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
mediawiki1>Xiplus
(翻譯分類)
(測試)
行 1: 行 1:
  
local z = {
+
local cs1 ={};
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 cfg = {}; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration
+
local dates, year_date_check, reformat_dates, date_hyphen_to_dash, -- functions in Module:Citation/CS1/Date_validation
local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist
+
date_name_xlate
  
--[[--------------------------< I S _ S E T >------------------------------------------------------------------
+
local is_set, in_array, substitute, error_comment, set_error, select_one, -- functions in Module:Citation/CS1/Utilities
 +
add_maint_cat, wrap_style, safe_for_italics, is_wikilink, make_wikilink;
  
Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not an empty string.
+
local z ={}; -- tables in Module:Citation/CS1/Utilities
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 )
+
 
return not (var == nil or var == '');
+
local added_deprecated_cat; -- boolean flag so that the category is added only once
end
+
local added_prop_cats = {}; -- list of property categories that have been added to z.properties_cats
 +
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 46: 行 56:
 
end
 
end
  
--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------
 
  
Whether needle is in haystack
+
--[[--------------------------< A D D _ P R O P _ C A T >--------------------------------------------------------
 +
 
 +
Adds a category to z.properties_cats using names from the configuration file with additional text if any.
 +
 
 +
foreign_lang_source and foreign_lang_source_2 keys have a language code appended to them so that multiple languages
 +
may be categorized but multiples of the same language are not categorized.
 +
 
 +
added_prop_cats is a table declared in page scope variables above
  
 
]]
 
]]
  
local function in_array( needle, haystack )
+
local function add_prop_cat (key, arguments)
if needle == nil then
+
if not added_prop_cats [key] then
return false;
+
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
for n,v in ipairs( haystack ) do
 
if v == needle then
 
return n;
 
end
 
end
 
return false;
 
 
end
 
end
  
--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
 
  
Populates numbered arguments in a message string using an argument table.
+
--[[--------------------------< A D D _ V A N C _ E R R O R >----------------------------------------------------
  
]]
+
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.
  
local function substitute( msg, args )
+
added_vanc_errs is a boolean declared in page scope variables above
return args and mw.message.newRawMessage( msg, args ):plain() or msg;
 
end
 
  
--[[--------------------------< E R R O R _ C O M M E N T >----------------------------------------------------
+
]]
  
Wraps error messages with css markup according to the state of hidden.
+
local function add_vanc_error (source)
 
+
if not added_vanc_errs then
]]
+
added_vanc_errs = true; -- note that we've added this category
local function error_comment( content, hidden )
+
table.insert( z.message_tail, { set_error( 'vancouver', {source}, true ) } );
return substitute( hidden and cfg.presentation['hidden-error'] or cfg.presentation['visible-error'], content );
+
end
 
end
 
end
  
--[[--------------------------< S E T _ E R R O R >--------------------------------------------------------------
 
  
Sets an error condition and returns the appropriate error message.  The actual placement of the error message in the output is
+
--[[--------------------------< I S _ S C H E M E >------------------------------------------------------------
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
local function set_error( error_id, arguments, raw, prefix, suffix )
+
is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says:
local error_state = cfg.error_conditions[ error_id ];
+
Scheme names consist of a sequence of characters beginning with a
+
  letter and followed by any combination of letters, digits, plus
prefix = prefix or "";
+
  ("+"), period ("."), or hyphen ("-").
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 added_maint_cats = {} -- list of maintenance categories that have been added to z.maintenance_cats
+
local function is_scheme (scheme)
local function add_maint_cat (key, arguments)
+
return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern
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 >--------------------------------------------------------
 
  
Adds a category to z.properties_cats using names from the configuration file with additional text if any.
+
--[=[-------------------------< I S _ D O M A I N _ N A M E >--------------------------------------------------
  
]]
+
Does this thing that purports to be a domain name seem to be a valid domain name?
  
local added_prop_cats = {} -- list of property categories that have been added to z.properties_cats
+
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
local function add_prop_cat (key, arguments)
+
BNF defined here: https://tools.ietf.org/html/rfc4234
if not added_prop_cats [key] then
+
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
added_prop_cats [key] = true; -- note that we've added this category
+
see also [[Single-letter second-level domain]]
table.insert( z.properties_cats, substitute (cfg.prop_cats [key], arguments)); -- make name then add to table
+
list of tlds: https://www.iana.org/domains/root/db
end
 
end
 
  
--[[--------------------------< A D D _ V A N C _ E R R O R >----------------------------------------------------
+
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.
  
Adds a single Vancouver system error message to the template's output regardless of how many error actually exist.
+
Also allowed are IPv4 addresses. IPv6 not supported
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.
  
local added_vanc_errs; -- flag so we only emit one Vancouver error / category
+
There are several tests:
local function add_vanc_error ()
+
the first character of the whole domain name including subdomains must be a letter or a digit
if not added_vanc_errs then
+
internationalized domain name (ascii characters with .xn-- ASCII Compatible Encoding (ACE) prefix xn-- in the tld) see https://tools.ietf.org/html/rfc3490
added_vanc_errs = true; -- note that we've added this category
+
single-letter/digit second-level domains in the .org and .cash TLDs
table.insert( z.message_tail, { set_error( 'vancouver', {}, true ) } );
+
q, x, and z SL domains in the .com TLD
end
+
i and q SL domains in the .net TLD
end
+
single-letter SL domains in the ccTLDs (where the ccTLD is two letters)
 +
two-character SL domains in gTLDs (where the gTLD is two or more letters)
 +
three-plus-character SL domains in gTLDs (where the gTLD is two or more letters)
 +
IPv4 dot-decimal address format; TLD not allowed
  
 +
returns true if domain appears to be a proper name and tld or IPv4 address, else false
  
--[[--------------------------< I S _ S C H E M E >------------------------------------------------------------
+
]=]
  
does this thing that purports to be a uri scheme seem to be a valid scheme?  The scheme is checked to see if it
+
local function is_domain_name (domain)
is in agreement with http://tools.ietf.org/html/std66#section-3.1 which says:
+
if not domain then
Scheme names consist of a sequence of characters beginning with a
+
return false; -- if not set, abandon
  letter and followed by any combination of letters, digits, plus
+
end
  ("+"), period ("."), or hyphen ("-").
+
 +
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
 +
-- Do most common case first
 +
if domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.%a%a+$') then -- three or more character hostname.hostname or hostname.tld
 +
return true;
 +
elseif domain:match ('%f[%a%d][%a%d][%a%d%-]+[%a%d]%.xn%-%-[%a%d]+$') 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
  
returns true if it does, else false
 
  
]]
+
--[[--------------------------< I S _ U R L >------------------------------------------------------------------
  
local function is_scheme (scheme)
+
returns true if the scheme and domain parts of a url appear to be a valid url; else false.
return scheme and scheme:match ('^%a[%a%d%+%.%-]*:'); -- true if scheme is set and matches the pattern
+
 
 +
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
 
end
  
  
--[=[-------------------------< I S _ D O M A I N _ N A M E >--------------------------------------------------
+
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
 +
 
 +
Split a url into a scheme, authority indicator, and domain.
  
Does this thing that purports to be a domain name seem to be a valid domain name?
+
First remove Fully Qualified Domain Name terminator (a dot following tld) (if any) and any path(/), query(?) or fragment(#).
  
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
+
If protocol relative url, return nil scheme and domain else return nil for both scheme and domain.
BNF defined here: https://tools.ietf.org/html/rfc4234
 
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
 
see also [[Single-letter second-level domain]]
 
list of tlds: https://www.iana.org/domains/root/db
 
  
rfc952 (modified by rfc 1123) requires the first and last character of a hostname to be a letter or a digitBetween
+
When not protocol relative, get scheme, authority indicator, and domainIf there is an authority indicator (one
the first and last characters the name may use letters, digits, and the hyphen.
+
or more '/' characters immediately following the scheme's colon), make sure that there are only 2.
  
Also allowed are IPv4 addresses. IPv6 not supported
+
Strip off any port and path;
  
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.
+
local function split_url (url_str)
 +
local scheme, authority, domain;
 +
 +
url_str = url_str:gsub ('([%a%d])%.?[/%?#].*$', '%1'); -- strip FQDN terminator and path(/), query(?), fragment (#) (the capture prevents false replacement of '//')
 +
 
 +
if url_str:match ('^//%S*') then -- if there is what appears to be a protocol relative url
 +
domain = url_str:match ('^//(%S*)')
 +
elseif url_str:match ('%S-:/*%S+') then -- if there is what appears to be a scheme, optional authority indicator, and domain name
 +
scheme, authority, domain = url_str:match ('(%S-:)(/*)(%S+)'); -- extract the scheme, authority indicator, and domain portions
 +
authority = authority:gsub ('//', '', 1); -- replace place 1 pair of '/' with nothing;
 +
if is_set(authority) then -- if anything left (1 or 3+ '/' where authority should be) then
 +
return scheme; -- return scheme only making domain nil which will cause an error message
 +
end
 +
domain = domain:gsub ('(%a):%d+', '%1'); -- strip port number if present
 +
end
 +
 +
return scheme, domain;
 +
end
  
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
 
  
returns true if domain appears to be a proper name and tld or IPv4 address, else false
+
--[[--------------------------< L I N K _ P A R A M _ O K >---------------------------------------------------
  
]=]
+
checks the content of |title-link=, |series-link=, |author-link= etc for properly formatted content: no wikilinks, no urls
  
local function is_domain_name (domain)
+
Link parameters are to hold the title of a wikipedia article so none of the WP:TITLESPECIALCHARACTERS are allowed:
if not domain then
+
# < > [ ] | { } _
return false; -- if not set, abandon
+
except the underscore which is used as a space in wiki urls and # which is used for section links
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
 
  
 +
returns false when the value contains any of these characters.
  
--[[--------------------------< I S _ U R L >------------------------------------------------------------------
+
When there are no illegal characters, this function returns TRUE if value DOES NOT appear to be a valid url (the
 +
|<param>-link= parameter is ok); else false when value appears to be a valid url (the |<param>-link= parameter is NOT ok).
  
returns true if the scheme and domain parts of a url appear to be a valid url; else false.
+
]]
  
This function is the last step in the validation process.  This function is separate because there are cases that
+
local function link_param_ok (value)
are not covered by split_url(), for example is_parameter_ext_wikilink() which is looking for bracketted external
+
local scheme, domain;
wikilinks.
+
if value:find ('[<>%[%]|{}]') then -- if any prohibited characters
 +
return false;
 +
end
  
]]
+
scheme, domain = split_url (value); -- get scheme or nil and domain or nil from url;  
 
+
return not is_url (scheme, domain); -- return true if value DOES NOT appear to be a valid url
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
 
end
  
  
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
+
--[[--------------------------< L I N K _ T I T L E _ O K >---------------------------------------------------
  
Split a url into a scheme, authority indicator, and domain.
+
Use link_param_ok() to validate |<param>-link= value and its matching |<title>= value.
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 domainIf there is an authority indicator (one
+
|<title>= may be wikilinked but not when |<param>-link= has a valueThis function emits an error message when
or more '/' characters following the scheme's colon), make sure that there are only 2.
+
that condition exists
  
 
]]
 
]]
  
local function split_url (url_str)
+
local function link_title_ok (link, lorig, title, torig)
local scheme, authority, domain;
+
local orig;
 
url_str = url_str:gsub ('([%a%d])%.?[/%?#].*$', '%1'); -- strip FQDN terminator and path(/), query(?), fragment (#) (the capture prevents false replacement of '//')
 
  
if url_str:match ('^//%S*') then -- if there is what appears to be a protocol relative url
+
if is_set (link) then -- don't bother if <param>-link doesn't have a value
domain = url_str:match ('^//(%S*)')
+
if not link_param_ok (link) then -- check |<param>-link= markup
elseif url_str:match ('%S-:/*%S+') then -- if there is what appears to be a scheme, optional authority indicator, and domain name
+
orig = lorig; -- identify the failing link parameter
scheme, authority, domain = url_str:match ('(%S-:)(/*)(%S+)'); -- extract the scheme, authority indicator, and domain portions
+
elseif title:find ('%[%[') then -- check |title= for wikilink markup
authority = authority:gsub ('//', '', 1); -- replace place 1 pair of '/' with nothing;
+
orig = torig; -- identify the failing |title= parameter
if is_set(authority) then -- if anything left (1 or 3+ '/' where authority should be) then
 
return scheme; -- return scheme only making domain nil which will cause an error message
 
 
end
 
end
domain = domain:gsub ('(%a):%d+', '%1'); -- strip port number if present
 
 
end
 
end
 
return scheme, domain;
 
end
 
  
 
+
if is_set (orig) then
--[[--------------------------< L I N K _ P A R A M _ O K >---------------------------------------------------
+
table.insert( z.message_tail, { set_error( 'bad_paramlink', orig)}); -- url or wikilink in |title= with |title-link=;
 
 
checks the content of |title-link=, |series-link=, |author-link= etc for properly formatted content: no wikilinks, no urls
 
 
 
Link parameters are to hold the title of a wikipedia article so none of the WP:TITLESPECIALCHARACTERS are allowed:
 
# < > [ ] | { } _
 
except the underscore which is used as a space in wiki urls and # which is used for section links
 
 
 
returns false when the value contains any of these characters.
 
 
 
When there are no illegal characters, this function returns TRUE if value DOES NOT appear to be a valid url (the
 
|<param>-link= parameter is ok); else false when value appears to be a valid url (the |<param>-link= parameter is NOT ok).
 
 
 
]]
 
 
 
local function link_param_ok (value)
 
local scheme, domain;
 
if value:find ('[<>%[%]|{}]') then -- if any prohibited characters
 
return false;
 
 
end
 
end
 
scheme, domain = split_url (value); -- get scheme or nil and domain or nil from url;
 
return not is_url (scheme, domain); -- return true if value DOES NOT appear to be a valid url
 
 
end
 
end
  
行 334: 行 293:
 
portions, or for protocol relative (//example.com) urls, just the domain.  Use is_url() to validate the two
 
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.
 
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
  
 
]]
 
]]
行 343: 行 306:
 
local scheme, domain;
 
local scheme, domain;
  
scheme, domain = split_url (url_str); -- get scheme or nil and domain or nil from url;  
+
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
 
return is_url (scheme, domain); -- return true if value appears to be a valid url
 
end
 
end
行 363: 行 331:
 
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 = value:match ('%f[%[]%[(%a%S*:)(%S+).*%]')
+
scheme, domain = split_url (value:match ('%f[%[]%[(%a%S*:%S+).*%]'));
elseif value:match ('%f[%[]%[//%S*%.%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
+
elseif value:match ('%f[%[]%[//%S+.*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
domain = value:match ('%f[%[]%[//(%S*%.%S+).*%]');
+
scheme, domain = split_url (value:match ('%f[%[]%[(//%S+).*%]'));
 
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
 
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
scheme, domain = value:match ('(%a%S*:)(%S+)');
+
scheme, domain = split_url (value:match ('(%a%S*:%S+)'));
elseif value:match ('//%S*%.%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
+
elseif value:match ('//%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
domain = value:match ('//(%S*%.%S+)'); -- what is left should be the domain
+
scheme, domain = split_url (value:match ('(//%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 404: 行 370:
  
  
--[[--------------------------< S A F E _ F O R _ I T A L I C S >----------------------------------------------
+
--[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------
  
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_italics( str )
+
local function safe_for_url( str )
if not is_set(str) then
+
if str:match( "%[%[.-%]%]" ) ~= nil then  
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 ) } );
 
table.insert( z.message_tail, { set_error( 'wikilink_in_url', {}, true ) } );
 
end
 
end
行 443: 行 387:
 
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 467: 行 394:
 
]]
 
]]
  
local function external_link( URL, label, source )
+
local function external_link( URL, label, source, access)
 
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 480: 行 411:
 
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 });
+
end
+
domain, path = URL:match ('^([/%.%-%+:%a%d]+)([/%?#].*)$'); -- split the url into scheme plus domain and path
 +
if path then -- if there is a path portion
 +
path = path:gsub ('[%[%]]', {['[']='%5b',[']']='%5d'}); -- replace '[' and ']' with their percent encoded values
 +
URL=domain..path; -- and reassemble
 +
end
 +
 +
if is_set (access) then -- access level (subscription, registration, limited)
 +
label = safe_for_url (label); -- replace square brackets and newlines
  
--[[--------------------------< E X T E R N A L _ L I N K _ I D >----------------------------------------------
+
base_url = table.concat ( -- assemble external link with access signal
 
+
{
Formats a wiki style external link
+
'<span class="plainlinks">[', -- opening css and url markup
 
+
URL, -- the url
]]
+
' ', -- the required space
 
+
label,
local function external_link_id(options)
+
'<span style="padding-left:0.15em">', -- signal spacing css
local url_string = options.id;
+
cfg.presentation[access], -- the appropriate icon
if options.encode == true or options.encode == nil then
+
'</span>', -- close signal spacing span
url_string = mw.uri.encode( url_string );
+
']</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\>]',
+
options.prefix, url_string, options.suffix or "",
+
return table.concat({ base_url, error_str });
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 506: 行 445:
 
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 page_in_deprecated_cat then
+
if not added_deprecated_cat then
page_in_deprecated_cat = true; -- note that we've added this category
+
added_deprecated_cat = true; -- note that we've added this category
 
table.insert( z.message_tail, { set_error( 'deprecated_params', {name}, true ) } ); -- add error message
 
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 >--------------------------------------------------------
 
  
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.
+
--[=[-------------------------< 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:
 
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]]
 
 
cap, cap2 = str:match ("^([\"\'])([^\'].+)"); -- match leading double or single quote but not double single quotes
+
if 1 == wl_type then -- [[D]] simple wikilink with or without quote marks
if is_set (cap) then
+
if mw.ustring.match (str, '%[%[[\" “”\'‘’].+[\" “”\' ‘’]%]%]') then -- leading and trailing quote marks
str = substitute (cfg.presentation['kern-left'], {cap, cap2});
+
str = substitute (cfg.presentation['kern-wl-both'], str);
end
+
elseif mw.ustring.match (str, '%[%[[\"“”\' ‘’].+%]%]') then -- leading quote marks
 +
str = substitute (cfg.presentation['kern-wl-left'], str);
 +
elseif mw.ustring.match (str, '%[%[.+[\"“”\'‘’]%]%]') then -- trailing quote marks
 +
str = substitute (cfg.presentation['kern-wl-right'], str);
 +
end
 +
 
 +
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 = str:match ("^(.+[^\'])([\"\'])$")
+
cap, cap2 = mw.ustring.match (label, "^([\"\'])([^\'].+)"); -- match leading double or single quote but not doubled single quotes (italic markup)
if is_set (cap) then
+
if is_set (cap) then
str = substitute (cfg.presentation['kern-right'], {cap, cap2});
+
label = substitute (cfg.presentation['kern-left'], {cap, cap2});
 +
end
 +
 +
cap, cap2 = mw.ustring.match (label, "^(.+[^\'])([\"\'])$") -- 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 579: 行 549:
 
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, mw.getContentLanguage():getCode() ); -- get language name so that we can use it to categorize
+
name = mw.language.fetchLanguageName( lang, "en" ); -- get language name so that we can use it to categorize
 
if is_set (name) then -- is prefix a proper ISO 639-1 language code?
 
