Module:Lang: Difference between revisions

From Vigyanwiki
Template>Trappist the monk
(make is_latn() accessible to Module:Lang/utilities)
m (209 revisions imported from alpha:Module:Lang)
 
(71 intermediate revisions by 12 users not shown)
Line 5: Line 5:
]=]
]=]


require('Module:No globals');
-- require('strict');
local p = {};


local initial_style_state; -- set by lang_xx_normal() and lang_xx_italic()
local getArgs = require ('Module:Arguments').getArgs;
local unicode = require ("Module:Unicode data"); -- for is_latin() and is_rtl()
local yesno = require ('Module:Yesno');


local getArgs = require ('Module:Arguments').getArgs;
local lang_data = mw.loadData ('Module:Lang/data'); -- language name override and transliteration tool-tip tables
local lang_name_table = mw.loadData ('Module:Language/name/data');
local lang_name_table = lang_data.lang_name_table; -- language codes, names, regions, scripts, suppressed scripts
local lang_table = lang_data.lang_name_table.lang;
local lang_dep_table = lang_data.lang_name_table.lang_dep;
local script_table = lang_data.lang_name_table.script;
local region_table = lang_data.lang_name_table.region;
local variant_table = lang_data.lang_name_table.variant;
local suppressed_table = lang_data.lang_name_table.suppressed;
local override_table = lang_data.override;


local synonym_table = mw.loadData ('Module:Lang/ISO 639 synonyms'); -- ISO 639-2/639-2T code translation to 639-1 code
local synonym_table = mw.loadData ('Module:Lang/ISO 639 synonyms'); -- ISO 639-2/639-2T code translation to 639-1 code


local lang_data = mw.loadData ('Module:Lang/data'); -- language name override and transliteration tool-tip tables
local namespace = mw.title.getCurrentTitle().namespace; -- used for categorization
 
local content_lang = mw.language.getContentLanguage();
local this_wiki_lang_tag = content_lang.code; -- get this wiki's language tag
local this_wiki_lang_dir = content_lang:getDir(); -- get this wiki's language direction


local namespace = mw.title.getCurrentTitle().namespace; -- used for categorization
local initial_style_state; -- set by lang_xx_normal() and lang_xx_italic()


local maint_cats = {}; -- maintenance categories go here
local maint_cats = {}; -- maintenance categories go here
Line 34: Line 46:




--[[--------------------------< I S _ L A T N >----------------------------------------------------------------
--[[--------------------------< I N V E R T  _ I T A L I C S >-------------------------------------------------
 
This function attempts to invert the italic markup a args.text by adding/removing leading/trailing italic markup
in args.text.  Like |italic=unset, |italic=invert disables automatic italic markup.  Individual leading/trailing
apostrophes are converted to their html numeric entity equivalent so that the new italic markup doesn't become
bold markup inadvertently.
 
Leading and trailing wiki markup is extracted from args.text into separate table elements.  Addition, removal,
replacement of wiki markup is handled by a string.gsub() replacement table operating only on these separate elements.
In the string.gsub() matching pattern, '.*' matches empty string as well as the three expected wiki markup patterns.
 
This function expects that markup in args.text is complete and correct; if it is not, oddness may result.


Returns true if all of text argument is written using Latn script for letters, numbers and punctuationset; false else.
]]


For the purposes of this function, Latn script is characters less control characters from these Unicode 10.0 Character Code Charts:
local function invert_italics (source)
[http://www.unicode.org/charts/PDF/U0000.pdf C0 Controls and Basic Latin] U+0020–U+007E (20 - 7E) + see note about <poem>...</poem> support
local invert_pattern_table = { -- leading/trailing markup add/remove/replace patterns
[http://www.unicode.org/charts/PDF/U0080.pdf C1 Controls and Latin-1 Supplement] U+00A0-U+00AC, U+00C0–U+00FF (C2 A0 - C2 AC, C3 80 - C3 BF: \194\160-\194\172)
[""]="\'\'", -- empty string becomes italic markup
[http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] U+0100–U+017F (C4 80 - C5 BF)
["\'\'"]="", -- italic markup becomes empty string
[http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] U+0180–U+024F (C6 80 - C9 8F)
["\'\'\'"]="\'\'\'\'\'", -- bold becomes bold italic
[http://www.unicode.org/charts/PDF/U1E00.pdf Latin Extended Additional] U+1E00-U+1EFF (E1 B8 80 - E1 BB BF)
["\'\'\'\'\'"]="\'\'\'", -- bold italic become bold
[http://www.unicode.org/charts/PDF/U2C60.pdf Latin Extended-C] U+2C60–U+2C7F (E2 B1 A0 - E2 B1 BF)
};
[http://www.unicode.org/charts/PDF/UA720.pdf Latin Extended-D] U+A720-U+A7FF (EA 9C A0 - EA 9F BF)
local seg = {};
[http://www.unicode.org/charts/PDF/UAB30.pdf Latin Extended-E] U+AB30-U+AB6F (EA AC B0 - EA AD AF)
[http://www.unicode.org/charts/PDF/UFB00.pdf Alphabetic Presentaion Forms] U+FB00-U+FB06 (EF AC 80 - EF AC 86)
[http://www.unicode.org/charts/PDF/UFF00.pdf Halfwidth and Fullwidth Forms] U+FF01-U+FF3C (EF BC 81 EF BC BC)


does not include:
source = source:gsub ("%f[\']\'%f[^\']", '&#39;'); -- protect single quote marks from being interpreted as bold markup
[http://www.unicode.org/charts/PDF/U1D00.pdf Phonetic Extensions] U+1D00-U+1D7F (E1 B4 80 - E1 B5 BF)
[http://www.unicode.org/charts/PDF/U0250.pdf IPA Extensions] U+0250-U+02AF (C9 90 - CA AF)
[http://www.unicode.org/charts/PDF/U1D80.pdf Phonetic Extensions Supplement] U+1D80-U+1DBF (E1 B6 80 - E1 B6 BF)


{{lang}} is used inside <poem>...</poem> tags for song lyrics, poetry, etc.  <poem>...</poem> replaces newlines with
seg[1] = source:match ('^(\'\'+%f[^\']).+') or ''; -- get leading markup, if any; ignore single quote
poem stripmarkers.  These have the form:
seg[3] = source:match ('.+(%f[\']\'\'+)$') or ''; -- get trailing markup, if any; ignore single quote
?'"`UNIQ--poem-67--QINU`"'?
where the '?' character is actually the delete character (U+007F, \127). Including the '\n' (U+0010) and 'del' (U+007F)
characters in the latn character table allows {{lang}} to auto-italicize text within <poem>...</poem> tags.


]]
if '' ~= seg[1] and '' ~= seg[3] then -- extract the 'text'
seg[2] = source:match ('^\'\'+%f[^\'](.+)%f[\']\'\'+$') -- from between leading and trailing markup
elseif '' ~= seg[1] then
seg[2] = source:match ('^\'\'+%f[^\'](.+)') -- following leading markup
elseif '' ~= seg[3] then
seg[2] = source:match ('(.+)%f[\']\'\'+$') -- preceding trailing markup
else
seg[2] = source -- when there is no markup
end


function p.is_latn (text)
seg[1] = invert_pattern_table[seg[1]] or seg[1]; -- replace leading markup according to pattern table
local latn = table.concat (
seg[3] = invert_pattern_table[seg[3]] or seg[3]; -- replace leading markup according to pattern table
{
'[', -- this is a set so include opening bracket
'\n\32-\127', -- C0 Controls and Basic Latin U+0020–U+007E (20 - 7E) + (U+0010 and U+007F <poem>...</poem> support)
'\194\160-\194\172', -- C1 Controls and Latin-1 Supplement U+00A0-U+00AC (C2 A0 - C2 AC)
'\195\128-\195\191', -- (skip shy) U+00C0–U+00FF (C3 80 - C3 BF)
'\196\128-\197\191', -- Latin Extended-A U+0100–U+017F (C4 80 - C5 BF)
'\198\128-\201\143', -- Latin Extended-B U+0180–U+024F (C6 80 - C9 8F)
'\225\184\128-\225\187\191', -- Latin Extended Additional U+1E00-U+1EFF (E1 B8 80 - E1 BB BF)
'\226\177\160-\226\177\191', -- Latin Extended-C U+2C60–U+2C7F (E2 B1 A0 - E2 B1 BF)
'\234\156\160-\234\159\191', -- Latin Extended-D U+A720-U+A7FF (EA 9C A0 - EA 9F BF)
'\234\172\176-\234\173\175', -- Latin Extended-E U+AB30-U+AB6F (EA AC B0 - EA AD AF)
'\239\172\128-\239\172\134', -- Alphabetic Presentaion Forms U+FB00-U+FB06 (EF AC 80 - EF AC 86)
'\239\188\129-\239\188\188', -- Halfwidth and Fullwidth Forms U+FF01-U+FF3C (EF BC 81 - EF BC BC)
'–', -- ndash
'—', -- mdash
'«', '»', -- guillemets commonly used in several 'Latn' languages
']', -- close the set
});


text = mw.text.decode (text, true); -- replace numeric and named html entities with their unicode characters
return table.concat (seg); -- put it all back together and done
text = mw.ustring.gsub (text, '%[%[[^|]+|([^%]]+)%]%]', '%1'); -- remove the link and markup from complex wikilink in case interwiki to non-Latn wikipedia
return not is_set (mw.ustring.gsub (text, latn, '')); -- replace all latn characters with empty space; if result is all empty space, text is latn
end
end


