Module:Citation/CS1/Identifiers

< Module:Citation/CS1
Revision as of 16:03, 15 December 2015 by m>Trappist the monk

Documentation for this module may be created at Module:Citation/CS1/Identifiers/doc

local utilities = require ('Module:Citation/CS1/Utilities');

local identifiers = {};
local z ={};																	-- tables in Module:Citation/CS1/Utilities

local cfg;			-- table of configuration tables that are defined in Module:Citation/CS1/Configuration

--[[--------------------------< E X T E R N A L _ L I N K _ I D >----------------------------------------------

Formats a wiki style external link

]]

local function external_link_id(options)
	local url_string = options.id;
	if options.encode == true or options.encode == nil then
		url_string = mw.uri.encode( url_string );
	end
	return mw.ustring.format( '[[%s|%s]]%s[%s%s%s %s]',
		options.link, options.label, options.separator or "&nbsp;",
		options.prefix, url_string, options.suffix or "",
		mw.text.nowiki(options.id)
	);
end


--[[--------------------------< I N T E R N A L _ L I N K _ I D >----------------------------------------------

Formats a wiki style internal link

]]

local function internal_link_id(options)
	return mw.ustring.format( '[[%s|%s]]%s[[%s%s%s|%s]]',
		options.link, options.label, options.separator or "&nbsp;",
		options.prefix, options.id, options.suffix or "",
		mw.text.nowiki(options.id)
	);
end


--[[--------------------------< IS _ V A L I D _ I S X N >-----------------------------------------------------

ISBN-10 and ISSN validator code calculates checksum across all isbn/issn digits including the check digit.
ISBN-13 is checked in check_isbn().

If the number is valid the result will be 0. Before calling this function, issbn/issn must be checked for length
and stripped of dashes, spaces and other non-isxn characters.

]]

local function is_valid_isxn (isxn_str, len)
	local temp = 0;
	isxn_str = { isxn_str:byte(1, len) };	-- make a table of byte values '0' → 0x30 .. '9'  → 0x39, 'X' → 0x58
	len = len+1;							-- adjust to be a loop counter
	for i, v in ipairs( isxn_str ) do		-- loop through all of the bytes and calculate the checksum
		if v == string.byte( "X" ) then		-- if checkdigit is X (compares the byte value of 'X' which is 0x58)
			temp = temp + 10*( len - i );	-- it represents 10 decimal
		else
			temp = temp + tonumber( string.char(v) )*(len-i);
		end
	end
	return temp % 11 == 0;					-- returns true if calculation result is zero
end


--[[--------------------------< IS _ V A L I D _ I S X N  _ 1 3 >----------------------------------------------

ISBN-13 and ISMN validator code calculates checksum across all 13 isbn/ismn digits including the check digit.
If the number is valid, the result will be 0. Before calling this function, isbn-13/ismn must be checked for length
and stripped of dashes, spaces and other non-isxn-13 characters.

]]

local function is_valid_isxn_13 (isxn_str)
	local temp=0;
	
	isxn_str = { isxn_str:byte(1, 13) };										-- make a table of byte values '0' → 0x30 .. '9'  → 0x39
	for i, v in ipairs( isxn_str ) do
		temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) );				-- multiply odd index digits by 1, even index digits by 3 and sum; includes check digit
	end
	return temp % 10 == 0;														-- sum modulo 10 is zero when isbn-13/ismn is correct
end


--[[--------------------------< C H E C K _ I S B N >------------------------------------------------------------

Determines whether an ISBN string is valid

]]

local function check_isbn( isbn_str )
	if nil ~= isbn_str:match("[^%s-0-9X]") then return false; end		-- fail if isbn_str contains anything but digits, hyphens, or the uppercase X
	isbn_str = isbn_str:gsub( "-", "" ):gsub( " ", "" );	-- remove hyphens and spaces
	local len = isbn_str:len();
 
	if len ~= 10 and len ~= 13 then
		return false;
	end

	if len == 10 then
		if isbn_str:match( "^%d*X?$" ) == nil then return false; end
		return is_valid_isxn(isbn_str, 10);
	else
		local temp = 0;
		if isbn_str:match( "^97[89]%d*$" ) == nil then return false; end		-- isbn13 begins with 978 or 979; ismn begins with 979
		return is_valid_isxn_13 (isbn_str);
	end
end


--[[--------------------------< I S M N >----------------------------------------------------------------------

Determines whether an ISMN string is valid.  Similar to isbn-13, ismn is 13 digits begining 979-0-... and uses the
same check digit calculations.  See http://www.ismn-international.org/download/Web_ISMN_Users_Manual_2008-6.pdf
section 2, pages 9–12.