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, {'ar', 'bg', 'bs', 'dv', 'el', 'fa', 'he', 'hy', 'ja', 'ka', 'ko', 'ku', 'mk', 'ps', 'ru', 'sd', 'sr', 'th', 'uk', 'ug', 'yi', 'zh'}) then
+
if in_array (lang, cfg.script_lang_codes) then
 
add_prop_cat ('script_with_name', {name, lang})
 
add_prop_cat ('script_with_name', {name, lang})
 
else
 
else
行 592: 行 562:
 
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 600: 行 567:
 
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 608: 行 576:
  
 
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 643: 行 608:
  
  
--[[-------------------------< I S _ A L I A S _ U S E D >-----------------------------------------------------
+
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
  
This function is used by select_one() to determine if one of a list of alias parameters is in the argument list
+
Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
provided by the template.
+
parameter (chapter_url_source used for error messages).
 
 
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
 
 
 
Returns:
 
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 is_alias_used (args, alias, index, enumerated, value, selected, error_list)
+
local function format_chapter_title (scriptchapter, chapter, transchapter, chapterurl, chapter_url_source, no_quotes, access)
if enumerated then -- is this a test for an enumerated parameters?
+
local chapter_error = '';
alias = alias:gsub ('#', index); -- replace '#' with the value in index
+
 +
if not is_set (chapter) then
 +
chapter = ''; -- to be safe for concatenation
 
else
 
else
alias = alias:gsub ('#', ''); -- remove '#' if it exists
+
if false == no_quotes then
 +
chapter = kern_quotes (chapter); -- if necessary, separate chapter title's leading and trailing quote marks from Module provided quote marks
 +
chapter = wrap_style ('quoted-title', chapter);
 +
end
 
end
 
end
  
if is_set(args[alias]) then -- alias is in the template's argument list
+
chapter = script_concatenate (chapter, scriptchapter) -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
if value ~= nil and selected ~= alias then -- if we have already selected one of the aliases
+
 
local skip;
+
if is_set (transchapter) then
for _, v in ipairs(error_list) do -- spin through the error list to see if we've added this alias
+
transchapter = wrap_style ('trans-quoted-title', transchapter);
if v == alias then
+
if is_set (chapter) then
skip = true;
+
chapter = chapter ..  ' ' .. transchapter;
break; -- has been added so stop looking
+
else -- here when transchapter without chapter or script-chapter
end
+
chapter = transchapter; --  
end
+
chapter_error = ' ' .. set_error ('trans_missing_title', {'chapter'});
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
  
--[[--------------------------< S E L E C T _ O N E >----------------------------------------------------------
+
return chapter .. chapter_error;
 +
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'.
 
  
Because enumerated parameter |<param>1= is an alias of |<param>= we must test for both possibilities.
+
--[[--------------------------< H A S _ I N V I S I B L E _ C H A R S >----------------------------------------
  
 +
This function searches a parameter's value for nonprintable or invisible characters.  The search stops at the
 +
first match.
  
Generates an error if more than one match is present.
+
This function will detect the visible replacement character when it is part of the wikisource.
  
]]
+
Detects but ignores nowiki and math stripmarkers.  Also detects other named stripmarkers (gallery, math, pre, ref)
 +
and identifies them with a slightly different error message.  See also coins_cleanup().
  
local function select_one( args, aliases_list, error_condition, index )
+
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
local value = nil; -- the value assigned to the selected parameter
 
local selected = ''; -- the name of the parameter we have chosen
 
local error_list = {};
 
  
if index ~= nil then index = tostring(index); end
+
Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker
 +
that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the
 +
parameter value.
  
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
 
  
if #error_list > 0 and 'none' ~= error_condition then -- for cases where this code is used outside of extract_names()
+
local function has_invisible_chars (param, v)
local error_str = "";
+
local position = ''; -- position of invisible char or starting position of stripmarker
for _, k in ipairs( error_list ) do
+
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
if error_str ~= "" then error_str = error_str .. cfg.messages['parameter-separator'] end
+
local capture; -- used by stripmarker detection to hold name of the stripmarker
error_str = error_str .. wrap_style ('parameter', k);
+
local i=1;
end
+
local stripmarker, apostrophe;
if #error_list > 1 then
+
error_str = error_str .. cfg.messages['parameter-final-separator'];
+
capture = string.match (v, '[%w%p ]*'); -- Test for values that are simple ASCII text and bypass other tests if true
else
+
if capture == v then -- if same there are no unicode characters
error_str = error_str .. cfg.messages['parameter-pair-separator'];
+
return;
 +
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
error_str = error_str .. wrap_style ('parameter', selected);
+
i=i+1; -- bump our index
table.insert( z.message_tail, { set_error( error_condition, {error_str}, true ) } );
 
 
end
 
end
 
return value, selected;
 
 
end
 
end
  
  
--[[--------------------------< F O R M A T _ C H A P T E R _ T I T L E >--------------------------------------
+
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
  
Format the four chapter parameters: |script-chapter=, |chapter=, |trans-chapter=, and |chapter-url= into a single Chapter meta-
+
Argument wrapper.  This function provides support for argument mapping defined in the configuration file so that
parameter (chapter_url_source used for error messages).
+
multiple names can be transparently aliased to single internal variable.
  
 
]]
 
]]
  
local function format_chapter_title (scriptchapter, chapter, transchapter, chapterurl, chapter_url_source, no_quotes)
+
local function argument_wrapper( args )
local chapter_error = '';
+
local origin = {};
 
 
if not is_set (chapter) then
+
return setmetatable({
chapter = ''; -- to be safe for concatenation
+
ORIGIN = function( self, k )
else
+
local dummy = self[k]; --force the variable to be loaded.
if false == no_quotes then
+
return origin[k];
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
+
__index = function ( tbl, k )
 
+
if origin[k] ~= nil then
if is_set (transchapter) then
+
return nil;
transchapter = wrap_style ('trans-quoted-title', transchapter);
+
end
if is_set (chapter) then
+
chapter = chapter ..  ' ' .. transchapter;
+
local args, list, v = args, cfg.aliases[k];
else -- here when transchapter without chapter or script-chapter
+
chapter = transchapter; --  
+
if type( list ) == 'table' then
chapter_error = ' ' .. set_error ('trans_missing_title', {'chapter'});
+
v, origin[k] = select_one( args, list, 'redundant_parameters' );
end
+
if origin[k] == nil then
end
+
origin[k] = ''; -- Empty string, not nil
 
+
end
if is_set (chapterurl) then
+
elseif list ~= nil then
chapter = external_link (chapterurl, chapter, chapter_url_source); -- adds bare_url_missing_title error if appropriate
+
v, origin[k] = args[list], list;
end
+
else
 
+
-- maybe let through instead of raising an error?
return chapter .. chapter_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 >----------------------------------------
 
  
This function searches a parameter's value for nonprintable or invisible characters.  The search stops at the
+
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
first match.
 
  
This function will detect the visible replacement character when it is part of the wikisource.
+
Looks for a parameter's name in one of several whitelists.
  
Detects but ignores nowiki and math stripmarkers.  Also detects other named stripmarkers (gallery, math, pre, ref)
+
Parameters in the whitelist can have three values:
and identifies them with a slightly different error message.  See also coins_cleanup().
+
true - active, supported parameters
 +
false - deprecated, supported parameters
 +
nil - unsupported parameters
 +
 +
]]
  
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
+
local function validate( name, cite_class )
 +
local name = tostring( name );
 +
local state;
 +
 +
if in_array (cite_class, {'arxiv', 'biorxiv', 'citeseerx'}) then -- limited parameter sets allowed for these templates
 +
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
  
Output of this function is an error message that identifies the character or the Unicode group, or the stripmarker
+
if 'arxiv' == cite_class then -- basic parameters unique to these templates
that was detected along with its position (or, for multi-byte characters, the position of its first byte) in the
+
state = whitelist.arxiv_basic_arguments[name];
parameter value.
+
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
 +
if false == state then
 +
deprecated_parameter (name); -- parameter is deprecated but still supported
 +
return true;
 +
end
 +
-- limited enumerated parameters list
 +
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
  
local function has_invisible_chars (param, v)
+
return false; -- not supported because not found or name is set to nil
local position = ''; -- position of invisible char or starting position of stripmarker
+
end -- end limited parameter-set templates
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
+
local capture; -- used by stripmarker detection to hold name of the stripmarker
+
state = whitelist.basic_arguments[ name ]; -- all other templates; all normal parameters allowed
local i=1;
+
local stripmarker, apostrophe;
+
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
 +
-- all enumerated parameters allowed
 +
name = name:gsub( "%d+", "#" ); -- replace digit(s) with # (last25 becomes last#
 +
state = whitelist.numbered_arguments[ name ];
  
while cfg.invisible_chars[i] do
+
if true == state then return true; end -- valid actively supported parameter
local char=cfg.invisible_chars[i][1] -- the character or group name
+
if false == state then
local pattern=cfg.invisible_chars[i][2] -- the pattern used to find it
+
deprecated_parameter (name); -- parameter is deprecated but still supported
position, dummy, capture = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
+
return true;
 
if position then
 
-- if 'nowiki' == capture or 'math' == capture or ('ref' == capture and 'quote' == param) then -- nowiki, math, or quote param and ref stripmarker (not an error condition)
 
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
 
elseif 'apostrophe' == char then -- apostrophe template uses &zwj;, hair space and zero-width space
 
apostrophe = true;
 
elseif true == apostrophe and in_array (char, {'zero width joiner', 'zero width space', 'hair space'}) then
 
position = nil; -- unset
 
else
 
local err_msg;
 
if capture then
 
err_msg = capture .. ' ' .. cfg.invisible_chars[i][3] or char;
 
else
 
err_msg = cfg.invisible_chars[i][3] or (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
 
i=i+1; -- bump our index
 
 
end
 
end
 +
 +
return false; -- not supported because not found or name is set to nil
 
end
 
end
  
  
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
+
--[[--------------------------< 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
 +
MMMM DD, YYYY then wrap in nowrap span: <span ...>DD MMMM</span> YYYY or <span ...>MMMM DD,</span> YYYY
  
Argument wrapper.  This function provides support for argument mapping defined in the configuration file so that
+
DOES NOT yet support MMMM YYYY or any of the date ranges.
multiple names can be transparently aliased to single internal variable.
 
  
 
]]
 
]]
  
local function argument_wrapper( args )
+
local function nowrap_date (date)
local origin = {};
+
local cap='';
 +
local cap2='';
 +
 
 +
if date:match("^%d%d%d%d%-%d%d%-%d%d$") then
 +
date = substitute (cfg.presentation['nowrap1'], date);
 
 
return setmetatable({
+
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
ORIGIN = function( self, k )
+
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
local dummy = self[k]; --force the variable to be loaded.
+
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
return origin[k];
+
end
 +
 +
return date;
 +
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
},
+
return title_type; -- if |type= has been set to any other value use that value
{
+
end
__index = function ( tbl, k )
+
 
if origin[k] ~= nil then
+
return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation
return nil;
+
end
end
+
 
+
 
local args, list, v = args, cfg.aliases[k];
+
--[[--------------------------< H Y P H E N _ T O _ D A S H >--------------------------------------------------
 
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
 
  
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
+
Converts a hyphen to a dash
Looks for a parameter's name in the whitelist.
 
  
Parameters in the whitelist can have three values:
 
true - active, supported parameters
 
false - deprecated, supported parameters
 
nil - unsupported parameters
 
 
 
]]
 
]]
  
local function validate( name )
+
local function hyphen_to_dash( str )
local name = tostring( name );
+
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
local state = whitelist.basic_arguments[ name ];
+
return str;
+
end
-- Normal arguments
+
return str:gsub( '-', '–' );
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
 
 
return false; -- Not supported because not found or name is set to nil
 
 
end
 
end
  
  
-- Formats a wiki style internal link
+
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
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
 
  
 
+
Joins a sequence of strings together while checking for duplicate separation characters.
--[[--------------------------< 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
 
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 nowrap_date (date)
+
local function safe_join( tbl, duplicate_char )
local cap='';
+
--[[
local cap2='';
+
Note: we use string functions here, rather than ustring functions.
 
 
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
+
This has considerably faster performance and should work correctly as
cap, cap2 = string.match (date, "^(.*)%s+(%d%d%d%d)$");
+
long as the duplicate_char is strict ASCII. The strings
date = substitute (cfg.presentation['nowrap2'], {cap, cap2});
+
in tbl may be ASCII or UTF8.
end
+
]]
 
 
return date;
+
local str = ''; -- the output string
end
+
local comp = ''; -- what does 'comp' mean?
 
+
local end_chr = '';
--[[--------------------------< IS _ V A L I D _ I S X N >-----------------------------------------------------
+
local trim;
 
+
for _, value in ipairs( tbl ) do
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 value == nil then value = ''; end
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.
+
if str == '' then -- if output string is empty
 
+
str = value; -- assign value to it (first time through the loop)
]]
+
elseif value ~= '' then
 
+
if value:sub(1,1) == '<' then -- Special case of values enclosed in spans and other markup.
local function is_valid_isxn (isxn_str, len)
+
comp = value:gsub( "%b<>", "" ); -- remove html markup (<span>string</span> -> string)
local temp = 0;
+
else
isxn_str = { isxn_str:byte(1, len) }; -- make a table of byte values '0' → 0x30 .. '9'  → 0x39, 'X' → 0x58
+
comp = value;
len = len+1; -- adjust to be a loop counter
+
end
for i, v in ipairs( isxn_str ) do -- loop through all of the bytes and calculate the checksum
+
-- typically duplicate_char is sepc
if v == string.byte( "X" ) then -- if checkdigit is X (compares the byte value of 'X' which is 0x58)
+
if comp:sub(1,1) == duplicate_char then -- is first charactier same as duplicate_char? why test first character?
temp = temp + 10*( len - i ); -- it represents 10 decimal
+
--   Because individual string segments often (always?) begin with terminal punct for th
else
+
--   preceding segment: 'First element' .. 'sepc next element' .. etc?
temp = temp + tonumber( string.char(v) )*(len-i);
+
trim = false;
end
+
end_chr = str:sub(-1,-1); -- get the last character of the output string
end
+
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
return temp % 11 == 0; -- returns true if calculation result is zero
+
if end_chr == duplicate_char then -- if same as separator
end
+
str = str:sub(1,-2); -- remove it
 
+
elseif end_chr == "'" then -- if it might be wikimarkup
 
+
if str:sub(-3,-1) == duplicate_char .. "''" then -- if last three chars of str are sepc''
--[[--------------------------< IS _ V A L I D _ I S X N  _ 1 3 >----------------------------------------------
+
str = str:sub(1, -4) .. "''"; -- remove them and add back ''
 +
elseif str:sub(-5,-1) == duplicate_char .. "]]''" then -- if last five chars of str are sepc]]''
 +
trim = true; -- why? why do this and next differently from previous?
 +
elseif str:sub(-4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''
 +
trim = true; -- same question
 +
end
 +
elseif end_chr == "]" then -- if it might be wikimarkup
 +
if str:sub(-3,-1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink
 +
trim = true;
 +
elseif str:sub(-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
  
ISBN-13 and ISMN validator code calculates checksum across all 13 isbn/ismn digits including the check digit.
+
if trim then
If the number is valid, the result will be 0. Before calling this function, isbn-13/ismn must be checked for length
+
if value ~= comp then -- value does not equal comp when value contains html markup
and stripped of dashes, spaces and other non-isxn-13 characters.
+
local dup2 = duplicate_char;
 
+
if dup2:match( "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it
]]
+
 +
value = value:gsub( "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows html markup
 +
else
 +
value = value:sub( 2, -1 ); -- remove duplicate_char when it is first character
 +
end
 +
end
 +
end
 +
str = str .. value; --add it to the output string
 +
end
 +
end
 +
return str;
 +
end 
  
local function is_valid_isxn_13 (isxn_str)
 
local temp=0;
 
 
isxn_str = { isxn_str:byte(1, 13) }; -- make a table of byte values '0' → 0x30 .. '9'  → 0x39
 
for i, v in ipairs( isxn_str ) do
 
temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) ); -- multiply odd index digits by 1, even index digits by 3 and sum; includes check digit
 
end
 
return temp % 10 == 0; -- sum modulo 10 is zero when isbn-13/ismn is correct
 
end
 
  
--[[--------------------------< C H E C K _ I S B N >------------------------------------------------------------
+
--[[--------------------------< I S _ S U F F I X >------------------------------------------------------------
  
Determines whether an ISBN string is valid
+
returns true is suffix is properly formed Jr, Sr, or ordinal in the range 2–9.  Puncutation not allowed.
  
