Module:DateI18n: Difference between revisions

Content deleted Content added
Reverted to revision 879901483 by Ferret (talk): Restore last working version (TW)
Undid revision 954131094 by RexxS (talk) errors should be fixed (sandboxes working) - try new version again
Line 1:
--[[
__ __ _ _ ____ _ ___ _ ___
| \/ | ___ __| |_ _| | ___ _| _ \ __ _| |_ ___|_ _/ |( _ ) _ __
| |\/| |/ _ \ / _` | | | | |/ _ (_) | | |/ _` | __/ _ \| || |/ _ \| '_ \
| | | | (_) | (_| | |_| | | __/_| |_| | (_| | || __/| || | (_) | | | |
|_| |_|\___/ \__,_|\__,_|_|\___(_)____/ \__,_|\__\___|___|_|\___/|_| |_|
This module is intended for processing of date strings.
 
Please do not modify this code without applying the changes first at Module:Date/sandbox and testing
Copied from https://commons.wikimedia.org/w/index.php?title=Module:Date&oldid=224728211
at Module:Date/sandbox/testcases and Module talk:Date/sandbox/testcases.
Used by Module:ISOdate, Module:Complex date, Module:WikidataIB
 
Authors and maintainers:
* User:Parent5446 - original version of the function mimicking template:ISOdate
* User:Jarekt - original version of the functions mimicking template:Date and template:ISOyear
 
]]
require('Module:No globals')
 
-- ==================================================
-- === Internal functions ===========================
-- ==================================================
 
-- Function allowing for consistent treatment of boolean-like wikitext input.
local p = {}
-- It works similarly to Module:Yesno
 
local function yesno(val, default)
-- =======================================
if type(val) == 'boolean' then
-- === Dependencies ======================
return val
-- =======================================
elseif type(val) == 'number' then
local i18n = require('Module:I18n/date') -- get localized translations of date formats
if val==1 then
local yesno = require('Module:Yesno')
return true
 
elseif val==0 then
local function langSwitch(list,lang)
return false
local langList = mw.language.getFallbacksFor(lang)
table.insert(langList,1,lang)
for i,language in ipairs(langList) do
if list[language] then
return list[language]
end
elseif type(val) == 'string' then
end
val = mw.ustring.lower(val) -- put in lower case
if val == 'no' or val == 'n' or val == 'false' or tonumber(val) == 0 then
return false
elseif val == 'yes' or val == 'y' or val == 'true' or tonumber(val) == 1 then
return true
end
end
return default
end
 
---------------------------------------------------------------------------------------
--[[
-- String replacement that ignores part of the string in "..."
Date
local function strReplace(String, old, new)
if String:find('"') then
This function can be used to provide an ISOdate template.
local T={}
for i, str in ipairs(mw.text.split( String, '"', true )) do
Usage:
if i%2==1 then
{{#invoke:Date|Date|year=|month=|day=|hour=|minute=|second=|tzhour=|tzmin=|lang=en}}
str = str:gsub(old, new)
 
end
Parameters:
table.insert(T, str)
year,month,day,hour,minute,second: broken down date-time component strings
end
tzhour, tzmin: timezone offset from UTC, hours and minutes
return table.concat(T,'"')
lang: The language to display it in
else
case: Language format (genitive, etc.) for some languages
return String:gsub(old, new)
class: CSS class for the <time> node, use "" for no metadata at all
 
Error Handling:
 
]]
function p.Date(frame)
local args = frame.args
if not (args.lang and mw.language.isSupportedLanguage(args.lang)) then
args.lang = frame:callParserFunction( "int", "lang" ) -- get user's chosen language
end
return p._Date(
{
args.year or '',
args.month or '',
args.day or '',
args.hour or '',
args.minute or '',
args.second or '',
args.tzhour or '',
args.tzmin or ''
},
args.lang, -- language
args.case or '', -- allows to specify grammatical case for the month for languages that use them
args.class or 'dtstart', -- allows to set the html class of the time node where the date is included. This is useful for microformats.
args.trim_year or '100-999' -- by default pad one and 2 digit years to be 4 digit long, while keeping 3 digit years as is
)
end
 
