|
local zcs1 = {};
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 >--------------------------------------
]]
local dates, year_date_check -- functions in Module:Citation/CS1/Date_validation
local cfgdates, =year_date_check, {};reformat_dates -- table of configuration tables that are definedfunctions in Module:Citation/CS1/ConfigurationDate_validation
local whitelist = {}; -- table of tables listing valid template parameter names; defined in Module:Citation/CS1/Whitelist
local is_set, in_array, substitute, error_comment, set_error, select_one, -- functions in Module:Citation/CS1/Utilities
--[[--------------------------< I S _ S E T >------------------------------------------------------------------
add_maint_cat, wrap_style, safe_for_italics, remove_wiki_link;
local z ={}; -- tables in Module:Citation/CS1/Utilities
Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not an empty string.
This function is global because it is called from both this module and from Date validation
local extract_ids, build_id_list, is_embargoed; -- functions in Module:Citation/CS1/Identifiers
]]
function is_set( var )
local make_coins_title, get_coins_pages, COinS; -- functions in Module:Citation/CS1/COinS
return not (var == nil or var == '');
end
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
--[[--------------------------< F I R S T _ S E T >------------------------------------------------------------
end
i = i + 1; -- point to next
end
end
--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------
Whether needle is in haystack
]]
local function in_array( needle, haystack )
if needle == nil then
return false;
end
for n,v in ipairs( haystack ) do
if v == needle then
return n;
end
end
return false;
end
--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
Populates numbered arguments in a message string using an argument table.
]]
local function substitute( msg, args )
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 error_comment( content, hidden )
return substitute( hidden and cfg.presentation['hidden-error'] or cfg.presentation['visible-error'], content );
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
the responsibility of the calling function.
]]
local function set_error( error_id, arguments, raw, prefix, suffix )
local error_state = cfg.error_conditions[ error_id ];
prefix = prefix or "";
suffix = suffix or "";
if error_state == nil then
error( cfg.messages['undefined_error'] );
elseif is_set( error_state.category ) then
table.insert( z.error_categories, error_state.category );
end
local message = substitute( error_state.message, arguments );
message = message .. " ([[" .. cfg.messages['help page link'] ..
"#" .. error_state.anchor .. "|" ..
cfg.messages['help page label'] .. "]])";
z.error_ids[ error_id ] = true;
if in_array( error_id, { 'bare_url_missing_title', 'trans_missing_title' } )
and z.error_ids['citation_missing_title'] then
return '', false;
end
message = table.concat({ prefix, message, suffix });
if raw == true then
return message, error_state.hidden;
end
return error_comment( message, error_state.hidden );
end
--[[--------------------------< A D D _ M A I N T _ C A T >------------------------------------------------------
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 add_maint_cat (key, arguments)
if not added_maint_cats [key] then
added_maint_cats [key] = true; -- note that we've added this category
table.insert( z.maintenance_cats, substitute (cfg.maint_cats [key], arguments)); -- make name then add to table
end
end
--[=[--------------------------< 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?
Syntax defined here: http://tools.ietf.org/html/rfc1034#section-3.5
BNF defined here: https://tools.ietf.org/html/rfc4234
Single character names are generally reserved; see https://tools.ietf.org/html/draft-ietf-dnsind-iana-dns-01#page-15;
see also [[Single-letter second-level domain]]
list of tlds: https://www.iana.org/domains/root/db
rfc952 (modified by rfc 1123) requires the first and last character of a hostname to be a letter or a digit. Between
the first and last characters the name may use letters, digits, and the hyphen. Single character names are not allowed.
Also allowed are IPv4 addresses. IPv6 not supported
domain is expected to be stripped of any path so that the last character in the last character of the tld. tld
There are three tests: the first is looking for a hostname that is 2 to n letters or digits followed by a dot and a
is two or more alpha characters. Any preceding '//' (from splitting a url with a scheme) will be stripped
letter (tld); the second looks for a hostname that is 3 to n characters where the first and last are letters or
here. Perhaps not necessary but retained incase it is necessary for IPv4 dot decimal.
digits and the middle characters are letters, digits, or the hyphen; the whole followed by a dot and a letter or digit.
The third test is for IPv4 dot-decimal address format; tld not allowed.
There are several tests:
the first character of the whole domain name including subdomains must be a letter or a digit
internationalized domain name (ascii characters with .xn-- ASCII Compatible Encoding (ACE) prefix xn-- in the tld) see https://tools.ietf.org/html/rfc3490
single-letter/digit second-level domains in the .org 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
]=]
local function is_domain_name (domain)
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][%a%d]+%.%a') then -- twofirst character hostnamemust be letter andor tlddigit
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]%.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$') then -- assigned one character .net hostname (q.net registered but not active 2015-12-10)
return true;
elseif domain:match ('%f[%a%d][%a%d]%.%a%a$') then -- one character hostname and cctld (2 chars)
return true;
elseif domain:match ('^%f[%a%d][%a%d%-]+[%a%d]%.[%a%d]a+$') then -- three or moretwo character hostname.hostname orand hostname.tld
return true;
elseif domain:match ('^%d%d?%d?%.%d%d?%d?%.%d%d?%d?%.%d%d?%d?') then -- IPv4 address
--[[--------------------------< S P L I T _ U R L >------------------------------------------------------------
Split a url into a scheme, andauthority domain pairindicator, and return both partsdomain. If protocol relative url, return nil for scheme
and domain else return nil for both scheme and domain.
First remove Fully Qualified Domain Name terminator (a dot following tld) (if any) and any path(/), query(?) or fragment(#).
If protocol relative url, return nil scheme and domain else return nil for both scheme and domain.
When not protocol relative, get scheme, authority indicator, and domain. If there is an authority indicator (one
or more '/' characters immediately following the scheme's colon), make sure that there are only 2.
Strip off any port and path;
]]
local function split_url (url_str)
local scheme, authority, domain;
url_str = url_str:gsub ('([%a%d])%.?[/%?#].*$', '%1'); -- strip FQDN terminator and path(/), query(?), fragment (#) (the capture prevents false replacement of '//')
if url_str:match ('%S-:%S+') then -- if there is what appears to be a scheme domain pair
scheme, domain = url_str:match ('(%S-:)(%S+)'); -- extract the scheme and domain portions
elseifif 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
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.
local function link_param_ok (value)
local scheme, domain;
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
end
--[[--------------------------< L I N K _ T I T L E _ O K >---------------------------------------------------
Use link_param_ok() to validate |<param>-link= value and its matching |<title>= value.
|<title>= may be wikilinked but not when |<param>-link= has a value. This function emits an error message when
that condition exists
]]
local function link_title_ok (link, lorig, title, torig)
local orig;
if is_set (link) then -- don't bother if <param>-link doesn't have a value
if not link_param_ok (link) then -- check |<param>-link= markup
orig = lorig; -- identify the failing link parameter
elseif title:find ('%[%[') then -- check |title= for wikilink markup
orig = torig; -- identify the failing |title= parameter
end
end
if is_set (orig) then
table.insert( z.message_tail, { set_error( 'bad_paramlink', orig)}); -- url or wikilink in |title= with |title-link=;
end
end
First we test for space characters. If any are found, return false. Then split the url into scheme and domain
portions, or for protocol relative (//example.com) urls, just the domain. Use is_schemeis_url() andto is_domain()validate tothe two
validate the two portions of the url. If both are valid, or for protocol relative if domain is valid, return true, else false.
Because it is different from a standard url, and because this module used external_link() to make external links
that work for standard and news: links, we validate newsgroup names here. The specification for a newsgroup name
is at https://tools.ietf.org/html/rfc5536#section-3.1.4
]]
local scheme, domain;
scheme, domain = split_url (url_str); -- get scheme or nil and domain or nil from url;
if 'news:' == scheme then -- special case for newsgroups
return domain:match('^[%a%d%+%-_]+%.[%a%d%+%-_%.]*[%a%d%+%-_]$');
end
return is_url (scheme, domain); -- return true if value appears to be a valid url
end
Return true if a parameter value has a string that begins and ends with square brackets [ and ] and the first
non-space characters following the opening bracket obeyappear the rulesto ofbe a uri scheme (see check_url())url. The test will also find external wikilinks
external wikilinks that use protocol relative urls. Also finds bare urls.
The frontier pattern prevents a match on interwiki links which are similar to scheme:path urls. The tests that
find bracketed urls are required because the parameters that call this test (currently |title=, |chapter=, and|work=,
and |workpublisher=) may have wikilinks and there are articles or redirects like '//Hus' so, while uncommon, |title=[[//Hus]] is
is possible as might be [[en://Hus]].
