Jump to content

Module:Citation/CS1/COinS: Difference between revisions

33 revisions imported from templatewiki:Module:Citation/CS1/COinS
No edit summary
m (33 revisions imported from templatewiki:Module:Citation/CS1/COinS)
(31 intermediate revisions by 6 users not shown)
History of changes since last sync: 2015-12-12
local coins = {};
--[[--------------------------< F O R W A R D D E C L A R A T I O N S >--------------------------------------
local is_set, in_array, remove_wiki_link, strip_apostrophe_markup; -- functions in Module:Citation/CS1/Utilities
local cfg; -- table of configuration tables that are defined in Module:Citation/CS1/Configuration
--[[--------------------------< 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
if argument:find ( "''", 1, true ) == nil then -- Is there at least one double apostrophe? If not, exit.
return argument;
while true do
if argument:find ( "'''''", 1, true ) then -- bold italic (5)
argument=argument:gsub("%'%'%'%'%'", ""); -- remove all instances of it
elseif argument:find ( "''''", 1, true ) then -- italic start and end without content (4)
argument=argument:gsub("%'%'%'%'", "");
elseif argument:find ( "'''", 1, true ) then -- bold (3)
argument=argument:gsub("%'%'%'", "");
elseif argument:find ( "''", 1, true ) then -- italic (2)
argument=argument:gsub("%'%'", "");
return argument; -- done
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 correuptedcorrupted with strings
of %27%27...
return title .. script; -- return the concatenation
--[[--------------------------< 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("([%^%$%(%)%.%[%]%*%+%-%?])", "%%%1"); -- replace all other lua magic pattern characters
return argument;
pages = pages:gsub("&%w+;", "-" ); -- and replace html entities (&ndash; etc.) with hyphens; do we need to replace numerical entities like &#32; and the like?
return pages;
--[=[-------------------------< C O I N S _ R E P L A C E _ M A T H _ S T R I P M A R K E R >------------------
There are three options for math markup rendering that depend on the editor's math preference settings. These
settings are at [[Special:Preferences#mw-prefsection-rendering]] and are
PNG images
TeX source
MathML with SVG or PNG fallback
All three are heavy with html and css which doesn't belong in the metadata.
Without this function, the metadata saved in the raw wikitext contained the rendering determined by the settings
of the last editor to save the page.
This function gets the rendered form of an equation according to the editor's preference before the page is saved. It
then searches the rendering for the text equivalent of the rendered equation and replaces the rendering with that so
that the page is saved without extraneous html/css markup and with a reasonably readable text form of the equation.
When a replacement is made, this function returns true and the value with replacement; otherwise false and the intital
value. To replace multipe equations it is necesary to call this function from within a loop.
local function coins_replace_math_stripmarker (value)
local stripmarker = cfg.stripmarkers['math'];
local rendering = value:match (stripmarker); -- is there a math stripmarker
if not rendering then -- when value doesn't have a math stripmarker, abandon this test
return false, value;
rendering = mw.text.unstripNoWiki (rendering); -- convert stripmarker into rendered value (or nil? ''? when math render error)
if rendering:match ('alt="[^"]+"') then -- if PNG math option
rendering = rendering:match ('alt="([^"]+)"'); -- extract just the math text
elseif rendering:match ('$%s+.+%s+%$') then -- if TeX math option; $ is legit character that is escapes as \$
rendering = rendering:match ('$%s+(.+)%s+%$') -- extract just the math text
elseif rendering:match ('<annotation[^>]+>.+</annotation>') then -- if MathML math option
rendering = rendering:match ('<annotation[^>]+>(.+)</annotation>') -- extract just the math text
return false, value; -- had math stripmarker but not one of the three defined forms
return true, value:gsub (stripmarker, rendering, 1);
Cleanup parameter values for the metadata by removing or replacing invisible characters and certain html entities.
2015-12-10: there is a bug in mw.text.unstripNoWiki (). It replacedreplaces math stripmarkers with the appropriate content
when it shouldn't. See https://phabricator.wikimedia.org/T121085 and Wikipedia_talk:Lua#stripmarkers_and_mw.text.unstripNoWiki.28.29
TODO: move the replacement patterns and replacement values into a table in /Configuration similar to the invisible
characters table?
local function coins_cleanup (value)
local replaced = true; -- default state to get the do loop running
while replaced do -- loop until all math stripmarkers replaced
replaced, value = coins_replace_math_stripmarker (value); -- replace math stripmarker with text representation of the equation
value = value:gsub (cfg.stripmarkers['math'], "MATH RENDER ERROR"); -- one or more couldn't be replaced; insert vague error message
value = mw.text.unstripNoWiki (value); -- replace nowiki stripmarkers with their content
value = value:gsub ('<span class="nowrap" style="padding%-left:0%.1em;">&#39;(s?)</span>', "'s%1"); -- replace {{'s}} templateor {{'s}} with simple apostrophe or apostrophe-s
value = value:gsub ('&zwj;\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe
value = value:gsub ('\226\128\138\039\226\128\139', "'"); -- replace {{'}} with simple apostrophe (as of 2015-12-11)
value = value:gsub ('&nbsp;', ' '); -- replace &nbsp; entity with plain space
value = value:gsub ('\226\128\138', ' '); -- replace hair space with plain space
if not mw.ustring.find (value, cfg.indic_script) then -- don't remove zero width joiner characters from indic script
value = value:gsub ('&zwj;', ''); -- remove &zwj; entities
value = value:gsub ('[\226\128\141\226\128\139]&zwj;', ''); -- remove zero-width joiner, zero-width&zwj; spaceentities
value = value:mw.ustring.gsub (value, '[\194226\173128\009141\010226\013128\139\194\173]', ' '); -- replaceremove softzero-width hyphenjoiner, horizontalzero-width tabspace, linesoft feed, carriage return with plain spacehyphen
value = value:gsub ('[\009\010\013]', ' '); -- replace horizontal tab, line feed, carriage return with plain space
return value;
if in_array (class, {'arxiv', 'biorxiv', 'citeseerx', 'ssrn', 'journal', 'news', 'magazine'}) 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 in_array (class, {'arxiv', =='biorxiv', class'citeseerx', 'ssrn'}) then -- set genre according to the type of citation template we are rendering
OCinSoutput["rft.genre"] = "preprint"; -- cite arxiv, cite biorxiv, cite citeseerx, cite ssrn
elseif 'conference' == class then
OCinSoutput["rft.genre"] = "conference"; -- cite conference (when Periodical set)
elseif 'conference' == class then -- cite conference when Periodical not set
OCinSoutput["rft.genre"] = "conference";
OCinSoutput["rft.atitle"] = data.Chapter; -- conference paper as chapter in proceedings (book)
elseif in_array (class, {'book', 'citation', 'encyclopaedia', 'interview', 'map'}) then
if is_set (data.Chapter) then
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_isbnv:gsub( v"[^-0-9X]", "" ); 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
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 };
OCinSoutput[ id ] = value;
local last, first;
for k, v in ipairs( data.Authors ) do
-- 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, "&");
cfg = cfg_table_ptr;
is_set = utilities_page_ptr.is_set; -- import functions from selectselected Module:Citation/CS1/Utilities module
in_array = utilities_page_ptr.in_array;
remove_wiki_link = utilities_page_ptr.remove_wiki_link;
strip_apostrophe_markup = utilities_page_ptr.strip_apostrophe_markup;
--[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------
return {
Cookies help us deliver our services. By using our services, you agree to our use of cookies.