---------------------------------------------------------------------------------------
 
-- process datevec
function p._Date(datevec, lang, case, class, trim_year)
-- INPUT:
-- make sure inputs are in the right format
-- * datevec - Array of {year,month,day,hour,minute,second, tzhour, tzmin} containing broken
for i = #datevec + 1, 8 do
-- down date-time component strings or numbers
datevec[i] = ''
-- OUTPUT:
-- * datecode - a code specifying content of the array where Y' is year, 'M' is month,
-- 'D' is day, 'H' is hour, 'M' minute, 'S' is second. output has to be one of YMDHMS, YMDHM, YMD, YM, MD, Y
-- * datenum - same array but holding only numbers or nuls
local function parserDatevec(datevec)
-- if month is not a number than check if it is a month name in project's language
local month = datevec[2]
if month and month~='' and not tonumber(month) then
datevec[2] = mw.getContentLanguage():formatDate( "n", month)
end
if not case then case = '' end
if not class then class = '' end
if not trim_year then trim_year = '100-999' end
 
-- ifcreate languagedatecode isbased noton which variables are provided thanand lookcheck upfor usersout-of-bound languagevalues
local maxval = {nil, 12, 31, 23, 59, 59, 23, 59} -- max values for year, month, ...
-- WARNING: This step should be done by the template as it does not seem to work as well here (cache issues?)
local minval = {nil, 1, 1, 0, 0, 0, -23, 0} -- min values for year, month, ...
if not lang or not mw.language.isValidCode( lang ) then
lang = 'en'
end
-- Just in case someone broke the internationalization code than fix the english defaults
if i18n.DateLang['en'] == nil then
i18n.DateLang['en'] = 'en-form'
end
if i18n.DateFormat['en-form'] == nil then
i18n.DateFormat['en-form'] = {YMDHMS='j F Y, H:i:s', YMDHM='j F Y, H:i', YMD='j F Y', YM='F Y', MD='j F', Y='Y'}
end
 
-- create datecode based on which variables are provided and check for out of bound values
local maxval = {9999, 12, 31, 23, 59, 60, 23, 59} -- max values for year, month, ...
local c = {'Y', 'M', 'D', 'H', 'M', 'S', '', ''}
local datecode = '' -- a string signifying which combination of variables was provided
local datenum = {} -- date-time encoded as a vector = [year, month, ... , second]
for i, v= in ipairs( datevec )1,8 do
datenum[i] = tonumber(datevec[i])
if v~=nil and v~='' then
if datenum[i] and (i==1 or (datenum[i]>=minval[i] and datenum[i]<=maxval[i])) then
datecode = datecode .. c[i]
datenum[i] = tonumber(v)
if datenum[i]==nil and i==2 then
-- month is not a number -> check if it is a month name in English
v = mw.language.new('en'):formatDate( "n", v)
datenum[i] = tonumber(v)
end
if datenum[i]==nil or datenum[i]>maxval[i] then
-- Some numbers are out of range -> abort and return the empty string
return ''
end
end
end
return datecode, datenum
end
---------------------------------------------------------------------------------------
-- process datevec
-- INPUT:
-- * datecode - a code specifying content of the array where Y' is year, 'M' is month,
-- 'D' is day, 'H' is hour, 'M' minute, 'S' is second. output has to be one of YMDHMS, YMDHM, YMD, YM, MD, Y
-- * datenum - Array of {year,month,day,hour,minute,second, tzhour, tzmin} as numbers or nuls
-- OUTPUT:
-- * timeStamp - date string in the format taken by mw.language:formatDate lua function and {{#time}} perser function
-- https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#mw.language:formatDate
-- https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions#.23time
-- * datecode - with possible corrections
local function getTimestamp(datecode, datenum)
-- create time stamp string (for example 2000-02-20 02:20:20) based on which variables were provided
local timeStamp
Line 133 ⟶ 119:
elseif datecode == 'M' then
timeStamp = string.format('%04i-%02i-%02i', 2000, datenum[2], 1 )
class = '' -- date not complete -> no html formating or micro-tagging of date string
elseif datecode == 'MD' then
timeStamp = string.format('%04i-%02i-%02i', 2000, datenum[2], datenum[3] )
class = '' -- date not complete -> no html formating or micro-tagging of date string
else
returntimeStamp ''= nil -- format not supported
end
return timeStamp, datecode
end
 