]]

local function ismn (id)
	local handler = cfg.id_handlers['ISMN'];
	local text;
	local valid_ismn = true;

	id=id:gsub( "[%s-–]", "" );													-- strip spaces, hyphens, and endashes from the ismn

	if 13 ~= id:len() or id:match( "^9790%d*$" ) == nil then					-- ismn must be 13 digits and begin 9790
		valid_ismn = false;
	else
		valid_ismn=is_valid_isxn_13 (id);										-- validate ismn
	end

--	text = internal_link_id({link = handler.link, label = handler.label,		-- use this (or external version) when there is some place to link to
--		prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 
	text="[[" .. handler.link .. "|" .. handler.label .. "]]" .. handler.separator .. id;		-- because no place to link to yet

	if false == valid_ismn then
		text = text .. ' ' .. set_error( 'bad_ismn' )							-- add an error message if the issn is invalid
	end 
	
	return text;
end


--[[--------------------------< I S S N >----------------------------------------------------------------------

Validate and format an issn.  This code fixes the case where an editor has included an ISSN in the citation but
has separated the two groups of four digits with a space.  When that condition occurred, the resulting link looked
like this:

	|issn=0819 4327 gives: [http://www.worldcat.org/issn/0819 4327 0819 4327]  -- can't have spaces in an external link
	
This code now prevents that by inserting a hyphen at the issn midpoint.  It also validates the issn for length
and makes sure that the checkdigit agrees with the calculated value.  Incorrect length (8 digits), characters
other than 0-9 and X, or checkdigit / calculated value mismatch will all cause a check issn error message.  The
issn is always displayed with a hyphen, even if the issn was given as a single group of 8 digits.

]]

local function issn(id)
	local issn_copy = id;		-- save a copy of unadulterated issn; use this version for display if issn does not validate
	local handler = cfg.id_handlers['ISSN'];
	local text;
	local valid_issn = true;

	id=id:gsub( "[%s-–]", "" );									-- strip spaces, hyphens, and endashes from the issn

	if 8 ~= id:len() or nil == id:match( "^%d*X?$" ) then		-- validate the issn: 8 digits long, containing only 0-9 or X in the last position
		valid_issn=false;										-- wrong length or improper character
	else
		valid_issn=is_valid_isxn(id, 8);						-- validate issn
	end

	if true == valid_issn then
		id = string.sub( id, 1, 4 ) .. "-" .. string.sub( id, 5 );	-- if valid, display correctly formatted version
	else
		id = issn_copy;											-- if not valid, use the show the invalid issn with error message
	end
	
	text = external_link_id({link = handler.link, label = handler.label,
		prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 
	if false == valid_issn then
		text = text .. ' ' .. set_error( 'bad_issn' )			-- add an error message if the issn is invalid
	end 
	
	return text
end


--[[--------------------------< A M A Z O N >------------------------------------------------------------------

Formats a link to Amazon.  Do simple error checking: asin must be mix of 10 numeric or uppercase alpha
characters.  If a mix, first character must be uppercase alpha; if all numeric, asins must be 10-digit
isbn. If 10-digit isbn, add a maintenance category so a bot or awb script can replace |asin= with |isbn=.
Error message if not 10 characters, if not isbn10, if mixed and first character is a digit.

