Difference between revisions of "Module:Citation/CS1"

From WikiMD
Jump to navigation Jump to search
(Created page with " --[[--------------------------< F O R W A R D D E C L A R A T I O N S >-------------------------------------- ]] local dates, year_date_check, reformat_dates, date_hyphen_...")
Line 1: Line 1:
  
--[[--------------------------< F O R W A R D  D E C L A R A T I O N S >--------------------------------------
+
local z = {
]]
+
error_categories = {}; -- for categorizing citations that contain errors
 +
error_ids = {};
 +
message_tail = {};
 +
maintenance_cats = {}; -- for categorizing citations that aren't erroneous per se, but could use a little work
 +
properties_cats = {}; -- for categorizing citations based on certain properties, language of source for instance
 +
}
  
local dates, year_date_check, reformat_dates, date_hyphen_to_dash, -- functions in Module:Citation/CS1/Date_validation
+
-- Whether variable is set or not
date_name_xlate
+
function is_set( var )
 +
return not (var == nil or var == '');
 +
end
  
local is_set, in_array, substitute, error_comment, set_error, select_one, -- functions in Module:Citation/CS1/Utilities
+
-- First set variable or nil if none
add_maint_cat, wrap_style, safe_for_italics, is_wikilink, make_wikilink;
+
function first_set(...)
 +
local list = {...};
 +
for _, var in pairs(list) do
 +
if is_set( var ) then
 +
return var;
 +
end
 +
end
 +
end
  
local z ={}; -- tables in Module:Citation/CS1/Utilities
+
-- Whether needle is in haystack
 +
function inArray( needle, haystack )
 +
if needle == nil then
 +
return false;
 +
end
 +
for n,v in ipairs( haystack ) do
 +
if v == needle then
 +
return n;
 +
end
 +
end
 +
return false;
 +
end
  
local extract_ids, extract_id_access_levels, build_id_list, is_embargoed; -- functions in Module:Citation/CS1/Identifiers
+
--[[
 +
Categorize and emit an error message when the citation contains one or more deprecated parameters.  Because deprecated parameters (currently |month=,
 +
|coauthor=, and |coauthors=) aren't related to each other and because these parameters may be concatenated into the variables used by |date= and |author#= (and aliases)
 +
details of which parameter caused the error message are not provided.  Only one error message is emitted regardless of the number of deprecated parameters in the citation.
 +
]]
 +
function deprecated_parameter(name)
 +
if true ~= Page_in_deprecated_cat then -- if we haven't been here before then set a
 +
Page_in_deprecated_cat=true; -- sticky flag so that if there are more than one deprecated parameter the category is added only once
 +
table.insert( z.message_tail, { seterror( 'deprecated_params', {name}, true ) } ); -- add error message
 +
end
 +
end
  
local make_coins_title, get_coins_pages, COinS; -- functions in Module:Citation/CS1/COinS
+
-- Populates numbered arguments in a message string using an argument table.
 +
function substitute( msg, args )
 +
return args and mw.message.newRawMessage( msg, args ):plain() or msg;
 +
end
  
local cfg = {}; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration
+
--[[--------------------------< K E R N _ Q U O T E S >--------------------------------------------------------
local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist
 
  
 +
Apply kerning to open the space between the quote mark provided by the Module and a leading or trailing quote mark contained in a |title= or |chapter= parameter's value.
 +
This function will positive kern either single or double quotes:
 +
"'Unkerned title with leading and trailing single quote marks'"
 +
" 'Kerned title with leading and trailing single quote marks' " (in real life the kerning isn't as wide as this example)
 +
Double single quotes (italic or bold wikimarkup) are not kerned.
  
--[[--------------------------< P A G E  S C O P E  V A R I A B L E S >--------------------------------------
+
Call this function for chapter titles, for website titles, etc; not for book titles.
 
 
delare variables here that have page-wide scope that are not brought in from other modules; thatare created here
 
and used here
 
  
 
]]
 
]]
  
local added_deprecated_cat; -- boolean flag so that the category is added only once
+
function kern_quotes (str)
local added_prop_cats = {}; -- list of property categories that have been added to z.properties_cats
+
local cap='';
local added_vanc_errs; -- boolean flag so we only emit one Vancouver error / category
+
local cap2='';
 +
 +
cap, cap2 = str:match ("^([\"\'])([^\'].+)"); -- match leading double or single quote but not double single quotes
 +
if is_set (cap) then
 +
str = substitute (cfg.presentation['kern-left'], {cap, cap2});
 +
end
  
local Frame; -- holds the module's frame table
+
cap, cap2 = str:match ("^(.+[^\'])([\"\'])$")
 +
if is_set (cap) then
 +
str = substitute (cfg.presentation['kern-right'], {cap, cap2});
 +
end
 +
return str;
 +
end
  
 +
--[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >----------------------------------------
  
--[[--------------------------< F I R S T _ S E T >------------------------------------------------------------
+
|script-title= holds title parameters that are not written in Latin based scripts: Chinese, Japanese, Arabic, Hebrew, etc. These scripts should
 +
not be italicized and may be written right-to-left.  The value supplied by |script-title= is concatenated onto Title after Title has been wrapped
 +
in italic markup.
  
Locates and returns the first set value in a table of values where the order established in the table,
+
Regardless of language, all values provided by |script-title= are wrapped in <bdi>...</bdi> tags to isolate rtl languages from the English left to right.
left-to-right (or top-to-bottom), is the order in which the values are evaluated.  Returns nil if none are set.
 
  
This version replaces the original 'for _, val in pairs do' and a similar version that used ipairs.  With the pairs
+
|script-title= provides a unique feature.  The value in |script-title= may be prefixed with a two-character ISO639-1 language code and a colon:
version the order of evaluation could not be guaranteed.  With the ipairs version, a nil value would terminate
+
|script-title=ja:*** *** (where * represents a Japanese character)
the for-loop before it reached the actual end of the list.
+
Spaces between the two-character code and the colon and the colon and the first script character are allowed:
 +
|script-title=ja : *** ***
 +
|script-title=ja: *** ***
 +
|script-title=ja :*** ***
 +
Spaces preceding the prefix are allowed: |script-title = ja:*** ***
  
]]
+
The prefix is checked for validity.  If it is a valid ISO639-1 language code, the lang attribute (lang="ja") is added to the <bdi> tag so that browsers can
 +
know the language the tag contains.  This may help the browser render the script more correctly.  If the prefix is invalid, the lang attribute
 +
is not added.  At this time there is no error message for this condition.
  
local function first_set (list, count)
+
At this writing, only |script-title= is supported. It is anticipated that additional parameters will be created to use this function.
local i = 1;
 
while i <= count do -- loop through all items in list
 
if is_set( list[i] ) then
 
return list[i]; -- return the first set list member
 
end
 
i = i + 1; -- point to next
 
end
 
end
 
 
 
 
 
--[[--------------------------< A D D _ P R O P _ C A T >--------------------------------------------------------
 
 
 
Adds a category to z.properties_cats using names from the configuration file with additional text if any.
 
 
 
foreign_lang_source and foreign_lang_source_2 keys have a language code appended to them so that multiple languages
 
may be categorized but multiples of the same language are not categorized.
 
 
 
added_prop_cats is a table declared in page scope variables above
 
  
 +
TODO: error messages when prefix is invalid ISO639-1 code; when script_value has prefix but no script;
 
]]
 
]]
  
local function add_prop_cat (key, arguments)
+
function format_script_value (script_value)
if not added_prop_cats [key] then
+
local lang=''; -- initialize to empty string
added_prop_cats [key] = true; -- note that we've added this category
+
local name;
key = key:gsub ('(foreign_lang_source_?2?)%a%a%a?', '%1'); -- strip lang code from keyname
+
if script_value:match('^%l%l%s*:') then -- if first 3 non-space characters are script language prefix
table.insert( z.properties_cats, substitute (cfg.prop_cats [key], arguments)); -- make name then add to table
+
lang = script_value:match('^(%l%l)%s*:%s*%S.*'); -- get the language prefix or nil if there is no script
end
+
if not is_set (lang) then
 +
return ''; -- script_value was just the prefix so return empty string
 +
end
 +
-- if we get this far we have prefix and script
 +
name = mw.language.fetchLanguageName( lang, "en" ); -- get language name so that we can use it to categorize
 +
if is_set (name) then -- is prefix a proper ISO 639-1 language code?
 +
script_value = script_value:gsub ('^%l%l%s*:%s*', ''); -- strip prefix from script
 +
-- is prefix one of these language codes?
 +
if inArray (lang, {'ar', 'bs', 'dv', 'el', 'fa', 'hy', 'ja', 'ko', 'ku', 'he', 'ps', 'ru', 'sd', 'sr', 'th', 'uk', 'ug', 'yi', 'zh'}) then
 +
table.insert( z.properties_cats, 'CS1 uses ' .. name .. '-language script ('..lang..')'); -- categorize in language-specific categories
 +
else
 +
table.insert( z.properties_cats, 'CS1 uses foreign language script'); -- use this category as a catchall until language-specific category is available
 +
end
 +
lang = ' lang="' .. lang .. '" '; -- convert prefix into a lang attribute
 +
else
 +
lang = ''; -- invalid so set lang to empty string
 +
end
 +
end
 +
script_value = substitute (cfg.presentation['bdi'], {lang, script_value}); -- isolate in case script is rtl
 +
 
 +
return script_value;
 
end
 
end
  
 +
--[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------
  
--[[--------------------------< A D D _ V A N C _ E R R O R >----------------------------------------------------
+
Initially for |title= and |script-title=, this function concatenates those two parameter values after the script value has been
 
+
wrapped in <bdi> tags.
Adds a single Vancouver system error message to the template's output regardless of how many error actually exist.
 
To prevent duplication, added_vanc_errs is nil until an error message is emitted.
 
 
 
added_vanc_errs is a boolean declared in page scope variables above
 
 
 
 
]]
 
]]
  
local function add_vanc_error (source)
+
function script_concatenate (title, script)
if not added_vanc_errs then
+
if is_set (script) then
added_vanc_errs = true; -- note that we've added this category
+
script = format_script_value (script); -- <bdi> tags, lang atribute, categorization, etc; returns empty string on error
table.insert( z.message_tail, { set_error( 'vancouver', {source}, true ) } );
+
if is_set (script) then
end
+
title = title .. ' ' .. script; -- concatenate title and script title
 +
end
 +
end
 +
return title;
 
end
 
end
  
  
--[[--------------------------< I S _ S C H E M E >------------------------------------------------------------
+
--[[--------------------------< W R A P _ S T Y L E >----------------------------------------------------------
  
does this thing that purports to be a uri scheme seem to be a valid scheme? The scheme is checked to see if it
+
Applies styling to various parameters. Supplied string is wrapped using a message_list configuration taking one
is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says:
+
argument; protects italic styled parameters. Additional text taken from citation_config.presentation - the reason
Scheme names consist of a sequence of characters beginning with a
+
this function is similar to but separate from wrap_msg().
  letter and followed by any combination of letters, digits, plus
 
  ("+"), period ("."), or hyphen ("-").
 
  
returns true if it does, else false
+
]]
  
]]
+
function wrap_style (key, str)
 +
if not is_set( str ) then
 +
return "";
 +
elseif inArray( key, { 'italic-title', 'trans-italic-title' } ) then
 +
str = safeforitalics( str );
 +
end
  
local function is_scheme (scheme)
+
return substitute( cfg.presentation[key], {str} );
return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern
 
 
end
 
end
  
  
--[=[-------------------------< I S _ D O M A I N _ N A M E >--------------------------------------------------
+
--[[--------------------------< W R A P _ M S G >--------------------------------------------------------------
  
Does this thing that purports to be a domain name seem to be a valid domain name?
+
Applies additional message text to various parameter values. Supplied string is wrapped using a message_list
 +
configuration taking one argument.  Supports lower case text for {{citation}} templates.  Additional text taken
 +
from citation_config.messages - the reason this function is similar to but separate from wrap_style().
  
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
+
]]
BNF defined here: https://tools.ietf.org/html/rfc4234
 
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
 
see also [[Single-letter second-level domain]]
 
list of tlds: https://www.iana.org/domains/root/db
 
  
rfc952 (modified by rfc 1123) requires the first and last character of a hostname to be a letter or a digit.  Between
+
function wrap_msg (key, str, lower)
the first and last characters the name may use letters, digits, and the hyphen.
+
if not is_set( str ) then
 +
return "";
 +
end
 +
if true == lower then
 +
local msg;
 +
msg = cfg.messages[key]:lower(); -- set the message to lower case before
 +
str = substitute( msg, {str} ); -- including template text
 +
return str;
 +
else
 +
return substitute( cfg.messages[key], {str} );
 +
end
 +
end
  
Also allowed are IPv4 addresses. IPv6 not supported
+
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
  
domain is expected to be stripped of any path so that the last character in the last character of the tld.  tld
+
Format the three chapter parameters: |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
is two or more alpha characters.  Any preceding '//' (from splitting a url with a scheme) will be stripped
+
parameter (chapter_url_source used for error messages).
here.  Perhaps not necessary but retained incase it is necessary for IPv4 dot decimal.
 
  
There are several tests:
+
]]
the first character of the whole domain name including subdomains must be a letter or a digit
 
internationalized domain name (ascii characters with .xn-- ASCII Compatible Encoding (ACE) prefix xn-- in the tld) see https://tools.ietf.org/html/rfc3490
 
single-letter/digit second-level domains in the .org, .cash, and .today TLDs
 
q, x, and z SL domains in the .com TLD
 
i and q SL domains in the .net TLD
 
single-letter SL domains in the ccTLDs (where the ccTLD is two letters)
 
two-character SL domains in gTLDs (where the gTLD is two or more letters)
 
three-plus-character SL domains in gTLDs (where the gTLD is two or more letters)
 
IPv4 dot-decimal address format; TLD not allowed
 
  
returns true if domain appears to be a proper name and tld or IPv4 address, else false
+
function format_chapter_title (chapter, transchapter, chapterurl, chapter_url_source)
 
+
local chapter_error = '';
]=]
 
 
 
local function is_domain_name (domain)
 
if not domain then
 
return false; -- if not set, abandon
 
end
 
 
 
domain = domain:gsub ('^//', ''); -- strip '//' from domain name if present; done here so we only have to do it once
+
if not is_set (chapter) then
+
chapter = ''; -- just to be safe for concatenation
if not domain:match ('^[%a%d]') then -- first character must be letter or digit
+
if is_set (transchapter) then
return false;
+
chapter = wrap_style ('trans-quoted-title', transchapter);
end
+
chapter_error = " " .. seterror ('trans_missing_chapter');
 
+
end
if domain:match ('^%a+:') then -- hack to detect things that look like s:Page:Title where Page: is namespace at wikisource
+
if is_set (chapterurl) then
return false;
+
chapter = externallink (chapterurl, chapter, chapter_url_source); -- adds bare_url_missing_title error if appropriate
end
+
end
-- Do most common case first
+
return chapter .. chapter_error;
if domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.%a%a+$') then -- three or more character hostname.hostname or hostname.tld
+
else -- here when chapter is set
return true;
+
chapter = kern_quotes (chapter); -- if necessary, separate chapter title's leading and trailing quote marks from Module provided quote marks
elseif domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.xn%-%-[%a%d]+$') then -- internationalized domain name with ACE prefix
+
chapter = wrap_style ('quoted-title', chapter);
return true;
+
if is_set (transchapter) then
elseif domain:match ('%f[%a%d][%a%d]%.cash$') then -- one character/digit .cash hostname
+
transchapter = wrap_style ('trans-quoted-title', transchapter);
return true;
+
chapter = chapter .. ' ' .. transchapter;
elseif domain:match ('%f[%a%d][%a%d]%.today') then -- one character/digit .today hostname
+
end
return true;
+
if is_set (chapterurl) then
elseif domain:match ('%f[%a%d][%a%d]%.org$') then -- one character/digit .org hostname
+
chapter = externallink (chapterurl, chapter); -- adds bare_url_missing_title error if appropriate
return true;
+
end
elseif domain:match ('%f[%a][qxz]%.com$') then -- assigned one character .com hostname (x.com times out 2015-12-10)
 
return true;
 
elseif domain:match ('%f[%a][iq]%.net$') then -- assigned one character .net hostname (q.net registered but not active 2015-12-10)
 
return true;
 
elseif domain:match ('%f[%a%d][%a%d]%.%a%a$') then -- one character hostname and cctld (2 chars)
 
return true;
 
elseif domain:match ('%f[%a%d][%a%d][%a%d]%.%a%a+$') then -- two character hostname and tld
 
return true;
 
elseif domain:match ('^%d%d?%d?%.%d%d?%d?%.%d%d?%d?%.%d%d?%d?') then -- IPv4 address
 
return true;
 
else
 
return false;
 
 
end
 
end
 +
return chapter;
 
end
 
end
  
 
+
--[[
--[[--------------------------< I S _ U R L >------------------------------------------------------------------
+
Argument wrapper.  This function provides support for argument
 
+
mapping defined in the configuration file so that multiple names
returns true if the scheme and domain parts of a url appear to be a valid url; else false.
+
can be transparently aliased to single internal variable.
 
+
]]
This function is the last step in the validation process.  This function is separate because there are cases that
+
function argument_wrapper( args )
are not covered by split_url(), for example is_parameter_ext_wikilink() which is looking for bracketted external
+
local origin = {};
wikilinks.
 
 
 
]]
 
 
 
local function is_url (scheme, domain)
 
if is_set (scheme) then -- if scheme is set check it and domain
 
return is_scheme (scheme) and is_domain_name (domain);
 
else
 
return is_domain_name (domain); -- scheme not set when url is protocol relative
 
end
 
end
 
 
 
 
 
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
 
 
 
Split a url into a scheme, authority indicator, and domain.
 
 
 
First remove Fully Qualified Domain Name terminator (a dot following tld) (if any) and any path(/), query(?) or fragment(#).
 
 
 
If protocol relative url, return nil scheme and domain else return nil for both scheme and domain.
 
 
 
When not protocol relative, get scheme, authority indicator, and domain.  If there is an authority indicator (one
 
or more '/' characters immediately following the scheme's colon), make sure that there are only 2.
 
 
 
Strip off any port and path;
 
 
 
]]
 
 
 
local function split_url (url_str)
 
local scheme, authority, domain;
 
 
 
url_str = url_str:gsub ('([%a%d])%.?[/%?#].*$', '%1'); -- strip FQDN terminator and path(/), query(?), fragment (#) (the capture prevents false replacement of '//')
+
return setmetatable({
 
+
ORIGIN = function( self, k )
if url_str:match ('^//%S*') then -- if there is what appears to be a protocol relative url
+
local dummy = self[k]; --force the variable to be loaded.
domain = url_str:match ('^//(%S*)')
+
return origin[k];
elseif url_str:match ('%S-:/*%S+') then -- if there is what appears to be a scheme, optional authority indicator, and domain name
 
scheme, authority, domain = url_str:match ('(%S-:)(/*)(%S+)'); -- extract the scheme, authority indicator, and domain portions
 
authority = authority:gsub ('//', '', 1); -- replace place 1 pair of '/' with nothing;
 
if is_set(authority) then -- if anything left (1 or 3+ '/' where authority should be) then
 
return scheme; -- return scheme only making domain nil which will cause an error message
 
 
end
 
end
domain = domain:gsub ('(%a):%d+', '%1'); -- strip port number if present
+
},
end
+
{
+
__index = function ( tbl, k )
return scheme, domain;
+
if origin[k] ~= nil then
end
+
return nil;
 
+
end
 
+
--[[--------------------------< L I N K _ P A R A M _ O K >---------------------------------------------------
+
local args, list, v = args, cfg.aliases[k];
 
+
checks the content of |title-link=, |series-link=, |author-link= etc for properly formatted content: no wikilinks, no urls
+
if type( list ) == 'table' then
 
+
v, origin[k] = selectone( args, list, 'redundant_parameters' );
Link parameters are to hold the title of a wikipedia article so none of the WP:TITLESPECIALCHARACTERS are allowed:
+
if origin[k] == nil then
# < > [ ] | { } _
+
origin[k] = ''; -- Empty string, not nil
except the underscore which is used as a space in wiki urls and # which is used for section links
+
end
 
+
elseif list ~= nil then
returns false when the value contains any of these characters.
+
v, origin[k] = args[list], list;
 
+
else
When there are no illegal characters, this function returns TRUE if value DOES NOT appear to be a valid url (the
+
-- maybe let through instead of raising an error?
|<param>-link= parameter is ok); else false when value appears to be a valid url (the |<param>-link= parameter is NOT ok).
+
-- v, origin[k] = args[k], k;
 
+
error( cfg.messages['unknown_argument_map'] );
]]
+
end
 
+
local function link_param_ok (value)
+
-- Empty strings, not nil;
local scheme, domain;
+
if v == nil then
if value:find ('[<>%[%]|{}]') then -- if any prohibited characters
+
v = cfg.defaults[k] or '';
return false;
+
origin[k] = '';
end
+
end
 
+
scheme, domain = split_url (value); -- get scheme or nil and domain or nil from url;  
+
tbl = rawset( tbl, k, v );
return not is_url (scheme, domain); -- return true if value DOES NOT appear to be a valid url
+
return v;
 +
end,
 +
});
 
end
 
end
  
 +
--[[
 +
Looks for a parameter's name in the whitelist.
  
--[[--------------------------< L I N K _ T I T L E _ O K >---------------------------------------------------
+
Parameters in the whitelist can have three values:
 
+
true - active, supported parameters
Use link_param_ok() to validate |<param>-link= value and its matching |<title>= value.
+
false - deprecated, supported parameters
 
+
nil - unsupported parameters
|<title>= may be wikilinked but not when |<param>-link= has a value.  This function emits an error message when
 
that condition exists
 
 
 
 
]]
 
]]
 
+
function validate( name )
local function link_title_ok (link, lorig, title, torig)
+
local name = tostring( name );
local orig;
+
local state = whitelist.basic_arguments[ name ];
 
+
if is_set (link) then -- don't bother if <param>-link doesn't have a value
+
-- Normal arguments
if not link_param_ok (link) then -- check |<param>-link= markup
+
if true == state then return true; end -- valid actively supported parameter
orig = lorig; -- identify the failing link parameter
+
if false == state then
elseif title:find ('%[%[') then -- check |title= for wikilink markup
+
deprecated_parameter (name); -- parameter is deprecated but still supported
orig = torig; -- identify the failing |title= parameter
+
return true;
end
 
 
end
 
end
 
+
if is_set (orig) then
+
-- Arguments with numbers in them
table.insert( z.message_tail, { set_error( 'bad_paramlink', orig)}); -- url or wikilink in |title= with |title-link=;
+
name = name:gsub( "%d+", "#" ); -- replace digit(s) with # (last25 becomes last#
 +
state = whitelist.numbered_arguments[ name ];
 +
if true == state then return true; end -- valid actively supported parameter
 +
if false == state then
 +
deprecated_parameter (name); -- parameter is deprecated but still supported
 +
return true;
 
end
 
end
 +
 +
return false; -- Not supported because not found or name is set to nil
 
end
 
end
  
 +
--[[--------------------------< E R R O R C O M M E N T >------------------------------------------------------
  
--[[--------------------------< C H E C K _ U R L >------------------------------------------------------------
+
Wraps error messages with css markup according to the state of hidden.
  
Determines whether a URL string appears to be valid.
+
]]
 
+
function errorcomment( content, hidden )
First we test for space characters.  If any are found, return false.  Then split the url into scheme and domain
+
return substitute( hidden and cfg.presentation['hidden-error'] or cfg.presentation['visible-error'], content );
portions, or for protocol relative (//example.com) urls, just the domain.  Use is_url() to validate the two
+
end
portions of the url. If both are valid, or for protocol relative if domain is valid, return true, else false.
 
 
 
Because it is different from a standard url, and because this module used external_link() to make external links
 
that work for standard and news: links, we validate newsgroup names here.  The specification for a newsgroup name
 
is at https://tools.ietf.org/html/rfc5536#section-3.1.4
 
  
 +
--[[
 +
Sets an error condition and returns the appropriate error message.  The actual placement
 +
of the error message in the output is the responsibility of the calling function.
 
