Module:Citation/CS1
From Timelines
Revision as of 07:06, 13 October 2014 by Trappist the monk (talk) (Synch from sandbox; script error fix in strip_apostrophe_markup();)
Documentation for this module may be created at Module:Citation/CS1/doc
local z = {
error_categories = {}; -- for categorizing citations that contain errors
error_ids = {};
message_tail = {};
maintenance_cats = {}; -- for categorizing citations that aren't erroneous per se, but could use a little work
properties_cats = {}; -- for categorizing citations based on certain properties, language of source for instance
}
-- Whether variable is set or not
function is_set( var )
return not (var == nil or var == '');
end
-- First set variable or nil if none
function first_set(...)
local list = {...};
for _, var in pairs(list) do
if is_set( var ) then
return var;
end
end
end
-- Whether needle is in haystack
function inArray( needle, haystack )
if needle == nil then
return false;
end
for n,v in ipairs( haystack ) do
if v == needle then
return n;
end
end
return false;
end
--[[
Categorize and emit an error message when the citation contains one or more deprecated parameters. Because deprecated parameters (currently |day=, |month=,
|coauthor=, and |coauthors=) aren't related to each other and because these parameters may be concatenated into the variables used by |date= and |author#= (and aliases)
details of which parameter caused the error message are not provided. Only one error message is emitted regardless of the number of deprecated parameters in the citation.
]]
function deprecated_parameter()
if true ~= Page_in_deprecated_cat then -- if we haven't been here before then set a
Page_in_deprecated_cat=true; -- sticky flag so that if there are more than one deprecated parameter the category is added only once
table.insert( z.message_tail, { seterror( 'deprecated_params', {}, true ) } ); -- add error message
end
end
-- Populates numbered arguments in a message string using an argument table.
function substitute( msg, args )
-- return args and tostring( mw.message.newRawMessage( msg, args ) ) or msg;
return args and mw.message.newRawMessage( msg, args ):plain() or msg;
end
--[[
Apply kerning to open the space between the quote mark provided by the Module and a leading or trailing quote mark contained in a |title= or |chapter= parameter's value.
This function will positive kern either single or double quotes:
"'Unkerned title with leading and trailing single quote marks'"
" 'Kerned title with leading and trailing single quote marks' " (in real life the kerning isn't as wide as this example)
]]
function kern_quotes (str)
local left='<span style="padding-left:0.2em;">%1</span>'; -- spacing to use when title contains leading single or double quote mark
local right='<span style="padding-right:0.2em;">%1</span>'; -- spacing to use when title contains trailing single or double quote mark
if str:match ("^[\"\'][^\']") then
str = string.gsub( str, "^[\"\']", left, 1 ); -- replace (captured) leading single or double quote with left-side <span>
end
if str:match ("[^\'][\"\']$") then
str = string.gsub( str, "[\"\']$", right, 1 ); -- replace (captured) trailing single or double quote with right-side <span>
end
return str;
end
--[[--------------------------< F O R M A T _ S C R I P T _ V A L U E >----------------------------------------
|script-title= holds title parameters that are not written in Latin based scripts: Chinese, Japanese, Arabic, Hebrew, etc. These scripts should
not be italicized and may be written right-to-left. The value supplied by |script-title= is concatenated onto Title after Title has been wrapped
in italic markup.
Regardless of language, all values provided by |script-title= are wrapped in <bdi>...</bdi> tags to isolate rtl languages from the English left to right.
|script-title= provides a unique feature. The value in |script-title= may be prefixed with a two-character ISO639-1 language code and a colon:
|script-title=ja:*** *** (where * represents a Japanese character)
Spaces between the two-character code and the colon and the colon and the first script character are allowed:
|script-title=ja : *** ***
|script-title=ja: *** ***
|script-title=ja :*** ***
Spaces preceding the prefix are allowed: |script-title = ja:*** ***
The prefix is checked for validity. If it is a valid ISO639-1 language code, the lang attribute (lang="ja") is added to the <bdi> tag so that browsers can
know the language the tag contains. This may help the browser render the script more correctly. If the prefix is invalid, the lang attribute
is not added. At this time there is no error message for this condition.
At this writing, only |script-title= is supported. It is anticipated that additional parameters will be created to use this function.
TODO: error messages when prefix is invalid ISO639-1 code; when script_value has prefix but no script;
]]
function format_script_value (script_value)
local lang=''; -- initialize to empty string
local name;
if script_value:match('^%l%l%s*:') then -- if first 3 non-space characters are script language prefix
lang = script_value:match('^(%l%l)%s*:%s*%S.*'); -- get the language prefix or nil if there is no script
if not is_set (lang) then
return ''; -- script_value was just the prefix so return empty string
end
-- if we get this far we have prefix and script
name = mw.language.fetchLanguageName( lang, "en" ); -- get language name so that we can use it to categorize
if is_set (name) then -- is prefix a proper ISO 639-1 language code?
script_value = script_value:gsub ('^%l%l%s*:%s*', ''); -- strip prefix from script
if inArray (lang, {'ar', 'ja', 'ko', 'he', 'zh'}) then -- is prefix one of these language codes?
table.insert( z.properties_cats, 'CS1 uses ' .. name .. '-language script ('..lang..')'); -- categorize in language-specific categories
else
table.insert( z.properties_cats, 'CS1 uses foreign language script'); -- use this category as a catchall until language-specific category is available
end
lang = ' lang="' .. lang .. '" '; -- convert prefix into a lang attribute
else
lang = ''; -- invalid so set lang to empty string
end
end
script_value = '<bdi' .. lang .. '>' .. script_value .. '</bdi>'; -- isolate incase script is rlt
return script_value;
end
--[[--------------------------< S C R I P T _ C O N C A T E N A T E >------------------------------------------
Initially for |title= and |script-title=, this function concatenates those two parameter values after the script value has been
wrapped in <bdi> tags.
]]
function script_concatenate (title, script)
if is_set (script) then
script = format_script_value (script); -- <bdi> tags, lang atribute, categorization, etc; returns empty string on error
if is_set (script) then
title = title .. ' ' .. script; -- concatenate title and script title
end
end
return title;
end
-- Wraps a string using a message_list configuration taking one argument
function wrap( key, str, lower )
if not is_set( str ) then
return "";
elseif inArray( key, { 'italic-title', 'trans-italic-title' } ) then
str = safeforitalics( str );
end
if lower == true then
return substitute( cfg.messages[key]:lower(), {str} );
else
return substitute( cfg.messages[key], {str} );
end
end
--[[
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.
]]
function argument_wrapper( args )
local origin = {};
return setmetatable({
ORIGIN = function( self, k )
local dummy = self[k]; --force the variable to be loaded.
return origin[k];
end
},
{
__index = function ( tbl, k )
if origin[k] ~= nil then
return nil;
end
local args, list, v = args, cfg.aliases[k];
if type( list ) == 'table' then
v, origin[k] = selectone( args, list, 'redundant_parameters' );
if origin[k] == nil then
origin[k] = ''; -- Empty string, not nil
end
elseif list ~= nil then
v, origin[k] = args[list], list;
else
-- maybe let through instead of raising an error?
-- v, origin[k] = args[k], k;
error( cfg.messages['unknown_argument_map'] );
end
-- Empty strings, not nil;
if v == nil then
v = cfg.defaults[k] or '';
origin[k] = '';
end
tbl = rawset( tbl, k, v );
return v;
end,
});
end
--[[
Looks for a parameter's name in the whitelist.
Parameters in the whitelist can have three values:
true - active, supported parameters
false - deprecated, supported parameters
nil - unsupported parameters
]]
function validate( name )
local name = tostring( name );
local state = whitelist.basic_arguments[ name ];
-- Normal arguments
if true == state then return true; end -- valid actively supported parameter
if false == state then
deprecated_parameter (); -- parameter is deprecated but still supported
return true;
end
-- Arguments with numbers in them
name = name:gsub( "%d+", "#" ); -- replace digit(s) with # (last25 becomes last#
state = whitelist.numbered_arguments[ name ];
if true == state then return true; end -- valid actively supported parameter
if false == state then
deprecated_parameter (); -- parameter is deprecated but still supported
return true;
end
return false; -- Not supported because not found or name is set to nil
end
-- Formats a comment for error trapping
function errorcomment( content, hidden )
return wrap( hidden and 'hidden-error' or 'visible-error', content );
end
--[[
Sets an error condition and returns the appropriate error message. The actual placement
of the error message in the output is the responsibility of the calling function.
