() behaves.
* The library exports a new function iconv_open_into() that stores the
  conversion descriptor in pre-allocated memory, rather than allocating fresh
  memory for it.
* Added CP1131 converter.

New in 1.12:
* The iconv program is now licensed under the GPL version 3, instead of the
  GPL version 2. The libiconv library continues to be licensed under LGPL.
* Added RK1048 converter.
* On AIX, an existing system libiconv no longer causes setlocale() to fail.
* Upgraded EUC-KR, JOHAB to include the Korean postal code sign.

New in 1.11:
* The iconv program has new options --unicode-subst, --byte-subst,
  --widechar-subst that allow to specify substitutions for characters that
  cannot be converted.
* The iconv program now understands long options:
    long option    equivalent to
    --from-code    -f
    --to-code      -t
    --list         -l
    --silent       -s
* The CP936 converter is now different from the GBK converter: it has changed
  to include the Euro sign and private area characters. CP936 is no longer an
  alias of GBK.
* Updated GB18030 converter to include all private area characters.
* Updated CP950 converter to include the Euro sign and private area characters.
* Updated CP949 converter to include private area characters.
* Updated the BIG5-HKSCS converter. The old BIG5-HKSCS converter is renamed to
  BIG5-HKSCS:1999 and updated to Unicode 4. New converters BIG5-HKSCS:2001 and
  BIG5-HKSCS:2004 are added. BIG5-HKSCS is now an alias for BIG5-HKSCS:2004.
* Added a few irreversible mappings to the CP932 converter.
* Tidy up the list of symbols exported from libiconv (assumes gcc >= 4.0).

New in 1.10:
* Added ISO-8859-11 converter.
* Updated the ISO-8859-7 converter.
* Added ATARIST converter, available through --enable-extra-encodings.
* Added BIG5-2003 converter (experimental), available through
  --enable-extra-encodings.
* Updated EUC-TW converter to include the Euro sign.
* The preloadable library has been renamed from libiconv_plug.so to
  preloadable_libiconv.so.
* Portability to mingw.

New in 1.9:
* Many more transliterations.
* New configuration option --enable-relocatable.  See the INSTALL.generic file
  for details.

New in 1.8:
* The iconv program has new options -l, -c, -s.
* The iconv program is internationalized.
* Added C99 converter.
* Added KOI8-T converter.
* New configuration option --enable-extra-encodings that enables a bunch of
  additional encodings; see the README for details.
* Updated the ISO-8859-16 converter.
* Upgraded BIG5-HKSCS, EUC-TW, ISO-2022-CN, ISO-2022-CN-EXT converters to
  Unicode 3.2.
* Upgraded EUC-KR, CP949, JOHAB converters to include the Euro sign.
* Changed the ARMSCII-8 converter.
* Extended the EUC-JP encoder so that YEN SIGN characters don't cause failures
  in Shift_JIS to EUC-JP conversion.
* The JAVA converter now handles characters outside the Unicode BMP correctly.
* Fixed a bug in the CP1255, CP1258, TCVN decoders: The base characters of
  combining characters could be dropped at the end of the conversion buffer.
* Fixed a bug in the transliteration that could lead to excessive memory
  allocations in libintl when transliteration was needed.
* Portability to BSD/OS and SCO 3.2.5.

New in 1.7:
* Added UTF-32, UTF-32BE, UTF-32LE converters.
* Changed CP1255, CP1258 and TCVN converters to handle combining characters.
* Changed EUC-JP, SHIFT_JIS, CP932, ISO-2022-JP, ISO-2022-JP-2, ISO-2022-JP-1
  converters to use fullwidth Yen sign instead of halfwidth Yen sign, and
  fullwidth tilde instead of halfwidth tilde.
* Upgraded EUC-TW, ISO-2022-CN, ISO-2022-CN-EXT converters to Unicode 3.1.
* Changed the GB18030 converter to not reject unassigned and private-use
  Unicode characters.
* Fixed a bug in the byte order mark treatment of the UCS-4 decoder.
* The manual pages are now distributed also in HTML format.