Line 112: Line 114:
no - force args.text to be rendered in normal font; returns 'normal'
no - force args.text to be rendered in normal font; returns 'normal'
unset - disables font control so that font-style applied to text is dictated by markup inside or outside the template; returns 'inherit'
unset - disables font control so that font-style applied to text is dictated by markup inside or outside the template; returns 'inherit'
invert - disables font control so that font-style applied to text is dictated by markup outside or inverted inside the template; returns 'invert'


]]
]]


local function validate_italic (italic, italics)
local function validate_italic (args)
local properties = {['yes'] = 'italic', ['no'] = 'normal', ['unset'] = 'inherit', ['default'] = nil};
local properties = {['yes'] = 'italic', ['no'] = 'normal', ['unset'] = 'inherit', ['invert'] = 'invert', ['default'] = nil};
local count = 0
for _, arg in pairs {'italic', 'italics', 'i'} do
if args[arg] then
count = count + 1
end
end


if italic and italics then -- return nil and an error message if both are set  
if count > 1 then -- return nil and an error message if more than one is set  
return nil, 'conflicting: &#124;italic= and &#124;italics=';
return nil, 'only one of |italic=, |italics=, or |i= can be specified';
end
end
return properties[italic or italics], nil; -- return an appropriate value and a nil error message
return properties[args.italic or args.italics or args.i], nil; -- return an appropriate value and a nil error message
end
 
 
--[=[--------------------------< V A L I D A T E _ C A T _ A R G S >----------------------------------------------------------
 
Default behavior of the {{lang}} and {{lang-xx}} templates is to add categorization when the templates are used in mainspace.
This default functionality may be suppressed by setting |nocat=yes or |cat=no.  This function selects one of these two parameters
to control categorization.
 
Because having two parameters with 'opposite' names and 'opposite' values is confusing, this function accepts only affirmative
values for |nocat= and only negative values for |cat=; in both cases the 'other' sense (and non-sense) is not accepted and the
parameter is treated as if it were not set in the template.
 
Sets args.nocat to true if categorization is to be turned off; to nil if the default behavior should apply.
 
Accepted values for |nocat= are the text strings:
'yes', 'y', 'true', 't', on, '1' -- [[Module:Yesno]] returns logical true for all of these; false or nil else
for |cat=
'no', 'n', 'false', 'f', 'off', '0' -- [[Module:Yesno]] returns logical false for all of these; true or nil else
 
]=]
 
local function validate_cat_args (args)
if not (args.nocat or args.cat) then -- both are nil, so categorize
return;
end
if false == yesno (args.cat) or true == yesno (args.nocat) then
args.nocat = true; -- set to true when args.nocat is affirmative; nil else (as if the parameter were not set in the template)
else -- args.nocat is the parameter actually used.
args.nocat = nil;
end
end
end


Line 132: Line 174:
]]
]]


local function in_array( needle, haystack )
local function in_array ( needle, haystack )
if needle == nil then
if needle == nil then
return false;
return false;
Line 152: Line 194:
region: upper case
region: upper case
variant: lower case
variant: lower case
 
private: lower case prefixed with -x-
]]
]]


local function format_ietf_tag (code, script, region, variant)
local function format_ietf_tag (code, script, region, variant, private)
local out = {};
local out = {};


if is_set (private) then
return table.concat ({code:lower(), 'x', private:lower()}, '-'); -- if private, all other tags ignored
end
table.insert (out, code:lower());
table.insert (out, code:lower());
if is_set (script) then
if is_set (script) then
Line 182: Line 229:
script subtag - four character IANA script code
script subtag - four character IANA script code
region subtag - two-letter or three digit IANA region code
region subtag - two-letter or three digit IANA region code
variant subtag - four digit or 5-8 alnum variant code
variant subtag - four digit or 5-8 alnum variant code; only one variant subtag supported
private subtag - x- followed by 1-8 alnum private code; only supported with the primary language tag
private subtag - x- followed by 1-8 alnum private code; only supported with the primary language tag


Line 201: Line 248:
because those parameters are superfluous to the IETF subtags in |code=)
because those parameters are superfluous to the IETF subtags in |code=)


returns six  values.  Valid parts are returned as themselves; omitted parts are returned as empty strings, invalid
returns six  values; all lower case.  Valid parts are returned as themselves; omitted parts are returned as empty strings, invalid
parts are returned as nil; the sixth returned item is an error message (if an error detected) or nil.
parts are returned as nil; the sixth returned item is an error message (if an error detected) or nil.


Line 209: Line 256:


local function get_ietf_parts (source, args_script, args_region, args_variant)
local function get_ietf_parts (source, args_script, args_region, args_variant)
local code;
local code, script, region, variant, private; -- ietf tag parts
local script = '';
local region = '';
local variant = '';
local private = '';


if not is_set (source) then
if not is_set (source) then
Line 219: Line 262:
end
end


if source:match ('^%a%a%a?%-%a%a%a%a%-%a%a%-%d%d%d%d$') then -- ll-Ssss-RR-variant (where variant is 4 digits)
local pattern = { -- table of tables holding acceptibe ietf tag patterns and short names of the ietf part captured by the pattern
code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%d%d%d%d)$');
{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%d%d%d%d)$', 's', 'r', 'v'}, -- 1 - ll-Ssss-RR-variant (where variant is 4 digits)
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%d%d%d%-%d%d%d%d$') then -- ll-Ssss-DDD-variant (where region is 3 digits; variant is 4 digits)
{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%d%d%d%d)$', 's', 'r', 'v'}, -- 2 -  ll-Ssss-DDD-variant (where region is 3 digits; variant is 4 digits)
code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%d%d%d%d)$');
{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'r', 'v'}, -- 3 -  ll-Ssss-RR-variant (where variant is 5-8 alnum characters)
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-Ssss-RR-variant (where variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'r', 'v'}, -- 4 - ll-Ssss-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%d%d%d%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-Ssss-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d%d)$', 's', 'v'}, -- 5 -  ll-Ssss-variant (where variant is 4 digits)
code, script, region, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
{'^(%a%a%a?)%-(%a%a%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'v'}, -- 6 -  ll-Ssss-variant (where variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%a%a)%-(%d%d%d%d)$', 'r', 'v'}, -- 7 -  ll-RR-variant (where variant is 4 digits)
{'^(%a%a%a?)%-(%d%d%d)%-(%d%d%d%d)$', 'r', 'v'}, -- 8 -  ll-DDD-variant (where region is 3 digits; variant is 4 digits)
{'^(%a%a%a?)%-(%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 'r', 'v'}, -- 9 - ll-RR-variant (where variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%d%d%d)%-(%w%w%w%w%w%w?%w?%w?)$', 'r', 'v'}, -- 10 - ll-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%d%d%d%d)$', 'v'}, -- 11 - ll-variant (where variant is 4 digits)
{'^(%a%a%a?)%-(%w%w%w%w%w%w?%w?%w?)$', 'v'}, -- 12 - ll-variant (where variant is 5-8 alnum characters)
{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$', 's', 'r'}, -- 13 - ll-Ssss-RR
{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$', 's', 'r'}, -- 14 - ll-Ssss-DDD (region is 3 digits)
{'^(%a%a%a?)%-(%a%a%a%a)$', 's'}, -- 15 - ll-Ssss
{'^(%a%a%a?)%-(%a%a)$', 'r'}, -- 16 - ll-RR
{'^(%a%a%a?)%-(%d%d%d)$', 'r'}, -- 17 - ll-DDD (region is 3 digits)
{'^(%a%a%a?)$'}, -- 18 - ll
{'^(%a%a%a?)%-x%-(%w%w?%w?%w?%w?%w?%w?%w?)$', 'p'}, -- 19 - ll-x-pppppppp (private is 1-8 alnum characters)
}


elseif source:match ('^%a%a%a?%-%a%a%a%a%-%d%d%d%d$') then -- ll-Ssss-variant (where variant is 4 digits)
local t = {}; -- table of captures; serves as a translator between captured ietf tag parts and named variables
code, script, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d%d)$');
elseif source:match ('^%a%a%a?%-%a%a%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-Ssss-variant (where variant is 5-8 alnum characters)
code, script, variant = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');