]]
 
]]
 
+
function seterror( error_id, arguments, raw, prefix, suffix )
local function check_url( url_str )
+
local error_state = cfg.error_conditions[ error_id ];
if nil == url_str:match ("^%S+$") then -- if there are any spaces in |url=value it can't be a proper url
+
return false;
+
prefix = prefix or "";
 +
suffix = suffix or "";
 +
 +
if error_state == nil then
 +
error( cfg.messages['undefined_error'] );
 +
elseif is_set( error_state.category ) then
 +
table.insert( z.error_categories, error_state.category );
 
end
 
end
local scheme, domain;
 
 
scheme, domain = split_url (url_str); -- get scheme or nil and domain or nil from url;
 
 
 
if 'news:' == scheme then -- special case for newsgroups
+
local message = substitute( error_state.message, arguments );
return domain:match('^[%a%d%+%-_]+%.[%a%d%+%-_%.]*[%a%d%+%-_]$');
+
 +
message = message .. " ([[" .. cfg.messages['help page link'] ..
 +
"#" .. error_state.anchor .. "|" ..
 +
cfg.messages['help page label'] .. "]])";
 +
 +
z.error_ids[ error_id ] = true;
 +
if inArray( error_id, { 'bare_url_missing_title', 'trans_missing_title' } )
 +
and z.error_ids['citation_missing_title'] then
 +
return '', false;
 
end
 
end
 
 
return is_url (scheme, domain); -- return true if value appears to be a valid url
+
message = table.concat({ prefix, message, suffix });
 +
 +
if raw == true then
 +
return message, error_state.hidden;
 +
end
 +
 +
return errorcomment( message, error_state.hidden );
 
end
 
end
  
 
+
-- Formats a wiki style external link
--[=[-------------------------< I S _ P A R A M E T E R _ E X T _ W I K I L I N K >----------------------------
+
function externallinkid(options)
 
+
local url_string = options.id;
Return true if a parameter value has a string that begins and ends with square brackets [ and ] and the first
+
if options.encode == true or options.encode == nil then
non-space characters following the opening bracket appear to be a url.  The test will also find external wikilinks
+
url_string = mw.uri.encode( url_string );
that use protocol relative urls. Also finds bare urls.
 
 
 
The frontier pattern prevents a match on interwiki links which are similar to scheme:path urls.  The tests that
 
find bracketed urls are required because the parameters that call this test (currently |title=, |chapter=, |work=,
 
and |publisher=) may have wikilinks and there are articles or redirects like '//Hus' so, while uncommon, |title=[[//Hus]]
 
is possible as might be [[en://Hus]].
 
 
 
]=]
 
 
 
local function is_parameter_ext_wikilink (value)
 
local scheme, domain;
 
 
 
if value:match ('%f[%[]%[%a%S*:%S+.*%]') then -- if ext wikilink with scheme and domain: [xxxx://yyyyy.zzz]
 
scheme, domain = split_url (value:match ('%f[%[]%[(%a%S*:%S+).*%]'));
 
elseif value:match ('%f[%[]%[//%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
 
scheme, domain = split_url (value:match ('%f[%[]%[(//%S+).*%]'));
 
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
 
scheme, domain = split_url (value:match ('(%a%S*:%S+)'));
 
elseif value:match ('//%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
 
scheme, domain = split_url (value:match ('(//%S+)')); -- what is left should be the domain
 
else
 
return false; -- didn't find anything that is obviously a url
 
 
end
 
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
  
return is_url (scheme, domain); -- return true if value appears to be a valid url
+
-- Formats a wiki style internal link
 +
function internallinkid(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
 
end
  
 +
-- Format an external link with error checking
 +
function externallink( URL, label, source )
 +
local error_str = "";
 +
if not is_set( label ) then
 +
label = URL;
 +
if is_set( source ) then
 +
error_str = seterror( 'bare_url_missing_title', { wrap_style ('parameter', source) }, false, " " );
 +
else
 +
error( cfg.messages["bare_url_no_origin"] );
 +
end
 +
end
 +
if not checkurl( URL ) then
 +
error_str = seterror( 'bad_url', {}, false, " " ) .. error_str;
 +
end
 +
return table.concat({ "[", URL, " ", safeforurl( label ), "]", error_str });
 +
end
  
--[[-------------------------< C H E C K _ F O R _ U R L >-----------------------------------------------------
+
--[[--------------------------< N O W R A P _ D A T E >--------------------------------------------------------
  
loop through a list of parameters and their valuesLook at the value and if it has an external link, emit an error message.
+
When date is YYYY-MM-DD format wrap in nowrap span: <span ...>YYYY-MM-DD</span>When date is DD MMMM YYYY or is
 +
MMMM DD, YYYY then wrap in nowrap span: <span ...>DD MMMM</span> YYYY or <span ...>MMMM DD,</span> YYYY
 +
 
 +
DOES NOT yet support MMMM YYYY or any of the date ranges.
  
 
]]
 
]]
  
local function check_for_url (parameter_list)
+
function nowrap_date (date)
local error_message = '';
+
local cap='';
for k, v in pairs (parameter_list) do -- for each parameter in the list
+
local cap2='';
if is_parameter_ext_wikilink (v) then -- look at the value; if there is a url add an error message
+
 
if is_set(error_message) then -- once we've added the first portion of the error message ...
+
if date:match("^%d%d%d%d%-%d%d%-%d%d$") then
error_message=error_message .. ", "; -- ... add a comma space separator
+
date = substitute (cfg.presentation['nowrap1'], date);
end
+
error_message=error_message .. "&#124;" .. k .. "="; -- add the failed parameter
+
elseif date:match("%a+%s*%d%d?,%s*%d%d%d%d") or date:match ("%d%d?%s*%a+%s*%d%d%d%d") then
end
+
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
end
+
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
if is_set (error_message) then -- done looping, if there is an error message, display it
 
table.insert( z.message_tail, { set_error( 'param_has_ext_link', {error_message}, true ) } );
 
 
end
 
end
 +
 +
return date;
 
end
 
end
  
 +
--[[--------------------------< A M A Z O N >------------------------------------------------------------------
  
--[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------
+
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
Escape sequences for content that will be used for URL descriptions
+
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 safe_for_url( str )
+
function amazon(id, domain)
if str:match( "%[%[.-%]%]" ) ~= nil then  
+
local err_cat = ""
table.insert( z.message_tail, { set_error( 'wikilink_in_url', {}, true ) } );
+
 
 +
if not id:match("^[%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u]$") then
 +
err_cat =  ' ' .. seterror ('bad_asin'); -- asin is not a mix of 10 uppercase alpha and numeric characters
 +
else
 +
if id:match("^%d%d%d%d%d%d%d%d%d[%dX]$") then -- if 10-digit numeric (or 9 digits with terminal X)
 +
if checkisbn( id ) then -- see if asin value is isbn10
 +
table.insert( z.maintenance_cats, "CS1 maint: ASIN uses ISBN"); -- add to maint category
 +
elseif not is_set (err_cat) then
 +
err_cat =  ' ' .. seterror ('bad_asin'); -- asin is not isbn10
 +
end
 +
elseif not id:match("^%u[%d%u]+$") then
 +
err_cat =  ' ' .. seterror ('bad_asin'); -- asin doesn't begin with uppercase alpha
 +
end
 +
end
 +
if not is_set(domain) then
 +
domain = "com";
 +
elseif inArray (domain, {'jp', 'uk'}) then -- Japan, United Kingdom
 +
domain = "co." .. domain;
 +
elseif inArray (domain, {'au', 'br', 'mx'}) then -- Australia, Brazil, Mexico
 +
domain = "com." .. domain;
 
end
 
end
+
local handler = cfg.id_handlers['ASIN'];
return str:gsub( '[%[%]\n]', {
+
return externallinkid({link = handler.link,
['['] = '&#91;',
+
label=handler.label , prefix="//www.amazon."..domain.."/dp/",id=id,
[']'] = '&#93;',
+
encode=handler.encode, separator = handler.separator}) .. err_cat;
['\n'] = ' ' } );
 
 
end
 
end
  
 +
--[[--------------------------< A R X I V >--------------------------------------------------------------------
  
--[[--------------------------< E X T E R N A L _ L I N K >----------------------------------------------------
+
See: http://arxiv.org/help/arxiv_identifier
  
Format an external link with error checking
+
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 external_link( URL, label, source, access)
+
function arxiv (id)
local error_str = "";
+
local handler = cfg.id_handlers['ARXIV'];
local domain;
+
local year, month, version;
local path;
+
local err_cat = ""
local base_url;
 
 
 
if not is_set( label ) then
 
label = URL;
 
if is_set( source ) then
 
error_str = set_error( 'bare_url_missing_title', { wrap_style ('parameter', source) }, false, " " );
 
else
 
error( cfg.messages["bare_url_no_origin"] );
 
end
 
end
 
if not check_url( URL ) then
 
error_str = set_error( 'bad_url', {wrap_style ('parameter', source)}, false, " " ) .. error_str;
 
end
 
 
 
domain, path = URL:match ('^([/%.%-%+:%a%d]+)([/%?#].*)$'); -- split the url into scheme plus domain and path
+
if id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%d$") or id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%dv%d+$") then -- test for the 9108-0703 format w/ & w/o version
if path then -- if there is a path portion
+
year, month = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d[v%d]*$");
path = path:gsub ('[%[%]]', {['[']='%5b',[']']='%5d'}); -- replace '[' and ']' with their percent encoded values
+
year = tonumber(year);
URL = table.concat ({domain, path}); -- and reassemble
+
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 = ' ' .. seterror( '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]*$");
 +
year = tonumber(year);
 +
month = tonumber(month);
 +
if ((7 > year) or (14 < year) or (1 > month or 12 < month)) or -- is year invalid or is month invalid? (doesn't test for future years)
 +
((7 == year) and (4 > month)) then --or -- when year is 07, is month invalid (before April)?
 +
err_cat = ' ' .. seterror( 'bad_arxiv' ); -- set error message
 +
end
 +
elseif id:match("^%d%d[01]%d%.%d%d%d%d%d$") 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 = ' ' .. seterror( 'bad_arxiv' ); -- set error message
 +
end
 +
else
 +
err_cat = ' ' .. seterror( 'bad_arxiv' ); -- arXiv id doesn't match any format
 
end
 
end
  
base_url = table.concat({ "[", URL, " ", safe_for_url (label), "]" }); -- assemble a wikimarkup url
+
return externallinkid({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 +
end
  
if is_set (access) then -- access level (subscription, registration, limited)
 
base_url = substitute (cfg.presentation['ext-link-access-signal'], {cfg.presentation[access].class, cfg.presentation[access].title, base_url}); -- add the appropriate icon
 
end
 
 
return table.concat ({base_url, error_str});
 
end
 
  
 +
--[[
 +
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.
 +
]]
  
--[[--------------------------< D E P R E C A T E D _ P A R A M E T E R >--------------------------------------
+
function normalize_lccn (lccn)
 +
lccn = lccn:gsub ("%s", ""); -- 1. strip whitespace
  
Categorize and emit an error message when the citation contains one or more deprecated parameters. The function includes the
+
if nil ~= string.find (lccn,'/') then
offending parameter name to the error message.  Only one error message is emitted regardless of the number of deprecated
+
lccn = lccn:match ("(.-)/"); -- 2. remove forward slash and all character to the right of it
parameters in the citation.
+
end
  
added_deprecated_cat is a boolean declared in page scope variables above
+
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
local function deprecated_parameter(name)
+
lccn=prefix..suffix; -- reassemble the lccn
if not added_deprecated_cat then
+
end
added_deprecated_cat = true; -- note that we've added this category
+
table.insert( z.message_tail, { set_error( 'deprecated_params', {name}, true ) } ); -- add error message
+
return lccn;
 
end
 
end
end
 
  
 +
--[[
 +
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/
  
--[=[-------------------------< K E R N _ Q U O T E S >--------------------------------------------------------
+
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
  
Apply kerning to open the space between the quote mark provided by the Module and a leading or trailing quote
+
]]
mark contained in a |title= or |chapter= parameter's value.
+
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
  
This function will positive kern either single or double quotes:
+
id = normalize_lccn (id); -- get canonical form (no whitespace, hyphens, forward slashes)
"'Unkerned title with leading and trailing single quote marks'"
+
local len = id:len(); -- get the length of the lccn
" 'Kerned title with leading and trailing single quote marks' " (in real life the kerning isn't as wide as this example)
 
Double single quotes (italic or bold wikimarkup) are not kerned.
 
  
Replaces unicode quotemarks in plain text or in the label portion of a [[L|D]] style wikilink with typewriter
+
if 8 == len then
quote marks regardless of the need for kerning. Unicode quote marks are not replaced in simple [[D]] wikilinks.
+
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits)
 +
err_cat = ' ' .. seterror( '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 = ' ' .. seterror( '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 = ' ' .. seterror( '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 = ' ' .. seterror( '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 = ' ' .. seterror( 'bad_lccn' ); -- no match, set an error message
 +
end
 +
else
 +
err_cat = ' ' .. seterror( 'bad_lccn' ); -- wrong length, set an error message
 +
end
  
Call this function for chapter titles, for website titles, etc; not for book titles.
+
if not is_set (err_cat) and nil ~= lccn:find ('%s') then
 +
err_cat = ' ' .. seterror( 'bad_lccn' ); -- lccn contains a space, set an error message
 +
end
  
]=]
+
return externallinkid({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
 +
end
  
local function kern_quotes (str)
+
--[[
local cap='';
+
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
local cap2='';
+
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 wl_type, label, link;
+
]]
 +
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 = ' ' .. seterror( '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 = ' ' .. seterror( 'bad_pmid' ); -- set an error message
 +
end
 +
end
 +
 +
return externallinkid({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 +
end
  
wl_type, label, link = is_wikilink (str); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|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 true; otherwise, returns false because the embargo has expired or |embargo= not set in this cite.
 +
]]
 +
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 1 == wl_type then -- [[D]] simple wikilink with or without quote marks
+
if good1 and good2 and tonumber( embargo_date ) >= tonumber( todays_date ) then --is embargo date is in the future?
if mw.ustring.match (str, '%[%[[\"“”\'‘’].+[\"“”\'‘’]%]%]') then -- leading and trailing quote marks
+
return true; -- still embargoed
str = substitute (cfg.presentation['kern-wl-both'], str);
 
elseif mw.ustring.match (str, '%[%[[\"“”\'‘’].+%]%]') then -- leading quote marks
 
str = substitute (cfg.presentation['kern-wl-left'], str);
 
elseif mw.ustring.match (str, '%[%[.+[\"“”\'‘’]%]%]') then -- trailing quote marks
 
str = substitute (cfg.presentation['kern-wl-right'], str);
 
 
end
 
end
 +
end
 +
return false; -- embargo expired or |embargo= not set
 +
end
 +
 +
--[[
 +
Format a PMC, do simple error checking, and check for embargoed articles.
  
else -- plain text or [[L|D]]; text in label variable
+
The embargo parameter takes a date for a value. If the embargo date is in the future
label= mw.ustring.gsub (label, '[“”]', '\"'); -- replace “” (U+201C & U+201D) with " (typewriter double quote mark)
+
the PMC identifier will not be linked to the article.  If the embargo specifies a date in the past, or if it is empty or omitted, then
label= mw.ustring.gsub (label, '[‘’]', '\''); -- replace ‘’ (U+2018 & U+2019) with ' (typewriter single quote mark)
+
the PMC identifier is linked to the article through the link at cfg.id_handlers['PMC'].prefix.
 +
 
 +
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.
 +
]]
 +
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;
  
cap, cap2 = mw.ustring.match (label, "^([\"\'])([^\'].+)"); -- match leading double or single quote but not doubled single quotes (italic markup)
+
if id:match("[^%d]") then -- if PMC has anything but digits
if is_set (cap) then
+
err_cat = ' ' .. seterror( 'bad_pmc' ); -- set an error message
label = substitute (cfg.presentation['kern-left'], {cap, cap2});
+
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 = ' ' .. seterror( 'bad_pmc' ); -- set an error message
 
end
 
end
 +
end
 
 
cap, cap2 = mw.ustring.match (label, "^(.+[^\'])([\"\'])$") -- match trailing double or single quote but not doubled single quotes (italic markup)
+
if is_embargoed(embargo) then
if is_set (cap) then
+
text="[[" .. handler.link .. "|" .. handler.label .. "]]:" .. handler.separator .. id .. err_cat; --still embargoed so no external link
label = substitute (cfg.presentation['kern-right'], {cap, cap2});
+
else
end
+
text = externallinkid({link = handler.link, label = handler.label, --no embargo date, ok to link to article
+
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
if 2 == wl_type then
 
str = make_wikilink (link, label); -- reassemble the wikilink
 
else
 
str = label;
 
end
 
 
end
 
end
return str;
+
return text;
 
end
 
end
  
 +
-- Formats a DOI and checks for DOI errors.
  
--[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >----------------------------------------
+
-- 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
  
|script-title= holds title parameters that are not written in Latin based scripts: Chinese, Japanese, Arabic, Hebrew, etc. These scripts should
+
-- This function checks a DOI name for: prefix/suffix.  If the doi name contains spaces or endashes,
not be italicized and may be written right-to-left.  The value supplied by |script-title= is concatenated onto Title after Title has been wrapped
+
-- or, if it ends with a period or a comma, this function will emit a bad_doi error message.
in italic markup.
 
  
Regardless of language, all values provided by |script-title= are wrapped in <bdi>...</bdi> tags to isolate rtl languages from the English left to right.
+
-- 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.
  
|script-title= provides a unique feature. The value in |script-title= may be prefixed with a two-character ISO639-1 language code and a colon:
+
function doi(id, inactive)
|script-title=ja:*** *** (where * represents a Japanese character)
+
local cat = ""
Spaces between the two-character code and the colon and the colon and the first script character are allowed:
+
local handler = cfg.id_handlers['DOI'];
|script-title=ja : *** ***
+
|script-title=ja: *** ***
+
local text;
|script-title=ja :*** ***
+
if is_set(inactive) then
Spaces preceding the prefix are allowed: |script-title = ja:*** ***
+
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 = externallinkid({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 = ' ' .. seterror( 'bad_doi' );
 +
end
 +
return text .. inactive .. cat
 +
end
 +
 
 +
-- Formats an OpenLibrary link, and checks for associated errors.
 +
function openlibrary(id)
 +
local code = id:match("^%d+([AMW])$"); -- only digits followed by 'A', 'M', or 'W'
 +
local handler = cfg.id_handlers['OL'];
  
The prefix is checked for validity. If it is a valid ISO639-1 language code, the lang attribute (lang="ja") is added to the <bdi> tag so that browsers can
+
if ( code == "A" ) then
know the language the tag contains. This may help the browser render the script more correctly. If the prefix is invalid, the lang attribute
+
return externallinkid({link=handler.link, label=handler.label,
is not added. At this time there is no error message for this condition.
+
prefix="http://openlibrary.org/authors/OL",id=id, separator=handler.separator,
 +
encode = handler.encode})
 +
elseif ( code == "M" ) then
 +
return externallinkid({link=handler.link, label=handler.label,
 +
prefix="http://openlibrary.org/books/OL",id=id, separator=handler.separator,
 +
encode = handler.encode})
 +
elseif ( code == "W" ) then
 +
return externallinkid({link=handler.link, label=handler.label,
 +
prefix= "http://openlibrary.org/works/OL",id=id, separator=handler.separator,
 +
encode = handler.encode})
 +
else
 +
return externallinkid({link=handler.link, label=handler.label,
 +
prefix= "http://openlibrary.org/OL",id=id, separator=handler.separator,
 +
encode = handler.encode}) ..
 +
' ' .. seterror( 'bad_ol' );
 +
end
 +
end
  
Supports |script-title= and |script-chapter=
+
--[[
 +
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:
  
TODO: error messages when prefix is invalid ISO639-1 code; when script_value has prefix but no script;
+
|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.
 
]]
 
]]
 +
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;
  
local function format_script_value (script_value)
+
id=id:gsub( "[%s-]", "" ); -- strip spaces, hyphens, and endashes from the issn
local lang=''; -- initialize to empty string
+
 
local name;
+
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
if script_value:match('^%l%l%s*:') then -- if first 3 non-space characters are script language prefix
+
valid_issn=false; -- wrong length or improper character
lang = script_value:match('^(%l%l)%s*:%s*%S.*'); -- get the language prefix or nil if there is no script
+
else
if not is_set (lang) then
+
valid_issn=is_valid_isxn(id, 8); -- validate issn
return ''; -- script_value was just the prefix so return empty string
 
end
 
-- if we get this far we have prefix and script
 
name = cfg.lang_code_remap[lang] or mw.language.fetchLanguageName( lang, "en" ); -- get language name so that we can use it to categorize
 
if is_set (name) then -- is prefix a proper ISO 639-1 language code?
 
script_value = script_value:gsub ('^%l%l%s*:%s*', ''); -- strip prefix from script
 
-- is prefix one of these language codes?
 
if in_array (lang, cfg.script_lang_codes) then
 
add_prop_cat ('script_with_name', {name, lang})
 
else
 
add_prop_cat ('script')
 
end
 
lang = ' lang="' .. lang .. '" '; -- convert prefix into a lang attribute
 
else
 
lang = ''; -- invalid so set lang to empty string
 
end
 
 
end
 
end
script_value = substitute (cfg.presentation['bdi'], {lang, script_value}); -- isolate in case script is rtl
 
  
return script_value;
+
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 = externallinkid({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 +
 +
if false == valid_issn then
 +
text = text .. ' ' .. seterror( 'bad_issn' ) -- add an error message if the issn is invalid
 +
end
 +
 +
return text
 
end
 
end
  
 +
--[[--------------------------< M E S S A G E _ I D >----------------------------------------------------------
  
--[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------
+
Validate and format a usenet message id.  Simple error checking, looks for 'id-[email protected]-right' not enclosed in
 +
'<' and/or '>' angle brackets.
  
Initially for |title= and |script-title=, this function concatenates those two parameter values after the script value has been
 
wrapped in <bdi> tags.
 
