求真百科歡迎當事人提供第一手真實資料,洗刷冤屈,終結網路霸凌。

「模块:Citation/CS1」修訂間的差異檢視原始碼檢視歷史

事實揭露 揭密真相
前往: 導覽搜尋
(导入1个版本)
(导入1个版本)
 
(未顯示由 3 位使用者於中間所作的 5 次修訂)
行 1: 行 1:
  
local cs1 ={};
+
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
 +
}
  
 
--[[--------------------------< F O R W A R D  D E C L A R A T I O N S >--------------------------------------
 
--[[--------------------------< F O R W A R D  D E C L A R A T I O N S >--------------------------------------
 
]]
 
]]
 +
local dates, year_date_check -- functions in Module:Citation/CS1/Date_validation
  
local dates, year_date_check, reformat_dates, date_hyphen_to_dash, -- functions in Module:Citation/CS1/Date_validation
+
local cfg = {}; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration
date_name_xlate
+
local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist
  
local is_set, in_array, substitute, error_comment, set_error, select_one, -- functions in Module:Citation/CS1/Utilities
+
--[[--------------------------< I S _ S E T >------------------------------------------------------------------
add_maint_cat, wrap_style, safe_for_italics, is_wikilink, make_wikilink;
 
  
local z ={}; -- tables in Module:Citation/CS1/Utilities
+
Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not an empty string.
 
+
This function is global because it is called from both this module and from Date validation
local extract_ids, extract_id_access_levels, build_id_list, is_embargoed; -- functions in Module:Citation/CS1/Identifiers
 
 
 
local make_coins_title, get_coins_pages, COinS; -- functions in Module:Citation/CS1/COinS
 
 
 
local cfg = {}; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration
 
local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist
 
 
 
 
 
--[[--------------------------< P A G E  S C O P E  V A R I A B L E S >--------------------------------------
 
 
 
delare variables here that have page-wide scope that are not brought in from other modules; thatare created here
 
and used here
 
  
 
]]
 
]]
 
+
function is_set( var )
local added_deprecated_cat; -- boolean flag so that the category is added only once
+
return not (var == nil or var == '');
local added_prop_cats = {}; -- list of property categories that have been added to z.properties_cats
+
end
local added_vanc_errs; -- boolean flag so we only emit one Vancouver error / category
 
 
 
local Frame; -- holds the module's frame table
 
 
 
  
 
--[[--------------------------< F I R S T _ S E T >------------------------------------------------------------
 
--[[--------------------------< F I R S T _ S E T >------------------------------------------------------------
行 56: 行 46:
 
end
 
end
  
 +
--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------
  
--[[--------------------------< A D D _ P R O P _ C A T >--------------------------------------------------------
+
Whether needle is in haystack
 +
 
 +
]]
  
Adds a category to z.properties_cats using names from the configuration file with additional text if any.
+
local function in_array( 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
  
foreign_lang_source and foreign_lang_source_2 keys have a language code appended to them so that multiple languages
+
--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
may be categorized but multiples of the same language are not categorized.
 
  
added_prop_cats is a table declared in page scope variables above
+
Populates numbered arguments in a message string using an argument table.
  
 
]]
 
]]
  
local function add_prop_cat (key, arguments)
+
local function substitute( msg, args )
if not added_prop_cats [key] then
+
return args and mw.message.newRawMessage( msg, args ):plain() or msg;
added_prop_cats [key] = true; -- note that we've added this category
 
key = key:gsub ('(foreign_lang_source_?2?)%a%a%a?', '%1'); -- strip lang code from keyname
 
table.insert( z.properties_cats, substitute (cfg.prop_cats [key], arguments)); -- make name then add to table
 
end
 
 
end
 
end
  
 +
--[[--------------------------< E R R O R _ C O M M E N T >----------------------------------------------------
  
--[[--------------------------< A D D _ V A N C _ E R R O R >----------------------------------------------------
+
Wraps error messages with css markup according to the state of hidden.
 
 
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 error_comment( content, hidden )
local function add_vanc_error (source)
+
return substitute( hidden and cfg.presentation['hidden-error'] or cfg.presentation['visible-error'], content );
if not added_vanc_errs then
 
added_vanc_errs = true; -- note that we've added this category
 
table.insert( z.message_tail, { set_error( 'vancouver', {source}, true ) } );
 
end
 
 
end
 
end
  
 +
--[[--------------------------< S E T _ E R R O R >--------------------------------------------------------------
  
--[[--------------------------< I S _ S C H E M E >------------------------------------------------------------
+
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.
  
does this thing that purports to be a uri scheme seem to be a valid scheme?  The scheme is checked to see if it
+
]]
is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says:
+
local function set_error( error_id, arguments, raw, prefix, suffix )
Scheme names consist of a sequence of characters beginning with a
+
local error_state = cfg.error_conditions[ error_id ];
  letter and followed by any combination of letters, digits, plus
+
  ("+"), period ("."), or hyphen ("-").
+
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
 +
 +
local message = substitute( error_state.message, arguments );
 +
 +
message = message .. " ([[" .. cfg.messages['help page link'] ..
 +
"#" .. error_state.anchor .. "|" ..
 +
cfg.messages['help page label'] .. "]])";
 +
 +
z.error_ids[ error_id ] = true;
 +
if in_array( error_id, { 'bare_url_missing_title', 'trans_missing_title' } )
 +
and z.error_ids['citation_missing_title'] then
 +
return '', false;
 +
end
 +
 +
message = table.concat({ prefix, message, suffix });
 +
 +
if raw == true then
 +
return message, error_state.hidden;
 +
end
 +
 +
return error_comment( message, error_state.hidden );
 +
end
 +
 
 +
--[[--------------------------< A D D _ M A I N T _ C A T >------------------------------------------------------
  
returns true if it does, else false
+
Adds a category to z.maintenance_cats using names from the configuration file with additional text if any.
 +
To prevent duplication, the added_maint_cats table lists the categories by key that have been added to z.maintenance_cats.
  
 
]]
 
]]
  
local function is_scheme (scheme)
+
local added_maint_cats = {} -- list of maintenance categories that have been added to z.maintenance_cats
return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern
+
local function add_maint_cat (key, arguments)
 +
if not added_maint_cats [key] then
 +
added_maint_cats [key] = true; -- note that we've added this category
 +
table.insert( z.maintenance_cats, substitute (cfg.maint_cats [key], arguments)); -- make name then add to table
 +
end
 
end
 
end
  
 +
--[[--------------------------< A D D _ P R O P _ C A T >--------------------------------------------------------
  
--[=[-------------------------< I S _ D O M A I N _ N A M E >--------------------------------------------------
+
Adds a category to z.properties_cats using names from the configuration file with additional text if any.
  
Does this thing that purports to be a domain name seem to be a valid domain name?
+
]]
  
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
+
local added_prop_cats = {} -- list of property categories that have been added to z.properties_cats
BNF defined here: https://tools.ietf.org/html/rfc4234
+
local function add_prop_cat (key, arguments)
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
+
if not added_prop_cats [key] then
see also [[Single-letter second-level domain]]
+
added_prop_cats [key] = true; -- note that we've added this category
list of tlds: https://www.iana.org/domains/root/db
+
table.insert( z.properties_cats, substitute (cfg.prop_cats [key], arguments)); -- make name then add to table
 +
end
 +
end
  
rfc952 (modified by rfc 1123) requires the first and last character of a hostname to be a letter or a digit.  Between
+
--[[--------------------------< A D D _ V A N C _ E R R O R >----------------------------------------------------
the first and last characters the name may use letters, digits, and the hyphen.
 
  
Also allowed are IPv4 addresses. IPv6 not supported
+
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.
  
domain is expected to be stripped of any path so that the last character in the last character of the tld.  tld
+
]]
is two or more alpha characters.  Any preceding '//' (from splitting a url with a scheme) will be stripped
 
here.  Perhaps not necessary but retained incase it is necessary for IPv4 dot decimal.
 
  
There are several tests:
+
local added_vanc_errs; -- flag so we only emit one Vancouver error / category
the first character of the whole domain name including subdomains must be a letter or a digit
+
local function add_vanc_error ()
internationalized domain name (ascii characters with .xn-- ASCII Compatible Encoding (ACE) prefix xn-- in the tld) see https://tools.ietf.org/html/rfc3490
+
if not added_vanc_errs then
single-letter/digit second-level domains in the .org and .cash TLDs
+
added_vanc_errs = true; -- note that we've added this category
q, x, and z SL domains in the .com TLD
+
table.insert( z.message_tail, { set_error( 'vancouver', {}, true ) } );
i and q SL domains in the .net TLD
+
end
single-letter SL domains in the ccTLDs (where the ccTLD is two letters)
+
end
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
 
  
]=]
+
--[[--------------------------< I S _ S C H E M E >------------------------------------------------------------
  
local function is_domain_name (domain)
+
does this thing that purports to be a uri scheme seem to be a valid scheme?  The scheme is checked to see if it
if not domain then
+
is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says:
return false; -- if not set, abandon
+
Scheme names consist of a sequence of characters beginning with a
end
+
  letter and followed by any combination of letters, digits, plus
+
  ("+"), period ("."), or hyphen ("-").
domain = domain:gsub ('^//', ''); -- strip '//' from domain name if present; done here so we only have to do it once
+
 
+
returns true if it does, else false
if not domain:match ('^[%a%d]') then -- first character must be letter or digit
+
 
return false;
+
]]
end
+
 
-- Do most common case first
+
local function is_scheme (scheme)
if domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.%a%a+$') then -- three or more character hostname.hostname or hostname.tld
+
return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern
return true;
 
elseif domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.xn%-%-[%a%d]+$') then -- internationalized domain name with ACE prefix
 
return true;
 
elseif domain:match ('%f[%a%d][%a%d]%.cash$') then -- one character/digit .cash hostname
 
return true;
 
elseif domain:match ('%f[%a%d][%a%d]%.org$') then -- one character/digit .org hostname
 
return true;
 
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
 
end
  
  
--[[--------------------------< I S _ U R L >------------------------------------------------------------------
+
--[=[-------------------------< I S _ D O M A I N _ N A M E >--------------------------------------------------
  
returns true if the scheme and domain parts of a url appear to be a valid url; else false.
+
Does this thing that purports to be a domain name seem to be a valid domain name?
  
This function is the last step in the validation process. This function is separate because there are cases that
+
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
are not covered by split_url(), for example is_parameter_ext_wikilink() which is looking for bracketted external
+
BNF defined here: https://tools.ietf.org/html/rfc4234
wikilinks.
+
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
 +
the first and last characters the name may use letters, digits, and the hyphen.
  
local function is_url (scheme, domain)
+
Also allowed are IPv4 addresses. IPv6 not supported
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
 
  
 +
domain is expected to be stripped of any path so that the last character in the last character of the tld.  tld
 +
is two or more alpha characters.  Any preceding '//' (from splitting a url with a scheme) will be stripped
 +
here.  Perhaps not necessary but retained incase it is necessary for IPv4 dot decimal.
  
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
+
There are several tests:
 +
the first character of the whole domain name including subdomains must be a letter or a digit
 +
single-letter/digit second-level domains in the .org TLD
 +
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
  
Split a url into a scheme, authority indicator, and domain.
+
returns true if domain appears to be a proper name and tld or IPv4 address, else false
  
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.
+
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 domain:match ('^[%a%d]') then -- first character must be letter or digit
 +
return false;
 +
end
 +
 +
if domain:match ('%f[%a%d][%a%d]%.org$') then -- one character .org hostname
 +
return true;
 +
elseif domain:match ('%f[%a][qxz]%.com$') then -- assigned one character .com hostname (x.com times out 2015-12-10)
 +
return true;
 +
elseif domain:match ('%f[%a][iq]%.net$') then -- assigned one character .net hostname (q.net registered but not active 2015-12-10)
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.xn%-%-[%a%d]+$') then -- internationalized domain name with ACE prefix
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d]%.cash$') then -- one character/digit .cash hostname
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d]%.%a%a$') then -- one character hostname and cctld (2 chars)
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d][%a%d]%.%a%a+$') then -- two character hostname and tld
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.%a%a+$') then -- three or more character hostname.hostname or hostname.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
  
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;
+
--[[--------------------------< I S _ U R L >------------------------------------------------------------------
  
]]
+
returns true if the scheme and domain parts of a url appear to be a valid url; else false.
  
local function split_url (url_str)
+
This function is the last step in the validation process.  This function is separate because there are cases that
 +
are not covered by split_url(), for example is_parameter_ext_wikilink() which is looking for bracketted external
 +
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.
 +
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 following the scheme's colon), make sure that there are only 2.
 +
 
 +
]]
 +
 
 +
local function split_url (url_str)
 
local scheme, authority, domain;
 
local scheme, authority, domain;
 
 
行 260: 行 327:
  
  
--[[--------------------------< L I N K _ T I T L E _ O K >---------------------------------------------------
+
--[[--------------------------< C H E C K _ U R L >------------------------------------------------------------
  
Use link_param_ok() to validate |<param>-link= value and its matching |<title>= value.
+
Determines whether a URL string appears to be valid.
  
|<title>= may be wikilinked but not when |<param>-link= has a valueThis function emits an error message when
+
First we test for space charactersIf any are found, return false.  Then split the url into scheme and domain
that condition exists
+
portions, or for protocol relative (//example.com) urls, just the domain.  Use is_url() to validate the two
 +
portions of the url.  If both are valid, or for protocol relative if domain is valid, return true, else false.
  
 
]]
 
]]
  
local function link_title_ok (link, lorig, title, torig)
+
local function check_url( url_str )
local orig;
+
if nil == url_str:match ("^%S+$") then -- if there are any spaces in |url=value it can't be a proper url
 
+
return false;
if is_set (link) then -- don't bother if <param>-link doesn't have a value
 
if not link_param_ok (link) then -- check |<param>-link= markup
 
orig = lorig; -- identify the failing link parameter
 
elseif title:find ('%[%[') then -- check |title= for wikilink markup
 
orig = torig; -- identify the failing |title= parameter
 
end
 
 
end
 
end
 +
local scheme, domain;
  
if is_set (orig) then
+
scheme, domain = split_url (url_str); -- get scheme or nil and domain or nil from url;  
table.insert( z.message_tail, { set_error( 'bad_paramlink', orig)}); -- url or wikilink in |title= with |title-link=;
+
return is_url (scheme, domain); -- return true if value appears to be a valid url
end
 
end
 
 
 
 
 
--[[--------------------------< C H E C K _ U R L >------------------------------------------------------------
 
 
 
Determines whether a URL string appears to be valid.
 
 
 
First we test for space characters.  If any are found, return false.  Then split the url into scheme and domain
 
portions, or for protocol relative (//example.com) urls, just the domain.  Use is_url() to validate the two
 
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
 
 
 
]]
 
 
 
local function check_url( url_str )
 
if nil == url_str:match ("^%S+$") then -- if there are any spaces in |url=value it can't be a proper url
 
return false;
 
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
 
return domain:match('^[%a%d%+%-_]+%.[%a%d%+%-_%.]*[%a%d%+%-_]$');
 
end
 
 
return is_url (scheme, domain); -- return true if value appears to be a valid url
 
 
end
 
end
  
行 331: 行 363:
 
local function is_parameter_ext_wikilink (value)
 
local function is_parameter_ext_wikilink (value)
 
local scheme, domain;
 
local scheme, domain;
 +
 +
value = value:gsub ('([^%s/])/[%a%d].*', '%1'); -- strip path information (the capture prevents false replacement of '//')
  
 
if value:match ('%f[%[]%[%a%S*:%S+.*%]') then -- if ext wikilink with scheme and domain: [xxxx://yyyyy.zzz]
 
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+).*%]'));
+
scheme, domain = value:match ('%f[%[]%[(%a%S*:)(%S+).*%]')
elseif value:match ('%f[%[]%[//%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
+
elseif value:match ('%f[%[]%[//%S*%.%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
scheme, domain = split_url (value:match ('%f[%[]%[(//%S+).*%]'));
+
domain = value:match ('%f[%[]%[//(%S*%.%S+).*%]');
 
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
 
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+)'));
+
scheme, domain = value:match ('(%a%S*:)(%S+)');
elseif value:match ('//%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
+
elseif value:match ('//%S*%.%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
+
domain = value:match ('//(%S*%.%S+)'); -- what is left should be the domain
 
else
 
else
 
return false; -- didn't find anything that is obviously a url
 
return false; -- didn't find anything that is obviously a url
行 370: 行 404:
  
  
--[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------
+
--[[--------------------------< S A F E _ F O R _ I T A L I C S >----------------------------------------------
 +
 
 +
Protects a string that will be wrapped in wiki italic markup '' ... ''
  
Escape sequences for content that will be used for URL descriptions
+
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.
  
 
]]
 
]]
  
local function safe_for_url( str )
+
local function safe_for_italics( str )
if str:match( "%[%[.-%]%]" ) ~= nil then  
+
if not is_set(str) then
table.insert( z.message_tail, { set_error( 'wikilink_in_url', {}, true ) } );
+
return str;
 +
else
 +
if str:sub(1,1) == "'" then str = "<span></span>" .. str; end
 +
if str:sub(-1,-1) == "'" then str = str .. "<span></span>"; end
 +
 +
-- Remove newlines as they break italics.
 +
return str:gsub( '\n', ' ' );
 +
end
 +
end
 +
 
 +
--[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------
 +
 
 +
Escape sequences for content that will be used for URL descriptions
 +
 
 +
]]
 +
 
 +
local function safe_for_url( str )
 +
if str:match( "%[%[.-%]%]" ) ~= nil then  
 +
table.insert( z.message_tail, { set_error( 'wikilink_in_url', {}, true ) } );
 
end
 
end
 
 
行 387: 行 443:
 
end
 
end
  
 +
--[[--------------------------< W R A P _ S T Y L E >----------------------------------------------------------
 +
 +
Applies styling to various parameters.  Supplied string is wrapped using a message_list configuration taking one
 +
argument; protects italic styled parameters.  Additional text taken from citation_config.presentation - the reason
 +
this function is similar to but separate from wrap_msg().
 +
 +
]]
 +
 +
local function wrap_style (key, str)
 +
if not is_set( str ) then
 +
return "";
 +
elseif in_array( key, { 'italic-title', 'trans-italic-title' } ) then
 +
str = safe_for_italics( str );
 +
end
 +
 +
return substitute( cfg.presentation[key], {str} );
 +
end
  
 
--[[--------------------------< E X T E R N A L _ L I N K >----------------------------------------------------
 
--[[--------------------------< E X T E R N A L _ L I N K >----------------------------------------------------
行 394: 行 467:
 
]]
 
]]
  
local function external_link( URL, label, source, access)
+
local function external_link( URL, label, source )
 
local error_str = "";
 
local error_str = "";
local domain;
 
local path;
 
local base_url;
 
 
 
if not is_set( label ) then
 
if not is_set( label ) then
 
label = URL;
 
label = URL;
行 411: 行 480:
 
error_str = set_error( 'bad_url', {wrap_style ('parameter', source)}, false, " " ) .. error_str;
 
error_str = set_error( 'bad_url', {wrap_style ('parameter', source)}, false, " " ) .. error_str;
 
end
 
end
+
return table.concat({ "[", URL, " ", safe_for_url( label ), "]", error_str });
domain, path = URL:match ('^([/%.%-%+:%a%d]+)([/%?#].*)$'); -- split the url into scheme plus domain and path
+
end
if path then -- if there is a path portion
+
 
path = path:gsub ('[%[%]]', {['[']='%5b',[']']='%5d'}); -- replace '[' and ']' with their percent encoded values
+
--[[--------------------------< E X T E R N A L _ L I N K _ I D >----------------------------------------------
URL=domain..path; -- and reassemble
+
 
end
+
Formats a wiki style external link
+
 
if is_set (access) then -- access level (subscription, registration, limited)
+
]]
label = safe_for_url (label); -- replace square brackets and newlines
 
  
base_url = table.concat ( -- assemble external link with access signal
+
local function external_link_id(options)
{
+
local url_string = options.id;
'<span class="plainlinks">[', -- opening css and url markup
+
if options.encode == true or options.encode == nil then
URL, -- the url
+
url_string = mw.uri.encode( url_string );
' ', -- the required space
 
label,
 
'<span style="padding-left:0.15em">', -- signal spacing css
 
cfg.presentation[access], -- the appropriate icon
 
'</span>', -- close signal spacing span
 
']</span>' -- close url markup and plain links span
 
});
 
else
 
base_url = table.concat({ "[", URL, " ", safe_for_url( label ), "]" }); -- no signal markup
 
 
end
 
end
+
return mw.ustring.format( '[%s%s%s \<span title\=\"%s\"\>%s%s%s\<\/span\>]',
return table.concat({ base_url, error_str });
+
options.prefix, url_string, options.suffix or "",
 +
options.link, options.label, options.separator or "&nbsp;",
 +
mw.text.nowiki(options.id)
 +
);
 
end
 
end
 
  
 
--[[--------------------------< D E P R E C A T E D _ P A R A M E T E R >--------------------------------------
 
--[[--------------------------< D E P R E C A T E D _ P A R A M E T E R >--------------------------------------
行 445: 行 506:
 
offending parameter name to the error message.  Only one error message is emitted regardless of the number of deprecated
 
offending parameter name to the error message.  Only one error message is emitted regardless of the number of deprecated
 
parameters in the citation.
 
parameters in the citation.
 
added_deprecated_cat is a boolean declared in page scope variables above
 
  
 
]]
 
]]
  
 +
local page_in_deprecated_cat; -- sticky flag so that the category is added only once
 
local function deprecated_parameter(name)
 
local function deprecated_parameter(name)
if not added_deprecated_cat then
+
if not page_in_deprecated_cat then
added_deprecated_cat = true; -- note that we've added this category
+
page_in_deprecated_cat = true; -- note that we've added this category
 
table.insert( z.message_tail, { set_error( 'deprecated_params', {name}, true ) } ); -- add error message
 
table.insert( z.message_tail, { set_error( 'deprecated_params', {name}, true ) } ); -- add error message
 
end
 
end
 
end
 
end
  
 +
--[[--------------------------< K E R N _ Q U O T E S >--------------------------------------------------------
  
--[=[-------------------------< K E R N _ Q U O T E S >--------------------------------------------------------
+
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:
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'"
 
"'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)
 
" '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.
 