elseif source:match ('^%a%a%a?%-%a%a%-%d%d%d%d$') then -- ll-RR-variant (where variant is 4 digits)
for i, v in ipairs (pattern) do -- spin through the pattern table looking for a match
code, region, variant = source:match ('^(%a%a%a?)%-(%a%a)%-(%d%d%d%d)$');
local c1, c2, c3, c4; -- captures in the 'pattern' from the pattern table go here
elseif source:match ('^%a%a%a?%-%d%d%d%-%d%d%d%d$') then -- ll-DDD-variant (where region is 3 digits; variant is 4 digits)
code, region, variant = source:match ('^(%a%a%a?)%-(%d%d%d)%-(%d%d%d%d)$');
c1, c2, c3, c4 = source:match (pattern[i][1]); -- one or more captures set if source matches pattern[i])
elseif source:match ('^%a%a%a?%-%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-RR-variant (where variant is 5-8 alnum characters)
if c1 then -- c1 always set on match
code, region, variant = source:match ('^(%a%a%a?)%-(%a%a)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
code = c1; -- first capture is always code
elseif source:match ('^%a%a%a?%-%d%d%d%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
t = {
code, region, variant = source:match ('^(%a%a%a?)%-(%d%d%d)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
[pattern[i][2] or 'x'] = c2, -- fill the table of captures with the rest of the captures
 
[pattern[i][3] or 'x'] = c3, -- take index names from pattern table and assign sequential captures
elseif source:match ('^%a%a%a?%-%d%d%d%d$') then -- ll-variant (where variant is 4 digits)
[pattern[i][4] or 'x'] = c4, -- index name may be nil in pattern[i] table so "or 'x'" spoofs a name for this index in this table
code, variant = source:match ('^(%a%a%a?)%-(%d%d%d%d)$');
};
elseif source:match ('^%a%a%a?%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') then -- ll-variant (where variant is 5-8 alnum characters)
script = t.s or ''; -- translate table contents to named variables;
code, variant = source:match ('^(%a%a%a?)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
region = t.r or ''; -- absent table entries are nil so set named ietf parts to empty string for concatenation
 
variant= t.v or '';
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%a%a$') then -- ll-Ssss-RR
private = t.p or '';
code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$');
break; -- and done
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%d%d%d$') then -- ll-Ssss-DDD (region is 3 digits)
end
code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$');
end
 
elseif source:match ('^%a%a%a?%-%a%a%a%a$') then -- ll-Ssss
code, script = source:match ('^(%a%a%a?)%-(%a%a%a%a)$');
 
elseif source:match ('^%a%a%a?%-%a%a$') then -- ll-RR
code, region = source:match ('^(%a%a%a?)%-(%a%a)$');
elseif source:match ('^%a%a%a?%-%d%d%d$') then -- ll-DDD (region is 3 digits)
code, region = source:match ('^(%a%a%a?)%-(%d%d%d)$');
 
elseif source:match ('^%a%a%a?$') then -- ll
code = source:match ('^(%a%a%a?)$');
 
elseif source:match ('^%a%a%a?%-x%-[%a%d][%a%d]?[%a%d]?[%a%d]?[%a%d]?[%a%d]?[%a%d]?[%a%d]?$') then -- ll-x-pppppppp)
code, private = source:match ('^(%a%a%a?)%-x%-([%a%d]+)$');


else
if not code then
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language tag: ', source}); -- don't know what we got but it is malformed
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language tag: ', source}); -- don't know what we got but it is malformed
end
end
Line 272: Line 319:
code = code:lower(); -- ensure that we use and return lower case version of this
code = code:lower(); -- ensure that we use and return lower case version of this
if not (lang_data.override[code] or lang_name_table.lang[code]) then
if not (override_table[code] or lang_table[code] or synonym_table[code] or lang_dep_table[code]) then
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language code: ', code}); -- invalid language code, don't know about the others (don't care?)
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language code: ', code}); -- invalid language code, don't know about the others (don't care?)
end
end
Line 292: Line 339:
if is_set (script) then
if is_set (script) then
script = script:lower(); -- ensure that we use and return lower case version of this
script = script:lower(); -- ensure that we use and return lower case version of this
if not lang_name_table.script[script] then
if not script_table[script] then
return code, nil, nil, nil, nil, table.concat ({'unrecognized script: ', script, ' for code: ', code}); -- language code ok, invalid script, don't know about the others (don't care?)
return code, nil, nil, nil, nil, table.concat ({'unrecognized script: ', script, ' for code: ', code}); -- language code ok, invalid script, don't know about the others (don't care?)
end
end
end
end
if lang_name_table.suppressed[script] then -- ensure that code-script does not use a suppressed script
if suppressed_table[script] then -- ensure that code-script does not use a suppressed script
if in_array (code, lang_name_table.suppressed[script]) then
if in_array (code, suppressed_table[script]) then
return code, nil, nil, nil, nil, table.concat ({'script: ', script, ' not supported for code: ', code}); -- language code ok, script is suppressed for this code
return code, nil, nil, nil, nil, table.concat ({'script: ', script, ' not supported for code: ', code}); -- language code ok, script is suppressed for this code
end
end
Line 312: Line 359:
if is_set (region) then
if is_set (region) then
region = region:lower(); -- ensure that we use and return lower case version of this
region = region:lower(); -- ensure that we use and return lower case version of this
if not lang_name_table.region[region] then
if not region_table[region] then
return code, script, nil, nil, nil, table.concat ({'unrecognized region: ', region, ' for code: ', code});
return code, script, nil, nil, nil, table.concat ({'unrecognized region: ', region, ' for code: ', code});
end
end
Line 327: Line 374:
if is_set (variant) then
if is_set (variant) then
variant = variant:lower(); -- ensure that we use and return lower case version of this
variant = variant:lower(); -- ensure that we use and return lower case version of this
if not lang_name_table.variant[variant] then -- make sure variant is valid
if not variant_table[variant] then -- make sure variant is valid
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant});
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant});
end -- does this duplicate/replace tests in lang() and lang_xx()?
end -- does this duplicate/replace tests in lang() and lang_xx()?
if is_set (script) then -- if script set it must be part of the 'prefix'
if is_set (script) then -- if script set it must be part of the 'prefix'
if not in_array (table.concat ({code, '-', script}), lang_name_table.variant[variant]['prefixes']) then
if not in_array (table.concat ({code, '-', script}), variant_table[variant]['prefixes']) then
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-script pair: ', code, '-', script});
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-script pair: ', code, '-', script});
end
end
else
elseif is_set (region) then -- if region set, there are some prefixes that require lang code and region (en-CA-newfound)
if not in_array (code, lang_name_table.variant[variant]['prefixes']) then
if not in_array (code, variant_table[variant]['prefixes']) then -- first see if lang code is all that's required (en-oxendict though en-GB-oxendict is preferred)
if not in_array (table.concat ({code, '-', region}), variant_table[variant]['prefixes']) then -- now try for lang code and region (en-CA-newfound)
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-region pair: ', code, '-', region});
end
end
else -- cheap way to determine if there are prefixes; fonipa and others don't have prefixes; # operator always returns 0
if variant_table[variant]['prefixes'][1] and not in_array (code, variant_table[variant]['prefixes']) then
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code: ', code});
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code: ', code});
end
end
end
end
end
end
 
if is_set (private) then
if is_set (private) then
private = private:lower(); -- ensure that we use and return lower case version of this
private = private:lower(); -- ensure that we use and return lower case version of this
if not lang_data.override[table.concat ({code, '-x-', private})] then -- make sure private tag is valid; note that index  
if not override_table[table.concat ({code, '-x-', private})] then -- make sure private tag is valid; note that index  
return code, script, region, nil, nil, table.concat ({'unrecognized private tag: ', private});
return code, script, region, nil, nil, table.concat ({'unrecognized private tag: ', private});
end
end
Line 361: Line 414:
local category;
local category;
if 'transl' == template then
if 'Transl' == template then
category = 'transl';
category = 'Transl';
else
else
category = 'lang and lang-xx'
category = 'Lang and lang-xx'
end
end
table.insert (out, table.concat ({'&#x5B;', args.text or 'undefined', '&#x5D; '})); -- for error messages output args.text if available
table.insert (out, table.concat ({'[', args.text or 'undefined', '] '})); -- for error messages output args.text if available
table.insert (out, table.concat ({'<span style=\"font-size:100%; font-style:normal;\" class=\"error\">error: {{', template, '}}: '}));
table.insert (out, table.concat ({'<span style=\"color:#d33\">Error: {{', template, '}}: '}));
table.insert (out, msg);
table.insert (out, msg);
table.insert (out, table.concat ({' ([[:Category:', category, ' template errors|help]])'}));
table.insert (out, table.concat ({' ([[:Category:', category, ' template errors|help]])'}));
table.insert (out, '</span>');
table.insert (out, '</span>');
if (0 == namespace) and not is_set (args.nocat) then -- only categorize in article space
if (0 == namespace or 10 == namespace) and not args.nocat then -- categorize in article space (and template space to take care of broken usages)
table.insert (out, table.concat ({'[[Category:', category, ' template errors]]'}));
table.insert (out, table.concat ({''}));
end
end


