2000-05-15  Miguel de Icaza  <miguel@gnu.org>

	* Makefile.am: Generate the unicodeConf.sh file, include
	unicodeConf.sh.in file in the distribution.

	* configure.in (UNICODE_INCLUDEDIR, UNICODE_LIBDIR, UNICODE_LIBS):
	New variables, used to susbtitute in the unicodeConf.sh file

	* unicodeConf.sh.in: New file, template for generating
	unicodeConf.sh file. 

2000-04-26    <alves@hungry.com>

	* configure.in: no need for c++ here, removing check.

2000-02-11 Owen Taylor  <otaylor@redhat.com>

	* libunicode.spec.in: Add a .spec file

	* autogen.sh: Run libtoolize

	* ltmain.sh ltconfig config.guess config.sub: Removed
	from CVS

2000-02-10  Tom Tromey  <tromey@cygnus.com>

	* unicode.h (EILSEQ): Don't use EBADMSG.  Fixed comment.

2000-02-09  Tom Tromey  <tromey@cygnus.com>

	* sjis.c (sjis_write): Use empty brace pair, not semicolon, to
	avoid compiler warning.

	* convert.c (EILSEQ): Don't define.
	* unicode.h: Include <errno.h>
	(EILSEQ): Define if not defined.

2000-02-09  Dan Winship  <danw@helixcode.com>

	* convert.c (unicode_iconv): Change return type to ssize_t
	since it can return -1.
	* utf8.h (UTF8_COMPUTE): 0xFE and 0xFF are invalid.
	(UTF8_GET): break after setting result to -1 on error.

2000-01-07  Owen Taylor  <otaylor@redhat.com>

	* Makefile.am (AUTOMAKE_OPTIONS): Change AUTOMAKE_OPTIONS from
	gnits to gnu, since gnits prohibits the LGPL, at least in
	currently released versions of automake.

1999-11-07  Tom Tromey  <tromey@cygnus.com>

	* prop.c (unicode_iswide): Reject some characters that aren't
	wide.  From Markus Kuhn.

1999-11-06  Tor Lillqvist  <tml@iki.fi>

	* gen-formata.pl
	* sjis.c
	* jis/shiftjis.h
	* msft/cp932.h: Slight cosmetic cleanup, no changes in functionality.

1999-10-31  Tom Tromey  <tromey@cygnus.com>

	* unicode.h (unicode_iswide): Declare.
	* prop.c (unicode_iswide): New function.

1999-10-29  Tom Tromey  <tromey@cygnus.com>

	* iso8859.c (iso8859_write): Special case old mapping of 8859-7.
	* iso/iso8859-7.h: Remapped 0xa0 and 0xa1 per latest Unicode
	tables.

1999-10-25  Tom Tromey  <tromey@cygnus.com>

	* testsuite/Makefile.am (tinyiconv_SOURCES): New macro.
	(tinyiconv_LDADD): Likewise.
	* configure.in: Create msft/Makefile, jis/Makefile.

1999-10-24  Tor Lillqvist  <tml@iki.fi>

	* gen-formata.pl: New file. Generates tables for converting
	CJKV encodings with "Format A" style table text files available
	from www.unicode.org.

	* sjis.c: New file. Handles Shift-JIS encoding of JIS X 0208 (and
	the closely related DBCS encoding of Microsoft code page 932).
	* Makefile.am: Add it.

	* convert.h
	* init.c: Add the two new encodings.

	* jis/shiftjis.h
	* msft/cp932.h: New files, generated by gen-formata.

	* jis/Makefile.am
	* msft/Makefile.am: New files.

	* testsuite/tinyiconv.c: New file, for generic tests of encodings.
	* testsuite/Makefile.am: Add it.

1999-10-16  Tor Lillqvist  <tml@iki.fi>

	* configure.in: Check for langinfo.h. Add AC_LIBTOOL_WIN32_DLL to
	enable dll builds on Win32. Note: To actually build libunicode as
	a DLL on Win32 using the auto* and libtool mechanism, you will
	currently need some pretty hairy manual intervention, and still it
	probably won't work. I actually use hand-written makefiles
	anyway...