 
]]
 
]]
  
local function script_concatenate (title, script)
+
function message_id (id)
if is_set (script) then
+
local handler = cfg.id_handlers['USENETID'];
script = format_script_value (script); -- <bdi> tags, lang atribute, categorization, etc; returns empty string on error
+
 
if is_set (script) then
+
text = externallinkid({link = handler.link, label = handler.label,
title = title .. ' ' .. script; -- concatenate title and script title
+
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
end
+
end
+
if not id:match('^[email protected]+$') or not id:match('^[^<].*[^>]$')then -- doesn't have '@' or has one or first or last character is '< or '>'
return title;
+
text = text .. ' ' .. seterror( 'bad_message_id' ) -- add an error message if the message id is invalid
 +
end  
 +
 +
return text
 
end
 
end
  
  
--[[--------------------------< W R A P _ M S G >--------------------------------------------------------------
 
  
Applies additional message text to various parameter values. Supplied string is wrapped using a message_list
+
--[[--------------------------< S E T _ T I T L E T Y P E >----------------------------------------------------
configuration taking one argument.  Supports lower case text for {{citation}} templates. Additional text taken
+
 
from citation_config.messages - the reason this function is similar to but separate from wrap_style().
+
This function sets default title types (equivalent to the citation including |type=<default value>) for those citations that have defaults.
 +
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
  
 
]]
 
]]
 
+
function set_titletype(cite_class, title_type)
local function wrap_msg (key, str, lower)
+
if is_set(title_type) then
if not is_set( str ) then
+
if "none" == title_type then
return "";
+
title_type = ""; -- if |type=none then type parameter not displayed
 +
end
 +
return title_type; -- if |type= has been set to any other value use that value
 
end
 
end
if true == lower then
 
local msg;
 
msg = cfg.messages[key]:lower(); -- set the message to lower case before
 
return substitute( msg, str ); -- including template text
 
else
 
return substitute( cfg.messages[key], str );
 
end
 
end
 
  
 +
if "AV-media-notes" == cite_class or "DVD-notes" == cite_class then -- if this citation is cite AV media notes or cite DVD notes
 +
return "Media notes"; -- display AV media notes / DVD media notes annotation
  
--[[--------------------------< W I K I S O U R C E _ U R L _ M A K E >----------------------------------------
+
elseif "mailinglist" == cite_class then -- if this citation is cite mailing list
 +
return "Mailing list"; -- display mailing list annotation
  
makes a wikisource url from wikisource interwiki link.  returns the url and appropriate label; nil else.
+
elseif "podcast" == cite_class then -- if this citation is cite podcast
 +
return "Podcast"; -- display podcast annotation
  
str is the value assigned to |chapter= (or aliases) or |title= or |title-link=
+
elseif "pressrelease" == cite_class then -- if this citation is cite press release
 +
return "Press release"; -- display press release annotation
  
]]
+
elseif "report" == cite_class then -- if this citation is cite report
 
+
return "Report"; -- display report annotation
local function wikisource_url_make (str)
+
local wl_type, D, L;
+
elseif "techreport" == cite_class then -- if this citation is cite techreport
local ws_url, ws_label;
+
return "Technical report"; -- display techreport annotation
 +
 +
elseif "thesis" == cite_class then -- if this citation is cite thesis (degree option handled after this function returns)
 +
return "Thesis"; -- display simple thesis annotation (without |degree= modification)
 +
end
 +
end
  
wl_type, D, L = is_wikilink (str); -- wl_type is 0 (not a wikilink), 1 (simple wikilink), 2 (complex wikilink)
+
--[[
 +
Determines whether a URL string is valid
  
if 0 == wl_type then -- not a wikilink; might be from |title-link=
+
At present the only check is whether the string appears to
str = D:match ('^[Ww]ikisource:(.+)') or D:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace
+
be prefixed with a URI scheme. It is not determined whether
if is_set (str) then
+
the URI scheme is valid or whether the URL is otherwise well
ws_url = table.concat ({ -- build a wikisource url
+
formed.
'https://en.wikisource.org/wiki/', -- prefix
+
]]
str, -- article title
+
function checkurl( url_str )
});
+
-- Protocol-relative or URL scheme
ws_label = str; -- label for the url
+
return url_str:sub(1,2) == "//" or url_str:match( "^[^/]*:" ) ~= nil;
end
+
end
elseif 1 == wl_type then -- simple wikilink: [[Wikisource:ws article]]
 
str = D:match ('^[Ww]ikisource:(.+)') or D:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace
 
if is_set (str) then
 
ws_url = table.concat ({ -- build a wikisource url
 
'https://en.wikisource.org/wiki/', -- prefix
 
str, -- article title
 
});
 
ws_label = str; -- label for the url
 
end
 
elseif 2 == wl_type then -- non-so-simple wikilink: [[Wikisource:ws article|displayed text]] ([[L|D]])
 
str = L:match ('^[Ww]ikisource:(.+)') or L:match ('^[Ss]:(.+)'); -- article title from interwiki link with long-form or short-form namespace
 
if is_set (str) then
 
ws_label = D; -- get ws article name from display portion of interwiki link
 
ws_url = table.concat ({ -- build a wikisource url
 
'https://en.wikisource.org/wiki/', -- prefix
 
str, -- article title without namespace from link portion of wikilink
 
});
 
end
 
end
 
 
if ws_url then
 
ws_url = mw.uri.encode (ws_url, 'WIKI'); -- make a usable url
 
ws_url = ws_url:gsub ('%%23', '#'); -- undo percent encoding of anchor
 
end
 
  
return ws_url, ws_label, L or D; -- return proper url or nil and a label or nil
+
-- Removes irrelevant text and dashes from ISBN number
 +
-- Similar to that used for Special:BookSources
 +
function cleanisbn( isbn_str )
 +
return isbn_str:gsub( "[^-0-9X]", "" );
 
end
 
end
  
 +
--[[--------------------------< E S C A P E _ L U A _ M A G I C _ C H A R S >----------------------------------
  
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
+
Returns a string where all of lua's magic characters have been escaped.  This is important because functions like
 +
string.gsub() treat their pattern and replace strings as patterns, not literal strings.
 +
]]
 +
function escape_lua_magic_chars (argument)
 +
argument = argument:gsub("%%", "%%%%"); -- replace % with %%
 +
argument = argument:gsub("([%^%$%(%)%.%[%]%*%+%-%?])", "%%%1"); -- replace all other lua magic pattern characters
 +
return argument;
 +
end
 +
 
 +
--[[--------------------------< S T R I P _ A P O S T R O P H E _ M A R K U P >--------------------------------
  
Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
+
Strip wiki italic and bold markup from argument so that it doesn't contaminate COinS metadata.
parameter (chapter_url_source used for error messages).
+
This function strips common patterns of apostrophe markup.  We presume that editors who have taken the time to
 +
markup a title have, as a result, provided valid markup. When they don't, some single apostrophes are left behind.
  
 
]]
 
]]
 +
function strip_apostrophe_markup (argument)
 +
if not is_set (argument) then return argument; end
  
local function format_chapter_title (scriptchapter, chapter, transchapter, chapterurl, chapter_url_source, no_quotes, access)
+
while true do
local chapter_error = '';
+
if argument:match ("%'%'%'%'%'") then -- bold italic (5)
 
+
argument=argument:gsub("%'%'%'%'%'", ""); -- remove all instances of it
local ws_url, ws_label, L = wikisource_url_make (chapter); -- make a wikisource url and label from a wikisource interwiki link
+
elseif argument:match ("%'%'%'%'") then -- italic start and end without content (4)
if ws_url then
+
argument=argument:gsub("%'%'%'%'", "");
ws_label = ws_label:gsub ('_', ''); -- replace underscore separaters with space characters
+
elseif argument:match ("%'%'%'") then -- bold (3)
chapter = ws_label;
+
argument=argument:gsub("%'%'%'", "");
end
+
elseif argument:match ("%'%'") then -- italic (2)
 
+
argument=argument:gsub("%'%'", "");
if not is_set (chapter) then
+
else
chapter = ''; -- to be safe for concatenation
+
break;
else
 
if false == no_quotes then
 
chapter = kern_quotes (chapter); -- if necessary, separate chapter title's leading and trailing quote marks from Module provided quote marks
 
chapter = wrap_style ('quoted-title', chapter);
 
 
end
 
end
 
end
 
end
 +
return argument; -- done
 +
end
  
chapter = script_concatenate (chapter, scriptchapter) -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
+
--[[--------------------------< M A K E _ C O I N S _ T I T L E >----------------------------------------------
  
if is_set (chapterurl) then
+
Makes a title for COinS from Title and / or ScriptTitle (or any other name-script pairs)
chapter = external_link (chapterurl, chapter, chapter_url_source, access); -- adds bare_url_missing_title error if appropriate
 
elseif ws_url then
 
chapter = external_link (ws_url, chapter .. '&nbsp;', 'ws link in chapter'); -- adds bare_url_missing_title error if appropriate; space char to move icon away from chap text; TODO: better way to do this?
 
chapter = substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, chapter});
 
end
 
  
if is_set (transchapter) then
+
Apostrophe markup (bold, italics) is stripped from each value so that the COinS metadata isn't correupted with strings
transchapter = wrap_style ('trans-quoted-title', transchapter);
+
of %27%27...
if is_set (chapter) then
+
]]
chapter = chapter .. ' ' .. transchapter;
 
else -- here when transchapter without chapter or script-chapter
 
chapter = transchapter; --
 
chapter_error = ' ' .. set_error ('trans_missing_title', {'chapter'});
 
end
 
end
 
  
-- if is_set (chapterurl) then
+
function make_coins_title (title, script)
-- chapter = external_link (chapterurl, chapter, chapter_url_source, access); -- adds bare_url_missing_title error if appropriate
+
if is_set (title) then
-- end
+
title = strip_apostrophe_markup (title); -- strip any apostrophe markup
 
+
else
return chapter .. chapter_error;
+
title=''; -- if not set, make sure title is an empty string
 +
end
 +
if is_set (script) then
 +
script = script:gsub ('^%l%l%s*:%s*', ''); -- remove language prefix if present (script value may now be empty string)
 +
script = strip_apostrophe_markup (script); -- strip any apostrophe markup
 +
else
 +
script=''; -- if not set, make sure script is an empty string
 +
end
 +
if is_set (title) and is_set (script) then
 +
script = ' ' .. script; -- add a space before we concatenate
 +
end
 +
return title .. script; -- return the concatenation
 
end
 
end
  
 +
--[[--------------------------< G E T _ C O I N S _ P A G E S >------------------------------------------------
  
--[[--------------------------< H A S _ I N V I S I B L E _ C H A R S >----------------------------------------
+
Extract page numbers from external wikilinks in any of the |page=, |pages=, or |at= parameters for use in COinS.
 
 
This function searches a parameter's value for nonprintable or invisible characters.  The search stops at the
 
first match.
 
 
 
This function will detect the visible replacement character when it is part of the wikisource.
 
 
 
Detects but ignores nowiki and math stripmarkers.  Also detects other named stripmarkers (gallery, math, pre, ref)
 
and identifies them with a slightly different error message.  See also coins_cleanup().
 
 
 
Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker
 
that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the
 
parameter value.
 
  
 
]]
 
]]
 
+
function get_coins_pages (pages)
local function has_invisible_chars (param, v)
+
local pattern;
local position = ''; -- position of invisible char or starting position of stripmarker
+
if not is_set (pages) then return pages; end -- if no page numbers then we're done
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
 
local capture; -- used by stripmarker detection to hold name of the stripmarker
 
local i=1;
 
local stripmarker, apostrophe;
 
 
 
capture = string.match (v, '[%w%p ]*'); -- Test for values that are simple ASCII text and bypass other tests if true
+
while true do
if capture == v then -- if same there are no unicode characters
+
pattern = pages:match("%[(%w*:?//[^ ]+%s+)[%w%d].*%]"); -- pattern is the opening bracket, the url and following space(s): "[url "
return;
+
if nil == pattern then break; end -- no more urls
 +
pattern = escape_lua_magic_chars (pattern); -- pattern is not a literal string; escape lua's magic pattern characters
 +
pages = pages:gsub(pattern, ""); -- remove as many instances of pattern as possible
 
end
 
end
 +
pages = pages:gsub("[%[%]]", ""); -- remove the brackets
 +
pages = pages:gsub("–", "-" ); -- replace endashes with hyphens
 +
pages = pages:gsub("&%w+;", "-" ); -- and replace html entities (&ndash; etc.) with hyphens; do we need to replace numerical entities like &#32; and the like?
 +
return pages;
 +
end
  
while cfg.invisible_chars[i] do
+
--[[
local char=cfg.invisible_chars[i][1] -- the character or group name
+
ISBN-10 and ISSN validator code calculates checksum across all isbn/issn digits including the check digit. ISBN-13 is checked in checkisbn().
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
+
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,
position, dummy, capture = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
+
spaces and other non-isxn characters.
+
]]
if position and (char == 'zero width joiner') then -- if we found a zero width joiner character
+
function is_valid_isxn (isxn_str, len)
if mw.ustring.find (v, cfg.indic_script) then -- its ok if one of the indic scripts
+
local temp = 0;
position = nil; -- unset position
+
isxn_str = { isxn_str:byte(1, len) }; -- make a table of bytes
end
+
len = len+1; -- adjust to be a loop counter
end
+
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
if position then
+
temp = temp + 10*( len - i ); -- it represents 10 decimal
if 'nowiki' == capture or 'math' == capture or -- nowiki and math stripmarkers (not an error condition)
+
else
('templatestyles' == capture and in_array (param, {'id', 'quote'})) then -- templatestyles stripmarker allowed in these parameters
+
temp = temp + tonumber( string.char(v) )*(len-i);
stripmarker = true; -- set a flag
 
elseif true == stripmarker and 'delete' == char then -- because stripmakers begin and end with the delete char, assume that we've found one end of a stripmarker
 
position = nil; -- unset
 
else
 
local err_msg;
 
if capture then
 
err_msg = capture .. ' ' .. char;
 
else
 
err_msg = char .. ' ' .. 'character';
 
end
 
 
 
table.insert( z.message_tail, { set_error( 'invisible_char', {err_msg, wrap_style ('parameter', param), position}, true ) } ); -- add error message
 
return; -- and done with this parameter
 
end
 
 
end
 
end
i=i+1; -- bump our index
 
 
end
 
end
 +
return temp % 11 == 0; -- returns true if calculation result is zero
 
end
 
end
  
 +
-- Determines whether an ISBN string is valid
 +
function checkisbn( 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
  
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
+
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
 +
isbn_str = { isbn_str:byte(1, len) };
 +
for i, v in ipairs( isbn_str ) do
 +
temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) );
 +
end
 +
return temp % 10 == 0;
 +
end
 +
end
  
Argument wrapper. This function provides support for argument mapping defined in the configuration file so that
+
-- Gets the display text for a wikilink like [[A|B]] or [[B]] gives B
multiple names can be transparently aliased to single internal variable.
+
function removewikilink( str )
 +
return (str:gsub( "%[%[([^%[%]]*)%]%]", function(l)
 +
return l:gsub( "^[^|]*|(.*)$", "%1" ):gsub("^%s*(.-)%s*$", "%1");
 +
end));
 +
end
 +
 
 +
-- Escape sequences for content that will be used for URL descriptions
 +
function safeforurl( str )
 +
if str:match( "%[%[.-%]%]" ) ~= nil then
 +
table.insert( z.message_tail, { seterror( 'wikilink_in_url', {}, true ) } );
 +
end
 +
 +
return str:gsub( '[%[%]\n]', {
 +
['['] = '&#91;',
 +
[']'] = '&#93;',
 +
['\n'] = ' ' } );
 +
end
  
]]
+
-- Converts a hyphen to a dash
 +
function hyphentodash( str )
 +
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
 +
return str;
 +
end
 +
return str:gsub( '-', '–' );
 +
end
  
local function argument_wrapper( args )
+
-- Protects a string that will be wrapped in wiki italic markup '' ... ''
local origin = {};
+
function safeforitalics( str )
 +
--[[ Note: We cannot use <i> for italics, as the expected behavior for
 +
italics specified by ''...'' in the title is that they will be inverted
 +
(i.e. unitalicized) in the resulting references.  In addition, <i> and ''
 +
tend to interact poorly under Mediawiki's HTML tidy. ]]
 
 
return setmetatable({
+
if not is_set(str) then
ORIGIN = function( self, k )
+
return str;
local dummy = self[k]; --force the variable to be loaded.
+
else
return origin[k];
+
if str:sub(1,1) == "'" then str = "<span></span>" .. str; end
end
+
if str:sub(-1,-1) == "'" then str = str .. "<span></span>"; end
},
+
{
+
-- Remove newlines as they break italics.
__index = function ( tbl, k )
+
return str:gsub( '\n', ' ' );
if origin[k] ~= nil then
+
end
return nil;
+
end
 +
 
 +
--[[--------------------------< S A F E J O I N >--------------------------------------------------------------
 +
 
 +
Joins a sequence of strings together while checking for duplicate separation characters.
 +
 
 +
]]
 +
function safejoin( tbl, duplicate_char )
 +
--[[
 +
Note: we use string functions here, rather than ustring functions.
 +
 +
This has considerably faster performance and should work correctly as
 +
long as the duplicate_char is strict ASCII.  The strings
 +
in tbl may be ASCII or UTF8.
 +
]]
 +
 +
local str = ''; -- the output string
 +
local comp = ''; -- what does 'comp' mean?
 +
local end_chr = '';
 +
local trim;
 +
for _, value in ipairs( tbl ) do
 +
if value == nil then value = ''; end
 +
 +
if str == '' then -- if output string is empty
 +
str = value; -- assign value to it (first time through the loop)
 +
elseif value ~= '' then
 +
if value:sub(1,1) == '<' then -- Special case of values enclosed in spans and other markup.
 +
comp = value:gsub( "%b<>", "" ); -- remove html markup (<span>string</span> -> string)
 +
else
 +
comp = value;
 
end
 
end
+
-- typically duplicate_char is sepc
local args, list, v = args, cfg.aliases[k];
+
if comp:sub(1,1) == duplicate_char then -- is first charactier same as duplicate_char? why test first character?
+
--  Because individual string segments often (always?) begin with terminal punct for th
if type( list ) == 'table' then
+
--  preceding segment: 'First element' .. 'sepc next element' .. etc?
v, origin[k] = select_one( args, list, 'redundant_parameters' );
+
trim = false;
if origin[k] == nil then
+
end_chr = str:sub(-1,-1); -- get the last character of the output string
origin[k] = ''; -- Empty string, not nil
+
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
 +
if end_chr == duplicate_char then -- if same as separator
 +
str = str:sub(1,-2); -- remove it
 +
elseif end_chr == "'" then -- if it might be wikimarkup
 +
if str:sub(-3,-1) == duplicate_char .. "''" then -- if last three chars of str are sepc''
 +
str = str:sub(1, -4) .. "''"; -- remove them and add back ''
 +
elseif str:sub(-5,-1) == duplicate_char .. "]]''" then -- if last five chars of str are sepc]]''
 +
trim = true; -- why? why do this and next differently from previous?
 +
elseif str:sub(-4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''  
 +
trim = true; -- same question
 +
end
 +
elseif end_chr == "]" then -- if it might be wikimarkup
 +
if str:sub(-3,-1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink
 +
trim = true;
 +
elseif str:sub(-2,-1) == duplicate_char .. "]" then -- if last two chars of str are sepc] external link
 +
trim = true;
 +
elseif str:sub(-4,-1) == duplicate_char .. "'']" then -- normal case when |url=something & |title=Title.
 +
trim = true;
 +
end
 +
elseif end_chr == " " then -- if last char of output string is a space
 +
if str:sub(-2,-1) == duplicate_char .. " " then -- if last two chars of str are <sepc><space>
 +
str = str:sub(1,-3); -- remove them both
 +
end
 +
end
 +
 
 +
if trim then
 +
if value ~= comp then -- value does not equal comp when value contains html markup
 +
local dup2 = duplicate_char;
 +
if dup2:match( "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it
 +
 +
value = value:gsub( "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows html markup
 +
else
 +
value = value:sub( 2, -1 ); -- remove duplicate_char when it is first character
 +
end
 
end
 
end
elseif list ~= nil then
 
v, origin[k] = args[list], list;
 
else
 
-- maybe let through instead of raising an error?
 
-- v, origin[k] = args[k], k;
 
error( cfg.messages['unknown_argument_map'] );
 
 
end
 
end
+
str = str .. value; --add it to the output string
-- Empty strings, not nil;
+
end
if v == nil then
+
end
v = cfg.defaults[k] or '';
+
return str;
origin[k] = '';
+
end 
end
+
 
+
-- Attempts to convert names to initials.
tbl = rawset( tbl, k, v );
+
function reducetoinitials(first)
return v;
+
local initials = {}
end,
+
local i = 0; -- counter for number of initials
});
+
for word in string.gmatch(first, "%S+") do
 +
table.insert(initials, string.sub(word,1,1)) -- Vancouver format does not include full stops.
 +
i = i + 1; -- bump the counter
 +
if 2 <= i then break; end -- only two initials allowed in Vancouver system; if 2, quit
 +
end
 +
return table.concat(initials) -- Vancouver format does not include spaces.
 
end
 
end
  
 
+
-- Formats a list of people (e.g. authors / editors)
--[[--------------------------< N O W R A P _ D A T E >--------------------------------------------------------
+
function listpeople(control, people)
 
+
local sep;
When date is YYYY-MM-DD format wrap in nowrap span: <span ...>YYYY-MM-DD</span>.  When date is DD MMMM YYYY or is
+
local namesep;
MMMM DD, YYYY then wrap in nowrap span: <span ...>DD MMMM</span> YYYY or <span ...>MMMM DD,</span> YYYY
+
local format = control.format
 
+
local maximum = control.maximum
DOES NOT yet support MMMM YYYY or any of the date ranges.
+
local lastauthoramp = control.lastauthoramp;
 
+
local text = {}
]]
+
local etal = false;
 
 
local function nowrap_date (date)
 
local cap='';
 
local cap2='';
 
 
 
if date:match("^%d%d%d%d%-%d%d%-%d%d$") then
 
date = substitute (cfg.presentation['nowrap1'], date);
 
 
 
elseif date:match("^%a+%s*%d%d?,%s+%d%d%d%d$") or date:match ("^%d%d?%s*%a+%s+%d%d%d%d$") then
+
if 'vanc' == format then -- Vancouver-like author/editor name styling?
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
+
sep = ','; -- name-list separator between authors is a comma
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
+
namesep = ' '; -- last/first separator is a space
 +
else
 +
sep = ';' -- name-list separator between authors is a semicolon
 +
namesep = ', ' -- last/first separator is <comma><space>
 
end
 
end
 
 
return date;
+
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
end
+
if maximum ~= nil and maximum < 1 then return "", 0; end
 
+
 +
for i,person in ipairs(people) do
 +
if is_set(person.last) then
 +
local mask = person.mask
 +
local one
 +
local sep_one = sep;
 +
if maximum ~= nil and i > maximum then
 +
etal = true;
 +
break;
 +
elseif (mask ~= nil) then
 +
local n = tonumber(mask)
 +
if (n ~= nil) then
 +
one = string.rep("&mdash;",n)
 +
else
 +
one = mask;
 +
sep_one = " ";
 +
end
 +
else
 +
one = person.last
 +
local first = person.first
 +
if is_set(first) then
 +
if ( "vanc" == format ) then first = reducetoinitials(first) end
 +
one = one .. namesep .. first
 +
end
 +
if is_set(person.link) and person.link ~= control.page_name then
 +
one = "[[" .. person.link .. "|" .. one .. "]]" -- link author/editor if this page is not the author's/editor's page
 +
end
  
--[[--------------------------< S E T _ T I T L E T Y P E >----------------------------------------------------
+
if is_set(person.link) and ((nil ~= person.link:find("//")) or (nil ~= person.link:find("[%[%]]"))) then
 +
one = one .. " " .. seterror( 'bad_authorlink' ) end -- url or wikilink in author link;
 +
end
 +
table.insert( text, one )
 +
table.insert( text, sep_one )
 +
end
 +
end
  
This function sets default title types (equivalent to the citation including |type=<default value>) for those templates that have defaults.
+
local count = #text / 2;
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
+
if count > 0 then  
 
+
if count > 1 and is_set(lastauthoramp) and not etal then
]]
+
text[#text-2] = " & ";
 
 
local function set_titletype (cite_class, title_type)
 
if is_set(title_type) then
 
if "none" == title_type then
 
title_type = ""; -- if |type=none then type parameter not displayed
 
 
end
 
end
return title_type; -- if |type= has been set to any other value use that value
+
text[#text] = nil;  
 +
end
 +
 +
local result = table.concat(text) -- construct list
 +
if etal then
 +
local etal_text = cfg.messages['et al'];
 +
result = result .. " " .. etal_text;
 
end
 
end
 
+
return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation
+
return result, count
 
end
 
end
  
 +
--[[--------------------------< A N C H O R I D >--------------------------------------------------------------
 +
Generates a CITEREF anchor ID if we have at least one name or a date.  Otherwise returns an empty string.
  
--[[--------------------------< H Y P H E N _ T O _ D A S H >--------------------------------------------------
+
]]
  
Converts a hyphen to a dash under certain conditions.  The hyphen must separate like items; unlike items are
+
function anchorid( options )
returned unmodified. These forms are modified:
+
local id = table.concat( options ); -- concatenate names and year for CITEREF id
letter - letter (A - B)
+
if is_set (id) then -- if concatenation is not an empty string
digit - digit (4-5)
+
return "CITEREF" .. id; -- add the CITEREF portion
digit separator digit - digit separator digit (4.1-4.5 or 4-1-4-5)
+
else
letterdigit - letterdigit (A1-A5) (an optional separator between letter and digit is supported – a.1-a.5 or a-1-a-5)
+
return ''; -- return an empty string; no reason to include CITEREF id in this citation
digitletter - digitletter (5a - 5d) (an optional separator between letter and digit is supported – 5.a-5.d or 5-a-5-d)
+
end
 +
end
  
any other forms are returned unmodified.
+
--[[
 +
Gets name list from the input arguments
  
str may be a comma- or semicolon-separated list
+
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
 +
Stops searching when both |lastn= and |firstn= are not found in args after two sequential attempts: found |last1=, |last2=, and |last3= but doesn't
 +
find |last4= and |last5= then the search is done.
  
