Module:Citation/CS1/Identifiers

    From Nonbinary Wiki

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

    
    --[[--------------------------< F O R W A R D   D E C L A R A T I O N S >--------------------------------------
    ]]
    
    local has_accept_as_written, is_set, in_array, set_message, select_one,			-- functions in Module:Citation/CS1/Utilities
    		substitute, make_wikilink;
    
    local z;																		-- table of tables defined in Module:Citation/CS1/Utilities
    
    local cfg;																		-- table of configuration tables that are defined in Module:Citation/CS1/Configuration
    
    
    --[[--------------------------< P A G E   S C O P E   V A R I A B L E S >--------------------------------------
    
    declare variables here that have page-wide scope that are not brought in from other modules; that are created here and used here
    
    ]]
    
    local auto_link_urls = {};														-- holds identifier URLs for those identifiers that can auto-link |title=
    
    
    --============================<< H E L P E R   F U N C T I O N S >>============================================
    
    --[[--------------------------< W I K I D A T A _ A R T I C L E _ N A M E _ G E T >----------------------------
    
    as an aid to internationalizing identifier-label wikilinks, gets identifier article names from Wikidata.
    
    returns :<lang code>:<article title> when <q> has an <article title> for <lang code>; nil else
    
    for identifiers that do not have q, returns nil
    
    for wikis that do not have mw.wikibase installed, returns nil
    
    ]]
    
    local function wikidata_article_name_get (q)
    	if not is_set (q) or (q and not mw.wikibase) then							-- when no q number or when a q number but mw.wikibase not installed on this wiki
    		return nil;																-- abandon
    	end
    
    	local wd_article;
    	local this_wiki_code = cfg.this_wiki_code;									-- Wikipedia subdomain; 'en' for en.wikipedia.org
    
    	wd_article = mw.wikibase.getSitelink (q, this_wiki_code .. 'wiki');			-- fetch article title from WD; nil when no title available at this wiki
    
    	if wd_article then
    		wd_article = table.concat ({':', this_wiki_code, ':', wd_article});		-- interwiki-style link without brackets if taken from WD; leading colon required
    	end
    
    	return wd_article;															-- article title from WD; nil else
    end
    
    
    --[[--------------------------< L I N K _ L A B E L _ M A K E >------------------------------------------------
    
    common function to create identifier link label from handler table or from Wikidata
    
    returns the first available of
    	1. redirect from local wiki's handler table (if enabled)
    	2. Wikidata (if there is a Wikidata entry for this identifier in the local wiki's language)
    	3. label specified in the local wiki's handler table
    	
    ]]
    
    local function link_label_make (handler)
    	local wd_article;
    	
    	if not (cfg.use_identifier_redirects and is_set (handler.redirect)) then	-- redirect has priority so if enabled and available don't fetch from Wikidata because expensive
    		wd_article = wikidata_article_name_get (handler.q);						-- if Wikidata has an article title for this wiki, get it;
    	end
    	
    	return (cfg.use_identifier_redirects and is_set (handler.redirect) and handler.redirect) or wd_article or handler.link;
    end
    
    
    --[[--------------------------< 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;
    	local ext_link;
    	local this_wiki_code = cfg.this_wiki_code;									-- Wikipedia subdomain; 'en' for en.wikipedia.org
    	local wd_article;															-- article title from Wikidata
    	
    	if options.encode == true or options.encode == nil then
    		url_string = mw.uri.encode (url_string, 'PATH');
    	end
    
    	if options.auto_link and is_set (options.access) then
    		auto_link_urls[options.auto_link] = table.concat ({options.prefix, url_string, options.suffix});
    	end
    
    	ext_link = mw.ustring.format ('[%s%s%s %s]', options.prefix, url_string, options.suffix or "", mw.text.nowiki (options.id));
    	if is_set (options.access) then
    		ext_link = substitute (cfg.presentation['ext-link-access-signal'], {cfg.presentation[options.access].class, cfg.presentation[options.access].title, ext_link});	-- add the free-to-read / paywall lock
    	end
    
    	return table.concat	({
    		make_wikilink (link_label_make (options), options.label),				-- redirect, Wikidata link, or locally specified link (in that order)
    		options.separator or '&nbsp;',
    		ext_link
    		});
    end
    
