ChangeLog

2018-05-28: beta 3 release:
	* src: numbertext renamed to spellout
	* libnumbertext: fix cppcheck warnings by Julien Nabet
	* java: add -encoding utf8 to javac call by Rene Engelhard
	* LO extension: default capitalization, missing capitalization
	          was reported by TileStonePro
	* bg.sor: new Bulgarian module by Mihail Balabanov
	* cs.sor: fix for ordinal numbers, bug report by Stanislav Horáček
	* hu.sor: add number, letter, day and month affixation
	          for extended number, cross reference and date formats
	* is.sor: new Icelandic module by László Németh
	* it.sor: add missing groups in ordinal numbers,
	          the problem was reported by Valter Mura
	* lt.sor: fix typo, correct ordinal-number,
	          problem was reported by Modestas Rimkus
	* ru.sor: ordinal numbers and fixes by Mike Kaganski
	* sl.sor: ordinals, also genders and adverbials,
	          using also information by Martin Srebotnjak
	* sv.sor: fix typos
	* zh.sor: real native Simplified Chinese expressions by Suokunlong
	* *.xcu:
	  - Bulgarian translation by Mihail Balabanov
	  - Spanish translation by Adolfo Jayme Barrientos
	  - fix typo by Rene Engelhard
	* web:
	  - fix typo by Adolfo Jayme Barrientos
	  - add Bulgarian and Icelandic modules

2018-04-26: László Németh (nemeth numbertext org):
	* rename numbertext executable to spellout,
	  add default data path: /usr/share/numbertext
	  and support environmental variable NUMBERTEXTPATH
	* fix platform incompabilities
	* Greek cardinal and ordinal genders

2018-04-19: László Németh (nemeth numbertext org):
	* Soros programming language: revised draft:
	- conditional text, prefix macros, language
	  variant program lines
	* improved interpreters
	* C++11 port
	* language data:
	- More functions:
	  ordinal numbers (prefix: ordinal)
	  ordinal indicator (prefix: ordinal-number)
	  year (prefix: year)
	  gender variants and other functions
	- list all functions of a module with the input "help"
	  (in Calc: =NUMBERTEXT("help"), but enlarge the
	   row height to see every line of the help text)
	- new Malaysian and Swiss Standard German
	- compressed language modules:
	- instead of hu_HU_2, ja_JP_2, zh_ZH_2,
	  "formal" prefix allows formal number names here.
	  For example, MONEYTEXT(1200;"money HUF";"hu")
	  results "egyezerkettőszáz forint" instead of
	  "ezerkétszáz forint",
	  MONEYTEXT(123;"formal";"ja") results "参百弐拾壱"
	  instead of "三百二十一"
	- instead of en_US_2, use "money" prefix combined
	  with currencies, eg. MONEYTEXT(1.01;"money USD")
	  results "one and 1/100 U.S. dollar" in LibreOffice
	  instead of "one U.S. dollar and one cent".

2016-05-24: Liudas Ališauskas
	* Fix Lithuanian module

2014-05-10: Mihovil Stanić (mihovil.stanic gmail com):
	* new Croatian module

2013-03-16: Jerzy Moruś
	* Polish translation of LibO extension resource data

2012-12-06: László Németh (nemeth numbertext org):
        * src/Soros.py, soroside.py, __init__.py, test.py:
          Python 3 port (back compatible with Python 2)
        * reg.uno.py, bin/*.py: initial patches for tabulator
          inconsistency and terminating whitespaces in Python 3
          reported by Julien2412 at libreoffice-dev.

2012-10-09: László Németh (nemeth numbertext org):
	* data/lt_LT.sor: Lithuanian module (hints by Mantas Kriaučiūnas
	  and Laimonas Vėbra (laimis) at LaunchPad #936205)

	* data/fr_*: fix un/une + currency, based on the patch by
	  Jmzambon at OOO Wiki

	* data/es_ES.sor: euro: y dos cents -> con dos céntimos, reported by
	  mike8p at launchpad.net

	* data/pt_*.sor: ciento(s) -> cento(s), reported by Gwenaël Quenette
	  (sent by Zoltán Reizinger and Andrea Pescetti).