 +
This function emits an error message when there is a |firstn= without a matching |lastn=.  When there are 'holes' in the list of last names, |last1= and |last3=
 +
are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=.
 
]]
 
]]
 
+
function extractnames(args, list_name)
local function hyphen_to_dash( str )
+
local names = {}; -- table of names
if not is_set (str) then
+
local last; -- individual name components
return str;
+
local first;
end
+
local link;
 +
local mask;
 +
local i = 1; -- loop counter/indexer
 +
local n = 1; -- output table indexer
 +
local count = 0; -- used to count the number of times we haven't found a |last= (or alias for authors, |editor-last or alias for editors)
 
 
str, count = str:gsub ('^%(%((.+)%)%)$', '%1'); -- remove accept-this-as-written markup when it wraps all of str
+
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
if 0 ~= count then -- non-zero when markup removed; zero else
+
 
return str; -- nothing to do, we're done
+
while true do
end
+
last = selectone( args, cfg.aliases[list_name .. '-Last'], 'redundant_parameters', i ); -- search through args for name components beginning at 1
+
first = selectone( args, cfg.aliases[list_name .. '-First'], 'redundant_parameters', i );
str = str:gsub ('&[nm]dash;', {['&ndash;'] = '', ['&mdash;'] = ''}); -- replace &mdash; and &ndash; entities  with their characters; semicolon mucks up the text.split
+
link = selectone( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i );
+
mask = selectone( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
local out = {};
 
local list = mw.text.split (str, '%s*[,;]%s*'); -- split str at comma or semicolon separators if there are any
 
  
for _, item in ipairs (list) do -- for each item in the list
+
if first and not last then -- if there is a firstn without a matching lastn
if mw.ustring.match (item, '^%w*[%.%-]?%w+%s*[%-–—]%s*%w*[%.%-]?%w+$') then -- if a hyphenated range or has endash or emdash separators
+
table.insert( z.message_tail, { seterror( 'first_missing_last', {err_msg_list_name, i}, true ) } ); -- add this error message
if item:match ('%a+[%.%-]?%d+%s*%-%s*%a+[%.%-]?%d+') or -- letterdigit hyphen letterdigit (optional separator between letter and digit)
+
elseif not first and not last then -- if both firstn and lastn aren't found, are we done?
item:match ('%d+[%.%-]?%a+%s*%-%s*%d+[%.%-]?%a+') or -- digitletter hyphen digitletter (optional separator between digit and letter)
+
count = count + 1; -- number of times we haven't found last and first
item:match ('%d+[%.%-]%d+%s*%-%s*%d+[%.%-]%d+') or -- digit separator digit hyphen digit separator digit
+
if 2 == count then -- two missing names and we give up
item:match ('%d+%s*%-%s*%d+') or -- digit hyphen digit
+
break; -- normal exit or there is a two-name hole in the list; can't tell which
item:match ('%a+%s*%-%s*%a+') then -- letter hyphen letter
+
end
item = item:gsub ('(%w*[%.%-]?%w+)%s*%-%s*(%w*[%.%-]?%w+)', '%1–%2'); -- replace hyphen, remove extraneous space characters
+
else -- we have last with or without a first
else
+
names[n] = {last = last, first = first, link = link, mask = mask}; -- add this name to our names list
item = mw.ustring.gsub (item, '%s*[–—]%s*', '–'); -- for endash or emdash separated ranges, replace em with en, remove extraneous white space
+
n = n + 1; -- point to next location in the names table
 +
if 1 == count then -- if the previous name was missing
 +
table.insert( z.message_tail, { seterror( 'missing_name', {err_msg_list_name, i-1}, true ) } ); -- add this error message
 
end
 
end
 +
count = 0; -- reset the counter, we're looking for two consecutive missing names
 
end
 
end
item = item:gsub ('^%(%((.+)%)%)$', '%1'); -- remove the accept-this-as-written markup
+
i = i + 1; -- point to next args location
table.insert (out, item); -- add the (possibly modified) item to the output table
 
 
end
 
end
 +
return names; -- all done, return our list of names
 +
end
  
return table.concat (out, ', '); -- concatenate the output table into a comma separated string
+
-- Populates ID table from arguments using configuration settings
 +
function extractids( args )
 +
local id_list = {};
 +
for k, v in pairs( cfg.id_handlers ) do
 +
v = selectone( args, v.parameters, 'redundant_parameters' );
 +
if is_set(v) then id_list[k] = v; end
 +
end
 +
return id_list;
 
end
 
end
  
 +
--[[--------------------------< B U I L D I D L I S T >--------------------------------------------------------
 +
Takes a table of IDs and turns it into a table of formatted ID outputs.
  
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
+
]]
 +
function buildidlist( id_list, options )
 +
local new_list, handler = {};
  
Joins a sequence of strings together while checking for duplicate separation characters.
+
function fallback(k) return { __index = function(t,i) return cfg.id_handlers[k][i] end } end;
 
+
]]
+
for k, v in pairs( id_list ) do
 
+
-- fallback to read-only cfg
local function safe_join( tbl, duplicate_char )
+
handler = setmetatable( { ['id'] = v }, fallback(k) );
local f = {}; -- create a function table appropriate to type of 'dupicate character'
 
if 1 == #duplicate_char then -- for single byte ascii characters use the string library functions
 
f.gsub=string.gsub
 
f.match=string.match
 
f.sub=string.sub
 
else -- for multi-byte characters use the ustring library functions
 
f.gsub=mw.ustring.gsub
 
f.match=mw.ustring.match
 
f.sub=mw.ustring.sub
 
end
 
 
 
local str = ''; -- the output string
 
local comp = ''; -- what does 'comp' mean?
 
local end_chr = '';
 
local trim;
 
for _, value in ipairs( tbl ) do
 
if value == nil then value = ''; end
 
 
 
if str == '' then -- if output string is empty
+
if handler.mode == 'external' then
str = value; -- assign value to it (first time through the loop)
+
table.insert( new_list, {handler.label, externallinkid( handler ) } );
elseif value ~= '' then
+
elseif handler.mode == 'internal' then
if value:sub(1,1) == '<' then -- Special case of values enclosed in spans and other markup.
+
table.insert( new_list, {handler.label, internallinkid( handler ) } );
comp = value:gsub( "%b<>", "" ); -- remove html markup (<span>string</span> -> string)
+
elseif handler.mode ~= 'manual' then
else
+
error( cfg.messages['unknown_ID_mode'] );
comp = value;
+
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 ) } );
 +
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' 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 == 'ISSN' then
 +
table.insert( new_list, {handler.label, issn( v ) } );
 +
elseif k == 'ISBN' then
 +
local ISBN = internallinkid( handler );
 +
if not checkisbn( v ) and not is_set(options.IgnoreISBN) then
 +
ISBN = ISBN .. seterror( 'bad_isbn', {}, false, " ", "" );
 
end
 
end
-- typically duplicate_char is sepc
+
table.insert( new_list, {handler.label, ISBN } );
if f.sub(comp, 1,1) == duplicate_char then -- is first character same as duplicate_char? why test first character?
+
elseif k == 'USENETID' then
--  Because individual string segments often (always?) begin with terminal punct for the
+
table.insert( new_list, {handler.label, message_id( v ) } );
--  preceding segment: 'First element' .. 'sepc next element' .. etc?
+
else
trim = false;
+
error( cfg.messages['unknown_manual_ID'] );
end_chr = f.sub(str, -1,-1); -- get the last character of the output string
+
end
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
+
end
if end_chr == duplicate_char then -- if same as separator
+
str = f.sub(str, 1,-2); -- remove it
+
function comp( a, b ) -- used in following table.sort()
elseif end_chr == "'" then -- if it might be wikimarkup
+
return a[1] < b[1];
if f.sub(str, -3,-1) == duplicate_char .. "''" then -- if last three chars of str are sepc''
+
end
str = f.sub(str, 1, -4) .. "''"; -- remove them and add back ''
+
elseif  f.sub(str, -5,-1) == duplicate_char .. "]]''" then -- if last five chars of str are sepc]]''
+
table.sort( new_list, comp );
trim = true; -- why? why do this and next differently from previous?
+
for k, v in ipairs( new_list ) do
elseif f.sub(str, -4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''
+
new_list[k] = v[2];
trim = true; -- same question
+
end
end
+
elseif end_chr == "]" then -- if it might be wikimarkup
+
return new_list;
if f.sub(str, -3,-1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink
+
end
trim = true;
+
 
elseif f.sub(str, -3,-1) == duplicate_char .. '"]' then -- if last three chars of str are sepc"] quoted external link
+
-- Chooses one matching parameter from a list of parameters to consider
trim = true;
+
-- Generates an error if more than one match is present.
elseif  f.sub(str, -2,-1) == duplicate_char .. "]" then -- if last two chars of str are sepc] external link
+
function selectone( args, possible, error_condition, index )
trim = true;
+
local value = nil;
elseif f.sub(str, -4,-1) == duplicate_char .. "'']" then -- normal case when |url=something & |title=Title.
+
local selected = '';
trim = true;
+
local error_list = {};
end
+
elseif end_chr == " " then -- if last char of output string is a space
+
if index ~= nil then index = tostring(index); end
if f.sub(str, -2,-1) == duplicate_char .. " " then -- if last two chars of str are <sepc><space>
+
str = f.sub(str, 1,-3); -- remove them both
+
-- Handle special case of "#" replaced by empty string
end
+
if index == '1' then
end
+
for _, v in ipairs( possible ) do
 
+
v = v:gsub( "#", "" );
if trim then
+
if is_set(args[v]) then
if value ~= comp then -- value does not equal comp when value contains html markup
+
if value ~= nil and selected ~= v then
local dup2 = duplicate_char;
+
table.insert( error_list, v );
if f.match(dup2, "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it
+
else
+
value = args[v];
value = f.gsub(value, "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows html markup
+
selected = v;
else
 
value = f.sub(value, 2, -1 ); -- remove duplicate_char when it is first character
 
end
 
 
end
 
end
 
end
 
end
str = str .. value; --add it to the output string
+
end
 +
end
 +
 +
for _, v in ipairs( possible ) do
 +
if index ~= nil then
 +
v = v:gsub( "#", index );
 +
end
 +
if is_set(args[v]) then
 +
if value ~= nil and selected ~=  v then
 +
table.insert( error_list, v );
 +
else
 +
value = args[v];
 +
selected = v;
 +
end
 +
end
 +
end
 +
 +
if #error_list > 0 then
 +
local error_str = "";
 +
for _, k in ipairs( error_list ) do
 +
if error_str ~= "" then error_str = error_str .. cfg.messages['parameter-separator'] end
 +
error_str = error_str .. wrap_style ('parameter', k);
 +
end
 +
if #error_list > 1 then
 +
error_str = error_str .. cfg.messages['parameter-final-separator'];
 +
else
 +
error_str = error_str .. cfg.messages['parameter-pair-separator'];
 
end
 
end
 +
error_str = error_str .. wrap_style ('parameter', selected);
 +
table.insert( z.message_tail, { seterror( error_condition, {error_str}, true ) } );
 
end
 
end
return str;
+
 +
return value, selected;
 
end
 
end
  
 
+
-- COinS metadata (see <http://ocoins.info/>) allows automated tools to parse
--[[--------------------------< I S _ S U F F I X >------------------------------------------------------------
+
-- the citation information.
 
+
function COinS(data)
returns true is suffix is properly formed Jr, Sr, or ordinal in the range 2–9.  Puncutation not allowed.
+
if 'table' ~= type(data) or nil == next(data) then
 
+
return '';
]]
 
 
 
local function is_suffix (suffix)
 
if in_array (suffix, {'Jr', 'Sr', '2nd', '3rd'}) or suffix:match ('^%dth$') then
 
return true;
 
 
end
 
end
return false;
+
end
+
local ctx_ver = "Z39.88-2004";
 
+
 
+
-- treat table strictly as an array with only set values.
--[[--------------------------< I S _ G O O D _ V A N C _ N A M E >--------------------------------------------
+
local OCinSoutput = setmetatable( {}, {
 
+
__newindex = function(self, key, value)
For Vancouver Style, author/editor names are supposed to be rendered in Latin (read ASCII) characters.  When a name
+
if is_set(value) then
uses characters that contain diacritical marks, those characters are to converted to the corresponding Latin character.
+
rawset( self, #self+1, table.concat{ key, '=', mw.uri.encode( removewikilink( value ) ) } );
When a name is written using a non-Latin alphabet or logogram, that name is to be transliterated into Latin characters.
+
end
These things are not currently possible in this module so are left to the editor to do.
+
end
 
+
});
This test allows |first= and |last= names to contain any of the letters defined in the four Unicode Latin character sets
+
[http://www.unicode.org/charts/PDF/U0000.pdf C0 Controls and Basic Latin] 0041–005A, 0061–007A
+
if is_set(data.Chapter) then
[http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] 00C0–00D6, 00D8–00F6, 00F8–00FF
+
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:book";
[http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] 0100–017F
+
OCinSoutput["rft.genre"] = "bookitem";
[http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] 0180–01BF, 01C4–024F
+
OCinSoutput["rft.atitle"] = data.Chapter;
 
+
OCinSoutput["rft.btitle"] = data.Title;
|lastn= also allowed to contain hyphens, spaces, and apostrophes. (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
+
elseif is_set(data.Periodical) then
|firstn= also allowed to contain hyphens, spaces, apostrophes, and periods
+
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:journal";
 
+
OCinSoutput["rft.genre"] = "article";
This original test:
+
OCinSoutput["rft.jtitle"] = data.Periodical;
if nil == mw.ustring.find (last, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ-ɏ%-%s%']*$") or nil == mw.ustring.find (first, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ-ɏ%-%s%'%.]+[2-6%a]*$") then
+
OCinSoutput["rft.atitle"] = data.Title;
was written ouside of the code editor and pasted here because the code editor gets confused between character insertion point and cursor position.
+
else
The test has been rewritten to use decimal character escape sequence for the individual bytes of the unicode characters so that it is not necessary
+
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:book";
to use an external editor to maintain this code.
+
OCinSoutput["rft.genre"] = "book"
 
+
OCinSoutput["rft.btitle"] = data.Title;
\195\128-\195\150 – À-Ö (U+00C0–U+00D6 – C0 controls)
+
end
\195\152-\195\182 – Ø-ö (U+00D8-U+00F6 – C0 controls)
+
\195\184-\198\191 – ø-ƿ (U+00F8-U+01BF – C0 controls, Latin extended A & B)
+
OCinSoutput["rft.place"] = data.PublicationPlace;
\199\132-\201\143 – DŽ-ɏ (U+01C4-U+024F – Latin extended B)
+
OCinSoutput["rft.date"] = data.Date;
 
+
OCinSoutput["rft.series"] = data.Series;
]]
+
OCinSoutput["rft.volume"] = data.Volume;
 
+
OCinSoutput["rft.issue"] = data.Issue;
local function is_good_vanc_name (last, first)
+
OCinSoutput["rft.pages"] = data.Pages;
local first, suffix = first:match ('(.-),?%s*([%dJS][%drndth]+)%.?$') or first; -- if first has something that looks like a generational suffix, get it
+
OCinSoutput["rft.edition"] = data.Edition;
 
+
OCinSoutput["rft.pub"] = data.PublisherName;
if is_set (suffix) then
+
if not is_suffix (suffix) then
+
for k, v in pairs( data.ID_list ) do
add_vanc_error ('suffix');
+
local id, value = cfg.id_handlers[k].COinS;
return false; -- not a name with an appropriate suffix
+
if k == 'ISBN' then value = cleanisbn( v ); else value = v; end
 +
if string.sub( id or "", 1, 4 ) == 'info' then
 +
OCinSoutput["rft_id"] = table.concat{ id, "/", v };
 +
else
 +
OCinSoutput[ id ] = value;
 +
end
 +
end
 +
 +
local last, first;
 +
for k, v in ipairs( data.Authors ) do
 +
last, first = v.last, v.first;
 +
if k == 1 then
 +
if is_set(last) then
 +
OCinSoutput["rft.aulast"] = last;
 +
end
 +
if is_set(first) then  
 +
OCinSoutput["rft.aufirst"] = first;
 +
end
 +
end
 +
if is_set(last) and is_set(first) then
 +
OCinSoutput["rft.au"] = table.concat{ last, ", ", first };
 +
elseif is_set(last) then
 +
OCinSoutput["rft.au"] = last;
 
end
 
end
 
end
 
end
if nil == mw.ustring.find (last, "^[A-Za-z\195\128-\195\150\195\152-\195\182\195\184-\198\191\199\132-\201\143%-%s%']*$") or
+
nil == mw.ustring.find (first, "^[A-Za-z\195\128-\195\150\195\152-\195\182\195\184-\198\191\199\132-\201\143%-%s%'%.]*$") then
+
OCinSoutput.rft_id = data.URL;
add_vanc_error ('non-Latin character');
+
OCinSoutput.rfr_id = table.concat{ "info:sid/", mw.site.server:match( "[^/]*$" ), ":", data.RawPage };
return false; -- not a string of latin characters; Vancouver requires Romanization
+
OCinSoutput = setmetatable( OCinSoutput, nil );
end;
+
return true;
+
-- sort with version string always first, and combine.
 +
table.sort( OCinSoutput );
 +
table.insert( OCinSoutput, 1, "ctx_ver=" .. ctx_ver ); -- such as "Z39.88-2004"
 +
return table.concat(OCinSoutput, "&");
 
end
 
end
  
  
--[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------
+
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
  
Attempts to convert names to initials in support of |name-list-format=vanc.
+
Validates language names provided in |language= parameter if not an ISO639-1 code.  Handles the special case that is Norwegian where
 +
ISO639-1 code 'no' is mapped to language name 'Norwegian Bokmål' by Extention:CLDR.
  
Names in |firstn= may be separated by spaces or hyphens, or for initials, a period. See http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35062/.
+
Returns the language name and associated ISO639-1 code.  Because case of the source may be incorrect or different from the case that Wikimedia
 +
uses, the name comparisons are done in lower case and when a match is found, the Wikimedia version (assumed to be correct) is returned along
 +
with the code. When there is no match, we return the original language name string.
  
Vancouver style requires family rank designations (Jr, II, III, etc) to be rendered as Jr, 2nd, 3rd, etcSee http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35085/.
+
mw.language.fetchLanguageNames() will return a list of languages that aren't part of ISO639-1. Names that aren't ISO639-1 but that are included
This code only accepts and understands generational suffix in the Vancouver format because Roman numerals look like, and can be mistaken for, initials.
+
in the list will be found if that name is provided in the |language= parameter.  For example, if |language=Samaritan Aramaic, that name will be
 
+
found with the associated code 'sam', not an ISO639-1 codeWhen names are found and the associated code is not two characters, this function
This function uses ustring functions because firstname initials may be any of the unicode Latin characters accepted by is_good_vanc_name ().
+
returns only the Wikimedia language name.
  
 +
Adapted from code taken from Module:Check ISO 639-1.
 
]]
 
]]
  
local function reduce_to_initials(first)
+
function get_iso639_code (lang)
local name, suffix = mw.ustring.match(first, "^(%u+) ([%dJS][%drndth]+)$");
+
if 'norwegian' == lang:lower() then -- special case related to Wikimedia remap of code 'no' at Extension:CLDR
 
+
return 'Norwegian', 'no'; -- Make sure rendered version is properly capitalized
if not name then -- if not initials and a suffix
 
name = mw.ustring.match(first, "^(%u+)$"); -- is it just intials?
 
 
end
 
end
 
+
if name then -- if first is initials with or without suffix
+
local languages = mw.language.fetchLanguageNames('en', 'all') -- get a list of language names known to Wikimedia
if 3 > mw.ustring.len (name) then -- if one or two initials
+
-- ('all' is required for North Ndebele, South Ndebele, and Ojibwa)
if suffix then -- if there is a suffix
+
local langlc = mw.ustring.lower(lang); -- lower case version for comparisons
if is_suffix (suffix) then -- is it legitimate?
+
return first; -- one or two initials and a valid suffix so nothing to do
+
for code, name in pairs(languages) do -- scan the list to see if we can find our language
else
+
if langlc == mw.ustring.lower(name) then
add_vanc_error ('suffix'); -- one or two initials with invalid suffix so error message
+
if 2 ~= code:len() then -- ISO639-1 codes only
return first; -- and return first unmolested
+
return name; -- so return the name but not the code
end
 
else
 
return first; -- one or two initials without suffix; nothing to do
 
 
end
 
end
 +
return name, code; -- found it, return name to ensure proper capitalization and the ISO639-1 code
 
end
 
end
end -- if here then name has 3 or more uppercase letters so treat them as a word
+
end
 +
return lang; -- not valid language; return language in original case and nil for ISO639-1 code
 +
end
  
 +
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
  
local initials, names = {}, {}; -- tables to hold name parts and initials
+
Get language name from ISO639-1 code value provided.  If a code is valid use the returned name; if not, then use the value that was provided with the language parameter.
local i = 1; -- counter for number of initials
 
  
names = mw.text.split (first, '[%s,]+'); -- split into a table of names and possible suffix
+
There is an exception. There are three ISO639-1 codes for Norewegian language variants. There are two official variants: Norwegian Bokmål (code 'nb') and
 +
Norwegian Nynorsk (code 'nn').  The third, code 'no', is defined by ISO639-1 as 'Norwegian' though in Norway this is pretty much meaningless.  However, it appears
 +
that on enwiki, editors are for the most part unaware of the nb and nn variants (compare page counts for these variants at Category:Articles with non-English-language external links.
  
while names[i] do -- loop through the table
+
Because Norwegian Bokmål is the most common language variant, Media wiki has been modified to return Norwegian Bokmål for ISO639-1 code 'no'. Here we undo that and
if 1 < i and names[i]:match ('[%dJS][%drndth]+%.?$') then -- if not the first name, and looks like a suffix (may have trailing dot)
+
return 'Norwegian' when editors use |language=no. We presume that editors don't know about the variants or can't descriminate between them.
names[i] = names[i]:gsub ('%.', ''); -- remove terminal dot if present
 
if is_suffix (names[i]) then -- if a legitimate suffix
 
table.insert (initials, ' ' .. names[i]); -- add a separator space, insert at end of initials table
 
break; -- and done because suffix must fall at the end of a name
 
end -- no error message if not a suffix; possibly because of Romanization
 
end
 
if 3 > i then
 
table.insert (initials, mw.ustring.sub(names[i],1,1)); -- insert the intial at end of initials table
 
end
 
i = i+1; -- bump the counter
 
end
 
 
return table.concat(initials) -- Vancouver format does not include spaces.
 
end
 
  
 +
See Help talk:Citation Style_1#An ISO 639-1 language name test
  
--[[--------------------------< L I S T  _ P E O P L E >-------------------------------------------------------
+
When |language= contains a valid ISO639-1 code, the page is assigned to the category for that code: Category:Norwegian-language sources (no) if
 
+
the page is a mainspace page and the ISO639-1 code is not 'en'. Similarly, if the  parameter is |language=Norwegian, it will be categorized in the same way.
Formats a list of people (e.g. authors / editors)
 
  
 
]]
 
]]
  
local function list_people(control, people, etal)
+
function language_parameter (lang, namespace)
local sep;
+
local code; -- the ISO639-1 two character code
local namesep;
+
local name; -- the language name
local format = control.format
+
local test='';
local maximum = control.maximum
+
local lastauthoramp = control.lastauthoramp;
+
if 0 == namespace and (('en' == lang:lower()) or ('english' == lang:lower())) then
local text = {}
+
table.insert (z.maintenance_cats, 'CS1 maint: English language specified'); -- add maintenance category if |language=English or |language=en in article space
 +
end
  
if 'vanc' == format then -- Vancouver-like author/editor name styling?
+
if 2 == lang:len() then -- ISO639-1 language code are 2 characters (fetchLanguageName also supports 3 character codes)
sep = cfg.presentation['sep_nl_vanc']; -- name-list separator between authors is a comma
+
name = mw.language.fetchLanguageName( lang:lower(), "en" ); -- get ISO 639-1 language name if Language is a proper code
namesep = cfg.presentation['sep_name_vanc']; -- last/first separator is a space
+
end
 +
 