]]

local function amazon(id, domain)
	local err_cat = ""

	if not id:match("^[%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u]$") then
		err_cat =  ' ' .. set_error ('bad_asin');								-- asin is not a mix of 10 uppercase alpha and numeric characters
	else
		if id:match("^%d%d%d%d%d%d%d%d%d[%dX]$") then								-- if 10-digit numeric (or 9 digits with terminal X)
			if check_isbn( id ) then												-- see if asin value is isbn10
				add_maint_cat ('ASIN');
			elseif not is_set (err_cat) then
				err_cat =  ' ' .. set_error ('bad_asin');						-- asin is not isbn10
			end
		elseif not id:match("^%u[%d%u]+$") then
			err_cat =  ' ' .. set_error ('bad_asin');							-- asin doesn't begin with uppercase alpha
		end
	end
	if not is_set(domain) then 
		domain = "com";
	elseif in_array (domain, {'jp', 'uk'}) then			-- Japan, United Kingdom
		domain = "co." .. domain;
	elseif in_array (domain, {'au', 'br', 'mx'}) then	-- Australia, Brazil, Mexico
		domain = "com." .. domain;
	end
	local handler = cfg.id_handlers['ASIN'];
	return external_link_id({link=handler.link,
		label=handler.label, prefix=handler.prefix .. domain .. "/dp/",
		id=id, encode=handler.encode, separator = handler.separator}) .. err_cat;
end


--[[--------------------------< A R X I V >--------------------------------------------------------------------

See: http://arxiv.org/help/arxiv_identifier

format and error check arXiv identifier.  There are three valid forms of the identifier:
the first form, valid only between date codes 9108 and 0703 is:
	arXiv:<archive>.<class>/<date code><number><version>
where:
	<archive> is a string of alpha characters - may be hyphenated; no other punctuation
	<class> is a string of alpha characters - may be hyphenated; no other punctuation
	<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
		first digit of YY for this form can only 9 and 0
	<number> is a three-digit number
	<version> is a 1 or more digit number preceded with a lowercase v; no spaces (undocumented)
	
the second form, valid from April 2007 through December 2014 is:
	arXiv:<date code>.<number><version>
where:
	<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
	<number> is a four-digit number
	<version> is a 1 or more digit number preceded with a lowercase v; no spaces

the third form, valid from January 2015 is:
	arXiv:<date code>.<number><version>
where:
	<date code> and <version> are as defined for 0704-1412
	<number> is a five-digit number
]]

local function arxiv (id, class)
	local handler = cfg.id_handlers['ARXIV'];
	local year, month, version;
	local err_cat = '';
	local text;
	
	if id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%d$") or id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%dv%d+$") then	-- test for the 9108-0703 format w/ & w/o version
		year, month = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d[v%d]*$");
		year = tonumber(year);
		month = tonumber(month);
		if ((not (90 < year or 8 > year)) or (1 > month or 12 < month)) or		-- if invalid year or invalid month
			((91 == year and 7 > month) or (7 == year and 3 < month)) then		-- if years ok, are starting and ending months ok?
				err_cat = ' ' .. set_error( 'bad_arxiv' );						-- set error message
		end
	elseif id:match("^%d%d[01]%d%.%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%dv%d+$") then	-- test for the 0704-1412 w/ & w/o version
		year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d[v%d]*$");
		year = tonumber(year);
		month = tonumber(month);
		if ((7 > year) or (14 < year) or (1 > month or 12 < month)) or			-- is year invalid or is month invalid? (doesn't test for future years)
			((7 == year) and (4 > month)) then --or									-- when year is 07, is month invalid (before April)?
				err_cat = ' ' .. set_error( 'bad_arxiv' );						-- set error message
		end
	elseif id:match("^%d%d[01]%d%.%d%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%d%dv%d+$") then	-- test for the 1501- format w/ & w/o version
		year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d%d[v%d]*$");
		year = tonumber(year);
		month = tonumber(month);
		if ((15 > year) or (1 > month or 12 < month)) then						-- is year invalid or is month invalid? (doesn't test for future years)
			err_cat = ' ' .. set_error( 'bad_arxiv' );							-- set error message
		end
	else
		err_cat = ' ' .. set_error( 'bad_arxiv' );								-- arXiv id doesn't match any format
	end

	text = external_link_id({link = handler.link, label = handler.label,
			prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;

	if is_set (class) then
		class = ' [[' .. '//arxiv.org/archive/' .. class .. ' ' .. class .. ']]';	-- external link within square brackets, not wikilink
	else
		class = '';																-- empty string for concatenation
	end
	