]]
function seterror( error_id, arguments, raw, prefix, suffix )
local 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 inArray( 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 errorcomment( message, error_state.hidden );
end
-- Formats a wiki style external link
function externallinkid(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
-- Formats a wiki style internal link
function internallinkid(options)
return mw.ustring.format( '[[%s|%s]]%s[[%s%s%s|%s]]',
options.link, options.label, options.separator or " ",
options.prefix, options.id, options.suffix or "",
mw.text.nowiki(options.id)
);
end
-- Format an external link with error checking
function externallink( URL, label, source )
local error_str = "";
if not is_set( label ) then
label = URL;
if is_set( source ) then
error_str = seterror( 'bare_url_missing_title', { wrap( 'parameter', source ) }, false, " " );
else
error( cfg.messages["bare_url_no_origin"] );
end
end
if not checkurl( URL ) then
error_str = seterror( 'bad_url', {}, false, " " ) .. error_str;
end
return table.concat({ "[", URL, " ", safeforurl( label ), "]", error_str });
end
-- Formats a link to Amazon
function amazon(id, domain)
if checkisbn( id ) then -- see if asin value is a 10-digit isbn
table.insert( z.maintenance_cats, "CS1 maint: ASIN uses ISBN"); -- add to maint category so a bot or awb script can replace |asin= with |isbn=
end
if not is_set(domain) then
domain = "com";
elseif inArray (domain, {'jp', 'uk'}) then -- Japan, United Kingdom
domain = "co." .. domain;
elseif inArray (domain, {'au', 'br', 'mx'}) then -- Australia, Brazil, Mexico
domain = "com." .. domain;
end
local handler = cfg.id_handlers['ASIN'];
return externallinkid({link = handler.link,
label=handler.label , prefix="//www.amazon."..domain.."/dp/",id=id,
encode=handler.encode, separator = handler.separator})
end
--[[
format and error check arXiv identifier. There are two 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 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
]]
function arxiv (id)
local handler = cfg.id_handlers['ARXIV'];
local year, month, version;
local err_cat = ""
year, month, version = id:match("^%a[%a%.%-]+/([90]%d)([01]%d)%d%d%d([v%d]*)$"); -- test for the 9108-0703 format
if not year then -- arXiv id is not proper 9108-0703 form
year, month, version = id:match("^(%d%d)([01]%d)%.%d%d%d%d([v%d]*)$"); -- test for the 0704- format
if not year then
err_cat = ' ' .. seterror( 'bad_arxiv' ); -- arXiv id doesn't match either format
else -- id is the 0704- format
year = tonumber(year);
month = tonumber(month);
if ((7 > year) or (1 > month and 12 < month)) or -- is year invalid or is month invalid? (doesn't test for future years)
((7 == year) and (4 > month)) or -- when year is 07, is month invalid (before April)?
is_set (version) and nil == version:match("v%d+") then -- is version proper format of single 'v' followed by digits?
err_cat = ' ' .. seterror( 'bad_arxiv' ); -- set error message
end
end
else -- id is the 9108-0703 format; are the date values ok
year = tonumber(year);
month = tonumber(month);
if ((91 > year and 7 < year) or (1 > month and 12 < month)) or -- if invalid year or invalid month
((91 == year and 8 > month) or (7 == year and 3 < month)) or -- if years ok, are starting and ending months ok?
is_set (version) and nil == version:match("v%d+") then -- is version proper format of single 'v' followed by digits?
err_cat = ' ' .. seterror( 'bad_arxiv' ); -- set error message
end
end
return externallinkid({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
--[[
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.
]]
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
]]
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 = ' ' .. seterror( 'bad_lccn' ); -- set an error message
end
elseif 9 == len then -- LCCN should be adddddddd
if nil == id:match("%l%d%d%d%d%d%d%d%d") then -- does it match our pattern?
err_cat = ' ' .. seterror( 'bad_lccn' ); -- set an error message
end
elseif 10 == len then -- LCCN should be aadddddddd or dddddddddd
if id:match("[^%d]") then -- if LCCN has anything but digits (nil if only digits) ...
if nil == id:match("^%l%l%d%d%d%d%d%d%d%d") then -- ... see if it matches our pattern
err_cat = ' ' .. seterror( 'bad_lccn' ); -- no match, set an error message
end
end
elseif 11 == len then -- LCCN should be aaadddddddd or adddddddddd
if not (id:match("^%l%l%l%d%d%d%d%d%d%d%d") or id:match("^%l%d%d%d%d%d%d%d%d%d%d")) then -- see if it matches one of our patterns
err_cat = ' ' .. seterror( 'bad_lccn' ); -- no match, set an error message
end
elseif 12 == len then -- LCCN should be aadddddddddd
if not id:match("^%l%l%d%d%d%d%d%d%d%d%d%d") then -- see if it matches our pattern
err_cat = ' ' .. seterror( 'bad_lccn' ); -- no match, set an error message
end
else
err_cat = ' ' .. seterror( 'bad_lccn' ); -- wrong length, set an error message
end
if not is_set (err_cat) and nil ~= lccn:find ('%s') then
err_cat = ' ' .. seterror( 'bad_lccn' ); -- lccn contains a space, set an error message
end
return externallinkid({link = handler.link, label = handler.label,
prefix=handler.prefix,id=lccn,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
--[[
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.
]]
function pmid(id)
local test_limit = 30000000; -- update this value as PMIDs approach
local handler = cfg.id_handlers['PMID'];
local err_cat = ''; -- presume that PMID is valid
if id:match("[^%d]") then -- if PMID has anything but digits
err_cat = ' ' .. seterror( 'bad_pmid' ); -- set an error message
else -- PMID is only digits
local id_num = tonumber(id); -- convert id to a number for range testing
if 1 > id_num or test_limit < id_num then -- if PMID is outside test limit boundaries
err_cat = ' ' .. seterror( 'bad_pmid' ); -- set an error message
end
end
return externallinkid({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
end
--[[
Determines if a PMC identifier's online version is embargoed. Compares the date in |embargo= against today's date. If embargo date is
in the future, returns true; otherwise, returns false because the embargo has expired or |embargo= not set in this cite.
]]
function is_embargoed(embargo)
if is_set(embargo) then
local lang = mw.getContentLanguage();
local good1, embargo_date, good2, todays_date;
good1, embargo_date = pcall( lang.formatDate, lang, 'U', embargo );
good2, todays_date = pcall( lang.formatDate, lang, 'U' );
if good1 and good2 and tonumber( embargo_date ) >= tonumber( todays_date ) then --is embargo date is in the future?
return true; -- still embargoed
end
end
return false; -- embargo expired or |embargo= not set
end
--[[
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 specifies a date 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.
PMCs are sequential numbers beginning at 1 and counting up. This code checks the PMC to see that it contains only digits and is less
than test_limit; the value in local variable test_limit will need to be updated periodically as more PMCs are issued.
]]
function pmc(id, embargo)
local test_limit = 5000000; -- update this value as PMCs approach
local handler = cfg.id_handlers['PMC'];
local err_cat = ''; -- presume that PMC is valid
local text;
if id:match("[^%d]") then -- if PMC has anything but digits
err_cat = ' ' .. seterror( '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 = ' ' .. seterror( 'bad_pmc' ); -- set an error message
end
end
if is_embargoed(embargo) then
text="[[" .. handler.link .. "|" .. handler.label .. "]]:" .. handler.separator .. id .. err_cat; --still embargoed so no external link
else
text = externallinkid({link = handler.link, label = handler.label, --no embargo date, ok to link to article
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode}) .. err_cat;
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.
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 = externallinkid({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
inactive = ""
end
if nil == id:match("^10%.[^%s–]-/[^%s–]-[^%.,]$") then -- doi must begin with '10.', must contain a fwd slash, must not contain spaces or endashes, and must not end with period or comma
cat = ' ' .. seterror( 'bad_doi' );
end
return text .. inactive .. cat
end
-- Formats an OpenLibrary link, and checks for associated errors.
function openlibrary(id)
local code = id:match("^%d+([AMW])$"); -- only digits followed by 'A', 'M', or 'W'
local handler = cfg.id_handlers['OL'];
if ( code == "A" ) then
return externallinkid({link=handler.link, label=handler.label,
prefix="http://openlibrary.org/authors/OL",id=id, separator=handler.separator,
encode = handler.encode})
elseif ( code == "M" ) then
return externallinkid({link=handler.link, label=handler.label,
prefix="http://openlibrary.org/books/OL",id=id, separator=handler.separator,
encode = handler.encode})
elseif ( code == "W" ) then
return externallinkid({link=handler.link, label=handler.label,
prefix= "http://openlibrary.org/works/OL",id=id, separator=handler.separator,
encode = handler.encode})
else
return externallinkid({link=handler.link, label=handler.label,
prefix= "http://openlibrary.org/OL",id=id, separator=handler.separator,
encode = handler.encode}) ..
' ' .. seterror( 'bad_ol' );
end
end
--[[
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.
]]
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 = externallinkid({link = handler.link, label = handler.label,
prefix=handler.prefix,id=id,separator=handler.separator, encode=handler.encode})
if false == valid_issn then
text = text .. ' ' .. seterror( 'bad_issn' ) -- add an error message if the issn is invalid
end
return text
end
--[[
This function sets default title types (equivalent to the citation including |type=<default value>) for those citations that have defaults.
Also handles the special case where it is desirable to omit the title type from the rendered citation (|type=none).
]]
function set_titletype(cite_class, title_type)
if is_set(title_type) then
if "none" == title_type then
title_type = ""; -- if |type=none then type parameter not displayed
end
return title_type; -- if |type= has been set to any other value use that value
end
if "AV-media-notes" == cite_class or "DVD-notes" == cite_class then -- if this citation is cite AV media notes or cite DVD notes
return "Media notes"; -- display AV media notes / DVD media notes annotation
elseif "podcast" == cite_class then -- if this citation is cite podcast
return "Podcast"; -- display podcast annotation
elseif "pressrelease" == cite_class then -- if this citation is cite press release
return "Press release"; -- display press release annotation
elseif "techreport" == cite_class then -- if this citation is cite techreport
return "Technical report"; -- display techreport annotation
elseif "thesis" == cite_class then -- if this citation is cite thesis (degree option handled after this function returns)
return "Thesis"; -- display simple thesis annotation (without |degree= modification)
end
end
--[[
Determines whether a URL string is valid
At present the only check is whether the string appears to
be prefixed with a URI scheme. It is not determined whether
the URI scheme is valid or whether the URL is otherwise well
formed.
]]
function checkurl( url_str )
-- Protocol-relative or URL scheme
return url_str:sub(1,2) == "//" or url_str:match( "^[^/]*:" ) ~= nil;
end
-- Removes irrelevant text and dashes from ISBN number
-- Similar to that used for Special:BookSources
function cleanisbn( isbn_str )
return isbn_str:gsub( "[^-0-9X]", "" );
end
--[[--------------------------< 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.
]]
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.