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
 
quote marks regardless of the need for kerning.  Unicode quote marks are not replaced in simple [[D]] wikilinks.
 
  
 
Call this function for chapter titles, for website titles, etc; not for book titles.
 
Call this function for chapter titles, for website titles, etc; not for book titles.
  
]=]
+
]]
  
 
local function kern_quotes (str)
 
local function kern_quotes (str)
 
local cap='';
 
local cap='';
 
local cap2='';
 
local cap2='';
local wl_type, label, link;
 
 
wl_type, label, link = is_wikilink (str); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]]
 
 
 
if 1 == wl_type then -- [[D]] simple wikilink with or without quote marks
+
cap, cap2 = str:match ("^([\"\'])([^\'].+)"); -- match leading double or single quote but not double single quotes
if mw.ustring.match (str, '%[%[[\" “”\'‘’].+[\" “”\' ‘’]%]%]') then -- leading and trailing quote marks
+
if is_set (cap) then
str = substitute (cfg.presentation['kern-wl-both'], str);
+
str = substitute (cfg.presentation['kern-left'], {cap, cap2});
elseif mw.ustring.match (str, '%[%[[\"“”\' ‘’].+%]%]') then -- leading quote marks
+
end
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
 
 
 
else -- plain text or [[L|D]]; text in label variable
 
label= mw.ustring.gsub (label, '[“”]', '\"'); -- replace “” (U+201C & U+201D) with " (typewriter double quote mark)
 
label= mw.ustring.gsub (label, '[‘’]', '\''); -- replace ‘’ (U+2018 & U+2019) with ' (typewriter single quote mark)
 
  
cap, cap2 = mw.ustring.match (label, "^([\"\'])([^\'].+)"); -- match leading double or single quote but not doubled single quotes (italic markup)
+
cap, cap2 = str:match ("^(.+[^\'])([\"\'])$")
if is_set (cap) then
+
if is_set (cap) then
label = substitute (cfg.presentation['kern-left'], {cap, cap2});
+
str = substitute (cfg.presentation['kern-right'], {cap, cap2});
end
 
 
cap, cap2 = mw.ustring.match (label, "^(.+[^\'])([\"\'])$") -- match trailing double or single quote but not doubled single quotes (italic markup)
 
if is_set (cap) then
 
label = substitute (cfg.presentation['kern-right'], {cap, cap2});
 
end
 
 
if 2 == wl_type then
 
str = make_wikilink (link, label); -- reassemble the wikilink
 
else
 
str = label;
 
end
 
 
end
 
end
 
return str;
 
return str;
 
end
 
end
 
  
 
--[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >----------------------------------------
 
--[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >----------------------------------------
行 549: 行 579:
 
end
 
end
 
-- if we get this far we have prefix and script
 
-- 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
+
name = mw.language.fetchLanguageName( lang, mw.getContentLanguage():getCode() ); -- 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?
 
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
 
script_value = script_value:gsub ('^%l%l%s*:%s*', ''); -- strip prefix from script
 
-- is prefix one of these language codes?
 
-- is prefix one of these language codes?
if in_array (lang, cfg.script_lang_codes) then
+
if in_array (lang, {'ar', 'bg', 'bs', 'dv', 'el', 'fa', 'he', 'hy', 'ja', 'ka', 'ko', 'ku', 'mk', 'ps', 'ru', 'sd', 'sr', 'th', 'uk', 'ug', 'yi', 'zh'}) then
 
add_prop_cat ('script_with_name', {name, lang})
 
add_prop_cat ('script_with_name', {name, lang})
 
else
 
else
行 562: 行 592:
 
lang = ''; -- invalid so set lang to empty string
 
lang = ''; -- invalid so set lang to empty string
 
end
 
end
 +
end
 +
if is_set(script_value) then
 +
script_value = '-{R|' .. script_value .. '}-';
 
end
 
end
 
script_value = substitute (cfg.presentation['bdi'], {lang, script_value}); -- isolate in case script is rtl
 
script_value = substitute (cfg.presentation['bdi'], {lang, script_value}); -- isolate in case script is rtl
行 567: 行 600:
 
return script_value;
 
return script_value;
 
end
 
end
 
  
 
--[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------
 
--[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------
行 576: 行 608:
  
 
local function script_concatenate (title, script)
 
local function script_concatenate (title, script)
 +
if is_set(title) then
 +
title = '-{zh;zh-hans;zh-hant|' .. title .. '}-';
 +
end
 
if is_set (script) then
 
if is_set (script) then
 
script = format_script_value (script); -- <bdi> tags, lang atribute, categorization, etc; returns empty string on error
 
script = format_script_value (script); -- <bdi> tags, lang atribute, categorization, etc; returns empty string on error
行 608: 行 643:
  
  
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
+
--[[-------------------------< I S _ A L I A S _ U S E D >-----------------------------------------------------
 +
 
 +
This function is used by select_one() to determine if one of a list of alias parameters is in the argument list
 +
provided by the template.
 +
 
 +
Input:
 +
args – pointer to the arguments table from calling template
 +
alias – one of the list of possible aliases in the aliases lists from Module:Citation/CS1/Configuration
 +
index – for enumerated parameters, identifies which one
 +
enumerated – true/false flag used choose how enumerated aliases are examined
 +
value – value associated with an alias that has previously been selected; nil if not yet selected
 +
selected – the alias that has previously been selected; nil if not yet selected
 +
error_list – list of aliases that are duplicates of the alias already selected
  
Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
+
Returns:
parameter (chapter_url_source used for error messages).
+
value – value associated with alias we selected or that was previously selected or nil if an alias not yet selected
 +
selected – the alias we selected or the alias that was previously selected or nil if an alias not yet selected
  
 
]]
 
]]
  
local function format_chapter_title (scriptchapter, chapter, transchapter, chapterurl, chapter_url_source, no_quotes, access)
+
local function is_alias_used (args, alias, index, enumerated, value, selected, error_list)
local chapter_error = '';
+
if enumerated then -- is this a test for an enumerated parameters?
+
alias = alias:gsub ('#', index); -- replace '#' with the value in index
if not is_set (chapter) then
 
chapter = ''; -- to be safe for concatenation
 
 
else
 
else
if false == no_quotes then
+
alias = alias:gsub ('#', ''); -- remove '#' if it exists
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
  
chapter = script_concatenate (chapter, scriptchapter) -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
+
if is_set(args[alias]) then -- alias is in the template's argument list
 
+
if value ~= nil and selected ~= alias then -- if we have already selected one of the aliases
if is_set (transchapter) then
+
local skip;
transchapter = wrap_style ('trans-quoted-title', transchapter);
+
for _, v in ipairs(error_list) do -- spin through the error list to see if we've added this alias
if is_set (chapter) then
+
if v == alias then
chapter = chapter ..  ' ' .. transchapter;
+
skip = true;
else -- here when transchapter without chapter or script-chapter
+
break; -- has been added so stop looking
chapter = transchapter; --  
+
end
chapter_error = ' ' .. set_error ('trans_missing_title', {'chapter'});
+
end
 +
if not skip then -- has not been added so
 +
table.insert( error_list, alias ); -- add error alias to the error list
 +
end
 +
else
 +
value = args[alias]; -- not yet selected an alias, so select this one
 +
selected = alias;
 
end
 
end
 
end
 
end
 +
return value, selected; -- return newly selected alias, or previously selected alias
 +
end
  
if is_set (chapterurl) then
 
chapter = external_link (chapterurl, chapter, chapter_url_source, access); -- adds bare_url_missing_title error if appropriate
 
end
 
  
return chapter .. chapter_error;
+
--[[--------------------------< S E L E C T _ O N E >----------------------------------------------------------
end
 
  
 +
Chooses one matching parameter from a list of parameters to consider.  The list of parameters to consider is just
 +
names.  For parameters that may be enumerated, the position of the numerator in the parameter name is identified
 +
by the '#' so |author-last1= and |author1-last= are represented as 'author-last#' and 'author#-last'.
  
--[[--------------------------< H A S _ I N V I S I B L E _ C H A R S >----------------------------------------
+
Because enumerated parameter |<param>1= is an alias of |<param>= we must test for both possibilities.
  
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.
+
Generates an error if more than one match is present.
  
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().
 
  
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
+
local function select_one( args, aliases_list, error_condition, index )
 +
local value = nil; -- the value assigned to the selected parameter
 +
local selected = ''; -- the name of the parameter we have chosen
 +
local error_list = {};
  
Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker
+
if index ~= nil then index = tostring(index); end
that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the
 
parameter value.
 
  
]]
+
for _, alias in ipairs( aliases_list ) do -- for each alias in the aliases list
 +
if alias:match ('#') then -- if this alias can be enumerated
 +
if '1' == index then -- when index is 1 test for enumerated and non-enumerated aliases
 +
value, selected = is_alias_used (args, alias, index, false, value, selected, error_list); -- first test for non-enumerated alias
 +
end
 +
value, selected = is_alias_used (args, alias, index, true, value, selected, error_list); -- test for enumerated alias
 +
else
 +
value, selected = is_alias_used (args, alias, index, false, value, selected, error_list); --test for non-enumerated alias
 +
end
 +
end
  
local function has_invisible_chars (param, v)
+
if #error_list > 0 and 'none' ~= error_condition then -- for cases where this code is used outside of extract_names()
local position = ''; -- position of invisible char or starting position of stripmarker
+
local error_str = "";
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
+
for _, k in ipairs( error_list ) do
local capture; -- used by stripmarker detection to hold name of the stripmarker
+
if error_str ~= "" then error_str = error_str .. cfg.messages['parameter-separator'] end
local i=1;
+
error_str = error_str .. wrap_style ('parameter', k);
local stripmarker, apostrophe;
+
end
+
if #error_list > 1 then
capture = string.match (v, '[%w%p ]*'); -- Test for values that are simple ASCII text and bypass other tests if true
+
error_str = error_str .. cfg.messages['parameter-final-separator'];
if capture == v then -- if same there are no unicode characters
+
else
return;
+
error_str = error_str .. cfg.messages['parameter-pair-separator'];
end
 
 
 
while cfg.invisible_chars[i] do
 
local char=cfg.invisible_chars[i][1] -- the character or group name
 
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
 
position, dummy, capture = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
 
 
if position and (char == 'zero width joiner') then -- if we found a zero width joiner character
 
if mw.ustring.find (v, cfg.indic_script) then -- its ok if one of the indic scripts
 
position = nil; -- unset position
 
end
 
end
 
 
if position then
 
if 'nowiki' == capture or 'math' == capture then -- nowiki, math stripmarker (not an error condition)
 
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
+
error_str = error_str .. wrap_style ('parameter', selected);
 +
table.insert( z.message_tail, { set_error( error_condition, {error_str}, true ) } );
 
end
 
end
 +
 +
return value, selected;
 
end
 
end
  
  
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
+
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
  
Argument wrapper.  This function provides support for argument mapping defined in the configuration file so that
+
Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
multiple names can be transparently aliased to single internal variable.
+
parameter (chapter_url_source used for error messages).
  
 
]]
 
]]
  
local function argument_wrapper( args )
+
local function format_chapter_title (scriptchapter, chapter, transchapter, chapterurl, chapter_url_source, no_quotes)
local origin = {};
+
local chapter_error = '';
 
 
return setmetatable({
+
if not is_set (chapter) then
ORIGIN = function( self, k )
+
chapter = ''; -- to be safe for concatenation
local dummy = self[k]; --force the variable to be loaded.
+
else
return origin[k];
+
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
 +
 
 +
chapter = script_concatenate (chapter, scriptchapter) -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
 +
 
 +
if is_set (transchapter) then
 +
transchapter = wrap_style ('trans-quoted-title', transchapter);
 +
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
},
+
end
{
+
 
__index = function ( tbl, k )
+
if is_set (chapterurl) then
if origin[k] ~= nil then
+
chapter = external_link (chapterurl, chapter, chapter_url_source); -- adds bare_url_missing_title error if appropriate
return nil;
+
end
end
+
 
+
return chapter .. chapter_error;
local args, list, v = args, cfg.aliases[k];
 
 
if type( list ) == 'table' then
 
v, origin[k] = select_one( args, list, 'redundant_parameters' );
 
if origin[k] == nil then
 
origin[k] = ''; -- Empty string, not nil
 
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
 
 
-- Empty strings, not nil;
 
if v == nil then
 
v = cfg.defaults[k] or '';
 
origin[k] = '';
 
end
 
 
tbl = rawset( tbl, k, v );
 
return v;
 
end,
 
});
 
 
end
 
end
  
 +
--[[--------------------------< H A S _ I N V I S I B L E _ C H A R S >----------------------------------------
  
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
+
This function searches a parameter's value for nonprintable or invisible characters.  The search stops at the
 +
first match.
  
Looks for a parameter's name in one of several whitelists.
+
This function will detect the visible replacement character when it is part of the wikisource.
  
Parameters in the whitelist can have three values:
+
Detects but ignores nowiki and math stripmarkers.  Also detects other named stripmarkers (gallery, math, pre, ref)
true - active, supported parameters
+
and identifies them with a slightly different error message.  See also coins_cleanup().
false - deprecated, supported parameters
 
nil - unsupported parameters
 
 
]]
 
  
local function validate( name, cite_class )
+
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
local name = tostring( name );
+
 
local state;
+
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
if in_array (cite_class, {'arxiv', 'biorxiv', 'citeseerx'}) then -- limited parameter sets allowed for these templates
+
parameter value.
state = whitelist.limited_basic_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
 
  
if 'arxiv' == cite_class then -- basic parameters unique to these templates
+
]]
state = whitelist.arxiv_basic_arguments[name];
 
end
 
if 'biorxiv' == cite_class then
 
state = whitelist.biorxiv_basic_arguments[name];
 
end
 
if 'citeseerx' == cite_class then
 
state = whitelist.citeseerx_basic_arguments[name];
 
end
 
  
if true == state then return true; end -- valid actively supported parameter
+
local function has_invisible_chars (param, v)
if false == state then
+
local position = ''; -- position of invisible char or starting position of stripmarker
deprecated_parameter (name); -- parameter is deprecated but still supported
+
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
return true;
+
local capture; -- used by stripmarker detection to hold name of the stripmarker
end
+
local i=1;
-- limited enumerated parameters list
+
local stripmarker, apostrophe;
name = name:gsub( "%d+", "#" ); -- replace digit(s) with # (last25 becomes last#)
 
state = whitelist.limited_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
 
  
return false; -- not supported because not found or name is set to nil
+
while cfg.invisible_chars[i] do
end -- end limited parameter-set templates
+
local char=cfg.invisible_chars[i][1] -- the character or group name
 +
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
 +
position, dummy, capture = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
 
 
state = whitelist.basic_arguments[ name ]; -- all other templates; all normal parameters allowed
+
if position then
+
if 'nowiki' == capture or 'math' == capture or -- nowiki and math stripmarkers (not an error condition)
if true == state then return true; end -- valid actively supported parameter
+
('templatestyles' == capture) then -- templatestyles stripmarker allowed
if false == state then
+
stripmarker = true; -- set a flag
deprecated_parameter (name); -- parameter is deprecated but still supported
+
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
return true;
+
position = nil; -- unset
end
+
elseif 'apostrophe' == char then -- apostrophe template uses &zwj;, hair space and zero-width space
-- all enumerated parameters allowed
+
apostrophe = true;
name = name:gsub( "%d+", "#" ); -- replace digit(s) with # (last25 becomes last#
+
elseif true == apostrophe and in_array (char, {'zero width joiner', 'zero width space', 'hair space'}) then
state = whitelist.numbered_arguments[ name ];
+
position = nil; -- unset
 
+
else
if true == state then return true; end -- valid actively supported parameter
+
local err_msg;
if false == state then
+
if capture then
deprecated_parameter (name); -- parameter is deprecated but still supported
+
err_msg = capture .. ' ' .. cfg.invisible_chars[i][3] or char;
return true;
+
else
end
+
err_msg = cfg.invisible_chars[i][3] or (char .. ' character');
+
end
return false; -- not supported because not found or name is set to nil
 
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
 +
i=i+1; -- bump our index
 +
end
 +
end
  
--[[--------------------------< N O W R A P _ D A T E >--------------------------------------------------------
 
  
When date is YYYY-MM-DD format wrap in nowrap span: <span ...>YYYY-MM-DD</span>.  When date is DD MMMM YYYY or is
+
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
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.
+
Argument wrapper.  This function provides support for argument mapping defined in the configuration file so that
 +
multiple names can be transparently aliased to single internal variable.
  
 
]]
 
]]
  
local function nowrap_date (date)
+
local function argument_wrapper( args )
local cap='';
+
local origin = {};
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
+
return setmetatable({
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
+
ORIGIN = function( self, k )
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
+
local dummy = self[k]; --force the variable to be loaded.
end
+
return origin[k];
+
end
return date;
+
},
 +
{
 +
__index = function ( tbl, k )
 +
if origin[k] ~= nil then
 +
return nil;
 +
end
 +
 +
local args, list, v = args, cfg.aliases[k];
 +
 +
if type( list ) == 'table' then
 +
v, origin[k] = select_one( args, list, 'redundant_parameters' );
 +
if origin[k] == nil then
 +
origin[k] = ''; -- Empty string, not nil
 +
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
 +
 +
-- Empty strings, not nil;
 +
if v == nil then
 +
v = cfg.defaults[k] or '';
 +
origin[k] = '';
 +
end
 +
 +
tbl = rawset( tbl, k, v );
 +
return v;
 +
end,
 +
});
 
end
 
end
  
 +
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
 +
Looks for a parameter's name in the whitelist.
  