	return text .. class;
end


--[[--------------------------< N O R M A L I Z E _ L C C N >--------------------------------------------------

lccn normalization (http://www.loc.gov/marc/lccn-namespace.html#normalization)
1. Remove all blanks.
2. If there is a forward slash (/) in the string, remove it, and remove all characters to the right of the forward slash.
3. If there is a hyphen in the string:
	a. Remove it.
	b. Inspect the substring following (to the right of) the (removed) hyphen. Then (and assuming that steps 1 and 2 have been carried out):
		1. All these characters should be digits, and there should be six or less. (not done in this function)
		2. If the length of the substring is less than 6, left-fill the substring with zeroes until the length is six.

Returns a normalized lccn for lccn() to validate.  There is no error checking (step 3.b.1) performed in this function.
]]

local function normalize_lccn (lccn)
	lccn = lccn:gsub ("%s", "");									-- 1. strip whitespace

	if nil ~= string.find (lccn,'/') then
		lccn = lccn:match ("(.-)/");								-- 2. remove forward slash and all character to the right of it
	end

	local prefix
	local suffix
	prefix, suffix = lccn:match ("(.+)%-(.+)");						-- 3.a remove hyphen by splitting the string into prefix and suffix

	if nil ~= suffix then											-- if there was a hyphen
		suffix=string.rep("0", 6-string.len (suffix)) .. suffix;	-- 3.b.2 left fill the suffix with 0s if suffix length less than 6
		lccn=prefix..suffix;										-- reassemble the lccn
	end
	
	return lccn;
	end


--[[--------------------------< L C C N >----------------------------------------------------------------------

Format LCCN link and do simple error checking.  LCCN is a character string 8-12 characters long. The length of
the LCCN dictates the character type of the first 1-3 characters; the rightmost eight are always digits.
http://info-uri.info/registry/OAIHandler?verb=GetRecord&metadataPrefix=reg&identifier=info:lccn/

length = 8 then all digits
length = 9 then lccn[1] is lower case alpha
length = 10 then lccn[1] and lccn[2] are both lower case alpha or both digits
length = 11 then lccn[1] is lower case alpha, lccn[2] and lccn[3] are both lower case alpha or both digits
length = 12 then lccn[1] and lccn[2] are both lower case alpha

]]

local function lccn(lccn)
	local handler = cfg.id_handlers['LCCN'];
	local err_cat =  '';								-- presume that LCCN is valid
	local id = lccn;									-- local copy of the lccn

	id = normalize_lccn (id);							-- get canonical form (no whitespace, hyphens, forward slashes)
	local len = id:len();								-- get the length of the lccn

	if 8 == len then
		if id:match("[^%d]") then						-- if LCCN has anything but digits (nil if only digits)
			err_cat = ' ' .. set_error( 'bad_lccn' );	-- set an error message
		end
	elseif 9 == len then								-- LCCN should be adddddddd
		if nil == id:match("%l%d%d%d%d%d%d%d%d") then			-- does it match our pattern?
			err_cat = ' ' .. set_error( 'bad_lccn' );	-- set an error message
		end
	elseif 10 == len then								-- LCCN should be aadddddddd or dddddddddd
		if id:match("[^%d]") then							-- if LCCN has anything but digits (nil if only digits) ...
			if nil == id:match("^%l%l%d%d%d%d%d%d%d%d") then	-- ... see if it matches our pattern
				err_cat = ' ' .. set_error( 'bad_lccn' );	-- no match, set an error message
			end
		end
	elseif 11 == len then								-- LCCN should be aaadddddddd or adddddddddd
		if not (id:match("^%l%l%l%d%d%d%d%d%d%d%d") or id:match("^%l%d%d%d%d%d%d%d%d%d%d")) then	-- see if it matches one of our patterns
			err_cat = ' ' .. set_error( 'bad_lccn' );	-- no match, set an error message
		end
	elseif 12 == len then								-- LCCN should be aadddddddddd
		if not id:match("^%l%l%d%d%d%d%d%d%d%d%d%d") then	-- see if it matches our pattern
			err_cat = ' ' .. set_error( 'bad_lccn' );	-- no match, set an error message
		end
	else
		err_cat = ' ' .. set_error( 'bad_lccn' );		-- wrong length, set an error message
	end

	if not is_set (err_cat) and nil ~= lccn:find ('%s') then
		err_cat = ' ' .. set_error( 'bad_lccn' );		-- lccn contains a space, set an error message
	end

	return external_link_id({link = handler.link, label = handler.label,
			prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
end