 
]]
 
]]
  
local function check_isbn( isbn_str )
+
local function is_suffix (suffix)
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
+
if in_array (suffix, {'Jr', 'Sr', '2nd', '3rd'}) or suffix:match ('^%dth$') then
isbn_str = isbn_str:gsub( "-", "" ):gsub( " ", "" ); -- remove hyphens and spaces
+
return true;
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
 +
return false;
 
end
 
end
  
--[[--------------------------< C H E C K _ I S M N >------------------------------------------------------------
 
  
Determines whether an ISMN string is valid.  Similar to isbn-13, ismn is 13 digits begining 979-0-... and uses the
+
--[[--------------------------< I S _ G O O D _ V A N C _ N A M E >--------------------------------------------
same check digit calculations.  See http://www.ismn-international.org/download/Web_ISMN_Users_Manual_2008-6.pdf
 
section 2, pages 9–12.
 
  
]]
+
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.
  
local function ismn (id)
+
This test allows |first= and |last= names to contain any of the letters defined in the four Unicode Latin character sets
local handler = cfg.id_handlers['ISMN'];
+
[http://www.unicode.org/charts/PDF/U0000.pdf C0 Controls and Basic Latin] 0041–005A, 0061–007A
local text;
+
[http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] 00C0–00D6, 00D8–00F6, 00F8–00FF
local valid_ismn = true;
+
[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
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
+
|lastn= also allowed to contain hyphens, spaces, and apostrophes. (http://www.ncbi.nlm.nih.gov/books/NBK7271/box/A35029/)
valid_ismn = false;
+
|firstn= also allowed to contain hyphens, spaces, apostrophes, and periods
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
+
This original test:
-- prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
+
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
+
was written ouside of the code editor and pasted here because the code editor gets confused between character insertion point and cursor position.
text="[[" .. handler.link .. "|" .. handler.label .. "]]" .. handler.separator .. id; -- because no place to link to yet
+
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.
  
if false == valid_ismn then
+
\195\128-\195\150 – À-Ö (U+00C0–U+00D6 – C0 controls)
text = text .. ' ' .. set_error( 'bad_ismn' ) -- add an error message if the issn is invalid
+
\195\152-\195\182 – Ø- ö (U+00D8-U+00F6 – C0 controls)
end
+
\195\184-\198\191 – ø-ƿ (U+00F8-U+01BF – C0 controls, Latin extended A & B)
+
\199\132-\201\143 – DŽ-ɏ (U+01C4-U+024F – Latin extended B)
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
+
local function is_good_vanc_name (last, first)
digits with a space.  When that condition occurred, the resulting link looked like this:
+
local first, suffix = first:match ('(.-),?%s*([%dJS][%drndth]+)%.?$') or first; -- if first has something that looks like a generational suffix, get it
  
|issn=0819 4327 gives: [http://www.worldcat.org/issn/0819 4327 0819 4327]  -- can't have spaces in an external link
+
if is_set (suffix) then
+
if not is_suffix (suffix) then
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
+
add_vanc_error ('suffix');
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
+
return false; -- not a name with an appropriate suffix
error message. The issn is always displayed with a hyphen, even if the issn was given as a single group of 8 digits.
+
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
  
]]
 
  
local function issn(id, e)
+
--[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------
local issn_copy = id; -- save a copy of unadulterated issn; use this version for display if issn does not validate
 
local handler;
 
local text;
 
local valid_issn = true;
 
 
if e then
 
handler = cfg.id_handlers['EISSN'];
 
else
 
handler = cfg.id_handlers['ISSN'];
 
end
 
  
id=id:gsub( "[%s- –]", "" ); -- strip spaces, hyphens, and endashes from the issn
+
Attempts to convert names to initials in support of |name-list-format=vanc. 
  
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
+
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/.
valid_issn=false; -- wrong length or improper character
 
else
 
valid_issn=is_valid_isxn(id, 8); -- validate issn
 
end
 
  
if true == valid_issn then
+
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/.
id = string.sub( id, 1, 4 ) .. "-" .. string.sub( id, 5 ); -- if valid, display correctly formatted version
+
This code only accepts and understands generational suffix in the Vancouver format because Roman numerals look like, and can be mistaken for, initials.
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 >------------------------------------------------------------------
+
This function uses ustring functions because firstname initials may be any of the unicode Latin characters accepted by is_good_vanc_name ().
 
 
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 function reduce_to_initials(first)
local err_cat = ""
+
local name, suffix = mw.ustring.match(first, "^(%u+) ([%dJS][%drndth]+)$");
  
if not id:match("^[%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u]$") then
+
if not name then -- if not initials and a suffix
err_cat =  ' ' .. set_error ('bad_asin'); -- asin is not a mix of 10 uppercase alpha and numeric characters
+
name = mw.ustring.match(first, "^(%u+)$"); -- is it just intials?
else
+
end
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
+
if name then -- if first is initials with or without suffix
add_maint_cat ('ASIN');
+
if 3 > mw.ustring.len (name) then -- if one or two initials
elseif not is_set (err_cat) then
+
if suffix then -- if there is a suffix
err_cat =  ' ' .. set_error ('bad_asin'); -- asin is not isbn10
+
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
 +
return first; -- one or two initials without suffix; nothing to do
 
end
 
end
elseif not id:match("^%u[%d%u]+$") then
 
err_cat =  ' ' .. set_error ('bad_asin'); -- asin doesn't begin with uppercase alpha
 
 
end
 
end
end
+
end -- if here then name has 3 or more uppercase letters so treat them as a word
if not is_set(domain) then
 
domain = "com";
 
elseif in_array (domain, {'jp', 'uk'}) then -- Japan, United Kingdom
 
domain = "co." .. domain;
 
elseif in_array (domain, {'au', 'br', 'mx'}) then -- Australia, Brazil, Mexico
 
domain = "com." .. domain;
 
end
 
local handler = cfg.id_handlers['ASIN'];
 
return external_link_id({link=handler.link,
 
label=handler.label, prefix=handler.prefix .. domain .. "/dp/",
 
id=id, encode=handler.encode, separator = handler.separator}) .. err_cat;
 
end
 
 
 
--[[--------------------------< A R X I V >--------------------------------------------------------------------
 
  
See: http://arxiv.org/help/arxiv_identifier
 
  
format and error check arXiv identifier.  There are three valid forms of the identifier:
+
local initials, names = {}, {}; -- tables to hold name parts and initials
the first form, valid only between date codes 9108 and 0703 is:
+
local i = 1; -- counter for number of initials
arXiv:<archive>.<class>/<date code><number><version>
 
where:
 
<archive> is a string of alpha characters - may be hyphenated; no other punctuation
 
<class> is a string of alpha characters - may be hyphenated; no other punctuation
 
<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
 
first digit of YY for this form can only 9 and 0
 
<number> is a three-digit number
 
<version> is a 1 or more digit number preceded with a lowercase v; no spaces (undocumented)
 
 
the second form, valid from April 2007 through December 2014 is:
 
arXiv:<date code>.<number><version>
 
where:
 
<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
 
<number> is a four-digit number
 
<version> is a 1 or more digit number preceded with a lowercase v; no spaces
 
  
the third form, valid from January 2015 is:
+
names = mw.text.split (first, '[%s,]+'); -- split into a table of names and possible suffix
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)
+
while names[i] do -- loop through the table
local handler = cfg.id_handlers['ARXIV'];
+
if 1 < i and names[i]:match ('[%dJS][%drndth]+%.?$') then -- if not the first name, and looks like a suffix (may have trailing dot)
local year, month, version;
+
names[i] = names[i]:gsub ('%.', ''); -- remove terminal dot if present
local err_cat = '';
+
if is_suffix (names[i]) then -- if a legitimate suffix
local text;
+
table.insert (initials, ' ' .. names[i]); -- add a separator space, insert at end of initials table
+
break; -- and done because suffix must fall at the end of a name
if id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%d$") or id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%dv%d+$") then -- test for the 9108-0703 format w/ & w/o version
+
end -- no error message if not a suffix; possibly because of Romanization
year, month = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d[v%d]*$");
 
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
 
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
+
if 3 > i then
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d[v%d]*$");
+
table.insert (initials, mw.ustring.sub(names[i],1,1)); -- insert the intial at end of initials table
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
 
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
+
i = i+1; -- bump the counter
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d%d[v%d]*$");
 
year = tonumber(year);
 
month = tonumber(month);
 
if ((15 > year) or (1 > month or 12 < month)) then -- is year invalid or is month invalid? (doesn't test for future years)
 
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
 
end
 
else
 
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- arXiv id doesn't match any format
 
 
end
 
end
 +
 +
return table.concat(initials) -- Vancouver format does not include spaces.
 +
end
  
text = external_link_id({link = handler.link, label = handler.label,
 
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 
  
if is_set (class) then
+
--[[--------------------------< L I S T  _ P E O P L E >-------------------------------------------------------
class = ' [[' .. '//arxiv.org/archive/' .. class .. ' ' .. class .. ']]'; -- external link within square brackets, not wikilink
 
else
 
class = ''; -- empty string for concatenation
 
end
 
 
return text .. class;
 
end
 
  
--[[
+
Formats a list of people (e.g. authors / editors)  
lccn normalization (http://www.loc.gov/marc/lccn-namespace.html#normalization)
 
1. Remove all blanks.
 
2. If there is a forward slash (/) in the string, remove it, and remove all characters to the right of the forward slash.
 
3. If there is a hyphen in the string:
 
a. Remove it.
 
b. Inspect the substring following (to the right of) the (removed) hyphen. Then (and assuming that steps 1 and 2 have been carried out):
 
1. All these characters should be digits, and there should be six or less. (not done in this function)
 
2. If the length of the substring is less than 6, left-fill the substring with zeroes until the length is six.
 
  
Returns a normalized lccn for lccn() to validate.  There is no error checking (step 3.b.1) performed in this function.
 
 
]]
 
]]
  
local function normalize_lccn (lccn)
+
local function list_people(control, people, etal)
lccn = lccn:gsub ("%s", ""); -- 1. strip whitespace
+
local sep;
 +
local namesep;
 +
local format = control.format
 +
local maximum = control.maximum
 +
local lastauthoramp = control.lastauthoramp;
 +
local text = {}
  
if nil ~= string.find (lccn,'/') then
+
if 'vanc' == format then -- Vancouver-like author/editor name styling?
lccn = lccn:match ("(.-)/"); -- 2. remove forward slash and all character to the right of it
+
sep = ','; -- name-list separator between authors is a comma
end
+
namesep = ' '; -- last/first separator is a space
 
 
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
 
 
 
--[[
 
Format LCCN link and do simple error checking.  LCCN is a character string 8-12 characters long. The length of the LCCN dictates the character type of the first 1-3 characters; the
 
rightmost eight are always digits. http://info-uri.info/registry/OAIHandler?verb=GetRecord&metadataPrefix=reg&identifier=info:lccn/
 
 
 
length = 8 then all digits
 
length = 9 then lccn[1] is lower case alpha
 
length = 10 then lccn[1] and lccn[2] are both lower case alpha or both digits
 
length = 11 then lccn[1] is lower case alpha, lccn[2] and lccn[3] are both lower case alpha or both digits
 
length = 12 then lccn[1] and lccn[2] are both lower case alpha
 
 
 
]]
 
 
 
local function lccn(lccn)
 
local handler = cfg.id_handlers['LCCN'];
 
local err_cat =  ''; -- presume that LCCN is valid
 
local id = lccn; -- local copy of the lccn
 
 
 
id = normalize_lccn (id); -- get canonical form (no whitespace, hyphens, forward slashes)
 
local len = id:len(); -- get the length of the lccn
 
 
 
if 8 == len then
 
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits)
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
 
end
 
elseif 9 == len then -- LCCN should be adddddddd
 
if nil == id:match("%l%d%d%d%d%d%d%d%d") then -- does it match our pattern?
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
 
end
 
elseif 10 == len then -- LCCN should be aadddddddd or dddddddddd
 
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits) ...
 
if nil == id:match("^%l%l%d%d%d%d%d%d%d%d") then -- ... see if it matches our pattern
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 
end
 
end
 
elseif 11 == len then -- LCCN should be aaadddddddd or adddddddddd
 
if not (id:match("^%l%l%l%d%d%d%d%d%d%d%d") or id:match("^%l%d%d%d%d%d%d%d%d%d%d")) then -- see if it matches one of our patterns
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 
end
 
elseif 12 == len then -- LCCN should be aadddddddddd
 
if not id:match("^%l%l%d%d%d%d%d%d%d%d%d%d") then -- see if it matches our pattern
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
 
end
 
 
else
 
else
err_cat = ' ' .. set_error( 'bad_lccn' ); -- wrong length, set an error message
+
sep = ';' -- name-list separator between authors is a semicolon
 +
namesep = ', ' -- last/first separator is <comma><space>
 
end
 
end
 
if not is_set (err_cat) and nil ~= lccn:find ('%s') then
 
err_cat = ' ' .. set_error( 'bad_lccn' ); -- lccn contains a space, set an error message
 
end
 
 
return external_link_id({link = handler.link, label = handler.label,
 
prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
 
end
 
 
--[[
 
Format PMID and do simple error checking.  PMIDs are sequential numbers beginning at 1 and counting up.  This code checks the PMID to see that it
 
contains only digits and is less than test_limit; the value in local variable test_limit will need to be updated periodically as more PMIDs are issued.
 
]]
 
 
local function pmid(id)
 
local test_limit = 30000000; -- update this value as PMIDs approach
 
local handler = cfg.id_handlers['PMID'];
 
local err_cat =  ''; -- presume that PMID is valid
 
 
 
if id:match("[^%d]") then -- if PMID has anything but digits
+
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
err_cat = ' ' .. set_error( 'bad_pmid' ); -- set an error message
+
if is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for authors
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,
+
for i,person in ipairs(people) do
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
+
if is_set(person.last) then
end
+
local mask = person.mask
 
+
local one
--[[--------------------------< I S _ E M B A R G O E D >------------------------------------------------------
+
local sep_one = sep;
 
+
if is_set (maximum) and i > maximum then
Determines if a PMC identifier's online version is embargoed. Compares the date in |embargo= against today's date.  If embargo date is
+
etal = true;
in the future, returns the content of |embargo=; otherwise, returns and empty string because the embargo has expired or because
+
break;
|embargo= was not set in this cite.
+
elseif (mask ~= nil) then
 
+
local n = tonumber(mask)
]]
+
if (n ~= nil) then
 
+
one = string.rep("&mdash;",n)
local function is_embargoed (embargo)
+
else
if is_set (embargo) then
+
one = mask;
local lang = mw.getContentLanguage();
+
sep_one = " ";
local good1, embargo_date, good2, todays_date;
+
end
good1, embargo_date = pcall( lang.formatDate, lang, 'U', embargo );
 
good2, todays_date = pcall( lang.formatDate, lang, 'U' );
 
 
if good1 and good2 then -- if embargo date and today's date are good dates
 
if tonumber( embargo_date ) >= tonumber( todays_date ) then -- is embargo date is in the future?
 
return embargo; -- still embargoed
 
 
else
 
else
add_maint_cat ('embargo')
+
one = person.last
return ''; -- unset because embargo has expired
+
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 = make_wikilink (person.link, one); -- link author/editor if this page is not the author's/editor's page
 +
end
 
end
 
end
 +
table.insert( text, one )
 +
table.insert( text, sep_one )
 
end
 
end
 
end
 
end
return ''; -- |embargo= not set return empty string
 
end
 
  
--[[--------------------------< P M C >------------------------------------------------------------------------
+
local count = #text / 2; -- (number of names + number of separators) divided by 2
 
+
if count > 0 then
Format a PMC, do simple error checking, and check for embargoed articles.
+
if count > 1 and is_set(lastauthoramp) and not etal then
 
+
text[#text-2] = " & "; -- replace last separator with ampersand text
The embargo parameter takes a date for a value. If the embargo date is in the future the PMC identifier will not
+
end
be linked to the article. If the embargo date is today or in the past, or if it is empty or omitted, then the
+
text[#text] = nil; -- erase the last separator
PMC identifier is linked to the article through the link at cfg.id_handlers['PMC'].prefix.
+
end
 +
 +
local result = table.concat(text) -- construct list
 +
if etal and is_set (result) then -- etal may be set by |display-authors=etal but we might not have a last-first list
 +
result = result .. sep .. ' ' .. cfg.messages['et al']; -- we've go a last-first list and etal so add et al.
 +
end
 +
 +
return result, count
 +
end
  
PMC embargo date testing is done in function is_embargoed () which is called earlier because when the citation
 
has |pmc=<value> but does not have a |url= then |title= is linked with the PMC link.  Function is_embargoed ()
 
returns the embargo date if the PMC article is still embargoed, otherwise it returns an empty string.
 