]=]
local scheme, domain;
if value:match ('%f[%[]%[%a%S*:%S+.*%]') then -- if ext wikilink with scheme and domain: [xxxx://yyyyy.zzz]
scheme, domain = split_url (value:match ('%f[%[]%[(%a%S*:)(%S+).*)%]'));
elseif value:match ('%f[%[]%[//%S*%+.%S*%]') then -- if protocol relative ext wikilink: [//yyyyy.zzz]
scheme, domain = split_url (value:match ('%f[%[]%[(//(%S*%+).%S*)%]'));
elseif value:match ('%a%S*:%S+') then -- if bare url with scheme; may have leading or trailing plain text
scheme, domain = split_url (value:match ('(%a%S*:%S+)'));
elseif value:match ('//%S+') then -- if protocol relative bare url: //yyyyy.zzz; may have leading or trailing plain text
scheme, domain = split_url (value:match ('(//%S+)')); -- what is left should be the domain
else
scheme,return domain = split_url (value)false; -- getdidn't schemefind oranything nilthat andis domainobviously or nil froma url;
end
end
end
if is_set (error_message) then -- done looping, if there is an error message, display it
table.insert( z.message_tail, { set_error( 'param_has_ext_link', {error_message}, true ) } );
end
end
--[[--------------------------< S A F E _ F O R _ I T A L I C S >----------------------------------------------
Protects a string that will be wrapped in wiki italic markup '' ... ''
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 )
if not is_set(str) then
return str;
else
if str:sub(1,1) == "'" then str = "<span />" .. str; end
if str:sub(-1,-1) == "'" then str = str .. "<span />"; end
-- Remove newlines as they break italics.
return str:gsub( '\n', ' ' );
end
end
--[[--------------------------< S A F E _ F O R _ U R L >------------------------------------------------------
[']'] = ']',
['\n'] = ' ' } );
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
local function external_link( URL, label, source )
local error_str = "";
local domain;
local path;
if not is_set( label ) then
label = URL;
error_str = set_error( 'bad_url', {wrap_style ('parameter', source)}, false, " " ) .. error_str;
end
domain, path = URL:match ('^([/%.%-%+:%a%d]+)([/%?#].*)$'); -- split the url into scheme plus domain and path
if 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
return table.concat({ "[", URL, " ", safe_for_url( label ), "]", error_str });
end
--[[--------------------------< E X T E R N A L _ L I N K _ I D >----------------------------------------------
Formats a wiki style external link
]]
local function external_link_id(options)
local url_string = options.id;
if options.encode == true or options.encode == nil then
url_string = mw.uri.encode( url_string );
end
return mw.ustring.format( '[[%s|%s]]%s[%s%s%s %s]',
options.link, options.label, options.separator or " ",
options.prefix, url_string, options.suffix or "",
mw.text.nowiki(options.id)
);
end
--[[--------------------------< D E P R E C A T E D _ P A R A M E T E R >--------------------------------------
script_value = script_value:gsub ('^%l%l%s*:%s*', ''); -- strip prefix from script
-- is prefix one of these language codes?
if in_array (lang, {'am', 'ar', 'be', 'bg', 'bs', 'dv', 'el', 'fa', 'he', 'hy', 'ja', 'ka', 'ko', 'ku', 'mk', 'ml', 'ps', 'ru', 'sd', 'sr', 'th', 'uk', 'ug', 'yi', 'zh'}) then
add_prop_cat ('script_with_name', {name, lang})
else
return substitute( cfg.messages[key], str );
end
end
--[[-------------------------< I S _ A L I A S _ U S E D >-----------------------------------------------------
This function is used by select_one() to determine if one of a list of alias parameters is in the argument list
provided by the template.
Input:
args – pointer to the arguments table from calling template
alias – one of the list of possible aliases in the aliases lists from Module:Citation/CS1/Configuration
index – for enumerated parameters, identifies which one
enumerated – true/false flag used choose how enumerated aliases are examined
value – value associated with an alias that has previously been selected; nil if not yet selected
selected – the alias that has previously been selected; nil if not yet selected
error_list – list of aliases that are duplicates of the alias already selected
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)
if enumerated then -- is this a test for an enumerated parameters?
alias = alias:gsub ('#', index); -- replace '#' with the value in index
else
alias = alias:gsub ('#', ''); -- remove '#' if it exists
end
if is_set(args[alias]) then -- alias is in the template's argument list
if value ~= nil and selected ~= alias then -- if we have already selected one of the aliases
local skip;
for _, v in ipairs(error_list) do -- spin through the error list to see if we've added this alias
if v == alias then
skip = true;
break; -- has been added so stop looking
end
end
if not skip then -- has not been added so
table.insert( error_list, alias ); -- add error alias to the error list
end
else
value = args[alias]; -- not yet selected an alias, so select this one
selected = alias;
end
end
return value, selected; -- return newly selected alias, or previously selected alias
end
--[[--------------------------< S E L E C T _ O N E >----------------------------------------------------------
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.
Generates an error if more than one match is present.
]]
local function select_one( args, aliases_list, error_condition, index )
local value = nil; -- the value assigned to the selected parameter
local selected = ''; -- the name of the parameter we have chosen
local error_list = {};
if index ~= nil then index = tostring(index); end
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 error_str = "";
for _, k in ipairs( error_list ) do
if error_str ~= "" then error_str = error_str .. cfg.messages['parameter-separator'] end
error_str = error_str .. wrap_style ('parameter', k);
end
if #error_list > 1 then
error_str = error_str .. cfg.messages['parameter-final-separator'];
else
error_str = error_str .. cfg.messages['parameter-pair-separator'];
end
error_str = error_str .. wrap_style ('parameter', selected);
table.insert( z.message_tail, { set_error( error_condition, {error_str}, true ) } );
end
return value, selected;
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
first match.
This function will detect the visible replacement character when it is part of the wikisource.
Detects but ignores nowiki and math stripmarkers. Also detects other named stripmarkers (gallery, math, pre, ref)
and identifies them with a slightly different error message. See also coins_cleanup().
Detects but ignores the character pattern that results from the transclusion of {{'}} templates.
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.
]]
local function has_invisible_chars (param, v)
local position = ''; -- position of invisible char or starting position of stripmarker
local dummy; -- end of matching string; not used but required to hold end position when a capture is returned
local capture; -- used by stripmarker detection to hold name of the stripmarker
local i=1;
local stripmarker, apostrophe;
capture = string.match (v, '[%w%p ]*'); -- Test for values that are simple ASCII text and bypass other tests if true
if capture == v then -- if same there are no unicode characters
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 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
i=i+1; -- bump our index
end
end
--[[--------------------------< A R G U M E N T _ W R A P P E R >----------------------------------------------
Argument wrapper. This function provides support for argument mapping defined in the configuration file so that
multiple names can be transparently aliased to single internal variable.
--[[
Argument wrapper. This function provides support for argument
mapping defined in the configuration file so that multiple names
can be transparently aliased to single internal variable.
]]
end
--[[--------------------------< V A L I D A T E >--------------------------------------------------------------
--[[
Looks for a parameter's name in the whitelist.
false - deprecated, supported parameters
nil - unsupported parameters
]]
return false; -- Not supported because not found or name is set to nil
end
-- Formats a wiki style internal link
local function internal_link_id(options)
return mw.ustring.format( '[[%s|%s]]%s[[%s%s%s|%s]]',
options.link, options.label, options.separator or " ",
options.prefix, options.id, options.suffix or "",
mw.text.nowiki(options.id)
);
end
return date;
end
--[[--------------------------< IS _ V A L I D _ I S X N >-----------------------------------------------------
ISBN-10 and ISSN validator code calculates checksum across all isbn/issn digits including the check digit. ISBN-13 is checked in check_isbn().
If the number is valid the result will be 0. Before calling this function, issbn/issn must be checked for length and stripped of dashes,
spaces and other non-isxn characters.
]]
local function is_valid_isxn (isxn_str, len)
local temp = 0;
isxn_str = { isxn_str:byte(1, len) }; -- make a table of byte values '0' → 0x30 .. '9' → 0x39, 'X' → 0x58
len = len+1; -- adjust to be a loop counter
for i, v in ipairs( isxn_str ) do -- loop through all of the bytes and calculate the checksum
if v == string.byte( "X" ) then -- if checkdigit is X (compares the byte value of 'X' which is 0x58)
temp = temp + 10*( len - i ); -- it represents 10 decimal
else
temp = temp + tonumber( string.char(v) )*(len-i);
end
end
return temp % 11 == 0; -- returns true if calculation result is zero
end
--[[--------------------------< IS _ V A L I D _ I S X N _ 1 3 >----------------------------------------------
ISBN-13 and ISMN validator code calculates checksum across all 13 isbn/ismn digits including the check digit.
If the number is valid, the result will be 0. Before calling this function, isbn-13/ismn must be checked for length
and stripped of dashes, spaces and other non-isxn-13 characters.