	* acconfig.h: Add HAVE_LANGINFO_H.

	* utf8.c: Guard inclusion of langinfo.h

	* testsuite/utf8.c: Don't include netinet/in.h just for one htonl,
	do it manually. (Good for system without netinet/in.h.)

1999-09-13  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* unicode.h, utf8.c: Add prototypes for and implement :
	  unicode_offset_to_index, unicode_index_to_offset,
	  unicode_strchr, unicode_strrchr, unicode_strncpy
	  unicode_pad_string

	* testsuite/utf8.c: Add more test cases for unicode_strncpy,
	  unicode_offset_to_index, unicode_index_to_offset,
	  unicode_strchr, unicode_strncpy, unicode_last_utf8, 
	  unicode_string_width.
	
1999-09-07  Tom Tromey  <tromey@cygnus.com>

	* COPYING.LIB: Added.
	* COPYING: Removed.

1999-09-02  Tom Tromey  <tromey@cygnus.com>

	* unicode.h: Document errno substitution.
	* convert.c (unicode_iconv): Use bcopy if memmove doesn't exist.
	(EILSEQ): Define if not defined.
	* configure.in: Check for memmove.

	* ucs2.c (get_one): Pull out correct byte according to
	endianness.
	(write_one): Likewise.
	(ucs2_native_init): Correctly initialize for big-endianness.

	* testsuite/utf8.c (test_split_utf8_results): Translate to
	UCS4-native.

	* ucs4.c (ucs4_native_init): Reverted previous change.
	(ubn): Use names UCS4 and UCS-4.
	(un): Use name UCS4-native only.

	* testsuite/utf8.c (test_split_utf8_results): Correctly check
	result of unicode_iconv().

	* testsuite/Makefile.am (ucs4_SOURCES): New macro.
	(ucs4_LDADD): Likewise.
	(std): Added ucs4.
	* testsuite/ucs4.c: New file.
	* ucs4.c (ucs4_read): Fixed big-endian code.
	(ucs4_write): Likewise.

	* unicode.h: Coding-standard cleanup.

	* testsuite/utf8.c (test_utf8_sequence): Use big-endian (network)
	byte order for UCS-4.

	* ucs4.c (ucs4_native_init): Always use big-endian.

1999-08-24  Tom Tromey  <tromey@cygnus.com>

	* Makefile.am (cvs-dist): New target.

1999-08-20  Tom Tromey  <tromey@cygnus.com>

	* utf8.c: Indentation and other pedantry.
	* unicode.h: Likewise.

1999-08-20  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* unicode.h, init.c (unicode_init): ANSIfied function.

	* unicode.h (unicode_get_charset, unicode_string_width): Added
	prototype.
	
	* utf8.c (unicode_get_charset): New function. Returns TRUE if
	we are in a UTF-8 locale, or if CHARSET is set to "UTF-8". Puts
	the charset into the parameter if non-NULL.

	(unicode_string_width): Returns the probable visual width of a string.
	
1999-08-18  Tom Tromey  <tromey@cygnus.com>

	* Makefile.am (BUILT_SOURCES): Removed.
	(unicode-config): Removed.

1999-08-18  Raja R Harinath  <harinath@cs.umn.edu>

	Include everything to fallback on for systems with native iconv.
	* Makefile.am (libunicode_la_SOURCES): Include all files
	unconditionally. 
	* init.c (unicode_init): Register all encodings unconditionally.
	
1999-08-18  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* testsuite/utf8.c (test_utf8_strlen): New test, to check if
 	unicode_strlen returns the correct value.
	(main): Call it.

	* utf8.c (unicode_strlen): Fix off by one error.
	