--[[--------------------------< S E T _ T I T L E T Y P E >----------------------------------------------------
+
Parameters in the whitelist can have three values:
 
+
true - active, supported parameters
This function sets default title types (equivalent to the citation including |type=<default value>) for those templates that have defaults.
+
false - deprecated, supported parameters
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
+
nil - unsupported parameters
 
+
 
]]
 
]]
  
local function set_titletype (cite_class, title_type)
+
local function validate( name )
if is_set(title_type) then
+
local name = tostring( name );
if "none" == title_type then
+
local state = whitelist.basic_arguments[ name ];
title_type = ""; -- if |type=none then type parameter not displayed
+
end
+
-- Normal arguments
return title_type; -- if |type= has been set to any other value use that value
+
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
 +
 +
-- Arguments with numbers in them
 +
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
  
return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation
+
 
 +
-- Formats a wiki style internal link
 +
local function internal_link_id(options)
 +
return mw.ustring.format( '[[%s%s%s|\<span title\=\"%s\"\>%s\<\/span\>%s%s]]',
 +
options.prefix, options.id, options.suffix or "",
 +
options.link, options.label, options.separator or "&nbsp;",
 +
mw.text.nowiki(options.id)
 +
);
 
end
 
end
  
  
--[[--------------------------< H Y P H E N _ T O _ D A S H >--------------------------------------------------
+
--[[--------------------------< N O W R A P _ D A T E >--------------------------------------------------------
  
Converts a hyphen to a dash
+
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 hyphen_to_dash( str )
+
local function nowrap_date (date)
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
+
local cap='';
return str;
+
local cap2='';
end
+
 
return str:gsub( '-', ' ' );
+
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$") then
 +
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
 +
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
 +
end
 +
 +
return date;
 +
end
 +
 
 +
--[[--------------------------< IS _ V A L I D _ I S X N >-----------------------------------------------------
 +
 
 +
ISBN-10 and ISSN validator code calculates checksum across all isbn/issn digits including the check digit. ISBN-13 is checked in check_isbn().
 +
If the number is valid the result will be 0. Before calling this function, issbn/issn must be checked for length and stripped of dashes,
 +
spaces and other non-isxn characters.
 +
 
 +
]]
 +
 
 +
local function is_valid_isxn (isxn_str, len)
 +
local temp = 0;
 +
isxn_str = { isxn_str:byte(1, len) }; -- make a table of byte values '0' → 0x30 .. '9'  → 0x39, 'X' → 0x58
 +
len = len+1; -- adjust to be a loop counter
 +
for i, v in ipairs( isxn_str ) do -- loop through all of the bytes and calculate the checksum
 +
if v == string.byte( "X" ) then -- if checkdigit is X (compares the byte value of 'X' which is 0x58)
 +
temp = temp + 10*( len - i ); -- it represents 10 decimal
 +
else
 +
temp = temp + tonumber( string.char(v) )*(len-i);
 +
end
 +
end
 +
return temp % 11 == 0; -- returns true if calculation result is zero
 
end
 
end
  
  
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
+
--[[--------------------------< IS _ V A L I D _ I S X N _ 1 3 >----------------------------------------------
  
Joins a sequence of strings together while checking for duplicate separation characters.
+
ISBN-13 and ISMN validator code calculates checksum across all 13 isbn/ismn digits including the check digit.
 +
If the number is valid, the result will be 0. Before calling this function, isbn-13/ismn must be checked for length
 +
and stripped of dashes, spaces and other non-isxn-13 characters.
  
 
]]
 
]]
  
local function safe_join( tbl, duplicate_char )
+
local function is_valid_isxn_13 (isxn_str)
--[[
+
local temp=0;
Note: we use string functions here, rather than ustring functions.
 
 
 
This has considerably faster performance and should work correctly as
+
isxn_str = { isxn_str:byte(1, 13) }; -- make a table of byte values '0' → 0x30 .. '9'  → 0x39
long as the duplicate_char is strict ASCIIThe strings
+
for i, v in ipairs( isxn_str ) do
in tbl may be ASCII or UTF8.
+
temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) ); -- multiply odd index digits by 1, even index digits by 3 and sum; includes check digit
]]
+
end
 +
return temp % 10 == 0; -- sum modulo 10 is zero when isbn-13/ismn is correct
 +
end
 +
 
 +
--[[--------------------------< C H E C K _ I S B N >------------------------------------------------------------
 +
 
 +
Determines whether an ISBN string is valid
 +
 
 +
]]
 +
 
 +
local function check_isbn( isbn_str )
 +
if nil ~= isbn_str:match("[^%s-0-9X]") then return false; end -- fail if isbn_str contains anything but digits, hyphens, or the uppercase X
 +
isbn_str = isbn_str:gsub( "-", "" ):gsub( " ", "" ); -- remove hyphens and spaces
 +
local len = isbn_str:len();
 +
 +
if len ~= 10 and len ~= 13 then
 +
return false;
 +
end
 +
 
 +
if len == 10 then
 +
if isbn_str:match( "^%d*X?$" ) == nil then return false; end
 +
return is_valid_isxn(isbn_str, 10);
 +
else
 +
local temp = 0;
 +
if isbn_str:match( "^97[89]%d*$" ) == nil then return false; end -- isbn13 begins with 978 or 979; ismn begins with 979
 +
return is_valid_isxn_13 (isbn_str);
 +
end
 +
end
 +
 
 +
--[[--------------------------< C H E C K _ I S M N >------------------------------------------------------------
 +
 
 +
Determines whether an ISMN string is validSimilar to isbn-13, ismn is 13 digits begining 979-0-... and uses the
 +
same check digit calculations.  See http://www.ismn-international.org/download/Web_ISMN_Users_Manual_2008-6.pdf
 +
section 2, pages 9–12.
 +
 
 +
]]
 +
 
 +
local function ismn (id)
 +
local handler = cfg.id_handlers['ISMN'];
 +
local text;
 +
local valid_ismn = true;
 +
 
 +
id=id:gsub( "[%s-–]", "" ); -- strip spaces, hyphens, and endashes from the ismn
 +
 
 +
if 13 ~= id:len() or id:match( "^9790%d*$" ) == nil then -- ismn must be 13 digits and begin 9790
 +
valid_ismn = false;
 +
else
 +
valid_ismn=is_valid_isxn_13 (id); -- validate ismn
 +
end
 +
 
 +
-- text = internal_link_id({link = handler.link, label = handler.label, -- use this (or external version) when there is some place to link to
 +
-- prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 +
 +
text="[[" .. handler.link .. "|" .. handler.label .. "]]" .. handler.separator .. id; -- because no place to link to yet
 +
 
 +
if false == valid_ismn then
 +
text = text .. ' ' .. set_error( 'bad_ismn' ) -- add an error message if the issn is invalid
 +
end
 +
 +
return text;
 +
end
 +
 
 +
--[[--------------------------< I S S N >----------------------------------------------------------------------
 +
 
 +
Validate and format an issn.  This code fixes the case where an editor has included an ISSN in the citation but has separated the two groups of four
 +
digits with a space.  When that condition occurred, the resulting link looked like this:
 +
 
 +
|issn=0819 4327 gives: [http://www.worldcat.org/issn/0819 4327 0819 4327]  -- can't have spaces in an external link
 
 
local str = ''; -- the output string
+
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
local comp = ''; -- what does 'comp' mean?
+
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
local end_chr = '';
+
error message.  The issn is always displayed with a hyphen, even if the issn was given as a single group of 8 digits.
local trim;
+
 
for _, value in ipairs( tbl ) do
+
]]
if value == nil then value = ''; end
+
 
+
local function issn(id, e)
if str == '' then -- if output string is empty
+
local issn_copy = id; -- save a copy of unadulterated issn; use this version for display if issn does not validate
str = value; -- assign value to it (first time through the loop)
+
local handler;
elseif value ~= '' then
+
local text;
if value:sub(1,1) == '<' then -- Special case of values enclosed in spans and other markup.
+
local valid_issn = true;
comp = value:gsub( "%b<>", "" ); -- remove html markup (<span>string</span> -> string)
+
else
+
if e then
comp = value;
+
handler = cfg.id_handlers['EISSN'];
 +
else
 +
handler = cfg.id_handlers['ISSN'];
 +
end
 +
 
 +
id=id:gsub( "[%s-–]", "" ); -- strip spaces, hyphens, and endashes from the issn
 +
 
 +
if 8 ~= id:len() or nil == id:match( "^%d*X?$" ) then -- validate the issn: 8 digits long, containing only 0-9 or X in the last position
 +
valid_issn=false; -- wrong length or improper character
 +
else
 +
valid_issn=is_valid_isxn(id, 8); -- validate issn
 +
end
 +
 
 +
if true == valid_issn then
 +
id = string.sub( id, 1, 4 ) .. "-" .. string.sub( id, 5 ); -- if valid, display correctly formatted version
 +
else
 +
id = issn_copy; -- if not valid, use the show the invalid issn with error message
 +
end
 +
 +
text = external_link_id({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 +
 +
if false == valid_issn then
 +
text = text .. ' ' .. set_error( 'bad_issn' ) -- add an error message if the issn is invalid
 +
end
 +
 +
return text
 +
end
 +
 
 +
--[[--------------------------< A M A Z O N >------------------------------------------------------------------
 +
 
 +
Formats a link to Amazon.  Do simple error checking: asin must be mix of 10 numeric or uppercase alpha
 +
characters.  If a mix, first character must be uppercase alpha; if all numeric, asins must be 10-digit
 +
isbn. If 10-digit isbn, add a maintenance category so a bot or awb script can replace |asin= with |isbn=.
 +
Error message if not 10 characters, if not isbn10, if mixed and first character is a digit.
 +
 
 +
]]
 +
 
 +
local function amazon(id, domain)
 +
local err_cat = ""
 +
 
 +
if not id:match("^[%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u]$") then
 +
err_cat = ' ' .. set_error ('bad_asin'); -- asin is not a mix of 10 uppercase alpha and numeric characters
 +
else
 +
if id:match("^%d%d%d%d%d%d%d%d%d[%dX]$") then -- if 10-digit numeric (or 9 digits with terminal X)
 +
if check_isbn( id ) then -- see if asin value is isbn10
 +
add_maint_cat ('ASIN');
 +
elseif not is_set (err_cat) then
 +
err_cat = ' ' .. set_error ('bad_asin'); -- asin is not isbn10
 
end
 
end
-- typically duplicate_char is sepc
+
elseif not id:match("^%u[%d%u]+$") then
if comp:sub(1,1) == duplicate_char then -- is first charactier same as duplicate_char? why test first character?
+
err_cat =  ' ' .. set_error ('bad_asin'); -- asin doesn't begin with uppercase alpha
--  Because individual string segments often (always?) begin with terminal punct for th
+
end
--  preceding segment: 'First element' .. 'sepc next element' .. etc?
+
end
trim = false;
+
if not is_set(domain) then
end_chr = str:sub(-1,-1); -- get the last character of the output string
+
domain = "com";
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
+
elseif in_array (domain, {'jp', 'uk'}) then -- Japan, United Kingdom
if end_chr == duplicate_char then -- if same as separator
+
domain = "co." .. domain;
str = str:sub(1,-2); -- remove it
+
elseif in_array (domain, {'au', 'br', 'mx'}) then -- Australia, Brazil, Mexico
elseif end_chr == "'" then -- if it might be wikimarkup
+
domain = "com." .. domain;
if str:sub(-3,-1) == duplicate_char .. "''" then -- if last three chars of str are sepc''
+
end
str = str:sub(1, -4) .. "''"; -- remove them and add back ''
+
local handler = cfg.id_handlers['ASIN'];
elseif str:sub(-5,-1) == duplicate_char .. "]]''" then -- if last five chars of str are sepc]]''  
+
return external_link_id({link=handler.link,
trim = true; -- why? why do this and next differently from previous?
+
label=handler.label, prefix=handler.prefix .. domain .. "/dp/",
elseif str:sub(-4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''
+
id=id, encode=handler.encode, separator = handler.separator}) .. err_cat;
trim = true; -- same question
+
end
end
+
 
elseif end_chr == "]" then -- if it might be wikimarkup
+
--[[--------------------------< A R X I V >--------------------------------------------------------------------
if str:sub(-3,-1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink
+
 
trim = true;
+
See: http://arxiv.org/help/arxiv_identifier
elseif str:sub(-3,-1) == duplicate_char .. '"]' then -- if last three chars of str are sepc"] quoted external link  
 
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
+
format and error check arXiv identifier.  There are three valid forms of the identifier:
if value ~= comp then -- value does not equal comp when value contains html markup
+
the first form, valid only between date codes 9108 and 0703 is:
local dup2 = duplicate_char;
+
arXiv:<archive>.<class>/<date code><number><version>
if dup2:match( "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it
+
where:
+
<archive> is a string of alpha characters - may be hyphenated; no other punctuation
value = value:gsub( "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows html markup
+
<class> is a string of alpha characters - may be hyphenated; no other punctuation
else
+
<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
value = value:sub( 2, -1 ); -- remove duplicate_char when it is first character
+
first digit of YY for this form can only 9 and 0
end
+
<number> is a three-digit number
end
+
<version> is a 1 or more digit number preceded with a lowercase v; no spaces (undocumented)
end
+
str = str .. value; --add it to the output string
+
the second form, valid from April 2007 through December 2014 is:
 +
arXiv:<date code>.<number><version>
 +
where:
 +
<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
 +
<number> is a four-digit number
 +
<version> is a 1 or more digit number preceded with a lowercase v; no spaces
 +
 
 +
the third form, valid from January 2015 is:
 +
arXiv:<date code>.<number><version>
 +
where:
 +
<date code> and <version> are as defined for 0704-1412
 +
<number> is a five-digit number
 +
]]
 +
 
 +
local function arxiv (id, class)
 +
local handler = cfg.id_handlers['ARXIV'];
 +
local year, month, version;
 +
local err_cat = '';
 +
local text;
 +
 +
if id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%d$") or id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%dv%d+$") then -- test for the 9108-0703 format w/ & w/o version
 +
year, month = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d[v%d]*$");
 +
year = tonumber(year);
 +
month = tonumber(month);
 +
if ((not (90 < year or 8 > year)) or (1 > month or 12 < month)) or -- if invalid year or invalid month
 +
((91 == year and 7 > month) or (7 == year and 3 < month)) then -- if years ok, are starting and ending months ok?
 +
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
 +
end
 +
elseif id:match("^%d%d[01]%d%.%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%dv%d+$") then -- test for the 0704-1412 w/ & w/o version
 +
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d[v%d]*$");
 +
year = tonumber(year);
 +
month = tonumber(month);
 +
if ((7 > year) or (14 < year) or (1 > month or 12 < month)) or -- is year invalid or is month invalid? (doesn't test for future years)
 +
((7 == year) and (4 > month)) then --or -- when year is 07, is month invalid (before April)?
 +
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
 +
end
 +
elseif id:match("^%d%d[01]%d%.%d%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%d%dv%d+$") then -- test for the 1501- format w/ & w/o version
 +
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d%d[v%d]*$");
 +
year = tonumber(year);
 +
month = tonumber(month);
 +
if ((15 > year) or (1 > month or 12 < month)) then -- is year invalid or is month invalid? (doesn't test for future years)
 +
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
 
end
 
end
 +
else
 +
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- arXiv id doesn't match any format
 
end
 
end
return str;
 
end 
 
  
 +
text = external_link_id({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
  
--[[--------------------------< I S _ S U F F I X >------------------------------------------------------------
+
if is_set (class) then
 +
class = ' [[' .. '//arxiv.org/archive/' .. class .. ' ' .. class .. ']]'; -- external link within square brackets, not wikilink
 +
else
 +
class = ''; -- empty string for concatenation
 +
end
 +
 +
return text .. class;
 +
end
  
returns true is suffix is properly formed Jr, Sr, or ordinal in the range 2 –9. Puncutation not allowed.
+
--[[
 +
lccn normalization (http://www.loc.gov/marc/lccn-namespace.html#normalization)
 +
1. Remove all blanks.
 +
2. If there is a forward slash (/) in the string, remove it, and remove all characters to the right of the forward slash.
 +
3. If there is a hyphen in the string:
 +
a. Remove it.
 +
b. Inspect the substring following (to the right of) the (removed) hyphen. Then (and assuming that steps 1 and 2 have been carried out):
 +
1. All these characters should be digits, and there should be six or less. (not done in this function)
 +
2. If the length of the substring is less than 6, left-fill the substring with zeroes until the length is six.
  
 +
Returns a normalized lccn for lccn() to validate.  There is no error checking (step 3.b.1) performed in this function.
 
]]
 
]]
  
local function is_suffix (suffix)
+
local function normalize_lccn (lccn)
if in_array (suffix, {'Jr', 'Sr', '2nd', '3rd'}) or suffix:match ('^%dth$') then
+
lccn = lccn:gsub ("%s", ""); -- 1. strip whitespace
return true;
+
 
 +
if nil ~= string.find (lccn,'/') then
 +
lccn = lccn:match ("(.-)/"); -- 2. remove forward slash and all character to the right of it
 
end
 
end
return false;
 
end
 
  
 +
local prefix
 +
local suffix
 +
prefix, suffix = lccn:match ("(.+)%-(.+)"); -- 3.a remove hyphen by splitting the string into prefix and suffix
 +
 +
if nil ~= suffix then -- if there was a hyphen
 +
suffix=string.rep("0", 6-string.len (suffix)) .. suffix; -- 3.b.2 left fill the suffix with 0s if suffix length less than 6
 +
lccn=prefix..suffix; -- reassemble the lccn
 +
end
 +
 +
return lccn;
 +
end
  
--[[--------------------------< I S _ G O O D _ V A N C _ N A M E >--------------------------------------------
+
--[[
 +
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/
  
For Vancouver Style, author/editor names are supposed to be rendered in Latin (read ASCII) characters.  When a name
+
length = 8 then all digits
uses characters that contain diacritical marks, those characters are to converted to the corresponding Latin character.
+
length = 9 then lccn[1] is lower case alpha
When a name is written using a non-Latin alphabet or logogram, that name is to be transliterated into Latin characters.
+
length = 10 then lccn[1] and lccn[2] are both lower case alpha or both digits
These things are not currently possible in this module so are left to the editor to do.
+
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
  
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
 
[http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] 00C0–00D6, 00D8–00F6, 00F8–00FF
 
[http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] 0100–017F
 
[http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] 0180–01BF, 01C4–024F
 
  
|lastn= also allowed to contain hyphens, spaces, and apostrophes. (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
+
local function lccn(lccn)
|firstn= also allowed to contain hyphens, spaces, apostrophes, and periods
+
local handler = cfg.id_handlers['LCCN'];
 +
local err_cat =  ''; -- presume that LCCN is valid
 +
local id = lccn; -- local copy of the lccn
  
This original test:
+
id = normalize_lccn (id); -- get canonical form (no whitespace, hyphens, forward slashes)
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
+
local len = id:len(); -- get the length of the lccn
was written ouside of the code editor and pasted here because the code editor gets confused between character insertion point and cursor position.
 
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
 
to use an external editor to maintain this code.
 
  
\195\128-\195\150 – À- Ö (U+00C0–U+00D6 – C0 controls)
+
if 8 == len then
\195\152-\195\182 – Ø- ö (U+00D8-U+00F6 – C0 controls)
+
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits)
\195\184-\198\191 – ø- ƿ (U+00F8-U+01BF – C0 controls, Latin extended A & B)
+
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
\199\132-\201\143 – DŽ- ɏ (U+01C4-U+024F – Latin extended B)
+
end
 +
elseif 9 == len then -- LCCN should be adddddddd
 +
if nil == id:match("%l%d%d%d%d%d%d%d%d") then -- does it match our pattern?
 +
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
 +
end
 +
elseif 10 == len then -- LCCN should be aadddddddd or dddddddddd
 +
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits) ...
 +
if nil == id:match("^%l%l%d%d%d%d%d%d%d%d") then -- ... see if it matches our pattern
 +
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 +
end
 +
end
 +
elseif 11 == len then -- LCCN should be aaadddddddd or adddddddddd
 +
if not (id:match("^%l%l%l%d%d%d%d%d%d%d%d") or id:match("^%l%d%d%d%d%d%d%d%d%d%d")) then -- see if it matches one of our patterns
 +
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 +
end
 +
elseif 12 == len then -- LCCN should be aadddddddddd
 +
if not id:match("^%l%l%d%d%d%d%d%d%d%d%d%d") then -- see if it matches our pattern
 +
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 +
end
 +
else
 +
err_cat = ' ' .. set_error( 'bad_lccn' ); -- wrong length, set an error message
 +
end
  
]]
+
if not is_set (err_cat) and nil ~= lccn:find ('%s') then
 
+
err_cat = ' ' .. set_error( 'bad_lccn' ); -- lccn contains a space, set an error message
local function is_good_vanc_name (last, first)
+
end
local first, suffix = first:match ('(.-),?%s*([%dJS][%drndth]+)%.?$') or first; -- if first has something that looks like a generational suffix, get it
 
  
if is_set (suffix) then
+
return external_link_id({link = handler.link, label = handler.label,
if not is_suffix (suffix) then
+
prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
add_vanc_error ('suffix');
 
return false; -- not a name with an appropriate suffix
 
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
 
add_vanc_error ('non-Latin character');
 
return false; -- not a string of latin characters; Vancouver requires Romanization
 
end;
 
return true;
 
 
end
 
end
  
 +
--[[--------------------------< P M I D >----------------------------------------------------------------------
  
--[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------
+
Format PMID and do simple error checking.  PMIDs are sequential numbers beginning at 1 and counting up.  This
 +
code checks the PMID to see that it contains only digits and is less than test_limit; the value in local variable
 +
test_limit will need to be updated periodically as more PMIDs are issued.
  