]]
function strip_apostrophe_markup (argument)
local pattern, cap, cap2;
if not is_set (argument) then return argument; end
while true do -- look for and remove all 5-apostrophe (bold and italic) markup
if argument:match ("%'%'%'%'%'") then -- is there an instance of bold-italic?
if argument:match ("%'%'%'%'%'.*%'%'%'%'%'") then -- 5, 5
pattern, cap = argument:match ("(%'%'%'%'%'(.*)%'%'%'%'%')");
cap2 = ""; -- set to empty string so we do only one replacement at end
elseif argument:match ("%'%'%'%'%'.*%'%'%'.*%'%'") then -- bold italic followed by italic (5, 3, 2)
pattern, cap, cap2 = argument:match ("(%'%'%'%'%'(.*)%'%'%'(.*)%'%')");
elseif argument:match ("%'%'%'%'%'.*%'%'.*%'%'%'") then -- bold italic followed by bold (5, 2, 3)
pattern, cap, cap2 = argument:match ("(%'%'%'%'%'(.*)%'%'(.*)%'%'%')");
elseif argument:match ("%'%'%'.*%'%'.*%'%'%'%'%'") then -- bold italic followed by italic (3, 2, 5)
pattern, cap, cap2 = argument:match ("(%'%'%'(.*)%'%'(.*)%'%'%'%'%')");
elseif argument:match ("%'%'.*%'%'%'.*%'%'%'%'%'") then -- italic followed by bold (2, 3, 5)
pattern, cap, cap2 = argument:match ("(%'%'(.*)%'%'%'(.*)%'%'%'%'%')");
end
cap = escape_lua_magic_chars (cap); -- replace lua magic characters
cap2 = escape_lua_magic_chars (cap2); -- replace lua magic characters
pattern = escape_lua_magic_chars (pattern); -- replace lua magic characters
argument=argument:gsub(pattern, cap..cap2); -- remove the markup
else
break; -- none or no more 5-apostrophe matches
end
end
while true do -- look for and remove all 3-apostrophe (bold) markup
if argument:match ("%'%'%'.*%'%'%'") then -- is there an instance of bold?
pattern, cap = argument:match ("(%'%'%'(.*)%'%'%')")
cap = escape_lua_magic_chars (cap); -- replace lua magic characters
pattern = escape_lua_magic_chars (pattern); -- replace lua magic characters
argument=argument:gsub(pattern, cap); -- remove the markup
else
break; -- none or no more 3 matches
end
end
while true do -- look for and remove all 2-apostrophe (italic) markup
if argument:match ("%'%'.*%'%'") then -- is there an instance of italic?
pattern, cap = argument:match ("(%'%'(.*)%'%')")
cap = escape_lua_magic_chars (cap); -- replace lua magic characters
pattern = escape_lua_magic_chars (pattern); -- replace lua magic characters
argument=argument:gsub(pattern, cap); -- remove the markup
else
break; -- none or no more 2 matches
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...
]]
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.
]]
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
--[[
ISBN-10 and ISSN validator code calculates checksum across all isbn/issn digits including the check digit. ISBN-13 is checked in checkisbn().
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.
]]
function is_valid_isxn (isxn_str, len)
local temp = 0;
isxn_str = { isxn_str:byte(1, len) }; -- make a table of bytes
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
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
-- Determines whether an ISBN string is valid
function checkisbn( isbn_str )
if nil ~= isbn_str:match("[^%s-0-9X]") then return false; end -- fail if isbn_str contains anything but digits, hyphens, or the uppercase X
isbn_str = isbn_str:gsub( "-", "" ):gsub( " ", "" ); -- remove hyphens and spaces
local len = isbn_str:len();
if len ~= 10 and len ~= 13 then
return false;
end
if len == 10 then
if isbn_str:match( "^%d*X?$" ) == nil then return false; end
return is_valid_isxn(isbn_str, 10);
else
local temp = 0;
if isbn_str:match( "^97[89]%d*$" ) == nil then return false; end -- isbn13 begins with 978 or 979
isbn_str = { isbn_str:byte(1, len) };
for i, v in ipairs( isbn_str ) do
temp = temp + (3 - 2*(i % 2)) * tonumber( string.char(v) );
end
return temp % 10 == 0;
end
end
-- Gets the display text for a wikilink like [[A|B]] or [[B]] gives B
function removewikilink( str )
return (str:gsub( "%[%[([^%[%]]*)%]%]", function(l)
return l:gsub( "^[^|]*|(.*)$", "%1" ):gsub("^%s*(.-)%s*$", "%1");
end));
end
-- Escape sequences for content that will be used for URL descriptions
function safeforurl( str )
if str:match( "%[%[.-%]%]" ) ~= nil then
table.insert( z.message_tail, { seterror( 'wikilink_in_url', {}, true ) } );
end
return str:gsub( '[%[%]\n]', {
['['] = '[',
[']'] = ']',
['\n'] = ' ' } );
end
-- Converts a hyphen to a dash
function hyphentodash( str )
if not is_set(str) or str:match( "[%[%]{}<>]" ) ~= nil then
return str;
end
return str:gsub( '-', '–' );
end
-- Protects a string that will be wrapped in wiki italic markup '' ... ''
function safeforitalics( str )
--[[ Note: We cannot use <i> for italics, as the expected behavior for
italics specified by ''...'' in the title is that they will be inverted
(i.e. unitalicized) in the resulting references. In addition, <i> and ''
tend to interact poorly under Mediawiki's HTML tidy. ]]
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
--[[
Joins a sequence of strings together while checking for duplicate separation characters.
TODO: safejoin() has a flaw where it won't remove the duplicate character from a |title= / |url= combination.
This is because by the time we get here, |title=http://somesite.com and |title=Document Title. have been combined:
[http://somesite.com and ''Document Title.'']
so that now, the last character is not sepc but is ] (unless sepc == ']' which breaks the external link)
]]
function safejoin( tbl, duplicate_char )
--[[
Note: we use string functions here, rather than ustring functions.
This has considerably faster performance and should work correctly as
long as the duplicate_char is strict ASCII. The strings
in tbl may be ASCII or UTF8.
]]
local str = '';
local comp = '';
local end_chr = '';
local trim;
for _, value in ipairs( tbl ) do
if value == nil then value = ''; end
if str == '' then
str = value;
elseif value ~= '' then
if value:sub(1,1) == '<' then
-- Special case of values enclosed in spans and other markup.
comp = value:gsub( "%b<>", "" );
else
comp = value;
end
if comp:sub(1,1) == duplicate_char then
trim = false;
end_chr = str:sub(-1,-1);
-- str = str .. "<HERE(enchr=" .. end_chr.. ")"
if end_chr == duplicate_char then
str = str:sub(1,-2);
elseif end_chr == "'" then
if str:sub(-3,-1) == duplicate_char .. "''" then
str = str:sub(1, -4) .. "''";
elseif str:sub(-5,-1) == duplicate_char .. "]]''" then
trim = true;
elseif str:sub(-4,-1) == duplicate_char .. "]''" then
trim = true;
end
elseif end_chr == "]" then
if str:sub(-3,-1) == duplicate_char .. "]]" then
trim = true;
elseif str:sub(-2,-1) == duplicate_char .. "]" then
trim = true;
end
elseif end_chr == " " then
if str:sub(-2,-1) == duplicate_char .. " " then
str = str:sub(1,-3);
end
end
if trim then
if value ~= comp then
local dup2 = duplicate_char;
if dup2:match( "%A" ) then dup2 = "%" .. dup2; end
value = value:gsub( "(%b<>)" .. dup2, "%1", 1 )
else
value = value:sub( 2, -1 );
end
end
end
str = str .. value;
end
end
return str;
end
-- Attempts to convert names to initials.
function reducetoinitials(first)
local initials = {}
local i = 0; -- counter for number of initials
for word in string.gmatch(first, "%S+") do
table.insert(initials, string.sub(word,1,1)) -- Vancouver format does not include full stops.
i = i + 1; -- bump the counter
if 2 <= i then break; end -- only two initials allowed in Vancouver system; if 2, quit
end
return table.concat(initials) -- Vancouver format does not include spaces.
end
-- Formats a list of people (e.g. authors / editors)
function listpeople(control, people)
local sep = control.sep;
local namesep = control.namesep
local format = control.format
local maximum = control.maximum
local lastauthoramp = control.lastauthoramp;
local text = {}
local etal = false;
if sep:sub(-1,-1) ~= " " then sep = sep .. " " end
if maximum ~= nil and maximum < 1 then return "", 0; end
for i,person in ipairs(people) do
if is_set(person.last) then
local mask = person.mask
local one
local sep_one = sep;
if maximum ~= nil and i > maximum then
etal = true;
break;
elseif (mask ~= nil) then
local n = tonumber(mask)
if (n ~= nil) then
one = string.rep("—",n)
else
one = mask;
sep_one = " ";
end
else
one = person.last
local first = person.first
if is_set(first) then
if ( "vanc" == format ) then first = reducetoinitials(first) end
one = one .. namesep .. first
end
if is_set(person.link) and person.link ~= control.page_name then
one = "[[" .. person.link .. "|" .. one .. "]]" -- link author/editor if this page is not the author's/editor's page
end
if is_set(person.link) and nil ~= person.link:find("//") then one = one .. " " .. seterror( 'bad_authorlink' ) end -- check for url in author link;
end
table.insert( text, one )
table.insert( text, sep_one )
end
end
local count = #text / 2;
if count > 0 then
if count > 1 and is_set(lastauthoramp) and not etal then
text[#text-2] = " & ";
end
text[#text] = nil;
end
local result = table.concat(text) -- construct list
if etal then
local etal_text = cfg.messages['et al'];
result = result .. " " .. etal_text;
end
-- if necessary wrap result in <span> tag to format in Small Caps
if ( "scap" == format ) then result =
'<span class="smallcaps" style="font-variant:small-caps">' .. result .. '</span>';
end
return result, count
end
-- Generates a CITEREF anchor ID.
function anchorid( options )
return "CITEREF" .. table.concat( options );
end
--[[
Gets name list from the input arguments
Searches through args in sequential order to find |lastn= and |firstn= parameters (or their aliases), and their matching link and mask parameters.
Stops searching when both |lastn= and |firstn= are not found in args after two sequential attempts: found |last1=, |last2=, and |last3= but doesn't
find |last4= and |last5= then the search is done.