 +
if is_set (name) then -- if Language specified a valid ISO639-1 code
 +
code = lang:lower(); -- save it
 +
else
 +
name, code = get_iso639_code (lang); -- attempt to get code from name (assign name here so that we are sure of proper capitalization)
 +
end
 +
 
 +
if is_set (code) then
 +
if 'no' == code then name = 'Norwegian' end; -- override wikimedia when code is 'no'
 +
if 0 == namespace and 'en' ~= code then -- is this page main / article space and English not the language?
 +
table.insert( z.properties_cats, 'CS1 ' .. name .. '-language sources (' .. code .. ')'); -- in main space and not English: categorize
 +
end
 
else
 
else
sep = cfg.presentation['sep_nl']; -- name-list separator between authors is a semicolon
+
table.insert (z.maintenance_cats, 'CS1 maint: Unrecognized language'); -- add maintenance category when |language= does not appear to be ISO 639-1 language
namesep = cfg.presentation['sep_name']; -- last/first separator is <comma><space>
 
 
end
 
end
+
return (" " .. wrap_msg ('language', name)); -- wrap with '(in ...)'
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
+
end
if is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for authors
+
 
+
--[[--------------------------< G E T _ S E T T I N G S _ F R O M _ C I T E _ C L A S S >----------------------
for i,person in ipairs(people) do
+
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
if is_set(person.last) then
+
rendered style.
local mask = person.mask
+
 
local one
+
]]
local sep_one = sep;
+
 
if is_set (maximum) and i > maximum then
+
function get_settings_from_cite_class (ps, ref, cite_class)
etal = true;
+
local sep;
break;
+
if (cite_class == "citation") then -- for citation templates (CS2)
elseif (mask ~= nil) then
+
sep = ','; -- set citation separator to its default (comma)
local n = tonumber(mask)
+
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
if (n ~= nil) then
+
ps = ''; -- make sure it isn't nil
one = string.rep("&mdash;",n)
+
end
else
+
if not is_set (ref) then -- if |ref= is not set
one = mask;
+
ref = "harv"; -- set default |ref=harv
sep_one = " ";
+
end
end
+
else -- not a citation template so CS1
else
+
sep = '.'; -- set cite xxx separator to its default (period)
one = person.last
+
if not is_set (ps) then -- if |postscript= has not been set
local first = person.first
+
ps = '.'; -- set cs1 default
if is_set(first) then
 
if ( "vanc" == format ) then -- if vancouver format
 
one = one:gsub ('%.', ''); -- remove periods from surnames (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
 
if not person.corporate and is_good_vanc_name (one, first) then -- and name is all Latin characters; corporate authors not tested
 
first = reduce_to_initials(first) -- attempt to convert first name(s) to initials
 
end
 
end
 
one = one .. namesep .. first;
 
end
 
if is_set(person.link) and person.link ~= control.page_name then
 
one = make_wikilink (person.link, one); -- link author/editor if this page is not the author's/editor's page
 
end
 
end
 
table.insert( text, one )
 
table.insert( text, sep_one )
 
 
end
 
end
 
end
 
end
  
local count = #text / 2; -- (number of names + number of separators) divided by 2
+
return sep, ps, ref -- return them all
if count > 0 then
 
if count > 1 and is_set(lastauthoramp) and not etal then
 
text[#text-2] = " & "; -- replace last separator with ampersand text
 
end
 
text[#text] = nil; -- erase the last separator
 
end
 
 
local result = table.concat(text) -- construct list
 
if etal and is_set (result) then -- etal may be set by |display-authors=etal but we might not have a last-first list
 
result = result .. sep .. ' ' .. cfg.messages['et al']; -- we've go a last-first list and etal so add et al.
 
end
 
 
return result, count
 
 
end
 
end
  
 +
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
  
--[[--------------------------< A N C H O R _ I D >------------------------------------------------------------
+
Establish basic style settings to be used when rendering the citation.  Uses |mode= if set and valid or uses
 +
config.CitationClass from the template's #invoke: to establish style.
  
Generates a CITEREF anchor ID if we have at least one name or a date.  Otherwise returns an empty string.
+
]]
  
namelist is one of the contributor-, author-, or editor-name lists chosen in that order. year is Year or anchor_year.
+
function set_style (mode, ps, ref, cite_class)
 
+
local sep;
]]
+
if is_set (mode) then
local function anchor_id (namelist, year)
+
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
local names={}; -- a table for the one to four names and year
+
sep = ','; -- separate elements with a comma
for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names
+
if not is_set (ps) then -- unless explicitely set to something
names[i] = v.last
+
ps = ''; -- make sure it isn't nil
if i == 4 then break end -- if four then done
+
end
 +
if not is_set (ref) then -- unless explicitely set to something
 +
ref = 'harv'; -- set so this template renders with CITEREF anchor id
 +
end
 +
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
 +
sep = '.'; -- separate elements with a period
 +
if not is_set (ps) then -- unless explicitely set to something
 +
ps = '.'; -- terminate the rendered citation with a period
 +
end
 +
else -- anything but cs1 or cs2
 +
table.insert( z.message_tail, { seterror( 'invalid_param_val', {'mode', mode}, true ) } ); -- add error message
 +
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
 +
end
 +
else -- when |mode= empty or omitted
 +
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
 
end
 
end
table.insert (names, year); -- add the year at the end
+
if 'none' == ps:lower() then -- if assigned value is 'none' then
local id = table.concat(names); -- concatenate names and year for CITEREF id
+
ps = ''; -- set to empty string
if is_set (id) then -- if concatenation is not an empty string
 
return "CITEREF" .. id; -- add the CITEREF portion
 
else
 
return ''; -- return an empty string; no reason to include CITEREF id in this citation
 
 
end
 
end
 +
 +
return sep, ps, ref
 
end
 
end
  
  
--[[--------------------------< N A M E _ H A S _ E T A L >----------------------------------------------------
+
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
 
+
This is the main function doing the majority of the citation
Evaluates the content of name parameters (author, editor, etc) for variations on the theme of et al.  If found,
+
formatting.
the et al. is removed, a flag is set to true and the function returns the modified name and the flag.
 
 
 
This function never sets the flag to false but returns it's previous state because it may have been set by
 
previous passes through this function or by the associated |display-<names>=etal parameter
 
 
 
 
]]
 
]]
 +
function citation0( config, args)
 +
--[[
 +
Load Input Parameters
 +
The argument_wrapper facilitates the mapping of multiple
 +
aliases to single internal variable.
 +
]]
 +
local A = argument_wrapper( args );
  
local function name_has_etal (name, etal, nocat, param)
+
local i
 +
local PPrefix = A['PPrefix']
 +
local PPPrefix = A['PPPrefix']
 +
if is_set( A['NoPP'] ) then PPPrefix = "" PPrefix = "" end
 +
 +
-- Pick out the relevant fields from the arguments.  Different citation templates
 +
-- define different field names for the same underlying things.
 +
local Authors = A['Authors'];
 +
local a = extractnames( args, 'AuthorList' );
  
if is_set (name) then -- name can be nil in which case just return
+
local Coauthors = A['Coauthors'];
local patterns = cfg.et_al_patterns; --get patterns from configuration
+
local Others = A['Others'];
+
local Editors = A['Editors'];
for _, pattern in ipairs (patterns) do -- loop through all of the patterns
+
local e = extractnames( args, 'EditorList' );
if name:match (pattern) then -- if this 'et al' pattern is found in name
 
name = name:gsub (pattern, ''); -- remove the offending text
 
etal = true; -- set flag (may have been set previously here or by |display-<names>=etal)
 
if not nocat then -- no categorization for |vauthors=
 
table.insert( z.message_tail, {set_error ('etal', {param})}); -- and set an error if not added
 
end
 
end
 
end
 
end
 
 
 
return name, etal; --
 
end
 
 
 
 
 
--[[--------------------------< N A M E _ H A S _ E D _ M A R K U P >------------------------------------------
 
 
 
Evaluates the content of author and editor parameters for extranious editor annotations: ed, ed., eds, (Ed.), etc.
 
These annotation do not belong in author parameters and are redundant in editor parameters.  If found, the function
 
adds the editor markup maintenance category.
 
 
 
]]
 
  
local function name_has_ed_markup (name, list_name)
+
local NameListFormat = A['NameListFormat']; -- replaces |author-format= and |editor-format=
local _, pattern;
+
if is_set (NameListFormat) and ('vanc' ~= NameListFormat) then -- only accepted value for this parameter is 'vanc'
local patterns = cfg.editor_markup_patterns; -- get patterns from configuration
+
table.insert( z.message_tail, { seterror( 'invalid_param_val', {'name-list-format', NameListFormat}, true ) } ); -- not vanc so add error message
 
+
NameListFormat = ''; -- set to empty string
if is_set (name) then
 
for _, pattern in ipairs (patterns) do -- spin through patterns table and
 
if name:match (pattern) then
 
add_maint_cat ('extra_text_names', cfg.special_case_translation [list_name]); -- add a maint cat for this template
 
break;
 
end
 
end
 
 
end
 
end
return name; -- and done
 
end
 
  
 +
local Year = A['Year'];
 +
local PublicationDate = A['PublicationDate'];
 +
local OrigYear = A['OrigYear'];
 +
local Date = A['Date'];
 +
local LayDate = A['LayDate'];
 +
------------------------------------------------- Get title data
 +
local Title = A['Title'];
 +
local ScriptTitle = A['ScriptTitle'];
 +
local BookTitle = A['BookTitle'];
 +
local Conference = A['Conference'];
 +
local TransTitle = A['TransTitle'];
 +
local TitleNote = A['TitleNote'];
 +
local TitleLink = A['TitleLink'];
 +
local Chapter = A['Chapter'];
 +
local ChapterLink = A['ChapterLink']; -- deprecated
 +
local TransChapter = A['TransChapter'];
 +
local TitleType = A['TitleType'];
 +
local Degree = A['Degree'];
 +
local Docket = A['Docket'];
 +
local ArchiveURL = A['ArchiveURL'];
 +
local URL = A['URL']
 +
local URLorigin = A:ORIGIN('URL'); -- get name of parameter that holds URL
 +
local ChapterURL = A['ChapterURL'];
 +
local ChapterURLorigin = A:ORIGIN('ChapterURL'); -- get name of parameter that holds ChapterURL
 +
local ConferenceURL = A['ConferenceURL'];
 +
local ConferenceURLorigin = A:ORIGIN('ConferenceURL'); -- get name of parameter that holds ConferenceURL
 +
local Periodical = A['Periodical'];
 +
 +
local Series = A['Series'];
 +
local Volume = A['Volume'];
 +
local Issue = A['Issue'];
 +
local Position = '';
 +
local Page = A['Page'];
 +
local Pages = hyphentodash( A['Pages'] );
 +
local At = A['At'];
  
--[[--------------------------< N A M E _ H A S _ M U L T _ N A M E S >----------------------------------------
+
local Edition = A['Edition'];
 
+
local PublicationPlace = A['PublicationPlace']
Evaluates the content of author and editor (surnames only) parameters for multiple names.  Multiple names are
+
local Place = A['Place'];
indicated if there is more than one comma and or semicolon.  If found, the function adds the multiple name
+
(author or editor) maintenance category.
+
local PublisherName = A['PublisherName'];
 
+
local RegistrationRequired = A['RegistrationRequired'];
]]
+
local SubscriptionRequired = A['SubscriptionRequired'];
 
+
local Via = A['Via'];
local function name_has_mult_names (name, list_name)
+
local AccessDate = A['AccessDate'];
local count, _;
+
local ArchiveDate = A['ArchiveDate'];
if is_set (name) then
+
local Agency = A['Agency'];
_, count = name:gsub ('[;,]', ''); -- count the number of separator-like characters
+
local DeadURL = A['DeadURL']
+
local Language = A['Language'];
if 1 < count then -- param could be |author= or |editor= so one separator character is acceptable
+
local Format = A['Format'];
add_maint_cat ('mult_names', cfg.special_case_translation [list_name]); -- more than one separator indicates multiple names so add a maint cat for this template
+
local ChapterFormat = A['ChapterFormat'];
end
+
local DoiBroken = A['DoiBroken'];
end
+
local ID = A['ID'];
return name; -- and done
+
local ASINTLD = A['ASINTLD'];
end
+
local IgnoreISBN = A['IgnoreISBN'];
 +
local Embargo = A['Embargo'];
  
 +
local ID_list = extractids( args );
  
--[[--------------------------< N A M E _ C H E C K S >--------------------------------------------------------
+
local Quote = A['Quote'];
This function calls various name checking functions used to validate the content of the various name-holding
 
parameters.
 
  
]]
+
local LayURL = A['LayURL'];
 +
local LaySource = A['LaySource'];
 +
local Transcript = A['Transcript'];
 +
local TranscriptURL = A['TranscriptURL']
 +
local TranscriptURLorigin = A:ORIGIN('TranscriptURL'); -- get name of parameter that holds TranscriptURL
  
local function name_checks (last, first, list_name)
+
local LastAuthorAmp = A['LastAuthorAmp'];
if is_set (last) then
+
local no_tracking_cats = A['NoTracking'];
if last:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
+
 
last = last:match ('^%(%((.*)%)%)$'); -- strip parens
+
--these are used by cite interview
else
+
local Callsign = A['Callsign'];
last = name_has_mult_names (last, list_name); -- check for multiple names in the parameter (last only)
+
local City = A['City'];
last = name_has_ed_markup (last, list_name); -- check for extraneous 'editor' annotation
+
local Program = A['Program'];
 +
 
 +
--local variables that are not cs1 parameters
 +
local use_lowercase; -- controls capitalization of certain static text
 +
local this_page = mw.title.getCurrentTitle(); -- also used for COinS and for language
 +
local anchor_year; -- used in the CITEREF identifier
 +
local COinS_date; -- used in the COinS metadata
 +
 
 +
-- set default parameter values defined by |mode= parameter.  If |mode= is empty or omitted, use CitationClass to set these values
 +
local sepc; -- separator between citation elements for CS1 a period, for CS2, a comma
 +
local PostScript;
 +
local Ref;
 +
sepc, PostScript, Ref = set_style (A['Mode']:lower(), A['PostScript'], A['Ref'], config.CitationClass);
 +
use_lowercase = ( sepc == ',' ); -- used to control capitalization for certain static text
 +
 
 +
--check this page to see if it is in one of the namespaces that cs1 is not supposed to add to the error categories.
 +
if not is_set(no_tracking_cats) then -- ignore if we are already not going to categorize this page
 +
if inArray (this_page.nsText, cfg.uncategorized_namespaces) then
 +
no_tracking_cats = "true"; -- set no_tracking_cats
 
end
 
end
 
end
 
end
if is_set (first) then
+
 
if first:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
+
-- check for extra |page=, |pages= or |at= parameters.
first = first:match ('^%(%((.*)%)%)$'); -- strip parens
+
if is_set(Page) then
else
+
if is_set(Pages) or is_set(At) then
first = name_has_ed_markup (first, list_name); -- check for extraneous 'editor' annotation
+
Page = Page .. " " .. seterror('extra_pages'); -- add error message
 +
Pages = ''; -- unset the others
 +
At = '';
 +
end
 +
elseif is_set(Pages) then
 +
if is_set(At) then
 +
Pages = Pages .. " " .. seterror('extra_pages'); -- add error messages
 +
At = ''; -- unset
 
end
 
end
 +
end
 +
 +
-- both |publication-place= and |place= (|location=) allowed if different
 +
if not is_set(PublicationPlace) and is_set(Place) then
 +
PublicationPlace = Place; -- promote |place= (|location=) to |publication-place
 
end
 
end
return last, first; -- done
+
end
+
if PublicationPlace == Place then Place = ''; end -- don't need both if they are the same
 +
 +
--[[
 +
Parameter remapping for cite encyclopedia:
 +
When the citation has these parameters:
 +
|encyclopedia and |title then map |title to |article and |encyclopedia to |title
 +
|encyclopedia and |article then map |encyclopedia to |title
 +
|encyclopedia then map |encyclopedia to |title
  
 +
|trans_title maps to |trans_chapter when |title is re-mapped
 +
|url maps to |chapterurl when |title is remapped
  
--[[--------------------------< E X T R A C T _ N A M E S >----------------------------------------------------
+
All other combinations of |encyclopedia, |title, and |article are not modified
Gets name list from the input arguments
+
TODO: script-title to script-chapter if and when we support script-chapter
 +
]]
  
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
+
local Encyclopedia = A['Encyclopedia'];
Stops searching when both |lastn= and |firstn= are not found in args after two sequential attempts: found |last1=, |last2=, and |last3= but doesn't
 
find |last4= and |last5= then the search is done.
 
  
This function emits an error message when there is a |firstn= without a matching |lastn=.  When there are 'holes' in the list of last names, |last1= and |last3=
+
if ( config.CitationClass == "encyclopaedia" ) or ( config.CitationClass == "citation" and is_set (Encyclopedia)) then -- test code for citation
are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=.
+
if is_set(Periodical) then -- Periodical is set when |encyclopedia is set
 +
if is_set(Title) then
 +
if not is_set(Chapter) then
 +
Chapter = Title; -- |encyclopedia and |title are set so map |title to |article and |encyclopedia to |title
 +
TransChapter = TransTitle;
 +
ChapterURL = URL;
 +
Title = Periodical;
 +
ChapterFormat = Format;
 +
Periodical = ''; -- redundant so unset
 +
TransTitle = ''; -- redundant so unset
 +
URL = ''; -- redundant so unset
 +
Format = ''; -- redundant so unset
 +
end
 +
else -- |title not set
 +
Title = Periodical; -- |encyclopedia set and |article set or not set so map |encyclopedia to |title
 +
Periodical = ''; -- redundant so unset
 +
end
 +
end
 +
end
  
When an author or editor parameter contains some form of 'et al.', the 'et al.' is stripped from the parameter and a flag (etal) returned
+
-- Special case for cite techreport.
that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration.  This keeps 'et al.' out of the
+
if (config.CitationClass == "techreport") then -- special case for cite techreport
template's metadata.  When this occurs, the page is added to a maintenance category.
+
if is_set(Issue) then -- cite techreport uses 'number', which other citations aliase to 'issue'
 +
if not is_set(ID) then -- can we use ID for the "number"?
 +
ID = Issue; -- yes, use it
 +
Issue = ""; -- unset Issue so that "number" isn't duplicated in the rendered citation or COinS metadata
 +
else -- can't use ID so emit error message
 +
ID = ID .. " " .. seterror('redundant_parameters', '<code>&#124;id=</code> and <code>&#124;number=</code>');
 +
end
 +
end
 +
end
  
]]
+
-- special case for cite interview
 +
if (config.CitationClass == "interview") then
 +
if is_set(Program) then
 +
ID = ' ' .. Program;
 +
end
 +
if is_set(Callsign) then
 +
if is_set(ID) then
 +
ID = ID .. sepc .. ' ' .. Callsign;
 +
else
 +
ID = ' ' .. Callsign;
 +
end
 +
end
 +
if is_set(City) then
 +
if is_set(ID) then
 +
ID = ID .. sepc .. ' ' .. City;
 +
else
 +
ID = ' ' .. City;
 +
end
 +
end
  
local function extract_names(args, list_name)
+
if is_set(Others) then
local names = {}; -- table of names
+
if is_set(TitleType) then
local last; -- individual name components
+
Others = ' ' .. TitleType .. ' with ' .. Others;
local first;
+
TitleType = '';
local link;
+
else
local mask;
+
Others = ' ' .. 'Interview with ' .. Others;
local i = 1; -- loop counter/indexer
+
end
local n = 1; -- output table indexer
+
else
local count = 0; -- used to count the number of times we haven't found a |last= (or alias for authors, |editor-last or alias for editors)
+
Others = '(Interview)';
local etal=false; -- return value set to true when we find some form of et al. in an author parameter
+
end
 +
end
  
local last_alias, first_alias; -- selected parameter aliases used in error messaging
+
-- special case for cite mailing list
while true do
+
if (config.CitationClass == "mailinglist") then
last, last_alias = select_one( args, cfg.aliases[list_name .. '-Last'], 'redundant_parameters', i ); -- search through args for name components beginning at 1
+
Periodical = A ['MailingList'];
first, first_alias = select_one( args, cfg.aliases[list_name .. '-First'], 'redundant_parameters', i );
+
end
link = select_one( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i );
 
mask = select_one( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
 
  
last, etal = name_has_etal (last, etal, false, last_alias); -- find and remove variations on et al.
+
--Account for the oddity that is {{cite journal}} with |pmc= set and |url= not set
first, etal = name_has_etal (first, etal, false, first_alias); -- find and remove variations on et al.
+
if config.CitationClass == "journal" and not is_set(URL) and is_set(ID_list['PMC']) then
last, first= name_checks (last, first, list_name); -- multiple names, extraneous annotation, etc checks
+
if not is_embargoed(Embargo) then
+
URL=cfg.id_handlers['PMC'].prefix .. ID_list['PMC']; -- set url to be the same as the PMC external link if not embargoed
if first and not last then -- if there is a firstn without a matching lastn
+
URLorigin = cfg.id_handlers['PMC'].parameters[1]; -- set URLorigin to parameter name for use in error message if citation is missing a |title=
table.insert( z.message_tail, { set_error( 'first_missing_last', {first_alias, first_alias:gsub('first', 'last')}, true ) } ); -- add this error message
 
elseif not first and not last then -- if both firstn and lastn aren't found, are we done?
 
count = count + 1; -- number of times we haven't found last and first
 
if 2 <= count then -- two missing names and we give up
 
break; -- normal exit or there is a two-name hole in the list; can't tell which
 
end
 
else -- we have last with or without a first
 
link_title_ok (link, list_name:match ("(%w+)List"):lower() .. '-link' .. i, last, list_name:match ("(%w+)List"):lower() .. '-last' .. i); -- check for improper wikimarkup
 
 
 
names[n] = {last = last, first = first, link = link, mask = mask, corporate=false}; -- add this name to our names list (corporate for |vauthors= only)
 
n = n + 1; -- point to next location in the names table
 
if 1 == count then -- if the previous name was missing
 
table.insert( z.message_tail, { set_error( 'missing_name', {list_name:match ("(%w+)List"):lower(), i-1}, true ) } ); -- add this error message
 
end
 
count = 0; -- reset the counter, we're looking for two consecutive missing names
 
 
end
 
end
i = i + 1; -- point to next args location
 
 
end
 
end
 
return names, etal; -- all done, return our list of names
 
end
 
  
 
+
-- Account for the oddity that is {{cite conference}}, before generation of COinS data.
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
+
if 'conference' == config.CitationClass then
 
+
if is_set(BookTitle) then
Validates language names provided in |language= parameter if not an ISO639-1 or 639-2 code.
+
Chapter = Title;
 
+
-- ChapterLink = TitleLink; -- |chapterlink= is deprecated
Returns the language name and associated two- or three-character code.  Because case of the source may be incorrect
+
ChapterURL = URL;
or different from the case that WikiMedia uses, the name comparisons are done in lower case and when a match is
+
ChapterURLorigin = URLorigin;
found, the Wikimedia version (assumed to be correct) is returned along with the code. When there is no match, we
+
URLorigin = '';
return the original language name string.
+
ChapterFormat = Format;
 
+
TransChapter = TransTitle;
mw.language.fetchLanguageNames(<local wiki language>, 'all') returns a list of languages that in some cases may include
+
Title = BookTitle;
extensions. For example, code 'cbk-zam' and its associated name 'Chavacano de Zamboanga' (MediaWiki does not support
+
Format = '';
code 'cbk' or name 'Chavacano'.  Most (all?) of these languages are not used a 'language' codes per se, rather they
+
-- TitleLink = '';
are used as sub-domain names: cbk-zam.wikipedia.org.  A list of language names and codes supported by fetchLanguageNames()
+
TransTitle = '';
can be found at Template:Citation Style documentation/language/doc
+
URL = '';
 
+
end
Names that are included in the list will be found if that name is provided in the |language= parameter.  For example,
+
else
if |language=Chavacano de Zamboanga, that name will be found with the associated code 'cbk-zam'.  When names are found
+
Conference = ''; -- not cite conference so make sure this is empty string
and the associated code is not two or three characters, this function returns only the WikiMedia language name.
 