Attempts to convert names to initials in support of |name-list-format=vanc. 
+
]]
  
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/.
+
local function pmid(id)
 +
local test_limit = 32000000; -- update this value as PMIDs approach
 +
local handler = cfg.id_handlers['PMID'];
 +
local err_cat = ''; -- presume that PMID is valid
 +
 +
if id:match("[^%d]") then -- if PMID has anything but digits
 +
err_cat = ' ' .. set_error( 'bad_pmid' ); -- set an error message
 +
else -- PMID is only digits
 +
local id_num = tonumber(id); -- convert id to a number for range testing
 +
if 1 > id_num or test_limit < id_num then -- if PMID is outside test limit boundaries
 +
err_cat = ' ' .. set_error( 'bad_pmid' ); -- set an error message
 +
end
 +
end
 +
 +
return external_link_id({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 +
end
  
Vancouver style requires family rank designations (Jr, II, III, etc) to be rendered as Jr, 2nd, 3rd, etc.  See http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35085/.
+
--[[--------------------------< I S _ E M B A R G O E D >------------------------------------------------------
This code only accepts and understands generational suffix in the Vancouver format because Roman numerals look like, and can be mistaken for, initials.
 
  
This function uses ustring functions because firstname initials may be any of the unicode Latin characters accepted by is_good_vanc_name ().
+
Determines if a PMC identifier's online version is embargoed. Compares the date in |embargo= against today's date.  If embargo date is
 +
in the future, returns the content of |embargo=; otherwise, returns and empty string because the embargo has expired or because
 +
|embargo= was not set in this cite.
  
 
]]
 
]]
  
local function reduce_to_initials(first)
+
local function is_embargoed (embargo)
local name, suffix = mw.ustring.match(first, "^(%u+) ([%dJS][%drndth]+)$");
+
if is_set (embargo) then
 
+
local lang = mw.getContentLanguage();
if not name then -- if not initials and a suffix
+
local good1, embargo_date, good2, todays_date;
name = mw.ustring.match(first, "^(%u+)$"); -- is it just intials?
+
good1, embargo_date = pcall( lang.formatDate, lang, 'U', embargo );
end
+
good2, todays_date = pcall( lang.formatDate, lang, 'U' );
 
+
if name then -- if first is initials with or without suffix
+
if good1 and good2 then -- if embargo date and today's date are good dates
if 3 > mw.ustring.len (name) then -- if one or two initials
+
if tonumber( embargo_date ) >= tonumber( todays_date ) then -- is embargo date is in the future?
if suffix then -- if there is a suffix
+
return embargo; -- still embargoed
if is_suffix (suffix) then -- is it legitimate?
 
return first; -- one or two initials and a valid suffix so nothing to do
 
else
 
add_vanc_error ('suffix'); -- one or two initials with invalid suffix so error message
 
return first; -- and return first unmolested
 
end
 
 
else
 
else
return first; -- one or two initials without suffix; nothing to do
+
add_maint_cat ('embargo')
 +
return ''; -- unset because embargo has expired
 
end
 
end
 
end
 
end
end -- if here then name has 3 or more uppercase letters so treat them as a word
+
end
 +
return ''; -- |embargo= not set return empty string
 +
end
 +
 
 +
--[[--------------------------< P M C >------------------------------------------------------------------------
 +
 
 +
Format a PMC, do simple error checking, and check for embargoed articles.
 +
 
 +
The embargo parameter takes a date for a value. If the embargo date is in the future the PMC identifier will not
 +
be linked to the article.  If the embargo date is today or in the past, or if it is empty or omitted, then the
 +
PMC identifier is linked to the article through the link at cfg.id_handlers['PMC'].prefix.
  
 +
PMC embargo date testing is done in function is_embargoed () which is called earlier because when the citation
 +
has |pmc=<value> but does not have a |url= then |title= is linked with the PMC link.  Function is_embargoed ()
 +
returns the embargo date if the PMC article is still embargoed, otherwise it returns an empty string.
  
local initials, names = {}, {}; -- tables to hold name parts and initials
+
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
local i = 1; -- counter for number of initials
+
than test_limit; the value in local variable test_limit will need to be updated periodically as more PMCs are issued.
  
names = mw.text.split (first, '[%s,]+'); -- split into a table of names and possible suffix
+
]]
  
while names[i] do -- loop through the table
+
local function pmc(id, embargo)
if 1 < i and names[i]:match ('[%dJS][%drndth]+%.?$') then -- if not the first name, and looks like a suffix (may have trailing dot)
+
local test_limit = 7000000; -- update this value as PMCs approach
names[i] = names[i]:gsub ('%.', ''); -- remove terminal dot if present
+
local handler = cfg.id_handlers['PMC'];
if is_suffix (names[i]) then -- if a legitimate suffix
+
local err_cat = ''; -- presume that PMC is valid
table.insert (initials, ' ' .. names[i]); -- add a separator space, insert at end of initials table
+
local id_num;
break; -- and done because suffix must fall at the end of a name
+
local text;
end -- no error message if not a suffix; possibly because of Romanization
+
 +
id_num = id:match ('^[Pp][Mm][Cc](%d+)$'); -- identifier with pmc prefix
 +
 
 +
if is_set (id_num) then
 +
add_maint_cat ('pmc_format');
 +
else -- plain number without pmc prefix
 +
id_num = id:match ('^%d+$'); -- if here id is all digits
 +
end
 +
 
 +
if is_set (id_num) then -- id_num has a value so test it
 +
id_num = tonumber(id_num); -- convert id_num to a number for range testing
 +
if 1 > id_num or test_limit < id_num then -- if PMC is outside test limit boundaries
 +
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
 +
else
 +
id = tostring (id_num); -- make sure id is a string
 
end
 
end
if 3 > i then
+
else -- when id format incorrect
table.insert (initials, mw.ustring.sub(names[i],1,1)); -- insert the intial at end of initials table
+
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
end
+
end
i = i+1; -- bump the counter
+
 +
if is_set (embargo) then -- is PMC is still embargoed?
 +
text = table.concat ( -- still embargoed so no external link
 +
{
 +
make_wikilink (handler.link, handler.label),
 +
handler.separator,
 +
id,
 +
err_cat
 +
});
 +
else
 +
text = external_link_id({link = handler.link, label = handler.label, -- no embargo date or embargo has expired, ok to link to article
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 
end
 
end
+
return text;
return table.concat(initials) -- Vancouver format does not include spaces.
 
 
end
 
end
  
 +
--[[--------------------------< D O I >------------------------------------------------------------------------
 +
 +
Formats a DOI and checks for DOI errors.
 +
 +
DOI names contain two parts: prefix and suffix separated by a forward slash.
 +
Prefix: directory indicator '10.' followed by a registrant code
 +
Suffix: character string of any length chosen by the registrant
  
--[[--------------------------< L I S T _ P E O P L E >-------------------------------------------------------
+
This function checks a DOI name for: prefix/suffix. If the doi name contains spaces or endashes, or, if it ends
 +
with a period or a comma, this function will emit a bad_doi error message.
  
Formats a list of people (e.g. authors / editors)
+
DOI names are case-insensitive and can incorporate any printable Unicode characters so the test for spaces, endash,
 +
and terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely
 +
if ever used in doi names.
  
 
]]
 
]]
  
local function list_people(control, people, etal)
+
local function doi(id, inactive)
local sep;
+
local cat = ""
local namesep;
+
local handler = cfg.id_handlers['DOI'];
local format = control.format
+
local maximum = control.maximum
+
local text;
local lastauthoramp = control.lastauthoramp;
+
if is_set(inactive) then
local text = {}
+
local inactive_year = inactive:match("%d%d%d%d") or ''; -- try to get the year portion from the inactive date
 +
if is_set(inactive_year) then
 +
table.insert( z.error_categories, "自" .. inactive_year .. "年含有不活躍DOI的頁面" );
 +
else
 +
table.insert( z.error_categories, "含有不活躍DOI的頁面" );                 -- when inactive doesn't contain a recognizable year
 +
end
 +
inactive = " (" .. cfg.messages['inactive'] .. " " .. inactive .. ")"
 +
end
 +
text = external_link_id({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. (inactive or '')
  
if 'vanc' == format then -- Vancouver-like author/editor name styling?
+
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
sep = ','; -- name-list separator between authors is a comma
+
cat = ' ' .. set_error( 'bad_doi' );
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 text .. cat
 +
end
 +
 +
 +
--[[--------------------------< H D L >------------------------------------------------------------------------
 +
 +
Formats an HDL with minor error checking.
 +
 +
HDL names contain two parts: prefix and suffix separated by a forward slash.
 +
Prefix: character string using any character in the UCS-2 character set except '/'
 +
Suffix: character string of any length using any character in the UCS-2 character set chosen by the registrant
 +
 +
This function checks a HDL name for: prefix/suffix.  If the HDL name contains spaces, endashes, or, if it ends
 +
with a period or a comma, this function will emit a bad_hdl error message.
 +
 +
HDL names are case-insensitive and can incorporate any printable Unicode characters so the test for endashes and
 +
terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely
 +
if ever used in HDLs.
 +
 +
]]
 +
 +
local function hdl(id)
 +
local handler = cfg.id_handlers['HDL'];
 
 
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
+
local text = external_link_id({link = handler.link, label = handler.label,
if is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for authors
+
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
+
 
for i,person in ipairs(people) do
+
if nil == id:match("^[^%s–]-/[^%s–]-[^%.,]$") then -- hdl must contain a fwd slash, must not contain spaces, endashes, and must not end with period or comma
if is_set(person.last) then
+
text = text .. ' ' .. set_error( 'bad_hdl' );
local mask = person.mask
+
end
local one
+
return text;
local sep_one = sep;
+
end
if is_set (maximum) and i > maximum then
+
 
etal = true;
+
--[[--------------------------< O P E N L I B R A R Y >--------------------------------------------------------
break;
+
 
elseif (mask ~= nil) then
+
Formats an OpenLibrary link, and checks for associated errors.
local n = tonumber(mask)
+
 
if (n ~= nil) then
+
]]
one = string.rep("&mdash;",n)
+
local function openlibrary(id)
else
+
local code = id:match("^%d+([AMW])$"); -- only digits followed by 'A', 'M', or 'W'
one = mask;
+
local handler = cfg.id_handlers['OL'];
sep_one = " ";
+
 
end
+
if ( code == "A" ) then
else
+
return external_link_id({link=handler.link, label=handler.label,
one = person.last
+
prefix=handler.prefix .. 'authors/OL',
local first = person.first
+
id=id, separator=handler.separator, encode = handler.encode})
if is_set(first) then
+
elseif ( code == "M" ) then
if ( "vanc" == format ) then -- if vancouver format
+
return external_link_id({link=handler.link, label=handler.label,
one = one:gsub ('%.', ''); -- remove periods from surnames (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
+
prefix=handler.prefix .. 'books/OL',
if not person.corporate and is_good_vanc_name (one, first) then -- and name is all Latin characters; corporate authors not tested
+
id=id, separator=handler.separator, encode = handler.encode})
first = reduce_to_initials(first) -- attempt to convert first name(s) to initials
+
elseif ( code == "W" ) then
end
+
return external_link_id({link=handler.link, label=handler.label,
end
+
prefix=handler.prefix .. 'works/OL',
one = one .. namesep .. first;
+
id=id, separator=handler.separator, encode = handler.encode})
end
+
else
if is_set(person.link) and person.link ~= control.page_name then
+
return external_link_id({link=handler.link, label=handler.label,
one = make_wikilink (person.link, one); -- link author/editor if this page is not the author's/editor's page
+
prefix=handler.prefix .. 'OL',
end
+
id=id, separator=handler.separator, encode = handler.encode}) .. ' ' .. set_error( 'bad_ol' );
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
+
 
if count > 0 then  
+
--[[--------------------------< M E S S A G E _ I D >----------------------------------------------------------
if count > 1 and is_set(lastauthoramp) and not etal then
+
 
text[#text-2] = " & "; -- replace last separator with ampersand text
+
Validate and format a usenet message id.  Simple error checking, looks for 'id-left@id-right' not enclosed in
 +
'<' and/or '>' angle brackets.
 +
 
 +
]]
 +
 
 +
local function message_id (id)
 +
local handler = cfg.id_handlers['USENETID'];
 +
 
 +
text = external_link_id({link = handler.link, label = handler.label,
 +
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 +
 +
if not id:match('^.+@.+$') or not id:match('^[^<].*[^>]$')then -- doesn't have '@' or has one or first or last character is '< or '>'
 +
text = text .. ' ' .. set_error( 'bad_message_id' ) -- add an error message if the message id is invalid
 +
end
 +
 +
return text
 +
end
 +
 
 +
--[[--------------------------< S E T _ T I T L E T Y P E >----------------------------------------------------
 +
 
 +
This function sets default title types (equivalent to the citation including |type=<default value>) for those templates that have defaults.
 +
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
 +
 
 +
]]
 +
 
 +
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
text[#text] = nil; -- erase the last separator
+
return title_type; -- if |type= has been set to any other value use that value
 
end
 
end
+
 
local result = table.concat(text) -- construct list
+
return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation
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
  
 +
--[[--------------------------< C L E A N _ I S B N >----------------------------------------------------------
  
--[[--------------------------< A N C H O R _ I D >------------------------------------------------------------
+
Removes irrelevant text and dashes from ISBN number
 +
Similar to that used for Special:BookSources
 +
 
 +
]]
  
Generates a CITEREF anchor ID if we have at least one name or a date.  Otherwise returns an empty string.
+
local function clean_isbn( isbn_str )
 +
return isbn_str:gsub( "[^-0-9X]", "" );
 +
end
  
namelist is one of the contributor-, author-, or editor-name lists chosen in that order.  year is Year or anchor_year.
+
--[[--------------------------< E S C A P E _ L U A _ M A G I C _ C H A R S >----------------------------------
  
 +
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.
 
]]
 
]]
local function anchor_id (namelist, year)
+
local function escape_lua_magic_chars (argument)
local names={}; -- a table for the one to four names and year
+
argument = argument:gsub("%%", "%%%%"); -- replace % with %%
for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names
+
argument = argument:gsub("([%^%$%(%)%.%[%]%*%+%-%?])", "%%%1"); -- replace all other lua magic pattern characters
names[i] = v.last
+
return argument;
if i == 4 then break end -- if four then done
+
end
 +
 
 +
--[[--------------------------< S T R I P _ A P O S T R O P H E _ M A R K U P >--------------------------------
 +
 
 +
Strip wiki italic and bold markup from argument so that it doesn't contaminate COinS metadata.
 +
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.
 +
 
 +
]]
 +
 
 +
local function strip_apostrophe_markup (argument)
 +
if not is_set (argument) then return argument; end
 +
 
 +
while true do
 +
if argument:match ("%'%'%'%'%'") then -- bold italic (5)
 +
argument=argument:gsub("%'%'%'%'%'", ""); -- remove all instances of it
 +
elseif argument:match ("%'%'%'%'") then -- italic start and end without content (4)
 +
argument=argument:gsub("%'%'%'%'", "");
 +
elseif argument:match ("%'%'%'") then -- bold (3)
 +
argument=argument:gsub("%'%'%'", "");
 +
elseif argument:match ("%'%'") then -- italic (2)
 +
argument=argument:gsub("%'%'", "");
 +
else
 +
break;
 +
end
 +
end
 +
return argument; -- done
 +
end
 +
 
 +
--[[--------------------------< M A K E _ C O I N S _ T I T L E >----------------------------------------------
 +
 
 +
Makes a title for COinS from Title and / or ScriptTitle (or any other name-script pairs)
 +
 
 +
Apostrophe markup (bold, italics) is stripped from each value so that the COinS metadata isn't correupted with strings
 +
of %27%27...
 +
]]
 +
 
 +
local function make_coins_title (title, script)
 +
if is_set (title) then
 +
title = strip_apostrophe_markup (title); -- strip any apostrophe markup
 +
else
 +
title=''; -- if not set, make sure title is an empty string
 
end
 
end
table.insert (names, year); -- add the year at the end
+
if is_set (script) then
local id = table.concat(names); -- concatenate names and year for CITEREF id
+
script = script:gsub ('^%l%l%s*:%s*', ''); -- remove language prefix if present (script value may now be empty string)
if is_set (id) then -- if concatenation is not an empty string
+
script = strip_apostrophe_markup (script); -- strip any apostrophe markup
return "CITEREF" .. id; -- add the CITEREF portion
 
 
else
 
else
return ''; -- return an empty string; no reason to include CITEREF id in this citation
+
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
 
end
 +
return title .. script; -- return the concatenation
 
end
 
end
  
 +
--[[--------------------------< G E T _ C O I N S _ P A G E S >------------------------------------------------
  
--[[--------------------------< N A M E _ H A S _ E T A L >----------------------------------------------------
+
Extract page numbers from external wikilinks in any of the |page=, |pages=, or |at= parameters for use in COinS.
 
 
Evaluates the content of author and editor name parameters for variations on the theme of et al.  If found,
 
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 parameters |display-authors=etal or |display-editors=etal
 
  
 
]]
 
]]
  
local function name_has_etal (name, etal, nocat)
+
local function get_coins_pages (pages)
 +
local pattern;
 +
if not is_set (pages) then return pages; end -- if no page numbers then we're done
 +
 +
while true do
 +
pattern = pages:match("%[(%w*:?//[^ ]+%s+)[%w%d].*%]"); -- pattern is the opening bracket, the url and following space(s): "[url "
 +
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
 +
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
  
if is_set (name) then -- name can be nil in which case just return
+
-- Gets the display text for a wikilink like [[A|B]] or [[B]] gives B
local etal_pattern = "[;,]? *[\"']*%f[%a][Ee][Tt] *[Aa][Ll][%.\"']*$" -- variations on the 'et al' theme
+
local function remove_wiki_link( str )
local others_pattern = "[;,]? *%f[%a]and [Oo]thers"; -- and alternate to et al.
+
return (str:gsub( "%[%[([^%[%]]*)%]%]", function(l)
+
return l:gsub( "^[^|]*|(.*)$", "%1" ):gsub("^%s*(.-)%s*$", "%1");
if name:match (etal_pattern) then -- variants on et al.
+
end));
name = name:gsub (etal_pattern, ''); -- if found, remove
 
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
 
if not nocat then -- no categorization for |vauthors=
 
add_maint_cat ('etal'); -- and add a category if not already added
 
end
 
elseif name:match (others_pattern) then -- if not 'et al.', then 'and others'?
 
name = name:gsub (others_pattern, ''); -- if found, remove
 
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
 
if not nocat then -- no categorization for |vauthors=
 
add_maint_cat ('etal'); -- and add a category if not already added
 
end
 
end
 
end
 
return name, etal; --
 
 
end
 
end
  
 +
-- Converts a hyphen to a dash
 +
local function hyphen_to_dash( str )
 +
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
 +
return str;
 +
end
 +
return str:gsub( '-', '–' );
 +
end
  
--[[--------------------------< N A M E _ H A S _ E D _ M A R K U P >------------------------------------------
+
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
  
Evaluates the content of author and editor parameters for extranious editor annotations: ed, ed., eds, (Ed.), etc.
+
Joins a sequence of strings together while checking for duplicate separation characters.
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 function safe_join( tbl, duplicate_char )
local _, pattern;
+
--[[
local patterns = { -- these patterns match annotations at end of name
+
Note: we use string functions here, rather than ustring functions.
'%f[%(%[][%(%[]%s*[Ee][Dd][Ss]?%.?%s*[%)%]]?$', -- (ed) or (eds): leading '(', case insensitive 'ed', optional 's', '.' and/or ')'
+
'[,%.%s]%f[e]eds?%.?$', -- ed or eds: without '('or ')'; case sensitive (ED could be initials Ed could be name)
+
This has considerably faster performance and should work correctly as
'%f[%(%[][%(%[]%s*[Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%.?%s*[%)%]]?$', -- (editor) or (editors): leading '(', case insensitive, optional '.' and/or ')'
+
long as the duplicate_char is strict ASCII. The strings
'[,%.%s]%f[Ee][Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%.?$', -- editor or editors: without '('or ')'; case insensitive
+
in tbl may be ASCII or UTF8.
+
]]
-- these patterns match annotations at beginning of name
+
'^eds?[%.,;]', -- ed. or eds.: lower case only, optional 's', requires '.'
+
local str = ''; -- the output string
'^[%(%[]%s*[Ee][Dd][Ss]?%.?%s*[%)%]]', -- (ed) or (eds): also sqare brackets, case insensitive, optional 's', '.'
+
local comp = ''; -- what does 'comp' mean?
'^[%(%[]?%s*[Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%A', -- (editor or (editors: also sq brackets, case insensitive, optional brackets, 's'
+
local end_chr = '';
'^[%(%[]?%s*[Ee][Dd][Ii][Tt][Ee][Dd]%A', -- (edited: also sq brackets, case insensitive, optional brackets
+
local trim;
}
+
for _, value in ipairs( tbl ) do
 
+
if value == nil then value = ''; end
if is_set (name) then
+
for _, pattern in ipairs (patterns) do -- spin through patterns table and
+
if str == '' then -- if output string is empty
if name:match (pattern) then
+
str = value; -- assign value to it (first time through the loop)
add_maint_cat ('extra_text_names', cfg.special_case_translation [list_name]); -- add a maint cat for this template
+
elseif value ~= '' then
break;
+
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
end
+
-- typically duplicate_char is sepc
end
+
if comp:sub(1,1) == duplicate_char then -- is first charactier same as duplicate_char? why test first character?
return name; -- and done
+
--   Because individual string segments often (always?) begin with terminal punct for th
end
+
--   preceding segment: 'First element' .. 'sepc next element' .. etc?
 
