Module:Date time

Documentation for this module may be created at Module:Date time/doc

--[[
Module:Date time – Date formatting and validation module.

This module provides functions for validating and formatting dates in templates such as
{{Start date}}, {{End date}}, {{Start date and age}}, and {{End date and age}}. 
It handles:
- Validation of date components (year, month, day)
- Validation of time components (hour, minute, second)
- Timezone formatting and validation
- Generation of appropriate hCalendar microformat markup
- "time-ago" calculations

Design notes:
- Functions are organized into helper, validation, and formatting sections
- Error handling uses a consistent pattern with centralized error messages
- Timezone validation supports standard ISO 8601 formats
- Leap year calculation is cached for performance
]]

require("strict")

local p = {}

---------------
-- Constants --
---------------

local HTML_SPACE = " "
local HTML_NBSP = " "

-- Error message constants
local ERROR_MESSAGES = {
	integers = "All values must be integers",
	has_leading_zeros = "Values cannot have unnecessary leading zeros",
	missing_year = "Year value is required",
	invalid_month = "Value is not a valid month",
	missing_month = "Month value is required when a day is provided",
	invalid_day = "Value is not a valid day (Month %d has %d days)",
	invalid_hour = "Value is not a valid hour",
	invalid_minute = "Value is not a valid minute",
	invalid_second = "Value is not a valid second",
	timezone_incomplete_date = "A timezone cannot be set without a day and hour",
	invalid_timezone = "Value is not a valid timezone",
	yes_value_parameter = '%s must be either "yes" or "y"',
	duplicate_parameters = 'Duplicate parameters used: %s and %s',
	template = "Template not supported",
	time_without_hour = "Minutes and seconds require an hour value"
}

-- Template class mapping
-- "itvstart" and "itvend" are unique classes used by the TV infoboxes,
-- which only allow the usage of {{Start date}} and {{End date}}.
local TEMPLATE_CLASSES = {
	["start date"] = "bday dtstart published updated itvstart",
	["start date and age"] = "bday dtstart published updated",
	["end date"] = "dtend itvend",
	["end date and age"] = "dtend"
}

-- Templates that require "time ago" calculations
local TIME_AGO = {
	["start date and age"] = true,
	["end date and age"] = true
}

-- English month names
local MONTHS = {
	"January", "February", "March", "April", "May", "June",
	"July", "August", "September", "October", "November", "December"
}

-- Error category
local ERROR_CATEGORY = "[[Category:Pages using Module:Date time with invalid values]]"

-- Namespaces where error categories should be applied
local CATEGORY_NAMESPACES = {
	[0] = true,		-- Article
	[1] = true,		-- Article talk
	[4] = true,		-- Wikipedia
	[10] = true,	-- Template
	[100] = true,	-- Portal
	[118] = true	-- Draft
}

-- Cached leap year calculations for performance
local leap_year_cache = {}

-- Local variables for error handling
local help_link

----------------------
-- Helper Functions --
----------------------

--- Pads a number with leading zeros to ensure a minimum of two digits.
-- @param value (number|string) The value to pad with leading zeros
-- @return string The value padded to at least two digits, or nil if input is nil
local function pad_left_zeros(value)
	if value == nil then
		return nil
	end

	local str = tostring(value)
	return string.rep("0", math.max(0, 2 - #str)) .. str
end

--- Replaces [[U+2212]] (Unicode minus) with [[U+002D]] (ASCII hyphen) or vice versa.
-- @param value (string) The string value to process
-- @param to_unicode (boolean) If true, converts ASCII hyphen to Unicode minus;
--                            If false, converts Unicode minus to ASCII hyphen
-- @return string The processed string with appropriate minus characters, or nil if input is nil
local function replace_minus_character(value, to_unicode)
	if not value then
		return nil
	end

	if to_unicode then
		return value:gsub("-", "−")
	end

	return value:gsub("−", "-")
end

--- Normalizes timezone format by ensuring proper padding of hours.
-- @param timezone (string) The timezone string to normalize
-- @return string The normalized timezone string with properly padded hours, or nil if input is nil
local function fix_timezone(timezone)
	if not timezone then
		return nil
	end