---------------------------------------------------------------------------------------
-- ==========================================================
-- trim leading zeros in years prior to year 1000
-- === Create Date String using in chosen language
-- INPUT:
-- ==========================================================
-- * datestr - translated date string
-- * lang - language of translation
-- OUTPUT:
-- * datestr - updated date string
 
local function trimYear(datestr, year, lang)
local yearStr0, yearStr1, yearStr2, zeroStr
yearStr0 = string.format('%04i', year ) -- 4 digit year in standard form "0123"
yearStr1 = mw.language.new(lang):formatDate( 'Y', yearStr0) -- same as calling {{#time}} parser function
--yearStr1 = mw.getCurrentFrame():callParserFunction( "#time", { 'Y', yearStr0, lang } ) -- translate to a language
if yearStr0==yearStr1 then -- most of languages use standard form of year
yearStr2 = tostring(year)
else -- some languages use different characters for numbers
yearStr2 = yearStr1
zeroStr = mw.ustring.sub(yearStr1,1,1) -- get "0" in whatever language
for i=1,3 do -- trim leading zeros
if mw.ustring.sub(yearStr2,1,1)==zeroStr then
yearStr2 = mw.ustring.sub(yearStr2, 2, 5-i)
else
break
end
end
end
return string.gsub(datestr, yearStr1, yearStr2 ) -- in datestr replace long year with trimmed one
end
 
---------------------------------------------------------------------------------------
-- Look up proper format string to be passed to {{#time}} parser function
-- INPUTS:
-- * datecode: YMDHMS, YMDHM, YMD, YM, MD, Y, or M
-- * day : Number between 1 and 31 (not needed for most languages)
-- * lang : language
-- OUTPUT:
-- * dFormat : input to {{#time}} function
local function getDateFormat(datecode, day, lang)
local function parseFormat(dFormat, day)
if dFormat:find('default') and #dFormat>10 then
-- special (and messy) case of dFormat code depending on a day number
-- then json contains a string with more json containing "default" field and 2 digit day keys
-- if desired day is not in that json than use "default" case
dFormat = dFormat:gsub('”','"') -- change fancy double quote to a straight one, used for json marking
local D = mw.text.jsonDecode( dFormat ) --com = mw.dumpObject(D)
day = string.format('d%02i',day) -- create day key
dFormat = D[day] or D.default
dFormat = dFormat:gsub("'", '"') -- change single quote to a double quote, used for {{#time}} marking
end
return dFormat
end
local T = {}
-- which form should the date take?
local tab = mw.ext.data.get('DateI18n.tab', lang)
-- Use langSwitch to pick formating for each language
for _, row in pairs(tab.data) do -- convert the output into a dictionary table
local langDateForm = langSwitch(i18n.DateLang, lang)
local id, _, msg = unpack(row)
T[id] = msg
-- special case of French and Gallic dates, which require different date format for the 1st day of the month
if datenum[3]==1 and (langDateForm=='fr-form' or langDateForm=='ga-form') then
langDateForm = langDateForm .. '1' -- ordinal form for the first day of the month
end
local dFormat = T[datecode]
-- special case of Basque dates, which require different date format for the 1st, 11th, 21st and 31st day of the month
if dFormat=='default' and (datecode=='YMDHMS' or datecode=='YMDHM') then
if langDateForm=='eu-form' then
-- for most languages adding hour:minute:second is done by adding ", HH:MM:SS to the
if (datenum[3]==1 or datenum[3]==21) then
-- day precission date, those languages are skipped in DateI18n.tab and default to
langDateForm = 'eu-form01'
-- English which stores word "default"
elseif (datenum[3]==11 or datenum[3]==31) then
dFormat = parseFormat(T['YMD'], day).. ', H:i'
langDateForm = 'eu-form11'
if datecode=='YMDHMS' then
dFormat = dFormat .. ':s'
end
else
dFormat = parseFormat(dFormat, day)
end
return dFormat
end
 