--[[--------------------------< P M I D >----------------------------------------------------------------------

Format PMID and do simple error checking.  PMIDs are sequential numbers beginning at 1 and counting up.  This
code checks the PMID to see that it contains only digits and is less than test_limit; the value in local variable
test_limit will need to be updated periodically as more PMIDs are issued.

]]

local function pmid(id)
	local test_limit = 30000000;						-- update this value as PMIDs approach
	local handler = cfg.id_handlers['PMID'];
	local err_cat =  '';								-- presume that PMID is valid
	
	if id:match("[^%d]") then							-- if PMID has anything but digits
		err_cat = ' ' .. set_error( 'bad_pmid' );		-- set an error message
	else												-- PMID is only digits
		local id_num = tonumber(id);					-- convert id to a number for range testing
		if 1 > id_num or test_limit < id_num then		-- if PMID is outside test limit boundaries
			err_cat = ' ' .. set_error( 'bad_pmid' );	-- set an error message
		end
	end
	
	return external_link_id({link = handler.link, label = handler.label,
			prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
end


--[[--------------------------< I S _ E M B A R G O E D >------------------------------------------------------

Determines if a PMC identifier's online version is embargoed. Compares the date in |embargo= against today's date.  If embargo date is
in the future, returns the content of |embargo=; otherwise, returns and empty string because the embargo has expired or because
|embargo= was not set in this cite.

]]

local function is_embargoed (embargo)
	if is_set (embargo) then
		local lang = mw.getContentLanguage();
		local good1, embargo_date, good2, todays_date;
		good1, embargo_date = pcall( lang.formatDate, lang, 'U', embargo );
		good2, todays_date = pcall( lang.formatDate, lang, 'U' );
	
		if good1 and good2 then													-- if embargo date and today's date are good dates
			if tonumber( embargo_date ) >= tonumber( todays_date ) then			-- is embargo date is in the future?
				return embargo;													-- still embargoed
			else
				add_maint_cat ('embargo')
				return '';														-- unset because embargo has expired
			end
		end
	end
	return '';																	-- |embargo= not set return empty string
end


--[[--------------------------< P M C >------------------------------------------------------------------------

Format a PMC, do simple error checking, and check for embargoed articles.

The embargo parameter takes a date for a value. If the embargo date is in the future the PMC identifier will not
be linked to the article.  If the embargo date is today or in the past, or if it is empty or omitted, then the
PMC identifier is linked to the article through the link at cfg.id_handlers['PMC'].prefix.

PMC embargo date testing is done in function is_embargoed () which is called earlier because when the citation
has |pmc=<value> but does not have a |url= then |title= is linked with the PMC link.  Function is_embargoed ()
returns the embargo date if the PMC article is still embargoed, otherwise it returns an empty string.

PMCs are sequential numbers beginning at 1 and counting up.  This code checks the PMC to see that it contains only digits and is less
than test_limit; the value in local variable test_limit will need to be updated periodically as more PMCs are issued.

]]

local function pmc(id, embargo)
	local test_limit = 5000000;							-- update this value as PMCs approach
	local handler = cfg.id_handlers['PMC'];
	local err_cat =  '';								-- presume that PMC is valid
	
	local text;

	if id:match("[^%d]") then							-- if PMC has anything but digits
		err_cat = ' ' .. set_error( 'bad_pmc' );			-- set an error message
	else												-- PMC is only digits
		local id_num = tonumber(id);					-- convert id to a number for range testing
		if 1 > id_num or test_limit < id_num then		-- if PMC is outside test limit boundaries
			err_cat = ' ' .. set_error( 'bad_pmc' );		-- set an error message
		end
	end
	
	if is_set (embargo) then													-- is PMC is still embargoed?
		text="[[" .. handler.link .. "|" .. handler.label .. "]]:" .. handler.separator .. id .. err_cat;	-- still embargoed so no external link
	else
		text = external_link_id({link = handler.link, label = handler.label,			-- no embargo date or embargo has expired, ok to link to article
			prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
	end
	return text;
end


--[[--------------------------< D O I >------------------------------------------------------------------------

Formats a DOI and checks for DOI errors.