Line 402: Line 455:




--[[--------------------------< M A K E _ T E X T _ S P A N >--------------------------------------------------
--[[--------------------------< D I V _ M A R K U P _ A D D >--------------------------------------------------
 
adds <i> and </i> tags to list-item text or to implied <p>..</p> text.  mixed not supported
 
]]


TODO: add support for block: div tags instead of span tags; would need some sort of proper parameter to control the switch
local function div_markup_add (text, style)
local implied_p = {};


For italic style, can't do ''{{lang|xx|text}}'' without using <span/> tags when text is italic because of -Latn, |italic=yes,
if text:find ('^\n[%*:;#]') then -- look for list markup; list markup must begin at start of text
or auto-italics because the wrapping wikimarkup produces this:
if 'italic' == style then
<i><i lang="xx">text</i></i>
return text:gsub ('(\n[%*:;#]+)([^\n]+)', '%1<i>%2</i>'); -- insert italic markup at each list item
which is later reduced to this:
else
<i>text</i>
return text;
This reduction happens in some sort of cleanup process outside the scope of this template/module.
end
end


Until or unless this is fixed italic text must be:
if text:find ('\n+') then -- look for any number of \n characters in text
<i><span lang="xx">text</span></i>
text = text:gsub ('([^\n])\n([^\n])', '%1 %2'); -- replace single newline characters with a space character which mimics mediawiki


if 'italic' == style then
text = text:gsub('[^\n]+', '<p><i>%1</i></p>'); -- insert p and italic markup tags at each impled p (two or more consecutive '\n\n' sequences)
else
text = text:gsub ('[^\n]+', '<p>%1</p>'); -- insert p markup at each impled p
text = text:gsub ('\n', ''); -- strip newline characters
end
end
return text;
end
--[[--------------------------< T I T L E _ W R A P P E R _ M A K E >------------------------------------------
makes a <span title="<title text>"><content_text></span> or <div title="<title text>"><content_text></div> where
<title text> is in the tool-tip in the wiki's local language and <content_text> is non-local-language text in
html markup.  This because the lang= attibute applies to the content of its enclosing tag.
<tag> holds a string 'div' or 'span' used to choose the correct wrapping tag
]]
]]


local function make_text_span (code, text, rtl, style, size, language)
local function title_wrapper_make (title_text, content_text, tag)
local span = {};
local wrapper_t = {};
table.insert (wrapper_t, table.concat ({'<', tag})); -- open opening wrapper tag
table.insert (wrapper_t, ' title=\"'); -- begin title attribute
table.insert (wrapper_t, title_text); -- add <title_text>
table.insert (wrapper_t, '\">'); -- end title attribute and close opening wrapper tag
table.insert (wrapper_t, content_text); -- add <content_text>
table.insert (wrapper_t, table.concat ({'</', tag, '>'})); -- add closing wrapper tag
 
return table.concat (wrapper_t); -- make a big string and done
end
 
 
--[[--------------------------< M A K E _ T E X T _ H T M L >--------------------------------------------------
 
Add the html markup to text according to the type of content that it is: <span> or <i> tags for inline content or
<div> tags for block content
 
The lang= attribute also applies to the content of the tag where it is placed so this is wrong because 'Spanish
language text' is English:
<i lang="es" title="Spanish language text">casa</i>
should be:
<span title="Spanish language text"><i lang="es">casa</i></span>
or for <div>...</div>:
<div title="Spanish language text"><div lang="es"><spanish-language-text></div></div>
 
]]
 
local function make_text_html (code, text, tag, rtl, style, size, language)
local html_t = {};
local style_added = '';
local style_added = '';
local wrapper_tag = tag; -- <tag> gets modified so save a copy for use when/if we create a wrapper span or div


if text:match ('^%*') then
if text:match ('^%*') then
table.insert (span, '&#42;'); -- move proto language text prefix outside of italic markup if any; use numeric entity because plan splat confuses MediaWiki
table.insert (html_t, '&#42;'); -- move proto language text prefix outside of italic markup if any; use numeric entity because plain splat confuses MediaWiki
text = text:gsub ('^%*', ''); -- remove the spat from the text
text = text:gsub ('^%*', ''); -- remove the splat from the text
end
end


if 'italic' == style then
if 'span' == tag then -- default html tag for inline content
table.insert (span, '<i>'); -- open italic style tag
if 'italic' == style then -- but if italic
tag = 'i'; -- change to <i> tags
end
else -- must be div so go
text = div_markup_add (text, style); -- handle implied <p>, implied <p> with <i>, and list markup (*;:#) with <i>
end
end
table.insert (span, table.concat ({'<span lang="'})); -- open <span> tag
table.insert (span, table.concat ({code, '\"'})); -- add language attribute


if rtl then
table.insert (html_t, table.concat ({'<', tag})); -- open the <i>, <span>, or <div> html tag
table.insert (span, ' dir="rtl"'); -- add direction attribute for right to left languages
code = code:gsub ('%-x%-.*', ''); -- strip private use subtag from code tag because meaningless outside of wikipedia
table.insert (html_t, table.concat ({' lang="', code, '\"'})); -- add language attribute
 
if (rtl or unicode.is_rtl(text)) and ('ltr' == this_wiki_lang_dir) then -- text is right-to-left on a left-to-right wiki
table.insert (html_t, ' dir="rtl"'); -- add direction attribute for right-to-left languages
elseif not (rtl or unicode.is_rtl(text)) and ('rtl' == this_wiki_lang_dir) then -- text is left-to-right on a right-to-left wiki
table.insert (html_t, ' dir="ltr"'); -- add direction attribute for left-to-right languages
end
end


if 'normal' == style then -- when |italic=no
if 'normal' == style then -- when |italic=no
table.insert (span, ' style=\"font-style:normal;'); -- override external markup, if any
table.insert (html_t, ' style=\"font-style: normal;'); -- override external markup, if any
style_added = '\"'; -- remember that style attribute added and is not yet closed
style_added = '\"'; -- remember that style attribute added and is not yet closed
end
end
Line 445: Line 560:
if is_set (size) then -- when |size=<something>
if is_set (size) then -- when |size=<something>
if is_set (style_added) then
if is_set (style_added) then
table.insert (span, table.concat ({' font-size:', size, ';'})); -- add when style attribute already inserted
table.insert (html_t, table.concat ({' font-size: ', size, ';'})); -- add when style attribute already inserted
else
else
table.insert (span, table.concat ({' style=\"font-size:', size, ';'})); -- create style attribute
table.insert (html_t, table.concat ({' style=\"font-size: ', size, ';'})); -- create style attribute
style_added = '\"'; -- remember that style attribute added and is not yet closed
style_added = '\"'; -- remember that style attribute added and is not yet closed
end
end
end
end


if is_set (language) then
table.insert (html_t, table.concat ({style_added, '>'})); -- close the opening html tag
table.insert (span, table.concat ({style_added, ' title=\"', language, ' language text\"', '>'})); -- add title text and close opening span tag
table.insert (html_t, text); -- insert the text
 
table.insert (html_t, table.concat ({'</', tag, '>'})); -- close the 'text' <i>, <span>, or <div> html tag
 
if is_set (language) then -- create a <title_text> string for the title= attribute in a wrapper span or div
local title_text;
if 'zxx' == code then -- special case for this tag 'no linguistic content'
title_text = table.concat ({language, ' text'}); -- not a language so don't use 'language' in title text
elseif mw.ustring.find (language, 'languages', 1, true) then
title_text = table.concat ({language, ' collective text'}); -- for collective languages
else
title_text = table.concat ({language, '-language text'}); -- for individual languages
end
 
return title_wrapper_make (title_text, table.concat (html_t), wrapper_tag);
else
else
table.insert (span, table.concat ({style_added, '>'})); -- close the style attribute and close opening span tag
return table.concat (html_t);
end
table.insert (span, text); -- insert the text
 
table.insert (span, '</span>'); -- close the <span> tag
if 'italic' == style then
table.insert (span, '</i>'); -- close italic style tag
end
if rtl then -- legacy; shouldn't be necessary because all of the rtl text is wrapped in <span dir="rtl">text</span>
table.insert (span, '&lrm;'); -- make sure the browser knows that we're at the end of the rtl
end
end
return table.concat (span); -- put it all together and done
end
end