    
    --[[--------------------------< I N T E R N A L _ L I N K _ I D >----------------------------------------------
    
    Formats a wiki-style internal link
    
    TODO: Does not currently need to support options.access, options.encode, auto-linking and COinS (as in external_link_id),
    but may be needed in the future for :m:Interwiki_map custom-prefixes like :arxiv:, :bibcode:, :DOI:, :hdl:, :ISSN:,
    :JSTOR:, :Openlibrary:, :PMID:, :RFC:.
    
    ]]
    
    local function internal_link_id (options)
    	local id = mw.ustring.gsub (options.id, '%d', cfg.date_names.local_digits);	-- translate 'local' digits to Western 0-9
    
    	return table.concat (
    		{
    		make_wikilink (link_label_make (options), options.label),				-- wiki-link the identifier label
    		options.separator or '&nbsp;',											-- add the separator
    		make_wikilink (
    			table.concat (
    				{
    				options.prefix,
    				id,																-- translated to Western digits
    				options.suffix or ''
    				}),
    			substitute (cfg.presentation['bdi'], {'', mw.text.nowiki (options.id)})	-- bdi tags to prevent Latin script identifiers from being reversed at RTL language wikis
    			);																	-- nowiki because MediaWiki still has magic links for ISBN and the like; TODO: is it really required?
    		});
    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 |pmc-embargo-date= against
    today's date.  If embargo date is in the future, returns the content of |pmc-embargo-date=; otherwise, returns
    an empty string because the embargo has expired or because |pmc-embargo-date= was not set in this cite.
    
    ]]
    
    local function is_embargoed (embargo)
    	if is_set (embargo) then
    		local lang = mw.getContentLanguage();
    		local good1, embargo_date, todays_date;
    		good1, embargo_date = pcall (lang.formatDate, lang, 'U', embargo);
    		todays_date = lang:formatDate ('U');
    	
    		if good1 then															-- if embargo date is a good date
    			if tonumber (embargo_date) >= tonumber (todays_date) then			-- is embargo date is in the future?
    				return embargo;													-- still embargoed
    			else
    				set_message ('maint_pmc_embargo');								-- embargo has expired; add main cat
    				return '';														-- unset because embargo has expired
    			end
    		end
    	end
    	return '';																	-- |pmc-embargo-date= not set return empty string
    end
    
    
    --[=[-------------------------< I S _ V A L I D _ B I O R X I V _ D A T E >------------------------------------
    
    returns true if:
    	2019-12-11T00:00Z <= biorxiv_date < today + 2 days
    	
    The dated form of biorxiv identifier has a start date of 2019-12-11.  The Unix timestamp for that date is {{#time:U|2019-12-11}} = 1576022400
    
    biorxiv_date is the date provided in those |biorxiv= parameter values that are dated at time 00:00:00 UTC
    today is the current date at time 00:00:00 UTC plus 48 hours
    	if today is 2015-01-01T00:00:00 then
    		adding 24 hours gives 2015-01-02T00:00:00 – one second more than today
    		adding 24 hours gives 2015-01-03T00:00:00 – one second more than tomorrow
    
    This function does not work if it is fed month names for languages other than English.  Wikimedia #time: parser
    apparently doesn't understand non-English date month names. This function will always return false when the date
    contains a non-English month name because good1 is false after the call to lang.formatDate().  To get around that
    call this function with YYYY-MM-DD format dates.
    