+
trim = false;
 
+
end_chr = str:sub(-1,-1); -- get the last character of the output string
--[[--------------------------< N A M E _ H A S _ M U L T _ N A M E S >----------------------------------------
+
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
 
+
if end_chr == duplicate_char then -- if same as separator
Evaluates the content of author and editor (surnames only) parameters for multiple names.  Multiple names are
+
str = str:sub(1,-2); -- remove it
indicated if there is more than one comma and or semicolon.  If found, the function adds the multiple name
+
elseif end_chr == "'" then -- if it might be wikimarkup
(author or editor) maintenance category.
+
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?
local function name_has_mult_names (name, list_name)
+
elseif str:sub(-4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''
local count, _;
+
trim = true; -- same question
if is_set (name) then
+
end
_, count = name:gsub ('[;,]', ''); -- count the number of separator-like characters
+
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
if 1 < count then -- param could be |author= or |editor= so one separator character is acceptable
+
trim = true;
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
+
elseif str:sub(-2,-1) == duplicate_char .. "]" then -- if last two chars of str are sepc] external link
end
+
trim = true;
end
+
elseif str:sub(-4,-1) == duplicate_char .. "'']" then -- normal case when |url=something & |title=Title.
return name; -- and done
+
trim = true;
end
+
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
--[[--------------------------< N A M E _ C H E C K S >--------------------------------------------------------
+
if value ~= comp then -- value does not equal comp when value contains html markup
This function calls various name checking functions used to validate the content of the various name-holding
+
local dup2 = duplicate_char;
parameters.
+
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
local function name_checks (last, first, list_name)
+
value = value:sub( 2, -1 ); -- remove duplicate_char when it is first character
if is_set (last) then
+
end
if last:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
+
end
last = last:match ('^%(%((.*)%)%)$'); -- strip parens
+
end
else
+
str = str .. value; --add it to the output string
last = name_has_mult_names (last, list_name); -- check for multiple names in the parameter (last only)
 
last = name_has_ed_markup (last, list_name); -- check for extraneous 'editor' annotation
 
 
end
 
end
 
end
 
end
if is_set (first) then
+
return str;
if first:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
+
end 
first = first:match ('^%(%((.*)%)%)$'); -- strip parens
+
 
else
+
--[[--------------------------< I S _ G O O D _ V A N C _ N A M E >--------------------------------------------
first = name_has_ed_markup (first, list_name); -- check for extraneous 'editor' annotation
 
end
 
end
 
return last, first; -- done
 
end
 
  
 +
For Vancouver Style, author/editor names are supposed to be rendered in Latin (read ASCII) characters.  When a name
 +
uses characters that contain diacritical marks, those characters are to converted to the corresponding Latin character.
 +
When a name is written using a non-Latin alphabet or logogram, that name is to be transliterated into Latin characters.
 +
These things are not currently possible in this module so are left to the editor to do.
  
--[[--------------------------< E X T R A C T _ N A M E S >----------------------------------------------------
+
This test allows |first= and |last= names to contain any of the letters defined in the four Unicode Latin character sets
Gets name list from the input arguments
+
[http://www.unicode.org/charts/PDF/U0000.pdf C0 Controls and Basic Latin] 0041–005A, 0061–007A
 +
[http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] 00C0–00D6, 00D8–00F6, 00F8–00FF
 +
[http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] 0100–017F
 +
[http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] 0180–01BF, 01C4–024F
  
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
+
|lastn= also allowed to contain hyphens, spaces, and apostrophes. (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
Stops searching when both |lastn= and |firstn= are not found in args after two sequential attempts: found |last1=, |last2=, and |last3= but doesn't
+
|firstn= also allowed to contain hyphens, spaces, apostrophes, and periods
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=
+
At the time of this writing, I had to write the 'if nil == mw.ustring.find ...' test ouside of the code editor and paste it here
are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=.
+
because the code editor gets confused between character insertion point and cursor position.
 
 
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
 
that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration. This keeps 'et al.' out of the  
 
template's metadata.  When this occurs, the page is added to a maintenance category.
 
  
 
]]
 
]]
  
local function extract_names(args, list_name)
+
local function is_good_vanc_name (last, first)
local names = {}; -- table of names
+
if nil == mw.ustring.find (last, "^[A-Za-zÀ-ÖØ- öø- ƿDŽ- ɏ%-%s%']*$") or nil == mw.ustring.find (first, "^[A-Za-- ÖØ- öø- ƿDŽ- ɏ%-%s%'%.]*$") then
local last; -- individual name components
+
add_vanc_error ();
local first;
+
return false; -- not a string of latin characters; Vancouver required Romanization
local link;
+
end;
local mask;
+
return true;
local i = 1; -- loop counter/indexer
+
end
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)
 
local etal=false; -- return value set to true when we find some form of et al. in an author parameter
 
  
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
+
--[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------
while true do
 
last = select_one( args, cfg.aliases[list_name .. '-Last'], 'redundant_parameters', i ); -- search through args for name components beginning at 1
 
first = select_one( args, cfg.aliases[list_name .. '-First'], 'redundant_parameters', i );
 
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); -- find and remove variations on et al.
+
Attempts to convert names to initials in support of |name-list-format=vanc.
first, etal = name_has_etal (first, etal, false); -- find and remove variations on et al.
 
last, first= name_checks (last, first, list_name); -- multiple names, extraneous annotation, etc checks
 
 
if first and not last then -- if there is a firstn without a matching lastn
 
table.insert( z.message_tail, { set_error( 'first_missing_last', {err_msg_list_name, i}, 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)
+
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/.
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', {err_msg_list_name, i-1}, true ) } ); -- add this error message
 
end
 
count = 0; -- reset the counter, we're looking for two consecutive missing names
 
end
 
i = i + 1; -- point to next args location
 
end
 
 
return names, etal; -- all done, return our list of names
 
end
 
  
 +
Vancouver style requires family rank designations (Jr, II, III, etc) to be rendered as Jr, 2nd, 3rd, etc.  This form is not
 +
currently supported by this code so correctly formed names like Smith JL 2nd are converted to Smith J2. See http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35085/.
  
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
+
This function uses ustring functions because firstname initials may be any of the unicode Latin characters accepted by is_good_vanc_name ().
 
 
Validates language names provided in |language= parameter if not an ISO639-1 or 639-2 code.
 
 
 
Returns the language name and associated two- or three-character 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.
 
 
 
mw.language.fetchLanguageNames(<local wiki language>, 'all') returns a list of languages that in some cases may include
 
extensions. For example, code 'cbk-zam' and its associated name 'Chavacano de Zamboanga' (MediaWiki does not support
 
code 'cbk' or name 'Chavacano'.  Most (all?) of these languages are not used a 'language' codes per se, rather they
 
are used as sub-domain names: cbk-zam.wikipedia.org.  These names can be found (for the time being) at
 
https://phabricator.wikimedia.org/diffusion/ECLD/browse/master/LocalNames/LocalNamesEn.php
 
 
 
Names but that are included in the list will be found if that name is provided in the |language= parameter.  For example,
 
if |language=Chavacano de Zamboanga, that name will be found with the associated code 'cbk-zam'.  When names are found
 
and the associated code is not two or three characters, this function returns only the Wikimedia language name.
 
 
 
Adapted from code taken from Module:Check ISO 639-1.
 
  
 
]]
 
]]
  
local function get_iso639_code (lang, this_wiki_code)
+
local function reduce_to_initials(first)
local remap = {
+
if mw.ustring.match(first, "^%u%u$") then return first end; -- when first contains just two upper-case letters, nothing to do
['bangla'] = {'Bengali', 'bn'}, -- MediaWiki returns Bangla (the endonym) but we want Bengali (the exonym); here we remap
+
local initials = {}
['bengali'] = {'Bengali', 'bn'}, -- MediaWiki doesn't use exonym so here we provide correct language name and 639-1 code
+
local i = 0; -- counter for number of initials
['bihari'] = {'Bihari', 'bh'}, -- MediaWiki replace 'Bihari' with 'Bhojpuri' so 'Bihari' cannot be found
+
for word in mw.ustring.gmatch(first, "[^%s%.%-]+") do -- names separated by spaces, hyphens, or periods
['bhojpuri'] = {'Bhojpuri', 'bho'}, -- MediaWiki uses 'bh' as a subdomain name for Bhojpuri wWikipedia: bh.wikipedia.org
+
table.insert(initials, mw.ustring.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
if remap[lang:lower()] then
 
return remap[lang:lower()][1], remap[lang:lower()][2]; -- for this language 'name', return a possibly new name and appropriate code
 
 
end
 
end
 
+
return table.concat(initials) -- Vancouver format does not include spaces.
local languages = mw.language.fetchLanguageNames(this_wiki_code, 'all') -- get a list of language names known to Wikimedia
 
-- ('all' is required for North Ndebele, South Ndebele, and Ojibwa)
 
local langlc = mw.ustring.lower(lang); -- lower case version for comparisons
 
 
for code, name in pairs(languages) do -- scan the list to see if we can find our language
 
if langlc == mw.ustring.lower(name) then
 
if 2 ~= code:len() and 3 ~= code:len() then -- two- or three-character codes only; extensions not supported
 
return name; -- so return the name but not the code
 
end
 
return name, code; -- found it, return name to ensure proper capitalization and the the code
 
end
 
end
 
return lang; -- not valid language; return language in original case and nil for the code
 
 
end
 
end
  
 +
--[[--------------------------< L I S T  _ P E O P L E >-------------------------------------------------------
  
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
+
Formats a list of people (e.g. authors / editors)
  
Gets language name from a provided two- or three-character ISO 639 code.  If a code is recognized by MediaWiki,
+
]]
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
+
local function list_people(control, people, etal, list_name) -- TODO: why is list_name here?  not used in this function
that code: Category:Norwegian-language sources (no). For valid three-character code languages, the page is assigned
+
local sep;
to the single category for '639-2' codes: Category:CS1 ISO 639-2 language sources.
+
local namesep;
 +
local format = control.format
 +
local maximum = control.maximum
 +
local lastauthoramp = control.lastauthoramp;
 +
local text = {}
  
Languages that are the same as the local wiki are not categorized.  MediaWiki does not recognize three-character
+
if 'vanc' == format then -- Vancouver-like author/editor name styling?
equivalents of two-character codes: code 'ar' is recognized bit code 'ara' is not.
+
sep = ','; -- name-list separator between authors is a comma
 
+
namesep = ' '; -- last/first separator is a space
This function supports multiple languages in the form |language=nb, French, th where the language names or codes are
+
else
separated from each other by commas.
+
sep = ';' -- name-list separator between authors is a semicolon
 
+
namesep = ', ' -- last/first separator is <comma><space>
]]
+
end
 
 
local function language_parameter (lang)
 
local code; -- the two- or three-character language code
 
local name; -- the language name
 
local language_list = {}; -- table of language names to be rendered
 
local names_table = {}; -- table made from the value assigned to |language=
 
 
 
local this_wiki = mw.getContentLanguage(); -- get a language object for this wiki
 
local this_wiki_code = this_wiki:getCode() -- get this wiki's language code
 
local this_wiki_name = mw.language.fetchLanguageName(this_wiki_code, this_wiki_code); -- get this wiki's language name
 
 
 
local remap = {
 
['bh'] = 'Bihari', -- MediaWiki uses 'bh' as a subdomain name for Bhojpuri wWikipedia: bh.wikipedia.org
 
['bn'] = 'Bengali', -- MediaWiki returns Bangla
 
}
 
 
 
names_table = mw.text.split (lang, '%s*,%s*'); -- names should be a comma separated list
 
 
 
for _, lang in ipairs (names_table) do -- reuse lang
 
 
 
if lang:match ('^%a%a%-') then -- strip ietf language 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
 
 
 
if is_set (name) then -- if |language= specified a valid code
+
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
code = lang:lower(); -- save it
+
if is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for authors
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
+
for i,person in ipairs(people) do
name = remap[code] or name; -- override wikimedia when they misuse language codes/names
+
if is_set(person.last) then
 
+
local mask = person.mask
if this_wiki_code ~= code then -- when the language is not the same as this wiki's language
+
local one
if 2 == code:len() then -- and is a two-character code
+
local sep_one = sep;
add_prop_cat ('foreign_lang_source' .. code, {name, code}) -- categorize it
+
if is_set (maximum) and i > maximum then
else -- or is a recognized language (but has a three-character code)
+
etal = true;
add_prop_cat ('foreign_lang_source_2' .. code, {code}) -- categorize it differently TODO: support mutliple three-character code categories per cs1|2 template
+
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 -- 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 = "[[" .. person.link .. "|" .. one .. "]]" -- link author/editor if this page is not the author's/editor's page
 
end
 
end
 
end
 
end
else
+
table.insert( text, one )
add_maint_cat ('unknown_lang'); -- add maint category if not already added
+
table.insert( text, sep_one )
 
end
 
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
+
local count = #text / 2; -- (number of names + number of separators) divided by 2
if 2 >= code then
+
if count > 0 then  
name = table.concat (language_list, ' and ') -- insert '<space>and<space>' between two language names
+
if count > 1 and is_set(lastauthoramp) and not etal then
elseif 2 < code then
+
text[#text-2] = " & "; -- replace last separator with ampersand text
language_list[code] = 'and ' .. language_list[code]; -- prepend last name with 'and<space>'
+
end
name = table.concat (language_list, ', ') -- and concatenate with '<comma><space>' separators
+
text[#text] = nil; -- erase the last separator
 
end
 
end
if this_wiki_name == name then
+
return ''; -- if one language and that language is this wiki's return an empty string (no annotation)
+
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
 
end
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
+
--[[ TODO: should only return blank or name rather than full list
+
return result, count
so we can clean up the bunched parenthetical elements Language, Type, Format
 
]]
 
 
end
 
end
  
 +
--[[--------------------------< A N C H O R _ I D >------------------------------------------------------------
  
--[[--------------------------< S E T _ C S 1 _ S T Y L E >----------------------------------------------------
+
Generates a CITEREF anchor ID if we have at least one name or a date.  Otherwise returns an empty string.
  
Set style settings for CS1 citation templates. Returns separator and postscript settings
+
namelist is one of the contributor-, author-, or editor-name lists chosen in that order.  year is Year or anchor_year.
  
 
]]
 
]]
 
+
local function anchor_id (namelist, year)
local function set_cs1_style (ps)
+
local names={}; -- a table for the one to four names and year
if not is_set (ps) then -- unless explicitely set to something
+
for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names
ps = '.'; -- terminate the rendered citation with a period
+
names[i] = v.last
 +
if i == 4 then break end -- if four then done
 
end
 
end
return '.', ps; -- separator is a full stop
+
table.insert (names, year); -- add the year at the end
end
+
local id = table.concat(names); -- concatenate names and year for CITEREF id
 
+
if is_set (id) then -- if concatenation is not an empty string
 
+
return "CITEREF" .. id; -- add the CITEREF portion
--[[--------------------------< S E T _ C S 2 _ S T Y L E >----------------------------------------------------
+
else
 
+
return ''; -- return an empty string; no reason to include CITEREF id in this citation
Set style settings for CS2 citation templates. Returns separator, postscript, ref settings
 
 
 
]]
 
 
 
local function set_cs2_style (ps, ref)
 
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
 
ps = ''; -- make sure it isn't nil
 
 
end
 
end
if not is_set (ref) then -- if |ref= is not set
 
ref = "harv"; -- set default |ref=harv
 
end
 
return ',', ps, ref; -- separator is a comma
 
 
end
 
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 >----------------------
+
--[[--------------------------< N A M E _ H A S _ E T A L >----------------------------------------------------
 +
 
 +
Evaluates the content of author and editor name parameters for variations on the theme of et al.  If found,
 +
the et al. is removed, a flag is set to true and the function returns the modified name and the flag.
  
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
+
This function never sets the flag to false but returns it's previous state because it may have been set by
rendered style.
+
previous passes through this function or by the parameters |display-authors=etal or |display-editors=etal
  
 
]]
 
]]
  
local function get_settings_from_cite_class (ps, ref, cite_class)
+
local function name_has_etal (name, etal, nocat)
local sep;
 
if (cite_class == "citation") then -- for citation templates (CS2)
 
sep, ps, ref = set_cs2_style (ps, ref);
 
else -- not a citation template so CS1
 
sep, ps = set_cs1_style (ps);
 
end
 
  
return sep, ps, ref -- return them all
+
if is_set (name) then -- name can be nil in which case just return
 +
local etal_pattern = "[;,]? *[\"']*%f[%a][Ee][Tt] *[Aa][Ll][%.\"']*$" -- variations on the 'et al' theme
 +
local others_pattern = "[;,]? *%f[%a]and [Oo]thers"; -- and alternate to et al.
 +
 +
if name:match (etal_pattern) then -- variants on et al.
 +
name = name:gsub (etal_pattern, ''); -- if found, remove
 +
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
 +
if not nocat then -- no categorization for |vauthors=
 +
add_maint_cat ('etal'); -- and add a category if not already added
 +
end
 +
elseif name:match (others_pattern) then -- if not 'et al.', then 'and others'?
 +
name = name:gsub (others_pattern, ''); -- if found, remove
 +
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
 +
if not nocat then -- no categorization for |vauthors=
 +
add_maint_cat ('etal'); -- and add a category if not already added
 +
end
 +
end
 +
end
 +
return name, etal; --  
 
end
 
end
  
 +
--[[--------------------------< E X T R A C T _ N A M E S >----------------------------------------------------
 +
Gets name list from the input arguments
  
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
+
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=.
  
Establish basic style settings to be used when rendering the citation. Uses |mode= if set and valid or uses
+
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
config.CitationClass from the template's #invoke: to establish style.
+
that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration.  This keeps 'et al.' out of the  
 +
template's metadata.  When this occurs, the page is added to a maintenance category.
  
