2007-01-27  Bill Poser  <poser@khunekcho.khunek.poser.org>
	
	* uninum.c:  (2.0.3)
	Fixed bug that prevents "all" and "any" number system names
	from being recognized.

2007-01-24  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: (2.0.2)
	Defined wcpcpy() internally instead of relying on glibc
	so as to avoid portability problems.
	Retracted replacement of unicode.h as too problematic.
	
	* UTF8in.c: (2.0.1)
	Replaced include of "unicode.h" with <uninum/uninum.h>
	so as to eliminate the need for unicode.h

2007-01-23  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: 
	Protected the Tcl command UNStrToWNStr against
	crashing when given an invalid number system name
	as argument.

	Added the function Tell_Library_Version which returns
	the library version number.

	* FindTclLibDir.tcl:
	Added script for finding out where to install Tcl library.


2007-01-22  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: 
	Added inverse functions, for generating strings.
	
	Eliminated need to include unicode.h by moving
	the necessary definitions into uninum.h.


2007-01-14  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.i: 
	Added Tcl API.

2007-01-05  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: 
	Modified Chinese to handle mixed Chinese Western numbers of the
	sort not infrequently found in Japan (e.g. 2500 man).

2007-01-04  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* numconv.c: 
	Added -i option which identifies number system.

	* uninum.c: 
	Added Cyrillic alphabetic.
	Added Glagolitic alphabetic.
	Added Sinhala in accordance with Unicode proposal.
	Extended Chinese to handle traditional units through 10^48.

2007-01-01  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: 
	Bugfix: removed declaration of ArabicAlphabeticToInt as static.
	
	Added Armenian alphabetic.
	Added characters to NormalizeChineseNumbers.
		

2006-12-31  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c: 
	Release 1.1.
	
	Added a third return type. For those who wish to do further
	arbitrary precision computation using the results of the
	conversion, the result may now be requested as an object
	of type mpz_t.

	Fixed bug in Chinese.

	Added support for Greek Alphabetic (only through 999).
	Added support for Arabic Alphabetic.
	Added WesternGeneralToInt.
	Added type NS_HEX.
	
	Made system-specific routines return a little earlier on
	  detecting an invalid character.

2006-12-30  Bill Poser  <poser@khunekcho.khunek.poser.org>

	* uninum.c:
	Added support for Limbu.
	Added support for New Tai Lue.

	GuessNumberSystem now tests for the case in which the string
	consists entirely of U+0030 and if it detects it returns
	the new value NS_ALLZERO. A string containing all zeroes
	has two special properties. One is that it can be assumed
	to have the value 0. The other is that it is ambiguous
	as to number system since some non-Western systems use
	the Western zero character. The number system specifier
	returned is distinct from NS_UNKNOWN both because the
	value of the string is determinate and because the number
	system can be determinate from additional data if it is
	available.

2006-12-28  Bill Poser  <poser@khunekcho.khunek.poser.org>
	*Release of version 1.0.

	* uninum.c: 
	Now do internal calculations using bignums and return results
	as decimal ascii strings. This allows lossless conversion regardless
	of the machine word sizes available on the machine.
	
	Added 0x0030 to NormalizeChinese since it may be used.
	
	Added Mongolian.
	Added Balinese.
	Added Khmer.
		