    ]=]
    
    local function is_valid_biorxiv_date (biorxiv_date)
    	local good1, good2;
    	local biorxiv_ts, tomorrow_ts;												-- to hold Unix timestamps representing the dates
    	local lang_object = mw.getContentLanguage();
    
    	good1, biorxiv_ts = pcall (lang_object.formatDate, lang_object, 'U', biorxiv_date);		-- convert biorxiv_date value to Unix timestamp 
    	good2, tomorrow_ts = pcall (lang_object.formatDate, lang_object, 'U', 'today + 2 days' );	-- today midnight + 2 days is one second more than all day tomorrow
    	
    	if good1 and good2 then														-- lang.formatDate() returns a timestamp in the local script which tonumber() may not understand
    		biorxiv_ts = tonumber (biorxiv_ts) or lang_object:parseFormattedNumber (biorxiv_ts);	-- convert to numbers for the comparison;
    		tomorrow_ts = tonumber (tomorrow_ts) or lang_object:parseFormattedNumber (tomorrow_ts);
    	else
    		return false;															-- one or both failed to convert to Unix timestamp
    	end
    
    	return ((1576022400 <= biorxiv_ts) and (biorxiv_ts < tomorrow_ts))			-- 2012-12-11T00:00Z <= biorxiv_date < tomorrow's date
    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 isbn().
    
    If the number is valid the result will be 0. Before calling this function, ISBN/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
    
    
    --[[--------------------------< 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
    
    
    --============================<< I D E N T I F I E R   F U N C T I O N S >>====================================
    
    --[[--------------------------< 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 9107 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; not the same as |class= parameter which is not supported in this form
    	<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 (options)
    	local id = options.id;
    	local class = options.Class;												-- TODO: lowercase?
    	local handler = options.handler;
    	local year, month, version;
    	local err_cat = false;														-- assume no error message
    	local text;																	-- output 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 9107-0703 format with or without 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 = true;													-- flag for 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 with or without 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									-- when year is 07, is month invalid (before April)?
    				err_cat = true;													-- flag for 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 with or without 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 = true;													-- flag for error message
    		end
    
    	else
    		err_cat = true;															-- not a recognized format; flag for error message
    	end
    
    	if err_cat then
    		options.coins_list_t['ARXIV'] = nil;									-- when error, unset so not included in COinS
    	end
    	
    	err_cat = err_cat and table.concat ({' ', set_message ('err_bad_arxiv')}) or '';	-- set error message if flag is true
    	
    	text = external_link_id ({link = handler.link, label = handler.label, q = handler.q, redirect = handler.redirect,
    			prefix = handler.prefix, id = id, separator = handler.separator, encode = handler.encode, access = handler.access}) .. err_cat;
    
    	if is_set (class) then
    		if id:match ('^%d+') then
    			text = table.concat ({text, ' [[//arxiv.org/archive/', class, ' ', class, ']]'});	-- external link within square brackets, not wikilink
    		else
    			text = table.concat ({text, ' ', set_message ('err_class_ignored')});		
    		end
    	end
    
    	return text;	
    end
    
    
    --[[--------------------------< B I B C O D E >--------------------------------------------------------------------
    
    Validates (sort of) and formats a bibcode ID.
    
    Format for bibcodes is specified here: http://adsabs.harvard.edu/abs_doc/help_pages/data.html#bibcodes
    
    But, this: 2015arXiv151206696F is apparently valid so apparently, the only things that really matter are length, 19 characters
    and first four digits must be a year.  This function makes these tests:
    	length must be 19 characters
    	characters in position
    		1–4 must be digits and must represent a year in the range of 1000 – next year
    		5 must be a letter
    		6–8 must be letter, digit, ampersand, or dot (ampersand cannot directly precede a dot; &. )
    		9–18 must be letter, digit, or dot
    		19 must be a letter or dot
    