Line 475: Line 592:


For individual language, <language>, returns:
For individual language, <language>, returns:
[[Category:Articles containing <language>-language text]]


for English:
for English:
[[Category:Articles containing explicitly cited English-language text]]
for artificial languages (code: art)
[[Category:Articles containing constructed-language text]]
 
for ISO 639-2 collective languages (and for 639-1 bh):
for ISO 639-2 collective languages (and for 639-1 bh):
[[Category:Articles with text from the <language> languages collective]]


]=]
]=]


local function make_category (code, language_name, nocat)
local function make_category (code, language_name, nocat, name_get)
local cat = {};
local cat = {};
local retval;
if (0 ~= namespace) or nocat then -- only categorize in article space
 
if ((0 ~= namespace) or nocat) and not name_get then -- only categorize in article space
return ''; -- return empty string for concatenation
return ''; -- return empty string for concatenation
end
end
 
if language_name:find ('languages') then
if mw.ustring.find (language_name, 'languages', 1, true) then
return table.concat ({'[[Category:Articles with text from the ', language_name, ' collective]]'});
return table.concat ({''});
end
end
table.insert (cat, '[[Category:Articles containing ');
table.insert (cat, '[[Category:Articles containing ');
 
if 'en' == code then
if this_wiki_lang_tag == code then
table.insert (cat, 'explicitly cited English');
table.insert (cat, 'explicitly cited ' .. language_name); -- unique category name for the local language
elseif 'art' == code then
table.insert (cat, 'constructed')
else
else
table.insert (cat, language_name);
table.insert (cat, language_name);
Line 511: Line 624:
table.insert (cat, '-language text]]');
table.insert (cat, '-language text]]');


return table.concat (cat);
return table.concat (cat);
end
end


Line 535: Line 648:
local function make_translit (code, language_name, translit, std, tscript, style)
local function make_translit (code, language_name, translit, std, tscript, style)
local title;
local title;
local tout = {};
local out_t = {};
local title_table = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
local title_t = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
local title_text = ''; -- tool tip text for title= attribute
if is_set (code) then -- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})
if not style then -- nil for is the default italic style
table.insert (tout, "<i lang=\""); -- so use <i> tag
else
table.insert (tout, table.concat ({'<span style=\"font-style:', style, '\" lang=\"'})); -- non-standard style, construct a span tag for it
end
table.insert (tout, code);
table.insert (tout, "-Latn\" title=\""); -- transliterations are always Latin script
else
table.insert (tout, "<span title=\""); -- when no language code: no lang= attribute, not italic ({{transl}} only)
end
std = std and std:lower(); -- lower case for table indexing
std = std and std:lower(); -- lower case for table indexing
if not is_set (std) and not is_set (tscript) then -- when neither standard nor script specified
if not is_set (std) and not is_set (tscript) then -- when neither standard nor script specified
table.insert (tout, language_name); -- write a generic tool tip
title_text = language_name; -- write a generic tool tip
if not language_name:find ('languages') then -- collective language names (plural 'languages' is part of the name)
if not mw.ustring.find (language_name, 'languages', 1, true) then -- collective language names (plural 'languages' is part of the name)
table.insert (tout, '-language') -- skip this text (individual and macro languages only)
title_text = title_text .. '-language'; -- skip this text (individual and macro languages only)
end
end
table.insert (tout, ' transliteration'); -- finish the tool tip
title_text = title_text .. ' romanization'; -- finish the tool tip; use romanization when neither script nor standard supplied
 
elseif is_set (std) and is_set (tscript) then -- when both are specified
elseif is_set (std) and is_set (tscript) then -- when both are specified
if title_table[std] then -- and if standard is legitimate
if title_t[std] then -- and if standard is legitimate
if title_table[std][tscript] then -- and if script for that standard is legitimate
if title_t[std][tscript] then -- and if script for that standard is legitimate
table.insert (tout, table.concat ({title_table[std][tscript:lower()], ' (', lang_name_table.script[tscript][1], ' script) transliteration'})); -- add the appropriate text to the tool tip
if script_table[tscript] then
title_text = title_text .. table.concat ({title_t[std][tscript:lower()], ' (', script_table[tscript], ' script) transliteration'}); -- add the appropriate text to the tool tip
else
title_text = title_text .. title_t[std]['default']; -- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?
end
else
else
table.insert (tout, title_table[std]['default']); -- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?
title_text = title_text .. title_t[std]['default']; -- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?
end
end
else
else
Line 570: Line 677:


elseif is_set (std) then -- translit-script not set, use language code
elseif is_set (std) then -- translit-script not set, use language code
if not title_table[std] then return ''; end -- invalid standard, setup for error message
if not title_t[std] then return ''; end -- invalid standard, setup for error message
if title_table[std][code] then -- if language code is in the table (transl may not provide a language code)
if title_t[std][code] then -- if language code is in the table (transl may not provide a language code)
table.insert (tout, table.concat ({title_table[std][code:lower()], ' (', lang_name_table.lang[code][1], ' language) transliteration'})); -- add the appropriate text to the tool tip
title_text = title_text .. table.concat ({title_t[std][code:lower()], ' (', language_name, ' language) transliteration'}); -- add the appropriate text to the tool tip
else -- code doesn't match
else -- code doesn't match
table.insert (tout, title_table[std]['default']); -- so use the standard's default
title_text = title_text .. title_t[std]['default']; -- so use the standard's default
end
end
else -- here if translit-script set but translit-std not set
else -- here if translit-script set but translit-std not set
if title_table['no_std'][tscript] then
if title_t['no_std'][tscript] then
table.insert (tout, title_table['no_std'][tscript]); -- use translit-script if set
title_text = title_text .. title_t['no_std'][tscript]; -- use translit-script if set
elseif title_table['no_std'][code] then
elseif title_t['no_std'][code] then
table.insert (tout, title_table['no_std'][code]); -- use language code
title_text = title_text .. title_t['no_std'][code]; -- use language code
else
else
if is_set (tscript) then
if is_set (tscript) then
table.insert (tout, table.concat ({language_name, '-script transliteration'})); -- write a script tool tip
title_text = title_text .. table.concat ({language_name, '-script transliteration'}); -- write a script tool tip
elseif is_set (code) then
elseif is_set (code) then
if not language_name:find ('languages') then -- collective language names (plural 'languages' is part of the name)
if not mw.ustring.find (language_name, 'languages', 1, true) then -- collective language names (plural 'languages' is part of the name)
table.insert (tout, '-language') -- skip this text (individual and macro languages only)
title_text = title_text .. '-language'; -- skip this text (individual and macro languages only)
end
end
table.insert (tout, ' transliteration'); -- finish the tool tip
title_text = title_text .. ' transliteration'; -- finish the tool tip
else
else
table.insert (tout, ' transliteration'); -- generic tool tip (can we ever get here?)
title_text = title_text .. ' transliteration'; -- generic tool tip (can we ever get here?)
end
end
end
end
end
end


table.insert (tout, '">');
local close_tag;
table.insert (tout, translit);
if is_set (code) then -- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})
if is_set (code) and not style then -- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})
if not style then -- nil for the default italic style
table.insert (tout, "</i>"); -- close the italic tag
table.insert (out_t, "<i lang=\""); -- so use <i> tag
close_tag = '</i>'; -- tag to be used when closing
else
table.insert (out_t, table.concat ({'<span style=\"font-style: ', style, '\" lang=\"'})); -- non-standard style, construct a span tag for it
close_tag = '</span>'; -- tag to be used when closing
end
table.insert (out_t, code);
table.insert (out_t, "-Latn\">"); -- transliterations are always Latin script
else
table.insert (out_t, "<span>"); -- when no language code: no lang= attribute, not italic ({{transl}} only)
close_tag = '</span>';
end
 
table.insert (out_t, translit); -- add the translit text
table.insert (out_t, close_tag); -- and add the appropriate </i> or </span>
 
if '' == title_text then -- when there is no need for a tool-tip
return table.concat (out_t); -- make a string and done
else
else
table.insert (tout, "</span>"); -- no language code so close the span tag
return title_wrapper_make (title_text, table.concat (out_t), 'span'); -- wrap with a tool-tip span and don
end
end
return table.concat (tout);
end
end




--[=[-------------------------< V A L I D A T E _ T E X T >---------------------------------------------------
--[[--------------------------< V A L I D A T E _ T E X T >----------------------------------------------------


This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
Line 613: Line 737:


Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
either as ''itself''' or as '''''bold italic'''''.
either as ''itself'' or as '''''bold italic''''' unless |italic=unset or |italic=invert.