 
 
Some language names have multiple entries under different codes:
 
Aromanian has code rup and code roa-rup
 
When this occurs, this function returns the language name and the 2- or 3-character code
 
 
 
Adapted from code taken from Module:Check ISO 639-1.
 
 
 
]]
 
 
 
local function get_iso639_code (lang, this_wiki_code)
 
if cfg.lang_name_remap[lang:lower()] then -- if there is a remapped name (because MediaWiki uses something that we don't think is correct)
 
return cfg.lang_name_remap[lang:lower()][1], cfg.lang_name_remap[lang:lower()][2]; -- for this language 'name', return a possibly new name and appropriate code
 
 
end
 
end
  
local ietf_code; -- because some languages have both ietf-like codes and iso 639-like codes
+
-- Account for the oddity that is {{cite episode}}, before generation of COinS data.
local ietf_name;
+
--[[ -- {{cite episode}} is not currently supported by this module
+
if config.CitationClass == "episode" then
local languages = mw.language.fetchLanguageNames(this_wiki_code, 'all') -- get a list of language names known to Wikimedia
+
local AirDate = A['AirDate'];
-- ('all' is required for North Ndebele, South Ndebele, and Ojibwa)
+
local SeriesLink = A['SeriesLink'];
local langlc = mw.ustring.lower(lang); -- lower case version for comparisons
+
local Season = A['Season'];
 +
local SeriesNumber = A['SeriesNumber'];
 +
local Network = A['Network'];
 +
local Station = A['Station'];
 +
local s, n = {}, {};
 +
local Sep = (first_set(A["SeriesSeparator"], A["Separator"]) or "") .. " ";
 +
 +
if is_set(Issue) then table.insert(s, cfg.messages["episode"] .. " " .. Issue); Issue = ''; end
 +
if is_set(Season) then table.insert(s, cfg.messages["season"] .. " " .. Season); end
 +
if is_set(SeriesNumber) then table.insert(s, cfg.messages["series"] .. " " .. SeriesNumber); end
 +
if is_set(Network) then table.insert(n, Network); end
 +
if is_set(Station) then table.insert(n, Station); end
 +
 +
Date = Date or AirDate;
 +
Chapter = Title;
 +
ChapterLink = TitleLink;
 +
TransChapter = TransTitle;
 +
Title = Series;
 +
TitleLink = SeriesLink;
 +
TransTitle = '';
 +
 +
Series = table.concat(s, Sep);
 +
ID = table.concat(n, Sep);
 +
end
 +
-- end of {{cite episode}} stuff]]
  
for code, name in pairs(languages) do -- scan the list to see if we can find our language
+
-- legacy: promote concatenation of |month=, and |year= to Date if Date not set; or, promote PublicationDate to Date if neither Date nor Year are set.
if langlc == mw.ustring.lower(name) then
+
if is_set(Date) then
if 2 == code:len() or 3 == code:len() then -- two- or three-character codes only; extensions not supported
+
if is_set (Year) then
return name, code; -- so return the name and the code
+
table.insert( z.maintenance_cats, "CS1 maint: Date and year"); -- add to maint category
 +
end
 +
else
 +
Date = Year; -- promote Year to Date
 +
Year = nil; -- make nil so Year as empty string isn't used for CITEREF
 +
if is_set(Date) then
 +
local Month = A['Month'];
 +
if is_set(Month) then
 +
Date = Month .. " " .. Date;
 
end
 
end
ietf_code = code; -- remember that we found an ietf-like code and save its name
+
elseif is_set(PublicationDate) then -- use PublicationDate when |date= and |year= are not set
ietf_name = name; -- but keep looking for a 2- or 3-char code
+
Date = PublicationDate; -- promote PublicationDate to Date
 +
PublicationDate = ''; -- unset, no longer needed
 
end
 
end
 
end
 
end
-- didn't find name with 2- or 3-char code; if ietf-like code found return
 
return ietf_code and ietf_name or lang; -- associated name; return original language text else
 
end
 
 
 
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
 
  
Gets language name from a provided two- or three-character ISO 639 code.  If a code is recognized by MediaWiki,
+
if PublicationDate == Date then PublicationDate = ''; end -- if PublicationDate is same as Date, don't display in rendered citation
use the returned name; if not, then use the value that was provided with the language parameter.
 
  
When |language= contains a recognized language (either code or name), the page is assigned to the category for
 
that code: Category:Norwegian-language sources (no).  For valid three-character code languages, the page is assigned
 
to the single category for '639-2' codes: Category:CS1 ISO 639-2 language sources.
 
  
Languages that are the same as the local wiki are not categorized.  MediaWiki does not recognize three-character
+
--[[
equivalents of two-character codes: code 'ar' is recognized bit code 'ara' is not.
+
Go test all of the date-holding parameters for valid MOS:DATE format and make sure that dates are real dates. This must be done before we do COinS because here is where
 
+
we get the date used in the metadata.
This function supports multiple languages in the form |language=nb, French, th where the language names or codes are
 
separated from each other by commas.
 
  
 +
Date validation supporting code is in Module:Citation/CS1/Date_validation
 
]]
 
]]
 +
anchor_year, COinS_date, error_message = dates({['accessdate']=AccessDate, ['airdate']=AirDate, ['archivedate']=ArchiveDate, ['date']=Date, ['doi_brokendate']=DoiBroken,
 +
['embargo']=Embargo, ['laydate']=LayDate, ['publicationdate']=PublicationDate, ['year']=Year});
 +
if is_set(error_message) then
 +
table.insert( z.message_tail, { seterror( 'bad_date', {error_message}, true ) } ); -- add this error message
 +
end
  
local function language_parameter (lang)
+
-- At this point fields may be nil if they weren't specified in the template use.  We can use that fact.
local code; -- the two- or three-character language code
+
-- Test if citation has no title
local name; -- the language name
+
if not is_set(Title) and
local language_list = {}; -- table of language names to be rendered
+
-- not is_set(Periodical) and -- not a title
local names_table = {}; -- table made from the value assigned to |language=
+
-- not is_set(Conference) and -- not a title
 
+
not is_set(TransTitle) and
local this_wiki = mw.getContentLanguage(); -- get a language object for this wiki
+
not is_set(ScriptTitle) then
local this_wiki_code = this_wiki:getCode() -- get this wiki's language code
+
table.insert( z.message_tail, { seterror( 'citation_missing_title', {}, true ) } );
local this_wiki_name = mw.language.fetchLanguageName(this_wiki_code, this_wiki_code); -- get this wiki's language name
 
 
 
names_table = mw.text.split (lang, '%s*,%s*'); -- names should be a comma separated list
 
 
 
for _, lang in ipairs (names_table) do -- reuse lang
 
name = cfg.lang_code_remap[lang:lower()]; -- first see if this is a code that is not supported by MediaWiki but is in remap
 
 
 
if name then -- there was a remapped code so
 
lang = lang:gsub ('^(%a%a%a?)%-.*', '%1'); -- strip ietf tags from code
 
else
 
if lang:match ('^%a%a%-') then -- strip ietf tags from code; TODO: is there a need to support 3-char with tag?
 
lang = lang:match ('(%a%a)%-') -- keep only 639-1 code portion to lang; TODO: do something with 3166 alpha 2 country code?
 
end
 
if 2 == lang:len() or 3 == lang:len() then -- if two-or three-character code
 
name = mw.language.fetchLanguageName (lang:lower(), this_wiki_code); -- get language name if |language= is a proper code
 
end
 
end
 
 
 
if is_set (name) then -- if |language= specified a valid code
 
code = lang:lower(); -- save it
 
else
 
name, code = get_iso639_code (lang, this_wiki_code); -- attempt to get code from name (assign name here so that we are sure of proper capitalization)
 
end
 
 
if is_set (code) then -- only 2- or 3-character codes
 
name = cfg.lang_code_remap[code] or name; -- override wikimedia when they misuse language codes/names
 
 
 
if this_wiki_code ~= code then -- when the language is not the same as this wiki's language
 
if 2 == code:len() then -- and is a two-character code
 
add_prop_cat ('foreign_lang_source' .. code, {name, code}) -- categorize it
 
else -- or is a recognized language (but has a three-character code)
 
add_prop_cat ('foreign_lang_source_2' .. code, {code}) -- categorize it differently TODO: support mutliple three-character code categories per cs1|2 template
 
end
 
end
 
else
 
add_maint_cat ('unknown_lang'); -- add maint category if not already added
 
end
 
 
table.insert (language_list, name);
 
name = ''; -- so we can reuse it
 
 
end
 
end
 
 
code = #language_list -- reuse code as number of languages in the list
+
if 'none' == Title and is_set(Periodical) and not (( config.CitationClass == "encyclopaedia" ) or ( config.CitationClass == "citation" and is_set (Encyclopedia))) then -- special case
if 2 >= code then
+
Title = ''; -- set title to empty string
name = table.concat (language_list, cfg.messages['parameter-pair-separator']) -- insert '<space>and<space>' between two language names
+
table.insert( z.maintenance_cats, "CS1 maint: Untitled periodical"); -- add to maint category
elseif 2 < code then
 
name = table.concat (language_list, ', '); -- and concatenate with '<comma><space>' separators
 
name = name:gsub (', ([^,]+)$', cfg.messages['parameter-final-separator'] .. '%1'); -- replace last '<comma><space>' separator with '<comma><space>and<space>' separator
 
end
 
if this_wiki_name == name then
 
return ''; -- if one language and that language is this wiki's return an empty string (no annotation)
 
 
end
 
end
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
 
--[[ TODO: should only return blank or name rather than full list
 
so we can clean up the bunched parenthetical elements Language, Type, Format
 
]]
 
end
 
 
 
--[[--------------------------< S E T _ C S 1 _ S T Y L E >----------------------------------------------------
 
 
Set style settings for CS1 citation templates. Returns separator and postscript settings
 
At en.wiki, for cs1:
 
ps gets: '.'
 
sep gets: '.'
 
  
 +
-- COinS metadata (see <http://ocoins.info/>) for automated parsing of citation information.
 +
-- handle the oddity that is cite encyclopedia and {{citation |encyclopedia=something}}. Here we presume that
 +
-- when Periodical, Title, and Chapter are all set, then Periodical is the book (encyclopedia) title, Title
 +
-- is the article title, and Chapter is a section within the article.  So, we remap
 +
 +
local coins_chapter = Chapter; -- default assuming that remapping not required
 +
local coins_title = Title; -- et tu
 +
if 'encyclopaedia' == config.CitationClass or ('citation' == config.CitationClass and is_set (Encyclopedia)) then
 +
if is_set (Chapter) and is_set (Title) and is_set (Periodical) then -- if all are used then
 +
coins_chapter = Title; -- remap
 +
coins_title = Periodical;
 +
end
 +
end
 +
 +
-- this is the function call to COinS()
 +
local OCinSoutput = COinS{
 +
['Periodical'] = Periodical,
 +
['Chapter'] = strip_apostrophe_markup (coins_chapter), -- Chapter stripped of bold / italic wikimarkup
 +
['Title'] = make_coins_title (coins_title, ScriptTitle), -- Title and ScriptTitle stripped of bold / italic wikimarkup
 +
['PublicationPlace'] = PublicationPlace,
 +
['Date'] = first_set(COinS_date, Date), -- COinS_date has correctly formatted date if Date is valid; any reason to keep Date here?  Should we be including invalid dates in metadata?
 +
['Series'] = Series,
 +
['Volume'] = Volume,
 +
['Issue'] = Issue,
 +
['Pages'] = get_coins_pages (first_set(Page, Pages, At)), -- pages stripped of external links
 +
['Edition'] = Edition,
 +
['PublisherName'] = PublisherName,
 +
['URL'] = first_set( URL, ChapterURL ),
 +
['Authors'] = a,
 +
['ID_list'] = ID_list,
 +
['RawPage'] = this_page.prefixedText,
 +
};
 +
--[[Why is this here?  Why are we mapping Title to Chapter when Periodical is set?
 +
if is_set(Periodical) and not is_set(Chapter) and is_set(Title) then
 +
Chapter = Title;
 +
ChapterLink = TitleLink;
 +
TransChapter = TransTitle;
 +
Title = '';
 +
TitleLink = '';
 +
TransTitle = '';
 +
end
 
]]
 
]]
  
local function set_cs1_style (ps)
+
-- special case for cite newsgroup.  Do this after COinS because we are modifying Publishername to include some static text
if not is_set (ps) then -- unless explicitely set to something
+
if 'newsgroup' == config.CitationClass then
ps = cfg.presentation['ps_cs1']; -- terminate the rendered citation
+
if is_set (PublisherName) then
 +
PublisherName = '[[Newsgroup]]:&nbsp;' .. externallink( 'news:' .. PublisherName, PublisherName );
 +
end
 
end
 
end
return cfg.presentation['sep_cs1'], ps; -- element separator
 
end
 
  
  
--[[--------------------------< S E T _ C S 2 _ S T Y L E >----------------------------------------------------
 
  
Set style settings for CS2 citation templates. Returns separator, postscript, ref settings
+
-- Now perform various field substitutions.
At en.wiki, for cs2:
+
-- We also add leading spaces and surrounding markup and punctuation to the
ps gets: '' (empty string - no terminal punctuation)
+
-- various parts of the citation, but only when they are non-nil.
sep gets: ','
+
if not is_set(Authors) then
 +
local Maximum = tonumber( A['DisplayAuthors'] );
  
]]
+
if is_set (Maximum) then
 +
if Maximum >= #a then -- if display-authors value greater than or equal to number of authors
 +
table.insert( z.maintenance_cats, "CS1 maint: display-authors"); -- add maintenance category because display-authors parameter may be removed
 +
end
 +
else
 +
Maximum = #a + 1; -- number of authors + 1
 +
end
  
local function set_cs2_style (ps, ref)
+
local control = {
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
+
format = NameListFormat, -- empty string or 'vanc'
ps = cfg.presentation['ps_cs2']; -- terminate the rendered citation
+
maximum = Maximum,
 +
lastauthoramp = LastAuthorAmp,
 +
page_name = this_page.text -- get current page name so that we don't wikilink to it via authorlinkn
 +
};
 +
 +
-- If the coauthor field is also used, prevent ampersand and et al. formatting.
 +
if is_set(Coauthors) then
 +
control.lastauthoramp = nil;
 +
control.maximum = #a + 1;
 +
end
 +
 +
Authors = listpeople(control, a)
 
end
 
end
if not is_set (ref) then -- if |ref= is not set
+
 
ref = "harv"; -- set default |ref=harv
+
if not is_set(Authors) and is_set(Coauthors) then -- coauthors aren't displayed if one of authors=, authorn=, or lastn= isn't specified
 +
table.insert( z.message_tail, { seterror('coauthors_missing_author', {}, true) } ); -- emit error message
 
end
 
end
return cfg.presentation['sep_cs2'], ps, ref; -- element separator
 
end
 
  
 +
local EditorCount
 +
if not is_set(Editors) then
 +
local Maximum = tonumber( A['DisplayEditors'] );
 +
-- Preserve old-style implicit et al.
 +
if not is_set(Maximum) and #e == 4 then
 +
Maximum = 3;
 +
table.insert( z.message_tail, { seterror('implict_etal_editor', {}, true) } );
 +
elseif not is_set(Maximum) then
 +
Maximum = #e + 1;
 +
end
  
--[[--------------------------< G E T _ S E T T I N G S _ F R O M _ C I T E _ C L A S S >----------------------
+
local control = {
 +
format = NameListFormat, -- empty string or 'vanc'
 +
maximum = Maximum,
 +
lastauthoramp = LastAuthorAmp,
 +
page_name = this_page.text -- get current page name so that we don't wikilink to it via editorlinkn
 +
};
  
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
+
Editors, EditorCount = listpeople(control, e);
rendered style.
+
else
 +
EditorCount = 1;
 +
end
  
]]
+
local Cartography = "";
 
+
local Scale = "";
local function get_settings_from_cite_class (ps, ref, cite_class)
+
if config.CitationClass == "map" then
local sep;
+
if not is_set( Authors ) and is_set( PublisherName ) then
if (cite_class == "citation") then -- for citation templates (CS2)
+
Authors = PublisherName;
sep, ps, ref = set_cs2_style (ps, ref);
+
PublisherName = "";
else -- not a citation template so CS1
+
end
sep, ps = set_cs1_style (ps);
+
Cartography = A['Cartography'];
 +
if is_set( Cartography ) then
 +
Cartography = sepc .. " " .. wrap_msg ('cartography', Cartography, use_lowercase);
 +
end
 +
Scale = A['Scale'];
 +
if is_set( Scale ) then
 +
Scale = sepc .. " " .. Scale;
 +
end
 
end
 
end
 +
 +
Format = is_set(Format) and " (" .. Format .. ")" or "";
  
return sep, ps, ref -- return them all
+
if  not is_set(URL) and
end
+
not is_set(ArchiveURL) and
 
+
not is_set(ConferenceURL) and -- TODO: keep this here? conference as part of cite web or cite podcast?
 
+
not is_set(TranscriptURL) then
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
+
 
+
-- Test if cite web or cite podcast |url= is missing or empty
Establish basic style settings to be used when rendering the citation.  Uses |mode= if set and valid or uses
+
if inArray(config.CitationClass, {"web","podcast", "mailinglist"}) then
config.CitationClass from the template's #invoke: to establish style.
+
table.insert( z.message_tail, { seterror( 'cite_web_url', {}, true ) } );
 
+
end
]]
+
 
+
-- Test if accessdate is given without giving a URL
local function set_style (mode, ps, ref, cite_class)
+
if is_set(AccessDate) and not is_set(ChapterURL)then -- ChapterURL may be set when the others are not set; TODO: move this to a separate test?
local sep;
+
table.insert( z.message_tail, { seterror( 'accessdate_missing_url', {}, true ) } );
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
+
AccessDate = '';
sep, ps, ref = set_cs2_style (ps, ref);
+
end
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
+
sep, ps = set_cs1_style (ps);
+
-- Test if format is given without giving a URL
else -- anything but cs1 or cs2
+
if is_set(Format) then
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
+
Format = Format .. seterror( 'format_missing_url', {'format', 'url'} );
 +
end
 
end
 
end
if 'none' == ps:lower() then -- if assigned value is 'none' then
+
--[[
ps = ''; -- set to empty string
+
-- Test if citation has no title
 +
if not is_set(Title) and
 +
-- not is_set(Periodical) and -- not a title
 +
-- not is_set(Conference) and -- not a title
 +
not is_set(TransTitle) and
 +
not is_set(ScriptTitle) then
 +
table.insert( z.message_tail, { seterror( 'citation_missing_title', {}, true ) } );
 
end
 
end
 
 
return sep, ps, ref
+
if 'none' == Title and is_set(Periodical) then -- special case
end
+
Title = ''; -- set title to empty string
 
+
table.insert( z.maintenance_cats, "CS1 maint: Untitled periodical"); -- add to maint category
 
+
end
--[=[-------------------------< I S _ P D F >------------------------------------------------------------------
+
]]
 
+
local OriginalURL;
Determines if a url has the file extension that is one of the pdf file extensions used by [[MediaWiki:Common.css]] when
+
DeadURL = DeadURL:lower(); -- used later when assembling archived text
applying the pdf icon to external links.
+
if is_set( ArchiveURL ) then
 
+
if is_set (URL) then
returns true if file extension is one of the recognized extensions, else false
+
OriginalURL = URL; -- save copy of original source URL
 +
if 'no' ~= DeadURL then -- if URL set then archive-url applies to it
 +
URL = ArchiveURL -- swap-in the archive's url
 +
URLorigin = A:ORIGIN('ArchiveURL') -- name of archive url parameter for error messages
 +
end
 +
elseif is_set (ChapterURL) then -- URL not set so if chapter-url is set apply archive url to it
 +
OriginalURL = ChapterURL; -- save copy of source chapter's url for archive text
 +
if 'no' ~= DeadURL then
 +
ChapterURL = ArchiveURL -- swap-in the archive's url
 +
URLorigin = A:ORIGIN('ArchiveURL') -- name of archive url parameter for error messages
 +
end
 +
end
 +
end
  
]=]
+
if inArray(config.CitationClass, {"web","news","journal","pressrelease","podcast", "newsgroup"}) or
 
+
('citation' == config.CitationClass and is_set (Periodical) and not is_set (Encyclopedia)) then
local function is_pdf (url)
+
if is_set (Chapter) or is_set (TransChapter) or is_set (ChapterURL)then -- chapter parameters not supported for these citation types
return url:match ('%.pdf$') or url:match ('%.PDF$') or
+
table.insert( z.message_tail, { seterror( 'chapter_ignored', {}, true ) } ); -- add error message
url:match ('%.pdf[%?#]') or url:match ('%.PDF[%?#]') or
+
Chapter = ''; -- set to empty string to be safe with concatenation
url:match ('%.PDF&#035') or url:match ('%.pdf&#035');
+
TransChapter = '';
end
+
ChapterURL = '';
 
+
end
 
+
else -- otherwise, format chapter / article title
--[[--------------------------< S T Y L E _ F O R M A T >------------------------------------------------------
+
Chapter = format_chapter_title (Chapter, TransChapter, ChapterURL, ChapterURLorigin);
 
+
if is_set (Chapter) then
Applies css style to |format=, |chapter-format=, etc.  Also emits an error message if the format parameter does
+
ChapterFormat = is_set(ChapterFormat) and " (" .. ChapterFormat .. ")" or "";
not have a matching url parameter.  If the format parameter is not set and the url contains a file extension that
+
if is_set(ChapterFormat) and not is_set (ChapterURL) then -- Test if |chapter-format= is given without giving a |chapter-url=
is recognized as a pdf document by MediaWiki's commons.css, this code will set the format parameter to (PDF) with
+
ChapterFormat = ChapterFormat .. seterror( 'format_missing_url', {'chapter-format', 'chapter-url'} );
the appropriate styling.
+
end
 
+
Chapter = Chapter .. ChapterFormat .. sepc .. ' ';
]]
 
 
 
local function style_format (format, url, fmt_param, url_param)
 
if is_set (format) then
 
format = wrap_style ('format', format); -- add leading space, parentheses, resize
 
if not is_set (url) then
 
format = format .. set_error( 'format_missing_url', {fmt_param, url_param} ); -- add an error message
 
 
end
 
end
elseif is_pdf (url) then -- format is not set so if url is a pdf file then
 
format = wrap_style ('format', 'PDF'); -- set format to pdf
 
else
 
format = ''; -- empty string for concatenation
 
 
end
 
end
return format;
 
end
 
  
 +
-- Format main title.
 +
if is_set(TitleLink) and is_set(Title) then
 +
Title = "[[" .. TitleLink .. "|" .. Title .. "]]"
 +
end
  
--[[--------------------------< G E T _ D I S P L A Y _ N A M E S >--------------------------------------------
+
if inArray(config.CitationClass, {"web","news","journal","pressrelease","podcast", "newsgroup", "mailinglist"}) or
 +
('citation' == config.CitationClass and is_set (Periodical) and not is_set (Encyclopedia)) then
 +
Title = kern_quotes (Title); -- if necessary, separate title's leading and trailing quote marks from Module provided quote marks
 +
Title = wrap_style ('quoted-title', Title);
 +
 +
Title = script_concatenate (Title, ScriptTitle); -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
 +
TransTitle= wrap_style ('trans-quoted-title', TransTitle );
 +
elseif 'report' == config.CitationClass then -- no styling for cite report
 +
Title = script_concatenate (Title, ScriptTitle); -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
 +
TransTitle= wrap_style ('trans-quoted-title', TransTitle ); -- for cite report, use this form for trans-title
 +
else
 +
Title = wrap_style ('italic-title', Title);
 +
Title = script_concatenate (Title, ScriptTitle); -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
 +
TransTitle = wrap_style ('trans-italic-title', TransTitle);
 +
end
  
Returns a number that defines the number of names displayed for author and editor name lists and a boolean flag
+
TransError = "";
to indicate when et al. should be appended to the name list.
+
if is_set(TransTitle) then
 
+
if is_set(Title) then
When the value assigned to |display-xxxxors= is a number greater than or equal to zero, return the number and
+
TransTitle = " " .. TransTitle;
the previous state of the 'etal' flag (false by default but may have been set to true if the name list contains
+
else
some variant of the text 'et al.').
+
TransError = " " .. seterror( 'trans_missing_title' );
 
 
When the value assigned to |display-xxxxors= is the keyword 'etal', return a number that is one greater than the
 
number of authors in the list and set the 'etal' flag true.  This will cause the list_people() to display all of
 
the names in the name list followed by 'et al.'
 