]]
local function is_valid_isxn_13 (isxn_str)
local temp=0;
isxn_str = { isxn_str:byte(1, 13) }; -- make a table of byte values '0' → 0x30 .. '9' → 0x39
for i, v in ipairs( isxn_str ) do
temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) ); -- multiply odd index digits by 1, even index digits by 3 and sum; includes check digit
end
return temp % 10 == 0; -- sum modulo 10 is zero when isbn-13/ismn is correct
end
--[[--------------------------< C H E C K _ I S B N >------------------------------------------------------------
Determines whether an ISBN string is valid
]]
local function check_isbn( isbn_str )
if nil ~= isbn_str:match("[^%s-0-9X]") then return false; end -- fail if isbn_str contains anything but digits, hyphens, or the uppercase X
isbn_str = isbn_str:gsub( "-", "" ):gsub( " ", "" ); -- remove hyphens and spaces
local len = isbn_str:len();
if len ~= 10 and len ~= 13 then
return false;
end
if len == 10 then
if isbn_str:match( "^%d*X?$" ) == nil then return false; end
return is_valid_isxn(isbn_str, 10);
else
local temp = 0;
if isbn_str:match( "^97[89]%d*$" ) == nil then return false; end -- isbn13 begins with 978 or 979; ismn begins with 979
return is_valid_isxn_13 (isbn_str);
end
end
--[[--------------------------< 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
same check digit calculations. See http://www.ismn-international.org/download/Web_ISMN_Users_Manual_2008-6.pdf
section 2, pages 9–12.
]]
local function ismn (id)
local handler = cfg.id_handlers['ISMN'];
local text;
local valid_ismn = true;
id=id:gsub( "[%s-–]", "" ); -- strip spaces, hyphens, and endashes from the ismn
if 13 ~= id:len() or id:match( "^9790%d*$" ) == nil then -- ismn must be 13 digits and begin 9790
valid_ismn = false;
else
valid_ismn=is_valid_isxn_13 (id); -- validate ismn
end
-- text = internal_link_id({link = handler.link, label = handler.label, -- use this (or external version) when there is some place to link to
-- prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
text="[[" .. handler.link .. "|" .. handler.label .. "]]" .. handler.separator .. id; -- because no place to link to yet
if false == valid_ismn then
text = text .. ' ' .. set_error( 'bad_ismn' ) -- add an error message if the issn is invalid
end
return text;
end
--[[--------------------------< I S S N >----------------------------------------------------------------------
Validate and format an issn. This code fixes the case where an editor has included an ISSN in the citation but has separated the two groups of four
digits with a space. When that condition occurred, the resulting link looked like this:
|issn=0819 4327 gives: [http://www.worldcat.org/issn/0819 4327 0819 4327] -- can't have spaces in an external link
This code now prevents that by inserting a hyphen at the issn midpoint. It also validates the issn for length and makes sure that the checkdigit agrees
with the calculated value. Incorrect length (8 digits), characters other than 0-9 and X, or checkdigit / calculated value mismatch will all cause a check issn
error message. The issn is always displayed with a hyphen, even if the issn was given as a single group of 8 digits.
]]
local function issn(id)
local issn_copy = id; -- save a copy of unadulterated issn; use this version for display if issn does not validate
local handler = cfg.id_handlers['ISSN'];
local text;
local valid_issn = true;
id=id:gsub( "[%s-–]", "" ); -- strip spaces, hyphens, and endashes from the issn
if 8 ~= id:len() or nil == id:match( "^%d*X?$" ) then -- validate the issn: 8 digits long, containing only 0-9 or X in the last position
valid_issn=false; -- wrong length or improper character
else
valid_issn=is_valid_isxn(id, 8); -- validate issn
end
if true == valid_issn then
id = string.sub( id, 1, 4 ) .. "-" .. string.sub( id, 5 ); -- if valid, display correctly formatted version
else
id = issn_copy; -- if not valid, use the show the invalid issn with error message
end
text = external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
if false == valid_issn then
text = text .. ' ' .. set_error( 'bad_issn' ) -- add an error message if the issn is invalid
end
return text
end
--[[--------------------------< A M A Z O N >------------------------------------------------------------------
Formats a link to Amazon. Do simple error checking: asin must be mix of 10 numeric or uppercase alpha
characters. If a mix, first character must be uppercase alpha; if all numeric, asins must be 10-digit
isbn. If 10-digit isbn, add a maintenance category so a bot or awb script can replace |asin= with |isbn=.
Error message if not 10 characters, if not isbn10, if mixed and first character is a digit.
]]
local function amazon(id, domain)
local err_cat = ""
if not id:match("^[%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u][%d%u]$") then
err_cat = ' ' .. set_error ('bad_asin'); -- asin is not a mix of 10 uppercase alpha and numeric characters
else
if id:match("^%d%d%d%d%d%d%d%d%d[%dX]$") then -- if 10-digit numeric (or 9 digits with terminal X)
if check_isbn( id ) then -- see if asin value is isbn10
add_maint_cat ('ASIN');
elseif not is_set (err_cat) then
err_cat = ' ' .. set_error ('bad_asin'); -- asin is not isbn10
end
elseif not id:match("^%u[%d%u]+$") then
err_cat = ' ' .. set_error ('bad_asin'); -- asin doesn't begin with uppercase alpha
end
end
if not is_set(domain) then
domain = "com";
elseif in_array (domain, {'jp', 'uk'}) then -- Japan, United Kingdom
domain = "co." .. domain;
elseif in_array (domain, {'au', 'br', 'mx'}) then -- Australia, Brazil, Mexico
domain = "com." .. domain;
end
local handler = cfg.id_handlers['ASIN'];
return external_link_id({link=handler.link,
label=handler.label, prefix=handler.prefix .. domain .. "/dp/",
id=id, encode=handler.encode, separator = handler.separator}) .. err_cat;
end
--[[--------------------------< A R X I V >--------------------------------------------------------------------
See: http://arxiv.org/help/arxiv_identifier
format and error check arXiv identifier. There are three valid forms of the identifier:
the first form, valid only between date codes 9108 and 0703 is:
arXiv:<archive>.<class>/<date code><number><version>
where:
<archive> is a string of alpha characters - may be hyphenated; no other punctuation
<class> is a string of alpha characters - may be hyphenated; no other punctuation
<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
first digit of YY for this form can only 9 and 0
<number> is a three-digit number
<version> is a 1 or more digit number preceded with a lowercase v; no spaces (undocumented)
the second form, valid from April 2007 through December 2014 is:
arXiv:<date code>.<number><version>
where:
<date code> is four digits in the form YYMM where YY is the last two digits of the four-digit year and MM is the month number January = 01
<number> is a four-digit number
<version> is a 1 or more digit number preceded with a lowercase v; no spaces
the third form, valid from January 2015 is:
arXiv:<date code>.<number><version>
where:
<date code> and <version> are as defined for 0704-1412
<number> is a five-digit number
]]
local function arxiv (id, class)
local handler = cfg.id_handlers['ARXIV'];
local year, month, version;
local err_cat = '';
local text;
if id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%d$") or id:match("^%a[%a%.%-]+/[90]%d[01]%d%d%d%dv%d+$") then -- test for the 9108-0703 format w/ & w/o version
year, month = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d[v%d]*$");
year = tonumber(year);
month = tonumber(month);
if ((not (90 < year or 8 > year)) or (1 > month or 12 < month)) or -- if invalid year or invalid month
((91 == year and 7 > month) or (7 == year and 3 < month)) then -- if years ok, are starting and ending months ok?
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
end
elseif id:match("^%d%d[01]%d%.%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%dv%d+$") then -- test for the 0704-1412 w/ & w/o version
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d[v%d]*$");
year = tonumber(year);
month = tonumber(month);
if ((7 > year) or (14 < year) or (1 > month or 12 < month)) or -- is year invalid or is month invalid? (doesn't test for future years)
((7 == year) and (4 > month)) then --or -- when year is 07, is month invalid (before April)?
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
end
elseif id:match("^%d%d[01]%d%.%d%d%d%d%d$") or id:match("^%d%d[01]%d%.%d%d%d%d%dv%d+$") then -- test for the 1501- format w/ & w/o version
year, month = id:match("^(%d%d)([01]%d)%.%d%d%d%d%d[v%d]*$");
year = tonumber(year);
month = tonumber(month);
if ((15 > year) or (1 > month or 12 < month)) then -- is year invalid or is month invalid? (doesn't test for future years)
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- set error message
end
else
err_cat = ' ' .. set_error( 'bad_arxiv' ); -- arXiv id doesn't match any format
end
text = external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
if is_set (class) then
class = ' [[' .. '//arxiv.org/archive/' .. class .. ' ' .. class .. ']]'; -- external link within square brackets, not wikilink
else
class = ''; -- empty string for concatenation
end
return text .. class;
end
--[[
lccn normalization (http://www.loc.gov/marc/lccn-namespace.html#normalization)
1. Remove all blanks.
2. If there is a forward slash (/) in the string, remove it, and remove all characters to the right of the forward slash.
3. If there is a hyphen in the string:
a. Remove it.
b. Inspect the substring following (to the right of) the (removed) hyphen. Then (and assuming that steps 1 and 2 have been carried out):
1. All these characters should be digits, and there should be six or less. (not done in this function)
2. If the length of the substring is less than 6, left-fill the substring with zeroes until the length is six.
Returns a normalized lccn for lccn() to validate. There is no error checking (step 3.b.1) performed in this function.