    ]]
    
    local function bibcode (options)
    	local id = options.id;
    	local access = options.access;
    	local handler = options.handler;
    	local err_type;
    	local year;
    
    	local text = external_link_id ({link = handler.link, label = handler.label, q = handler.q, redirect = handler.redirect,
    		prefix = handler.prefix, id = id, separator = handler.separator, encode = handler.encode,
    		access = access});
    	
    	if 19 ~= id:len() then
    		err_type = cfg.err_msg_supl.length;
    	else
    		year = id:match ("^(%d%d%d%d)[%a][%w&%.][%w&%.][%w&%.][%w.]+[%a%.]$");
    		if not year then														-- if nil then no pattern match
    			err_type = cfg.err_msg_supl.value;									-- so value error
    		else
    			local next_year = tonumber (os.date ('%Y')) + 1;					-- get the current year as a number and add one for next year
    			year = tonumber (year);												-- convert year portion of bibcode to a number
    			if (1000 > year) or (year > next_year) then
    				err_type = cfg.err_msg_supl.year;								-- year out of bounds
    			end
    			if id:find('&%.') then
    				err_type = cfg.err_msg_supl.journal;							-- journal abbreviation must not have '&.' (if it does it's missing a letter)
    			end
    		end
    	end
    
    	if is_set (err_type) then													-- if there was an error detected
    		text = text .. ' ' .. set_message ('err_bad_bibcode', {err_type});
    		options.coins_list_t['BIBCODE'] = nil;									-- when error, unset so not included in COinS
    
    	end
    	return text;
    end
    
    
    --[[--------------------------< B I O R X I V >-----------------------------------------------------------------
    
    Format bioRxiv ID and do simple error checking.  Before 2019-12-11, biorXiv IDs were 10.1101/ followed by exactly
    6 digits.  After 2019-12-11, biorXiv IDs retained the six-digit identifier but prefixed that with a yyyy.mm.dd. 
    date and suffixed with an optional version identifier.
    
    The bioRxiv ID is the string of characters:
    	https://doi.org/10.1101/078733 -> 10.1101/078733
    or a date followed by a six-digit number followed by an optional version indicator 'v' and one or more digits:
    	https://www.biorxiv.org/content/10.1101/2019.12.11.123456v2 -> 10.1101/2019.12.11.123456v2
    	
    see https://www.biorxiv.org/about-biorxiv
    
    ]]
    
    local function biorxiv (options)
    	local id = options.id;
    	local handler = options.handler;
    	local err_cat = true;														-- flag; assume that there will be an error
    	
    	local patterns = {
    		'^10.1101/%d%d%d%d%d%d$',												-- simple 6-digit identifier (before 2019-12-11)
    		'^10.1101/(20[1-9]%d)%.([01]%d)%.([0-3]%d)%.%d%d%d%d%d%dv%d+$',			-- y.m.d. date + 6-digit identifier + version (after 2019-12-11)
    		'^10.1101/(20[1-9]%d)%.([01]%d)%.([0-3]%d)%.%d%d%d%d%d%d$',				-- y.m.d. date + 6-digit identifier (after 2019-12-11)
    		}
    	
    	for _, pattern in ipairs (patterns) do										-- spin through the patterns looking for a match
    		if id:match (pattern) then
    			local y, m, d = id:match (pattern);									-- found a match, attempt to get year, month and date from the identifier
    
    			if m then															-- m is nil when id is the six-digit form
    				if not is_valid_biorxiv_date (y .. '-' .. m .. '-' .. d) then	-- validate the encoded date; TODO: don't ignore leap-year and actual month lengths ({{#time:}} is a poor date validator)
    					break;														-- date fail; break out early so we don't unset the error message
    				end
    			end
    			err_cat = nil;														-- we found a match so unset the error message
    			break;																-- and done
    		end
    	end																			-- err_cat remains set here when no match
    
    	if err_cat then
    		options.coins_list_t['BIORXIV'] = nil;									-- when error, unset so not included in COinS
    	end
    	
    	return external_link_id ({link = handler.link, label = handler.label, q = handler.q, redirect = handler.redirect,
    			prefix = handler.prefix, id = id, separator = handler.separator,
    			encode = handler.e