	-- Replace U+2212 (Unicode minus) with U+002D (ASCII hyphen)
	timezone = replace_minus_character(timezone, false)

	-- Match the timezone pattern for ±H:MM format
	local sign, hour, minutes = timezone:match("^([+-])(%d+):(%d+)$")

	if sign and hour and minutes then
		-- Pad the hour with a leading zero if necessary
		hour = pad_left_zeros(hour)
		return sign .. hour .. ":" .. minutes
	end

	-- If no match, return the original timezone (this handles invalid or already padded timezones)
	return timezone
end

--- Checks if a timezone string is valid according to standard timezone formats.
-- Valid timezones range from UTC-12:00 to UTC+14:00.
-- @param timezone (string) The timezone string to validate
-- @return boolean true if the timezone is valid, false otherwise
local function is_timezone_valid(timezone)
	-- Consolidated timezone pattern for better performance
	local valid_patterns = {
		-- Z (UTC)
		"^Z$",
		-- Full timezone with minutes ±HH:MM
		"^[+]0[1-9]:[0-5][0-9]$",
		"^[+-]0[1-9]:[0-5][0-9]$", 
		"^[+-]1[0-2]:[0-5][0-9]$",
		"^[+]1[34]:[0-5][0-9]$",
		-- Whole hour timezones ±HH
		"^[+-]0[1-9]$",
		"^[+-]1[0-2]$",
		"^[+]1[34]$",
		-- Special cases
		"^[+]00:00$",
		"^[+]00$"
	}

	-- Additional checks for invalid -00 and -00:00 cases
	if timezone == "-00" or timezone == "-00:00" then
		return false
	end

	for _, pattern in ipairs(valid_patterns) do
		if string.match(timezone, pattern) then
			return true
		end
	end

	return false
end

--- Checks if a given year is a leap year.
-- Uses a cache for better performance.
-- @param year (number) The year to check for leap year status
-- @return boolean true if the year is a leap year, false otherwise
local function is_leap_year(year)
	if leap_year_cache[year] == nil then
		leap_year_cache[year] = (year % 4 == 0 and year % 100 ~= 0) or (year % 400 == 0)
	end
	return leap_year_cache[year]
end

--- Returns the number of days in a given month of a specified year.
-- Handles leap years for February.
-- @param year (number) The year to check for leap year conditions
-- @param month (number) The month (1-12) for which to return the number of days
-- @return number The number of days in the specified month, accounting for leap years
local function get_days_in_month(year, month)
	local days_in_month = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }

	if month == 2 and is_leap_year(year) then
		return 29
	end

	return days_in_month[month] or 0
end

--- Checks if a given value has invalid leading zeros.
-- @param value (string) The value to check for leading zeros
-- @param field_type (string) Field type ("day", "month", "hour", "minute", "second")
-- @return boolean true if the value has invalid leading zeros, false otherwise
local function has_leading_zeros(value, field_type)
	value = tostring(value)

	-- Common checks for day and month
	if field_type == "day" or field_type == "month" then
		-- Reject "00" and values with leading zero followed by more than one digit
		return value == "00" or 
		       string.match(value, "^0[0-9][0-9]$") ~= nil or 
		       string.match(value, "^0[1-9][0-9]") ~= nil
	end

	-- Checks for hour, minute, second
	if field_type == "hour" or field_type == "minute" or field_type == "second" then
		-- Allow "00" and "01" to "09"
		if value == "00" or string.match(value, "^0[1-9]$") then
			return false
		end

		-- Reject values starting with "0" followed by more than one digit
		return string.match(value, "^0[0-9][0-9]+$") ~= nil
	end

	return false
end

--- Checks if a given value is an integer.
-- @param value (string|number) The value to check
-- @return boolean true if the value is a valid integer, false otherwise
local function is_integer(value)
	if not value then
		return false
	end
	-- Check if the value is a number first
	local num_value = tonumber(value)
	if not num_value then
		return false
	end
	-- Check if it's an integer by comparing floor with the original
	if math.floor(num_value) ~= num_value then
		return false
	end
	-- For string inputs, check for decimal point to reject values like "7."
	if type(value) == "string" then
		-- If the string contains a decimal point, it's not an integer
		if string.find(value, "%.") then
			return false
		end
	end
	return true
end