 
 
In all other cases, returns nil and the previous state of the 'etal' flag.
 
 
 
inputs:
 
max: A['DisplayAuthors'] or A['DisplayEditors']; a number or some flavor of etal
 
count: #a or #e
 
list_name: 'authors' or 'editors'
 
etal: author_etal or editor_etal
 
 
 
]]
 
 
 
local function get_display_names (max, count, list_name, etal)
 
if is_set (max) then
 
if 'etal' == max:lower():gsub("[ '%.]", '') then -- the :gsub() portion makes 'etal' from a variety of 'et al.' spellings and stylings
 
max = count + 1; -- number of authors + 1 so display all author name plus et al.
 
etal = true; -- overrides value set by extract_names()
 
elseif max:match ('^%d+$') then -- if is a string of numbers
 
max = tonumber (max); -- make it a number
 
if max >= count then -- if |display-xxxxors= value greater than or equal to number of authors/editors
 
add_maint_cat ('disp_name', cfg.special_case_translation [list_name]);
 
end
 
else -- not a valid keyword or number
 
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
 
max = nil; -- unset; as if |display-xxxxors= had not been set
 
 
end
 
end
 
end
 
end
 
 
return max, etal;
+
Title = Title .. TransTitle;
end
+
 +
if is_set(Title) then
 +
if not is_set(TitleLink) and is_set(URL) then
 +
Title = externallink( URL, Title ) .. TransError .. Format  
 +
URL = "";
 +
Format = "";
 +
else
 +
Title = Title .. TransError;
 +
end
 +
end
  
 
+
if is_set(Place) then
--[[--------------------------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >------------------------------
+
Place = " " .. wrap_msg ('written', Place, use_lowercase) .. sepc .. " ";
 
 
Adds page to Category:CS1 maint: extra text if |page= or |pages= has what appears to be some form of p. or pp.
 
abbreviation in the first characters of the parameter content.
 
 
 
check Page and Pages for extraneous p, p., pp, and pp. at start of parameter value:
 
good pattern: '^P[^%.P%l]' matches when |page(s)= begins PX or P# but not Px where x and X are letters and # is a dgiit
 
bad pattern: '^[Pp][Pp]' matches matches when |page(s)= begins pp or pP or Pp or PP
 
 
 
]]
 
 
 
local function extra_text_in_page_check (page)
 
local good_pattern = '^P[^%.Pp]'; -- ok to begin with uppercase P: P7 (pg 7 of section P) but not p123 (page 123) TODO: add Gg for PG or Pg?
 
local bad_pattern = '^[Pp]?[Pp]%.?[ %d]';
 
 
 
if not page:match (good_pattern) and (page:match (bad_pattern) or  page:match ('^[Pp]ages?')) then
 
add_maint_cat ('extra_text');
 
 
end
 
end
end
 
 
 
--[=[-------------------------< G E T _ V _ N A M E _ T A B L E >----------------------------------------------
 
 
split apart a |vauthors= or |veditors= parameter.  This function allows for corporate names, wrapped in doubled
 
parentheses to also have commas; in the old version of the code, the doubled parnetheses were included in the
 
rendered citation and in the metadata.  Individual author names may be wikilinked
 
 
|vauthors=Jones AB, [[E. B. White|White EB]], ((Black, Brown, and Co.))
 
 
]=]
 
 
local function get_v_name_table (vparam, output_table, output_link_table)
 
local name_table = mw.text.split(vparam, "%s*,%s*"); -- names are separated by commas
 
local wl_type, label, link; -- wl_type not used here; just a place holder
 
 
 
local i = 1;
+
if is_set(Conference) then
 +
if is_set(ConferenceURL) then
 +
Conference = externallink( ConferenceURL, Conference );
 +
end
 +
Conference = sepc .. " " .. Conference
 +
elseif is_set(ConferenceURL) then
 +
Conference = sepc .. " " .. externallink( ConferenceURL, nil, ConferenceURLorigin );
 +
end
 
 
while name_table[i] do
+
if not is_set(Position) then
if name_table[i]:match ('^%(%(.*[^%)][^%)]$') then -- first segment of corporate with one or more commas; this segment has the opening doubled parens
+
local Minutes = A['Minutes'];
local name = name_table[i];
+
if is_set(Minutes) then
i=i+1; -- bump indexer to next segment
+
Position = " " .. Minutes .. " " .. cfg.messages['minutes'];
while name_table[i] do
+
else
name = name .. ', ' .. name_table[i]; -- concatenate with previous segments
+
local Time = A['Time'];
if name_table[i]:match ('^.*%)%)$') then -- if this table member has the closing doubled parens
+
if is_set(Time) then
break; -- and done reassembling so
+
local TimeCaption = A['TimeCaption']
end
+
if not is_set(TimeCaption) then
i=i+1; -- bump indexer
+
TimeCaption = cfg.messages['event'];
end
+
if sepc ~= '.' then
table.insert (output_table, name); -- and add corporate name to the output table
+
TimeCaption = TimeCaption:lower();
table.insert (output_link_table, ''); -- no wikilink
+
end
else
+
end
wl_type, label, link = is_wikilink (name_table[i]); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]]
+
Position = " " .. TimeCaption .. " " .. Time;
table.insert (output_table, label); -- add this name
+
end
if 1 == wl_type then
+
end
table.insert (output_link_table, label); -- simple wikilink [[D]]
+
else
else
+
Position = " " .. Position;
table.insert (output_link_table, link); -- no wikilink or [[L|D]]; add this link if there is one, else empty string
+
At = '';
 +
end
 +
 +
if not is_set(Page) then
 +
if is_set(Pages) then
 +
if is_set(Periodical) and
 +
not inArray(config.CitationClass, {"encyclopaedia","web","book","news","podcast"}) then
 +
Pages = ": " .. Pages;
 +
elseif tonumber(Pages) ~= nil then
 +
Pages = sepc .." " .. PPrefix .. Pages;
 +
else
 +
Pages = sepc .." " .. PPPrefix .. Pages;
 
end
 
end
 
end
 
end
i = i+1;
+
else
end
+
if is_set(Periodical) and
return output_table;
+
not inArray(config.CitationClass, {"encyclopaedia","web","book","news","podcast"}) then
end
+
Page = ": " .. Page;
 
 
 
 
--[[--------------------------< P A R S E _ V A U T H O R S _ V E D I T O R S >--------------------------------
 
 
 
This function extracts author / editor names from |vauthors= or |veditors= and finds matching |xxxxor-maskn= and
 
|xxxxor-linkn= in args.  It then returns a table of assembled names just as extract_names() does.
 
 
 
Author / editor names in |vauthors= or |veditors= must be in Vancouver system style. Corporate or institutional names
 
may sometimes be required and because such names will often fail the is_good_vanc_name() and other format compliance
 
tests, are wrapped in doubled paranethese ((corporate name)) to suppress the format tests.
 
 
 
Supports generational suffixes Jr, 2nd, 3rd, 4th–6th.
 
 
 
This function sets the vancouver error when a reqired comma is missing and when there is a space between an author's initials.
 
 
 
]]
 
 
 
local function parse_vauthors_veditors (args, vparam, list_name)
 
local names = {}; -- table of names assembled from |vauthors=, |author-maskn=, |author-linkn=
 
local v_name_table = {};
 
local v_link_table = {}; -- when name is wikilinked, targets go in this table
 
local etal = false; -- return value set to true when we find some form of et al. vauthors parameter
 
local last, first, link, mask, suffix;
 
local corporate = false;
 
 
 
vparam, etal = name_has_etal (vparam, etal, true); -- find and remove variations on et al. do not categorize (do it here because et al. might have a period)
 
v_name_table = get_v_name_table (vparam, v_name_table, v_link_table); -- names are separated by commas
 
 
 
for i, v_name in ipairs(v_name_table) do
 
if v_name:match ('^%(%(.+%)%)$') then -- corporate authors are wrapped in doubled parentheses to supress vanc formatting and error detection
 
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
 
last = v_name:match ('^%(%((.+)%)%)$') -- remove doubled parntheses
 
corporate = true; -- flag used in list_people()
 
elseif string.find(v_name, "%s") then
 
if v_name:find('[;%.]') then -- look for commonly occurring punctuation characters;
 
add_vanc_error ('punctuation');
 
end
 
local lastfirstTable = {}
 
lastfirstTable = mw.text.split(v_name, "%s")
 
first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be author intials
 
if is_suffix (first) then -- if a valid suffix
 
suffix = first -- save it as a suffix and
 
first = table.remove(lastfirstTable); -- get what should be the initials from the table
 
end -- no suffix error message here because letter combination may be result of Romanization; check for digits?
 
last = table.concat(lastfirstTable, " ") -- returns a string that is the concatenation of all other names that are not initials
 
if mw.ustring.match (last, '%a+%s+%u+%s+%a+') then
 
add_vanc_error ('missing comma'); -- matches last II last; the case when a comma is missing
 
end
 
if mw.ustring.match (v_name, ' %u %u$') then -- this test is in the wrong place TODO: move or replace with a more appropriate test
 
add_vanc_error ('name'); -- matches a space between two intiials
 
end
 
 
else
 
else
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
+
Page = sepc .." " .. PPrefix .. Page;
last = v_name; -- last name or single corporate name?  Doesn't support multiword corporate names? do we need this?
 
 
end
 
end
+
end
if is_set (first) then
+
if not mw.ustring.match (first, "^%u?%u$") then -- first shall contain one or two upper-case letters, nothing else
+
At = is_set(At) and (sepc .. " " .. At) or "";
add_vanc_error ('initials'); -- too many initials; mixed case initials (which may be ok Romanization); hyphenated initials
+
Position = is_set(Position) and (sepc .. " " .. Position) or "";
 +
if config.CitationClass == 'map' then
 +
local Section = A['Section'];
 +
local Inset = A['Inset'];
 +
if first_set( Pages, Page, At ) ~= nil or sepc ~= '.' then
 +
if is_set( Section ) then
 +
Section = ", " .. wrap_msg ('section', Section, true);
 
end
 
end
is_good_vanc_name (last, first); -- check first and last before restoring the suffix which may have a non-Latin digit
+
if is_set( Inset ) then
if is_set (suffix) then
+
Inset = ", " .. wrap_msg ('inset', Inset, true);
first = first .. ' ' .. suffix; -- if there was a suffix concatenate with the initials
 
suffix = ''; -- unset so we don't add this suffix to all subsequent names
 
 
end
 
end
 
else
 
else
if not corporate then
+
if is_set( Section ) then
is_good_vanc_name (last, '');
+
Section = sepc .. " " .. wrap_msg ('section', Section, use_lowercase);
end
+
if is_set( Inset ) then
end
+
Inset = ", " .. wrap_msg ('inset', Inset, true);
 +
end
 +
elseif is_set( Inset ) then
 +
Inset = sepc .. " " .. wrap_msg ('inset', Inset, use_lowercase);
 +
end
 +
end
 +
At = At .. Section .. Inset;
 +
end
  
link = select_one( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i ) or v_link_table[i];
+
if is_set (Language) then
mask = select_one( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
+
Language = language_parameter (Language, this_page.namespace); -- format, categories (article namespace only), name from ISO639-1, etc
names[i] = {last = last, first = first, link = link, mask = mask, corporate=corporate}; -- add this assembled name to our names list
+
else
 +
Language=""; -- language not specified so make sure this is an empty string;
 
end
 
end
return names, etal; -- all done, return our list of names
 
end
 
  
 +
Others = is_set(Others) and (sepc .. " " .. Others) or "";
  
--[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------
+
-- handle type parameter for those CS1 citations that have default values
  
Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or
+
if inArray(config.CitationClass, {"AV-media-notes", "DVD-notes", "mailinglist", "podcast", "pressrelease", "report", "techreport", "thesis"}) then
select one of |editors=, |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list.
+
TitleType = set_titletype (config.CitationClass, TitleType);
 +
if is_set(Degree) and "Thesis" == TitleType then -- special case for cite thesis
 +
TitleType = Degree .. " thesis";
 +
end
 +
end
  
Only one of these appropriate three will be used.  The hierarchy is: |authorn= (and aliases) highest and |authors= lowest and
+
if is_set(TitleType) then -- if type parameter is specified
similarly, |editorn= (and aliases) highest and |editors= lowest
+
TitleType = " (" .. TitleType .. ")"; -- display it in parentheses
 +
end
  
When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second
+
TitleNote = is_set(TitleNote) and (sepc .. " " .. TitleNote) or "";
test which mimicks the test used in extract_names() when looking for a hole in the author name list. There may be a better
+
Edition = is_set(Edition) and (" " .. wrap_msg ('edition', Edition)) or "";
way to do this, I just haven't discovered what that way is.
+
Issue = is_set(Issue) and (" (" .. Issue .. ")") or "";
 +
Series = is_set(Series) and (sepc .. " " .. Series) or "";
 +
OrigYear = is_set(OrigYear) and (" [" .. OrigYear .. "]") or "";
 +
Agency = is_set(Agency) and (sepc .. " " .. Agency) or "";
  
Emits an error message when more than one xxxxor name source is provided.
+
if is_set(Volume) then
 
+
if ( mw.ustring.len(Volume) > 4 )
In this function, vxxxxors = vauthors or veditors; xxxxors = authors or editors as appropriate.
+
  then Volume = sepc .." " .. Volume;
 
+
  else Volume = " <b>" .. hyphentodash(Volume) .. "</b>";
]]
+
end
 
 
local function select_author_editor_source (vxxxxors, xxxxors, args, list_name)
 
local lastfirst = false;
 
if select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice incase we have a |first1= without a |last1=; this ...
 
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 1 ) or -- ... also catches the case where |first= is used with |vauthors=
 
select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) or
 
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 2 ) then
 
lastfirst=true;
 
 
end
 
end
  
if (is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions
+
------------------------------------ totally unrelated data
(is_set (vxxxxors) and is_set (xxxxors)) or
+
if is_set(Via) then
(true == lastfirst and is_set (xxxxors)) then
+
Via = " " .. wrap_msg ('via', Via);
local err_name;
 
if 'AuthorList' == list_name then -- figure out which name should be used in error message
 
err_name = 'author';
 
else
 
err_name = 'editor';
 
end
 
table.insert( z.message_tail, { set_error( 'redundant_parameters',
 
{err_name .. '-name-list parameters'}, true ) } ); -- add error message
 
 
end
 
end
  
if true == lastfirst then return 1 end; -- return a number indicating which author name source to use
+
--[[
if is_set (vxxxxors) then return 2 end;
+
Subscription implies paywall; Registration does not.  If both are used in a citation, the subscription required link
if is_set (xxxxors) then return 3 end;
+
note is displayed. There are no error messages for this condition.
return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last
 
end
 
 
 
 
 
--[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------
 
 
 
This function is used to validate a parameter's assigned value for those parameters that have only a limited number
 
of allowable values (yes, y, true, no, etc).  When the parameter value has not been assigned a value (missing or empty
 
in the source template) the function returns true. If the parameter value is one of the list of allowed values returns
 
true; else, emits an error message and returns false.
 
  
 
]]
 
]]
 
+
if inArray(SubscriptionRequired:lower(), {'yes', 'true', 'y'}) then
local function is_valid_parameter_value (value, name, possible)
+
SubscriptionRequired = sepc .. " " .. cfg.messages['subscription']; -- subscription required message
if not is_set (value) then
+
elseif inArray(RegistrationRequired:lower(), {'yes', 'true', 'y'}) then
return true; -- an empty parameter is ok
+
SubscriptionRequired = sepc .. " " .. cfg.messages['registration']; -- registration required message
elseif in_array(value:lower(), possible) then
 
return true;
 
 
else
 
else
table.insert( z.message_tail, { set_error( 'invalid_param_val', {name, value}, true ) } ); -- not an allowed value so add error message
+
SubscriptionRequired = ''; -- either or both might be set to something other than yes true y
return false
 
 
end
 
end
end
 
  
 +
if is_set(AccessDate) then
 +
local retrv_text = " " .. cfg.messages['retrieved']
  
--[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >----------------------------------------
+
AccessDate = nowrap_date (AccessDate); -- wrap in nowrap span if date in appropriate format
 +
if (sepc ~= ".") then retrv_text = retrv_text:lower() end -- if 'citation', lower case
 +
AccessDate = substitute (retrv_text, AccessDate); -- add retrieved text
 +
-- neither of these work; don't know why; it seems that substitute() isn't being called
 +
AccessDate = substitute (cfg.presentation['accessdate'], {sepc, AccessDate}); -- allow editors to hide accessdates
 +
end
 +
 +
if is_set(ID) then ID = sepc .." ".. ID; end
 +
  if "thesis" == config.CitationClass and is_set(Docket) then
 +
ID = sepc .." Docket ".. Docket .. ID;
 +
end
 +
  if "report" == config.CitationClass and is_set(Docket) then -- for cite report when |docket= is set
 +
ID = sepc .. ' ' .. Docket; -- overwrite ID even if |id= is set
 +
end
  
This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space
+
ID_list = buildidlist( ID_list, {DoiBroken = DoiBroken, ASINTLD = ASINTLD, IgnoreISBN = IgnoreISBN, Embargo=Embargo} );
when the last character is not a sepc character or when the last three characters are not sepc followed by two
 
closing square brackets (close of a wikilink).  When either of these is true, the name_list is terminated with a
 
single space character.
 
  
]]
+
if is_set(URL) then
 
+
URL = " " .. externallink( URL, nil, URLorigin );
local function terminate_name_list (name_list, sepc)
 
if (string.sub (name_list,-3,-1) == sepc .. '. ') then -- if already properly terminated
 
return name_list; -- just return the name list
 
elseif (string.sub (name_list,-1,-1) == sepc) or (string.sub (name_list,-3,-1) == sepc .. ']]') then -- if last name in list ends with sepc char
 
return name_list .. " "; -- don't add another
 
else
 
return name_list .. sepc .. ' '; -- otherwise terninate the name list
 
 
end
 
end
end
 
  
 
+
if is_set(Quote) then
--[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------
+
if Quote:sub(1,1) == '"' and Quote:sub(-1,-1) == '"' then
 
+
Quote = Quote:sub(2,-2);
returns the concatenation of the formatted volume and issue parameters as a single string; or formatted volume
 
or formatted issue, or an empty string if neither are set.
 
 
 
]]
 
 
local function format_volume_issue (volume, issue, cite_class, origin, sepc, lower)
 
if not is_set (volume) and not is_set (issue) then
 
return '';
 
end
 
 
if 'magazine' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'magazine' == origin) then
 
if is_set (volume) and is_set (issue) then
 
return wrap_msg ('vol-no', {sepc, volume, issue}, lower);
 
elseif is_set (volume) then
 
return wrap_msg ('vol', {sepc, volume}, lower);
 
else
 
return wrap_msg ('issue', {sepc, issue}, lower);
 
 
end
 
end
 +
Quote = sepc .." " .. wrap_style ('quoted-text', Quote ); -- wrap in <q>...</q> tags
 +
PostScript = ""; -- CS1 does not supply terminal punctuation when |quote= is set
 
end
 
end
 
+
local vol = ''; -- here for all cites except magazine
+
local Archived
+
if is_set(ArchiveURL) then
if is_set (volume) then
+
if not is_set(ArchiveDate) then
if volume:match ('^[MDCLXVI]+$') or volume:match ('^%d+$')then -- volume value is all digits or all uppercase roman numerals
+
ArchiveDate = seterror('archive_missing_date');
vol = substitute (cfg.presentation['vol-bold'], {sepc, hyphen_to_dash(volume)}); -- render in bold face
 
elseif (4 < mw.ustring.len(volume)) then -- not all digits or roman numerals and longer than 4 characters
 
vol = substitute (cfg.messages['j-vol'], {sepc, volume}); -- not bold
 
add_prop_cat ('long_vol');
 
else -- four or less characters
 
vol = substitute (cfg.presentation['vol-bold'], {sepc, hyphen_to_dash(volume)}); -- bold
 
 
end
 
end
end
+
if "no" == DeadURL then
if is_set (issue) then
+
local arch_text = cfg.messages['archived'];
return vol .. substitute (cfg.messages['j-issue'], issue);
+
if sepc ~= "." then arch_text = arch_text:lower() end
end
+
Archived = sepc .. " " .. substitute( cfg.messages['archived-not-dead'],
return vol;
+
{ externallink( ArchiveURL, arch_text ), ArchiveDate } );
end
+
if not is_set(OriginalURL) then
 
+
Archived = Archived .. " " .. seterror('archive_missing_url');  
 
 
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
 
 
 
adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings.
 
The return order is:
 
page, pages, sheet, sheets
 
 
 
Singular has priority over plural when both are provided.
 
 
 
]]
 
 
 
local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower)
 
if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators
 
if is_set (sheet) then
 
if 'journal' == origin then
 
return '', '', wrap_msg ('j-sheet', sheet, lower), '';
 
else
 
return '', '', wrap_msg ('sheet', {sepc, sheet}, lower), '';
 
end
 
elseif is_set (sheets) then
 
if 'journal' == origin then
 
return '', '', '', wrap_msg ('j-sheets', sheets, lower);
 
else
 
return '', '', '', wrap_msg ('sheets', {sepc, sheets}, lower);
 
 
end
 
end
 +
elseif is_set(OriginalURL) then
 +
local arch_text = cfg.messages['archived-dead'];
 +
if sepc ~= "." then arch_text = arch_text:lower() end
 +
Archived = sepc .. " " .. substitute( arch_text,
 +
{ externallink( OriginalURL, cfg.messages['original'] ), ArchiveDate } );
 +
else
 +
local arch_text = cfg.messages['archived-missing'];
 +
if sepc ~= "." then arch_text = arch_text:lower() end
 +
Archived = sepc .. " " .. substitute( arch_text,
 +
{ seterror('archive_missing_url'), ArchiveDate } );
 
end
 
end
 +
else
 +
Archived = ""
 
end
 
end
 
local is_journal = 'journal' == cite_class or (in_array (cite_class, {'citation', 'map', 'interview'}) and 'journal' == origin);
 
 
 
if is_set (page) then
+
local Lay
if is_journal then
+
if is_set(LayURL) then
return substitute (cfg.messages['j-page(s)'], page), '', '', '';
+
if is_set(LayDate) then LayDate = " (" .. LayDate .. ")" end
elseif not nopp then
+
if is_set(LaySource) then  
return substitute (cfg.messages['p-prefix'], {sepc, page}), '', '', '';
+
LaySource = " &ndash; ''" .. safeforitalics(LaySource) .. "''";
 
else
 
else
return substitute (cfg.messages['nopp'], {sepc, page}), '', '', '';
+
LaySource = "";
 
end
 
end
elseif is_set(pages) then
+
if sepc == '.' then
if is_journal then
+
Lay = sepc .. " " .. externallink( LayURL, cfg.messages['lay summary'] ) .. LaySource .. LayDate
return substitute (cfg.messages['j-page(s)'], pages), '', '', '';
 
elseif tonumber(pages) ~= nil and not nopp then -- if pages is only digits, assume a single page number
 
return '', substitute (cfg.messages['p-prefix'], {sepc, pages}), '', '';
 
elseif not nopp then
 
return '', substitute (cfg.messages['pp-prefix'], {sepc, pages}), '', '';
 
 
else
 
else
return '', substitute (cfg.messages['nopp'], {sepc, pages}), '', '';
+
Lay = sepc .. " " .. externallink( LayURL, cfg.messages['lay summary']:lower() ) .. LaySource .. LayDate
end
+
end
 +
else
 +
Lay = "";
 +
end
 +
 +
if is_set(Transcript) then
 +
if is_set(TranscriptURL) then Transcript = externallink( TranscriptURL, Transcript ); end
 +
elseif is_set(TranscriptURL) then
 +
Transcript = externallink( TranscriptURL, nil, TranscriptURLorigin );
 
end
 
end
 
 
return '', '', '', ''; -- return empty strings
+
local Publisher;
end
+
if is_set(Periodical) and
 
+
not inArray(config.CitationClass, {"encyclopaedia","web","pressrelease","podcast"}) then
 
+
if is_set(PublisherName) then
--[[--------------------------< I N S O U R C E _ L O C _ G E T >----------------------------------------------
+
if is_set(PublicationPlace) then
 
+
Publisher = PublicationPlace .. ": " .. PublisherName;
returns one of the in-source locators: page, pages, or at.
+
else
 
+
Publisher = PublisherName; 
If any of these are interwiki links to wikisource, returns the label portion of the interwikilink as plain text
+
end
for use in COinS.  This COinS thing is done because here we convert an interwiki link to and external link and
+
elseif is_set(PublicationPlace) then
add an icon span around that; get_coins_pages() doesn't know about the span.  TODO: should it? 
+
Publisher= PublicationPlace;
 
+
else
TODO: add support for sheet and sheets?; streamline;
+
Publisher = "";
 
 
TODO: make it so that this function returns only one of the three as the single in-source (the return value assigned
 
to a new name)?
 