New in 1.6:
* The iconv program's -f and -t options are now optional.
* Many more transliterations.
* Added CP862 converter.
* Changed the GB18030 converter.
* Portability to DOS with DJGPP.

New in 1.5:
* Added an iconv(1) program.
* New locale dependent encodings "char", "wchar_t".
* Transliteration is now off by default. Use a //TRANSLIT suffix to enable it.
* The JOHAB encoding is documented again.
* Changed a few mappings in the CP950 converter.

New in 1.4:
* Added GB18030, BIG5HKSCS converters.
* Portability to OS/2 with emx+gcc.

New in 1.3:
* Added UCS-2BE, UCS-2LE, UCS-4BE, UCS-4LE converters.
* Fixed the definition of EILSEQ on SunOS4.
* Fixed a build problem on OSF/1.
* Support for building as a shared library on Woe32.

New in 1.2:
* Added UTF-16BE and UTF-16LE converters.
* Changed the UTF-16 encoder.
* Fixed the treatment of tab characters in the UTF-7 converter.
* Fixed an internal error when output buffer was not large enough.

New in 1.1:
* Added ISO-8859-16 converter.
* Added CP932 converter, a variant of SHIFT_JIS.
* Added CP949 converter, a variant of EUC-KR.
* Improved the ISO-2022-CN-EXT converter: It now covers the ISO-IR-165 range.
* Updated the ISO-8859-8 conversion table.
* The JOHAB encoding is deprecated and not documented any more.
* Fixed two build problems: 1. "make -n check" failed. 2. When libiconv was
  already installed, "make" failed.

New in 1.0:
* Added transliteration facilities.
* Added a test suite.
* Fixed the iconv(3) manual page and function: the return value was not
  described correctly.
* Fixed a bug in the CP1258 decoder: invalid bytes now yield EILSEQ instead of
  U+FFFD.
* Fixed a bug in the Georgian-PS encoder: accept U+00E6.
* Fixed a bug in the EUC-JP encoder: reject 0x8E5C and 0x8E7E.
* Fixed a bug in the KSC5601 and JOHAB converters: they recognized some Hangul
  characters at some invalid code positions.
* Fixed a bug in the EUC-TW decoder; it was severely broken.
* Fixed a bug in the CP950 converter: it recognized a dubious BIG5 range.

New in 0.3:
* Reduced the size of the tables needed for the JOHAB converter.
* Portability to Woe32.

New in 0.2:
* Added KOI8-RU, CP850, CP866, CP874, CP950, ISO-2022-CN-EXT, GBK and
  ISO-2022-JP-1 converters.
* Added MACINTOSH as an alias for MAC-ROMAN.
* Added ASMO-708 as an alias for ISO-8859-6.
* Added ELOT_928 as an alias for ISO-8859-7.
* Improved the EUC-TW converter: Treat CNS 11643 plane 3.
* Improved the ISO-2022-KR and EUC-KR converters: Hangul characters are
  decomposed into Jamo when needed.
* Improved the CP932 converter.
* Updated the CP1133, MULELAO-1 and ARMSCII-8 mappings.
* The EUC-JP and SHIFT_JIS converters now cover the user-defined range.
* Fixed a possible buffer overrun in the JOHAB converter.
* Fixed a bug in the UTF-7, ISO-2022-*, HZ decoders: a shift sequence a the
  end of the input no longer gives an error.
* The HZ encoder now always terminates its output in the ASCII state.
* Use a perfect hash table for looking up the aliases.

New in 0.1:
* Portability to Linux/glibc-2.0.x, Linux/libc5, OSF/1, FreeBSD.
* Fixed a bug in the EUC-JP decoder. Extended the ISO-2022-JP-2 converter.
* Made TIS-620 mapping consistent with glibc-2.1.
thodg/libiconv/NEWS

Commit

NEWS