]]
local function normalize_lccn (lccn)
lccn = lccn:gsub ("%s", ""); -- 1. strip whitespace
if nil ~= string.find (lccn,'/') then
lccn = lccn:match ("(.-)/"); -- 2. remove forward slash and all character to the right of it
end
local prefix
local suffix
prefix, suffix = lccn:match ("(.+)%-(.+)"); -- 3.a remove hyphen by splitting the string into prefix and suffix
if nil ~= suffix then -- if there was a hyphen
suffix=string.rep("0", 6-string.len (suffix)) .. suffix; -- 3.b.2 left fill the suffix with 0s if suffix length less than 6
lccn=prefix..suffix; -- reassemble the lccn
end
return lccn;
end
--[[
Format LCCN link and do simple error checking. LCCN is a character string 8-12 characters long. The length of the LCCN dictates the character type of the first 1-3 characters; the
rightmost eight are always digits. http://info-uri.info/registry/OAIHandler?verb=GetRecord&metadataPrefix=reg&identifier=info:lccn/
length = 8 then all digits
length = 9 then lccn[1] is lower case alpha
length = 10 then lccn[1] and lccn[2] are both lower case alpha or both digits
length = 11 then lccn[1] is lower case alpha, lccn[2] and lccn[3] are both lower case alpha or both digits
length = 12 then lccn[1] and lccn[2] are both lower case alpha
]]
local function lccn(lccn)
local handler = cfg.id_handlers['LCCN'];
local err_cat = ''; -- presume that LCCN is valid
local id = lccn; -- local copy of the lccn
id = normalize_lccn (id); -- get canonical form (no whitespace, hyphens, forward slashes)
local len = id:len(); -- get the length of the lccn
if 8 == len then
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits)
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
end
elseif 9 == len then -- LCCN should be adddddddd
if nil == id:match("%l%d%d%d%d%d%d%d%d") then -- does it match our pattern?
err_cat = ' ' .. set_error( 'bad_lccn' ); -- set an error message
end
elseif 10 == len then -- LCCN should be aadddddddd or dddddddddd
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits) ...
if nil == id:match("^%l%l%d%d%d%d%d%d%d%d") then -- ... see if it matches our pattern
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
end
end
elseif 11 == len then -- LCCN should be aaadddddddd or adddddddddd
if not (id:match("^%l%l%l%d%d%d%d%d%d%d%d") or id:match("^%l%d%d%d%d%d%d%d%d%d%d")) then -- see if it matches one of our patterns
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
end
elseif 12 == len then -- LCCN should be aadddddddddd
if not id:match("^%l%l%d%d%d%d%d%d%d%d%d%d") then -- see if it matches our pattern
err_cat = ' ' .. set_error( 'bad_lccn' ); -- no match, set an error message
end
else
err_cat = ' ' .. set_error( 'bad_lccn' ); -- wrong length, set an error message
end
if not is_set (err_cat) and nil ~= lccn:find ('%s') then
err_cat = ' ' .. set_error( 'bad_lccn' ); -- lccn contains a space, set an error message
end
return external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
--[[
Format PMID and do simple error checking. PMIDs are sequential numbers beginning at 1 and counting up. This code checks the PMID to see that it
contains only digits and is less than test_limit; the value in local variable test_limit will need to be updated periodically as more PMIDs are issued.
]]
local function pmid(id)
local test_limit = 30000000; -- update this value as PMIDs approach
local handler = cfg.id_handlers['PMID'];
local err_cat = ''; -- presume that PMID is valid
if id:match("[^%d]") then -- if PMID has anything but digits
err_cat = ' ' .. set_error( 'bad_pmid' ); -- set an error message
else -- PMID is only digits
local id_num = tonumber(id); -- convert id to a number for range testing
if 1 > id_num or test_limit < id_num then -- if PMID is outside test limit boundaries
err_cat = ' ' .. set_error( 'bad_pmid' ); -- set an error message
end
end
return external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
--[[--------------------------< I S _ E M B A R G O E D >------------------------------------------------------
Determines if a PMC identifier's online version is embargoed. Compares the date in |embargo= against today's date. If embargo date is
in the future, returns the content of |embargo=; otherwise, returns and empty string because the embargo has expired or because
|embargo= was not set in this cite.
]]
local function is_embargoed (embargo)
if is_set (embargo) then
local lang = mw.getContentLanguage();
local good1, embargo_date, good2, todays_date;
good1, embargo_date = pcall( lang.formatDate, lang, 'U', embargo );
good2, todays_date = pcall( lang.formatDate, lang, 'U' );
if good1 and good2 then -- if embargo date and today's date are good dates
if tonumber( embargo_date ) >= tonumber( todays_date ) then -- is embargo date is in the future?
return embargo; -- still embargoed
else
add_maint_cat ('embargo')
return ''; -- unset because embargo has expired
end
end
end
return ''; -- |embargo= not set return empty string
end
--[[--------------------------< P M C >------------------------------------------------------------------------
Format a PMC, do simple error checking, and check for embargoed articles.
The embargo parameter takes a date for a value. If the embargo date is in the future the PMC identifier will not
be linked to the article. If the embargo date is today or in the past, or if it is empty or omitted, then the
PMC identifier is linked to the article through the link at cfg.id_handlers['PMC'].prefix.
PMC embargo date testing is done in function is_embargoed () which is called earlier because when the citation
has |pmc=<value> but does not have a |url= then |title= is linked with the PMC link. Function is_embargoed ()
returns the embargo date if the PMC article is still embargoed, otherwise it returns an empty string.
PMCs are sequential numbers beginning at 1 and counting up. This code checks the PMC to see that it contains only digits and is less
than test_limit; the value in local variable test_limit will need to be updated periodically as more PMCs are issued.
]]
local function pmc(id, embargo)
local test_limit = 5000000; -- update this value as PMCs approach
local handler = cfg.id_handlers['PMC'];
local err_cat = ''; -- presume that PMC is valid
local text;
if id:match("[^%d]") then -- if PMC has anything but digits
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
else -- PMC is only digits
local id_num = tonumber(id); -- convert id to a number for range testing
if 1 > id_num or test_limit < id_num then -- if PMC is outside test limit boundaries
err_cat = ' ' .. set_error( 'bad_pmc' ); -- set an error message
end
end
if is_set (embargo) then -- is PMC is still embargoed?
text="[[" .. handler.link .. "|" .. handler.label .. "]]:" .. handler.separator .. id .. err_cat; -- still embargoed so no external link
else
text = external_link_id({link = handler.link, label = handler.label, -- no embargo date or embargo has expired, ok to link to article
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
return text;
end
-- Formats a DOI and checks for DOI errors.
-- DOI names contain two parts: prefix and suffix separated by a forward slash.
-- Prefix: directory indicator '10.' followed by a registrant code
-- Suffix: character string of any length chosen by the registrant
-- This function checks a DOI name for: prefix/suffix. If the doi name contains spaces or endashes,
-- or, if it ends with a period or a comma, this function will emit a bad_doi error message.
-- DOI names are case-insensitive and can incorporate any printable Unicode characters so the test for spaces, endash,
-- and terminal punctuation may not be technically correct but it appears, that in practice these characters are rarely if ever used in doi names.
local function doi(id, inactive)
local cat = ""
local handler = cfg.id_handlers['DOI'];
local text;
if is_set(inactive) then
local inactive_year = inactive:match("%d%d%d%d") or ''; -- try to get the year portion from the inactive date
text = "[[" .. handler.link .. "|" .. handler.label .. "]]:" .. id;
if is_set(inactive_year) then
table.insert( z.error_categories, "Pages with DOIs inactive since " .. inactive_year );
else
table.insert( z.error_categories, "Pages with inactive DOIs" ); -- when inactive doesn't contain a recognizable year
end
inactive = " (" .. cfg.messages['inactive'] .. " " .. inactive .. ")"
else
text = external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
inactive = ""
end
if nil == id:match("^10%.[^%s–]-/[^%s–]-[^%.,]$") then -- doi must begin with '10.', must contain a fwd slash, must not contain spaces or endashes, and must not end with period or comma
cat = ' ' .. set_error( 'bad_doi' );
end
return text .. inactive .. cat
end
--[[--------------------------< O P E N L I B R A R Y >--------------------------------------------------------
Formats an OpenLibrary link, and checks for associated errors.
]]
local function openlibrary(id)
local code = id:match("^%d+([AMW])$"); -- only digits followed by 'A', 'M', or 'W'
local handler = cfg.id_handlers['OL'];
if ( code == "A" ) then
return external_link_id({link=handler.link, label=handler.label,
prefix=handler.prefix .. 'authors/OL',
id=id, separator=handler.separator, encode = handler.encode})
elseif ( code == "M" ) then
return external_link_id({link=handler.link, label=handler.label,
prefix=handler.prefix .. 'books/OL',
id=id, separator=handler.separator, encode = handler.encode})
elseif ( code == "W" ) then
return external_link_id({link=handler.link, label=handler.label,
prefix=handler.prefix .. 'works/OL',
id=id, separator=handler.separator, encode = handler.encode})
else
return external_link_id({link=handler.link, label=handler.label,
prefix=handler.prefix .. 'OL',
id=id, separator=handler.separator, encode = handler.encode}) .. ' ' .. set_error( 'bad_ol' );
end
end
--[[--------------------------< M E S S A G E _ I D >----------------------------------------------------------
Validate and format a usenet message id. Simple error checking, looks for 'id-left@id-right' not enclosed in
'<' and/or '>' angle brackets.