--- Returns the name of a month based on its numerical representation.
-- @param month_number (number) The month number (1-12)
-- @return string|nil The name of the month, or nil if invalid
local function get_month_name(month_number)
	month_number = tonumber(month_number)
	return MONTHS[month_number]
end

--- Generates an error message wrapped in HTML.
-- @param message (string) The error message to format
-- @param add_tracking_category (boolean, optional) If false, omits the tracking category
-- @return string An HTML-formatted error message with help link and error category
local function generate_error(message, add_tracking_category)
	local category = ERROR_CATEGORY

	if add_tracking_category == false then
		category = ""
	end

	-- Get current page title object
	local article_title = mw.title.getCurrentTitle()
	
	-- Special case for testcases pages
	local is_test_page = article_title.subpageText == "testcases"
	local allow_this_test_page = article_title.fullText == "Module talk:Date time/testcases"

	-- Remove category if the page is not in a tracked namespace or is any other testcases other than this module
	if (not CATEGORY_NAMESPACES[article_title.namespace] and not allow_this_test_page)
	or (is_test_page and not allow_this_test_page) then
		category = ""
	end

	return '<strong class="error">Error: ' .. message .. '</strong> ' .. help_link .. category
end

--------------------------
-- Formatting Functions --
--------------------------

--- Formats the time portion of a datetime string.
-- @param hour (string) The hour component
-- @param minute (string) The minute component
-- @param second (string) The second component
-- @return string The formatted time string, or empty string if hour is nil
local function format_time_string(hour, minute, second)
	if not hour then
		return ""
	end

	local time_string = string.format("%s:%s", hour, minute)

	if second and second ~= "00" and minute ~= "00" then
		time_string = string.format("%s:%s", time_string, second)
	end

	return time_string .. "," .. HTML_SPACE
end

--- Formats the date portion of a datetime string based on the specified format.
-- @param year (string) The year component
-- @param month (string) The month component
-- @param day (string) The day component
-- @param date_format_dmy (string) The date format ("yes" or "y" for day-month-year, otherwise month-day-year)
-- @return string The formatted date string, or empty string if year is nil
local function format_date_string(year, month, day, date_format_dmy)
	if not year then
		return ""
	end

	local date_string
	if month then
		local month_name = get_month_name(month)
		
		if day then
			day = tonumber(day)
			if date_format_dmy then
				date_string = day .. HTML_NBSP .. month_name
			else
				date_string = month_name .. HTML_NBSP .. day .. ","
			end
			date_string = date_string .. HTML_NBSP .. year
		else
			date_string = month_name .. HTML_NBSP .. year
		end
	else
		date_string = year
	end

	return date_string
end

--- Formats the timezone portion of a datetime string.
-- @param timezone (string) The timezone component
-- @return string The formatted timezone string, or empty string if timezone is nil
local function format_timezone(timezone)
	if not timezone then
		return ""
	end

	return HTML_SPACE .. (timezone == "Z" and "(UTC)" or "(" .. timezone .. ")")
end

--- Generates an hCalendar microformat string for the given date-time values.
-- @param date_time_values (table) A table containing date and time components
-- @param classes (string) The CSS classes to apply to the microformat span
-- @return string The HTML for the hCalendar microformat
local function generate_h_calendar(date_time_values, classes)
	local parts = {}

	if date_time_values.year then
		table.insert(parts, date_time_values.year)

		if date_time_values.month then
			table.insert(parts, "-" .. date_time_values.month)

			if date_time_values.day then
				table.insert(parts, "-" .. date_time_values.day)
			end
		end

		if date_time_values.hour then
			table.insert(parts, "T" .. date_time_values.hour)

			if date_time_values.minute then
				table.insert(parts, ":" .. date_time_values.minute)

				if date_time_values.second then
					table.insert(parts, ":" .. date_time_values.second)
				end
			end
		end
	end

	local h_calendar_content = table.concat(parts) .. (date_time_values.timezone or "")
	local class_span = string.format('<span class="%s">', classes)

	return string.format(
		'<span style="display: none;">%s(%s)</span>', 
		HTML_NBSP, 
		class_span .. h_calendar_content .. '</span>'
	)
end