DOI names contain two parts: prefix and suffix separated by a forward slash.
	Prefix: directory indicator '10.' followed by a registrant code
	Suffix: character string of any length chosen by the registrant

This function checks a DOI name for: prefix/suffix.  If the doi name contains spaces or endashes, or, if it ends
with a period or a comma, this function will emit a bad_doi error message.

DOI names are case-insensitive and can incorporate any printable Unicode characters so the test for spaces, endash,
and terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely
if ever used in doi names.

]]

local function doi(id, inactive)
	local cat = ""
	local handler = cfg.id_handlers['DOI'];
	
	local text;
	if is_set(inactive) then
		local inactive_year = inactive:match("%d%d%d%d") or '';		-- try to get the year portion from the inactive date
		text = "[[" .. handler.link .. "|" .. handler.label .. "]]:" .. id;
		if is_set(inactive_year) then
			table.insert( z.error_categories, "Pages with DOIs inactive since " .. inactive_year );
		else
			table.insert( z.error_categories, "Pages with inactive DOIs" );	-- when inactive doesn't contain a recognizable year
		end
		inactive = " (" .. cfg.messages['inactive'] .. " " .. inactive .. ")" 
	else 
		text = external_link_id({link = handler.link, label = handler.label,
			prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
		inactive = "" 
	end

	if nil == id:match("^10%.[^%s–]-/[^%s–]-[^%.,]$") then	-- doi must begin with '10.', must contain a fwd slash, must not contain spaces or endashes, and must not end with period or comma
		cat = ' ' .. set_error( 'bad_doi' );
	end
	return text .. inactive .. cat 
end


--[[--------------------------< O P E N L I B R A R Y >--------------------------------------------------------

Formats an OpenLibrary link, and checks for associated errors.

]]

local function openlibrary(id)
	local code = id:match("^%d+([AMW])$");										-- only digits followed by 'A', 'M', or 'W'
	local handler = cfg.id_handlers['OL'];

	if ( code == "A" ) then
		return external_link_id({link=handler.link, label=handler.label,
			prefix=handler.prefix .. 'authors/OL',
			id=id, separator=handler.separator,	encode = handler.encode})
	elseif ( code == "M" ) then
		return external_link_id({link=handler.link, label=handler.label,
			prefix=handler.prefix .. 'books/OL',
			id=id, separator=handler.separator,	encode = handler.encode})
	elseif ( code == "W" ) then
		return external_link_id({link=handler.link, label=handler.label,
			prefix=handler.prefix .. 'works/OL',
			id=id, separator=handler.separator,	encode = handler.encode})
	else
		return external_link_id({link=handler.link, label=handler.label,
			prefix=handler.prefix .. 'OL',
			id=id, separator=handler.separator,	encode = handler.encode}) .. ' ' .. set_error( 'bad_ol' );
	end
end


--[[--------------------------< M E S S A G E _ I D >----------------------------------------------------------

Validate and format a usenet message id.  Simple error checking, looks for 'id-left@id-right' not enclosed in
'<' and/or '>' angle brackets.