]]
local function message_id (id)
local handler = cfg.id_handlers['USENETID'];
text = external_link_id({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
if not id:match('^.+@.+$') or not id:match('^[^<].*[^>]$')then -- doesn't have '@' or has one or first or last character is '< or '>'
text = text .. ' ' .. set_error( 'bad_message_id' ) -- add an error message if the message id is invalid
end
return text
end
end
--[[--------------------------< C L E A N _ I S B N >----------------------------------------------------------
--[[--------------------------< H Y P H E N _ T O _ D A S H >--------------------------------------------------
Removes irrelevant text and dashes from ISBN number
Similar to that used for Special:BookSources
Converts a hyphen to a dash
]]
local function clean_isbn( isbn_str )
return isbn_str:gsub( "[^-0-9X]", "" );
end
--[[--------------------------< 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 >--------------------------------
Strip wiki italic and bold markup from argument so that it doesn't contaminate COinS metadata.
This function strips common patterns of apostrophe markup. We presume that editors who have taken the time to
markup a title have, as a result, provided valid markup. When they don't, some single apostrophes are left behind.
]]
local function strip_apostrophe_markup (argument)
if not is_set (argument) then return argument; end
while true do
if argument:match ("%'%'%'%'%'") then -- bold italic (5)
argument=argument:gsub("%'%'%'%'%'", ""); -- remove all instances of it
elseif argument:match ("%'%'%'%'") then -- italic start and end without content (4)
argument=argument:gsub("%'%'%'%'", "");
elseif argument:match ("%'%'%'") then -- bold (3)
argument=argument:gsub("%'%'%'", "");
elseif argument:match ("%'%'") then -- italic (2)
argument=argument:gsub("%'%'", "");
else
break;
end
end
return argument; -- done
end
--[[--------------------------< M A K E _ C O I N S _ T I T L E >----------------------------------------------
Makes a title for COinS from Title and / or ScriptTitle (or any other name-script pairs)
Apostrophe markup (bold, italics) is stripped from each value so that the COinS metadata isn't correupted with strings
of %27%27...
]]
local function make_coins_title (title, script)
if is_set (title) then
title = strip_apostrophe_markup (title); -- strip any apostrophe markup
else
title=''; -- if not set, make sure title is an empty string
end
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 >------------------------------------------------
Extract page numbers from external wikilinks in any of the |page=, |pages=, or |at= parameters for use in COinS.
]]
local function get_coins_pages (pages)
local pattern;
if not is_set (pages) then return pages; end -- if no page numbers then we're done
while true do
pattern = pages:match("%[(%w*:?//[^ ]+%s+)[%w%d].*%]"); -- pattern is the opening bracket, the url and following space(s): "[url "
if nil == pattern then break; end -- no more urls
pattern = escape_lua_magic_chars (pattern); -- pattern is not a literal string; escape lua's magic pattern characters
pages = pages:gsub(pattern, ""); -- remove as many instances of pattern as possible
end
pages = pages:gsub("[%[%]]", ""); -- remove the brackets
pages = pages:gsub("–", "-" ); -- replace endashes with hyphens
pages = pages:gsub("&%w+;", "-" ); -- and replace html entities (– etc.) with hyphens; do we need to replace numerical entities like   and the like?
return pages;
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*$", "%1");
end));
end
-- Converts a hyphen to a dash
local function hyphen_to_dash( str )
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
return str:gsub( '-', '–' );
end
--[[--------------------------< S A F E _ J O I N >------------------------------------------------------------
]]
local function list_people(control, people, etal, list_name) -- TODO: why is list_name here? not used in this function
local sep;
local namesep;
end
return name, etal; --
end
--[[--------------------------< N A M E _ H A S _ M U L T _ N A M E S >----------------------------------------
Evaluates the content of author and editor (surnames only) parameters for multiple names. Multiple names are
indicated if there is more than one comma and or semicolon. If found, the function adds the multiple name
(author or editor) maintenance category.
]]
local function name_has_mult_names (name, list_name)
local count, _;
if is_set (name) then
if name:match ('^%(%(.*%)%)$') then -- if wrapped in doubled parentheses, ignore
name = name:match ('^%(%((.*)%)%)$'); -- strip parens
else
_, count = name:gsub ('[;,]', ''); -- count the number of separator-like characters
if 1 < count then -- param could be |author= or |editor= so one separactor character is acceptable
add_maint_cat ('mult_names', list_name:lower()); -- more than one separator indicates multiple names so add a maint cat for this template
end
end
end
return name; -- and done
end
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 = name_has_mult_names (last, err_msg_list_name); -- check for multiple names in last and its aliases
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
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
if is_set (link) and false == link_param_ok (link) then -- do this test here in case link is missing last
table.insert( z.message_tail, { set_error( 'bad_paramlink', list_name:match ("(%w+)List"):lower() .. '-link' .. i )}); -- url or wikilink in author link;
end
names[n] = {last = last, first = first, link = link, mask = mask, corporate=false}; -- add this name to our names list (corporate for |vauthors= only)
n = n + 1; -- point to next location in the names table
end
--[[--------------------------< BG UE T _ I LS DO _6 I3 D9 _ LC IO SD TE >--------------------------------------------------------
Validates language names provided in |language= parameter if not an ISO639-1 or 639-2 code.
Populates ID table from arguments using configuration settings. Loops through cfg.id_handlers and searches args for
any of the parameters listed in each cfg.id_handlers['...'].parameters. If found, adds the parameter and value to
the identifier list. Emits redundant error message is more than one alias exists in args
Returns the language name and associated two- or three-character code. Because case of the source may be incorrect
]]
or different from the case that WikiMedia uses, the name comparisons are done in lower case and when a match is
found, the Wikimedia version (assumed to be correct) is returned along with the code. When there is no match, we
return the original language name string.
mw.language.fetchLanguageNames(<local wiki language>, 'all') return a list of languages that in some cases may include
local function extract_ids( args )
extensions. For example, code 'cbk-zam' and its associated name 'Chavacano de Zamboanga' (MediaWiki does not support
local id_list = {}; -- list of identifiers found in args
code 'cbk' or name 'Chavacano'.
for k, v in pairs( cfg.id_handlers ) do -- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
v = select_one( args, v.parameters, 'redundant_parameters' ); -- v.parameters is a table of aliases for k; here we pick one from args if present
if is_set(v) then id_list[k] = v; end -- if found in args, add identifier to our list
end
return id_list;
end
Names but that are included in the list will be found if that name is provided in the |language= parameter. For example,
--[[--------------------------< B U I L D _ I D _ L I S T >--------------------------------------------------------
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.
Takes a table of IDs created by extract_ids() and turns it into a table of formatted ID outputs.
inputs:
id_list – table of identifiers built by extract_ids()
options – table of various template parameter values used to modify some manually handled identifiers
]]
local function build_id_list( id_list, options )
local new_list, handler = {};
function fallback(k) return { __index = function(t,i) return cfg.id_handlers[k][i] end } end;
for k, v in pairs( id_list ) do -- k is uc identifier name as index to cfg.id_handlers; e.g. cfg.id_handlers['ISBN'], v is a table
-- fallback to read-only cfg
handler = setmetatable( { ['id'] = v }, fallback(k) );
if handler.mode == 'external' then
table.insert( new_list, {handler.label, external_link_id( handler ) } );
elseif handler.mode == 'internal' then
table.insert( new_list, {handler.label, internal_link_id( handler ) } );
elseif handler.mode ~= 'manual' then
error( cfg.messages['unknown_ID_mode'] );
elseif k == 'DOI' then
table.insert( new_list, {handler.label, doi( v, options.DoiBroken ) } );
elseif k == 'ARXIV' then
table.insert( new_list, {handler.label, arxiv( v, options.Class ) } );
elseif k == 'ASIN' then
table.insert( new_list, {handler.label, amazon( v, options.ASINTLD ) } );
elseif k == 'LCCN' then
table.insert( new_list, {handler.label, lccn( v ) } );
elseif k == 'OL' or k == 'OLA' then
table.insert( new_list, {handler.label, openlibrary( v ) } );
elseif k == 'PMC' then
table.insert( new_list, {handler.label, pmc( v, options.Embargo ) } );
elseif k == 'PMID' then
table.insert( new_list, {handler.label, pmid( v ) } );
elseif k == 'ISMN' then
table.insert( new_list, {handler.label, ismn( v ) } );
elseif k == 'ISSN' then
table.insert( new_list, {handler.label, issn( v ) } );
elseif k == 'ISBN' then
local ISBN = internal_link_id( handler );
if not check_isbn( v ) and not is_set(options.IgnoreISBN) then
ISBN = ISBN .. set_error( 'bad_isbn', {}, false, " ", "" );
end
table.insert( new_list, {handler.label, ISBN } );
elseif k == 'USENETID' then
table.insert( new_list, {handler.label, message_id( v ) } );
else
error( cfg.messages['unknown_manual_ID'] );
end
end
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
--[[--------------------------< C O I N S >--------------------------------------------------------------------
COinS metadata (see <http://ocoins.info/>) allows automated tools to parse the citation information.