 
]]
 
]]
  
local function set_style (mode, ps, ref, cite_class)
+
local function extract_names(args, list_name)
local sep;
+
local names = {}; -- table of names
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
+
local last; -- individual name components
sep, ps, ref = set_cs2_style (ps, ref);
+
local first;
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
+
local link;
sep, ps = set_cs1_style (ps);
+
local mask;
else -- anything but cs1 or cs2
+
local i = 1; -- loop counter/indexer
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
+
local n = 1; -- output table indexer
end
+
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)
if 'none' == ps:lower() then -- if assigned value is 'none' then
+
local etal=false; -- return value set to true when we find some form of et al. in an author parameter
ps = ''; -- set to empty string
 
end
 
 
return sep, ps, ref
 
end
 
  
 +
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
 +
while true do
 +
last = select_one( args, cfg.aliases[list_name .. '-Last'], 'redundant_parameters', i ); -- search through args for name components beginning at 1
 +
first = select_one( args, cfg.aliases[list_name .. '-First'], 'redundant_parameters', i );
 +
link = select_one( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i );
 +
mask = select_one( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
  
--[=[-------------------------< I S _ P D F >------------------------------------------------------------------
+
last, etal = name_has_etal (last, etal, false); -- find and remove variations on et al.
 +
first, etal = name_has_etal (first, etal, false); -- find and remove variations on et al.
  
Determines if a url has the file extension that is one of the pdf file extensions used by [[MediaWiki:Common.css]] when
+
if first and not last then -- if there is a firstn without a matching lastn
applying the pdf icon to external links.
+
table.insert( z.message_tail, { set_error( 'first_missing_last', {err_msg_list_name, i}, true ) } ); -- add this error message
 
+
elseif not first and not last then -- if both firstn and lastn aren't found, are we done?
returns true if file extension is one of the recognized extensions, else false
+
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
local function is_pdf (url)
+
else -- we have last with or without a first
return url:match ('%.pdf$') or url:match ('%.PDF$') or url:match ('%.pdf[%?#]') or url:match ('%.PDF[%?#]');
+
if is_set (link) and false == link_param_ok (link) then -- do this test here in case link is missing last
 +
table.insert( z.message_tail, { set_error( 'bad_paramlink', list_name:match ("(%w+)List"):lower() .. '-link' .. i )}); -- url or wikilink in author link;
 +
end
 +
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', {err_msg_list_name, i-1}, true ) } ); -- add this error message
 +
end
 +
count = 0; -- reset the counter, we're looking for two consecutive missing names
 +
end
 +
i = i + 1; -- point to next args location
 +
end
 +
 +
return names, etal; -- all done, return our list of names
 
end
 
end
  
 +
--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------
  
--[[--------------------------< S T Y L E _ F O R M A T >------------------------------------------------------
+
Populates ID table from arguments using configuration settings. Loops through cfg.id_handlers and searches args for
 
+
any of the parameters listed in each cfg.id_handlers['...'].parameters.  If found, adds the parameter and value to
Applies css style to |format=, |chapter-format=, etc. Also emits an error message if the format parameter does
+
the identifier list. Emits redundant error message is more than one alias exists in args
not have a matching url parameter. If the format parameter is not set and the url contains a file extension that
 
is recognized as a pdf document by MediaWiki's commons.css, this code will set the format parameter to (PDF) with
 
the appropriate styling.
 
  
 
]]
 
]]
  
local function style_format (format, url, fmt_param, url_param)
+
local function extract_ids( args )
if is_set (format) then
+
local id_list = {}; -- list of identifiers found in args
format = wrap_style ('format', format); -- add leading space, parentheses, resize
+
for k, v in pairs( cfg.id_handlers ) do -- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
if not is_set (url) then
+
v = select_one( args, v.parameters, 'redundant_parameters' ); -- v.parameters is a table of aliases for k; here we pick one from args if present
format = format .. set_error( 'format_missing_url', {fmt_param, url_param} ); -- add an error message
+
if is_set(v) then id_list[k] = v; end -- if found in args, add identifier to our list
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;
+
return id_list;
 
end
 
end
  
 +
--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------
  
--[[--------------------------< G E T _ D I S P L A Y _ A U T H O R S _ E D I T O R S >------------------------
+
Takes a table of IDs created by extract_ids() and turns it into a table of formatted ID outputs.
 
 
Returns a number that defines the number of names displayed for author and editor name lists and a boolean flag
 
to indicate when et al. should be appended to the name list.
 
 
 
When the value assigned to |display-xxxxors= is a number greater than or equal to zero, return the number and
 
the previous state of the 'etal' flag (false by default but may have been set to true if the name list contains
 
some variant of the text 'et al.').
 
 
 
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:
 
inputs:
max: A['DisplayAuthors'] or A['DisplayEditors']; a number or some flavor of etal
+
id_list – table of identifiers built by extract_ids()
count: #a or #e
+
options – table of various template parameter values used to modify some manually handled identifiers
list_name: 'authors' or 'editors'
 
etal: author_etal or editor_etal
 
  
 
]]
 
]]
  
local function get_display_authors_editors (max, count, list_name, etal)
+
local function build_id_list( id_list, options )
if is_set (max) then
+
local new_list, handler = {};
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.
+
function fallback(k) return { __index = function(t,i) return cfg.id_handlers[k][i] end } end;
etal = true; -- overrides value set by extract_names()
+
elseif max:match ('^%d+$') then -- if is a string of numbers
+
for k, v in pairs( id_list ) do -- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
max = tonumber (max); -- make it a number
+
-- fallback to read-only cfg
if max >= count then -- if |display-xxxxors= value greater than or equal to number of authors/editors
+
handler = setmetatable( { ['id'] = v }, fallback(k) );
add_maint_cat ('disp_auth_ed', cfg.special_case_translation [list_name]);
+
 +
if handler.mode == 'external' then
 +
table.insert( new_list, {handler.label, external_link_id( handler ) } );
 +
elseif handler.mode == 'internal' then
 +
table.insert( new_list, {handler.label, internal_link_id( handler ) } );
 +
elseif handler.mode ~= 'manual' then
 +
error( cfg.messages['unknown_ID_mode'] );
 +
elseif k == 'DOI' then
 +
table.insert( new_list, {handler.label, doi( v, options.DoiBroken ) } );
 +
elseif k == 'HDL' then
 +
table.insert( new_list, {handler.label, hdl( v ) } );
 +
elseif k == 'ARXIV' then
 +
table.insert( new_list, {handler.label, arxiv( v, options.Class ) } );
 +
elseif k == 'ASIN' then
 +
table.insert( new_list, {handler.label, amazon( v, options.ASINTLD ) } );
 +
elseif k == 'LCCN' then
 +
table.insert( new_list, {handler.label, lccn( v ) } );
 +
elseif k == 'OL' or k == 'OLA' then
 +
table.insert( new_list, {handler.label, openlibrary( v ) } );
 +
elseif k == 'PMC' then
 +
table.insert( new_list, {handler.label, pmc( v, options.Embargo ) } );
 +
elseif k == 'PMID' then
 +
table.insert( new_list, {handler.label, pmid( v ) } );
 +
elseif k == 'ISMN' then
 +
table.insert( new_list, {handler.label, ismn( v ) } );
 +
elseif k == 'ISSN' then
 +
table.insert( new_list, {handler.label, issn( v ) } );
 +
elseif k == 'EISSN' then
 +
table.insert( new_list, {handler.label, issn( v, true ) } ); -- true distinguishes eissn from issn
 +
elseif k == 'ISBN' then
 +
local ISBN = internal_link_id( handler );
 +
if not check_isbn( v ) and not is_set(options.IgnoreISBN) then
 +
ISBN = ISBN .. set_error( 'bad_isbn', {}, false, " ", "" );
 
end
 
end
else -- not a valid keyword or number
+
table.insert( new_list, {handler.label, ISBN } );
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
+
elseif k == 'USENETID' then
max = nil; -- unset; as if |display-xxxxors= had not been set
+
table.insert( new_list, {handler.label, message_id( v ) } );
 +
else
 +
error( cfg.messages['unknown_manual_ID'] );
 
end
 
end
 
end
 
end
 
 
return max, etal;
+
function comp( a, b ) -- used in following table.sort()
 +
return a[1] < b[1];
 +
end
 +
 +
table.sort( new_list, comp );
 +
for k, v in ipairs( new_list ) do
 +
new_list[k] = v[2];
 +
end
 +
 +
return new_list;
 
end
 
end
 +
 
  
 +
--[[--------------------------< C O I N S _ C L E A N U P >----------------------------------------------------
  
--[[--------------------------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >------------------------------
+
Cleanup parameter values for the metadata by removing or replacing invisible characters and certain html entities.
  
Adds page to Category:CS1 maint: extra text if |page= or |pages= has what appears to be some form of p. or pp.  
+
2015-12-10: there is a bug in mw.text.unstripNoWiki (). It replaced math stripmarkers with the appropriate content
abbreviation in the first characters of the parameter content.
+
when it shouldn't. See https://phabricator.wikimedia.org/T121085 and Wikipedia_talk:Lua#stripmarkers_and_mw.text.unstripNoWiki.28.29
 
 
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
 
  
 +
TODO: move the replacement patterns and replacement values into a table in /Configuration similar to the invisible
 +
characters table?
 
]]
 
]]
  
local function extra_text_in_page_check (page)
+
local function coins_cleanup (value)
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?
+
value = mw.text.unstripNoWiki (value); -- replace nowiki stripmarkers with their content
local bad_pattern = '^[Pp]?[Pp]%.?[ %d]';
+
value = value:gsub ('<span class="nowrap" style="padding%-left:0%.1em;">&#39;s</span>', "'s"); -- replace {{'s}} template with simple apostrophe-s
 +
value = value:gsub ('&zwj;\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe
 +
value = value:gsub ('\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe (as of 2015-12-11)
 +
value = value:gsub ('&nbsp;', ' '); -- replace &nbsp; entity with plain space
 +
value = value:gsub ('\226\128\138', ' '); -- replace hair space with plain space
 +
value = value:gsub ('&zwj;', ''); -- remove &zwj; entities
 +
value = value:gsub ('[\226\128\141\226\128\139]', '') -- remove zero-width joiner, zero-width space
 +
value = value:gsub ('[\194\173\009\010\013]', ' '); -- replace soft hyphen, horizontal tab, line feed, carriage return with plain space
 +
return value;
 +
end
  
if not page:match (good_pattern) and (page:match (bad_pattern) or  page:match ('^[Pp]ages?')) then
 
add_maint_cat ('extra_text');
 
end
 
end
 
  
 +
--[[--------------------------< C O I N S >--------------------------------------------------------------------
  
--[=[-------------------------< G E T _ V _ N A M E _ T A B L E >----------------------------------------------
+
COinS metadata (see <http://ocoins.info/>) allows automated tools to parse the citation information.
  
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 COinS(data, class)
 +
if 'table' ~= type(data) or nil == next(data) then
 +
return '';
 +
end
  
]=]
+
for k, v in pairs (data) do -- spin through all of the metadata parameter values
 +
if 'ID_list' ~= k and 'Authors' ~= k then -- except the ID_list and Author tables (author nowiki stripmarker done when Author table processed)
 +
data[k] = coins_cleanup (v);
 +
end
 +
end
  
local function get_v_name_table (vparam, output_table, output_link_table)
+
local ctx_ver = "Z39.88-2004";
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;
+
-- treat table strictly as an array with only set values.
 +
local OCinSoutput = setmetatable( {}, {
 +
__newindex = function(self, key, value)
 +
if is_set(value) then
 +
rawset( self, #self+1, table.concat{ key, '=', mw.uri.encode( remove_wiki_link( value ) ) } );
 +
end
 +
end
 +
});
 
 
while name_table[i] do
+
if in_array (class, {'arxiv', 'journal', 'news'}) or (in_array (class, {'conference', 'interview', 'map', 'press release', 'web'}) and is_set(data.Periodical)) or
if name_table[i]:match ('^%(%(.*[^%)][^%)]$') then -- first segment of corporate with one or more commas; this segment has the opening doubled parens
+
('citation' == class and is_set(data.Periodical) and not is_set (data.Encyclopedia)) then
local name = name_table[i];
+
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:journal"; -- journal metadata identifier
i=i+1; -- bump indexer to next segment
+
if 'arxiv' == class then -- set genre according to the type of citation template we are rendering
while name_table[i] do
+
OCinSoutput["rft.genre"] = "preprint"; -- cite arxiv
name = name .. ', ' .. name_table[i]; -- concatenate with previous segments
+
elseif 'conference' == class then
if name_table[i]:match ('^.*%)%)$') then -- if this table member has the closing doubled parens
+
OCinSoutput["rft.genre"] = "conference"; -- cite conference (when Periodical set)
break; -- and done reassembling so
+
elseif 'web' == class then
end
+
OCinSoutput["rft.genre"] = "unknown"; -- cite web (when Periodical set)
i=i+1; -- bump indexer
+
else
 +
OCinSoutput["rft.genre"] = "article"; -- journal and other 'periodical' articles
 +
end
 +
OCinSoutput["rft.jtitle"] = data.Periodical; -- journal only
 +
if is_set (data.Map) then
 +
OCinSoutput["rft.atitle"] = data.Map; -- for a map in a periodical
 +
else
 +
OCinSoutput["rft.atitle"] = data.Title; -- all other 'periodical' article titles
 
end
 
end
table.insert (output_table, name); -- and add corporate name to the output table
+
-- these used onlu for periodicals
table.insert (output_link_table, ''); -- no wikilink
+
OCinSoutput["rft.ssn"] = data.Season; -- keywords: winter, spring, summer, fall
else
+
OCinSoutput["rft.chron"] = data.Chron; -- free-form date components
wl_type, label, link = is_wikilink (name_table[i]); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]]
+
OCinSoutput["rft.volume"] = data.Volume; -- does not apply to books
table.insert (output_table, label); -- add this name
+
OCinSoutput["rft.issue"] = data.Issue;
if 1 == wl_type then
+
OCinSoutput["rft.pages"] = data.Pages; -- also used in book metadata
table.insert (output_link_table, label); -- simple wikilink [[D]]
+
 
 +
elseif 'thesis' ~= class then -- all others except cite thesis are treated as 'book' metadata; genre distinguishes
 +
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:book"; -- book metadata identifier
 +
if 'report' == class or 'techreport' == class then -- cite report and cite techreport
 +
OCinSoutput["rft.genre"] = "report";
 +
elseif 'conference' == class then -- cite conference when Periodical not set
 +
OCinSoutput["rft.genre"] = "conference";
 +
elseif in_array (class, {'book', 'citation', 'encyclopaedia', 'interview', 'map'}) then
 +
if is_set (data.Chapter) then
 +
OCinSoutput["rft.genre"] = "bookitem";
 +
OCinSoutput["rft.atitle"] = data.Chapter; -- book chapter, encyclopedia article, interview in a book, or map title
 
else
 
else
table.insert (output_link_table, link); -- no wikilink or [[L|D]]; add this link if there is one, else empty string
+
if 'map' == class or 'interview' == class then
 +
OCinSoutput["rft.genre"] = 'unknown'; -- standalone map or interview
 +
else
 +
OCinSoutput["rft.genre"] = 'book'; -- book and encyclopedia
 +
end
 
end
 
end
 +
else --{'audio-visual', 'AV-media-notes', 'DVD-notes', 'episode', 'interview', 'mailinglist', 'map', 'newsgroup', 'podcast', 'press release', 'serial', 'sign', 'speech', 'web'}
 +
OCinSoutput["rft.genre"] = "unknown";
 +
end
 +
OCinSoutput["rft.btitle"] = data.Title; -- book only
 +
OCinSoutput["rft.place"] = data.PublicationPlace; -- book only
 +
OCinSoutput["rft.series"] = data.Series; -- book only
 +
OCinSoutput["rft.pages"] = data.Pages; -- book, journal
 +
OCinSoutput["rft.edition"] = data.Edition; -- book only
 +
OCinSoutput["rft.pub"] = data.PublisherName; -- book and dissertation
 +
 +
else -- cite thesis
 +
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:dissertation"; -- dissertation metadata identifier
 +
OCinSoutput["rft.title"] = data.Title; -- dissertation (also patent but that is not yet supported)
 +
OCinSoutput["rft.degree"] = data.Degree; -- dissertation only
 +
OCinSoutput['rft.inst'] = data.PublisherName; -- book and dissertation
 +
end
 +
-- and now common parameters (as much as possible)
 +
OCinSoutput["rft.date"] = data.Date; -- book, journal, dissertation
 +
 +
for k, v in pairs( data.ID_list ) do -- what to do about these? For now assume that they are common to all?
 +
if k == 'ISBN' then v = clean_isbn( v ) end
 +
local id = cfg.id_handlers[k].COinS;
 +
if string.sub( id or "", 1, 4 ) == 'info' then -- for ids that are in the info:registry
 +
OCinSoutput["rft_id"] = table.concat{ id, "/", v };
 +
elseif string.sub (id or "", 1, 3 ) == 'rft' then -- for isbn, issn, eissn, etc that have defined COinS keywords
 +
OCinSoutput[ id ] = v;
 +
elseif id then -- when cfg.id_handlers[k].COinS is not nil
 +
OCinSoutput["rft_id"] = table.concat{ cfg.id_handlers[k].prefix, v }; -- others; provide a url
 
end
 
end
i = i+1;
+
end
end
 
return output_table;
 
end
 
  
 +
--[[
 +
for k, v in pairs( data.ID_list ) do -- what to do about these? For now assume that they are common to all?
 +
local id, value = cfg.id_handlers[k].COinS;
 +
if k == 'ISBN' then value = clean_isbn( 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 = coins_cleanup (v.last), coins_cleanup (v.first or ''); -- replace any nowiki strip markers, non-printing or invisible characers
 +
if k == 1 then -- for the first author name only
 +
if is_set(last)  and is_set(first) then -- set these COinS values if |first= and |last= specify the first author name
 +
OCinSoutput["rft.aulast"] = last; -- book, journal, dissertation
 +
OCinSoutput["rft.aufirst"] = first; -- book, journal, dissertation
 +
elseif is_set(last) then
 +
OCinSoutput["rft.au"] = last; -- book, journal, dissertation -- otherwise use this form for the first name
 +
end
 +
else -- for all other authors
 +
if is_set(last) and is_set(first) then
 +
OCinSoutput["rft.au"] = table.concat{ last, ", ", first }; -- book, journal, dissertation
 +
elseif is_set(last) then
 +
OCinSoutput["rft.au"] = last; -- book, journal, dissertation
 +
end
 +
end
 +
end
 +
 +
OCinSoutput.rft_id = data.URL;
 +
OCinSoutput.rfr_id = table.concat{ "info:sid/", mw.site.server:match( "[^/]*$" ), ":", data.RawPage };
 +
OCinSoutput = setmetatable( OCinSoutput, nil );
 +
 +
-- 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
  
--[[--------------------------< 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
+
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
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.
+
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.
  
This function sets the vancouver error when a reqired comma is missing and when there is a space between an author's initials.
+
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.
  
]]
+
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
 +
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 code.  When names are found and the associated code is not two characters, this function
 +
returns only the Wikimedia language name.
  
local function parse_vauthors_veditors (args, vparam, list_name)
+
Adapted from code taken from Module:Check ISO 639-1.
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
+
local function get_iso639_code (lang)
if v_name:match ('^%(%(.+%)%)$') then -- corporate authors are wrapped in doubled parentheses to supress vanc formatting and error detection
+
if 'norwegian' == lang:lower() then -- special case related to Wikimedia remap of code 'no' at Extension:CLDR
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
+
return 'Norwegian', 'no'; -- Make sure rendered version is properly capitalized
last = v_name:match ('^%(%((.+)%)%)$') -- remove doubled parntheses
+
end
corporate = true; -- flag used in list_people()
+
elseif string.find(v_name, "%s") then
+
local languages = mw.language.fetchLanguageNames(mw.getContentLanguage():getCode(), 'all') -- get a list of language names known to Wikimedia
if v_name:find('[;%.]') then -- look for commonly occurring punctuation characters;
+
-- ('all' is required for North Ndebele, South Ndebele, and Ojibwa)
add_vanc_error ('punctuation');
+
local langlc = mw.ustring.lower(lang); -- lower case version for comparisons
 +
 +
for code, name in pairs(languages) do -- scan the list to see if we can find our language
 +
if langlc == mw.ustring.lower(name) then
 +
if 2 ~= code:len() then -- ISO639-1 codes only
 +
return name; -- so return the name but not the code
 
end
 
end
local lastfirstTable = {}
+
return name, code; -- found it, return name to ensure proper capitalization and the ISO639-1 code
lastfirstTable = mw.text.split(v_name, "%s")
+
end
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
 
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
 
last = v_name; -- last name or single corporate name?  Doesn't support multiword corporate names? do we need this?
 