]=]
]]


local function validate_text (template, args)
local function validate_text (template, args)
Line 626: Line 750:
end
end


if ('unset' ~= args.italic) and ('unset' ~= args.italics) then -- allow italic markup when |italic=unset or |italics=unset
local style = args.italic;
if args.text:match ("%f[\']\'\'[^\']+\'\'%f[^\']") or args.text:match ("\'\'\'\'\'[^\']+\'\'\'\'\'") then -- italic but not bold, or bold italic
 
if ('unset' ~= style) and ('invert' ~=style) then
if args.text:find ("%f[\']\'\'%f[^\']") or args.text:find ("%f[\']\'\'\'\'\'%f[^\']") then -- italic but not bold, or bold italic
return make_error_msg ('text has italic markup', args, template);
return make_error_msg ('text has italic markup', args, template);
end
end
Line 644: Line 770:
if 0 < #maint_msgs then -- when there are maintenance messages
if 0 < #maint_msgs then -- when there are maintenance messages
table.insert (maint, table.concat ({'<span class="lang-comment" style="font-style:normal; display:none; color:#33aa33; margin-left:0.3em">'})); -- opening <span> tag
table.insert (maint, table.concat ({'<span class="lang-comment" style="font-style: normal; display: none; color: #33aa33; margin-left: 0.3em;">'})); -- opening <span> tag
for _, msg in ipairs (maint_msgs) do
for _, msg in ipairs (maint_msgs) do
table.insert (maint, table.concat ({msg, ' '})); -- add message strings
table.insert (maint, table.concat ({msg, ' '})); -- add message strings
Line 651: Line 777:
end
end
if (0 < #maint_cats) and (0 == namespace) and not is_set (nocat) then -- when there are mainenance categories; article namespace only
if (0 < #maint_cats) and (0 == namespace) and not nocat then -- when there are maintenance categories; article namespace only
for _, cat in ipairs (maint_cats) do
for _, cat in ipairs (maint_cats) do
table.insert (maint, table.concat ({'[[Category:', cat, ']]'})); -- format and add the categories
table.insert (maint, table.concat ({''})); -- format and add the categories
end
end
end
end
Line 663: Line 789:
--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------
--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------


for proto languages, text is prefixed with a splat.  We do that here as a flag for make_text_span() so that a splat
for proto languages, text is prefixed with a splat.  We do that here as a flag for make_text_html() so that a splat
will be rendered outside of italic markup (if used).  If the first character in text here is already a splat, we
will be rendered outside of italic markup (if used).  If the first character in text here is already a splat, we
do nothing
do nothing
proto_param is boolean or nil; true adds splat prefix regardless of language name; false removes and / or inhibits
regardless of language name; nil does nothing; presumes that the value in text is correct but removes extra splac


]]
]]


local function proto_prefix (text, language_name)
local function proto_prefix (text, language_name, proto_param)
if language_name:find ('^Proto%-') and not text:find ('^*') then -- language is a proto and text does not already have leading splat
if false == proto_param then -- when forced by |proto=no
return table.concat ({'*', text}); -- prefix proto language text with a splat
return text:gsub ('^%**', ''); -- return text without splat prefix regardless of language name or existing splat prefix in text
elseif (language_name:find ('^Proto%-') or (true == proto_param)) then -- language is a proto or forced by |proto=yes
return text:gsub ('^%**', '*'); -- prefix proto-language text with a splat; also removes duplicate prefixing splats
end
end
return text;
return text:gsub ('^%*+', '*'); -- return text unmolested except multiple splats reduced to one splat
end
 
 
--[[--------------------------< H A S _ P O E M _ T A G >------------------------------------------------------
 
looks for a poem strip marker in text; returns true when found; false else
 
auto-italic detection disabled when text has poem stripmarker because it is not possible for this code to know
the content that will replace the stripmarker.
 
]]
 
local function has_poem_tag (text)
return text:find ('\127[^\127]*UNIQ%-%-poem%-[%a%d]+%-QINU[^\127]*\127') and true or false;
end
 
 
--[[--------------------------< H T M L _ T A G _ S E L E C T >------------------------------------------------
 
Inspects content of and selectively trims text.  Returns text and the name of an appropriate html tag for text.
 
If text contains:
\n\n text has implied <p>..</p> tags - trim leading and trailing whitespace and return
If text begins with list markup:
\n* unordered
\n; definition
\n: definition
\n# ordered
trim all leading whitespace except  \n and trim all trailing whitespace
 
If text contains <poem>...</poem> stripmarker, return text unmodified and choose <div>..</div> tags because
the stripmarker is replaced with text wrapped in <div>..</div> tags.
 
]]
 
local function html_tag_select (text)
local tag;
if has_poem_tag (text) then -- contains poem stripmarker (we can't know the content of that)
tag = 'div'; -- poem replacement is in div tags so lang must use div tags
elseif mw.text.trim (text):find ('\n\n+') then -- contains implied p tags
text = mw.text.trim (text); -- trim leading and trailing whitespace characters
tag = 'div'; -- must be div because span may not contain p tags (added later by MediaWiki); poem replacement is in div tags
elseif text:find ('\n[%*:;%#]') then -- if text has list markup
text = text:gsub ('^[\t\r\f ]*', ''):gsub ('%s*$', ''); -- trim all whitespace except leading newline character '\n'
tag = 'div'; -- must be div because span may not contain ul, dd, dl, ol tags (added later by MediaWiki)
else
text = mw.text.trim (text); -- plain text
tag = 'span'; -- so span is fine
end
return text, tag;
end
 
 
--[[--------------------------< V A L I D A T E _ P R O T O >--------------------------------------------------
 
validates value assigned to |proto=; permitted values are yes and no; yes returns as true, no returns as false,
empty string (or parameter omitted) returns as nil; any other value returns as nil with a second return value of
true indicating that some other value has been assigned to |proto=
 
]]
 
local function validate_proto (proto_param)
if 'yes' == proto_param then
return true;
elseif 'no' == proto_param then
return false;
elseif is_set (proto_param) then
return nil, true; -- |proto= something other than 'yes' or 'no'
else
return nil; -- missing or empty
end
end
 
 
--[[--------------------------< L A N G U A G E _ N A M E _ G E T >--------------------------------------------
 
common function to return language name from the data set according to IETF tag
 
returns language name if found in data tables; nil else
 
]]
 
local function language_name_get (ietf, code, cat)
ietf = ietf:lower(); -- ietf:lower() because format_ietf_tag() returns mixed case
local name; -- remains nil if not found
if override_table[ietf] then -- look for whole IETF tag in override table
name = override_table[ietf];
elseif override_table[code] then -- not there so try basic language tag
name = override_table[code];
elseif lang_table[code] then -- shift to iana active tag/name table
name = lang_table[code];
elseif lang_dep_table[code] then -- try the iana deprecated tag/name table
name = lang_dep_table[code];
end
 
if lang_dep_table[code] and cat then -- because deprecated code may have been overridden to en.wiki preferred name
table.insert (maint_cats, table.concat ({'Lang and lang-xx using deprecated ISO 639 codes|', code}));
table.insert (maint_msgs, table.concat ({'code: ', code, ' is deprecated'}));
end
 
return name; -- name from data tables or nil
end
end




--[[--------------------------< L A N G >----------------------------------------------------------------------
--[[--------------------------< _ L A N G >--------------------------------------------------------------------


entry point for {{lang}}
entry point for {{lang}}
Line 685: Line 921:
<includeonly>{{#invoke:lang|lang}}</includeonly>
<includeonly>{{#invoke:lang|lang}}</includeonly>


parameters are recieved from the template's frame (parent frame)
parameters are received from the template's frame (parent frame)