]]
local function COinS(data, class)
if 'table' ~= type(data) or nil == next(data) then
return '';
end
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
OCinSoutput["rft.genre"] = "article"; -- journal and other 'periodical' articles
end
OCinSoutput["rft.jtitle"] = data.Periodical; -- journal only
if is_set (data.Map) then
OCinSoutput["rft.atitle"] = data.Map; -- for a map in a periodical
else
OCinSoutput["rft.atitle"] = data.Title; -- all other 'periodical' article titles
end
-- 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
--[[
for k, v in pairs( data.ID_list ) do -- what to do about these? For now assume that they are common to all?
local id, value = cfg.id_handlers[k].COinS;
if k == 'ISBN' then value = clean_isbn( v ); else value = v; end
if string.sub( id or "", 1, 4 ) == 'info' then
OCinSoutput["rft_id"] = table.concat{ id, "/", v };
else
OCinSoutput[ id ] = value;
end
end
]]
local last, first;
for k, v in ipairs( data.Authors ) do
last, first = v.last, v.first;
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
--[[--------------------------< G E T _ I S O 6 3 9 _ C O D E >------------------------------------------------
Validates language names provided in |language= parameter if not an ISO639-1 code. Handles the special case that is Norwegian where
ISO639-1 code 'no' is mapped to language name 'Norwegian Bokmål' by Extention:CLDR.
Returns the language name and associated ISO639-1 code. Because case of the source may be incorrect or different from the case that Wikimedia
uses, the name comparisons are done in lower case and when a match is found, the Wikimedia version (assumed to be correct) is returned along
with the code. When there is no match, we return the original language name string.
mw.language.fetchLanguageNames() will return a list of languages that aren't part of ISO639-1. Names that aren't ISO639-1 but that are included
in the list will be found if that name is provided in the |language= parameter. For example, if |language=Samaritan Aramaic, that name will be
found with the associated code 'sam', not an ISO639-1 code. When names are found and the associated code is not two characters, this function
returns only the Wikimedia language name.
Adapted from code taken from Module:Check ISO 639-1.
]]
local function get_iso639_code (lang, this_wiki_code)
local languages = mw.language.fetchLanguageNames(this_wiki_code, 'all') -- get a list of language names known to Wikimedia
if 'norwegian' == lang:lower() then -- special case related to Wikimedia remap of code 'no' at Extension:CLDR
return 'Norwegian', 'no'; -- Make sure rendered version is properly capitalized
end
local languages = mw.language.fetchLanguageNames('en', '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 -- ISO639two-1 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 ISO639-1the code
end
end
return lang; -- not valid language; return language in original case and nil for ISO639-1the code
end
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
GetGets language name from ISO639a provided two-1 codeor valuethree-character ISO provided639 code. If a code is validrecognized useby the returned name; if notMediaWiki, then use the value that was provided with the language parameter.
use the returned name; if not, then use the value that was provided with the language parameter.
When |language= contains a recognized language (either code or name), the page is assigned to the category for
There is an exception. There are three ISO639-1 codes for Norewegian language variants. There are two official variants: Norwegian Bokmål (code 'nb') and
that code: Category:Norwegian-language sources (no). For valid three-character code languages, the page is assigned
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
to the single category for '639-2' codes: Category:CS1 ISO 639-2 language sources.
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.
Languages that are the same as the local wiki are not categorized. MediaWiki does not recognize three-character
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
equivalents of two-character codes: code 'ar' is recognized bit code 'ara' is not.
return 'Norwegian' when editors use |language=no. We presume that editors don't know about the variants or can't descriminate between them.
This function supports multiple languages in the form |language=nb, French, th where the language names or codes are
See Help talk:Citation Style_1#An ISO 639-1 language name test
separated from each other by commas.
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 ISO639two-1 twoor 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
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 -- ISO639if two-1or languagethree-character code are 2 characters (fetchLanguageName also supports 3 character codes)
name = mw.language.fetchLanguageName( lang:lower(), "en" this_wiki_code); -- get ISO 639-1 language name if Language|language= is a proper code
end
if is_set (name) then -- if Language|language= specified a valid ISO639-1 code
code = lang:lower(); -- save it
else
name, code = get_iso639_code (lang, this_wiki_code); -- attempt to get code from name (assign name here so that we are sure of proper capitalization)
end
if is_set (code) then -- only 2- or 3-character codes
if 'no'this_wiki_code =~= code then name = 'Norwegian' end; -- overridewhen wikimediathe when codelanguage is not the same as this wiki'no's language
if 'en'2 ~== code:len() then -- Englishand notis thea languagetwo-character code
add_prop_cat ('foreign_lang_source', {name, code}) -- categorize it
else -- or is a recognized language (but has a three-character code)
add_prop_cat ('foreign_lang_source_2', {code}) -- categorize it differently TODO: support mutliple three-character code categories per cs1|2 template
end
end
else
name = table.concat (language_list, ', ') -- and concatenate with '<comma><space>' separators
end
if 'English'this_wiki_name == name then
return ''; -- if one language and that language is Englishthis wiki's return an empty string (no annotation)
end
return (" " .. wrap_msg ('language', name)); -- otherwise wrap with '(in ...)'
--[[--------------------------< 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 may or may not limitdefines the lengthnumber of thenames displayed for author orand editor name lists. and a boolean flag
to indicate when et al. should be appended to the name list.
When the value assigned to |display-authorsxxxxors= 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-authorsxxxxors= is the keyword 'etal', return a number that is one greater than the
number of authors in the list and set the 'etal' flag true. This will cause the list_people() to display all of
the names in the name list followed by 'et al.'
In all other cases, returns nil and the previous state of the 'etal' flag.
inputs:
max: A['DisplayAuthors'] or A['DisplayEditors']; a number or some flavor of etal
count: #a or #e
list_name: 'authors' or 'editors'
etal: author_etal or editor_etal
]]
elseif max:match ('^%d+$') then -- if is a string of numbers
max = tonumber (max); -- make it a number
if max >= count and 'authors' == list_name then -- AUTHORS ONLY -- if |display-xxxxors= value greater than or equal to number of authors/editors
add_maint_cat ('disp_auth_ed', list_name);
end
else -- not a valid keyword or number
table.insert( z.message_tail, { set_error( 'invalid_param_val', {'display-' .. list_name, max}, true ) } ); -- add error message
max = nil; -- unset; as if |display-xxxxors= had not been set
end
elseif 'authors' == list_name then -- AUTHORS ONLY need to clear implicit et al category
max = count + 1; -- number of authors + 1
end
This function sets the vancouver error when a reqired comma is missing and when there is a space between an author's initials.
TODO: check for names like Coon V JS (Coon JS 5th at PMID 25205766, John S. Coon V at doi:10.1093/humupd/dmu048)?
]]
corporate = true;
elseif string.find(v_name, "%s") then
local lastfirstTable = {}
lastfirstTable = mw.text.split(v_name, "%s")
first = table.remove(lastfirstTable); -- removes and returns value of last element in table which should be author intials
last = table.concat(lastfirstTable, " ") -- returns a string that is the concatenation of all other names that are not initials
if mw.ustring.match (last, '%a+%s+%u+%s+%a+') or mw.ustring.match (v_name, ' %u %u$') then
add_vanc_error (); -- matches last II last; the case when a comma is missing or a space between two intiials
end
end
--[[-------------------------< N O R M A L I Z E _ P A G E _ L I S T >-----------------------------------------
not currently used
normalizes a comma, ampersand, and/or space separated list to be '<value>, <value>, ..., <value>'
returns list unchanged if there are no commas else strips whitespace and then reformats the list
]]
--[[
local function normalize_page_list (list)
if not list:find ('[,& ]') then return list end -- if list is not delimited with commas, ampersands, or spaces; done
list = mw.text.split (list, '[,&%s]+'); -- make a table of values
list = table.concat (list, ', '); -- and now make a normalized list
return list;
end
]]
--[[-------------------------< F O R M A T _ P A G E S _ S H E E T S >-----------------------------------------
return '', '', '', ''; -- return empty strings
end
--[=[-------------------------< A R C H I V E _ U R L _ C H E C K >--------------------------------------------
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. 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.
When the archive.org url does not have a complete timestamp, archive.org chooses a snapshot according to its own
algorithm or provides a 'search' result. [[WP:ELNO]] discourages links to search results.