This function emits an error message when there is a |firstn= without a matching |lastn=. When there are 'holes' in the list of last names, |last1= and |last3=
are present but |last2= is missing, an error message is emitted. |lastn= is not required to have a matching |firstn=.
]]
function extractnames(args, list_name)
local names = {}; -- table of names
local last; -- individual name components
local first;
local link;
local mask;
local i = 1; -- loop counter/indexer
local n = 1; -- output table indexer
local count = 0; -- used to count the number of times we haven't found a |last= (or alias for authors, |editor-last or alias for editors)
local err_msg_list_name = list_name:match ("(%w+)List") .. 's list'; -- modify AuthorList or EditorList for use in error messages if necessary
while true do
last = selectone( args, cfg.aliases[list_name .. '-Last'], 'redundant_parameters', i ); -- search through args for name components beginning at 1
first = selectone( args, cfg.aliases[list_name .. '-First'], 'redundant_parameters', i );
link = selectone( args, cfg.aliases[list_name .. '-Link'], 'redundant_parameters', i );
mask = selectone( args, cfg.aliases[list_name .. '-Mask'], 'redundant_parameters', i );
if first and not last then -- if there is a firstn without a matching lastn
table.insert( z.message_tail, { seterror( 'first_missing_last', {err_msg_list_name, i}, true ) } ); -- add this error message
elseif not first and not last then -- if both firstn and lastn aren't found, are we done?
count = count + 1; -- number of times we haven't found last and first
if 2 == count then -- two missing names and we give up
break; -- normal exit or there is a two-name hole in the list; can't tell which
end
else -- we have last with or without a first
names[n] = {last = last, first = first, link = link, mask = mask}; -- add this name to our names list
n = n + 1; -- point to next location in the names table
if 1 == count then -- if the previous name was missing
table.insert( z.message_tail, { seterror( 'missing_name', {err_msg_list_name, i-1}, true ) } ); -- add this error message
end
count = 0; -- reset the counter, we're looking for two consecutive missing names
end
i = i + 1; -- point to next args location
end
return names; -- all done, return our list of names
end
-- Populates ID table from arguments using configuration settings
function extractids( args )
local id_list = {};
for k, v in pairs( cfg.id_handlers ) do
v = selectone( args, v.parameters, 'redundant_parameters' );
if is_set(v) then id_list[k] = v; end
end
return id_list;
end
-- Takes a table of IDs and turns it into a table of formatted ID outputs.
function buildidlist( 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
-- fallback to read-only cfg
handler = setmetatable( { ['id'] = v }, fallback(k) );
if handler.mode == 'external' then
table.insert( new_list, {handler.label, externallinkid( handler ) } );
elseif handler.mode == 'internal' then
table.insert( new_list, {handler.label, internallinkid( 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 ) } );
elseif k == 'ASIN' then
table.insert( new_list, {handler.label, amazon( v, options.ASINTLD ) } );
elseif k == 'LCCN' then
table.insert( new_list, {handler.label, lccn( v ) } );
elseif k == 'OL' then
table.insert( new_list, {handler.label, openlibrary( v ) } );
elseif k == 'PMC' then
table.insert( new_list, {handler.label, pmc( v, options.Embargo ) } );
elseif k == 'PMID' then
table.insert( new_list, {handler.label, pmid( v ) } );
elseif k == 'ISSN' then
table.insert( new_list, {handler.label, issn( v ) } );
elseif k == 'ISBN' then
local ISBN = internallinkid( handler );
if not checkisbn( v ) and not is_set(options.IgnoreISBN) then
ISBN = ISBN .. seterror( 'bad_isbn', {}, false, " ", "" );
end
table.insert( new_list, {handler.label, ISBN } );
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
-- Chooses one matching parameter from a list of parameters to consider
-- Generates an error if more than one match is present.
function selectone( args, possible, error_condition, index )
local value = nil;
local selected = '';
local error_list = {};
if index ~= nil then index = tostring(index); end
-- Handle special case of "#" replaced by empty string
if index == '1' then
for _, v in ipairs( possible ) do
v = v:gsub( "#", "" );
if is_set(args[v]) then
if value ~= nil and selected ~= v then
table.insert( error_list, v );
else
value = args[v];
selected = v;
end
end
end
end
for _, v in ipairs( possible ) do
if index ~= nil then
v = v:gsub( "#", index );
end
if is_set(args[v]) then
if value ~= nil and selected ~= v then
table.insert( error_list, v );
else
value = args[v];
selected = v;
end
end
end
if #error_list > 0 then
local error_str = "";
for _, k in ipairs( error_list ) do
if error_str ~= "" then error_str = error_str .. cfg.messages['parameter-separator'] end
error_str = error_str .. wrap( '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( 'parameter', selected );
table.insert( z.message_tail, { seterror( error_condition, {error_str}, true ) } );
end
return value, selected;
end
-- COinS metadata (see <http://ocoins.info/>) allows automated tools to parse
-- the citation information.
function COinS(data)
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( removewikilink( value ) ) } );
end
end
});
if is_set(data.Chapter) then
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:book";
OCinSoutput["rft.genre"] = "bookitem";
OCinSoutput["rft.btitle"] = data.Chapter;
OCinSoutput["rft.atitle"] = data.Title;
elseif is_set(data.Periodical) then
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:journal";
OCinSoutput["rft.genre"] = "article";
OCinSoutput["rft.jtitle"] = data.Periodical;
OCinSoutput["rft.atitle"] = data.Title;
else
OCinSoutput.rft_val_fmt = "info:ofi/fmt:kev:mtx:book";
OCinSoutput["rft.genre"] = "book"
OCinSoutput["rft.btitle"] = data.Title;
end
OCinSoutput["rft.place"] = data.PublicationPlace;
OCinSoutput["rft.date"] = data.Date;
OCinSoutput["rft.series"] = data.Series;
OCinSoutput["rft.volume"] = data.Volume;
OCinSoutput["rft.issue"] = data.Issue;
OCinSoutput["rft.pages"] = data.Pages;
OCinSoutput["rft.edition"] = data.Edition;
OCinSoutput["rft.pub"] = data.PublisherName;
for k, v in pairs( data.ID_list ) do
local id, value = cfg.id_handlers[k].COinS;
if k == 'ISBN' then value = cleanisbn( v ); else value = v; end
if string.sub( id or "", 1, 4 ) == 'info' then
OCinSoutput["rft_id"] = table.concat{ id, "/", v };
else
OCinSoutput[ id ] = value;
end
end
local last, first;
for k, v in ipairs( data.Authors ) do
last, first = v.last, v.first;
if k == 1 then
if is_set(last) then
OCinSoutput["rft.aulast"] = last;
end
if is_set(first) then
OCinSoutput["rft.aufirst"] = first;
end
end
if is_set(last) and is_set(first) then
OCinSoutput["rft.au"] = table.concat{ last, ", ", first };
elseif is_set(last) then
OCinSoutput["rft.au"] = last;
end
end
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.
]]
function get_iso639_code (lang)
if 'Norwegian' == lang then
return lang, 'no'; -- special case related to Wikimedia remap of code 'no' at Extension:CLDR
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() then -- ISO639-1 codes only
return name; -- so return the name but not the code
end
return name, code; -- found it, return name to ensure proper capitalization and the ISO639-1 code
end
end
return lang; -- not valid language; return language in original case and nil for ISO639-1 code
end
--[[--------------------------< L A N G U A G E _ P A R A M E T E R >------------------------------------------
Get language name from ISO639-1 code value provided. If a code is valid use the returned name; if not, then use the value that was provided with the language parameter.
There is an exception. There are three ISO639-1 codes for Norewegian language variants. There are two official variants: Norwegian Bokmål (code 'nb') and
Norwegian Nynorsk (code 'nn'). The third, code 'no', is defined by ISO639-1 as 'Norwegian' though in Norway this is pretty much meaningless. However, it appears
that on enwiki, editors are for the most part unaware of the nb and nn variants (compare page counts for these variants at Category:Articles with non-English-language external links.
Because Norwegian Bokmål is the most common language variant, Media wiki has been modified to return Norwegian Bokmål for ISO639-1 code 'no'. Here we undo that and
return 'Norwegian' when editors use |language=no. We presume that editors don't know about the variants or can't descriminate between them.
See Help talk:Citation Style_1#An ISO 639-1 language name test
When |language= contains a valid ISO639-1 code, the page is assigned to the category for that code: Category:Norwegian-language sources (no) if
the page is a mainspace page and the ISO639-1 code is not 'en'. Similarly, if the parameter is |language=Norwegian, it will be categorized in the same way.
TODO: Error message when language or language code is not ISO639-1?
]]
function language_parameter (lang, namespace)
local code; -- the ISO639-1 two character code
local name; -- the language name
local test='';
if 0 == namespace and (('en' == lang:lower()) or ('english' == lang:lower())) then
table.insert (z.maintenance_cats, 'CS1 maint: English language specified'); -- add maintenance category if |language=English or |language=en in article space
end
if 2 == lang:len() then -- ISO639-1 language code are 2 characters (fetchLanguageName also supports 3 character codes)
name = mw.language.fetchLanguageName( lang:lower(), "en" ); -- get ISO 639-1 language name if Language is a proper code
end
if is_set (name) then -- if Language specified a valid ISO639-1 code
code = lang:lower(); -- save it
else
name, code = get_iso639_code (lang); -- attempt to get code from name (assign name here so that we are sure of proper capitalization)
end
if is_set (code) then
if 'no' == code then name = 'Norwegian' end; -- override wikimedia when code is 'no'
if 0 == namespace and 'en' ~= code then -- is this page main / article space and English not the language?
table.insert( z.properties_cats, 'CS1 ' .. name .. '-language sources (' .. code .. ')'); -- in main space and not English: categorize
end
end
return (" " .. wrap( 'language', name)); -- wrap with '(in ...)'
end
--[[--------------------------< C I T A T I O N 0 >------------------------------------------------------------
This is the main function doing the majority of the citation
formatting.
