Editing Module:Sandbox/Harry noob/Dates

--Harry noob Google Code-in 2019, Lua Task 9

--[[
date format
M= word month
m= number month

d/m/y
y/m/d
m/d/y
d M y
M d y
y M
M y
y
]]--

--[[
TODO:
Q. how to know it is invalid day or it just doesnt match the current pattern?
A. save when it matches some pattern
]]--

local p = {}
local log = mw.log

function p.canonicalDate(frame)
	local fmt = frame.args.format
	local text = frame.args.text
	local date, defaultFmt = p.parseDate(text)
	if not date then
		return "Invalid entry"
	end
	local formattedDate = p.formatDate(date, fmt, defaultFmt)
	local result = addPreSuffix(text, formattedDate)
	return result
end

function p.parseDate(text)
	local d, m, y, a, b, c, mName
	local mayMatch
	
	log("try to match y/m/d or d/m/y or m/d/y")
	a, b, c = string.match(text, "(%d+)[/%- ](%d+)[/%- ](%d+)") -- y/m/d or d/m/y or m/d/y
	log(a, b, c)
	mayMatch = a and b and c
	local possibleArrangement = {{a, b, c}, {c, b, a}, {c, a, b}}
	mw.logObject(possibleArrangement, "possibleArrangement: ")
	for _, arrange in ipairs(possibleArrangement) do
		y, m, d = unpack(arrange) -- this is not python :(, need to add unpack, destructing assignment doesnt works with table
		if isValidNumDate(y, m, d) then
			log("isValidNumDate passed 1")
			return mapToNum({y, m, d}), "iso"
		end
	end
	if mayMatch then
		return nil
	end
	
	local monthShortNameToNum = {jan = 1, feb = 2, mar = 3, apr = 4,
		jun = 6, jul = 7, aug = 8, sept = 9,
		oct = 10, nov = 11, dec = 12}
	local monthLongNameToNum = {january = 1, february = 2, march = 3,
		april = 4, may = 5, june = 6, july = 7, august = 8, september = 9,
		october = 10, november = 11, december = 12}
	
	-- match long name first, and then short name
	local monthFound = false
	local start, stop
	for monthName, monthNum in pairs(monthLongNameToNum) do
		start, stop = text:lower():find(monthName)
		log("month for loop 1:", monthName, monthNum, start, stop)
		if start and stop then
			monthFound = true
			mName = text:sub(start, stop) -- ~= monthName, == orginal month name
			log("mName:", mName)
			m = monthNum
			break
		end
	end
	if not monthFound then
		for monthName, monthNum in pairs(monthShortNameToNum) do
			if monthName == "sept" then
				start, stop = text:lower():find("sept?")
			else
				start, stop = text:lower():find(monthName)
			end
			if start and stop then
				monthFound = true
				mName = text:sub(start, stop) -- ~= monthName, == orginal month name
				log("mName:", mName)
				m = monthNum
				break
			end
		end
	end
	
	if monthFound then
		log("try to match d M y or y M d")
		d, y = text:match("(%d+)%D-" .. mName .. "%D-(%d+)") -- d M y
		if isValidNumDate(y, m, d) then
			log("isValidNumDate passed 2")
			return mapToNum({y, m, d}), "dmy"
		end
		if d and y then
			return nil
		end
		
		log("try to match M d y")
		d, y = text:match(mName .. "%D-(%d+)%D-(%d+)") -- M d y
		if isValidNumDate(y, m, d) then
			log("isValidNumDate passed 4")
			return mapToNum({y, m, d}), "mdy"
		end
		if d and y then
			return nil
		end
		
		
		log("try to match M y")
		y = text:match(mName .. "%D-(%d+)") -- M y
		log("y =", y)
		if y then
			return mapToNum({y, m}), "my"
		end
		
		log("try to match y M")
		y = text:match("(%d+)%D-" .. mName) -- y M
		log("y =", y)
		if y then
			return mapToNum({y, m}), "ym"
		end
	end
	
	log("try to match y")
	y = text:match("(%d+)%D-$") -- y
	if y then
		return {tonumber(y)}, "year"
	end
	
	log("no match")
	return nil
end

-- y, m, d are string number
function isValidNumDate(y, m, d)
	log("isValidNumDate receive:", y, m, d)
	y, m, d = tonumber(y), tonumber(m), tonumber(d)
	local NOfDays = {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31}
	local result = y and m and d and 1 <= m and m <= 12 and (
		isLeap(y) and m == 2 and 1 <= d and       -- if is leap year and February
		(d <= 29) or                        -- then test if it <= 29
		(d <= NOfDays[m]))                  -- else follow general case (use assert and pcall instead of returning 2 result can type less code)
	log("isValidNumDate result:" .. tostring(result))
	return result
end

function isLeap(y)
	return y%4 == 0 and (y%100 ~= 0 or y%400 == 0)
end

function mapToNum(arr)
	mw.logObject(arr, "mapToNum receive arr:")
	for i, v in pairs(arr) do
		arr[i] = tonumber(v)
	end
	mw.logObject(arr, "mapToNum result arr:")
	return arr
end

function p.formatDate(date, fmt, defaultFmt)
	local monthNames = {"January", "February", "March",
		"April", "May", "June", "July", "August", "September",
		"October", "November", "December"}
	local y, m, d = unpack(date)
	
	if (fmt or defaultFmt) == "iso" then
		return ("%d-%02d-%02d"):format(y, m, d)
	elseif (fmt or defaultFmt) == "dmy" then
		return ("%d %s %d"):format(d, monthNames[m], y)
	elseif (fmt or defaultFmt) == "mdy" then
		return ("%s %d, %d"):format(monthNames[m], d, y)
	elseif (fmt or defaultFmt) == "my" then
		return ("%s %d"):format(monthNames[m], y)
	elseif (fmt or defaultFmt) == "ym" then
		return ("%d %s"):format(y, monthNames[m])	
	else
		return tostring(y)
	end
end

function addPreSuffix(text, dateStr)
	local aboutList = {"about", "around", "uncertain", "roughly",
		"approximate", "close to", "near "}
	local suffixList = {"BCE", "CE", "BC", "AD"}
	
	for i, word in ipairs(aboutList) do
		if text:lower():find(word) then
			dateStr = "circa " .. dateStr
			break
		end
	end
	
	for i, word in ipairs(suffixList) do
		if text:find(" " .. word) then
			dateStr = dateStr .. " " .. word
			break
		end
	end
	
	return dateStr
end

return p