This function looks at the value assigned to |archive-url= and returns empty strings for |archive-url= and
|archive-date= and an error message when:
|archive-url= holds an archive.org save command url
|archive-url= is an archive.org url that does not have a complete timestamp (YYYYMMDDhhmmss 14 digits) in the correct place
otherwise returns |archive-url= and |archive-date=
]=]
local function archive_url_check (url, date)
if url:match('//web\.archive\.org/') then -- for archive.org urls:
if url:match('//web\.archive\.org/save/') then -- if a save command url, we don't want to save target page
table.insert( z.message_tail, { set_error( 'archive_url', {'save command'}, true ) } ); -- add error message
return '', ''; -- return empty strings for archiveURL and ArchiveDate
elseif url:match('//web\.archive\.org/web/%d%d%d%d%d%d%d%d%d%d%d%d%d%d/') then -- if there is what looks like a correct timestamp
return url, date; -- return archiveURL and ArchiveDate
else -- malformed url
table.insert( z.message_tail, { set_error( 'archive_url', {'timestamp'}, true ) } ); -- add error message
return '', ''; -- return empty strings for archiveURL and ArchiveDate
end
end
return url, date; -- not an archive.org archive
end
--[[--------------------------< M I S S I N G _ P I P E _ C H E C K >------------------------------------------
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 parameters. If 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 missing_pipe_check (value)
local capture;
value = value:gsub ('%b<>', ''); -- remove xml/html tags because attributes: class=, title=, etc
capture = value:match ('%s+(%a[%a%d]+)%s*=') or value:match ('^(%a[%a%d]+)%s*='); -- find and categorize parameters with possible missing pipes
if capture and validate (capture) then -- if the capture is a valid parameter name
add_maint_cat ('missing_pipe');
end
end
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
local Authors;
local NameListFormat = A['NameListFormat'];
local Collaboration = A['Collaboration'];
do -- to limit scope of selected
elseif 3 == selected then
Authors = A['Authors']; -- use content of |authors=
end
if is_set (Collaboration) then
author_etal = true; -- so that |display-authors=etal not required
end
end
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
if is_set (TitleLink) and false == link_param_ok (TitleLink) then
table.insert( z.message_tail, { set_error( 'bad_paramlink', A:ORIGIN('TitleLink'))}); -- url or wikilink in |title-link=;
end
local Chapter = A['Chapter'];
local Docket = A['Docket'];
local ArchiveFormat = A['ArchiveFormat'];
local ArchiveURL = A['ArchiveURL'];
local ArchiveDate;
local ArchiveURL;
ArchiveURL, ArchiveDate = archive_url_check (A['ArchiveURL'], A['ArchiveDate'])
-- local ArchiveDate = A['ArchiveDate'];
-- local ArchiveURL = A['ArchiveURL'];
-- if ArchiveURL:match('//web\.archive\.org/save/') then -- if an archive.org save command url, we don't want to save target page ...
-- ArchiveURL = ''; -- every time a reader clicks the link so
-- ArchiveDate = ''; -- unset these
-- table.insert( z.message_tail, { set_error( 'archive_save', {}, true ) } ); -- and add error message
-- end
local DeadURL = A['DeadURL']
if not is_valid_parameter_value (DeadURL, 'dead-url', cfg.keywords ['deadurl']) then -- set in config.defaults to 'yes'
DeadURL = ''; -- anything else, set to empty string
end
local URL = A['URL']
local URLorigin = A:ORIGIN('URL'); -- get name of parameter that holds URL
local Via = A['Via'];
local AccessDate = A['AccessDate'];
local ArchiveDate = A['ArchiveDate'];
local Agency = A['Agency'];
local DeadURL = A['DeadURL']
if not is_valid_parameter_value (DeadURL, 'dead-url', cfg.keywords ['deadurl']) then -- set in config.defaults to 'yes'
DeadURL = ''; -- anything else, set to empty string
end
local Language = A['Language'];
local anchor_year; -- used in the CITEREF identifier
local COinS_date = {}; -- holds date info extracted from |date= for the COinS metadata by Module:Date verification
local DF = A['DF']; -- date format set in cs1|2 template
if not is_valid_parameter_value (DF, 'df', cfg.keywords['date-format']) then -- validate reformatting keyword
DF = ''; -- not valid, set to empty string
end
-- set default parameter values defined by |mode= parameter. If |mode= is empty or omitted, use CitationClass to set these values
ID = A['Number']; -- yes, use it
else -- ID has a value so emit error message
-- ID = ID .. " " .. set_error('redundant_parameters', '<code>|id=</code> and <code>|number=</code>');
table.insert( z.message_tail, { set_error('redundant_parameters', {wrap_style ('parameter', 'id') .. ' and ' .. wrap_style ('parameter', 'number')}, true )});
end
local AirDate = A['AirDate'];
local SeriesLink = A['SeriesLink'];
if is_set (SeriesLink) and false == link_param_ok (SeriesLink) then
link_title_ok (SeriesLink, A:ORIGIN ('SeriesLink'), Series, 'series'); -- check for wikimarkup in |series-link= or wikimarkup in |series= when |series-link= is set
table.insert( z.message_tail, { set_error( 'bad_paramlink', A:ORIGIN('SeriesLink'))});
end
local Network = A['Network'];
local Station = A['Station'];
Date validation supporting code is in Module:Citation/CS1/Date_validation
]]
do -- create defined block to contain local variables error_message, anddate_parameters_list, mismatch
local error_message = '';
-- AirDate has been promoted to Date so not necessary to check it
anchor_year,local error_messagedate_parameters_list = dates({['access-date']=AccessDate, ['archive-date']=ArchiveDate, ['date']=Date, ['doi-broken-date']=DoiBroken,
['embargo']=Embargo, ['lay-date']=LayDate, ['publication-date']=PublicationDate, ['year']=Year}, COinS_date);
anchor_year, Embargo, error_message = dates(date_parameters_list, COinS_date);
if is_set (Year) and is_set (Date) then -- both |date= and |year= not normally needed;
if is_set(error_message) then
table.insert( z.message_tail, { set_error( 'bad_date', {error_message}, true ) } ); -- add this error message
elseif is_set (DF) then
if reformat_dates (date_parameters_list, DF, false) then -- reformat to DF format, use long month names if appropriate
AccessDate = date_parameters_list['access-date']; -- overwrite date holding parameters with reformatted values
ArchiveDate = date_parameters_list['archive-date'];
Date = date_parameters_list['date'];
DoiBroken = date_parameters_list['doi-broken-date'];
LayDate = date_parameters_list['lay-date'];
PublicationDate = date_parameters_list['publication-date'];
end
end
end -- end of do
[A:ORIGIN('Chapter')]=Chapter,
[A:ORIGIN('Periodical')]=Periodical,
[A:ORIGIN('PublisherName')] = PublisherName,
});
['Encyclopedia'] = Encyclopedia,
['Chapter'] = make_coins_title (coins_chapter, ScriptChapter), -- Chapter and ScriptChapter stripped of bold / italic wikimarkup
['Map'] = Map,
['Degree'] = Degree; -- cite thesis only
['Title'] = make_coins_title (coins_title, ScriptTitle), -- Title and ScriptTitle stripped of bold / italic wikimarkup
do
local last_first_list;
-- local maximum;
local control = {
format = NameListFormat, -- empty string or 'vanc'
do -- do editor name list first because coauthors can modify control table
control.maximum , editor_etal = get_display_authors_editors (A['DisplayEditors'], #e, 'editors', editor_etal);
last_first_list, EditorCount = list_people(control, e, editor_etal);
-- Preserve old-style implicit et al.
if not is_set(maximum) and #e == 4 then
maximum = 3;
table.insert( z.message_tail, { set_error('implict_etal_editor', {}, true) } );
end
control.maximum = maximum;
last_first_list, EditorCount = list_people(control, e, editor_etal, 'editor');
if is_set (Editors) then
do -- now do translators
control.maximum = #t; -- number of translators
Translators = list_people(control, t, false, 'translator'); -- et al not currently supported
end
do -- now do contributors
control.maximum = #c; -- number of contributors
Contributors = list_people(control, c, false, 'contributor'); -- et al not currently supported
end
do -- now do authors
end
last_first_list = list_people(control, a, author_etal, 'author');
if is_set (Authors) then
end
end -- end of do
if is_set (Authors) and is_set (Collaboration) then
Authors = Authors .. ' (' .. Collaboration .. ')'; -- add collaboration after et al.
end
if not is_set(Authors) and is_set(Coauthors) then -- coauthors aren't displayed if one of authors=, authorn=, or lastn= isn't specified
end
if not is_set(URL) then --and
if in_array(config.CitationClass, {"web","podcast", "mailinglist"}) then -- Test|url= ifrequired for cite web or, cite podcast, |url=and is missing or emptycite mailinglist
table.insert( z.message_tail, { set_error( 'cite_web_url', {}, true ) } );
end
-- Testdo ifwe have |accessdate is given= without givingeither |url= aor URL|chapter-url=?
if is_set(AccessDate) and not is_set(ChapterURL)then -- ChapterURL may be set when theURL others areis not set; TODO: move this to a separate test?