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
 
add_vanc_error ('initials'); -- too many initials; mixed case initials (which may be ok Romanization); hyphenated initials
 
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 (suffix) then
 
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
 
else
 
if not corporate then
 
is_good_vanc_name (last, '');
 
end
 
end
 
 
 
link = select_one( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i ) or v_link_table[i];
 
mask = select_one( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
 
names[i] = {last = last, first = first, link = link, mask = mask, corporate=corporate}; -- add this assembled name to our names list
 
 
end
 
end
return names, etal; -- all done, return our list of names
+
return lang; -- not valid language; return language in original case and nil for ISO639-1 code
 
end
 
end
  
 +
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
  
--[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------
+
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.
  
Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or
+
There is an exception.  There are three ISO639-1 codes for Norewegian language variants.  There are two official variants: Norwegian Bokmål (code 'nb') and
select one of |editors=, |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list.
+
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.
  
Only one of these appropriate three will be used. The hierarchy is: |authorn= (and aliases) highest and |authors= lowest and
+
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
similarly, |editorn= (and aliases) highest and |editors= lowest
+
return 'Norwegian' when editors use |language=no.  We presume that editors don't know about the variants or can't descriminate between them.
  
When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second
+
See Help talk:Citation Style_1#An ISO 639-1 language name test
test which mimicks the test used in extract_names() when looking for a hole in the author name list.  There may be a better
 
way to do this, I just haven't discovered what that way is.
 
  
Emits an error message when more than one xxxxor name source is provided.
+
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.
  
In this function, vxxxxors = vauthors or veditors; xxxxors = authors or editors as appropriate.
+
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.
  
 
]]
 
]]
  
local function select_author_editor_source (vxxxxors, xxxxors, args, list_name)
+
local function language_parameter (lang)
local lastfirst = false;
+
local code; -- the ISO639-1 two character code
if select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice incase we have a |first1= without a |last1=; this ...
+
local name; -- the language name
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 1 ) or -- ... also catches the case where |first= is used with |vauthors=
+
local language_list = {}; -- table of language names to be rendered
select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) or
+
local names_table = {}; -- table made from the value assigned to |language=
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 2 ) then
 
lastfirst=true;
 
end
 
  
if (is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions
+
names_table = mw.text.split (lang, '%s*,%s*'); -- names should be a comma separated list
(is_set (vxxxxors) and is_set (xxxxors)) or
+
 
(true == lastfirst and is_set (xxxxors)) then
+
for _, lang in ipairs (names_table) do -- reuse lang
local err_name;
+
 
if 'AuthorList' == list_name then -- figure out which name should be used in error message
+
if lang:match ('^%a%a%-') or 2 == lang:len() then -- ISO639-1 language code are 2 characters (fetchLanguageName also supports 3 character codes)
err_name = 'author';
+
if lang:match ('^zh-') then
 +
name = mw.language.fetchLanguageName( lang:lower(), lang:lower() );
 
else
 
else
err_name = 'editor';
+
name = mw.language.fetchLanguageName( lang:lower(), mw.getContentLanguage():getCode() ); -- get ISO 639-1 language name if Language is a proper code
 +
end
 +
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 = '挪威语' end; -- override wikimedia when code is 'no'
 +
if 'zh' ~= code and not code:match ('^zh-') then -- English not the language
 +
add_prop_cat ('foreign_lang_source', {name, code})
 
end
 
end
table.insert( z.message_tail, { set_error( 'redundant_parameters',
+
else
{err_name .. '-name-list parameters'}, true ) } ); -- add error message
+
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
 +
 +
code = #language_list -- reuse code as number of languages in the list
 +
if 2 >= code then
 +
name = table.concat (language_list, ' 及') -- insert '及' between two language names
 +
elseif 2 < code then
 +
language_list[code] = '及' .. language_list[code]; -- prepend last name with '及'
 +
name = table.concat (language_list, '、'); -- and concatenate with '<comma><space>' separators
 +
name = name:gsub ('、及', '及', 1);
 
end
 
end
 
+
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
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;
 
if is_set (xxxxors) then return 3 end;
 
return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last
 
 
end
 
end
  
 +
--[[--------------------------< S E T _ C S 1 _ S T Y L E >----------------------------------------------------
  
--[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------
+
Set style settings for CS1 citation templates. Returns separator and postscript settings
 
 
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.
 
  
 
]]
 
]]
  
local function is_valid_parameter_value (value, name, possible)
+
local function set_cs1_style (ps)
if not is_set (value) then
+
if not is_set (ps) then -- unless explicitely set to something
return true; -- an empty parameter is ok
+
ps = '.'; -- terminate the rendered citation with a period
elseif in_array(value:lower(), possible) then
 
return true;
 
else
 
table.insert( z.message_tail, { set_error( 'invalid_param_val', {name, value}, true ) } ); -- not an allowed value so add error message
 
return false
 
 
end
 
end
 +
return '.', ps; -- separator is a full stop
 
end
 
end
  
 +
--[[--------------------------< S E T _ C S 2 _ S T Y L E >----------------------------------------------------
  
--[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >----------------------------------------
+
Set style settings for CS2 citation templates. Returns separator, postscript, ref settings
 
 
This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space
 
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.
 
  
 
]]
 
]]
  
local function terminate_name_list (name_list, sepc)
+
local function set_cs2_style (ps, ref)
if (string.sub (name_list,-3,-1) == sepc .. '. ') then -- if already properly terminated
+
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
return name_list; -- just return the name list
+
ps = ''; -- make sure it isn't nil
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 not is_set (ref) then -- if |ref= is not set
 +
ref = "harv"; -- set default |ref=harv
 +
end
 +
return ',', ps, ref; -- separator is a comma
 +
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 >----------------------
  
--[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------
+
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
 +
rendered style.
  
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 get_settings_from_cite_class (ps, ref, cite_class)
+
local sep;
local function format_volume_issue (volume, issue, cite_class, origin, sepc, lower)
+
if (cite_class == "citation") then -- for citation templates (CS2)
if not is_set (volume) and not is_set (issue) then
+
sep, ps, ref = set_cs2_style (ps, ref);
return '';
+
else -- not a citation template so CS1
end
+
sep, ps = set_cs1_style (ps);
 
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
 
end
  
local vol = '';
+
return sep, ps, ref -- return them all
+
end
if is_set (volume) then
+
 
if (4 < mw.ustring.len(volume)) then
+
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
vol = substitute (cfg.messages['j-vol'], {sepc, volume});
+
 
else
+
Establish basic style settings to be used when rendering the citation.  Uses |mode= if set and valid or uses
vol = substitute (cfg.presentation['vol-bold'], {sepc, hyphen_to_dash(volume)});
+
config.CitationClass from the template's #invoke: to establish style.
end
+
 
 +
]]
 +
 
 +
local function set_style (mode, ps, ref, cite_class)
 +
local sep;
 +
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
 +
sep, ps, ref = set_cs2_style (ps, ref);
 +
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
 +
sep, ps = set_cs1_style (ps);
 +
else -- anything but cs1 or cs2
 +
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
 
end
 
end
if is_set (issue) then
+
if 'none' == ps:lower() then -- if assigned value is 'none' then
return vol .. substitute (cfg.messages['j-issue'], issue);
+
ps = ''; -- set to empty string
 
end
 
end
return vol;
+
 +
return sep, ps, ref
 
end
 
end
  
 +
--[=[-------------------------< I S _ P D F >------------------------------------------------------------------
  
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
+
Determines if a url has the file extension that is one of the pdf file extensions used by [[MediaWiki:Common.css]] when
 +
applying the pdf icon to external links.
  
adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings.
+
returns true if file extension is one of the recognized extension, else false
The return order is:
 
page, pages, sheet, sheets
 
  
Singular has priority over plural when both are provided.
+
]=]
 +
 
 +
local function is_pdf (url)
 +
return url:match ('%.pdf[%?#]?') or url:match ('%.PDF[%?#]?');
 +
end
 +
 
 +
--[[--------------------------< S T Y L E _ F O R M A T >------------------------------------------------------
 +
 
 +
Applies css style to |format=, |chapter-format=, etc.  Also emits an error message if the format parameter does
 +
not have a matching url parameter.  If the format parameter is not set and the url contains a file extension that
 +
is recognized as a pdf document by MediaWiki's commons.css, this code will set the format parameter to (PDF) with
 +
the appropriate styling.
  
 
]]
 
]]
  
local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower)
+
local function style_format (format, url, fmt_param, url_param)
if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators
+
if is_set (format) then
if is_set (sheet) then
+
format = wrap_style ('format', format); -- add leading space, parenthases, resize
if 'journal' == origin then
+
if not is_set (url) then
return '', '', wrap_msg ('j-sheet', sheet, lower), '';
+
format = format .. set_error( 'format_missing_url', {fmt_param, url_param} ); -- add an error message
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
 
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
  
local is_journal = 'journal' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'journal' == origin);
+
--[[--------------------------< G E T _ D I S P L A Y _ A U T H O R S _ E D I T O R S >------------------------
 
if is_set (page) then
 
if is_journal then
 
return substitute (cfg.messages['j-page(s)'], page), '', '', '';
 
elseif not nopp then
 
return substitute (cfg.messages['p-prefix'], {sepc, page}), '', '', '';
 
else
 
return substitute (cfg.messages['nopp'], {sepc, page}), '', '', '';
 
end
 
elseif is_set(pages) then
 
if is_journal then
 
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
 
return '', substitute (cfg.messages['nopp'], {sepc, pages}), '', '';
 
end
 
end
 
 
return '', '', '', ''; -- return empty strings
 
end
 
  
 +
Returns a number that may or may not limit the length of the author or editor name lists.
  
--[=[-------------------------< A R C H I V E _ U R L _ C H E C K >--------------------------------------------
+
When the value assigned to |display-authors= is a number greater than or equal to zero, return the number and
 +
the previous state of the 'etal' flag (false by default but may have been set to true if the name list contains
 +
some variant of the text 'et al.').
  
Check archive.org urls to make sure they at least look like they are pointing at valid archives and not to the  
+
When the value assigned to |display-authors= is the keyword 'etal', return a number that is one greater than the
save snapshot url or to calendar pagesWhen the archive url is 'https://web.archive.org/save/' (or http://...)
+
number of authors in the list and set the 'etal' flag trueThis will cause the list_people() to display all of
archive.org saves a snapshot of the target page in the url.  That is something that Wikipedia should not allow
+
the names in the name list followed by 'et al.'
unwitting readers to do.
 
  
When the archive.org url does not have a complete timestamp, archive.org chooses a snapshot according to its own
+
In all other cases, returns nil and the previous state of the 'etal' flag.
algorithm or provides a calendar 'search' result.  [[WP:ELNO]] discourages links to search results.
 
  
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:
+
local function get_display_authors_editors (max, count, list_name, etal)
//web.archive.org/<timestamp>... -- the old form
+
if is_set (max) then
//web.archive.org/web/<timestamp>... -- the new form
+
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.
The old form does not support or map to the new form when it contains a display flag.  There are four identified flags
+
etal = true; -- overrides value set by extract_names()
('id_', 'js_', 'cs_', 'im_') but since archive.org ignores others following the same form (two letters and an underscore)
+
elseif max:match ('^%d+$') then -- if is a string of numbers
we don't check for these specific flags but we do check the form.
+
max = tonumber (max); -- make it a number
 
+
if max >= count and 'authors' == list_name then -- AUTHORS ONLY -- if |display-xxxxors= value greater than or equal to number of authors/editors
This function supports a preview mode.  When the article is rendered in preview mode, this funct may return a modified
+
add_maint_cat ('disp_auth_ed', list_name);
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
 
return url, date; -- not an archive.org archive, return ArchiveURL and ArchiveDate
 
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
 
err_msg = 'save command';
 
url = url:gsub ('(//web%.archive%.org)/save/', '%1/*/', 1); -- for preview mode: modify ArchiveURL
 
elseif url:match('//liveweb%.archive%.org/') then
 
err_msg = 'liveweb';
 
else
 
path, timestamp, flag = url:match('//web%.archive%.org/([^%d]*)(%d+)([^/]*)/'); -- split out some of the url parts for evaluation
 
 
if not is_set(timestamp) or 14 ~= timestamp:len() then -- path and flag optional, must have 14-digit timestamp here
 
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
+
else -- not a valid keyword or number
err_msg = 'path';
+
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
elseif is_set (flag) and not is_set (path) then -- flag not allowed with the old form url (without the 'web/' path element)
+
max = nil; -- unset
err_msg = 'flag';
 
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)
 
err_msg = 'flag';
 
else
 
return url, date; -- return archiveURL and ArchiveDate
 
 
end
 
end
 +
elseif 'authors' == list_name then -- AUTHORS ONLY need to clear implicit et al category
 +
max = count + 1; -- number of authors + 1
 
end
 
end
-- if here, something not right so
+
table.insert( z.message_tail, { set_error( 'archive_url', {err_msg}, true ) } ); -- add error message and
+
return max, etal;
if is_set (Frame:preprocess('{{REVISIONID}}')) then
 
return '', ''; -- return empty strings for archiveURL and ArchiveDate
 
else
 
return url, date; -- preview mode so return archiveURL and ArchiveDate
 
end
 
 
end
 
end
  
 +
--[[--------------------------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >------------------------------
  
--[[--------------------------< M I S S I N G _ P I P E _ C H E C K >------------------------------------------
+
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.
  
Look at the contents of a parameter. If the content has a string of characters and digits followed by an equal
+
check Page and Pages for extraneous p, p., pp, and pp. at start of parameter value:
sign, compare the alphanumeric string to the list of cs1|2 parameters. If found, then the string is possibly a
+
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
parameter that is missing its pipe:
+
bad pattern: '^[Pp][Pp]' matches matches when |page(s)= begins pp or pP or Pp or PP
{{cite ... |title=Title access-date=2016-03-17}}
 
 
 
cs1|2 shares some parameter names with xml/html atributes: class=, title=, etc.  To prevent false positives xml/html
 
tags are removed before the search.
 
 
 
If a missing pipe is detected, this function adds the missing pipe maintenance category.
 
  
 
]]
 
]]
  
local function missing_pipe_check (value)
+
local function extra_text_in_page_check (page)
local capture;
+
-- local good_pattern = '^P[^%.P%l]';
value = value:gsub ('%b<>', ''); -- remove xml/html tags because attributes: class=, title=, etc
+
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]';
 +
local bad_pattern = '^[Pp]?[Pp]%.?[ %d]';
  
capture = value:match ('%s+(%a[%a%d]+)%s*=') or value:match ('^(%a[%a%d]+)%s*='); -- find and categorize parameters with possible missing pipes
+
if not page:match (good_pattern) and (page:match (bad_pattern) or page:match ('^[Pp]ages?')) then
if capture and validate (capture) then -- if the capture is a valid parameter name
+
add_maint_cat ('extra_text');
add_maint_cat ('missing_pipe');
 
 
end
 
end
 +
-- if Page:match ('^[Pp]?[Pp]%.?[ %d]') or  Page:match ('^[Pp]ages?[ %d]') or
 +
-- Pages:match ('^[Pp]?[Pp]%.?[ %d]') or  Pages:match ('^[Pp]ages?[ %d]') then
 +
-- add_maint_cat ('extra_text');
 +
-- end
 
end
 
end
  
  
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
+
--[[--------------------------< P A R S E _ V A U T H O R S _ V E D I T O R S >--------------------------------
  
This is the main function doing the majority of the citation formatting.
+
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.
 +
 
 +
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 citation0( config, args)
+
local function parse_vauthors_veditors (args, vparam, list_name)
--[[
+
local names = {}; -- table of names assembled from |vauthors=, |author-maskn=, |author-linkn=
Load Input Parameters
+
local v_name_table = {};
The argument_wrapper facilitates the mapping of multiple aliases to single internal variable.
+
local etal = false; -- return value set to true when we find some form of et al. vauthors parameter
]]
+
local last, first, link, mask;
local A = argument_wrapper( args );
+
local corporate = false;
local i
 
  
-- Pick out the relevant fields from the arguments. Different citation templates
+
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)
-- define different field names for the same underlying things.
+
if vparam:find ('%[%[') or vparam:find ('%]%]') then -- no wikilinking vauthors names
 
+
add_vanc_error ();
-- 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
 +
v_name_table = mw.text.split(vparam, "%s*,%s*") -- names are separated by commas
  
local author_etal;
+
for i, v_name in ipairs(v_name_table) do
local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors=
+
if v_name:match ('^%(%(.+%)%)$') then -- corporate authors are wrapped in doubled parenthese to supress vanc formatting and error detection
local Authors;
+
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
local NameListFormat = A['NameListFormat'];
+
last = v_name:match ('^%(%((.+)%)%)$')
local Collaboration = A['Collaboration'];
+
corporate = true;
 
+
elseif string.find(v_name, "%s") then
do -- to limit scope of selected
+
    lastfirstTable = {}
local selected = select_author_editor_source (A['Vauthors'], A['Authors'], args, 'AuthorList');
+
    lastfirstTable = mw.text.split(v_name, "%s")
if 1 == selected then
+
    first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be author intials
a, author_etal = extract_names (args, 'AuthorList'); -- fetch author list from |authorn= / |lastn= / |firstn=, |author-linkn=, and |author-maskn=
+
    last  = table.concat(lastfirstTable, " ") -- returns a string that is the concatenation of all other names that are not initials
elseif 2 == selected then
+
    if mw.ustring.match (last, '%a+%s+%u+%s+%a+') or mw.ustring.match (v_name, ' %u %u$') then
NameListFormat = 'vanc'; -- override whatever |name-list-format= might be
+
add_vanc_error (); -- matches last II last; the case when a comma is missing or a space between two intiials
a, author_etal = parse_vauthors_veditors (args, args.vauthors, 'AuthorList'); -- fetch author list from |vauthors=, |author-linkn=, and |author-maskn=
+
end
elseif 3 == selected then
+
else
Authors = A['Authors']; -- use content of |authors=
+
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
if 'authors' == A:ORIGIN('Authors') then -- but add a maint cat if the parameter is |authors=
+
last = v_name; -- last name or single corporate name?  Doesn't support multiword corporate names? do we need this?
add_maint_cat ('authors'); -- because use of this parameter is discouraged; what to do about the aliases is a TODO:
 
end
 
 
end
 
end
if is_set (Collaboration) then
+
author_etal = true; -- so that |display-authors=etal not required
+
if is_set (first) and not mw.ustring.match (first, "^%u?%u$") then -- first shall contain one or two upper-case letters, nothing else
 +
add_vanc_error ();
 
end
 
end
 +
-- this from extract_names ()
 +
link = select_one( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i );
 +
mask = select_one( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
 +
names[i] = {last = last, first = first, link = link, mask = mask, corporate=corporate}; -- add this assembled name to our names list
 
end
 
end
 +
return names, etal; -- all done, return our list of names
 +
end
  
local Others = A['Others'];
+
--[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------
  
local editor_etal;
+
Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or
local e = {}; -- editors list from |editor-lastn= / |editor-firstn= pairs or |veditors=
+
select one of |editors=, |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list.
local Editors;
 
  
do -- to limit scope of selected
+
Only one of these appropriate three will be used.  The hierarchy is: |authorn= (and aliases) highest and |authors= lowest and
local selected = select_author_editor_source (A['Veditors'], A['Editors'], args, 'EditorList');
+
similarly, |editorn= (and aliases) highest and |editors= lowest
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
 
  
local t = {}; -- translators list from |translator-lastn= / translator-firstn= pairs
+
When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second
local Translators; -- assembled translators name list
+
test which mimicks the test used in extract_names() when looking for a hole in the author name list.  There may be a better
t = extract_names (args, 'TranslatorList'); -- fetch translator list from |translatorn= / |translator-lastn=, -firstn=, -linkn=, -maskn=
+
way to do this, I just haven't discovered what that way is.
 +
 
 +
Emits an error message when more than one xxxxor name source is provided.
 +
 
 +
In this function, vxxxxors = vauthors or veditors; xxxxors = authors or editors as appropriate.
 +
 
