Module:ISBN RangeMessage xlate

require ('strict');

local range_message_raw = mw.title.getCurrentTitle ():getContent();				-- 

local group_pattern = '<Group>(.-)</Group>';									-- capture holds <Prefix>, <Agency>, and <Rules> groups
local prefix_pattern = '<Prefix>(%d+)%-(%d+)</Prefix>';							-- captures are GS1 prefix (978 or 979) and registration group
local agency_pattern = '<Agency>(.-)</Agency>';									-- capture holds language, country, territory
local rule_pattern = '<Rule>(.-)</Rule>';										-- capture holds registrant <Range> and <Length> groups
local range_pattern = '<Range>(%d+)%-(%d+)</Range>';							-- registrant; captures are min and max ranges right-padded to 7 digits; min not currently used
local length_pattern = '<Length>(%d)</Length>';									-- registrant; capture holds number of (left) digits to use from <Range>; for len=2 in max=1299999 -> 12


--[[--------------------------< R A N G E _ M E S S A G E _ X L A T E >----------------------------------------

{{#invoke:Sandbox/trappist the monk/ISBN RangeMessage xlate|range_message_xlate}}

This function translates a local copy of https://www.isbn-international.org/export_rangemessage.xml for use by
Module:Format ISBN by way of Module:Format ISBN/data.

To update <hyphen_pos_t> in Module:Format ISBN/data:
	1. copy the new RangeMessage.xml text from the ISBN international export_rangemessage.xml page (url above) to
		your clipboard and paste it over the existing xml data inside the html comments in this module's doc page.
	2. save.  The new translation is presented on this module's doc page.
	3. copy the new translation from the doc page to your clipboard
	4. edit Module:Format ISBN/data and paste the new <hyphen_pos_t> table over the old
	5. save

]]

local function range_message_xlate ()
	local out_t = {};															-- the base output goes here (before prettifying)
	for group in range_message_raw:gmatch (group_pattern) do					-- get a <Group>...</Group> block
		local prefix, registration_group = group:match (prefix_pattern);		-- get the prefix and registration group

		local agency = group:match (agency_pattern);							-- get the 'agency'; not currently used; TODO: should it be? as a comment? 
		
		for rule in group:gmatch (rule_pattern) do								-- get a <Rule>...</Rule> block
			local isbn_parts_t = {};											-- the prefix, registration_group, and registrant, then filled right with '9's to thirteen digits
			local element_lengths_t = {};										-- sequence of lengths for registration group, registrant, and publication elements in an isbn

			table.insert (isbn_parts_t, prefix);								-- add prefix (1)  This applies to each <Rule> in the <Group>
			table.insert (isbn_parts_t, registration_group);					-- add registration group (2)  This applies to each <Rule> in the <Group>

			table.insert (element_lengths_t, #registration_group);				-- add length of isbn registration group (1)  This applies to each <Rule> in the <Group>
			local range_min, range_max = rule:match (range_pattern);			-- get range limits; range_min not currently used
			local registrant_len = rule:match (length_pattern);					-- get the length of the registrant length

			if '0' ~= registrant_len then
				table.insert (element_lengths_t, tonumber(registrant_len));		-- add length of isbn registrant (2)
				table.insert (element_lengths_t, 9-#registration_group-registrant_len);	-- calculate and add length available for publication element (3)
				
				table.insert (isbn_parts_t, range_max:sub (1, registrant_len));	-- add <registrant_len> digits from left side of <range_max> (3)

				local fill_length = 0;											-- calculate how many '9's are needed to right-fill an isbn
				for _, v in ipairs (isbn_parts_t) do							-- for each part get its length
					fill_length = fill_length + #(tostring (v));				-- and accumulate
				end
				fill_length = 13 - fill_length;									-- and figure how many are needed
				table.insert (isbn_parts_t, tonumber (string.rep ('9', fill_length)));	-- and right fill to 13 digits
				
				local pretty_string = '\t[' .. table.concat (isbn_parts_t) ..'] = {' .. table.concat (element_lengths_t, ', ') .. '}';	-- prettify
				table.insert (out_t, pretty_string);							-- add the prettified string to the output sequence
			end
		end
	end

	table.insert (out_t, '\t}</pre>');											-- to close the table
	return '<pre>local hyphen_pos_t = {\n' .. table.concat (out_t, ',\n');		-- make a big string and done	
end


--[[--------------------------< E X P O R T S >----------------------------------------------------------------
]]

return {
	range_message_xlate = range_message_xlate,
	}