table.insert( z.message_tail, { set_error( 'accessdate_missing_url', {}, true ) } );
AccessDate = '';
DeadURL = DeadURL:lower(); -- used later when assembling archived text
if is_set( ArchiveURL ) then
if is_set (ChapterURL) then -- swapped -- URL not set so if chapter-url is set apply archive url to it
if is_set (URL) then
OriginalURL = URL; -- save copy of original source URL
OriginalURLorigin = URLorigin; -- name of url parameter for error messages
OriginalFormat = Format; -- and original |format=
if 'no' ~= DeadURL then -- if URL set then archive-url applies to it
URL = ArchiveURL -- swap-in the archive's url
URLorigin = A:ORIGIN('ArchiveURL') -- name of archive url parameter for error messages
Format = ArchiveFormat or ''; -- swap in archive's format
end
elseif is_set (ChapterURL) then -- URL not set so if chapter-url is set apply archive url to it
OriginalURL = ChapterURL; -- save copy of source chapter's url for archive text
OriginalURLorigin = ChapterURLorigin; -- name of chapter-url parameter for error messages
ChapterFormat = ArchiveFormat or ''; -- swap in archive's format
end
elseif is_set (URL) then
end
OriginalURL = URL; -- save copy of original source URL
OriginalURLorigin = URLorigin; -- name of url parameter for error messages
OriginalFormat = Format; -- and original |format=
if 'no' ~= DeadURL then -- if URL set then archive-url applies to it
URL = ArchiveURL -- swap-in the archive's url
URLorigin = A:ORIGIN('ArchiveURL') -- name of archive url parameter for error messages
Format = ArchiveFormat or ''; -- swap in archive's format
end
end
end
Chapter = format_chapter_title (ScriptChapter, Chapter, TransChapter, ChapterURL, ChapterURLorigin, no_quotes); -- Contribution is also in Chapter
if is_set (Chapter) then
Chapter = Chapter .. ChapterFormat ;
if 'map' == config.CitationClass and is_set (TitleType) then
Chapter = Chapter .. ' ' .. TitleType; -- map annotation here; not after title
end
Chapter = Chapter .. ChapterFormat .. sepc .. ' ';
elseif is_set (ChapterFormat) then -- |chapter= not set but |chapter-format= is so ...
Chapter = ChapterFormat .. sepc .. ' '; -- ... ChapterFormat has error message, we want to see it
end
local TransError = "";
if is_set(TransTitle) then
if is_set(Title) then
if is_set(Title) then
if not is_set(TitleLink) and is_set(URL) then
Title = external_link( URL, Title, URLorigin ) .. TransError .. Format;
-- this experiment hidden 2016-04-10; see Help_talk:Citation_Style_1#Recycled_urls
URL = "";
-- local temp_title = external_link( URL, Title, URLorigin ) .. TransError .. Format; -- do this so we get error message even if url is usurped no archive
-- if in_array (DeadURL, {'unfit no archive', 'usurped no archive'}) then -- when url links to inappropriate location and there is no archive of original source available
-- local err_msg
-- if temp_title:match ('%[%S+%s+(.+)%](<.+)') then -- if there is an error message
-- Title, err_msg = temp_title:match ('%[%S+%s+(.+)%](<.+)'); -- strip off external link; TODO: find a better to do this
-- Title = Title .. (err_msg or '');
-- end
-- else
-- Title = temp_title;
-- end
URL = ''; -- unset these because no longer needed
Format = "";
else
if ('harv' == Ref ) then
local namelist = {}; -- holds selected contributor, author, editor name list
-- local year = first_set (Year, anchor_year); -- Year first for legacy citations and for YMD dates that require disambiguation
local year = first_set ({Year, anchor_year}, 2); -- Year first for legacy citations and for YMD dates that require disambiguation
end
if is_set(options.id) then -- here we wrap the rendered citation in <cite ...>...</cite> tags
text = substitute (cfg.presentation['<cite -id="'], .. {mw.uri.anchorEncode(options.id) ..'" class="' .., mw.text.nowiki(options.class), ..text}); -- '">'when ..|ref= textis .. "</cite>";set
else
text = substitute (cfg.presentation['<cite class="'], .. {mw.text.nowiki(options.class), ..text}); -- '">'all ..other text .. "</cite>";cases
end
text = text .. substitute (cfg.presentation['ocins'], {OCinSoutput}); -- append metadata to the citation
local empty_span = '<span style="display:none;"> </span>';
-- Note: Using display: none on the COinS span breaks some clients.
local OCinS = '<span title="' .. OCinSoutput .. '" class="Z3988">' .. empty_span .. '</span>';
text = text .. OCinS;
if #z.message_tail ~= 0 then
end
--[[--------------------------< H AC S _1 I. N VC I ST IA BT LI EO _ C H A R SN >------------------------------------------------------
This function searches a parameter's value for nonprintable or invisible characters. The search stops at the first match.
Sometime after this module is done with rendering a citation, some C0 control characters are replaced with the
replacement character. That replacement character is not detected by this test though it is visible to readers
of the rendered citation. This function will detect the replacement character when it is part of the wikisource.
Output of this function is an error message that identifies the character or the Unicode group that the character
belongs to along with its position in the parameter value.
]]
local function has_invisible_chars (param, v)
local position = '';
local i=1;
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 = mw.ustring.find (v, pattern) -- see if the parameter value contains characters that match the pattern
if position then
table.insert( z.message_tail, { set_error( 'invisible_char', {char, wrap_style ('parameter', param), position}, true ) } ); -- add error message
return; -- and done with this parameter
end
i=i+1; -- bump our index
end
end
--[[--------------------------< Z . C I T A T I O N >----------------------------------------------------------
This is used by templates such as {{cite book}} to create the actual citation text.
]]
function zcs1.citation(frame)
local pframe = frame:getParent()
local validation, utilities, identifiers, metadata;
if nil ~= string.find (frame:getTitle(), 'sandbox', 1, true) then -- did the {{#invoke:}} use sandbox version?
cfg = mw.loadData ('Module:Citation/CS1/Configuration/sandbox'); -- load sandbox versions of Configurationsupport and Whitelist and ...modules
whitelist = mw.loadData ('Module:Citation/CS1/Whitelist/sandbox');
validationutilities = require ('Module:Citation/CS1/Date_validationUtilities/sandbox'); -- ... sandbox version of date validation code
validation = require ('Module:Citation/CS1/Date_validation/sandbox');
identifiers = require ('Module:Citation/CS1/Identifiers/sandbox');
metadata = require ('Module:Citation/CS1/COinS/sandbox');
else -- otherwise
cfg = mw.loadData ('Module:Citation/CS1/Configuration'); -- load live versions of Configurationsupport and Whitelist and ...modules
whitelist = mw.loadData ('Module:Citation/CS1/Whitelist');
validationutilities = require ('Module:Citation/CS1/Date_validationUtilities'); -- ... live version of date validation code
validation = require ('Module:Citation/CS1/Date_validation');
identifiers = require ('Module:Citation/CS1/Identifiers');
metadata = require ('Module:Citation/CS1/COinS');
end
dates = validationutilities.datesset_selected_modules (cfg); -- importedso that functions in Utilities can see the cfg tables
identifiers.set_selected_modules (cfg, utilities); -- so that functions in Identifiers can see the selected cfg tables and selected Utilities module
validation.set_selected_modules (utilities); -- so that functions in Date validataion can see the selected Utilities module
metadata.set_selected_modules (cfg, utilities); -- so that functions in COinS can see the selected cfg tables and selected Utilities module
dates = validation.dates; -- imported functions from Module:Citation/CS1/Date validation
year_date_check = validation.year_date_check;
reformat_dates = validation.reformat_dates;
is_set = utilities.is_set; -- imported functions from Module:Citation/CS1/Utilities
in_array = utilities.in_array;
substitute = utilities.substitute;
error_comment = utilities.error_comment;
set_error = utilities.set_error;
select_one = utilities.select_one;
add_maint_cat = utilities.add_maint_cat;
wrap_style = utilities.wrap_style;
safe_for_italics = utilities.safe_for_italics;
remove_wiki_link = utilities.remove_wiki_link;
z = utilities.z; -- table of error and category tables in Module:Citation/CS1/Utilities
extract_ids = identifiers.extract_ids; -- imported functions from Module:Citation/CS1/Utilities
build_id_list = identifiers.build_id_list;
is_embargoed = identifiers.is_embargoed;
make_coins_title = metadata.make_coins_title; -- imported functions from Module:Citation/CS1/COinS
get_coins_pages = metadata.get_coins_pages;
COinS = metadata.COinS;
local args = {};
local suggestions = {};
end
end
missing_pipe_check (v); -- do we think that there is a parameter that is missing a pipe?
args[k] = v;
elseif args[k] ~= nil or (k == 'postscript') then
for k, v in pairs( args ) do
if 'string' == type (k) then -- don't evaluate positional parameters
has_invisible_chars (k, v)
has_invisible_chars (k, v);
end
end
return citation0( config, args)
end
return zcs1;
| 