 +
]]
  
local interviewers_list = {};
+
local function select_author_editor_source (vxxxxors, xxxxors, args, list_name)
local Interviewers = A['Interviewers']
+
local lastfirst = false;
if is_set (Interviewers) then -- add a maint cat if the |interviewers= is used
+
if select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice incase we have a first 1 without a last1
add_maint_cat ('interviewers'); -- because use of this parameter is discouraged
+
select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) then
else
+
lastfirst=true;
interviewers_list = extract_names (args, 'InterviewerList'); -- else, process preferred interviewers parameters
 
 
end
 
end
  
local c = {}; -- contributors list from |contributor-lastn= / contributor-firstn= pairs
+
if (is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions
local Contributors; -- assembled contributors name list
+
(is_set (vxxxxors) and is_set (xxxxors)) or
local Contribution = A['Contribution'];
+
(true == lastfirst and is_set (xxxxors)) then
if in_array(config.CitationClass, {"book","citation"}) and not is_set(A['Periodical']) then -- |contributor= and |contribution= only supported in book cites
+
local err_name;
c = extract_names (args, 'ContributorList'); -- fetch contributor list from |contributorn= / |contributor-lastn=, -firstn=, -linkn=, -maskn=
+
if 'AuthorList' == list_name then -- figure out which name should be used in error message
+
err_name = 'author';
if 0 < #c then
+
else
if not is_set (Contribution) then -- |contributor= requires |contribution=
+
err_name = 'editor';
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
 
end
if 0 == #a then -- |contributor= requires |author=
+
table.insert( z.message_tail, { set_error( 'redundant_parameters',
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'author')}); -- add missing author error message
+
{err_name .. '-name-list parameters'}, true ) } ); -- add error message
c = {}; -- blank the contributors' table; it is used as a flag later
 
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
  
if not is_valid_parameter_value (NameListFormat, 'name-list-format', cfg.keywords['name-list-format']) then -- only accepted value for this parameter is 'vanc'
+
if true == lastfirst then return 1 end; -- return a number indicating which author name source to use
NameListFormat = ''; -- anything else, set to empty string
+
if is_set (vxxxxors) then return 2 end;
end
+
if is_set (xxxxors) then return 3 end;
 +
return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last
 +
end
 +
 
  
local Year = A['Year'];
+
--[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------
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'];
 
link_title_ok (TitleLink, A:ORIGIN ('TitleLink'), Title, 'title'); -- check for wikimarkup in |title-link= or wikimarkup in |title= when |title-link= is set
 
  
local Chapter = A['Chapter'];
+
This function is used to validate a parameter's assigned value for those parameters that have only a limited number
local ScriptChapter = A['ScriptChapter'];
+
of allowable values (yes, y, true, no, etc).  When the parameter value has not been assigned a value (missing or empty
local ChapterLink -- = A['ChapterLink']; -- deprecated as a parameter but still used internally by cite episode
+
in the source template) the function refurns true.  If the parameter value is one of the list of allowed values returns
local TransChapter = A['TransChapter'];
+
true; else, emits an error message and returns false.
local TitleType = A['TitleType'];
+
 
local Degree = A['Degree'];
+
]]
local Docket = A['Docket'];
 
local ArchiveFormat = A['ArchiveFormat'];
 
  
local ArchiveDate;
+
local function is_valid_parameter_value (value, name, possible)
local ArchiveURL;
+
if not is_set (value) then
 +
return true; -- an empty parameter is ok
 +
elseif in_array(value:lower(), possible) then
 +
return true;
 +
else
 +
table.insert( z.message_tail, { set_error( 'invalid_param_val', {name, value}, true ) } ); -- not an allowed value so add error message
 +
return false
 +
end
 +
end
  
ArchiveURL, ArchiveDate = archive_url_check (A['ArchiveURL'], A['ArchiveDate'])
 
 
local DeadURL = A['DeadURL']
 
if not is_valid_parameter_value (DeadURL, 'dead-url', cfg.keywords ['deadurl']) then -- set in config.defaults to 'yes'
 
DeadURL = ''; -- anything else, set to empty string
 
end
 
  
local URL = A['URL']
+
--[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >----------------------------------------
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 ConferenceFormat = A['ConferenceFormat'];
 
local ConferenceURL = A['ConferenceURL'];
 
local ConferenceURLorigin = A:ORIGIN('ConferenceURL'); -- get name of parameter that holds ConferenceURL
 
local Periodical = A['Periodical'];
 
local Periodical_origin = A:ORIGIN('Periodical'); -- get the name of the periodical parameter
 
  
local Series = A['Series'];
+
This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space
+
when the last character is not a sepc character or when the last three characters are not sepc followed by two
local Volume;
+
closing square brackets (close of a wikilink).  When either of these is true, the name_list is terminated with a
local Issue;
+
single space character.
local Page;
+
 
local Pages;
+
]]
local At;
 
  
if in_array (config.CitationClass, cfg.templates_using_volume) then
+
local function terminate_name_list (name_list, sepc)
Volume = A['Volume'];
+
if (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
end
+
return name_list .. " "; -- don't add another
-- conference & map books do not support issue
+
else
if in_array (config.CitationClass, cfg.templates_using_issue) and not (in_array (config.CitationClass, {'conference', 'map'}) and not is_set (Periodical))then
+
return name_list .. sepc .. ' '; -- otherwise terninate the name list
Issue = A['Issue'];
 
end
 
local Position = '';
 
if not in_array (config.CitationClass, cfg.templates_not_using_page) then
 
Page = A['Page'];
 
Pages = hyphen_to_dash( A['Pages'] );
 
At = A['At'];
 
 
end
 
end
 +
end
  
local Edition = A['Edition'];
 
local PublicationPlace = A['PublicationPlace']
 
local Place = A['Place'];
 
 
local PublisherName = A['PublisherName'];
 
local RegistrationRequired = A['RegistrationRequired'];
 
if not is_valid_parameter_value (RegistrationRequired, 'registration', cfg.keywords ['yes_true_y']) then
 
RegistrationRequired=nil;
 
end
 
  
local SubscriptionRequired = A['SubscriptionRequired'];
+
--[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------
if not is_valid_parameter_value (SubscriptionRequired, 'subscription', cfg.keywords ['yes_true_y']) then
 
SubscriptionRequired=nil;
 
end
 
  
local UrlAccess = A['UrlAccess'];
+
returns the concatenation of the formatted volume and issue parameters as a single string; or formatted volume
if not is_valid_parameter_value (UrlAccess, 'url-access', cfg.keywords ['url-access']) then
+
or formatted issue, or an empty string if neither are set.
UrlAccess = nil;
 
end
 
if not is_set(URL) and is_set(UrlAccess) then
 
UrlAccess = nil;
 
table.insert( z.message_tail, { set_error( 'param_access_requires_param', {'url'}, true ) } );
 
end
 
  
if is_set (UrlAccess) and is_set (SubscriptionRequired) then -- while not aliases, these are much the same so if both are set
+
]]
table.insert( z.message_tail, { set_error( 'redundant_parameters', {wrap_style ('parameter', 'url-access') .. ' and ' .. wrap_style ('parameter', 'subscription')}, true ) } ); -- add error message
+
SubscriptionRequired = nil; -- unset; prefer |access= over |subscription=
+
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
if is_set (UrlAccess) and is_set (RegistrationRequired) then -- these are not the same but contradictory so if both are set
+
end
table.insert( z.message_tail, { set_error( 'redundant_parameters', {wrap_style ('parameter', 'url-access') .. ' and ' .. wrap_style ('parameter', 'registration')}, true ) } ); -- add error message
+
RegistrationRequired = nil; -- unset; prefer |access= over |registration=
+
local vol = '';
 +
 +
if is_set (volume) then
 +
if (6 < mw.ustring.len(volume)) then
 +
vol = substitute (cfg.messages['j-vol'], {sepc, volume});
 +
else
 +
vol = wrap_style ('vol-bold', hyphen_to_dash(volume));
 
end
 
end
 +
end
 +
if is_set (issue) then
 +
return vol .. substitute (cfg.messages['j-issue'], issue);
 +
end
 +
return vol;
 +
end
  
local ChapterUrlAccess = A['ChapterUrlAccess'];
 
if not is_valid_parameter_value (ChapterUrlAccess, 'chapter-url-access', cfg.keywords ['url-access']) then -- same as url-access
 
ChapterUrlAccess = nil;
 
end
 
if not is_set(ChapterURL) and is_set(ChapterUrlAccess) then
 
ChapterUrlAccess = nil;
 
table.insert( z.message_tail, { set_error( 'param_access_requires_param', {'chapter-url'}, true ) } );
 
end
 
  
local Via = A['Via'];
+
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
local AccessDate = A['AccessDate'];
 
local Agency = A['Agency'];
 
  
local Language = A['Language'];
+
adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings.
local Format = A['Format'];
+
The return order is:
local ChapterFormat = A['ChapterFormat'];
+
page, pages, sheet, sheets
local DoiBroken = A['DoiBroken'];
 
local ID = A['ID'];
 
local ASINTLD = A['ASINTLD'];
 
local IgnoreISBN = A['IgnoreISBN'];
 
if not is_valid_parameter_value (IgnoreISBN, 'ignore-isbn-error', cfg.keywords ['yes_true_y']) then
 
IgnoreISBN = nil; -- anything else, set to empty string
 
end
 
local Embargo = A['Embargo'];
 
local Class = A['Class']; -- arxiv class identifier
 
  
local ID_list = extract_ids( args );
+
Singular has priority over plural when both are provided.
local ID_access_levels = extract_id_access_levels( args, ID_list );
 
  
local Quote = A['Quote'];
+
]]
  
local LayFormat = A['LayFormat'];
+
local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower)
local LayURL = A['LayURL'];
+
if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators
local LaySource = A['LaySource'];
+
if is_set (sheet) then
local Transcript = A['Transcript'];
+
if 'journal' == origin then
local TranscriptFormat = A['TranscriptFormat'];
+
return '', '', wrap_msg ('j-sheet', sheet, lower), '';
local TranscriptURL = A['TranscriptURL']
+
else
local TranscriptURLorigin = A:ORIGIN('TranscriptURL'); -- get name of parameter that holds TranscriptURL
+
return '', '', wrap_msg ('sheet', {sepc, sheet}, lower), '';
 
+
end
local LastAuthorAmp = A['LastAuthorAmp'];
+
elseif is_set (sheets) then
if not is_valid_parameter_value (LastAuthorAmp, 'last-author-amp', cfg.keywords ['yes_true_y']) then
+
if 'journal' == origin then
LastAuthorAmp = nil; -- set to empty string
+
return '', '', '', wrap_msg ('j-sheets', sheets, lower);
end
+
else
 
+
return '', '', '', wrap_msg ('sheets', {sepc, sheets}, lower);
local no_tracking_cats = A['NoTracking'];
+
end
if not is_valid_parameter_value (no_tracking_cats, 'no-tracking', cfg.keywords ['yes_true_y']) then
 
no_tracking_cats = nil; -- set to empty string
 
 
end
 
end
 
--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 = {}; -- holds date info extracted from |date= for the COinS metadata by Module:Date verification
 
 
local DF = A['DF']; -- date format set in cs1|2 template
 
if not is_valid_parameter_value (DF, 'df', cfg.keywords['date-format']) then -- validate reformatting keyword
 
DF = ''; -- not valid, set to empty string
 
 
end
 
end
  
local sepc; -- separator between citation elements for CS1 a period, for CS2, a comma
+
local is_journal = 'journal' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'journal' == origin);
local PostScript;
 
local Ref;
 
sepc, PostScript, Ref = set_style (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 is_set (page) then
if not is_set (no_tracking_cats) then -- ignore if we are already not going to categorize this page
+
if is_journal then
if in_array (this_page.nsText, cfg.uncategorized_namespaces) then
+
return substitute (cfg.messages['j-page(s)'], page), '', '', '';
no_tracking_cats = "true"; -- set no_tracking_cats
+
elseif not nopp then
 +
return substitute (cfg.messages['p-prefix'], {sepc, page}), '', '', '';
 +
else
 +
return substitute (cfg.messages['nopp'], {sepc, page}), '', '', '';
 
end
 
end
for _,v in ipairs (cfg.uncategorized_subpages) do -- cycle through page name patterns
+
elseif is_set(pages) then
if this_page.text:match (v) then -- test page name against each pattern
+
if is_journal then
no_tracking_cats = "true"; -- set no_tracking_cats
+
return substitute (cfg.messages['j-page(s)'], pages), '', '', '';
break; -- bail out if one is found
+
elseif tonumber(pages) ~= nil and not nopp then -- if pages is only digits, assume a single page number
end
+
return '', substitute (cfg.messages['p-prefix'], {sepc, pages}), '', '';
 +
elseif not nopp then
 +
return '', substitute (cfg.messages['pp-prefix'], {sepc, pages}), '', '';
 +
else
 +
return '', substitute (cfg.messages['nopp'], {sepc, pages}), '', '';
 
end
 
end
 
end
 
end
 +
 +
return '', '', '', ''; -- return empty strings
 +
end
  
-- check for extra |page=, |pages= or |at= parameters. (also sheet and sheets while we're at it)
+
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
select_one( args, {'page', 'p', 'pp', 'pages', 'at', 'sheet', 'sheets'}, 'redundant_parameters' ); -- this is a dummy call simply to get the error message and category
+
 
 +
This is the main function doing the majority of the citation formatting.
  
local NoPP = A['NoPP']  
+
]]
if is_set (NoPP) and is_valid_parameter_value (NoPP, 'nopp', cfg.keywords ['yes_true_y']) then
 
NoPP = true;
 
else
 
NoPP = nil; -- unset, used as a flag later
 
end
 
  
if is_set(Page) then
+
local function citation0( config, args)
if is_set(Pages) or is_set(At) then
+
--[[  
Pages = ''; -- unset the others
+
Load Input Parameters
At = '';
+
The argument_wrapper facilitates the mapping of multiple aliases to single internal variable.
end
 
extra_text_in_page_check (Page); -- add this page to maint cat if |page= value begins with what looks like p. or pp.
 
elseif is_set(Pages) then
 
if is_set(At) then
 
At = ''; -- unset
 
end
 
extra_text_in_page_check (Pages); -- add this page to maint cat if |pages= value begins with what looks like p. or pp.
 
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
 
 
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
 
 
All other combinations of |encyclopedia, |title, and |article are not modified
 
 
 
]]
 
]]
 +
local A = argument_wrapper( args );
 +
local i
  
local Encyclopedia = A['Encyclopedia'];
+
-- Pick out the relevant fields from the arguments.  Different citation templates
 +
-- define different field names for the same underlying things.
 +
local author_etal;
 +
local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors=
 +
local Authors;
 +
local NameListFormat = A['NameListFormat'];
  
if ( config.CitationClass == "encyclopaedia" ) or ( config.CitationClass == "citation" and is_set (Encyclopedia)) then -- test code for citation
+
do -- to limit scope of selected
if is_set(Periodical) then -- Periodical is set when |encyclopedia is set
+
local selected = select_author_editor_source (A['Vauthors'], A['Authors'], args, 'AuthorList');
if is_set(Title) or is_set (ScriptTitle) then
+
if 1 == selected then
if not is_set(Chapter) then
+
a, author_etal = extract_names (args, 'AuthorList'); -- fetch author list from |authorn= / |lastn= / |firstn=, |author-linkn=, and |author-maskn=
Chapter = Title; -- |encyclopedia and |title are set so map |title to |article and |encyclopedia to |title
+
elseif 2 == selected then
ScriptChapter = ScriptTitle;
+
NameListFormat = 'vanc'; -- override whatever |name-list-format= might be
TransChapter = TransTitle;
+
a, author_etal = parse_vauthors_veditors (args, args.vauthors, 'AuthorList'); -- fetch author list from |vauthors=, |author-linkn=, and |author-maskn=
ChapterURL = URL;
+
elseif 3 == selected then
ChapterUrlAccess = UrlAccess;
+
Authors = A['Authors']; -- use content of |authors=
 
 
if not is_set (ChapterURL) and is_set (TitleLink) then
 
Chapter = make_wikilink (TitleLink, Chapter);
 
end
 
Title = Periodical;
 
ChapterFormat = Format;
 
Periodical = ''; -- redundant so unset
 
TransTitle = '';
 
URL = '';
 
Format = '';
 
TitleLink = '';
 
ScriptTitle = '';
 
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
 
end
 
end
  
-- Special case for cite techreport.
+
local Coauthors = A['Coauthors'];
if (config.CitationClass == "techreport") then -- special case for cite techreport
+
local Others = A['Others'];
if is_set(A['Number']) then -- cite techreport uses 'number', which other citations alias to 'issue'
+
 
if not is_set(ID) then -- can we use ID for the "number"?
+
local editor_etal;
ID = A['Number']; -- yes, use it
+
local e = {}; -- editors list from |editor-lastn= / |editor-firstn= pairs or |veditors=
else -- ID has a value so emit error message
+
local Editors;
table.insert( z.message_tail, { set_error('redundant_parameters', {wrap_style ('parameter', 'id') .. ' and ' .. wrap_style ('parameter', 'number')}, true )});
 
end
 
end
 
end
 
  
-- special case for cite mailing list
+
do -- to limit scope of selected
if (config.CitationClass == "mailinglist") then
+
local selected = select_author_editor_source (A['Veditors'], A['Editors'], args, 'EditorList');
Periodical = A ['MailingList'];
+
if 1 == selected then
elseif 'mailinglist' == A:ORIGIN('Periodical') then
+
e, editor_etal = extract_names (args, 'EditorList'); -- fetch editor list from |editorn= / |editor-lastn= / |editor-firstn=, |editor-linkn=, and |editor-maskn=
Periodical = ''; -- unset because mailing list is only used for cite mailing list
+
elseif 2 == selected then
end
+
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=
-- Account for the oddity that is {{cite conference}}, before generation of COinS data.
+
elseif 3 == selected then
if 'conference' == config.CitationClass then
+
Editors = A['Editors']; -- use content of |editors=
if is_set(BookTitle) then
 
Chapter = Title;
 
-- ChapterLink = TitleLink; -- |chapterlink= is deprecated
 
ChapterURL = URL;
 
ChapterUrlAccess = UrlAccess;
 
ChapterURLorigin = URLorigin;
 
URLorigin = '';
 
ChapterFormat = Format;
 
TransChapter = TransTitle;
 
Title = BookTitle;
 
Format = '';
 
-- TitleLink = '';
 
TransTitle = '';
 
URL = '';
 
 
end
 
end
elseif 'speech' ~= config.CitationClass then
 
Conference = ''; -- not cite conference or cite speech so make sure this is empty string
 
 
end
 
end
  
-- cite map oddities
+
local t = {}; -- translators list from |translator-lastn= / translator-firstn= pairs
local Cartography = "";
+
local Translators; -- assembled translators name list
local Scale = "";
+
t = extract_names (args, 'TranslatorList'); -- fetch translator list from |translatorn= / |translator-lastn=, -firstn=, -linkn=, -maskn=
local Sheet = A['Sheet'] or '';
+
local Sheets = A['Sheets'] or '';
+
local c = {}; -- contributors list from |contributor-lastn= / contributor-firstn= pairs
if config.CitationClass == "map" then
+
local Contributors; -- assembled contributors name list
Chapter = A['Map'];
+
local Contribution = A['Contribution'];
ChapterURL = A['MapURL'];
+
if in_array(config.CitationClass, {"book","citation"}) and not is_set(A['Periodical']) then -- |contributor= and |contribution= only supported in book cites
ChapterUrlAccess = UrlAccess;
+
c = extract_names (args, 'ContributorList'); -- fetch contributor list from |contributorn= / |contributor-lastn=, -firstn=, -linkn=, -maskn=
TransChapter = A['TransMap'];
 
ChapterURLorigin = A:ORIGIN('MapURL');
 
ChapterFormat = A['MapFormat'];
 
 
 
Cartography = A['Cartography'];
+
if 0 < #c then
if is_set( Cartography ) then
+
if not is_set (Contribution) then -- |contributor= requires |contribution=
Cartography = sepc .. " " .. wrap_msg ('cartography', Cartography, use_lowercase);
+
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'contribution')}); -- add missing contribution error message
end
+
c = {}; -- blank the contributors' table; it is used as a flag later
Scale = A['Scale'];
+
end