 
 
]]
 
 
 
local function insource_loc_get (page, pages, at)
 
local ws_url, ws_label, coins_pages, L; -- for wikisource interwikilinks; TODO: this corrupts page metadata (span remains in place after cleanup; fix there?)
 
 
 
if is_set (page) then
 
if is_set (pages) or is_set(at) then
 
pages = ''; -- unset the others
 
at = '';
 
 
end
 
end
extra_text_in_page_check (page); -- add this page to maint cat if |page= value begins with what looks like p. or pp.
+
if is_set(PublicationDate) then
 
+
if is_set(Publisher) then
ws_url, ws_label, L = wikisource_url_make (page); -- make ws url from |page= interwiki link; link portion L becomes tool tip label
+
Publisher = Publisher .. ", " .. wrap_msg ('published', PublicationDate);
if ws_url then
+
else
page = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in page'); -- space char after label to move icon away from in-source text; TODO: a better way to do this?
+
Publisher = PublicationDate;
page = substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, page});
+
end
coins_pages = ws_label;
 
 
end
 
end
elseif is_set (pages) then
+
if is_set(Publisher) then
if is_set (at) then
+
Publisher = " (" .. Publisher .. ")";
at = ''; -- unset
 
 
end
 
end
extra_text_in_page_check (pages); -- add this page to maint cat if |pages= value begins with what looks like p. or pp.
+
else
 
+
if is_set(PublicationDate) then
ws_url, ws_label, L = wikisource_url_make (pages); -- make ws url from |pages= interwiki link; link portion L becomes tool tip label
+
PublicationDate = " (" .. wrap_msg ('published', PublicationDate) .. ")";
if ws_url then
 
pages = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in pages'); -- space char after label to move icon away from in-source text; TODO: a better way to do this?
 
pages = substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, pages});
 
coins_pages = ws_label;
 
 
end
 
end
elseif is_set (at) then
+
if is_set(PublisherName) then
ws_url, ws_label, L = wikisource_url_make (at); -- make ws url from |at= interwiki link; link portion L becomes tool tip label
+
if is_set(PublicationPlace) then
if ws_url then
+
Publisher = sepc .. " " .. PublicationPlace .. ": " .. PublisherName .. PublicationDate;
at = external_link (ws_url, ws_label .. '&nbsp;', 'ws link in at'); -- space char after label to move icon away from in-source text; TODO: a better way to do this?
+
else
at = substitute (cfg.presentation['interwiki-icon'], {cfg.presentation['class-wikisource'], L, at});
+
Publisher = sepc .. " " .. PublisherName .. PublicationDate;
coins_pages = ws_label;
+
end
 +
elseif is_set(PublicationPlace) then
 +
Publisher= sepc .. " " .. PublicationPlace .. PublicationDate;
 +
else
 +
Publisher = PublicationDate;
 
end
 
end
 
end
 
end
 
 
return page, pages, at, coins_pages;
+
-- Several of the above rely upon detecting this as nil, so do it last.
end
+
if is_set(Periodical) then
 +
if is_set(Title) or is_set(TitleNote) then
 +
Periodical = sepc .. " " .. wrap_style ('italic-title', Periodical)
 +
else
 +
Periodical = wrap_style ('italic-title', Periodical)
 +
end
 +
end
  
 +
--[[
 +
Handle the oddity that is cite speech.  This code overrides whatever may be the value assigned to TitleNote (through |department=) and forces it to be " (Speech)" so that
 +
the annotation directly follows the |title= parameter value in the citation rather than the |event= parameter value (if provided).
 +
]]
 +
if "speech" == config.CitationClass then -- cite speech only
 +
TitleNote = " (Speech)"; -- annotate the citation
 +
if is_set (Periodical) then -- if Periodical, perhaps because of an included |website= or |journal= parameter
 +
if is_set (Conference) then -- and if |event= is set
 +
Conference = Conference .. sepc .. " "; -- then add appropriate punctuation to the end of the Conference variable before rendering
 +
end
 +
end
 +
end
  
 +
-- Piece all bits together at last.  Here, all should be non-nil.
 +
-- We build things this way because it is more efficient in LUA
 +
-- not to keep reassigning to the same string variable over and over.
  
--[=[-------------------------< A R C H I V E _ U R L _ C H E C K >--------------------------------------------
+
local tcommon
 
+
if inArray(config.CitationClass, {"journal","citation"}) and is_set(Periodical) then
Check archive.org urls to make sure they at least look like they are pointing at valid archives and not to the
+
if is_set(Others) then Others = Others .. sepc .. " " end
save snapshot url or to calendar pages.  When the archive url is 'https://web.archive.org/save/' (or http://...)
+
tcommon = safejoin( {Others, Title, TitleNote, Conference, Periodical, Format, TitleType, Scale, Series,  
archive.org saves a snapshot of the target page in the url.  That is something that Wikipedia should not allow
+
Language, Cartography, Edition, Publisher, Agency, Volume, Issue}, sepc );
unwitting readers to do.
+
else
 
+
tcommon = safejoin( {Title, TitleNote, Conference, Periodical, Format, TitleType, Scale, Series, Language,
When the archive.org url does not have a complete timestamp, archive.org chooses a snapshot according to its own
+
Volume, Issue, Others, Cartography, Edition, Publisher, Agency}, sepc );
algorithm or provides a calendar 'search' result.  [[WP:ELNO]] discourages links to search results.
+
end
 
 
This function looks at the value assigned to |archive-url= and returns empty strings for |archive-url= and
 
|archive-date= and an error message when:
 
|archive-url= holds an archive.org save command url
 
|archive-url= is an archive.org url that does not have a complete timestamp (YYYYMMDDhhmmss 14 digits) in the
 
correct place
 
otherwise returns |archive-url= and |archive-date=
 
 
 
There are two mostly compatible archive.org urls:
 
//web.archive.org/<timestamp>... -- the old form
 
//web.archive.org/web/<timestamp>... -- the new form
 
 
 
The old form does not support or map to the new form when it contains a display flag.  There are four identified flags
 
('id_', 'js_', 'cs_', 'im_') but since archive.org ignores others following the same form (two letters and an underscore)
 
we don't check for these specific flags but we do check the form.
 
 
 
This function supports a preview mode.  When the article is rendered in preview mode, this funct may return a modified
 
archive url:
 
for save command errors, return undated wildcard (/*/)
 
for timestamp errors when the timestamp has a wildcard, return the url unmodified
 
for timestamp errors when the timestamp does not have a wildcard, return with timestamp limited to six digits plus wildcard (/yyyymm*/)
 
 
 
]=]
 
 
 
local function archive_url_check (url, date)
 
local err_msg = ''; -- start with the error message empty
 
local path, timestamp, flag; -- portions of the archive.or url
 
 
 
if (not url:match('//web%.archive%.org/')) and (not url:match('//liveweb%.archive%.org/')) then -- also deprecated liveweb Wayback machine url
+
if #ID_list > 0 then
return url, date; -- not an archive.org archive, return ArchiveURL and ArchiveDate
+
ID_list = safejoin( { sepc .. " ",  table.concat( ID_list, sepc .. " " ), ID }, sepc );
 +
else
 +
ID_list = ID;
 
end
 
end
 
+
if url:match('//web%.archive%.org/save/') then -- if a save command url, we don't want to allow saving of the target page
+
local idcommon = safejoin( { ID_list, URL, Archived, AccessDate, Via, SubscriptionRequired, Lay, Quote }, sepc );
err_msg = 'save command';
+
local text;
url = url:gsub ('(//web%.archive%.org)/save/', '%1/*/', 1); -- for preview mode: modify ArchiveURL
+
local pgtext = Position .. Page .. Pages .. At;
elseif url:match('//liveweb%.archive%.org/') then
+
err_msg = 'liveweb';
+
if is_set(Authors) then
else
+
if is_set(Coauthors) then
path, timestamp, flag = url:match('//web%.archive%.org/([^%d]*)(%d+)([^/]*)/'); -- split out some of the url parts for evaluation
+
local sep = '; ';
+
if 'vanc' == NameListFormat then
if not is_set(timestamp) or 14 ~= timestamp:len() then -- path and flag optional, must have 14-digit timestamp here
+
sep = ', ';
err_msg = 'timestamp';
 
if '*' ~= flag then
 
url=url:gsub ('(//web%.archive%.org/[^%d]*%d?%d?%d?%d?%d?%d?)[^/]*', '%1*', 1) -- for preview, modify ts to be yearmo* max (0-6 digits plus splat)
 
 
end
 
end
elseif is_set(path) and 'web/' ~= path then -- older archive urls do not have the extra 'web/' path element
+
Authors = Authors .. sep .. Coauthors;
err_msg = 'path';
+
end
elseif is_set (flag) and not is_set (path) then -- flag not allowed with the old form url (without the 'web/' path element)
+
if is_set(Date) then
err_msg = 'flag';
+
Date = " ("..Date..")" .. OrigYear .. sepc .. " "
elseif is_set (flag) and not flag:match ('%a%a_') then -- flag if present must be two alpha characters and underscore (requires 'web/' path element)
+
elseif string.sub(Authors,-1,-1) == sepc then
err_msg = 'flag';
+
Authors = Authors .. " "
 
else
 
else
return url, date; -- return archiveURL and ArchiveDate
+
Authors = Authors .. sepc .. " "
 
end
 
end
end
+
if is_set(Editors) then
-- if here, something not right so
+
local in_text = " ";
table.insert( z.message_tail, { set_error( 'archive_url', {err_msg}, true ) } ); -- add error message and
+
local post_text = "";
if is_set (Frame:preprocess('{{REVISIONID}}')) then
+
if is_set(Chapter) then
return '', ''; -- return empty strings for archiveURL and ArchiveDate
+
in_text = in_text .. cfg.messages['in'] .. " "
 +
else
 +
if EditorCount <= 1 then
 +
post_text = ", " .. cfg.messages['editor'];
 +
else
 +
post_text = ", " .. cfg.messages['editors'];
 +
end
 +
end
 +
if (sepc ~= '.') then in_text = in_text:lower() end
 +
Editors = in_text .. Editors .. post_text;
 +
if (string.sub(Editors,-1,-1) == sepc)
 +
then Editors = Editors .. " "
 +
else Editors = Editors .. sepc .. " "
 +
end
 +
end
 +
text = safejoin( {Authors, Date, Chapter, Place, Editors, tcommon }, sepc );
 +
text = safejoin( {text, pgtext, idcommon}, sepc );
 +
elseif is_set(Editors) then
 +
if is_set(Date) then
 +
if EditorCount <= 1 then
 +
Editors = Editors .. ", " .. cfg.messages['editor'];
 +
else
 +
Editors = Editors .. ", " .. cfg.messages['editors'];
 +
end
 +
Date = " (" .. Date ..")" .. OrigYear .. sepc .. " "
 +
else
 +
if EditorCount <= 1 then
 +
Editors = Editors .. " (" .. cfg.messages['editor'] .. ")" .. sepc .. " "
 +
else
 +
Editors = Editors .. " (" .. cfg.messages['editors'] .. ")" .. sepc .. " "
 +
end
 +
end
 +
text = safejoin( {Editors, Date, Chapter, Place, tcommon}, sepc );
 +
text = safejoin( {text, pgtext, idcommon}, sepc );
 
else
 
else
return url, date; -- preview mode so return archiveURL and ArchiveDate
+
if is_set(Date) then
 +
if ( string.sub(tcommon,-1,-1) ~= sepc )
 +
  then Date = sepc .." " .. Date .. OrigYear
 +
  else Date = " " .. Date .. OrigYear
 +
end
 +
end
 +
if config.CitationClass=="journal" and is_set(Periodical) then
 +
text = safejoin( {Chapter, Place, tcommon}, sepc );
 +
text = safejoin( {text, pgtext, Date, idcommon}, sepc );
 +
else
 +
text = safejoin( {Chapter, Place, tcommon, Date}, sepc );
 +
text = safejoin( {text, pgtext, idcommon}, sepc );
 +
end
 
end
 
end
end
+
 +
if is_set(PostScript) and PostScript ~= sepc then
 +
text = safejoin( {text, sepc}, sepc );  --Deals with italics, spaces, etc.
 +
text = text:sub(1,-sepc:len()-1);
 +
-- text = text:sub(1,-2); --Remove final separator (assumes that sepc is only one character)
 +
end
 +
 +
text = safejoin( {text, PostScript}, sepc );
  
 
+
-- Now enclose the whole thing in a <span/> element
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
+
local options = {};
 
+
This is the main function doing the majority of the citation formatting.
+
if is_set(config.CitationClass) and config.CitationClass ~= "citation" then
 
+
options.class = "citation " .. config.CitationClass;
]]
+
else
 
+
options.class = "citation";
local function citation0( config, args)
 
--[[
 
Load Input Parameters
 
The argument_wrapper facilitates the mapping of multiple aliases to single internal variable.
 
]]
 
local A = argument_wrapper( args );
 
local i
 
 
 
-- Pick out the relevant fields from the arguments. Different citation templates
 
-- define different field names for the same underlying things.
 
 
 
-- set default parameter values defined by |mode= parameter.
 
local Mode = A['Mode'];
 
if not is_valid_parameter_value (Mode, 'mode', cfg.keywords['mode']) then
 
Mode = '';
 
 
end
 
end
 
+
local author_etal;
+
if is_set(Ref) and Ref:lower() ~= "none" then
local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors=
+
local id = Ref
local Authors;
+
if ( "harv" == Ref ) then
local NameListFormat = A['NameListFormat'];
+
local names = {} --table of last names & year
local Collaboration = A['Collaboration'];
+
if #a > 0 then
 
+
for i,v in ipairs(a) do
do -- to limit scope of selected
+
names[i] = v.last
local selected = select_author_editor_source (A['Vauthors'], A['Authors'], args, 'AuthorList');
+
if i == 4 then break end
if 1 == selected then
+
end
a, author_etal = extract_names (args, 'AuthorList'); -- fetch author list from |authorn= / |lastn= / |firstn=, |author-linkn=, and |author-maskn=
+
elseif #e > 0 then
elseif 2 == selected then
+
for i,v in ipairs(e) do
NameListFormat = 'vanc'; -- override whatever |name-list-format= might be
+
names[i] = v.last
a, author_etal = parse_vauthors_veditors (args, args.vauthors, 'AuthorList'); -- fetch author list from |vauthors=, |author-linkn=, and |author-maskn=
+
if i == 4 then break end
elseif 3 == selected then
+
end
Authors = A['Authors']; -- use content of |authors=
 
if 'authors' == A:ORIGIN('Authors') then -- but add a maint cat if the parameter is |authors=
 
add_maint_cat ('authors'); -- because use of this parameter is discouraged; what to do about the aliases is a TODO:
 
 
end
 
end
 +
names[ #names + 1 ] = first_set(Year, anchor_year); -- Year first for legacy citations and for YMD dates that require disambiguation
 +
id = anchorid(names)
 
end
 
end
if is_set (Collaboration) then
+
options.id = id;
author_etal = true; -- so that |display-authors=etal not required
 
end
 
 
end
 
end
 
+
local Others = A['Others'];
+
if string.len(text:gsub("<span[^>/]*>.-</span>", ""):gsub("%b<>","")) <= 2 then
 
+
z.error_categories = {};
local editor_etal;
+
text = seterror('empty_citation');
local e = {}; -- editors list from |editor-lastn= / |editor-firstn= pairs or |veditors=
+
z.message_tail = {};
local Editors;
 
 
 
do -- to limit scope of selected
 
local selected = select_author_editor_source (A['Veditors'], A['Editors'], args, 'EditorList');
 
if 1 == selected then
 
e, editor_etal = extract_names (args, 'EditorList'); -- fetch editor list from |editorn= / |editor-lastn= / |editor-firstn=, |editor-linkn=, and |editor-maskn=
 
elseif 2 == selected then
 
NameListFormat = 'vanc'; -- override whatever |name-list-format= might be
 
e, editor_etal = parse_vauthors_veditors (args, args.veditors, 'EditorList'); -- fetch editor list from |veditors=, |editor-linkn=, and |editor-maskn=
 
elseif 3 == selected then
 
Editors = A['Editors']; -- use content of |editors=
 
add_maint_cat ('editors'); -- but add a maint cat because use of this parameter is discouraged
 
end
 
 
end
 
end
 +
 +
if is_set(options.id) then
 +
text = '<span id="' .. mw.uri.anchorEncode(options.id) ..'" class="' .. mw.text.nowiki(options.class) .. '">' .. text .. "</span>";
 +
else
 +
text = '<span class="' .. mw.text.nowiki(options.class) .. '">' .. text .. "</span>";
 +
end
  
local translator_etal;
+
local empty_span = '<span style="display:none;">&nbsp;</span>';
local t = {}; -- translators list from |translator-lastn= / translator-firstn= pairs
+
local Translators; -- assembled translators name list
+
-- Note: Using display: none on then COinS span breaks some clients.
t = extract_names (args, 'TranslatorList'); -- fetch translator list from |translatorn= / |translator-lastn=, -firstn=, -linkn=, -maskn=
+
local OCinS = '<span title="' .. OCinSoutput .. '" class="Z3988">' .. empty_span .. '</span>';
 
+
text = text .. OCinS;
local interviewer_etal;
+
local interviewers_list = {};
+
if #z.message_tail ~= 0 then
local Interviewers; -- used later
+
text = text .. " ";
interviewers_list = extract_names (args, 'InterviewerList'); -- process preferred interviewers parameters
+
for i,v in ipairs( z.message_tail ) do
 
+
if is_set(v[1]) then
local contributor_etal;
+
if i == #z.message_tail then
local c = {}; -- contributors list from |contributor-lastn= / contributor-firstn= pairs
+
text = text .. errorcomment( v[1], v[2] );
local Contributors; -- assembled contributors name list
+
else
local Contribution = A['Contribution'];
+
text = text .. errorcomment( v[1] .. "; ", v[2] );
if in_array(config.CitationClass, {"book","citation"}) and not is_set(A['Periodical']) then -- |contributor= and |contribution= only supported in book cites
+
end
c = extract_names (args, 'ContributorList'); -- fetch contributor list from |contributorn= / |contributor-lastn=, -firstn=, -linkn=, -maskn=
 
 
if 0 < #c then
 
if not is_set (Contribution) then -- |contributor= requires |contribution=
 
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'contribution')}); -- add missing contribution error message
 
c = {}; -- blank the contributors' table; it is used as a flag later
 
end
 
if 0 == #a then -- |contributor= requires |author=
 
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'author')}); -- add missing author error message
 
c = {}; -- blank the contributors' table; it is used as a flag later
 
 
end
 
end
 
end
 
end
else -- if not a book cite
 
if select_one (args, cfg.aliases['ContributorList-Last'], 'redundant_parameters', 1 ) then -- are there contributor name list parameters?
 
table.insert( z.message_tail, { set_error( 'contributor_ignored')}); -- add contributor ignored error message
 
end
 
Contribution = nil; -- unset
 
 
end
 
end
 +
 +
no_tracking_cats = no_tracking_cats:lower();
 +
if inArray(no_tracking_cats, {"", "no", "false", "n"}) then
 +
for _, v in ipairs( z.error_categories ) do
 +
text = text .. '[[Category:' .. v ..']]';
 +
end
 +
for _, v in ipairs( z.maintenance_cats ) do -- append maintenance categories
 +
text = text .. '[[Category:' .. v ..']]';
 +
end
 +
for _, v in ipairs( z.properties_cats ) do -- append maintenance categories
 +
text = text .. '[[Category:' .. v ..']]';
 +
end
 +
end
 +
 +
return text
 +
end
  
if not is_valid_parameter_value (NameListFormat, 'name-list-format', cfg.keywords['name-list-format']) then -- only accepted value for this parameter is 'vanc'
+
-- This is used by templates such as {{cite book}} to create the actual citation text.
NameListFormat = ''; -- anything else, set to empty string
+
function z.citation(frame)
 +
local pframe = frame:getParent()
 +
 +
if nil ~= string.find( frame:getTitle(), 'sandbox', 1, true ) then -- did the {{#invoke:}} use sandbox version?
 +
cfg = mw.loadData( 'Module:Citation/CS1/Configuration/sandbox' ); -- load sandbox versions of Configuration and Whitelist and ...
 +
whitelist = mw.loadData( 'Module:Citation/CS1/Whitelist/sandbox' );
 +
dates = require('Module:Citation/CS1/Date_validation/sandbox').dates -- ... sandbox version of date validation code
 +
else -- otherwise
 +
cfg = mw.loadData( 'Module:Citation/CS1/Configuration' ); -- load live versions of Configuration and Whitelist and ...
 +
whitelist = mw.loadData( 'Module:Citation/CS1/Whitelist' );
 +
dates = require('Module:Citation/CS1/Date_validation').dates -- ... live version of date validation code
 
end
 
end
 +
 +
local args = {};
 +
local suggestions = {};
 +
local error_text, error_state;
  
if is_set (Others) then
+
local config = {};
if 0 == #a and 0 == #e then -- add maint cat when |others= has value and used without |author=, |editor=
+
for k, v in pairs( frame.args ) do
add_maint_cat ('others');
+
config[k] = v;
end
+
args[k] = v;  
end
+
end
  
local Year = A['Year'];
+
for k, v in pairs( pframe.args ) do
local PublicationDate = A['PublicationDate'];
+
if v ~= '' then
local OrigYear = A['OrigYear'];
+
if not validate( k ) then
local Date = A['Date'];
+
error_text = "";
local LayDate = A['LayDate'];
+
if type( k ) ~= 'string' then
------------------------------------------------- Get title data
+
-- Exclude empty numbered parameters
local Title = A['Title'];
+
if v:match("%S+") ~= nil then
local ScriptTitle = A['ScriptTitle'];
+
error_text, error_state = seterror( 'text_ignored', {v}, true );
local BookTitle = A['BookTitle'];
+
end
local Conference = A['Conference'];
+
elseif validate( k:lower() ) then
local TransTitle = A['TransTitle'];
+
error_text, error_state = seterror( 'parameter_ignored_suggest', {k, k:lower()}, true );
local TitleNote = A['TitleNote'];
+
else
local TitleLink = A['TitleLink'];
+
if #suggestions == 0 then
link_title_ok (TitleLink, A:ORIGIN ('TitleLink'), Title, 'title'); -- check for wikimarkup in |title-link= or wikimarkup in |title= when |title-link= is set
+
suggestions = mw.loadData( 'Module:Citation/CS1/Suggestions' );
 
+
end
local Chapter = A['Chapter'];
+
if suggestions[ k:lower() ] ~= nil then
local ScriptChapter = A['ScriptChapter'];
+
error_text, error_state = seterror( 'parameter_ignored_suggest', {k, suggestions[ k:lower() ]}, true );
local ChapterLink -- = A['ChapterLink']; -- deprecated as a parameter but still used internally by cite episode
+
else
local TransChapter = A['TransChapter'];
+
error_text, error_state = seterror( 'parameter_ignored', {k}, true );
local TitleType = A['TitleType'];
+
end
local Degree = A['Degree'];
+
end  
local Docket = A['Docket'];
+
if error_text ~= '' then
local ArchiveFormat = A['ArchiveFormat'];
+
table.insert( z.message_tail, {error_text, error_state} );
 
+
end