  
PMCs are sequential numbers beginning at 1 and counting up.  This code checks the PMC to see that it contains only digits and is less
+
--[[--------------------------< A N C H O R _ I D >------------------------------------------------------------
than test_limit; the value in local variable test_limit will need to be updated periodically as more PMCs are issued.
 
  
]]
+
Generates a CITEREF anchor ID if we have at least one name or a date.  Otherwise returns an empty string.
  
local function pmc(id, embargo)
+
namelist is one of the contributor-, author-, or editor-name lists chosen in that orderyear is Year or anchor_year.
local test_limit = 6000000; -- update this value as PMCs approach
 
local handler = cfg.id_handlers['PMC'];
 
local err_cat = ''; -- presume that PMC is valid
 
 
local text;
 
  
if id:match("[^%d]") then -- if PMC has anything but digits
+
]]
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
+
local function anchor_id (namelist, year)
else -- PMC is only digits
+
local names={}; -- a table for the one to four names and year
local id_num = tonumber(id); -- convert id to a number for range testing
+
for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names
if 1 > id_num or test_limit < id_num then -- if PMC is outside test limit boundaries
+
names[i] = v.last
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
+
if i == 4 then break end -- if four then done
end
 
 
end
 
end
+
table.insert (names, year); -- add the year at the end
if is_set (embargo) then -- is PMC is still embargoed?
+
local id = table.concat(names); -- concatenate names and year for CITEREF id
text="[[" .. handler.link .. "|" .. handler.label .. "]]:" .. handler.separator .. id .. err_cat; -- still embargoed so no external link
+
if is_set (id) then -- if concatenation is not an empty string
 +
return "CITEREF" .. id; -- add the CITEREF portion
 
else
 
else
text = external_link_id({link = handler.link, label = handler.label, -- no embargo date or embargo has expired, ok to link to article
+
return ''; -- return an empty string; no reason to include CITEREF id in this citation
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
 
 
end
 
end
return text;
 
 
end
 
end
  
-- Formats a DOI and checks for DOI errors.
 
  
-- DOI names contain two parts: prefix and suffix separated by a forward slash.
+
--[[--------------------------< N A M E _ H A S _ E T A L >----------------------------------------------------
-- Prefix: directory indicator '10.' followed by a registrant code
 
-- Suffix: character string of any length chosen by the registrant
 
  
-- This function checks a DOI name for: prefix/suffix.  If the doi name contains spaces or endashes,
+
Evaluates the content of author and editor name parameters for variations on the theme of et al.  If found,
-- or, if it ends with a period or a comma, this function will emit a bad_doi error message.
+
the et al. is removed, a flag is set to true and the function returns the modified name and the flag.
  
-- DOI names are case-insensitive and can incorporate any printable Unicode characters so the test for spaces, endash,
+
This function never sets the flag to false but returns it's previous state because it may have been set by
-- and terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely if ever used in doi names.
+
previous passes through this function or by the parameters |display-authors=etal or |display-editors=etal
  
local function doi(id, inactive)
+
]]
local cat = ""
+
 
local handler = cfg.id_handlers['DOI'];
+
local function name_has_etal (name, etal, nocat)
+
 
local text;
+
if is_set (name) then -- name can be nil in which case just return
if is_set(inactive) then
+
local etal_pattern = "[;,]? *[\"']*%f[%a][Ee][Tt] *[Aa][Ll][%.\"']*$" -- variations on the 'et al' theme
local inactive_year = inactive:match("%d%d%d%d") or ''; -- try to get the year portion from the inactive date
+
local others_pattern = "[;,]? *%f[%a]and [Oo]thers"; -- and alternate to et al.
text = "[[" .. handler.link .. "|" .. handler.label .. "]]:" .. id;
+
if is_set(inactive_year) then
+
if name:match (etal_pattern) then -- variants on et al.
table.insert( z.error_categories, "自" .. inactive_year .. "年含有不活躍DOI的頁面");
+
name = name:gsub (etal_pattern, ''); -- if found, remove
else
+
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
table.insert( z.error_categories, "含有不活躍DOI的頁面" ); -- when inactive doesn't contain a recognizable year
+
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
inactive = " (" .. cfg.messages['inactive'] .. " " .. inactive .. ")"
 
else
 
text = external_link_id({link = handler.link, label = handler.label,
 
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
 
inactive = ""
 
 
end
 
end
 
+
return name, etal; --  
if nil == id:match("^10%.[^%s–]-/[^%s–]-[^%.,]$") then -- doi must begin with '10.', must contain a fwd slash, must not contain spaces or endashes, and must not end with period or comma
 
cat = ' ' .. set_error( 'bad_doi' );
 
end
 
return text .. inactive .. cat
 
 
end
 
end
  
  
--[[--------------------------< O P E N L I B R A R Y >--------------------------------------------------------
+
--[[--------------------------< N A M E _ H A S _ E D _ M A R K U P >------------------------------------------
  
Formats an OpenLibrary link, and checks for associated errors.
+
Evaluates the content of author and editor parameters for extranious editor annotations: ed, ed., eds, (Ed.), etc.
 +
These annotation do not belong in author parameters and are redundant in editor parameters.  If found, the function
 +
adds the editor markup maintenance category.
  
 
]]
 
]]
local function openlibrary(id)
 
local code = id:match("^%d+([AMW])$"); -- only digits followed by 'A', 'M', or 'W'
 
local handler = cfg.id_handlers['OL'];
 
  
if ( code == "A" ) then
+
local function name_has_ed_markup (name, list_name)
return external_link_id({link=handler.link, label=handler.label,
+
local _, pattern;
prefix=handler.prefix .. 'authors/OL',
+
local patterns = { -- these patterns match annotations at end of name
id=id, separator=handler.separator, encode = handler.encode})
+
'%f[%(%[][%(%[]%s*[Ee][Dd][Ss]?%.?%s*[%)%]]?$', -- (ed) or (eds): leading '(', case insensitive 'ed', optional 's', '.' and/or ')'
elseif ( code == "M" ) then
+
'[,%.%s]%f[e]eds?%.?$', -- ed or eds: without '('or ')'; case sensitive (ED could be initials Ed could be name)
return external_link_id({link=handler.link, label=handler.label,
+
'%f[%(%[][%(%[]%s*[Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%.?%s*[%)%]]?$', -- (editor) or (editors): leading '(', case insensitive, optional '.' and/or ')'
prefix=handler.prefix .. 'books/OL',
+
'[,%.%s]%f[Ee][Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%.?$', -- editor or editors: without '('or ')'; case insensitive
id=id, separator=handler.separator, encode = handler.encode})
+
elseif ( code == "W" ) then
+
-- these patterns match annotations at beginning of name
return external_link_id({link=handler.link, label=handler.label,
+
'^eds?[%.,;]', -- ed. or eds.: lower case only, optional 's', requires '.'
prefix=handler.prefix .. 'works/OL',
+
'^[%(%[]%s*[Ee][Dd][Ss]?%.?%s*[%)%]]', -- (ed) or (eds): also sqare brackets, case insensitive, optional 's', '.'
id=id, separator=handler.separator, encode = handler.encode})
+
'^[%(%[]?%s*[Ee][Dd][Ii][Tt][Oo][Rr][Ss]?%A', -- (editor or (editors: also sq brackets, case insensitive, optional brackets, 's'
else
+
'^[%(%[]?%s*[Ee][Dd][Ii][Tt][Ee][Dd]%A', -- (edited: also sq brackets, case insensitive, optional brackets
return external_link_id({link=handler.link, label=handler.label,
+
}
prefix=handler.prefix .. 'OL',
+
 
id=id, separator=handler.separator, encode = handler.encode}) .. ' ' .. set_error( 'bad_ol' );
+
if is_set (name) then
 +
for _, pattern in ipairs (patterns) do -- spin through patterns table and
 +
if name:match (pattern) then
 +
add_maint_cat ('extra_text_names', cfg.special_case_translation [list_name]); -- add a maint cat for this template
 +
break;
 +
end
 +
end
 
end
 
end
 +
return name; -- and done
 
end
 
end
  
  
--[[--------------------------< M E S S A G E _ I D >----------------------------------------------------------
+
--[[--------------------------< N A M E _ H A S _ M U L T _ N A M E S >----------------------------------------
  
Validate and format a usenet message idSimple error checking, looks for 'id-left@id-right' not enclosed in
+
Evaluates the content of author and editor (surnames only) parameters for multiple names.  Multiple names are
'<' and/or '>' angle brackets.
+
indicated if there is more than one comma and or semicolonIf found, the function adds the multiple name
 +
(author or editor) maintenance category.
  
 
]]
 
]]
  
local function message_id (id)
+
local function name_has_mult_names (name, list_name)
local handler = cfg.id_handlers['USENETID'];
+
local count, _;
 
+
if is_set (name) then
text = external_link_id({link = handler.link, label = handler.label,
+
_, count = name:gsub ('[;,]', ''); -- count the number of separator-like characters
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
+
+
if 1 < count then -- param could be |author= or |editor= so one separator character is acceptable
if not id:match('^.+@.+$') or not id:match('^[^<].*[^>]$')then -- doesn't have '@' or has one or first or last character is '< or '>'
+
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
text = text .. ' ' .. set_error( 'bad_message_id' ) -- add an error message if the message id is invalid
+
end
end  
+
end
+
return name; -- and done
return text
 
 
end
 
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.
+
--[[--------------------------< N A M E _ C H E C K S >--------------------------------------------------------
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
+
This function calls various name checking functions used to validate the content of the various name-holding
 +
parameters.
  
 
]]
 
]]
  
local function set_titletype (cite_class, title_type)
+
local function name_checks (last, first, list_name)
if is_set(title_type) then
+
if is_set (last) then
if "none" == title_type then
+
if last:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
title_type = ""; -- if |type=none then type parameter not displayed
+
last = last:match ('^%(%((.*)%)%)$'); -- strip parens
 +
else
 +
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
 +
if is_set (first) then
 +
if first:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, accept as written
 +
first = first:match ('^%(%((.*)%)%)$'); -- strip parens
 +
else
 +
first = name_has_ed_markup (first, list_name); -- check for extraneous 'editor' annotation
 
end
 
end
return title_type; -- if |type= has been set to any other value use that value
+
end
 +
return last, first; -- done
 
end
 
end
  
return cfg.title_types [cite_class] or ''; -- set template's default title type; else empty string for concatenation
 
end
 
  
--[[--------------------------< C L E A N _ I S B N >----------------------------------------------------------
+
--[[--------------------------< E X T R A C T _ N A M E S >----------------------------------------------------
 +
Gets name list from the input arguments
  
Removes irrelevant text and dashes from ISBN number
+
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
Similar to that used for Special:BookSources
+
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=.
  
local function clean_isbn( isbn_str )
+
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
return isbn_str:gsub( "[^-0-9X]", "" );
+
that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration.  This keeps 'et al.' out of the
end
+
template's metadata.  When this occurs, the page is added to a maintenance category.
 
 
--[[--------------------------< 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 escape_lua_magic_chars (argument)
 
argument = argument:gsub("%%", "%%%%"); -- replace % with %%
 
argument = argument:gsub("([%^%$%(%)%.%[%]%*%+%-%?])", "%%%1"); -- replace all other lua magic pattern characters
 
return argument;
 
end
 
  
--[[--------------------------< S T R I P _ A P O S T R O P H E _ M A R K U P >--------------------------------
+
local function extract_names(args, list_name)
 +
local names = {}; -- table of names
 +
local last; -- individual name components
 +
local first;
 +
local link;
 +
local mask;
 +
local i = 1; -- loop counter/indexer
 +
local n = 1; -- output table indexer
 +
local count = 0; -- used to count the number of times we haven't found a |last= (or alias for authors, |editor-last or alias for editors)
 +
local etal=false; -- return value set to true when we find some form of et al. in an author parameter
  
Strip wiki italic and bold markup from argument so that it doesn't contaminate COinS metadata.
+
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
This function strips common patterns of apostrophe markup. We presume that editors who have taken the time to
+
while true do
markup a title have, as a result, provided valid markup. When they don't, some single apostrophes are left behind.
+
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.
 +
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
  
local function strip_apostrophe_markup (argument)
+
names[n] = {last = last, first = first, link = link, mask = mask, corporate=false}; -- add this name to our names list (corporate for |vauthors= only)
if not is_set (argument) then return argument; end
+
n = n + 1; -- point to next location in the names table
 
+
if 1 == count then -- if the previous name was missing
while true do
+
table.insert( z.message_tail, { set_error( 'missing_name', {err_msg_list_name, i-1}, true ) } ); -- add this error message
if argument:match ("%'%'%'%'%'") then -- bold italic (5)
+
end
argument=argument:gsub("%'%'%'%'%'", ""); -- remove all instances of it
+
count = 0; -- reset the counter, we're looking for two consecutive missing names
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
 +
i = i + 1; -- point to next args location
 
end
 
end
return argument; -- done
+
 +
return names, etal; -- all done, return our list of names
 
end
 
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)
+
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
  
Apostrophe markup (bold, italics) is stripped from each value so that the COinS metadata isn't correupted with strings
+
Validates language names provided in |language= parameter if not an ISO639-1 or 639-2 code.
of %27%27...
+
 
]]
+
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.
  
local function make_coins_title (title, script)
+
mw.language.fetchLanguageNames(<local wiki language>, 'all') returns a list of languages that in some cases may include
if is_set (title) then
+
extensions. For example, code 'cbk-zam' and its associated name 'Chavacano de Zamboanga' (MediaWiki does not support
title = strip_apostrophe_markup (title); -- strip any apostrophe markup
+
code 'cbk' or name 'Chavacano'.  Most (all?) of these languages are not used a 'language' codes per se, rather they
else
+
are used as sub-domain names: cbk-zam.wikipedia.org.  These names can be found (for the time being) at
title=''; -- if not set, make sure title is an empty string
+
https://phabricator.wikimedia.org/diffusion/ECLD/browse/master/LocalNames/LocalNamesEn.php
end
 
if is_set (script) then
 
script = script:gsub ('^%l%l%s*:%s*', ''); -- remove language prefix if present (script value may now be empty string)
 
script = strip_apostrophe_markup (script); -- strip any apostrophe markup
 
else
 
script=''; -- if not set, make sure script is an empty string
 
end
 
if is_set (title) and is_set (script) then
 
script = ' ' .. script; -- add a space before we concatenate
 
end
 
return title .. script; -- return the concatenation
 
end
 
  
--[[--------------------------< G E T _ C O I N S _ P A G E S >------------------------------------------------
+
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.
  
Extract page numbers from external wikilinks in any of the |page=, |pages=, or |at= parameters for use in COinS.
+
Adapted from code taken from Module:Check ISO 639-1.
  
 
]]
 
]]
  
local function get_coins_pages (pages)
+
local function get_iso639_code (lang, this_wiki_code)
local pattern;
+
local remap = {
if not is_set (pages) then return pages; end -- if no page numbers then we're done
+
['bangla'] = {'Bengali', 'bn'}, -- MediaWiki returns Bangla (the endonym) but we want Bengali (the exonym); here we remap
 +
['bengali'] = {'Bengali', 'bn'}, -- MediaWiki doesn't use exonym so here we provide correct language name and 639-1 code
 +
['bihari'] = {'Bihari', 'bh'}, -- MediaWiki replace 'Bihari' with 'Bhojpuri' so 'Bihari' cannot be found
 +
['bhojpuri'] = {'Bhojpuri', 'bho'}, -- MediaWiki uses 'bh' as a subdomain name for Bhojpuri wWikipedia: bh.wikipedia.org
 +
}
 
 
while true do
+
if remap[lang:lower()] then
pattern = pages:match("%[(%w*:?//[^ ]+%s+)[%w%d].*%]"); -- pattern is the opening bracket, the url and following space(s): "[url "
+
return remap[lang:lower()][1], remap[lang:lower()][2]; -- for this language 'name', return a possibly new name and appropriate code
if nil == pattern then break; end -- no more urls
+
end
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
+
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
 
end
pages = pages:gsub("[%[%]]", ""); -- remove the brackets
+
return lang; -- not valid language; return language in original case and nil for the code
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
 
end
  
-- Gets the display text for a wikilink like [[A|B]] or [[B]] gives B
 
local function remove_wiki_link( str )
 
return (str:gsub( "%[%[([^%[%]]*)%]%]", function(l)
 
return l:gsub( "^[^|]*|(.*)$", "%1" ):gsub("^%s*(.-)%s*$", "%1");
 
end));
 
end
 
  
-- Converts a hyphen to a dash
+
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
local function hyphen_to_dash( str )
 
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
 
return str;
 
end
 
return str:gsub( '-', '–' );
 
end
 
  
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
+
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.
  