--- Generates a "time ago" string for age calculation templates.
-- @param date_time_values (table) Table containing date components (year, month, day)
-- @param br (boolean) Whether to include a line break before the time ago text
-- @param p (boolean) Whether to format with parentheses around the time ago text
-- @return string Formatted "time ago" text wrapped in a noprint span
local function get_time_ago(date_time_values, br, p)
	-- Build timestamp based on available date components
	local timestamp
	local min_magnitude
	
	if date_time_values.day then
		-- Format with padding for month and day if needed
		timestamp = string.format("%d-%02d-%02d", 
								 date_time_values.year, 
								 date_time_values.month, 
								 date_time_values.day)
		min_magnitude = "days"
	elseif date_time_values.month then
		-- Format with padding for month if needed
		timestamp = string.format("%d-%02d", 
								 date_time_values.year, 
								 date_time_values.month)

		-- Get the current date
		local current_date = os.date("*t")

		-- Compute the difference in months
		local year_diff = current_date.year - date_time_values.year
		local month_diff = (year_diff * 12) + (current_date.month - date_time_values.month)

		-- If the difference is less than 12 months, use "months", otherwise "years"
		if month_diff < 12 then
			min_magnitude = "months"
		else
			min_magnitude = "years"
		end
	else
		timestamp = tostring(date_time_values.year)
		min_magnitude = "years"
	end
	
	-- Calculate time ago using [[Module:Time]] ago
	local m_time_ago = require("Module:Time ago")._main
	local time_ago = m_time_ago({timestamp, ["min_magnitude"] = min_magnitude})
	
	-- Format the result based on br and p parameters
	if br then
		time_ago = p and ("<br/>(" .. time_ago .. ")") or (";<br/>" .. time_ago)
	else
		time_ago = p and (HTML_SPACE .. "(" .. time_ago .. ")") or (";" .. HTML_SPACE .. time_ago)
	end
	
	-- Wrap in noprint span
	return "<span class=\"noprint\">" .. time_ago .. "</span>"
end

--------------------------
-- Validation Functions --
--------------------------

--- Validates the date and time values provided.
-- @param args (table) Table containing date and time values and optional parameters
-- @return nil|string Nil if validation passes, or an error message if validation fails
local function _validate_date_time(args)
	local template_name = args.template or "start date"
	help_link = string.format("<small>[[:Template:%s|(help)]]</small>", template_name)

	-- Store and validate date-time values
	local date_time_values = {
		year = args[1], 
		month = args[2], 
		day = args[3],
		hour = args[4], 
		minute = args[5], 
		second = args[6]
	}

	-- Validate each value
	for key, value in pairs(date_time_values) do
		if value then
			-- Check for integer and leading zeros
			if not is_integer(value) then
				return generate_error(ERROR_MESSAGES.integers)
			end

			if has_leading_zeros(tostring(value), key) then
				return generate_error(ERROR_MESSAGES.has_leading_zeros)
			end

			-- Convert to number
			date_time_values[key] = tonumber(value)
		end
	end

	-- Validate date components
	if not date_time_values.year then
		return generate_error(ERROR_MESSAGES.missing_year)
	end

	if date_time_values.month and (date_time_values.month < 1 or date_time_values.month > 12) then
		return generate_error(ERROR_MESSAGES.invalid_month)
	end

	if date_time_values.day then
		if not date_time_values.month then
			return generate_error(ERROR_MESSAGES.missing_month)
		end

		local max_day = get_days_in_month(date_time_values.year, date_time_values.month)
		if date_time_values.day < 1 or date_time_values.day > max_day then
			return generate_error(string.format(ERROR_MESSAGES.invalid_day, date_time_values.month, max_day))
		end
	end