1999-08-18  James Blandy  <jimb@mule.m17n.org>

	* testsuite/utf8.c (test_split_utf8_results): New test, verifying
	that we handle incomplete encodings at the end of the input
	correctly.
	(main): Call it.

	Use enums for selectors, not #defines.
	* convert.h (struct unicode_iconv_i): Make `type' an enum, with
	the values unicode_iconv_type_native and unicode_iconv_type_ours.
	* convert.c (NATIVE, OURS): Deleted.
	(unicode_iconv_open, unicode_iconv_close, unicode_iconv): Use the
	unicode_iconv_type_ values.

	* convert.c: (list_encodings): Remove function, added by accident.

	* testsuite/utf8.c (test_utf8_results): New test function, to
	catch the bugs below.
	(main): Call it.
	* testsuite/Makefile.am: Doc fix.

	* utf8conv.c (utf8_read): Remember to advance inbuf and
	inbytesleft by len, not by one.

	* utf8.h (UTF8_GET): Don't forget to wrap all the arguments in
	parens, to avoid operator associativity problems.

	* convert.h (unicode_encoding_t): Add `outbuf' and `outbytesleft'
	arguments to the `reset' function, so it can actually produce some
	text.
	* convert.c (unicode_iconv): Pass those arguments.

1999-08-17  Tom Tromey  <tromey@cygnus.com>

	* testsuite/utf8.c: Use unicode_iconv_t.
	* testsuite/roundtrip.c: Use unicode_iconv_t.

	* unicode.h (unicode_iconv_t): New define; replaces old structure
	and include.
	* convert.h (unicode_iconv_i): Moved from unicode.h and renamed.
	(unicode_iconv_t): New structure.
	* convert.c (unicode_iconv_open): Use native iconv_open if
	available.
	(unicode_iconv_close): Use native iconv_close, if available.  Free
	buffer in `our' case.  Free descriptor.
	(unicode_iconv): Use native iconv, if available.
	(NATIVE): New macro.
	(OURS): Likewise.

1999-08-13  Raja R Harinath  <harinath@cs.umn.edu>

	* convert.c (unicode_iconv_open): Don't dereference pointer after
	freeing it.

1999-08-13  Tom Tromey  <tromey@cygnus.com>

	Efficiency change suggested by Jim Blandy:
	* utf8conv.c (utf8_read): New function.
	(utf8_write): Updated interface.
	Include "utf8.h".
	(unicode_utf8_encoding): Use utf8_read.
	(unicode_java_utf8_encoding): Likewise.
	* utf8.c (unicode_get_utf8): Rewrote.
	(unicode_utf8_read): Removed.
	Include "utf8.h".
	* Makefile.am (libunicode_la_SOURCES): Added utf8.h.
	* utf8.h: New file.
	* ucs4.c (ucs4_read): Updated interface.
	(ucs4_write): Likewise.
	* ucs2.c (ucs2_write): Updated interface.
	(ucs2_read): Likewise.
	* latin1.c (latin1_write): Updated interface.
	(latin1_read): Likewise.
	* iso8859.c (iso8859_read): Updated interface.
	(iso8859_write): Likewise.
	* convert.c (unicode_iconv): Rewrote to use new interface.
	(unicode_iconv_open): Allocate intermediate buffer.
	* unicode.h (iconv_s): Added `buffer', `valid', and `size'
	fields.
	* convert.h (unicode_read_result): New enum; removed READ_
	defines.
	(unicode_write_result): Likewise.
	(unicode_encoding_t): Changed interface to read and write.

1999-08-10  Tom Tromey  <tromey@cygnus.com>

	* ucs2.c (unicode_ucs2_native_encoding): Added aliases.
	* latin1.c (unicode_latin1_encoding): Added aliases.
	* iso8859.c (ISO_DEFINE): New macro.  Use it to define the
	standard ISO tables.
	(unicode_windows_1252_encoding): Added aliases.
	(P3): New macro.
	* convert.c (find_encoding): Search all names for an encoder.  Use
	case-insensitive compare.
	* convert.h (unicode_encoding_t): Replaced `name' with `names', an
	array of names.
	* Various files: Changed all users of unicode_encoding_t.