	* java/: sources of the Java jar distribution

2012-03-17: Asterisks at OOo Wiki:
	* data/lv_LV.sor: Latvian module

2010-11-03: László Németh (nemeth numbertext org):
	* data/vi_VN.sor: new Vietnamese module
	* data/ru_RU.sor:
	- fix for 2 (subunits of EUR, GBP, USD),
	- add currency UAH, reports and patch by wasow at extensions ooo

2010-08-03: László Németh (nemeth numbertext org):
	* pythonpath/org/Numbertext/__init__.py: fix rounding differences
	  between Calc and Numbertext functions: rounding by Calc round()
	  in Python code of Numbertext. The problem was reported by
	  Bernard Marcelly on Numbertext OOo Wiki page.
	
	* data/{el,en,es,fr,pt,tr}*.sor, : remove "zero cents" etc. from
	  Moneytext output, reported by Bernard Marcelly.

	* data/{af,sv,tr}*.sor: add missing zero to Moneytext (eg. "nul rand")

2010-08-03: Eko Prasetiyo (ekoprasetiyo at gmail):
	* data/id_ID.sor: Indonesian module

2010-06-07: László Németh (nemeth numbertext org):
	* data/ru_RU.sor: fix bad back references in Russian rules. Missing
	  conversion reported by Birioukoff at OOo Extension site.

2010-05-28: László Németh (nemeth numbertext org):
	* data/en_IN.sor: Indian numbering system. Lakh = 100.000, crore =
	  10.000.000 (requested by Dhandip at OOo Extensions site).

	* data/ca_ES.sor: fix Catalan MONEYTEXT in Calc Numbertext extension
	  (Launchpad bug 585782 reported by Bernard Marcelly).

	* data/en_US.sor, hu_HU.sor: fix 0th

	* data/sv_SE.sor: fix 0th, 100th, 1000th

	* data/fr_*.sor: fix ordinal numbers

	* doc/*odt: modifications: DollarText/BAHTTEXT field formatters of
	  MS Office and ECMA-376, Unicode CLDR, corrections for RBNF,
	  BSD license for example Python and Java interpreters

2010-05-20: László Németh (nemeth numbertext org):
	* data/sv_SE.sor: 
	- fix 30 (trettio)
	- reference implementation with grammar and year support,
	  also with gender parameters for currencies. Gender problem
	  with currencies reported by Kent Karlsson in CLDR ticket 2783

	* data/Roman.sor:
	- replaced with the more compact Roman_2.sor
	- replace "(" and ")" with "\(" and "\)" (parenthesized numerals)

	* data/Roman_2.sor:
	- variant with overline and bars
	- extra functions: historical with the special Unicode characters for
	  1000, 5000 and 10000

	* src/Soros.js:
	- fix character encoding for \(, \) etc.

2010-04-26: László Németh (nemeth numbertext org):
	* data/fr_*: ordf and ordm instead of ord-fem and ord-mas

	* data/ru_RU.sor: fix bugs reported by georgepyh on OOo Extension site

	* description.xml, CalcAddIn.xcu: add French translation by
	  Bernard Marcelly.

	* data/he_IL.sor: [^0] -> [1-9], modify function names for MONEYTEXT
	
	* src/soroside.py: rename soros.py, suggested by Alex Bodnaru

2010-04-26: Alex Bodnaru (alexbodn at 012 dot net dot il):
	* data/he_IL.sor:
	- add ordinal numbers (ord, ordf, ordm functions)
	- restructuring currencies, etc.

2010-04-22: László Németh (nemeth numbertext org):
	* data/sv_SE.sor: fix Swedish 11000 (elvatusen instead of elvausen)
	
	* data/{he_IL,pl_PL,ru_RU}.sor: replace the complementer bracket
	  expressions (eg. [^1]) to non-complementers (eg. [02-9]). The
	  problem with the Polish Moneytext reported by geo78 at OOo Extensions.

	* src/soros.py: extended syntax: range with steps by tilde (0-100~10)

	* data/fr_FR.sor: add ordinal numbers (ord, ord-fem, ord-mas):
	=MONEYTEXT(5;"ord";"fr") -> cinquième
	=MONEYTEXT(1;"ord-fem";"fr") -> première
	=MONEYTEXT(1;"ord-mas";"fr") -> premier

	* data/th_TH.sor: add ordinal numbers (ord)

2010-04-19: László Németh (nemeth numbertext org):
	* pythonpath/org/__init__.py:
	- remove Asian and CTL lang. detection temporarily (problems with
	  document language detection in OOo 3.0 reported by Petr Mladek and
	  mrw at OOo Extension site);
	
	- now MONEYTEXT handles unknown currency codes/new function names.
	