]]

local function message_id (id)
	local handler = cfg.id_handlers['USENETID'];

	local text = external_link_id({link = handler.link, label = handler.label,
		prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 
	if not id:match('^.+@.+$') or not id:match('^[^<].*[^>]$')then				-- doesn't have '@' or has one or first or last character is '< or '>'
		text = text .. ' ' .. set_error( 'bad_message_id' )						-- add an error message if the message id is invalid
	end 
	
	return text
end


--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------

Takes a table of IDs created by extract_ids() and turns it into a table of formatted ID outputs.

inputs:
	id_list – table of identifiers built by extract_ids()
	options – table of various template parameter values used to modify some manually handled identifiers

]]

local function build_id_list( id_list, options )
	local new_list, handler = {};

	local function fallback(k) return { __index = function(t,i) return cfg.id_handlers[k][i] end } end;
	
	for k, v in pairs( id_list ) do												-- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
		-- fallback to read-only cfg
		handler = setmetatable( { ['id'] = v }, fallback(k) );
		
		if handler.mode == 'external' then
			table.insert( new_list, {handler.label, external_link_id( handler ) } );
		elseif handler.mode == 'internal' then
			table.insert( new_list, {handler.label, internal_link_id( handler ) } );
		elseif handler.mode ~= 'manual' then
			error( cfg.messages['unknown_ID_mode'] );
		elseif k == 'DOI' then
			table.insert( new_list, {handler.label, doi( v, options.DoiBroken ) } );
		elseif k == 'ARXIV' then
			table.insert( new_list, {handler.label, arxiv( v, options.Class ) } ); 
		elseif k == 'ASIN' then
			table.insert( new_list, {handler.label, amazon( v, options.ASINTLD ) } ); 
		elseif k == 'LCCN' then
			table.insert( new_list, {handler.label, lccn( v ) } );
		elseif k == 'OL' or k == 'OLA' then
			table.insert( new_list, {handler.label, openlibrary( v ) } );
		elseif k == 'PMC' then
			table.insert( new_list, {handler.label, pmc( v, options.Embargo ) } );
		elseif k == 'PMID' then
			table.insert( new_list, {handler.label, pmid( v ) } );
		elseif k == 'ISMN' then
			table.insert( new_list, {handler.label, ismn( v ) } );
		elseif k == 'ISSN' then
			table.insert( new_list, {handler.label, issn( v ) } );
		elseif k == 'ISBN' then
			local ISBN = internal_link_id( handler );
			if not check_isbn( v ) and not is_set(options.IgnoreISBN) then
				ISBN = ISBN .. set_error( 'bad_isbn', {}, false, " ", "" );
			end
			table.insert( new_list, {handler.label, ISBN } );				
		elseif k == 'USENETID' then
			table.insert( new_list, {handler.label, message_id( v ) } );
		else
			error( cfg.messages['unknown_manual_ID'] );
		end
	end
	
	local function comp( a, b )	-- used in following table.sort()
		return a[1] < b[1];
	end
	
	table.sort( new_list, comp );
	for k, v in ipairs( new_list ) do
		new_list[k] = v[2];
	end
	
	return new_list;
end


--[[--------------------------< E X T R A C T _ I D S >------------------------------------------------------------

Populates ID table from arguments using configuration settings. Loops through cfg.id_handlers and searches args for
any of the parameters listed in each cfg.id_handlers['...'].parameters.  If found, adds the parameter and value to
the identifier list.  Emits redundant error message is more than one alias exists in args

]]

local function extract_ids( args )
	local id_list = {};															-- list of identifiers found in args
	for k, v in pairs( cfg.id_handlers ) do										-- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
		v = select_one( args, v.parameters, 'redundant_parameters' );			-- v.parameters is a table of aliases for k; here we pick one from args if present
		if is_set(v) then id_list[k] = v; end									-- if found in args, add identifier to our list
	end
	return id_list;
end


--[[--------------------------< S E T _ C F G >----------------------------------------------------------------

Sets local cfg table to same as that used by the other modules

]]

local function set_cfg (cfg_table_ptr)
	cfg = cfg_table_ptr;
	 utilities = require ('Module:Citation/CS1/Utilities')
end

--[[--------------------------< F O R W A R D   D E C L A R A T I O N S >--------------------------------------
]]

local is_set = utilities.is_set	;												-- functions in Module:Citation/CS1/Utilities
local in_array = utilities.in_array;
local set_error = utilities.set_error;
local select_one = utilities.select_one;
local add_maint_cat = utilities.add_maint_cat;



return {
	build_id_list = build_id_list,
	extract_ids = extract_ids,
	is_embargoed = is_embargoed;
	set_cfg = set_cfg;
	}