Joins a sequence of strings together while checking for duplicate separation characters.
+
When |language= contains a recognized language (either code or name), the page is assigned to the category for
 +
that code: Category:Norwegian-language sources (no).  For valid three-character code languages, the page is assigned
 +
to the single category for '639-2' codes: Category:CS1 ISO 639-2 language sources.
  
]]
+
Languages that are the same as the local wiki are not categorized.  MediaWiki does not recognize three-character
 +
equivalents of two-character codes: code 'ar' is recognized bit code 'ara' is not.
  
local function safe_join( tbl, duplicate_char )
+
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.
Note: we use string functions here, rather than ustring functions.
+
 
 +
]]
 +
 
 +
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
 
 
This has considerably faster performance and should work correctly as
+
if is_set (name) then -- if |language= specified a valid code
long as the duplicate_char is strict ASCII.  The strings
+
code = lang:lower(); -- save it
in tbl may be ASCII or UTF8.
+
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
 
 
local str = ''; -- the output string
+
if is_set (code) then -- only 2- or 3-character codes
local comp = ''; -- what does 'comp' mean?
+
name = remap[code] or name; -- override wikimedia when they misuse language codes/names
local end_chr = '';
+
 
local trim;
+
if this_wiki_code ~= code then -- when the language is not the same as this wiki's language
for _, value in ipairs( tbl ) do
+
if 2 == code:len() then -- and is a two-character code
if value == nil then value = ''; end
+
add_prop_cat ('foreign_lang_source' .. code, {name, code}) -- categorize it
 +
else -- or is a recognized language (but has a three-character code)
 +
add_prop_cat ('foreign_lang_source_2' .. code, {code}) -- categorize it differently TODO: support mutliple three-character code categories per cs1|2 template
 +
end
 +
end
 +
else
 +
add_maint_cat ('unknown_lang'); -- add maint category if not already added
 +
end
 
 
if str == '' then -- if output string is empty
+
table.insert (language_list, name);
str = value; -- assign value to it (first time through the loop)
+
name = ''; -- so we can reuse it
elseif value ~= '' then
+
end
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)
+
code = #language_list -- reuse code as number of languages in the list
else
+
if 2 >= code then
comp = value;
+
name = table.concat (language_list, ' and ') -- insert '<space>and<space>' between two language names
end
+
elseif 2 < code then
-- typically duplicate_char is sepc
+
language_list[code] = 'and ' .. language_list[code]; -- prepend last name with 'and<space>'
if comp:sub(1,1) == duplicate_char then -- is first charactier same as duplicate_char? why test first character?
+
name = table.concat (language_list, ', ') -- and concatenate with '<comma><space>' separators
--  Because individual string segments often (always?) begin with terminal punct for th
+
end
--  preceding segment: 'First element' .. 'sepc next element' .. etc?
+
if this_wiki_name == name then
trim = false;
+
return ''; -- if one language and that language is this wiki's return an empty string (no annotation)
end_chr = str:sub(-1,-1); -- get the last character of the output string
+
end
-- str = str .. "<HERE(enchr=" .. end_chr.. ")" -- debug stuff?
+
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
if end_chr == duplicate_char then -- if same as separator
+
--[[ TODO: should only return blank or name rather than full list
str = str:sub(1,-2); -- remove it
+
so we can clean up the bunched parenthetical elements Language, Type, Format
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''
+
end
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?
+
--[[--------------------------< S E T _ C S 1 _ S T Y L E >----------------------------------------------------
elseif str:sub(-4,-1) == duplicate_char .. "]''" then -- if last four chars of str are sepc]''
+
 
trim = true; -- same question
+
Set style settings for CS1 citation templates. Returns separator and postscript settings
end
+
 
elseif end_chr == "]" then -- if it might be wikimarkup
+
]]
if str:sub(-3,-1) == duplicate_char .. "]]" then -- if last three chars of str are sepc]] wikilink
 
trim = true;
 
elseif str:sub(-2,-1) == duplicate_char .. "]" then -- if last two chars of str are sepc] external link
 
trim = true;
 
elseif str:sub(-4,-1) == duplicate_char .. "'']" then -- normal case when |url=something & |title=Title.
 
trim = true;
 
end
 
elseif end_chr == " " then -- if last char of output string is a space
 
if str:sub(-2,-1) == duplicate_char .. " " then -- if last two chars of str are <sepc><space>
 
str = str:sub(1,-3); -- remove them both
 
end
 
end
 
  
if trim then
+
local function set_cs1_style (ps)
if value ~= comp then -- value does not equal comp when value contains html markup
+
if not is_set (ps) then -- unless explicitely set to something
local dup2 = duplicate_char;
+
ps = '.'; -- terminate the rendered citation with a period
if dup2:match( "%A" ) then dup2 = "%" .. dup2; end -- if duplicate_char not a letter then escape it
+
end
+
return '.', ps; -- separator is a full stop
value = value:gsub( "(%b<>)" .. dup2, "%1", 1 ) -- remove duplicate_char if it follows html markup
+
end
else
 
value = value:sub( 2, -1 ); -- remove duplicate_char when it is first character
 
end
 
end
 
end
 
str = str .. value; --add it to the output string
 
end
 
end
 
return str;
 
end
 
  
--[[--------------------------< I S _ G O O D _ V A N C _ N A M E >--------------------------------------------
 
  
For Vancouver Style, author/editor names are supposed to be rendered in Latin (read ASCII) characters.  When a name
+
--[[--------------------------< S E T _ C S 2 _ S T Y L E >----------------------------------------------------
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.
 
  
This test allows |first= and |last= names to contain any of the letters defined in the four Unicode Latin character sets
+
Set style settings for CS2 citation templates. Returns separator, postscript, ref settings
[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/)
 
|firstn= also allowed to contain hyphens, spaces, apostrophes, and periods
 
 
 
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
 
because the code editor gets confused between character insertion point and cursor position.
 
  
 
]]
 
]]
  
local function is_good_vanc_name (last, first)
+
local function set_cs2_style (ps, ref)
if nil == mw.ustring.find (last, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ-ɏ%-%s%']*$") or nil == mw.ustring.find (first, "^[A-Za-zÀ-ÖØ-öø-ƿDŽ- ɏ%-%s%'%.]*$") then
+
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
add_vanc_error ();
+
ps = ''; -- make sure it isn't nil
return false; -- not a string of latin characters; Vancouver required Romanization
+
end
end;
+
if not is_set (ref) then -- if |ref= is not set
return true;
+
ref = "harv"; -- set default |ref=harv
 +
end
 +
return ',', ps, ref; -- separator is a comma
 
end
 
end
  
--[[--------------------------< R E D U C E _ T O _ I N I T I A L S >------------------------------------------
 
  
Attempts to convert names to initials in support of |name-list-format=vanc. 
+
--[[--------------------------< G E T _ S E T T I N G S _ F R O M _ C I T E _ C L A S S >----------------------
  
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/.
+
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
 
+
rendered style.
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/.
 
 
 
This function uses ustring functions because firstname initials may be any of the unicode Latin characters accepted by is_good_vanc_name ().
 
  
 
]]
 
]]
  
local function reduce_to_initials(first)
+
local function get_settings_from_cite_class (ps, ref, cite_class)
if mw.ustring.match(first, "^%u%u$") then return first end; -- when first contains just two upper-case letters, nothing to do
+
local sep;
local initials = {}
+
if (cite_class == "citation") then -- for citation templates (CS2)
local i = 0; -- counter for number of initials
+
sep, ps, ref = set_cs2_style (ps, ref);
for word in mw.ustring.gmatch(first, "[^%s%.%-]+") do -- names separated by spaces, hyphens, or periods
+
else -- not a citation template so CS1
table.insert(initials, mw.ustring.sub(word,1,1)) -- Vancouver format does not include full stops.
+
sep, ps = set_cs1_style (ps);
i = i + 1; -- bump the counter
 
if 2 <= i then break; end -- only two initials allowed in Vancouver system; if 2, quit
 
 
end
 
end
return table.concat(initials) -- Vancouver format does not include spaces.
+
 
 +
return sep, ps, ref -- return them all
 
end
 
end
  
--[[--------------------------< L I S T  _ P E O P L E >-------------------------------------------------------
 
  
Formats a list of people (e.g. authors / editors)
+
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
 +
 
 +
Establish basic style settings to be used when rendering the citation. Uses |mode= if set and valid or uses
 +
config.CitationClass from the template's #invoke: to establish style.
  
 
]]
 
]]
  
local function list_people(control, people, etal, list_name) -- TODO: why is list_name here?  not used in this function
+
local function set_style (mode, ps, ref, cite_class)
 
local sep;
 
local sep;
local namesep;
+
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
local format = control.format
+
sep, ps, ref = set_cs2_style (ps, ref);
local maximum = control.maximum
+
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
local lastauthoramp = control.lastauthoramp;
+
sep, ps = set_cs1_style (ps);
local text = {}
+
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
if 'vanc' == format then -- Vancouver-like author/editor name styling?
+
end
sep = ','; -- name-list separator between authors is a comma
+
if 'none' == ps:lower() then -- if assigned value is 'none' then
namesep = ' '; -- last/first separator is a space
+
ps = ''; -- set to empty string
else
 
sep = ';' -- name-list separator between authors is a semicolon
 
namesep = ', ' -- last/first separator is <comma><space>
 
 
end
 
end
 
 
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
+
return sep, ps, ref
if is_set (maximum) and maximum < 1 then return "", 0; end -- returned 0 is for EditorCount; not used for authors
+
end
+
 
for i,person in ipairs(people) do
+
 
if is_set(person.last) then
+
--[=[-------------------------< I S _ P D F >------------------------------------------------------------------
local mask = person.mask
+
 
local one
+
Determines if a url has the file extension that is one of the pdf file extensions used by [[MediaWiki:Common.css]] when
local sep_one = sep;
+
applying the pdf icon to external links.
if is_set (maximum) and i > maximum then
+
 
etal = true;
+
returns true if file extension is one of the recognized extensions, else false
break;
+
 
elseif (mask ~= nil) then
+
]=]
local n = tonumber(mask)
+
 
if (n ~= nil) then
+
local function is_pdf (url)
one = string.rep("&mdash;",n)
+
return url:match ('%.pdf$') or url:match ('%.PDF$') or url:match ('%.pdf[%?#]') or url:match ('%.PDF[%?#]');
else
+
end
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
 
table.insert( text, one )
 
table.insert( text, sep_one )
 
end
 
end
 
  
local count = #text / 2; -- (number of names + number of separators) divided by 2
+
--[[--------------------------< S T Y L E _ F O R M A T >------------------------------------------------------
if count > 0 then  
+
 
if count > 1 and is_set(lastauthoramp) and not etal then
+
Applies css style to |format=, |chapter-format=, etc.  Also emits an error message if the format parameter does
text[#text-2] = " & "; -- replace last separator with ampersand text
+
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)
 +
if is_set (format) then
 +
format = wrap_style ('format', format); -- add leading space, parentheses, resize
 +
if not is_set (url) then
 +
format = format .. set_error( 'format_missing_url', {fmt_param, url_param} ); -- add an error message
 
end
 
end
text[#text] = nil; -- erase the last separator
+
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;
local result = table.concat(text) -- construct list
 
if etal and is_set (result) then -- etal may be set by |display-authors=etal but we might not have a last-first list
 
result = result .. sep .. ' ' .. cfg.messages['et al']; -- we've go a last-first list and etal so add et al.
 
end
 
 
return result, count
 
 
end
 
end
  
--[[--------------------------< A N C H O R _ I D >------------------------------------------------------------
 
  
Generates a CITEREF anchor ID if we have at least one name or a date. Otherwise returns an empty string.
+
--[[--------------------------< G E T _ D I S P L A Y _ A U T H O R S _ E D I T O R S >------------------------
 +
 
 +
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.
  
namelist is one of the contributor-, author-, or editor-name lists chosen in that order. year is Year or anchor_year.
+
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
local function anchor_id (namelist, year)
+
number of authors in the list and set the 'etal' flag true. This will cause the list_people() to display all of
local names={}; -- a table for the one to four names and year
+
the names in the name list followed by 'et al.'
for i,v in ipairs (namelist) do -- loop through the list and take up to the first four last names
 
names[i] = v.last
 
if i == 4 then break end -- if four then done
 
end
 
table.insert (names, year); -- add the year at the 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
 
else
 
return ''; -- return an empty string; no reason to include CITEREF id in this citation
 
end
 
end
 
  
 +
In all other cases, returns nil and the previous state of the 'etal' flag.
  
--[[--------------------------< N A M E _ H A S _ E T A L >----------------------------------------------------
+
inputs:
 
+
max: A['DisplayAuthors'] or A['DisplayEditors']; a number or some flavor of etal
Evaluates the content of author and editor name parameters for variations on the theme of et al.  If found,
+
count: #a or #e
the et al. is removed, a flag is set to true and the function returns the modified name and the flag.
+
list_name: 'authors' or 'editors'
 
+
etal: author_etal or editor_etal
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_display_authors_editors (max, count, list_name, etal)
 
+
if is_set (max) then
if is_set (name) then -- name can be nil in which case just return
+
if 'etal' == max:lower():gsub("[ '%.]", '') then -- the :gsub() portion makes 'etal' from a variety of 'et al.' spellings and stylings
local etal_pattern = "[;,]? *[\"']*%f[%a][Ee][Tt] *[Aa][Ll][%.\"']*$" -- variations on the 'et al' theme
+
max = count + 1; -- number of authors + 1 so display all author name plus et al.
local others_pattern = "[;,]? *%f[%a]and [Oo]thers"; -- and alternate to et al.
+
etal = true; -- overrides value set by extract_names()
+
elseif max:match ('^%d+$') then -- if is a string of numbers
if name:match (etal_pattern) then -- variants on et al.
+
max = tonumber (max); -- make it a number
name = name:gsub (etal_pattern, ''); -- if found, remove
+
if max >= count then -- if |display-xxxxors= value greater than or equal to number of authors/editors
etal = true; -- set flag (may have been set previously here or by |display-authors=etal)
+
add_maint_cat ('disp_auth_ed', cfg.special_case_translation [list_name]);
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
 +
else -- not a valid keyword or number
 +
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
 +
max = nil; -- unset; as if |display-xxxxors= had not been set
 
end
 
end
 
end
 
end
return name, etal; --
+
 +
return max, etal;
 
end
 
end
  
--[[--------------------------< E X T R A C T _ N A M E S >----------------------------------------------------
 
Gets name list from the input arguments
 
  
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
+
--[[--------------------------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >------------------------------
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=
+
Adds page to Category:CS1 maint: extra text if |page= or |pages= has what appears to be some form of p. or pp.
are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=.
+
abbreviation in the first characters of the parameter content.
  