	  Example:
	
	  = MONEYTEXT(num, "ord", "en") -> ordinal names
	  = MONEYTEXT(num, "ord2", "en") -> ordinal abbreviations (1st, 2nd etc.)

	Problems with unknown currency codes reported by Bernard Marcelly.

	* src/Soros.{py,js}: fix broken $n|$ = $n||$ equivalence (it worked only
	  for parenthesized form: $(...)|$... = $(...)||$...)

	* src/Soros.py: fix output encoding of \$, \(, \), \|

	* data/{de_DE, hu_HU, hu_HU_2}.sor:
	- function ord for ordinal numbers
	- function help for help

	* data/{en_US, en_US_2}.sor:
	- function ord for ordinal numbers
	- function ord2 for ordinal abbreaviations (1st, 2nd, 3rd, 4th etc.)
	- function help for help

	* data/hu_HU_2.sor: formal Hungarian rules (eg. egyezeregyszáz)

	* data/ru_RU.sor: grammatical fixes for Russian money amounts
	  (problems reported by rogerclc on OOo extension site):
	– for numbers ending in 1 that are not 'teens':
	   (1, 21, 31, 41, 101, 121, etc.): roubles = рубль; kopeks = копейка
	– for all numbers ending in 1 in the kopeks column (.01, .21, etc):
	  one = одна, eg. 
	  21.00 = двадцать один *рубль*
	  21.01 = двадцать один *рубль одна* копейка
	  21.21 = двадцать один *рубль* двадцать *одна копейка*

	* data/th_TH.sor: fix number 1 in 201, 5001 etc. 
	  1 is เอ็ด in all numbers greater than 1 (หนึ่ง).
	  This is an old Excel bug revived in OOo, too (see OOo Issue 83490),
	  reported by Supapong at OOo Numbertext Wiki.

	* data/fr_*.sor: add XPF (franc Pacifique), missing monetary
	  unit reported by Bernard Marcelly.

	* src/places.py: add RON, problem with the new Romanian monetary unit
	  reported by Bernard Marcelly.

2010-04-13: Alex Bodnaru (alexbodn at 012 dot net dot il):
	* data/he_IL.sor: Hebrew module

	* Makefile(s): add better clean

2009-11-13: László Németh (nemeth numbertext org):
	* data/{af_ZA,it_IT,nl_NL}.sor: fix bad conversion of
	  1 subunit of currencies. The bug in the Italian module reported by
	  Nicolas Pagni (Nicolas dot Pagni at npagni dot it)

	* data/sv_SE.sor: the Swedish 7 is "sju", not *sjo,
	  reported by Stephan Nilsson (Stephan at Nilsson dot tf)

2009-11-11: Tantai Thanakanok (tantai osdev co th):
	* data/th_TH.sor: extended Thai decimal support for 1 and
	  more than 2 digits with currencies

2009-11-10: László Németh (nemeth numbertext org):
	* data/da_DK.sor: new Danish module

	* data/ca_ES.sor, es_ES.sor, fr_*.sor: 0 euro (without plural),
	  reported by Bernard Marcelly

	* data/sv_SE.sor: fix bad language code (*sv_SV.sor) and missing
	  plural form of krone. Reported by Bernard Marcelly.

	* data/fr_*.sor: "moins" for negative. Reported by Bernard
	  Marcelly.

	* data/tr_TR.sor: add space removing, fix large numbers

	* src/Soros.{py,js}: replace bad [^0] pattern with [1-9] in the
	  __numbertext__ directive implementation. Missing decimals leading by
	  zero reported by Bernard Marcelly.
	* doc/sorosspec.odt: replace bad [^0] pattern with [1-9].

2009-09-25: Susana Sotelo Docío (Roebek at OpenOffice.org Wiki):
	* data/es_ES.sor: gender fix for numbers containing the string "ientos"
	  (doscientos/as, quinientos/as, novecientos/as, etc)

2009-09-12: Ramdem at OpenOffice.org Wiki:
	* data/tr_TR.sor: new Turkish module

2009-09-06: Joan Montané:
	* data/ca_ES.sor: fixes
	- "1000-1999" must be spelled only "mil ...", it can't be spelled "un mil" (the same fixed French bug)
	- decimal units for EUR changed from "cent" to "cèntim", as it's usual and accepted
	- decimal units for USD changed form "cent" to "centau", as it's usual and accepted
	- millions are spelled as in French:
	  EUR 1000000 --> un milió d'euros (note the " de/d' " preposition")
	  EUR 1000001 --> un milió un euros (it lacks " de/d' " preposition).
	  EUR 1000000,xx --> un milió d'euros amb xx cèntims
	  EUR 1000001,xx --> un milió un euros amb xx cèntims
	                