---------------------------------------------------------------------------------------
-- Look up country specific format input to {{#time}} function
-- Look up proper format string to be passed to {{#time}} parser function
local dFormat = i18n.DateFormat[langDateForm][datecode]
-- INPUTS:
-- * month : month number
-- overwrite default grammatical case of the month (applies mostly to Slavic languages)
-- * case : gramatic case abbriviation, like "ins", "loc"
if (case=='gen') then
-- * lang : language
-- CAUTION: at the moment i18n.DateFormat uses "F" only as month name, but this might change and this operation does not check if 'F' is in "" brackets or not, so if some language starts using 'F' in "" than this will not work for that language
-- OUTPUT:
dFormat = dFormat:gsub("F", "xg");
-- * dFormat : input to {{#time}} function
local function MonthCase(month, case, lang)
local T = {{},{},{},{},{},{},{},{},{},{},{},{}}
local tab = mw.ext.data.get('I18n/MonthCases.tab', lang)
for _, row in pairs(tab.data) do
local mth, cs, msg = unpack(row)
T[mth][cs] = msg
end
return T[month][case]
if (case=='nom') then
end
-- CAUTION: at the moment i18n.DateFormat uses "xg" only as month name, but this might change and this operation does not check if 'xg' is in "" brackets or not, so if some language starts using 'xg' in "" than this will not work for that language
 
dFormat = dFormat:gsub("xg", "F");
-- ==================================================
-- === External functions ===========================
-- ==================================================
local p = {}
 
--[[ ========================================================================================
Date
This function is the core part of the ISOdate template.
Usage:
local Date = require('Module:DateI18n')._Date
local dateStr = Date({2020, 12, 30, 12, 20, 11}, lang)
Parameters:
* {year,month,day,hour,minute,second, tzhour, tzmin}: broken down date-time component strings or numbers
tzhour, tzmin are timezone offsets from UTC, hours and minutes
* lang: The language to display it in
* case: Language format (genitive, etc.) for some languages
* class: CSS class for the <time> node, use "" for no metadata at all
]]
function p._Date(datevec, lang, case, class, trim_year)
-- make sure inputs are in the right format
if not lang or not mw.language.isValidCode( lang ) then
lang = mw.getCurrentFrame():callParserFunction( "int", "lang" ) -- get user's chosen language
end
if lang == 'be-tarsk' then
if ((lang=='ru' or lang=='pl' or lang=='cs' or lang=='sl' or lang=='sk') and (case=='loc' or case=='ins')) or
lang = 'be-x-old'
(lang=='fi' and (case=='ptv' or case=='ine'or case=='ela'or case=='ill') ) then
end
local monthEn = mw.language.new('en'):formatDate( "F", timeStamp) -- month name in English
-- month name using proper case and language. It relies on messages stored in MediaWiki namespace for some cases and languages
-- process datevec and extract timeStamp and datecode strings as well as numeric datenum array
-- That is why this IF statement uses "lang" not "langDateForm" variable to decide
local datecode, datenum = parserDatevec(datevec)
local monthMsg = mw.message.new( string.format('%s-%s', monthEn, case ) ):inLanguage( lang )
local year, month, day = datenum[1], datenum[2], datenum[3]
if not monthMsg:isDisabled() then -- make sure it exists
local timeStamp, datecode = getTimestamp(datecode, datenum)
local month=monthMsg:plain()
if not timeStamp then -- something went wrong in parserDatevec
dFormat = dFormat:gsub('F', '"'..month..'"'); -- replace default month with month name we already looked up
return ''
dFormat = dFormat:gsub('xg', '"'..month..'"');
end
-- Commons [[Data:DateI18n.tab]] page stores prefered formats for diferent
-- languages and datecodes (specifying year-month-day or just year of month-day, etc)
-- Look up country specific format input to {{#time}} function
local dFormat = getDateFormat(datecode, day, lang)
 