]]
function citation0( config, args)
--[[
Load Input Parameters
The argument_wrapper facilitates the mapping of multiple
aliases to single internal variable.
]]
local A = argument_wrapper( args );
local i
local PPrefix = A['PPrefix']
local PPPrefix = A['PPPrefix']
if is_set( A['NoPP'] ) then PPPrefix = "" PPrefix = "" end
-- Pick out the relevant fields from the arguments. Different citation templates
-- define different field names for the same underlying things.
local Authors = A['Authors'];
local a = extractnames( args, 'AuthorList' );
local Coauthors = A['Coauthors'];
local Others = A['Others'];
local Editors = A['Editors'];
local e = extractnames( args, 'EditorList' );
local Year = A['Year'];
local PublicationDate = A['PublicationDate'];
local OrigYear = A['OrigYear'];
local Date = A['Date'];
local LayDate = A['LayDate'];
------------------------------------------------- Get title data
local Title = A['Title'];
local ScriptTitle = A['ScriptTitle'];
local BookTitle = A['BookTitle'];
local Conference = A['Conference'];
local TransTitle = A['TransTitle'];
local TitleNote = A['TitleNote'];
local TitleLink = A['TitleLink'];
local Chapter = A['Chapter'];
local ChapterLink = A['ChapterLink'];
local TransChapter = A['TransChapter'];
local TitleType = A['TitleType'];
local Degree = A['Degree'];
local Docket = A['Docket'];
local ArchiveURL = A['ArchiveURL'];
local URL = A['URL']
local URLorigin = A:ORIGIN('URL');
local ChapterURL = A['ChapterURL'];
local ChapterURLorigin = A:ORIGIN('ChapterURL');
local ConferenceURL = A['ConferenceURL'];
local ConferenceURLorigin = A:ORIGIN('ConferenceURL');
local Periodical = A['Periodical'];
local Series = A['Series'];
local Volume = A['Volume'];
local Issue = A['Issue'];
local Position = '';
local Page = A['Page'];
local Pages = hyphentodash( A['Pages'] );
local At = A['At'];
local Edition = A['Edition'];
local PublicationPlace = A['PublicationPlace']
local Place = A['Place'];
local PublisherName = A['PublisherName'];
local RegistrationRequired = A['RegistrationRequired'];
local SubscriptionRequired = A['SubscriptionRequired'];
local Via = A['Via'];
local AccessDate = A['AccessDate'];
local ArchiveDate = A['ArchiveDate'];
local Agency = A['Agency'];
local DeadURL = A['DeadURL']
local Language = A['Language'];
local Format = A['Format'];
local Ref = A['Ref'];
local DoiBroken = A['DoiBroken'];
local ID = A['ID'];
local ASINTLD = A['ASINTLD'];
local IgnoreISBN = A['IgnoreISBN'];
local Embargo = A['Embargo'];
local ID_list = extractids( args );
--[[ Hide unfinished cite newsgroup code so that long delayed update can take place
-- special case for cite newsgroup which uses |id= for a usenet article or post id
-- |id= is not included in COinS so here we convert it to an ID that will be included in COinS
if ('newsgroup' == config.CitationClass) and (is_set (ID)) then
ID_list['USENETID']=ID; -- add this new 'id' to the list of IDs
ID = ''; -- and unset
end
]]
local Quote = A['Quote'];
local PostScript = A['PostScript'];
local LayURL = A['LayURL'];
local LaySource = A['LaySource'];
local Transcript = A['Transcript'];
local TranscriptURL = A['TranscriptURL']
local TranscriptURLorigin = A:ORIGIN('TranscriptURL');
local sepc = A['Separator'];
local LastAuthorAmp = A['LastAuthorAmp'];
local no_tracking_cats = A['NoTracking'];
--these are used by cite interview
local Callsign = A['Callsign'];
local City = A['City'];
local Cointerviewers = A['Cointerviewers']; -- deprecated
local Interviewer = A['Interviewer']; -- deprecated
local Program = A['Program'];
--local variables that are not cs1 parameters
local page_type; -- is this needed? Doesn't appear to be used anywhere;
local use_lowercase = ( sepc ~= '.' );
local this_page = mw.title.getCurrentTitle(); --Also used for COinS and for language
local anchor_year; -- used in the CITEREF identifier
local COinS_date; -- used in the COinS metadata
-- Set postscript default.
if not is_set (PostScript) then -- if |postscript= has not been set (Postscript is nil which is the default for {{citation}}) and
if (config.CitationClass ~= "citation") then -- this template is not a citation template
PostScript = '.'; -- must be a cite xxx template so set postscript to default (period)
end
else
if PostScript:lower() == 'none' then -- if |postscript=none then
PostScript = ''; -- no postscript
end
end
--check this page to see if it is in one of the namespaces that cs1 is not supposed to add to the error categories.
if not is_set(no_tracking_cats) then -- ignore if we are already not going to categorize this page
if inArray (this_page.nsText, cfg.uncategorized_namespaces) then
no_tracking_cats = "true"; -- set no_tracking_cats
end
end
-- check for extra |page=, |pages= or |at= parameters.
if is_set(Page) then
if is_set(Pages) or is_set(At) then
Page = Page .. " " .. seterror('extra_pages'); -- add error message
Pages = ''; -- unset the others
At = '';
end
elseif is_set(Pages) then
if is_set(At) then
Pages = Pages .. " " .. seterror('extra_pages'); -- add error messages
At = ''; -- unset
end
end
-- both |publication-place= and |place= (|location=) allowed if different
if not is_set(PublicationPlace) and is_set(Place) then
PublicationPlace = Place; -- promote |place= (|location=) to |publication-place
end
if PublicationPlace == Place then Place = ''; end -- don't need both if they are the same
--[[
Parameter remapping for cite encyclopedia:
When the citation has these parameters:
|encyclopedia and |title then map |title to |article and |encyclopedia to |title
|encyclopedia and |article then map |encyclopedia to |title
|encyclopedia then map |encyclopedia to |title
|trans_title maps to |trans_chapter when |title is re-mapped
All other combinations of |encyclopedia, |title, and |article are not modified
]]
if ( config.CitationClass == "encyclopaedia" ) then
if is_set(Periodical) then -- Periodical is set when |encyclopedia is set
if is_set(Title) then
if not is_set(Chapter) then
Chapter = Title; -- |encyclopedia and |title are set so map |title to |article and |encyclopedia to |title
TransChapter = TransTitle;
Title = Periodical;
Periodical = ''; -- redundant so unset
TransTitle = ''; -- redundant so unset
end
else -- |title not set
Title = Periodical; -- |encyclopedia set and |article set or not set so map |encyclopedia to |title
Periodical = ''; -- redundant so unset
end
end
end
--special cases for citation.
if (config.CitationClass == "citation") then -- for citation templates
if not is_set (Ref) then -- if |ref= is not set
Ref = "harv"; -- set default |ref=harv
end
if not is_set (sepc) then -- if |separator= is not set
sepc = ','; -- set citation separator to its default (comma)
end
else -- not a citation template
if not is_set (sepc) then -- if |separator= has not been set
sepc = '.'; -- set cite xxx separator to its default (period)
end
end
-- check for special case where |separator=none
if 'none' == sepc:lower() then -- if |separator=none
sepc = ''; -- then set it to an empty string
end
-- Special case for cite techreport.
if (config.CitationClass == "techreport") then -- special case for cite techreport
if is_set(Issue) then -- cite techreport uses 'number', which other citations aliase to 'issue'
if not is_set(ID) then -- can we use ID for the "number"?
ID = Issue; -- yes, use it
Issue = ""; -- unset Issue so that "number" isn't duplicated in the rendered citation or COinS metadata
else -- can't use ID so emit error message
ID = ID .. " " .. seterror('redundant_parameters', '<code>|id=</code> and <code>|number=</code>');
end
end
end
-- special case for cite interview
if (config.CitationClass == "interview") then
if is_set(Program) then
ID = ' ' .. Program;
end
if is_set(Callsign) then
if is_set(ID) then
ID = ID .. sepc .. ' ' .. Callsign;
else
ID = ' ' .. Callsign;
end
end
if is_set(City) then
if is_set(ID) then
ID = ID .. sepc .. ' ' .. City;
else
ID = ' ' .. City;
end
end
if is_set(Interviewer) then
if is_set(TitleType) then
Others = ' ' .. TitleType .. ' with ' .. Interviewer;
TitleType = '';
else
Others = ' ' .. 'Interview with ' .. Interviewer;
end
if is_set(Cointerviewers) then
Others = Others .. sepc .. ' ' .. Cointerviewers;
end
else
Others = '(Interview)';
end
end
--Account for the oddity that is {{cite journal}} with |pmc= set and |url= not set
if config.CitationClass == "journal" and not is_set(URL) and is_set(ID_list['PMC']) then
if not is_embargoed(Embargo) then
URL=cfg.id_handlers['PMC'].prefix .. ID_list['PMC']; -- set url to be the same as the PMC external link if not embargoed
URLorigin = cfg.id_handlers['PMC'].parameters[1]; -- set URLorigin to parameter name for use in error message if citation is missing a |title=
end
end
-- Account for the oddity that is {{cite conference}}, before generation of COinS data.
--TODO: if this is only for {{cite conference}}, shouldn't we be checking? (if config.CitationClass=='conference' then ...)
if is_set(BookTitle) then
Chapter = Title;
ChapterLink = TitleLink;
TransChapter = TransTitle;
Title = BookTitle;
TitleLink = '';
TransTitle = '';
end
-- Account for the oddity that is {{cite episode}}, before generation of COinS data.