	Improvements:
	- Added all the currencies replaced by EUR
	- Added all the actives code beginning by A (I'm working in B-Z)
	- Added some actives codes beginning by X

2009-09-04: László Németh (nemeth numbertext org):
	* data/en_US*: fix English decimal conversion by replacing bad [^0]
	  (matched decimal separators, too) with "[1-9]", reported by
	  Bernard Marcelly (bmarcelly at openoffice dot org).

2009-09-03: László Németh (nemeth numbertext org):
	* pythonpath/org/Numbertext/__init__.py: add short language name support:
	  now module names works without country code or with bad or non-available
	  country and variant codes, too. The missing support of short languages
	  was reported by Eduardo Moreno, too.

	* Makefile: target "ooo" for OOo extensions, first used in ooo-build
	  integration by Kálmán Szalai.
	
	* ca_ES.sor, es_ES.sor, pt_PT.sor, pt_BR.sor, ro_RO.sor: add gender
          support,
	- ca_ES and es_ES bug fixes, reported by Joan Montané
	- ro_RO bug fixes, reported by Lucian Adrian Grijincu

	* data/af_ZA.sor, nl_NL.sor: extended decimal support, reported by
	  Friedel Wolff (friedel at translate dot org dot za)

	* data/fr_FR.sor: bug fixes:

	- extended and fixed decimal support, reported by
	  Bernard Marcelly (bmarcelly at openoffice dot org)

	- using "et" instead of "y", reported by Bernard Marcelly and xhydrogene
	  at launchpad dot net (https://bugs.launchpad.net/numbertext/+bug/421151)

	- virgule instead of comma, reported by Bernand Marcelly and xhydrogene
	  at launchpad dot net: https://bugs.launchpad.net/numbertext/+bug/421148,
	  also patch by Olivier Ronez

	- cent and mille instead of un cent and un mille, reported by Bernand Marcelly and xhydrogene
	  at launchpad dot net: https://bugs.launchpad.net/numbertext/+bug/421164,
	  also patch by Olivier Ronez

	* data/fr_CH.sor: decimal point instead of comma, reported by
	  Bernard Marcelly (bmarcelly at openoffice dot org)

	* data/pl_PL.sor: add affix function and portable regexes, fix
	  regex problem reported by Marcin Miłkowski

2009-08-30: Michel Weimerskirch (michel at weimerskirch dot net):
	* data/lb_LU.sor: new Luxembourgish module

2009-08-29 Marcin Miłkowski (milek_pl at o2 dot pl):
	* data/pl_PL.sor: patch for Polish
	  https://bugs.launchpad.net/numbertext/+bug/421031

2009-08-28: Goran Rakić (grakic at devbase dot net):
	* CalcAddIn.xcu, description.xml: add Serbian localizations

	* data/sh_RS.sor, data/sr_RS: new Serbian modules

2009-08-27: Olivier Ronez:

	* data/fr_BE.sor, data/fr_CH: patches for new modules 

	* data/fr_FR.aff: other patches:
	- milliarde -> milliards (reported by Sophie Gautier and
	  Bernand Marcelly, too)
	- fix for 81

2009-08-27: Joan Montané (jmontane at openoffice dot org):

	- data/ca_ES.sor: fixes
	
	http://wiki.services.openoffice.org/wiki/Talk:NUMBERTEXT/MONEYTEXT_development#Some_fixes_on_Catalan_definition):
		- "amb" instead of "und"
		- "cèntim" instead of (euro)cent
		- "menys" for negative

2009-08-20: László Németh (nemeth numbertext org):
	* data/ko*, ja*, Hang.sor: Korean, Japanese and Old Hungarian modules.

	* data/en_US.*, hu_HU.sor: optimized by separator functions.

	* web/webroot/Hung.ttf: TrueType font with Old Hungarian numbers and
	  some letters for the native name of the script (ROVÁSÍRÁS)

	* web/*: new section (old number systems)

	* data/it_IT.sor: Using "virgola" for comma, reported by
	  Attilafdd at launchpad.net.