-- By default the gramatical case is not specified (case=='') allowing the format to be specified
-- in [[Data:DateI18n.tab]]. You can overwrite the default grammatical case of the month by
-- specifying "case" variable. This is needed mostly by Slavic languages to create more complex
-- phrases as it is done in [[c:Module:Complex date]]
case = case or ''
if (lang=='qu' or lang=='qug') and (case=='nom') then
-- Special case related to Quechua and Kichwa languages. The form in the I18n is
-- Genitive case with suffix "pi" added to month names provided by {#time}}
-- in Nominative case that "pi" should be removed
-- see https://commons.wikimedia.org/wiki/Template_talk:Date#Quechua from 2014
dFormat = dFormat:gsub('F"pi"', 'F')
elseif (case=='gen') then
dFormat = strReplace(dFormat, "F", "xg")
elseif (case=='nom') then
dFormat = strReplace(dFormat, "xg", "F")
elseif (case ~= '') then
-- see is page [[Data:I18n/MonthCases.tab]] on Commons have name of the month
-- in specific gramatic case in desired language. If we have it than replace
-- "F" and xg" in dFormat
local monthMsg = MonthCase(month, case, lang)
if monthMsg and monthMsg ~= '' then -- make sure it exists
dFormat = strReplace(dFormat, 'F', '"'..monthMsg..'"') -- replace default month with month name we already looked up
dFormat = strReplace(dFormat, 'xg', '"'..monthMsg..'"')
end
end
-- Special case related to Quechua and Kichwa languages
-- see https://commons.wikimedia.org/wiki/Template_talk:Date#Quechua from 2014
if (lang=='qu' or lang=='qug') and case=='nom' then
dFormat = dFormat:gsub('F"pi"', 'F');
end
 
-- LuaTranslate onlythe date formating using {{#time}} parser functionspecified (new)format
-- See https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#mw.language:formatDate and
-- prefered call which gives "Lua error: too many language codes requested." on the [[Module talk:Date/sandbox/testcases]] page
-- https://www.mediawiki.org/wiki/Help:Extension:ParserFunctions##time for explanation of the format
--local datestr = mw.language.new(lang):formatDate( dFormat, timeStamp)
local datestr = mw.getCurrentFramelanguage.new(lang):callParserFunctionformatDate( "#time", { dFormat, timeStamp,) lang-- same as using {{#time}} )parser function
-- Another specialSpecial case related to Thai solar calendar: prior to 1940 new-year was at different
-- time of year, so just year (datecode=='Y') is ambiguous and is replaced by "YYYY or YYYY" phrase
if lang=='th' and datenum[1]~= nil and datenum[1]<=1940 then
if lang=='th' and datecode=='Y' and year<=1940 then
-- As of 2014 {{#time}} parser function did not resolve those cases properly
datestr = string.format('%04i หรือ %04i', year+542, year+543 )
-- See https://en.wikipedia.org/wiki/Thai_solar_calendar#New_year for reference
-- Disable once https://bugzilla.wikimedia.org/show_bug.cgi?id=66648 is fixed
if datecode=='Y' then -- date is ambiguous
datestr = string.format('%04i หรือ %04i', datenum[1]+542, datenum[1]+543 )
elseif datenum[2]<=3 then -- year is wrong (one too many)
datestr = datestr:gsub( string.format('%04i', datenum[1]+543), string.format('%04i', datenum[1]+542 ) )
end
end
-- If year<1000 than either keep itthe date padded to the length of 4 digits or trim it
-- decide if the year will stay padded with zeros (for years in 0-999 range)
if datenum[1]~= nilyear and datenum[1]year<1000 then
localif trim = yesnotype(trim_year,nil)=='nil' then
trim_year = '100-999'
if trim == nil then
end
local trim = yesno(trim_year,nil) -- convert to boolean
if trim==nil and type(trim_year)=='string' then
-- if "trim_year" not a simple True/False than it is range of dates
-- for example '100-999' means to pad one and 2 digit years to be 4 digit long, while keeping 3 digit years as is
local YMin, YMax = trim_year:match( '(%d+)-(%d+)' )
trim = (YMin~=nil and datenum[1]year>=tonumber(YMin) and datenum[1]year<=tonumber(YMax))
end
if trim==true then
datestr = trimYear(datestr, year, lang) -- in datestr replace long year with trimmed one
-- If the date form isn't the Thai solar calendar, don't zero pad years in the range of 100-999.
-- If at some point support for Islamic/Hebrew/Japanese years is added, they may need to be skipped as well.
if trim then
--local yearStr1 = mw.language.new(lang):formatDate( 'Y', timeStamp)
local yearStr1 = mw.getCurrentFrame():callParserFunction( "#time", { 'Y', timeStamp, lang } )
--local yearStr1 = datestr:match( '%d%d%d%d' ) -- 4 digits in a row (in any language) - that must be a year
local yearStr2 = yearStr1
local zeroStr = mw.ustring.sub(yearStr1,1,1)
for i=1,3 do -- trim leading zeros
if mw.ustring.sub(yearStr2,1,1)==zeroStr then
yearStr2 = mw.ustring.sub(yearStr2, 2, 5-i)
else
break
end
end
datestr = datestr:gsub( yearStr1, yearStr2 )
--datestr = string.format('%s (%s, %s)', datestr, yearStr1, yearStr2 )
end
end
 