--[[ -- {{cite episode}} is not currently supported by this module
if config.CitationClass == "episode" then
local AirDate = A['AirDate'];
local SeriesLink = A['SeriesLink'];
local Season = A['Season'];
local SeriesNumber = A['SeriesNumber'];
local Network = A['Network'];
local Station = A['Station'];
local s, n = {}, {};
local Sep = (first_set(A["SeriesSeparator"], A["Separator"]) or "") .. " ";
if is_set(Issue) then table.insert(s, cfg.messages["episode"] .. " " .. Issue); Issue = ''; end
if is_set(Season) then table.insert(s, cfg.messages["season"] .. " " .. Season); end
if is_set(SeriesNumber) then table.insert(s, cfg.messages["series"] .. " " .. SeriesNumber); end
if is_set(Network) then table.insert(n, Network); end
if is_set(Station) then table.insert(n, Station); end
Date = Date or AirDate;
Chapter = Title;
ChapterLink = TitleLink;
TransChapter = TransTitle;
Title = Series;
TitleLink = SeriesLink;
TransTitle = '';
Series = table.concat(s, Sep);
ID = table.concat(n, Sep);
end
-- end of {{cite episode}} stuff]]
-- legacy: promote concatenation of |day=, |month=, and |year= to Date if Date not set; or, promote PublicationDate to Date if neither Date nor Year are set.
if not is_set(Date) then
Date = Year; -- promote Year to Date
Year = nil; -- make nil so Year as empty string isn't used for CITEREF
if is_set(Date) then
local Month = A['Month'];
if is_set(Month) then
Date = Month .. " " .. Date;
local Day = A['Day']
if is_set(Day) then Date = Day .. " " .. Date end
end
elseif is_set(PublicationDate) then -- use PublicationDate when |date= and |year= are not set
Date = PublicationDate; -- promote PublicationDate to Date
PublicationDate = ''; -- unset, no longer needed
end
end
if PublicationDate == Date then PublicationDate = ''; end -- if PublicationDate is same as Date, don't display in rendered citation
--[[
Go test all of the date-holding parameters for valid MOS:DATE format and make sure that dates are real dates. This must be done before we do COinS because here is where
we get the date used in the metadata.
Date validation supporting code is in Module:Citation/CS1/Date_validation
]]
anchor_year, COinS_date, error_message = dates({['accessdate']=AccessDate, ['airdate']=AirDate, ['archivedate']=ArchiveDate, ['date']=Date, ['doi_brokendate']=DoiBroken,
['embargo']=Embargo, ['laydate']=LayDate, ['publicationdate']=PublicationDate, ['year']=Year});
if is_set(error_message) then
table.insert( z.message_tail, { seterror( 'bad_date', {error_message}, true ) } ); -- add this error message
end
-- At this point fields may be nil if they weren't specified in the template use. We can use that fact.
-- COinS metadata (see <http://ocoins.info/>) for
-- automated parsing of citation information.
-- this is the function call to COinS()
local OCinSoutput = COinS{
['Periodical'] = Periodical,
['Chapter'] = strip_apostrophe_markup (Chapter), -- Chapter stripped of bold / italic wikimarkup
['Title'] = make_coins_title (Title, ScriptTitle), -- Title and ScriptTitle stripped of bold / italic wikimarkup
['PublicationPlace'] = PublicationPlace,
['Date'] = first_set(COinS_date, Date), -- COinS_date has correctly formatted date if Date is valid; any reason to keep Date here? Should we be including invalid dates in metadata?
['Series'] = Series,
['Volume'] = Volume,
['Issue'] = Issue,
['Pages'] = get_coins_pages (first_set(Page, Pages, At)), -- pages stripped of external links
['Edition'] = Edition,
['PublisherName'] = PublisherName,
['URL'] = first_set( URL, ChapterURL ),
['Authors'] = a,
['ID_list'] = ID_list,
['RawPage'] = this_page.prefixedText,
};
--[[Why is this here? Why are we mapping Title to Chapter when Periodical is set?
if is_set(Periodical) and not is_set(Chapter) and is_set(Title) then
Chapter = Title;
ChapterLink = TitleLink;
TransChapter = TransTitle;
Title = '';
TitleLink = '';
TransTitle = '';
end
]]
--[[ Hide unfinished cite newsgroup code so that long delayed update can take place
-- special case for cite newsgroup. Do this after COinS because we are modifying Publishername and ID
if 'newsgroup' == config.CitationClass then
if is_set (PublisherName) then
PublisherName = '[Newsgroup]: ' .. externallink( 'news:' .. PublisherName, PublisherName );
end
end
]]
-- Now perform various field substitutions.
-- We also add leading spaces and surrounding markup and punctuation to the
-- various parts of the citation, but only when they are non-nil.
if not is_set(Authors) then
local Maximum = tonumber( A['DisplayAuthors'] );
-- Preserve old-style implicit et al.
if not is_set(Maximum) and #a == 9 then
Maximum = 8;
table.insert( z.message_tail, { seterror('implict_etal_author', {}, true ) } );
elseif not is_set(Maximum) then
Maximum = #a + 1;
end
local control = {
sep = A["AuthorSeparator"] .. " ",
namesep = (first_set(A["AuthorNameSeparator"], A["NameSeparator"]) or "") .. " ",
format = A["AuthorFormat"],
maximum = Maximum,
lastauthoramp = LastAuthorAmp,
page_name = this_page.text -- get current page name so that we don't wikilink to it via authorlinkn
};
-- If the coauthor field is also used, prevent ampersand and et al. formatting.
if is_set(Coauthors) then
control.lastauthoramp = nil;
control.maximum = #a + 1;
end
Authors = listpeople(control, a)
end
if not is_set(Authors) and is_set(Coauthors) then -- coauthors aren't displayed if one of authors=, authorn=, or lastn= isn't specified
table.insert( z.message_tail, { seterror('coauthors_missing_author', {}, true) } ); -- emit error message
end
local EditorCount
if not is_set(Editors) then
local Maximum = tonumber( A['DisplayEditors'] );
-- Preserve old-style implicit et al.
if not is_set(Maximum) and #e == 4 then
Maximum = 3;
table.insert( z.message_tail, { seterror('implict_etal_editor', {}, true) } );
elseif not is_set(Maximum) then
Maximum = #e + 1;
end
local control = {
sep = A["EditorSeparator"] .. " ",
namesep = (first_set(A["EditorNameSeparator"], A["NameSeparator"]) or "") .. " ",
format = A['EditorFormat'],
maximum = Maximum,
lastauthoramp = LastAuthorAmp,
page_name = this_page.text -- get current page name so that we don't wikilink to it via authorlinkn
};
Editors, EditorCount = listpeople(control, e);
else
EditorCount = 1;
end
local Cartography = "";
local Scale = "";
if config.CitationClass == "map" then
if not is_set( Authors ) and is_set( PublisherName ) then
Authors = PublisherName;
PublisherName = "";
end
Cartography = A['Cartography'];
if is_set( Cartography ) then
Cartography = sepc .. " " .. wrap( 'cartography', Cartography, use_lowercase );
end
Scale = A['Scale'];
if is_set( Scale ) then
Scale = sepc .. " " .. Scale;
end
end
if not is_set(URL) and
not is_set(ChapterURL) and
not is_set(ArchiveURL) and
not is_set(ConferenceURL) and
not is_set(TranscriptURL) then
-- Test if cite web or cite podcast |url= is missing or empty
if inArray(config.CitationClass, {"web","podcast"}) then
table.insert( z.message_tail, { seterror( 'cite_web_url', {}, true ) } );
end
-- Test if accessdate is given without giving a URL
if is_set(AccessDate) then
table.insert( z.message_tail, { seterror( 'accessdate_missing_url', {}, true ) } );
AccessDate = '';
end
-- Test if format is given without giving a URL
if is_set(Format) then
Format = Format .. seterror( 'format_missing_url' );
end
end
-- Test if citation has no title
if not is_set(Title) and
not is_set(Periodical) and
not is_set(Conference) and
not is_set(TransTitle) and
not is_set(ScriptTitle) then
table.insert( z.message_tail, { seterror( 'citation_missing_title', {}, true ) } );
end
Format = is_set(Format) and " (" .. Format .. ")" or "";
local OriginalURL = URL
DeadURL = DeadURL:lower();
if is_set( ArchiveURL ) then
if ( DeadURL ~= "no" ) then
URL = ArchiveURL
URLorigin = A:ORIGIN('ArchiveURL')
end
end
-- Format chapter / article title
if is_set(Chapter) and is_set(ChapterLink) then
Chapter = "[[" .. ChapterLink .. "|" .. Chapter .. "]]";
end
Chapter = kern_quotes (Chapter); -- if necessary, separate chapter title's leading and trailing quote marks from Module provided quote marks
Chapter = wrap( 'quoted-title', Chapter );
TransChapter = wrap( 'trans-quoted-title', TransChapter );
local TransError = ""
if is_set(TransChapter) then
if not is_set(Chapter) then
TransError = " " .. seterror( 'trans_missing_chapter' ); -- add error message
else
TransChapter = " " .. TransChapter;
end
end
Chapter = Chapter .. TransChapter;
if is_set(Chapter) then
if not is_set(ChapterLink) then
if is_set(ChapterURL) then
Chapter = externallink( ChapterURL, Chapter ) .. TransError;
if not is_set(URL) then
Chapter = Chapter .. Format;
Format = "";
end
elseif is_set(URL) then
Chapter = externallink( URL, Chapter ) .. TransError .. Format;
URL = "";
Format = "";
else
Chapter = Chapter .. TransError;
end
elseif is_set(ChapterURL) then
Chapter = Chapter .. " " .. externallink( ChapterURL, nil, ChapterURLorigin ) ..