1999-08-10  Raja R Harinath  <harinath@cs.umn.edu>

	* testsuite/utf8.c: Include config.h.
	* testsuite/roundtrip.c: Likewise.

	* Makefile.am (libunicode_la_SOURCES) [NATIVE_ICONV]: 
	Add `init.c' for `unicode_init'.

1999-08-10  Tom Tromey  <tromey@cygnus.com>

	* Makefile.am (SUBDIRS): Put `.' at beginning.

	* testsuite/roundtrip.c (main): Fixed names of charsets to conform
	to libunicode.

	* utf8conv.c (utf8_write): Fixed direction of increment from `++'
	to `--'.  Only loop `len - 1' times.

	* testsuite/roundtrip.c (main): Call unicode_init().
	* testsuite/utf8.c (test_utf8_sequence): Test `i' against -1, not
	0.  Added \n to failure message.  Don't use network byte order for
	UCS4 characters.
	(main): Call unicode_init().

1999-08-10  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* testsuite/utf8.c: New test program.
	* testsuite/Makefile.am (std): added utf8.
	(utf8_SOURCES utf8_LDADD): New macros.

1999-08-09  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* testsuite/roundtrip.c: New test program.
	* testsuite/Makefile.am (std): Added roundtrip.
	(roundtrip_SOURCES cxxsmoke_LDADD): New macros.

	* iso8859.c: (windows_1252_encoding): Renamed 
	"WINDOWS-1252" to "CP1252", for compatibility with
	glibc.

	* testsuite/.cvsignore: Made CVS quieter.
	
1999-08-08  Tom Tromey  <tromey@cygnus.com>

	* testsuite/cxxsmoke.cc: New file.
	* testsuite/Makefile.am (std): Added cxxsmoke.
	(cxxsmoke_SOURCES): New macro.
	(cxxsmoke_LDADD): New macro.
	* configure.in: Added AC_PROG_CXX.

	* unicode.h: Updated prototypes for const-ification.
	* utf8.c (unicode_previous_utf8): Made arguments `const'.
	(unicode_next_utf8): Made argument `const'.
	(unicode_get_utf8): Likewise.
	(unicode_strlen): Removed unneeded cast.

1999-08-06  Raja R Harinath  <harinath@cs.umn.edu>

	* iso/Makefile.am (noinst_HEADERS): Add windows-1252.h.
	* Makefile.am (include_HEADERS): Move decomp.h and convert.h ...
	(noinst_HEADERS): ... here.

1999-08-06  Tom Tromey  <tromey@cygnus.com>

	* unicode.h: Added missing `{'.

	* iso8859.c (unicode_georgian_academy_encoding): Fixed typo in
	init function name.
	* convert.h (unicode_georgian_ps_encoding): Fixed typo.

1999-08-06  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* unicode.h, utf8.c: (unicode_strlen): Made const.

	* unicode.h: Made C++-safe.

1999-08-06  Pablo Saratxaga <srtxg@chanae.alphanet.ch>

	* added georgian encodings (this time whithout typos :) )

1999-08-04  Raja R Harinath  <harinath@cs.umn.edu>

	Fix typos.
	* iso/armscii-8.h (armscii_8_table): Was `koi8_u_table'. 
	* iso/Makefile.am (noinst_HEADERS): Missing .h extns.
	* iso8859.c (unicode_iso8859_10_encoding): 
	Was `unicode_iso8859_14_encoding'.

1999-08-04  Pablo Saratxaga <srtxg@chanae.alphanet.ch>

	* added encodings for iso-8859-10, koi8-u (used in Ukrainia),
	armscii-8 (armenian) and tis-620 (thai)

1999-08-02  Owen Taylor  <otaylor@redhat.com>

	* autogen.sh: Add an autogen.sh. (Simplified from the
	one in GTK+)

1999-08-02  Owen Taylor  <otaylor@redhat.com>

	* testsuite/Makefile.am: Unconditionalize a conditional
	to avoid bugs in automake-1.4a.