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
+
check Page and Pages for extraneous p, p., pp, and pp. at start of parameter value:
that will cause list_people() to add the static 'et al.' text from Module:Citation/CS1/Configuration.  This keeps 'et al.' out of the
+
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
template's metadata.  When this occurs, the page is added to a maintenance category.
+
bad pattern: '^[Pp][Pp]' matches matches when |page(s)= begins pp or pP or Pp or PP
  
 
]]
 
]]
  
local function extract_names(args, list_name)
+
local function extra_text_in_page_check (page)
local names = {}; -- table of names
+
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 last; -- individual name components
+
local bad_pattern = '^[Pp]?[Pp]%.?[ %d]';
local first;
+
 
local link;
+
if not page:match (good_pattern) and (page:match (bad_pattern) or  page:match ('^[Pp]ages?')) then
local mask;
+
add_maint_cat ('extra_text');
local i = 1; -- loop counter/indexer
+
end
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)
+
 
local etal=false; -- return value set to true when we find some form of et al. in an author parameter
+
 
 +
--[=[-------------------------< G E T _ V _ N A M E _ T A B L E >----------------------------------------------
 +
 
 +
split apart a |vauthors= or |veditors= parameter.  This function allows for corporate names, wrapped in doubled
 +
parentheses to also have commas; in the old version of the code, the doubled parnetheses were included in the
 +
rendered citation and in the metadata. Individual author names may be wikilinked
  
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
+
|vauthors=Jones AB, [[E. B. White|White EB]], ((Black, Brown, and Co.))
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.
+
]=]
first, etal = name_has_etal (first, etal, false); -- find and remove variations on et al.
 
  
if first and not last then -- if there is a firstn without a matching lastn
+
local function get_v_name_table (vparam, output_table, output_link_table)
table.insert( z.message_tail, { set_error( 'first_missing_last', {err_msg_list_name, i}, true ) } ); -- add this error message
+
local name_table = mw.text.split(vparam, "%s*,%s*"); -- names are separated by commas
elseif not first and not last then -- if both firstn and lastn aren't found, are we done?
+
local wl_type, label, link; -- wl_type not used here; just a place holder
count = count + 1; -- number of times we haven't found last and first
+
if 2 <= count then -- two missing names and we give up
+
local i = 1;
break; -- normal exit or there is a two-name hole in the list; can't tell which
+
 +
while name_table[i] do
 +
if name_table[i]:match ('^%(%(.*[^%)][^%)]$') then -- first segment of corporate with one or more commas; this segment has the opening doubled parens
 +
local name = name_table[i];
 +
i=i+1; -- bump indexer to next segment
 +
while name_table[i] do
 +
name = name .. ', ' .. name_table[i]; -- concatenate with previous segments
 +
if name_table[i]:match ('^.*%)%)$') then -- if this table member has the closing doubled parens
 +
break; -- and done reassembling so
 +
end
 +
i=i+1; -- bump indexer
 
end
 
end
else -- we have last with or without a first
+
table.insert (output_table, name); -- and add corporate name to the output table
if is_set (link) and false == link_param_ok (link) then -- do this test here in case link is missing last
+
table.insert (output_link_table, ''); -- no wikilink
table.insert( z.message_tail, { set_error( 'bad_paramlink', list_name:match ("(%w+)List"):lower() .. '-link' .. i )}); -- url or wikilink in author link;
+
else
end
+
wl_type, label, link = is_wikilink (name_table[i]); -- wl_type is: 0, no wl (text in label variable); 1, [[D]]; 2, [[L|D]]
names[n] = {last = last, first = first, link = link, mask = mask, corporate=false}; -- add this name to our names list (corporate for |vauthors= only)
+
table.insert (output_table, label); -- add this name
n = n + 1; -- point to next location in the names table
+
if 1 == wl_type then
if 1 == count then -- if the previous name was missing
+
table.insert (output_link_table, label); -- simple wikilink [[D]]
table.insert( z.message_tail, { set_error( 'missing_name', {err_msg_list_name, i-1}, true ) } ); -- add this error message
+
else
 +
table.insert (output_link_table, link); -- no wikilink or [[L|D]]; add this link if there is one, else empty string
 
end
 
end
count = 0; -- reset the counter, we're looking for two consecutive missing names
 
 
end
 
end
i = i + 1; -- point to next args location
+
i = i+1;
end
+
end
+
return output_table;
return names, etal; -- all done, return our list of names
 
 
end
 
end
  
--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------
 
  
Populates ID table from arguments using configuration settings. Loops through cfg.id_handlers and searches args for
+
--[[--------------------------< P A R S E _ V A U T H O R S _ V E D I T O R S >--------------------------------
any of the parameters listed in each cfg.id_handlers['...'].parameters.  If found, adds the parameter and value to
 
the identifier list.  Emits redundant error message is more than one alias exists in args
 
  
]]
+
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.
  
local function extract_ids( args )
+
Author / editor names in |vauthors= or |veditors= must be in Vancouver system style. Corporate or institutional names
local id_list = {}; -- list of identifiers found in args
+
may sometimes be required and because such names will often fail the is_good_vanc_name() and other format compliance
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
+
tests, are wrapped in doubled paranethese ((corporate name)) to suppress the format tests.
v = select_one( args, v.parameters, 'redundant_parameters' ); -- v.parameters is a table of aliases for k; here we pick one from args if present
 
if is_set(v) then id_list[k] = v; end -- if found in args, add identifier to our list
 
end
 
return id_list;
 
end
 
  
--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------
+
Supports generational suffixes Jr, 2nd, 3rd, 4th–6th.
  
Takes a table of IDs created by extract_ids() and turns it into a table of formatted ID outputs.
+
This function sets the vancouver error when a reqired comma is missing and when there is a space between an author's initials.
 
 
inputs:
 
id_list – table of identifiers built by extract_ids()
 
options – table of various template parameter values used to modify some manually handled identifiers
 
  
 
]]
 
]]
  
local function build_id_list( id_list, options )
+
local function parse_vauthors_veditors (args, vparam, list_name)
local new_list, handler = {};
+
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;
  
function fallback(k) return { __index = function(t,i) return cfg.id_handlers[k][i] end } end;
+
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 k, v in pairs( id_list ) do -- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
+
 
-- fallback to read-only cfg
+
for i, v_name in ipairs(v_name_table) do
handler = setmetatable( { ['id'] = v }, fallback(k) );
+
if v_name:match ('^%(%(.+%)%)$') then -- corporate authors are wrapped in doubled parentheses to supress vanc formatting and error detection
+
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
if handler.mode == 'external' then
+
last = v_name:match ('^%(%((.+)%)%)$') -- remove doubled parntheses
table.insert( new_list, {handler.label, external_link_id( handler ) } );
+
corporate = true; -- flag used in list_people()
elseif handler.mode == 'internal' then
+
elseif string.find(v_name, "%s") then
table.insert( new_list, {handler.label, internal_link_id( handler ) } );
+
if v_name:find('[;%.]') then -- look for commonly occurring punctuation characters;
elseif handler.mode ~= 'manual' then
+
add_vanc_error ('punctuation');
error( cfg.messages['unknown_ID_mode'] );
+
end
elseif k == 'DOI' then
+
local lastfirstTable = {}
table.insert( new_list, {handler.label, doi( v, options.DoiBroken ) } );
+
lastfirstTable = mw.text.split(v_name, "%s")
elseif k == 'ARXIV' then
+
first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be author intials
table.insert( new_list, {handler.label, arxiv( v, options.Class ) } );  
+
if is_suffix (first) then -- if a valid suffix
elseif k == 'ASIN' then
+
suffix = first -- save it as a suffix and
table.insert( new_list, {handler.label, amazon( v, options.ASINTLD ) } );
+
first = table.remove(lastfirstTable); -- get what should be the initials from the table
elseif k == 'LCCN' then
+
end -- no suffix error message here because letter combination may be result of Romanization; check for digits?
table.insert( new_list, {handler.label, lccn( v ) } );
+
last = table.concat(lastfirstTable, " ") -- returns a string that is the concatenation of all other names that are not initials
elseif k == 'OL' or k == 'OLA' then
+
if mw.ustring.match (last, '%a+%s+%u+%s+%a+') then
table.insert( new_list, {handler.label, openlibrary( v ) } );
+
add_vanc_error ('missing comma'); -- matches last II last; the case when a comma is missing
elseif k == 'PMC' then
+
end
table.insert( new_list, {handler.label, pmc( v, options.Embargo ) } );
+
if mw.ustring.match (v_name, ' %u %u$') then -- this test is in the wrong place TODO: move or replace with a more appropriate test
elseif k == 'PMID' then
+
add_vanc_error ('name'); -- matches a space between two intiials
table.insert( new_list, {handler.label, pmid( v ) } );
+
end
elseif k == 'ISMN' then
+
else
table.insert( new_list, {handler.label, ismn( v ) } );
+
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
elseif k == 'ISSN' then
+
last = v_name; -- last name or single corporate name?  Doesn't support multiword corporate names? do we need this?
table.insert( new_list, {handler.label, issn( v ) } );
+
end
elseif k == 'EISSN' then
+
table.insert( new_list, {handler.label, issn( v, true ) } ); -- true distinguishes eissn from issn
+
if is_set (first) then
elseif k == 'ISBN' then
+
if not mw.ustring.match (first, "^%u?%u$") then -- first shall contain one or two upper-case letters, nothing else
local ISBN = internal_link_id( handler );
+
add_vanc_error ('initials'); -- too many initials; mixed case initials (which may be ok Romanization); hyphenated initials
if not check_isbn( v ) and not is_set(options.IgnoreISBN) then
+
end
ISBN = ISBN .. set_error( 'bad_isbn', {}, false, " ", "" );
+
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
 
end
table.insert( new_list, {handler.label, ISBN } );
 
elseif k == 'USENETID' then
 
table.insert( new_list, {handler.label, message_id( v ) } );
 
 
else
 
else
error( cfg.messages['unknown_manual_ID'] );
+
if not corporate then
 +
is_good_vanc_name (last, '');
 +
end
 
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
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 >----------------------------------------------------
 
  
Cleanup parameter values for the metadata by removing or replacing invisible characters and certain html entities.
+
--[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------
  
2015-12-10: there is a bug in mw.text.unstripNoWiki ().  It replaced math stripmarkers with the appropriate content
+
Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or
when it shouldn't.  See https://phabricator.wikimedia.org/T121085 and Wikipedia_talk:Lua#stripmarkers_and_mw.text.unstripNoWiki.28.29
+
select one of |editors=, |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list.
  
TODO: move the replacement patterns and replacement values into a table in /Configuration similar to the invisible
+
Only one of these appropriate three will be used.  The hierarchy is: |authorn= (and aliases) highest and |authors= lowest and
characters table?
+
similarly, |editorn= (and aliases) highest and |editors= lowest
]]
 
  
local function coins_cleanup (value)
+
When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second
value = mw.text.unstripNoWiki (value); -- replace nowiki stripmarkers with their content
+
test which mimicks the test used in extract_names() when looking for a hole in the author name list.  There may be a better
value = value:gsub ('<span class="nowrap" style="padding%-left:0%.1em;">&#39;s</span>', "'s"); -- replace {{'s}} template with simple apostrophe-s
+
way to do this, I just haven't discovered what that way is.
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
 
  
 +
Emits an error message when more than one xxxxor name source is provided.
  
--[[--------------------------< C O I N S >--------------------------------------------------------------------
+
In this function, vxxxxors = vauthors or veditors; xxxxors = authors or editors as appropriate.
 
 
COinS metadata (see <http://ocoins.info/>) allows automated tools to parse the citation information.
 
  
 
]]
 
]]
  
local function COinS(data, class)
+
local function select_author_editor_source (vxxxxors, xxxxors, args, list_name)
if 'table' ~= type(data) or nil == next(data) then
+
local lastfirst = false;
return '';
+
if select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice incase we have a |first1= without a |last1=; this ...
 +
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 1 ) or -- ... also catches the case where |first= is used with |vauthors=
 +
select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) or
 +
select_one( args, cfg.aliases[list_name .. '-First'], 'none', 2 ) then
 +
lastfirst=true;
 
end
 
end
  
for k, v in pairs (data) do -- spin through all of the metadata parameter values
+
if (is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions
if 'ID_list' ~= k and 'Authors' ~= k then -- except the ID_list and Author tables (author nowiki stripmarker done when Author table processed)
+
(is_set (vxxxxors) and is_set (xxxxors)) or
data[k] = coins_cleanup (v);
+
(true == lastfirst and is_set (xxxxors)) then
end
+
local err_name;
end
+
if 'AuthorList' == list_name then -- figure out which name should be used in error message
 
+
err_name = 'author';
local ctx_ver = "Z39.88-2004";
 
 
-- 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
 
});
 
 
if in_array (class, {'arxiv', 'journal', 'news'}) or (in_array (class, {'conference', 'interview', 'map', 'press release', 'web'}) and is_set(data.Periodical)) or  
 
('citation' == class and is_set(data.Periodical) and not is_set (data.Encyclopedia)) then
 
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:journal"; -- journal metadata identifier
 
if 'arxiv' == class then -- set genre according to the type of citation template we are rendering
 
OCinSoutput["rft.genre"] = "preprint"; -- cite arxiv
 
elseif 'conference' == class then
 
OCinSoutput["rft.genre"] = "conference"; -- cite conference (when Periodical set)
 
elseif 'web' == class then
 
OCinSoutput["rft.genre"] = "unknown"; -- cite web (when Periodical set)
 
 
else
 
else
OCinSoutput["rft.genre"] = "article"; -- journal and other 'periodical' articles
+
err_name = 'editor';
 
end
 
end
OCinSoutput["rft.jtitle"] = data.Periodical; -- journal only
+
table.insert( z.message_tail, { set_error( 'redundant_parameters',
if is_set (data.Map) then
+
{err_name .. '-name-list parameters'}, true ) } ); -- add error message
OCinSoutput["rft.atitle"] = data.Map; -- for a map in a periodical
 
else
 
OCinSoutput["rft.atitle"] = data.Title; -- all other 'periodical' article titles
 
end
 
-- these used onlu for periodicals
 
OCinSoutput["rft.ssn"] = data.Season; -- keywords: winter, spring, summer, fall
 
OCinSoutput["rft.chron"] = data.Chron; -- free-form date components
 
OCinSoutput["rft.volume"] = data.Volume; -- does not apply to books
 
OCinSoutput["rft.issue"] = data.Issue;
 
OCinSoutput["rft.pages"] = data.Pages; -- also used in book metadata
 
 
 
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
 
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
 
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
 
end
  
--[[
+
if true == lastfirst then return 1 end; -- return a number indicating which author name source to use
for k, v in pairs( data.ID_list ) do -- what to do about these? For now assume that they are common to all?
+
if is_set (vxxxxors) then return 2 end;
local id, value = cfg.id_handlers[k].COinS;
+
if is_set (xxxxors) then return 3 end;
if k == 'ISBN' then value = clean_isbn( v ); else value = v; end
+
return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last
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
 
end
  
  
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
+
--[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------
  
Validates language names provided in |language= parameter if not an ISO639-1 codeHandles the special case that is Norwegian where
+
This function is used to validate a parameter's assigned value for those parameters that have only a limited number
ISO639-1 code 'no' is mapped to language name 'Norwegian Bokmål' by Extention:CLDR.
+
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 trueIf the parameter value is one of the list of allowed values returns
 +
true; else, emits an error message and returns false.
  
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
+
local function is_valid_parameter_value (value, name, possible)
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
+
if not is_set (value) then
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
+
return true; -- an empty parameter is ok
returns only the Wikimedia language name.
+
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
  
Adapted from code taken from Module:Check ISO 639-1.
 
  
]]
+
--[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >----------------------------------------
  
local function get_iso639_code (lang)
+
This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space
if 'norwegian' == lang:lower() then -- special case related to Wikimedia remap of code 'no' at Extension:CLDR
+
when the last character is not a sepc character or when the last three characters are not sepc followed by two
return 'Norwegian', 'no'; -- Make sure rendered version is properly capitalized
+
closing square brackets (close of a wikilink).  When either of these is true, the name_list is terminated with a
end
+
single space character.
+
 
local languages = mw.language.fetchLanguageNames(mw.getContentLanguage():getCode(), '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
+
local function terminate_name_list (name_list, sepc)
+
if (string.sub (name_list,-3,-1) == sepc .. '. ') then -- if already properly terminated
for code, name in pairs(languages) do -- scan the list to see if we can find our language
+
return name_list; -- just return the name list
if langlc == mw.ustring.lower(name) then
+
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
if 2 ~= code:len() then -- ISO639-1 codes only
+
return name_list .. " "; -- don't add another
return name; -- so return the name but not the code
+
else
end
+
return name_list .. sepc .. ' '; -- otherwise terninate the name list
return name, code; -- found it, return name to ensure proper capitalization and the ISO639-1 code
 
end
 
 
end
 
end
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 >------------------------------------------
 
  
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.
+
--[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------
  
There is an exception.  There are three ISO639-1 codes for Norewegian language variants.  There are two official variants: Norwegian Bokmål (code 'nb') and
+
returns the concatenation of the formatted volume and issue parameters as a single string; or formatted volume
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
+
or formatted issue, or an empty string if neither are set.
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.
 
 
 
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
 
return 'Norwegian' when editors use |language=no.  We presume that editors don't know about the variants or can't descriminate between them.
 
 
 
See Help talk:Citation Style_1#An ISO 639-1 language name test
 
 
 
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.
 
 
 
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 language_parameter (lang)
 
local code; -- the ISO639-1 two character 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=
 
 
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%-') or 2 == lang:len() then -- ISO639-1 language code are 2 characters (fetchLanguageName also supports 3 character codes)
 
if lang:match ('^zh-') then
 
name = mw.language.fetchLanguageName( lang:lower(), lang:lower() );
 
else
 
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
+
local function format_volume_issue (volume, issue, cite_class, origin, sepc, lower)
code = lang:lower(); -- save it
+
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
 
else
name, code = get_iso639_code (lang); -- attempt to get code from name (assign name here so that we are sure of proper capitalization)
+
return wrap_msg ('issue', {sepc, issue}, lower);
 
end
 
end
+
end
if is_set (code) then
+
 
if 'no' == code then name = '挪威语' end; -- override wikimedia when code is 'no'
+
local vol = '';
if 'zh' ~= code and not code:match ('^zh-') then -- English not the language
+
add_prop_cat ('foreign_lang_source', {name, code})
+
if is_set (volume) then
end
+
if (4 < mw.ustring.len(volume)) then
 +
vol = substitute (cfg.messages['j-vol'], {sepc, volume});
 
else
 
else
add_maint_cat ('unknown_lang'); -- add maint category if not already added
+
vol = substitute (cfg.presentation['vol-bold'], {sepc, hyphen_to_dash(volume)});
 
end
 
end
 
table.insert (language_list, name);
 
name = ''; -- so we can reuse it
 
 
end
 
end
+
if is_set (issue) then
code = #language_list -- reuse code as number of languages in the list
+
return vol .. substitute (cfg.messages['j-issue'], issue);
if 2 >= code then
 
name = table.concat (language_list, '及') -- insert '<space>及<space>' between two language names
 
elseif 2 < code then
 
language_list[code] = '及' .. language_list[code]; -- prepend last name with ' 及<space>'
 
name = table.concat (language_list, ',') -- and concatenate with '<comma><space>' separators
 
 
end
 
end
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
+
return vol;
 
end
 
end
  
--[[--------------------------< S E T _ C S 1 _ S T Y L E >----------------------------------------------------
 
  
Set style settings for CS1 citation templates. Returns separator and postscript settings
+
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
  
]]
+
adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings.
 +
The return order is:
 +
page, pages, sheet, sheets
  
local function set_cs1_style (ps)
+
Singular has priority over plural when both are provided.
if not is_set (ps) then -- unless explicitely set to something
 
ps = '.'; -- terminate the rendered citation with a period
 
end
 
return '.', ps; -- separator is a full stop
 
end
 
 
 
--[[--------------------------< S E T _ C S 2 _ S T Y L E >----------------------------------------------------
 
 
 
Set style settings for CS2 citation templates. Returns separator, postscript, ref settings
 
  
 
]]
 
]]
  
local function set_cs2_style (ps, ref)
+
local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower)
if not is_set (ps) then -- if |postscript= has not been set, set cs2 default
+
if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators
ps = ''; -- make sure it isn't nil
+
if is_set (sheet) then
end
+
if 'journal' == origin then
if not is_set (ref) then -- if |ref= is not set
+
return '', '', wrap_msg ('j-sheet', sheet, lower), '';
ref = "harv"; -- set default |ref=harv
+
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
 +
 
 +
local is_journal = 'journal' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'journal' == origin);
 +
 +
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
 
end
return ',', ps, ref; -- separator is a comma
+
 +
return '', '', '', ''; -- return empty strings
 
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 >----------------------
 
  
When |mode= is not set or when its value is invalid, use config.CitationClass and parameter values to establish
+
--[=[-------------------------< A R C H I V E _ U R L _ C H E C K >--------------------------------------------
rendered style.
 