TransError;
else
Chapter = Chapter .. TransError;
end
Chapter = Chapter .. sepc .. " " -- with end-space
elseif is_set(ChapterURL) then
Chapter = " " .. externallink( ChapterURL, nil, ChapterURLorigin ) .. sepc .. " ";
end
-- Format main title.
if is_set(TitleLink) and is_set(Title) then
Title = "[[" .. TitleLink .. "|" .. Title .. "]]"
end
if inArray(config.CitationClass, {"web","news","journal","pressrelease","conference","podcast"}) then
Title = kern_quotes (Title); -- if necessary, separate title's leading and trailing quote marks from Module provided quote marks
Title = wrap( 'quoted-title', Title );
Title = script_concatenate (Title, ScriptTitle); -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
TransTitle = wrap( 'trans-quoted-title', TransTitle );
-- Chapter = ''; -- chapter not allowed - disabled because doing this here doesn't display error msg and promoted url is lost
else
Title = wrap( 'italic-title', Title );
Title = script_concatenate (Title, ScriptTitle); -- <bdi> tags, lang atribute, categorization, etc; must be done after title is wrapped
TransTitle = wrap( 'trans-italic-title', TransTitle );
end
TransError = "";
if is_set(TransTitle) then
if is_set(Title) then
TransTitle = " " .. TransTitle;
else
TransError = " " .. seterror( 'trans_missing_title' );
end
end
Title = Title .. TransTitle;
if is_set(Title) then
if not is_set(TitleLink) and is_set(URL) then
Title = externallink( URL, Title ) .. TransError .. Format
URL = "";
Format = "";
else
Title = Title .. TransError;
end
end
if is_set(Place) then
Place = " " .. wrap( 'written', Place, use_lowercase ) .. sepc .. " ";
end
if is_set(Conference) then
if is_set(ConferenceURL) then
Conference = externallink( ConferenceURL, Conference );
end
Conference = sepc .. " " .. Conference
elseif is_set(ConferenceURL) then
Conference = sepc .. " " .. externallink( ConferenceURL, nil, ConferenceURLorigin );
end
if not is_set(Position) then
local Minutes = A['Minutes'];
if is_set(Minutes) then
Position = " " .. Minutes .. " " .. cfg.messages['minutes'];
else
local Time = A['Time'];
if is_set(Time) then
local TimeCaption = A['TimeCaption']
if not is_set(TimeCaption) then
TimeCaption = cfg.messages['event'];
if sepc ~= '.' then
TimeCaption = TimeCaption:lower();
end
end
Position = " " .. TimeCaption .. " " .. Time;
end
end
else
Position = " " .. Position;
At = '';
end
if not is_set(Page) then
if is_set(Pages) then
if is_set(Periodical) and
not inArray(config.CitationClass, {"encyclopaedia","web","book","news","podcast"}) then
Pages = ": " .. Pages;
elseif tonumber(Pages) ~= nil then
Pages = sepc .." " .. PPrefix .. Pages;
else
Pages = sepc .." " .. PPPrefix .. Pages;
end
end
else
if is_set(Periodical) and
not inArray(config.CitationClass, {"encyclopaedia","web","book","news","podcast"}) then
Page = ": " .. Page;
else
Page = sepc .." " .. PPrefix .. Page;
end
end
At = is_set(At) and (sepc .. " " .. At) or "";
Position = is_set(Position) and (sepc .. " " .. Position) or "";
if config.CitationClass == 'map' then
local Section = A['Section'];
local Inset = A['Inset'];
if first_set( Pages, Page, At ) ~= nil or sepc ~= '.' then
if is_set( Section ) then
Section = ", " .. wrap( 'section', Section, true );
end
if is_set( Inset ) then
Inset = ", " .. wrap( 'inset', Inset, true );
end
else
if is_set( Section ) then
Section = sepc .. " " .. wrap( 'section', Section, use_lowercase );
if is_set( Inset ) then
Inset = ", " .. wrap( 'inset', Inset, true );
end
elseif is_set( Inset ) then
Inset = sepc .. " " .. wrap( 'inset', Inset, use_lowercase );
end
end
At = At .. Section .. Inset;
end
if is_set (Language) then
Language = language_parameter (Language, this_page.namespace); -- format, categories (article namespace only), name from ISO639-1, etc
else
Language=""; -- language not specified so make sure this is an empty string;
end
Others = is_set(Others) and (sepc .. " " .. Others) or "";
-- handle type parameter for those CS1 citations that have default values
if inArray(config.CitationClass, {"AV-media-notes", "DVD-notes", "podcast", "pressrelease", "techreport", "thesis"}) then
TitleType = set_titletype (config.CitationClass, TitleType);
if is_set(Degree) and "Thesis" == TitleType then -- special case for cite thesis
TitleType = Degree .. " thesis";
end
end
if is_set(TitleType) then -- if type parameter is specified
TitleType = " (" .. TitleType .. ")"; -- display it in parentheses
end
TitleNote = is_set(TitleNote) and (sepc .. " " .. TitleNote) or "";
Edition = is_set(Edition) and (" " .. wrap( 'edition', Edition )) or "";
Issue = is_set(Issue) and (" (" .. Issue .. ")") or "";
Series = is_set(Series) and (sepc .. " " .. Series) or "";
OrigYear = is_set(OrigYear) and (" [" .. OrigYear .. "]") or "";
Agency = is_set(Agency) and (sepc .. " " .. Agency) or "";
if is_set(Volume) then
if ( mw.ustring.len(Volume) > 4 )
then Volume = sepc .." " .. Volume;
else Volume = " <b>" .. hyphentodash(Volume) .. "</b>";
end
end
--[[ This code commented out while discussion continues until after week of 2014-03-23 live module update;
if is_set(Volume) then
if ( mw.ustring.len(Volume) > 4 )
then Volume = sepc .. " " .. Volume;
else
Volume = " <b>" .. hyphentodash(Volume) .. "</b>";
if is_set(Series) then Volume = sepc .. Volume;
end
end
end
]]
------------------------------------ totally unrelated data
--[[ Loosely mimic {{subscription required}} template; Via parameter identifies a delivery source that is not the publisher; these sources often, but not always, exist
behind a registration or paywall. So here, we've chosen to decouple via from subscription (via has never been part of the registration required template).
Subscription implies paywall; Registration does not. If both are used in a citation, the subscription required link note is displayed. There are no error messages for this condition.
]]
if is_set(Via) then
Via = " " .. wrap( 'via', Via );
end
if is_set(SubscriptionRequired) then
SubscriptionRequired = sepc .. " " .. cfg.messages['subscription']; --here when 'via' parameter not used but 'subscription' is
elseif is_set(RegistrationRequired) then
SubscriptionRequired = sepc .. " " .. cfg.messages['registration']; --here when 'via' and 'subscription' parameters not used but 'registration' is
end
if is_set(AccessDate) then
local retrv_text = " " .. cfg.messages['retrieved']
if (sepc ~= ".") then retrv_text = retrv_text:lower() end
AccessDate = '<span class="reference-accessdate">' .. sepc
.. substitute( retrv_text, {AccessDate} ) .. '</span>'
end
if is_set(ID) then ID = sepc .." ".. ID; end
if "thesis" == config.CitationClass and is_set(Docket) then
ID = sepc .." Docket ".. Docket .. ID;
end
ID_list = buildidlist( ID_list, {DoiBroken = DoiBroken, ASINTLD = ASINTLD, IgnoreISBN = IgnoreISBN, Embargo=Embargo} );
if is_set(URL) then
URL = " " .. externallink( URL, nil, URLorigin );
end
if is_set(Quote) then
if Quote:sub(1,1) == '"' and Quote:sub(-1,-1) == '"' then
Quote = Quote:sub(2,-2);
end
Quote = sepc .." " .. wrap( 'quoted-text', Quote );
PostScript = ""; -- CS1 does not supply terminal punctuation when |quote= is set
end
local Archived
if is_set(ArchiveURL) then
if not is_set(ArchiveDate) then
ArchiveDate = seterror('archive_missing_date');
end
if "no" == DeadURL then
local arch_text = cfg.messages['archived'];
if sepc ~= "." then arch_text = arch_text:lower() end
Archived = sepc .. " " .. substitute( cfg.messages['archived-not-dead'],
{ externallink( ArchiveURL, arch_text ), ArchiveDate } );
if not is_set(OriginalURL) then
Archived = Archived .. " " .. seterror('archive_missing_url');
end
elseif is_set(OriginalURL) then
local arch_text = cfg.messages['archived-dead'];
if sepc ~= "." then arch_text = arch_text:lower() end
Archived = sepc .. " " .. substitute( arch_text,
{ externallink( OriginalURL, cfg.messages['original'] ), ArchiveDate } );
else
local arch_text = cfg.messages['archived-missing'];
if sepc ~= "." then arch_text = arch_text:lower() end
Archived = sepc .. " " .. substitute( arch_text,
{ seterror('archive_missing_url'), ArchiveDate } );
end
else
Archived = ""
end
local Lay
if is_set(LayURL) then
if is_set(LayDate) then LayDate = " (" .. LayDate .. ")" end
if is_set(LaySource) then
LaySource = " – ''" .. safeforitalics(LaySource) .. "''";
else
LaySource = "";
end
if sepc == '.' then
Lay = sepc .. " " .. externallink( LayURL, cfg.messages['lay summary'] ) .. LaySource .. LayDate
else
Lay = sepc .. " " .. externallink( LayURL, cfg.messages['lay summary']:lower() ) .. LaySource .. LayDate
end
else
Lay = "";
end
if is_set(Transcript) then
if is_set(TranscriptURL) then Transcript = externallink( TranscriptURL, Transcript ); end
elseif is_set(TranscriptURL) then
Transcript = externallink( TranscriptURL, nil, TranscriptURLorigin );
end
local Publisher;
if is_set(Periodical) and
not inArray(config.CitationClass, {"encyclopaedia","web","pressrelease","podcast"}) then
if is_set(PublisherName) then
if is_set(PublicationPlace) then
Publisher = PublicationPlace .. ": " .. PublisherName;
else
Publisher = PublisherName;
end
elseif is_set(PublicationPlace) then
Publisher= PublicationPlace;
else
Publisher = "";
end
if is_set(PublicationDate) then
if is_set(Publisher) then
Publisher = Publisher .. ", " .. wrap( 'published', PublicationDate );
else
Publisher = PublicationDate;
end
end
if is_set(Publisher) then
Publisher = " (" .. Publisher .. ")";
end
else
if is_set(PublicationDate) then
PublicationDate = " (" .. wrap( 'published', PublicationDate ) .. ")";
end
if is_set(PublisherName) then
if is_set(PublicationPlace) then
Publisher = sepc .. " " .. PublicationPlace .. ": " .. PublisherName .. PublicationDate;
else
Publisher = sepc .. " " .. PublisherName .. PublicationDate;
end
elseif is_set(PublicationPlace) then
Publisher= sepc .. " " .. PublicationPlace .. PublicationDate;
else
Publisher = PublicationDate;
end
end
-- Several of the above rely upon detecting this as nil, so do it last.
if is_set(Periodical) then
if is_set(Title) or is_set(TitleNote) then
Periodical = sepc .. " " .. wrap( 'italic-title', Periodical )
else
Periodical = wrap( 'italic-title', Periodical )
end
end
--[[
Handle the oddity that is cite speech. This code overrides whatever may be the value assigned to TitleNote (through |department=) and forces it to be " (Speech)" so that
the annotation directly follows the |title= parameter value in the citation rather than the |event= parameter value (if provided).