2009-08-10: László Németh (nemeth numbertext org):
	* pythonpath/org/Numbertext/__init__.py: add missing Numbertext module
	
	* bin/locales.py: fix registration of the first language module

	* data/af_ZA.py: fix helper functions (space handling)

2009-08-09: László Németh (nemeth numbertext org):
	* numbertext.py: new namespace (incompatible with previous versions):
	  ORG.NUMBERTEXT.NUMBERTEXT() and ORG.NUMBERTEXT.MONEYTEXT(),
	  suggested by Eike Rathke (er at OOo).

	* CalcAddIn.xcu: standard resource files for settings and
	  localizations (see Jan Holst Jensen's excellent Add-in example and
	  documentation: http://biochemfusion.com/doc/Calc_addin_howto.html)

	* CalcAddIn.xcu: add compatibility names: working XLS export-import
          (XLS portability bug reported by Mushtaque Asghar).
    
	* numbertext.py: now NUMBERTEXT() has no default rounding:

        =NUMBERTEXT(4.12345) -> four dot one two three four five
        =NUMBERTEXT(4.00) -> four
        =NUMBERTEXT("4.00") -> four dot zero zero

	* numbertext.py: MONEYTEXT() supports not integer decimal places:
	=MONEYTEXT(1.8, "MGA") -> one ariary and *four* iraimbilanja

	* src/places.py: add MGA, MRO with decimal place 0.7
	(eg. one ariary = 5 iraimbilanja, decimal places = 0.7 in places.py)

	* doc/numbertext.odt: NUMBERTEXT specification 1.1:
        - remove decimal place argument

	* doc/sorosspec.odt:
	- add number to Devanagari number example to the Soros specification;
	- add Python implementation

	* data/pt_PT.sor: Portuguese 
	- initial version from Eduardo Moreno (emoreno tokonhu com):
          "Nombres en portugues de los números, con ayuda de Flavia Viegas".
	- extension:
	  - simplification
	  - long scale support
	  - syntax of "e" based on Portuguese test data of PERL::Lingua Words2Nums
	  - extended with currency names of Portuguese speaking countries

	* data/pt_BR.sor: Brazilian Portuguese
	- based on pt-PT, but Brazilian Portuguese uses short scale numbers
	  and different ortography (number 16, 17, 19).

	* data/af_ZA.sor: Africaans
	- the orthography based on Bruce C. Donaldson: A grammar of Africaans,
	  MOUTON DE GRUYTER, 1993
	  
	* data/zh_ZH_2.sor: Simplified financial Chinese module.

	* data/fr_FR.sor, data/en_US*.sor, data/es_ES.sor: 
	
	- extended currency names for English, French and Spanish modules

	* data/*: - fixes for language data:

	- Romanian, Swedish, Slovenian, Czech, etc.: fix decimals (.0n)

        - Greek: removing unnecessary "de"

        - Spanish: 21 (veintidós) -> 22, negativa - -> menos,
          reported by Eduardo Moreno (emoreno tokonhu com).

	- Italian: unucento -> cento, reported also by Claudio Colombo.

	- Esperanto: unudek -> dek, unucent -> cent, miliono -> unu miliono.

	* tests/test0*: test number 1-1000 in all languages (see make check).

	* web/*: fix for IE, etc.

2009-07-25: László Németh (nemeth numbertext org):
	* data/*.sor: 11 new languages (see README)
	- syntax and orthography based mostly on Wikipedia articles
	- for Greek numerals, see Harry Foundalis' excellent introduction
	  (http://www.foundalis.com/lan/grknum.htm)
	
	* numbertext.py: handling language variants (eg. en-US-2
	  for USD 20.25 -> twenty and 25/100 dollars)

	* doc/sorosspec.odt: improved Soros language specification.

	* Soros.js, pythonpath/Soros.py: standard Soros modules
	  in Python and JavaScript

	* bin/*.py: new conversion utilities and more readable format using
	  standard Python doc string of the numbertext language modules.

	* numbertext.py: fix for language codes without country ("eo", "Roman")

	* data/*.sor: add function-like currency handling for language modules

	* es_ES.sor: fix for Spanish "... y uno mil" -> "... y un mil".
	  The bug reported by Eduardo Moreno (emoreno tokonhu com).

	* Roman*.sor: handle arbitrary large Roman numbers: 1000000000 -> ((M))

	* test/*: testing environment and tests

	* web/: web pages of http://numbertext.org
	
	* bin/shellhtml: preprocessor tool for HTML, XML data