1999-07-29  Owen Taylor  <otaylor@redhat.com>

	* Makefile.am (libunicode_la_SOURCES): Add utf8.c for
	NATIVE_ICONV.

	* unicode.h: Added <stdlib.h> for size_t

	* configure.in: Use AM_PROG_LIBTOOL instead of AC_PROG_LIBTOOL,
	since AC_PROG_LIBTOOL doesn't appear in any released
	version of libtool.

	* unicode-config.in: Generate and install a GTK+-style
	  unicode-config for CFLAGS.

1999-07-28  Robert Brady  <rwb197@ecs.soton.ac.uk>

	* init.c, iso8859.c, convert.h, iso/windows-1252.h, iso/koi8-r.h:
        Added support for important legacy encodings Windows-1252
	and KOI8-R.

1999-07-28  Tom Tromey  <tromey@cygnus.com>

	* Makefile.am (AM_CFLAGS): New macro.
	* configure.in (cflags): New subst

	* utf8conv.c (utf8_write): Make `len' a `size_t' to avoid
	warning.  Removed unused variable.
	* utf8.c (unicode_utf8_read): Make `len' a `size_t' to avoid
	warning.
	* prop.c (unicode_istitle): Use unsigned to avoid warning.
	(unicode_toupper): Likewise.
	(unicode_tolower): Likewise.
	(unicode_totitle): Likewise.
	* latin1.c (latin1_write): Used `unsigned int' to avoid warning.
	* convert.c (unicode_iconv): Removed unused variable.

	* ucs2.c (unicode_ucs2_native_encoding): Reference
	ucs2_native_init.

	* utf8.c (unicode_previous_utf8): Use `char', not `unsigned
	char'.
	(unicode_next_utf8): Likewise.
	(unicode_strlen): Likewise.
	(unicode_get_utf8): Likewise.

1999-07-27  Tom Tromey  <tromey@cygnus.com>

	* unicode.h (unicode_strlen): Updated.
	* utf8.c (unicode_strlen): Changed interface.

1999-07-26  Raja R Harinath  <harinath@cs.umn.edu>

	* ucs2.c (write_one): Fix off-by-one error.
	(get_one): Likewise.
	* ucs4.c (ucs4_read): Likewise.
	(ucs4_write): Likewise.

Wed Jul 28 00:52:27 1999  Owen Taylor  <otaylor@redhat.com>

	* utf8.c (unicode_next_utf8): Fixed reversed
	check.

1999-07-11  Tom Tromey  <tromey@cygnus.com>

	* testsuite/ordering.c (check): New funtion.
	(check_decomp): Likewise.
	(main): Use them.
	(t2_src): New array.
	(t3_src): Likewise.
	(t4_src): Likewise.
	(t5_src): Likewise.
	(t6_src): Likewise.
	(t5_dst): Likewise.
	(t7_dst): Likewise.
	(t8_src): Likewise.
	(t7_src): Likewise.

	* decomp.h: Rebuilt.
	* gen-table.pl (process_one): Note decompositions.
	(fetch_cclass): Renamed.
	(expand_decomp): New sub.
	(print_decomp): Print decomposition table.
	* unicode.h (unicode_canonical_decomposition): Declare.
	* decomp.c (asize): New macro.
	(unicode_canonical_decomposition): Wrote.

	* testsuite/ordering.c: New file.
	* configure.in: Create iso/Makefile and testsuite/Makefile.
	* testsuite/Makefile.am: New file.
	* iso/Makefile.am: New file.
	* Makefile.am (SUBDIRS): New macro.