  
]]
+
Check archive.org urls to make sure they at least look like they are pointing at valid archives and not to the
 +
save snapshot url or to calendar pages.  When the archive url is 'https://web.archive.org/save/' (or http://...)
 +
archive.org saves a snapshot of the target page in the url.  That is something that Wikipedia should not allow
 +
unwitting readers to do.
  
local function get_settings_from_cite_class (ps, ref, cite_class)
+
When the archive.org url does not have a complete timestamp, archive.org chooses a snapshot according to its own
local sep;
+
algorithm or provides a calendar 'search' result.  [[WP:ELNO]] discourages links to search results.
if (cite_class == "citation") then -- for citation templates (CS2)
+
 
sep, ps, ref = set_cs2_style (ps, ref);
+
This function looks at the value assigned to |archive-url= and returns empty strings for |archive-url= and
else -- not a citation template so CS1
+
|archive-date= and an error message when:
sep, ps = set_cs1_style (ps);
+
|archive-url= holds an archive.org save command url
 +
|archive-url= is an archive.org url that does not have a complete timestamp (YYYYMMDDhhmmss 14 digits) in the
 +
correct place
 +
otherwise returns |archive-url= and |archive-date=
 +
 
 +
There are two mostly compatible archive.org urls:
 +
//web.archive.org/<timestamp>... -- the old form
 +
//web.archive.org/web/<timestamp>... -- the new form
 +
 
 +
The old form does not support or map to the new form when it contains a display flag.  There are four identified flags
 +
('id_', 'js_', 'cs_', 'im_') but since archive.org ignores others following the same form (two letters and an underscore)
 +
we don't check for these specific flags but we do check the form.
 +
 
 +
This function supports a preview mode.  When the article is rendered in preview mode, this funct may return a modified
 +
archive url:
 +
for save command errors, return undated wildcard (/*/)
 +
for timestamp errors when the timestamp has a wildcard, return the url unmodified
 +
for timestamp errors when the timestamp does not have a wildcard, return with timestamp limited to six digits plus wildcard (/yyyymm*/)
 +
 
 +
]=]
 +
 
 +
local function archive_url_check (url, date)
 +
local err_msg = ''; -- start with the error message empty
 +
local path, timestamp, flag; -- portions of the archive.or url
 +
 +
if (not url:match('//web%.archive%.org/')) and (not url:match('//liveweb%.archive%.org/')) then -- also deprecated liveweb Wayback machine url
 +
return url, date; -- not an archive.org archive, return ArchiveURL and ArchiveDate
 
end
 
end
  
return sep, ps, ref -- return them all
+
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
 +
elseif is_set(path) and 'web/' ~= path then -- older archive urls do not have the extra 'web/' path element
 +
err_msg = 'path';
 +
elseif is_set (flag) and not is_set (path) then -- flag not allowed with the old form url (without the 'web/' path element)
 +
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
 +
-- if here, something not right so
 +
table.insert( z.message_tail, { set_error( 'archive_url', {err_msg}, true ) } ); -- add error message and
 +
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
  
--[[--------------------------< S E T _ S T Y L E >------------------------------------------------------------
 
  
Establish basic style settings to be used when rendering the citationUses |mode= if set and valid or uses
+
--[[--------------------------< M I S S I N G _ P I P E _ C H E C K >------------------------------------------
config.CitationClass from the template's #invoke: to establish style.
+
 
 +
Look at the contents of a parameter. If the content has a string of characters and digits followed by an equal
 +
sign, compare the alphanumeric string to the list of cs1|2 parametersIf found, then the string is possibly a
 +
parameter that is missing its pipe:
 +
{{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 set_style (mode, ps, ref, cite_class)
+
local function missing_pipe_check (value)
local sep;
+
local capture;
if 'cs2' == mode then -- if this template is to be rendered in CS2 (citation) style
+
value = value:gsub ('%b<>', ''); -- remove xml/html tags because attributes: class=, title=, etc
sep, ps, ref = set_cs2_style (ps, ref);
+
 
elseif 'cs1' == mode then -- if this template is to be rendered in CS1 (cite xxx) style
+
capture = value:match ('%s+(%a[%a%d]+)%s*=') or value:match ('^(%a[%a%d]+)%s*='); -- find and categorize parameters with possible missing pipes
sep, ps = set_cs1_style (ps);
+
if capture and validate (capture) then -- if the capture is a valid parameter name
else -- anything but cs1 or cs2
+
add_maint_cat ('missing_pipe');
sep, ps, ref = get_settings_from_cite_class (ps, ref, cite_class); -- get settings based on the template's CitationClass
+
end
end
+
end
if 'none' == ps:lower() then -- if assigned value is 'none' then
 
ps = ''; -- set to empty string
 
end
 
 
return sep, ps, ref
 
end
 
  
--[=[-------------------------< I S _ P D F >------------------------------------------------------------------
 
  
Determines if a url has the file extension that is one of the pdf file extensions used by [[MediaWiki:Common.css]] when
+
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
applying the pdf icon to external links.
 
  
returns true if file extension is one of the recognized extension, else false
+
This is the main function doing the majority of the citation formatting.
  
]=]
+
]]
  
local function is_pdf (url)
+
local function citation0( config, args)
return url:match ('%.pdf[%?#]?') or url:match ('%.PDF[%?#]?');
+
--[[
end
+
Load Input Parameters
 +
The argument_wrapper facilitates the mapping of multiple aliases to single internal variable.
 +
]]
 +
local A = argument_wrapper( args );
 +
local i
  
--[[--------------------------< S T Y L E _ F O R M A T >------------------------------------------------------
+
-- Pick out the relevant fields from the arguments.  Different citation templates
 +
-- define different field names for the same underlying things.
  
Applies css style to |format=, |chapter-format=, etc.  Also emits an error message if the format parameter does
+
-- set default parameter values defined by |mode= parameter.
not have a matching url parameter. If the format parameter is not set and the url contains a file extension that
+
local Mode = A['Mode'];
is recognized as a pdf document by MediaWiki's commons.css, this code will set the format parameter to (PDF) with
+
if not is_valid_parameter_value (Mode, 'mode', cfg.keywords['mode']) then
the appropriate styling.
+
Mode = '';
 +
end
  
]]
+
local author_etal;
 +
local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors=
 +
local Authors;
 +
local NameListFormat = A['NameListFormat'];
 +
local Collaboration = A['Collaboration'];
  
local function style_format (format, url, fmt_param, url_param)
+
do -- to limit scope of selected
if is_set (format) then
+
local selected = select_author_editor_source (A['Vauthors'], A['Authors'], args, 'AuthorList');
format = wrap_style ('format', format); -- add leading space, parenthases, resize
+
if 1 == selected then
if not is_set (url) then
+
a, author_etal = extract_names (args, 'AuthorList'); -- fetch author list from |authorn= / |lastn= / |firstn=, |author-linkn=, and |author-maskn=
format = format .. set_error( 'format_missing_url', {fmt_param, url_param} ); -- add an error message
+
elseif 2 == selected then
 +
NameListFormat = 'vanc'; -- override whatever |name-list-format= might be
 +
a, author_etal = parse_vauthors_veditors (args, args.vauthors, 'AuthorList'); -- fetch author list from |vauthors=, |author-linkn=, and |author-maskn=
 +
elseif 3 == selected then
 +
Authors = A['Authors']; -- use content of |authors=
 +
if 'authors' == A:ORIGIN('Authors') then -- but add a maint cat if the parameter is |authors=
 +
add_maint_cat ('authors'); -- because use of this parameter is discouraged; what to do about the aliases is a TODO:
 +
end
 +
end
 +
if is_set (Collaboration) then
 +
author_etal = true; -- so that |display-authors=etal not required
 
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
 
  
--[[--------------------------< G E T _ D I S P L A Y _ A U T H O R S _ E D I T O R S >------------------------
+
local Others = A['Others'];
  
Returns a number that may or may not limit the length of the author or editor name lists.
+
local editor_etal;
 +
local e = {}; -- editors list from |editor-lastn= / |editor-firstn= pairs or |veditors=
 +
local Editors;
  
When the value assigned to |display-authors= is a number greater than or equal to zero, return the number and
+
do -- to limit scope of selected
the previous state of the 'etal' flag (false by default but may have been set to true if the name list contains
+
local selected = select_author_editor_source (A['Veditors'], A['Editors'], args, 'EditorList');
some variant of the text 'et al.').
+
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
  
When the value assigned to |display-authors= is the keyword 'etal', return a number that is one greater than the
+
local t = {}; -- translators list from |translator-lastn= / translator-firstn= pairs
number of authors in the list and set the 'etal' flag true.  This will cause the list_people() to display all of
+
local Translators; -- assembled translators name list
the names in the name list followed by 'et al.'
+
t = extract_names (args, 'TranslatorList'); -- fetch translator list from |translatorn= / |translator-lastn=, -firstn=, -linkn=, -maskn=
  
In all other cases, returns nil and the previous state of the 'etal' flag.
+
local interviewers_list = {};
 +
local Interviewers = A['Interviewers']
 +
if is_set (Interviewers) then -- add a maint cat if the |interviewers= is used
 +
add_maint_cat ('interviewers'); -- because use of this parameter is discouraged
 +
else
 +
interviewers_list = extract_names (args, 'InterviewerList'); -- else, process preferred interviewers parameters
 +
end
  
]]
+
local c = {}; -- contributors list from |contributor-lastn= / contributor-firstn= pairs
 
+
local Contributors; -- assembled contributors name list
local function get_display_authors_editors (max, count, list_name, etal)
+
local Contribution = A['Contribution'];
if is_set (max) then
+
if in_array(config.CitationClass, {"book","citation"}) and not is_set(A['Periodical']) then -- |contributor= and |contribution= only supported in book cites
if 'etal' == max:lower():gsub("[ '%.]", '') then -- the :gsub() portion makes 'etal' from a variety of 'et al.' spellings and stylings
+
c = extract_names (args, 'ContributorList'); -- fetch contributor list from |contributorn= / |contributor-lastn=, -firstn=, -linkn=, -maskn=
max = count + 1; -- number of authors + 1 so display all author name plus et al.
+
etal = true; -- overrides value set by extract_names()
+
if 0 < #c then
elseif max:match ('^%d+$') then -- if is a string of numbers
+
if not is_set (Contribution) then -- |contributor= requires |contribution=
max = tonumber (max); -- make it a number
+
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'contribution')}); -- add missing contribution error message
if max >= count and 'authors' == list_name then -- AUTHORS ONLY -- if |display-xxxxors= value greater than or equal to number of authors/editors
+
c = {}; -- blank the contributors' table; it is used as a flag later
add_maint_cat ('disp_auth_ed', list_name);
+
end
 +
if 0 == #a then -- |contributor= requires |author=
 +
table.insert( z.message_tail, { set_error( 'contributor_missing_required_param', 'author')}); -- add missing author error message
 +
c = {}; -- blank the contributors' table; it is used as a flag later
 
end
 
end
else -- not a valid keyword or number
 
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
 
max = nil; -- unset
 
 
end
 
end
elseif 'authors' == list_name then -- AUTHORS ONLY need to clear implicit et al category
+
else -- if not a book cite
max = count + 1; -- number of authors + 1
+
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
 
return max, etal;
 
end
 
  
--[[--------------------------< E X T R A _ T E X T _ I N _ P A G E _ C H E C K >------------------------------
+
if not is_valid_parameter_value (NameListFormat, 'name-list-format', cfg.keywords['name-list-format']) then -- only accepted value for this parameter is 'vanc'
 +
NameListFormat = ''; -- anything else, set to empty string
 +
end
  
Adds page to Category:CS1 maint: extra text if |page= or |pages= has what appears to be some form of p. or pp.
+
local Year = A['Year'];
abbreviation in the first characters of the parameter content.
+
local PublicationDate = A['PublicationDate'];
 
+
local OrigYear = A['OrigYear'];
check Page and Pages for extraneous p, p., pp, and pp. at start of parameter value:
+
local Date = A['Date'];
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
+
local LayDate = A['LayDate'];
bad pattern: '^[Pp][Pp]' matches matches when |page(s)= begins pp or pP or Pp or PP
+
------------------------------------------------- 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'];
 +
local ScriptChapter = A['ScriptChapter'];
 +
local ChapterLink -- = A['ChapterLink']; -- deprecated as a parameter but still used internally by cite episode
 +
local TransChapter = A['TransChapter'];
 +
local TitleType = A['TitleType'];
 +
local Degree = A['Degree'];
 +
local Docket = A['Docket'];
 +
local ArchiveFormat = A['ArchiveFormat'];
  
local function extra_text_in_page_check (page)
+
local ArchiveDate;
-- local good_pattern = '^P[^%.P%l]';
+
local ArchiveURL;
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]';
 
  
if not page:match (good_pattern) and (page:match (bad_pattern) or  page:match ('^[Pp]ages?')) then
+
ArchiveURL, ArchiveDate = archive_url_check (A['ArchiveURL'], A['ArchiveDate'])
add_maint_cat ('extra_text');
+
end
+
local DeadURL = A['DeadURL']
-- if Page:match ('^[Pp]?[Pp]%.?[ %d]') or  Page:match ('^[Pp]ages?[ %d]') or
+
if not is_valid_parameter_value (DeadURL, 'dead-url', cfg.keywords ['deadurl']) then -- set in config.defaults to 'yes'
-- Pages:match ('^[Pp]?[Pp]%.?[ %d]') or  Pages:match ('^[Pp]ages?[ %d]') then
+
DeadURL = ''; -- anything else, set to empty string
-- add_maint_cat ('extra_text');
+
end
-- end
 
end
 
  
 +
local URL = A['URL']
 +
local URLorigin = A:ORIGIN('URL'); -- get name of parameter that holds URL
 +
local ChapterURL = A['ChapterURL'];
 +
local ChapterURLorigin = A:ORIGIN('ChapterURL'); -- get name of parameter that holds ChapterURL
 +
local 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
  
--[[--------------------------< P A R S E _ V A U T H O R S _ V E D I T O R S >--------------------------------
+
local Series = A['Series'];
 
+
This function extracts author / editor names from |vauthors= or |veditors= and finds matching |xxxxor-maskn= and
+
local Volume;
|xxxxor-linkn= in args.  It then returns a table of assembled names just as extract_names() does.
+
local Issue;
 
+
local Page;
Author / editor names in |vauthors= or |veditors= must be in Vancouver system style. Corporate or institutional names
+
local Pages;
may sometimes be required and because such names will often fail the is_good_vanc_name() and other format compliance
+
local At;
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.
+
if in_array (config.CitationClass, cfg.templates_using_volume) then
 
+
Volume = A['Volume'];
]]
+
end
 +
-- conference & map books do not support issue
 +
if in_array (config.CitationClass, cfg.templates_using_issue) and not (in_array (config.CitationClass, {'conference', 'map'}) and not is_set (Periodical))then
 +
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
  
local function parse_vauthors_veditors (args, vparam, list_name)
+
local Edition = A['Edition'];
local names = {}; -- table of names assembled from |vauthors=, |author-maskn=, |author-linkn=
+
local PublicationPlace = A['PublicationPlace']
local v_name_table = {};
+
local Place = A['Place'];
local etal = false; -- return value set to true when we find some form of et al. vauthors parameter
+
local last, first, link, mask;
+
local PublisherName = A['PublisherName'];
local corporate = false;
+
local RegistrationRequired = A['RegistrationRequired'];
 +
if not is_valid_parameter_value (RegistrationRequired, 'registration', cfg.keywords ['yes_true_y']) then
 +
RegistrationRequired=nil;
 +
end
  