]]
if "speech" == config.CitationClass then -- cite speech only
TitleNote = " (Speech)"; -- annotate the citation
if is_set (Periodical) then -- if Periodical, perhaps because of an included |website= or |journal= parameter
if is_set (Conference) then -- and if |event= is set
Conference = Conference .. sepc .. " "; -- then add appropriate punctuation to the end of the Conference variable before rendering
end
end
end
-- Piece all bits together at last. Here, all should be non-nil.
-- We build things this way because it is more efficient in LUA
-- not to keep reassigning to the same string variable over and over.
local tcommon
if inArray(config.CitationClass, {"journal","citation"}) and is_set(Periodical) then
if is_set(Others) then Others = Others .. sepc .. " " end
tcommon = safejoin( {Others, Title, TitleNote, Conference, Periodical, Format, TitleType, Scale, Series,
Language, Cartography, Edition, Publisher, Agency, Volume, Issue}, sepc );
else
tcommon = safejoin( {Title, TitleNote, Conference, Periodical, Format, TitleType, Scale, Series, Language,
Volume, Issue, Others, Cartography, Edition, Publisher, Agency}, sepc );
end
if #ID_list > 0 then
ID_list = safejoin( { sepc .. " ", table.concat( ID_list, sepc .. " " ), ID }, sepc );
else
ID_list = ID;
end
local idcommon = safejoin( { ID_list, URL, Archived, AccessDate, Via, SubscriptionRequired, Lay, Quote }, sepc );
local text;
local pgtext = Position .. Page .. Pages .. At;
if is_set(Authors) then
if is_set(Coauthors) then
Authors = Authors .. A['AuthorSeparator'] .. " " .. Coauthors
end
if is_set(Date) then
Date = " ("..Date..")" .. OrigYear .. sepc .. " "
elseif string.sub(Authors,-1,-1) == sepc then
Authors = Authors .. " "
else
Authors = Authors .. sepc .. " "
end
if is_set(Editors) then
local in_text = " ";
local post_text = "";
if is_set(Chapter) then
in_text = in_text .. cfg.messages['in'] .. " "
else
if EditorCount <= 1 then
post_text = ", " .. cfg.messages['editor'];
else
post_text = ", " .. cfg.messages['editors'];
end
end
if (sepc ~= '.') then in_text = in_text:lower() end
Editors = in_text .. Editors .. post_text;
if (string.sub(Editors,-1,-1) == sepc)
then Editors = Editors .. " "
else Editors = Editors .. sepc .. " "
end
end
text = safejoin( {Authors, Date, Chapter, Place, Editors, tcommon }, sepc );
text = safejoin( {text, pgtext, idcommon}, sepc );
elseif is_set(Editors) then
if is_set(Date) then
if EditorCount <= 1 then
Editors = Editors .. ", " .. cfg.messages['editor'];
else
Editors = Editors .. ", " .. cfg.messages['editors'];
end
Date = " (" .. Date ..")" .. OrigYear .. sepc .. " "
else
if EditorCount <= 1 then
Editors = Editors .. " (" .. cfg.messages['editor'] .. ")" .. sepc .. " "
else
Editors = Editors .. " (" .. cfg.messages['editors'] .. ")" .. sepc .. " "
end
end
text = safejoin( {Editors, Date, Chapter, Place, tcommon}, sepc );
text = safejoin( {text, pgtext, idcommon}, sepc );
else
if is_set(Date) then
if ( string.sub(tcommon,-1,-1) ~= sepc )
then Date = sepc .." " .. Date .. OrigYear
else Date = " " .. Date .. OrigYear
end
end
if config.CitationClass=="journal" and is_set(Periodical) then
text = safejoin( {Chapter, Place, tcommon}, sepc );
text = safejoin( {text, pgtext, Date, idcommon}, sepc );
else
text = safejoin( {Chapter, Place, tcommon, Date}, sepc );
text = safejoin( {text, pgtext, idcommon}, sepc );
end
end
if is_set(PostScript) and PostScript ~= sepc then
text = safejoin( {text, sepc}, sepc ); --Deals with italics, spaces, etc.
text = text:sub(1,-sepc:len()-1);
-- text = text:sub(1,-2); --Remove final separator (assumes that sepc is only one character)
end
text = safejoin( {text, PostScript}, sepc );
-- Now enclose the whole thing in a <span/> element
local options = {};
if is_set(config.CitationClass) and config.CitationClass ~= "citation" then
options.class = "citation " .. config.CitationClass;
else
options.class = "citation";
end
if is_set(Ref) and Ref:lower() ~= "none" then
local id = Ref
if ( "harv" == Ref ) then
local names = {} --table of last names & year
if #a > 0 then
for i,v in ipairs(a) do
names[i] = v.last
if i == 4 then break end
end
elseif #e > 0 then
for i,v in ipairs(e) do
names[i] = v.last
if i == 4 then break end
end
end
names[ #names + 1 ] = first_set(Year, anchor_year); -- Year first for legacy citations
id = anchorid(names)
end
options.id = id;
end
if string.len(text:gsub("<span[^>/]*>.-</span>", ""):gsub("%b<>","")) <= 2 then
z.error_categories = {};
text = seterror('empty_citation');
z.message_tail = {};
end
if is_set(options.id) then
text = '<span id="' .. mw.uri.anchorEncode(options.id) ..'" class="' .. mw.text.nowiki(options.class) .. '">' .. text .. "</span>";
else
text = '<span class="' .. mw.text.nowiki(options.class) .. '">' .. text .. "</span>";
end
local empty_span = '<span style="display:none;"> </span>';
-- Note: Using display: none on then COinS span breaks some clients.
local OCinS = '<span title="' .. OCinSoutput .. '" class="Z3988">' .. empty_span .. '</span>';
text = text .. OCinS;
if #z.message_tail ~= 0 then
text = text .. " ";
for i,v in ipairs( z.message_tail ) do
if is_set(v[1]) then
if i == #z.message_tail then
text = text .. errorcomment( v[1], v[2] );
else
text = text .. errorcomment( v[1] .. "; ", v[2] );
end
end
end
end
no_tracking_cats = no_tracking_cats:lower();
if inArray(no_tracking_cats, {"", "no", "false", "n"}) then
for _, v in ipairs( z.error_categories ) do
text = text .. '[[Category:' .. v ..']]';
end
for _, v in ipairs( z.maintenance_cats ) do -- append maintenance categories
text = text .. '[[Category:' .. v ..']]';
end
for _, v in ipairs( z.properties_cats ) do -- append maintenance categories
text = text .. '[[Category:' .. v ..']]';
end
end
return text
end
-- This is used by templates such as {{cite book}} to create the actual citation text.
function z.citation(frame)
local pframe = frame:getParent()
if nil ~= string.find( frame:getTitle(), 'sandbox', 1, true ) then -- did the {{#invoke:}} use sandbox version?
cfg = mw.loadData( 'Module:Citation/CS1/Configuration/sandbox' ); -- load sandbox versions of Configuration and Whitelist and ...
whitelist = mw.loadData( 'Module:Citation/CS1/Whitelist/sandbox' );
dates = require('Module:Citation/CS1/Date_validation/sandbox').dates -- ... sandbox version of date validation code
else -- otherwise
cfg = mw.loadData( 'Module:Citation/CS1/Configuration' ); -- load live versions of Configuration and Whitelist and ...
whitelist = mw.loadData( 'Module:Citation/CS1/Whitelist' );
dates = require('Module:Citation/CS1/Date_validation').dates -- ... live version of date validation code
end
local args = {};
local suggestions = {};
local error_text, error_state;
local config = {};
for k, v in pairs( frame.args ) do
config[k] = v;
args[k] = v;
end
for k, v in pairs( pframe.args ) do
if v ~= '' then
if not validate( k ) then
error_text = "";
if type( k ) ~= 'string' then
-- Exclude empty numbered parameters
if v:match("%S+") ~= nil then
error_text, error_state = seterror( 'text_ignored', {v}, true );
end
elseif validate( k:lower() ) then
error_text, error_state = seterror( 'parameter_ignored_suggest', {k, k:lower()}, true );
else
if #suggestions == 0 then
suggestions = mw.loadData( 'Module:Citation/CS1/Suggestions' );
end
if suggestions[ k:lower() ] ~= nil then
error_text, error_state = seterror( 'parameter_ignored_suggest', {k, suggestions[ k:lower() ]}, true );
else
error_text, error_state = seterror( 'parameter_ignored', {k}, true );
end
end
if error_text ~= '' then
table.insert( z.message_tail, {error_text, error_state} );
end
end
args[k] = v;
elseif args[k] ~= nil or (k == 'postscript') then
args[k] = v;
end
end
return citation0( config, args)
end
return z