	* unicode.h (unicode_canonical_ordering): Declare.
	* Makefile.am (libunicode_la_SOURCES): Added decomp.c.
	* decomp.h: New file.
	* chartables.h: Rebuilt.
	* gen-table.pl (fetch_type): New sub.
	(fetch_attr): Likewise.
	(mappings): Added UNICODE_ prefix to each value.
	(print_row): Rewrote.
	(print_attr_row): Removed.
	(print_tables): Changed to use new print_row.
	(print_decomp): New sub.
	(fetch_cclas): New sub.
	(process_one): Note combining class of character.
	* decomp.c: New file.

	* iso8859.c (iso8859_write): Mask off high bits for clarity.

	* iso8859.c (iso8859_read): Special-case digits in 8859-6.
	(iso8859_write): Likewise.  Also, initialize `xlate'.

	* Makefile.am (libunicode_la_SOURCES): Added iso8859.c.
	* init.c (unicode_init): Register 8859-* encoders.
	* convert.h: Declare 8859-* encoding structures.
	* iso/iso8859-14.h, iso/iso8859-15.h, iso/iso8859-2.h,
	iso/iso8859-3.h, iso/iso8859-4.h, iso/iso8859-5.h,
	iso/iso8859-6.h, iso/iso8859-7.h, iso/iso8859-8.h,
	iso/iso8859-9.h: New files.
	* iso8859.c: New file.
	* gen-iso.pl: New file.

	* init.c (unicode_init): Register unicode_ucs4_native_encoding.
	* convert.h (unicode_ucs4_native_encoding): Declare.
	* ucs4.c (ucs4_native_init): New function.
	(unicode_ucs4_native_encoding): New struct.

	* Makefile.am (libunicode_la_SOURCES): Removed ascii.c.
	* ascii.c: Deleted.
	* latin1.c (unicode_ascii_encoding): Moved from ascii.c.  Set
	`init' element.
	(latin1_init): New function.
	(ascii_init): Likewise.
	(latin1_write): Use stored mask, not constant.
	(unicode_latin1_encoding): Set `init' element.

	* acconfig.h: New file.

	* utf8.c (unicode_get_utf8): Don't bother to compute length
	explicitly.

	* ucs2.c: Removed erroneous comment.

	* ucs2.c (ucs2_read): Handle surrogate pairs.
	(get_one): New function.
	(write_one): Likewise.
	(ucs2_write): Write surrogate pairs.

	* chartables.h: Rebuilt.
	* gen-table.pl (%gaps): New global.
	(CJK, HANGUL, SURROGATES, PRIVATE_USE): Removed.

	* ucs2.c (HIGH_SURROGATE): New macro.
	(LOW_SURROGATE): Likewise.
	(LAST_SURROGATE): Likewise.

1999-07-10  Tom Tromey  <tromey@cygnus.com>

	* init.c (unicode_init): Initialize unicode_ucs2_native_encoding.
	* convert.h (unicode_ucs2_native_encoding): Declare.
	* ucs2.c (ucs2_native_init): New function.
	(unicode_ucs2_native_encoding): New structure.

1999-07-09  Tom Tromey  <tromey@cygnus.com>

	* acconfig.h: New file.
	* configure.in: Check for existence of native iconv.

	* init.c (unicode_init): Conditionalize on
	UNICODE_USE_SYSTEM_ICONV.

1999-07-08  Tom Tromey  <tromey@cygnus.com>

	* init.c (unicode_init): Register unicode_java_utf8_encoding.
	* convert.h (unicode_java_utf8_encoding): Declare.
	* utf8.c (unicode_utf8_read): Handle 4, 5, and 6 byte characters.
	* utf8conv.c (utf8_std_init): New function
	(unicode_utf8_encoding): Use it.
	(utf8_java_init): Likewise.
	(unicode_java_utf8_encoding): New global.
	(utf8_write): Handle Java and standard encodings.  Handle 4, 5,
	and 6 byte characters.

	* configure.in: Version 0.2.