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)
+
local SubscriptionRequired = A['SubscriptionRequired'];
if vparam:find ('%[%[') or vparam:find ('%]%]') then -- no wikilinking vauthors names
+
if not is_valid_parameter_value (SubscriptionRequired, 'subscription', cfg.keywords ['yes_true_y']) then
add_vanc_error ();
+
SubscriptionRequired=nil;
end
+
end
v_name_table = mw.text.split(vparam, "%s*,%s*") -- names are separated by commas
 
  
for i, v_name in ipairs(v_name_table) do
+
local UrlAccess = A['UrlAccess'];
if v_name:match ('^%(%(.+%)%)$') then -- corporate authors are wrapped in doubled parenthese to supress vanc formatting and error detection
+
if not is_valid_parameter_value (UrlAccess, 'url-access', cfg.keywords ['url-access']) then
first = ''; -- set to empty string for concatenation and because it may have been set for previous author/editor
+
UrlAccess = nil;
last = v_name:match ('^%(%((.+)%)%)$')
 
corporate = true;
 
elseif string.find(v_name, "%s") then
 
    lastfirstTable = {}
 
    lastfirstTable = mw.text.split(v_name, "%s")
 
    first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be author intials
 
    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+') or mw.ustring.match (v_name, ' %u %u$') then
 
add_vanc_error (); -- matches last II last; the case when a comma is missing or 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
 
end
+
if not is_set(URL) and is_set(UrlAccess) then
if is_set (first) and not mw.ustring.match (first, "^%u?%u$") then -- first shall contain one or two upper-case letters, nothing else
+
UrlAccess = nil;
add_vanc_error ();
+
table.insert( z.message_tail, { set_error( 'param_access_requires_param', {'url'}, true ) } );
 
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
 
return names, etal; -- all done, return our list of names
 
end
 
  
--[[--------------------------< S E L E C T _ A U T H O R _ E D I T O R _ S O U R C E >------------------------
+
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=
 +
end
 +
if is_set (UrlAccess) and is_set (RegistrationRequired) then -- these are not the same but contradictory so if both are set
 +
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=
 +
end
  
Select one of |authors=, |authorn= / |lastn / firstn=, or |vauthors= as the source of the author name list or
+
local ChapterUrlAccess = A['ChapterUrlAccess'];
select one of |editors=, |editorn= / editor-lastn= / |editor-firstn= or |veditors= as the source of the editor name list.
+
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
  
Only one of these appropriate three will be used.  The hierarchy is: |authorn= (and aliases) highest and |authors= lowest and
+
local Via = A['Via'];
similarly, |editorn= (and aliases) highest and |editors= lowest
+
local AccessDate = A['AccessDate'];
 +
local Agency = A['Agency'];
  
When looking for |authorn= / |editorn= parameters, test |xxxxor1= and |xxxxor2= (and all of their aliases); stops after the second
+
local Language = A['Language'];
test which mimicks the test used in extract_names() when looking for a hole in the author name list.  There may be a better
+
local Format = A['Format'];
way to do this, I just haven't discovered what that way is.
+
local ChapterFormat = A['ChapterFormat'];
 +
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
  
Emits an error message when more than one xxxxor name source is provided.
+
local ID_list = extract_ids( args );
 +
local ID_access_levels = extract_id_access_levels( args, ID_list );
  
In this function, vxxxxors = vauthors or veditors; xxxxors = authors or editors as appropriate.
+
local Quote = A['Quote'];
  
]]
+
local LayFormat = A['LayFormat'];
 +
local LayURL = A['LayURL'];
 +
local LaySource = A['LaySource'];
 +
local Transcript = A['Transcript'];
 +
local TranscriptFormat = A['TranscriptFormat'];
 +
local TranscriptURL = A['TranscriptURL']
 +
local TranscriptURLorigin = A:ORIGIN('TranscriptURL'); -- get name of parameter that holds TranscriptURL
  
local function select_author_editor_source (vxxxxors, xxxxors, args, list_name)
+
local LastAuthorAmp = A['LastAuthorAmp'];
local lastfirst = false;
+
if not is_valid_parameter_value (LastAuthorAmp, 'last-author-amp', cfg.keywords ['yes_true_y']) then
if select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 1 ) or -- do this twice incase we have a first 1 without a last1
+
LastAuthorAmp = nil; -- set to empty string
select_one( args, cfg.aliases[list_name .. '-Last'], 'none', 2 ) then
+
end
lastfirst=true;
 
end
 
  
if (is_set (vxxxxors) and true == lastfirst) or -- these are the three error conditions
+
local no_tracking_cats = A['NoTracking'];
(is_set (vxxxxors) and is_set (xxxxors)) or
+
if not is_valid_parameter_value (no_tracking_cats, 'no-tracking', cfg.keywords ['yes_true_y']) then
(true == lastfirst and is_set (xxxxors)) then
+
no_tracking_cats = nil; -- set to empty string
local err_name;
+
end
if 'AuthorList' == list_name then -- figure out which name should be used in error message
 
err_name = 'author';
 
else
 
err_name = 'editor';
 
end
 
table.insert( z.message_tail, { set_error( 'redundant_parameters',
 
{err_name .. '-name-list parameters'}, true ) } ); -- add error message
 
end
 
  
if true == lastfirst then return 1 end; -- return a number indicating which author name source to use
+
--local variables that are not cs1 parameters
if is_set (vxxxxors) then return 2 end;
+
local use_lowercase; -- controls capitalization of certain static text
if is_set (xxxxors) then return 3 end;
+
local this_page = mw.title.getCurrentTitle(); -- also used for COinS and for language
return 1; -- no authors so return 1; this allows missing author name test to run in case there is a first without last
+
local anchor_year; -- used in the CITEREF identifier
end
+
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
  
--[[--------------------------< I S _ V A L I D _ P A R A M E T E R _ V A L U E >------------------------------
+
local sepc; -- separator between citation elements for CS1 a period, for CS2, a comma
 +
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
  
This function is used to validate a parameter's assigned value for those parameters that have only a limited number
+
--check this page to see if it is in one of the namespaces that cs1 is not supposed to add to the error categories
of allowable values (yes, y, true, no, etc). When the parameter value has not been assigned a value (missing or empty
+
if not is_set (no_tracking_cats) then -- ignore if we are already not going to categorize this page
in the source template) the function refurns true.  If the parameter value is one of the list of allowed values returns
+
if in_array (this_page.nsText, cfg.uncategorized_namespaces) then
true; else, emits an error message and returns false.
+
no_tracking_cats = "true"; -- set no_tracking_cats
 +
end
 +
for _,v in ipairs (cfg.uncategorized_subpages) do -- cycle through page name patterns
 +
if this_page.text:match (v) then -- test page name against each pattern
 +
no_tracking_cats = "true"; -- set no_tracking_cats
 +
break; -- bail out if one is found
 +
end
 +
end
 +
end
  
]]
+
-- check for extra |page=, |pages= or |at= parameters. (also sheet and sheets while we're at it)
 +
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
  
local function is_valid_parameter_value (value, name, possible)
+
local NoPP = A['NoPP']
if not is_set (value) then
+
if is_set (NoPP) and is_valid_parameter_value (NoPP, 'nopp', cfg.keywords ['yes_true_y']) then
return true; -- an empty parameter is ok
+
NoPP = true;
elseif in_array(value:lower(), possible) then
 
return true;
 
 
else
 
else
table.insert( z.message_tail, { set_error( 'invalid_param_val', {name, value}, true ) } ); -- not an allowed value so add error message
+
NoPP = nil; -- unset, used as a flag later
return false
 
 
end
 
end
end
 
  
 +
if is_set(Page) then
 +
if is_set(Pages) or is_set(At) then
 +
Pages = ''; -- unset the others
 +
At = '';
 +
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
  
--[[--------------------------< T E R M I N A T E _ N A M E _ L I S T >----------------------------------------
+
-- both |publication-place= and |place= (|location=) allowed if different
 
+
if not is_set(PublicationPlace) and is_set(Place) then
This function terminates a name list (author, contributor, editor) with a separator character (sepc) and a space
+
PublicationPlace = Place; -- promote |place= (|location=) to |publication-place
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)
 
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
 
return name_list .. " "; -- don't add another
 
else
 
return name_list .. sepc .. ' '; -- otherwise terninate the name list
 
 
end
 
end
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 Encyclopedia = A['Encyclopedia'];
  
--[[-------------------------< F O R M A T _ V O L U M E _ I S S U E >----------------------------------------
+
if ( config.CitationClass == "encyclopaedia" ) or ( config.CitationClass == "citation" and is_set (Encyclopedia)) then -- test code for citation
 +
if is_set(Periodical) then -- Periodical is set when |encyclopedia is set
 +
if is_set(Title) or is_set (ScriptTitle) then
 +
if not is_set(Chapter) then
 +
Chapter = Title; -- |encyclopedia and |title are set so map |title to |article and |encyclopedia to |title
 +
ScriptChapter = ScriptTitle;
 +
TransChapter = TransTitle;
 +
ChapterURL = URL;
 +
ChapterUrlAccess = UrlAccess;
  
returns the concatenation of the formatted volume and issue parameters as a single string; or formatted volume
+
if not is_set (ChapterURL) and is_set (TitleLink) then
or formatted issue, or an empty string if neither are set.
+
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
  
]]
+
-- Special case for cite techreport.
+
if (config.CitationClass == "techreport") then -- special case for cite techreport
local function format_volume_issue (volume, issue, cite_class, origin, sepc, lower)
+
if is_set(A['Number']) then -- cite techreport uses 'number', which other citations alias to 'issue'
if not is_set (volume) and not is_set (issue) then
+
if not is_set(ID) then -- can we use ID for the "number"?
return '';
+
ID = A['Number']; -- yes, use it
 +
else -- ID has a value so emit error message
 +
table.insert( z.message_tail, { set_error('redundant_parameters', {wrap_style ('parameter', 'id') .. ' and ' .. wrap_style ('parameter', 'number')}, true )});
 +
end
 +
end
 
end
 
end
+
 
if 'magazine' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'magazine' == origin) then
+
-- special case for cite mailing list
if is_set (volume) and is_set (issue) then
+
if (config.CitationClass == "mailinglist") then
return wrap_msg ('vol-no', {sepc, volume, issue}, lower);
+
Periodical = A ['MailingList'];
elseif is_set (volume) then
+
elseif 'mailinglist' == A:ORIGIN('Periodical') then
return wrap_msg ('vol', {sepc, volume}, lower);
+
Periodical = ''; -- unset because mailing list is only used for cite mailing list
else
 
return wrap_msg ('issue', {sepc, issue}, lower);
 
end
 
 
end
 
end
 
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
 
if is_set (issue) then
 
return vol .. substitute (cfg.messages['j-issue'], issue);
 
end
 
return vol;
 
end
 
  
 
+
-- Account for the oddity that is {{cite conference}}, before generation of COinS data.
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
+
if 'conference' == config.CitationClass then
 
+
if is_set(BookTitle) then
adds static text to one of |page(s)= or |sheet(s)= values and returns it with all of the others set to empty strings.
+
Chapter = Title;
The return order is:
+
-- ChapterLink = TitleLink; -- |chapterlink= is deprecated
page, pages, sheet, sheets
+
ChapterURL = URL;
 
+
ChapterUrlAccess = UrlAccess;
Singular has priority over plural when both are provided.
+
ChapterURLorigin = URLorigin;
 
+
URLorigin = '';
]]
+
ChapterFormat = Format;
 
+
TransChapter = TransTitle;
local function format_pages_sheets (page, pages, sheet, sheets, cite_class, origin, sepc, nopp, lower)
+
Title = BookTitle;
if 'map' == cite_class then -- only cite map supports sheet(s) as in-source locators
+
Format = '';
if is_set (sheet) then
+
-- TitleLink = '';
if 'journal' == origin then
+
TransTitle = '';
return '', '', wrap_msg ('j-sheet', sheet, lower), '';
+
URL = '';
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 'speech' ~= config.CitationClass then
 +
Conference = ''; -- not cite conference or cite speech so make sure this is empty string
 
end
 
end
  
local is_journal = 'journal' == cite_class or (in_array (cite_class, {'citation', 'map'}) and 'journal' == origin);
+
-- cite map oddities
 +
local Cartography = "";
 +
local Scale = "";
 +
local Sheet = A['Sheet'] or '';
 +
local Sheets = A['Sheets'] or '';
 +
if config.CitationClass == "map" then
 +
Chapter = A['Map'];
 +
ChapterURL = A['MapURL'];
 +
ChapterUrlAccess = UrlAccess;
 +
TransChapter = A['TransMap'];
 +
ChapterURLorigin = A:ORIGIN('MapURL');
 +
ChapterFormat = A['MapFormat'];
 +
 +
Cartography = A['Cartography'];
 +
if is_set( Cartography ) then
 +
Cartography = sepc .. " " .. wrap_msg ('cartography', Cartography, use_lowercase);
 +
end
 +
Scale = A['Scale'];
 +
if is_set( Scale ) then
 +
Scale = sepc .. " " .. Scale;
 +
end
 +
end
 +
 
 +
-- Account for the oddities that are {{cite episode}} and {{cite serial}}, before generation of COinS data.
 +
if 'episode' == config.CitationClass or 'serial' == config.CitationClass then
 +
local AirDate = A['AirDate'];
 +
local SeriesLink = A['SeriesLink'];
 +
 
 +
link_title_ok (SeriesLink, A:ORIGIN ('SeriesLink'), Series, 'series'); -- check for wikimarkup in |series-link= or wikimarkup in |series= when |series-link= is set
  
if is_set (page) then
+
local Network = A['Network'];
if is_journal then
+
local Station = A['Station'];
return substitute (cfg.messages['j-page(s)'], page), '', '', '';
+
local s, n = {}, {};
elseif not nopp then
+
-- do common parameters first
return substitute (cfg.messages['p-prefix'], {sepc, page}), '', '', '';
+
if is_set(Network) then table.insert(n, Network); end
else
+
if is_set(Station) then table.insert(n, Station); end
return substitute (cfg.messages['nopp'], {sepc, page}), '', '', '';
+
ID = table.concat(n, sepc .. ' ');
end
+
elseif is_set(pages) then
+
if not is_set (Date) and is_set (AirDate) then -- promote airdate to date
if is_journal then
+
Date = AirDate;
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
end
 
 
return '', '', '', ''; -- return empty strings
 
end
 
  
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
+
if 'episode' == config.CitationClass then -- handle the oddities that are strictly {{cite episode}}
 +
local Season = A['Season'];
 +
local SeriesNumber = A['SeriesNumber'];
  
This is the main function doing the majority of the citation formatting.
+
if is_set (Season) and is_set (SeriesNumber) then -- these are mutually exclusive so if both are set
 
+
table.insert( z.message_tail, { set_error( 'redundant_parameters', {wrap_style ('parameter', 'season') .. ' and ' .. wrap_style ('parameter', 'seriesno')}, true ) } ); -- add error message
]]
+
SeriesNumber = ''; -- unset; prefer |season= over |seriesno=
 
+
end
local function citation0( config, args)
+
-- assemble a table of parts concatenated later into Series
--[[
+
if is_set(Season) then table.insert(s, wrap_msg ('season', Season, use_lowercase)); end
Load Input Parameters
+
if is_set(SeriesNumber) then table.insert(s, wrap_msg ('series', SeriesNumber, use_lowercase)); end
The argument_wrapper facilitates the mapping of multiple aliases to single internal variable.
+
if is_set(Issue) then table.insert(s, wrap_msg ('episode', Issue, use_lowercase)); end
]]
+
Issue = ''; -- unset because this is not a unique parameter
local A = argument_wrapper( args );
+
local i
+
Chapter = Title; -- promote title parameters to chapter
 +
ScriptChapter = ScriptTitle;
 +
ChapterLink = TitleLink; -- alias episodelink
 +
TransChapter = TransTitle;
 +
ChapterURL = URL;
 +
ChapterUrlAccess = UrlAccess;
 +
ChapterURLorigin = A:ORIGIN('URL');
 +
 +
Title = Series; -- promote series to title
 +
TitleLink = SeriesLink;
 +
Series = table.concat(s, sepc .. ' '); -- this is concatenation of season, seriesno, episode number
  
-- Pick out the relevant fields from the arguments.  Different citation templates
+
if is_set (ChapterLink) and not is_set (ChapterURL) then -- link but not URL
-- define different field names for the same underlying things.
+
Chapter = make_wikilink (ChapterLink, Chapter);
local author_etal;
+
elseif is_set (ChapterLink) and is_set (ChapterURL) then -- if both are set, URL links episode;
local a = {}; -- authors list from |lastn= / |firstn= pairs or |vauthors=
+
Series = make_wikilink (ChapterLink, Series);
local Authors;
+
end
local NameListFormat = A['NameListFormat'];
+
URL = ''; -- unset
 +
TransTitle = '';
 +
ScriptTitle = '';
 +
 +
else -- now oddities that are cite serial
 +
Issue = ''; -- unset because this parameter no longer supported by the citation/core version of cite serial
 +
Chapter = A['Episode']; -- TODO: make |episode= available to cite episode someday?
 +
if is_set (Series) and is_set (SeriesLink) then
 +
Series = make_wikilink (SeriesLink, Series);
 +
end
 +
Series = wrap_style ('italic-title', Series); -- series is italicized
 +
end
 +
end
 +
-- end of {{cite episode}} stuff
 +
 
 +
-- Account for the oddities that are {{cite arxiv}}, {{cite biorxiv}}, {{cite citeseerx}}, before generation of COinS data.
 +
do
 +
if in_array (config.CitationClass, {'arxiv', 'biorxiv', 'citeseerx'}) then
 +
if not is_set (ID_list[config.CitationClass:upper()]) then -- |arxiv= or |eprint= required for cite arxiv; |biorxiv= & |citeseerx= required for their templates
 +
table.insert(