-- append timezone if present
if datevecdatenum[7] ~= '' and (datecode == 'YMDHMS' or datecode == 'YMDHM') then
-- use {{#time}} parser function to create timezone string, so that we use correct character set
local tzstr, tzhournum = '', tonumber(datevec[7])
iflocal tzhournumsign = (datenum[7]< 0) then tzstr =and '−' else tzstr =or '+' end
tzstrtimeStamp = tzstr..string.format("2000-01-01 %02d02i:%02i:00", math.abs(tzhournumdatenum[7]), datenum[8] or 0)..':'
local timezone = mw.language.new(lang):formatDate( 'H:i', timeStamp) -- same as using {{#time}} parser function
if datevec[8] ~= '' then tzstr = tzstr..datevec[8] else tzstr = tzstr..'00' end
datestr = datestrstring..'format("%s '..tzstr%s%s", datestr, sign, timezone )
end
 
-- html formating and tagging of date string
if class and class ~= '' and datecode~='M' and datecode~='MD'then
local DateHtmlTags = '<span style="white-space:nowrap"><time class="%s" datetime="%s">%s</time></span>'
datestr = DateHtmlTags:format(class, timeStamp, datestr)
end
return datestr
end
 
--[[ ========================================================================================
Date
This function is the core part of the ISOdate template.
Usage:
{{#invoke:DateI18n|Date|year=|month=|day=|hour=|minute=|second=|tzhour=|tzmin=|lang=en}}
Parameters:
* year, month, day, hour, minute, second: broken down date-time component strings
* tzhour, tzmin: timezone offset from UTC, hours and minutes
* lang: The language to display it in
* case: Language format (genitive, etc.) for some languages
* class: CSS class for the <time> node, use "" for no metadata at all
]]
function p.Date(frame)
local args = {}
for name, value in pairs( frame.args ) do
name = string.gsub( string.lower(name), ' ', '_')
args[name] = value
end
return p._Date(
{ args.year, args.month, args.day, args.hour, args.minute, args.second, args.tzhour, args.tzmin },
args.lang, -- language
args.case, -- allows to specify grammatical case for the month for languages that use them
args.class or 'dtstart', -- allows to set the html class of the time node where the date is included. This is useful for microformats.
args.trim_year or '100-999' -- by default pad one and 2 digit years to be 4 digit long, while keeping 3 digit years as is
)
end