	* prop.c (TTYPE): New macro.
	(TYPE): Use it.
	All functions and macros: Use UNICODE_* name for types.
	(ATTTABLE): New macro.
	(unicode_toupper): Use it.
	(unicode_totitle): Likewise.
	(unicode_tolower): Likewise.
	(unicode_digit_value): Likewise.
	(unicode_xdigit_value): Likewise.

	* chartables.h: Rebuilt.
	* gen-table.pl (print_attr_row): New sub.
	(print_tables): Use it.
	(print_row): Start loop at $start, not 0.

	* utf8.c (unicode_utf8_read): Moved back from utf8conv.c.  No
	longer static.
	* utf8conv.c (unicode_utf8_len): Removed.
	(unicode_utf8_encoding): Mention unicode_utf8_read.

	* unicode.h: Declare unicode_previous_utf8.
	* utf8.c (unicode_previous_utf8): New function.
	(unicode_get_utf8): Rewrote.  Changed interface.
	(unicode_next_utf8): Rewrote.

	* chartables.h: Rebuilt.
	* gen-table.pl (print_tables): Don't print character
	classification constants.
	* unicode.h: Define UNICODE_* character classification constants.

	* Makefile.am (libunicode_la_SOURCES): Added utf8conv.c.
	* utf8.c: Moved converter functions to utf8conv.c.
	* utf8conv.c: New file.

	* prop.c (unicode_xdigit_value): New function.
	(unicode_type): Likewise
	* unicode.h (unicode_xdigit_value): Declare.
	(unicode_type): Likewise.
	Also, add comments describing functions.

1999-07-06  Tom Tromey  <tromey@cygnus.com>

	* gen-table.pl (print_row): New sub.
	(print_tables): Use it.

1999-07-04  Tom Tromey  <tromey@cygnus.com>

	* utf8.c (unicode_utf8_len): Removed.
	(unicode_to_utf8): Removed.
	(unicode_get_utf8_and_advance): Removed.
	(utf8_write): New function.
	(utf8_read): New function.
	(unicode_get_utf8): Rewrote.

	* init.c (unicode_init): Register UCS2 encoders.
	* convert.h (unicode_ucs2_big_encoding,
	unicode_ucs2_little_encoding): Declare.
	* Makefile.am (libunicode_la_SOURCES): Added ucs2.c.
	* ucs2.c: New file.

	* Makefile.am (libunicode_la_SOURCES): Added ucs4.c.
	* init.c (unicode_init): Register UCS4 encoders.
	* convert.h (unicode_ucs4_big_encoding,
	unicode_ucs4_little_encoding): Declare.
	* ucs4.c: New file.

	* Makefile.am (libunicode_la_SOURCES): Conditionalize on
	NATIVE_ICONV.
	* configure.in (NATIVE_ICONV): New conditional.

	* latin1.c: Rewrote.
	* unicode.h (iconv_s, unicode_iconv_open, unicode_iconv_close,
	unicode_iconv): Declare.
	* ascii.c: Rewrote.
	* convert.c: Include convert.h.
	(find_encoding): Renamed; now static.
	(unicode_buffer_to_encoding): Deleted.
	(unicode_buffer_from_encoding): Deleted.
	(unicode_iconv_open): New function.
	(unicode_iconv_close): Likewise.
	(unicode_iconv): Likewise.
	* init.c: Include convert.h.
	* convert.h: New file.

1999-06-30  Tom Tromey  <tromey@cygnus.com>

	* chartables.h: Rebuilt.
	* gen-table.pl (process_one): Handle titlecase.
	(print_tables): Likewise.

	* prop.c (ISDIGIT, ISALPHA, asize): New macros.
	(unicode_isalnum, unicode_isalpha, unicode_iscntrl,
	unicode_isdigit, unicode_isgraph, unicode_isprint,
	unicode_ispunct, unicode_isxdigit): New functions.
	(unicode_toupper): Rewrote.
	(unicode_tolower): Likewise.
	(unicode_totitle): Likewise.
	(unicode_istitle): Likewise.

1999-06-27  Tom Tromey  <tromey@cygnus.com>

	* Started.