	-- Validate time components
	if (date_time_values.minute or date_time_values.second) and not date_time_values.hour then
		return generate_error(ERROR_MESSAGES.time_without_hour)
	end

	if date_time_values.hour and (date_time_values.hour < 0 or date_time_values.hour > 23) then
		return generate_error(ERROR_MESSAGES.invalid_hour)
	end

	if date_time_values.minute and (date_time_values.minute < 0 or date_time_values.minute > 59) then
		return generate_error(ERROR_MESSAGES.invalid_minute)
	end

	if date_time_values.second and (date_time_values.second < 0 or date_time_values.second > 59) then
		return generate_error(ERROR_MESSAGES.invalid_second)
	end

	-- Timezone cannot be set without a specific date and hour
	if args[7] and not (date_time_values.day and date_time_values.hour) then
		return generate_error(ERROR_MESSAGES.timezone_incomplete_date)
	elseif args[7] and not is_timezone_valid(args[7]) then
		return generate_error(ERROR_MESSAGES.invalid_timezone)
	end

	-- Validate that there aren't any duplicate parameters
	if args.p and args.paren then
		return generate_error(string.format(ERROR_MESSAGES.duplicate_parameters, "p", "paren"))
	end

	-- Validate parameters that use "y" or "yes" for values	
	local boolean_params = {'df', 'p', 'paren', 'br'}
	for _, param_name in ipairs(boolean_params) do
		if args[param_name] and not (args[param_name] == "yes" or args[param_name] == "y") then
			return generate_error(string.format(ERROR_MESSAGES.yes_value_parameter, param_name))
		end
	end

	return nil
end

----------------------
-- Public Functions --
----------------------

--- Validates date-time values from template arguments.
-- @param frame (table) The MediaWiki frame containing template arguments
-- @return nil|string Result of date-time validation
function p.validate_date_time(frame)
	local get_args = require("Module:Arguments").getArgs
	local args = get_args(frame)
	
	-- Sanitize inputs
	args[7] = fix_timezone(args[7])
	
	return _validate_date_time(args)
end

--- Generates a formatted date string with microformat markup.
-- @param frame (table) The MediaWiki frame containing template arguments
-- @return string A formatted date string, or an error message if validation fails
function p.generate_date(frame)
	local get_args = require("Module:Arguments").getArgs
	local args = get_args(frame)

	-- Sanitize inputs
	args[7] = fix_timezone(args[7])

	local validation_error = _validate_date_time(args)
	if validation_error then
		return validation_error
	end

	local classes = TEMPLATE_CLASSES[args.template or "start date"]
	if not classes then
		return generate_error(ERROR_MESSAGES.template, false)
	end
	
	-- Process date-time values
	local date_time_values = {
		year = args[1], 
		month = pad_left_zeros(args[2]), 
		day = pad_left_zeros(args[3]),
		hour = pad_left_zeros(args[4]), 
		minute = args[5] and pad_left_zeros(args[5]) or "00",
		second = args[6] and pad_left_zeros(args[6]) or "00", 
		timezone = replace_minus_character(args[7], true) -- Restore U+2212 (Unicode minus)
	}

    -- Generate individual components
	local time_string = format_time_string(
		date_time_values.hour, 
		date_time_values.minute, 
		date_time_values.second
	)

	local date_string = format_date_string(
		date_time_values.year, 
		date_time_values.month, 
		date_time_values.day, 
		args.df
	)

	local timezone_string = format_timezone(date_time_values.timezone)

	local time_ago = ""
	if TIME_AGO[args.template] then
		time_ago = get_time_ago(
			date_time_values, 
			args.br,
			args.p or args.paren
		)
	end

	local h_calendar = generate_h_calendar(date_time_values, classes)

	-- Combine components
	return time_string .. date_string .. timezone_string .. time_ago .. h_calendar
end

return p