]]
]]


function p.lang (frame)
local function _lang (args)
local args = getArgs(frame);
local out = {};
local out = {};
local language_name; -- used to make category names
local language_name; -- used to make category names
local category_name; -- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
local subtags = {}; -- IETF subtags script, region, variant, and private
local subtags = {}; -- IETF subtags script, region, variant, and private
local code; -- the language code
local code; -- the language code
local msg; -- for error messages
local msg; -- for error messages
local tag = 'span'; -- initial value for make_text_html()
local template = args.template or 'Lang';
maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
maint_msgs = {};
validate_cat_args (args); -- determine if categorization should be suppressed


if args[1] and args.code then
if args[1] and args.code then
return make_error_msg ('conflicting: {{{1}}} and &#124;code=', args, 'lang');
return make_error_msg ('conflicting: {{{1}}} and |code=', args, template);
else
else
args.code = args[1] or args.code; -- prefer args.code
args.code = args[1] or args.code; -- prefer args.code
Line 704: Line 947:


if args[2] and args.text then
if args[2] and args.text then
return make_error_msg ('conflicting: {{{2}}} and &#124;text=', args, 'lang');
return make_error_msg ('conflicting: {{{2}}} and |text=', args, template);
else
else
args.text = args[2] or args.text; -- prefer args.text
args.text = args[2] or args.text; -- prefer args.text
end
end
msg = validate_text ('lang', args); -- ensure that |text= is set (italic test disabled for the time being)
msg = validate_text (template, args); -- ensure that |text= is set
if is_set (msg) then -- msg is an already-formatted error message
if is_set (msg) then -- msg is an already-formatted error message
return msg;
return msg;
end
end
args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly


args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
args.proto, msg = validate_proto (args.proto); -- return boolean, or nil, or nil and error message flag
if msg then
return make_error_msg (table.concat ({'invalid |proto=: ', args.proto}), args, template);
end


code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code); -- |script=, |region=, |variant= not supported because they should be part of args.code ({{{1}}} in {{lang}})
code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code); -- |script=, |region=, |variant= not supported because they should be part of args.code ({{{1}}} in {{lang}})


if msg then
if msg then
return make_error_msg ( msg, args, 'lang');
return make_error_msg ( msg, args, template);
end
end


args.italic, msg = validate_italic (args.italic, args.italics);
args.italic, msg = validate_italic (args);
if msg then
if msg then
return make_error_msg (msg, args, 'lang');
return make_error_msg (msg, args, template);
end
end


if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
if ('latn' == subtags.script) or (p.is_latn (args.text) and 'en' ~= code) then -- script set to latn or text is wholly latn script but not rendering English
if ('latn' == subtags.script) or -- script is latn
args.italic = 'italic'; -- DEFAULT for {{lang}} templates is upright; but if latn script set for font-style:italic
(this_wiki_lang_tag ~= code and not is_set (subtags.script) and not has_poem_tag (args.text) and unicode.is_Latin (args.text)) then -- text not this wiki's language, no script specified and not in poem markup but is wholly latn script (auto-italics)
args.italic = 'italic'; -- DEFAULT for {{lang}} templates is upright; but if latn script set for font-style:italic
else
else
args.italic = 'inherit'; -- italic not set; script not latn; inherit current style
args.italic = 'inherit'; -- italic not set; script not latn; inherit current style
Line 743: Line 994:
end
end


args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant); -- format to recommended subtag styles; private omitted because private
args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles
language_name = language_name_get (args.code, code, true); -- get language name; try ietf tag first, then code w/o variant then code w/ variant


if is_set (subtags.private) and lang_data.override[table.concat ({code, '-x-', subtags.private})] then -- look for private use tags; done this way because ...
if 'invert' == args.italic and 'span' == tag then -- invert only supported for in-line content
language_name = lang_data.override[table.concat ({code, '-x-', subtags.private})][1]; -- ... args.code does not get private subtag
args.text = invert_italics (args.text)
elseif lang_data.override[code] then -- get the language name for categorization
language_name = lang_data.override[code][1] -- prefer language names taken from the override table
elseif lang_name_table.lang[code] then
language_name = lang_name_table.lang[code][1]; -- table entries sometimes have multiple names, always take the first one
end
end


args.text = proto_prefix (args.text, language_name); -- prefix proto-language text with a splat
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat


table.insert (out, make_text_span (args.code, args.text, args.rtl, args.italic, args.size, language_name));
table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, language_name));
table.insert (out, make_category (code, language_name, args.nocat));
table.insert (out, make_category (code, language_name, args.nocat));
table.insert (out, render_maint(args.nocat)); -- maintenance messages and categories
table.insert (out, render_maint(args.nocat)); -- maintenance messages and categories


return table.concat (out); -- put it all together and done
return table.concat (out); -- put it all together and done
end
--[[--------------------------< L A N G >----------------------------------------------------------------------
entry point for {{lang}}
there should be no reason to set parameters in the {{lang}} {{#invoke:}}
<includeonly>{{#invoke:lang|lang}}</includeonly>
parameters are received from the template's frame (parent frame)
]]
local function lang (frame)
local args = getArgs (frame, { -- this code so that we can detect and handle wiki list markup in text
valueFunc = function (key, value)
if 2 == key or 'text' == key then -- the 'text' parameter; do not trim wite space
return value; -- return untrimmed 'text'
elseif value then -- all other values: if the value is not nil
value = mw.text.trim (value); -- trim whitespace
if '' ~= value then -- empty string when value was only whitespace
return value;
end
end
return nil; -- value was empty or contained only whitespace
end -- end of valueFunc
});
return _lang (args);
end
--[[--------------------------< T R A N S L A T I O N _ M A K E >----------------------------------------------
stand-alone function to create literal translation of main text
Also used by {{lang-x2}}
]]
local function translation_make (args_t)
local translation_t = {', '}; -- initialize output
if 'none' ~= args_t.label then -- if we want a label
table.insert (translation_t, '<small>'); -- open the <small> html tag
if 'no' == args_t.link then
table.insert (translation_t, '<abbr title="literal translation">lit.</abbr>'); -- unlinked form
else
table.insert (translation_t, make_wikilink ('Literal translation', 'lit.')); -- linked form
end
table.insert (translation_t, "&thinsp;</small>"); -- close the <small> html tag
end
table.insert (translation_t, table.concat ({'&#39;', args_t.translation, '&#39;'})); -- use html entities to avoid wiki markup confusion
return table.concat (translation_t); -- make a big string and done
end
end


Line 769: Line 1,072:
that can have multiple writing systems, it may be appropriate to set |script= as well.
that can have multiple writing systems, it may be appropriate to set |script= as well.


For each {{lang-xx}} template choose the appropriate entry-point function so that this function know the default
For each {{lang-xx}} template choose the appropriate entry-point function so that this function knows the default
styling that should be applied to text.
styling that should be applied to text.


Line 781: Line 1,084:
Supported parameters are:
Supported parameters are:
|code = (required) the IANA language code
|code = (required) the IANA language code
|script = IANA script code; especially for use with languages that use multiple writing systems; yields to the script subtag in |code= if present [not yet implemented]
|script = IANA script code; especially for use with languages that use multiple writing systems
|region = IANA region code
|region = IANA region code
|variant = IANA variant code
|variant = IANA variant code
|text = (required) the displayed text in language specified by code
|text = (required) the displayed text in language specified by code
|link = boolean false ('no') unlinks language specified by code to associated language article
|link = boolean false ('no') does not link code-spcified language name to associated language article
|rtl = boolean true ('yes') identifies the language specified by code as a right-to-left language
|rtl = boolean true ('yes') identifies the language specified by code as a right-to-left language
|nocat = boolean true ('yes') inhibits normal categorization; error categories are not affected
|nocat = boolean true ('yes') inhibits normal categorization; error categories are not affected
|cat = boolian false ('no') opposite form of |nocat=
|italic = boolean true ('yes') renders displayed text in italic font; boolean false ('no') renders displayed text in normal font; not set renders according to initial_style_state
|italic = boolean true ('yes') renders displayed text in italic font; boolean false ('no') renders displayed text in normal font; not set renders according to initial_style_state
|lit = text that is a literal translation of text
|lit = text that is a literal translation of text
|label = 'none' to suppress all labeling (language name, 'translit.', 'lit.')
|label = 'none' to suppress all labeling (language name, 'translit.', 'lit.')
any other text replaces language-name label - automatic wikilinking disabled
any other text replaces language-name label - automatic wikilinking disabled
for those {{lang-xx}} templates that support transliteration (those template where |text= is entirely latn script):
for those {{lang-xx}} templates that support transliteration (those templates where |text= is not entirely latn script):
|translit = text that is a transliteration of text
|translit = text that is a transliteration of text
|translit-std = the standard that applies to the transliteration
|translit-std = the standard that applies to the transliteration
Line 806: Line 1,110:
]]
]]


local function _lang_xx (frame)
local function _lang_xx (args)
local args = getArgs(frame, {parentFirst= true}); -- parameters in the template override parameters set in the {{#invoke:}}
local out = {};
local out = {};
local language_name; -- used to make display text, article links
local language_name; -- used to make display text, article links
Line 818: Line 1,121:
local translit_title;
local translit_title;
local msg; -- for error messages
local msg; -- for error messages
local tag = 'span'; -- initial value for make_text_html()
local template = args.template or 'Lang-xx';


maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
maint_msgs = {};
if args[1] and args.text then
if args[1] and args.text then
return make_error_msg ('conflicting: {{{1}}} and &#124;text=', args, 'lang-xx');
return make_error_msg ('conflicting: {{{1}}} and |text=', args, template);
else
else
args.text = args[1] or args.text; -- prefer args.text
args.text = args[1] or args.text; -- prefer args.text
end
end
msg = validate_text ('lang-xx', args); -- ensure that |text= is set, does not contain italic markup and is protected from improper bolding
msg = validate_text (template, args); -- ensure that |text= is set, does not contain italic markup and is protected from improper bolding
if is_set (msg) then
if is_set (msg) then
return msg;
return msg;
end
end
args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly


if args[2] and args.translit then
if args[2] and args.translit then
return make_error_msg ('conflicting: {{{2}}} and &#124;translit=', args, 'lang-xx');
return make_error_msg ('conflicting: {{{2}}} and |translit=', args, template);
else
else
args.translit = args[2] or args.translit -- prefer args.translit
args.translit = args[2] or args.translit -- prefer args.translit
Line 837: Line 1,147:
if args[3] and (args.translation or args.lit) then
if args[3] and (args.translation or args.lit) then
return make_error_msg ('conflicting: {{{3}}} and &#124;lit= or &#124;translation=', args, 'lang-xx');
return make_error_msg ('conflicting: {{{3}}} and |lit= or |translation=', args, template);
elseif args.translation and args.lit then
elseif args.translation and args.lit then
return make_error_msg ('conflicting: &#124;lit= and &#124;translation=', args, 'lang-xx');
return make_error_msg ('conflicting: |lit= and |translation=', args, template);
else
else
args.translation = args[3] or args.translation or args.lit; -- prefer args.translation
args.translation = args[3] or args.translation or args.lit; -- prefer args.translation
Line 845: Line 1,155:


if args.links and args.link then
if args.links and args.link then
return make_error_msg ('conflicting: &#124;links= and &#124;link=', args, 'lang-xx');
return make_error_msg ('conflicting: |links= and |link=', args, template);
else
else
args.link = args.link or args.links; -- prefer args.link
args.link = args.link or args.links; -- prefer args.link
end
end
validate_cat_args (args); -- determine if categorization should be suppressed


args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
Line 855: Line 1,167:


if msg then -- if an error detected then there is an error message
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args, 'lang-xx');
return make_error_msg (msg, args, template);
end
end
args.italic, msg = validate_italic (args.italic, args.italics);
args.italic, msg = validate_italic (args);
if msg then
if msg then
return make_error_msg (msg, args, 'lang-xx');
return make_error_msg (msg, args, template);
end
end


Line 883: Line 1,195:
end
end


args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant); -- format to recommended subtag styles
args.proto, msg = validate_proto (args.proto); -- return boolean, or nil, or nil and error message flag
if msg then
if is_set (subtags.private) and lang_data.override[table.concat ({code, '-x-', subtags.private})] then -- look for private use tags; done this way because ...
return make_error_msg (table.concat ({'invalid |proto=: ', args.proto}), args, template);
language_name = lang_data.override[table.concat ({code, '-x-', subtags.private})][1]; -- ... args.code does not get private subtag
elseif lang_data.override[args.code:lower()] then -- look for whole IETF tag in override table
language_name = lang_data.override[args.code:lower()][1]; -- args.code:lower() because format_ietf_tag() returns mixed case
elseif lang_data.override[code] then -- not there so try basic language code
language_name = lang_data.override[code][1];
elseif not is_set (subtags.variant) then
if lang_name_table.lang[code] then
language_name = lang_name_table.lang[code][1]; -- table entries sometimes have multiple names, always take the first one
end
else -- TODO: is this the right thing to do: take language display name from variants table?
if lang_name_table.variant[subtags.variant] then -- TODO: there is some discussion at Template talk:Lang about having a label parameter for use when variant name is not desired among other things
language_name = lang_name_table.variant[subtags.variant]['descriptions'][1]; -- table entries sometimes have multiple names, always take the first one
end
end
end
args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles
language_name = language_name_get (args.code, code, true); -- get language name; try ietf tag first, then code w/o variant then code w/ variant


category_name = language_name; -- category names retain IANA parenthetical diambiguators (if any)
category_name = language_name; -- category names retain IANA parenthetical diambiguators (if any)
Line 912: Line 1,214:
table.insert (out, language_name); -- language name without wikilink
table.insert (out, language_name); -- language name without wikilink
else
else
if language_name:find ('languages') then
if mw.ustring.find (language_name, 'languages', 1, true) then
table.insert (out, make_wikilink (language_name)); -- collective language name uses simple wikilink
table.insert (out, make_wikilink (language_name)); -- collective language name uses simple wikilink
elseif lang_data.article_name[code] then
table.insert (out, make_wikilink (lang_data.article_name[code], language_name)); -- language name with wikilink from override data
else
else
table.insert (out, make_wikilink (language_name .. ' language', language_name)); -- language name with wikilink
table.insert (out, make_wikilink (language_name .. ' language', language_name)); -- language name with wikilink
Line 921: Line 1,225:
end
end


args.text = proto_prefix (args.text, language_name); -- prefix proto-language text with a splat
if 'invert' == args.italic then
args.text = invert_italics (args.text)
end
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat


table.insert (out, make_text_span (args.code, args.text, args.rtl, args.italic, args.size))
table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, ('none' == args.label) and language_name or nil))


if is_set (args.translit) and not p.is_latn (args.text) then -- transliteration (not supported in {{lang}}); not supported when args.text is wholly latn text (this is an imperfect test)
if is_set (args.translit) and not unicode.is_Latin (args.text) then -- transliteration (not supported in {{lang}}); not supported when args.text is wholly latn text (this is an imperfect test)
table.insert (out, ', '); -- comma to separate text from translit
table.insert (out, ', '); -- comma to separate text from translit
if 'none' ~= args.label then
if 'none' ~= args.label then
table.insert (out, '<small>');
table.insert (out, '<small>');
if lang_name_table.script[args['translit-script']] then -- when |translit-script= is set, try to use the script's name
if script_table[args['translit-script']] then -- when |translit-script= is set, try to use the script's name
translit_script_name = lang_name_table.script[args['translit-script'][1]];
translit_script_name = script_table[args['translit-script']];
else
else
translit_script_name = language_name; -- fall back on language name
translit_script_name = language_name; -- fall back on language name
Line 936: Line 1,244:
translit_title = mw.title.makeTitle (0, table.concat ({'Romanization of ', translit_script_name})); -- make a title object
translit_title = mw.title.makeTitle (0, table.concat ({'Romanization of ', translit_script_name})); -- make a title object
if translit_title.exists and ('no' ~= args.link) then
if translit_title.exists and ('no' ~= args.link) then
table.insert (out, make_wikilink ('Romanization of ' .. translit_script_name or language_name, 'translit.')); -- make a wikilink if there is an article to link to
table.insert (out, make_wikilink ('Romanization of ' .. translit_script_name or language_name, 'romanized') .. ':'); -- make a wikilink if there is an article to link to
else
else
table.insert (out, '<abbr title="transliteration">translit.</abbr>'); -- else define the abbreviation
table.insert (out, 'romanized:'); -- else plain text
end
end
table.insert (out, '&nbsp;</small>'); -- close the small tag
table.insert (out, '&nbsp;</small>'); -- close the small tag
Line 947: Line 1,255:
table.insert (out, translit);
table.insert (out, translit);
else
else
return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args, 'lang-xx');
return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args, template);
end
end
end
end
if is_set (args.translation) then -- translation (not supported in {{lang}})
if is_set (args.translation) then -- translation (not supported in {{lang}})
table.insert (out, ', ');
table.insert (out, translation_make (args));
if 'none' ~= args.label then
table.insert (out, '<small>');
if 'no' == args.link then
table.insert (out, '<abbr title="literal translation">lit.</abbr>');
else
table.insert (out, make_wikilink ('Literal translation', 'lit.'));
end
table.insert (out, "&nbsp;</small>");
end
table.insert (out, table.concat ({"'", args.translation, "'"}));
end
end
Line 969: Line 1,267:


return table.concat (out); -- put it all together and done
return table.concat (out); -- put it all together and done
end
--[[--------------------------< L A N G _ X X _ A R G S _ G E T >----------------------------------------------
common function to get args table from {{lang-??}} templates
returns table of args
]]
local function lang_xx_args_get (frame)
local args = getArgs(frame,
{
parentFirst= true, -- parameters in the template override parameters set in the {{#invoke:}}
valueFunc = function (key, value)
if 1 == key then -- the 'text' parameter; do not trim wite space
return value; -- return untrimmed 'text'
elseif value then -- all other values: if the value is not nil
value = mw.text.trim (value); -- trim whitespace
if '' ~= value then -- empty string when value was only whitespace
return value;
end
end
return nil; -- value was empty or contained only whitespace
end -- end of valueFunc
});
return args;
end
end


Line 978: Line 1,305:
]]
]]


function p.lang_xx_italic (frame)
local function lang_xx_italic (frame)
local args = lang_xx_args_get (frame);
initial_style_state = 'italic';
initial_style_state = 'italic';
return _lang_xx (frame);
return _lang_xx (args);
end
 
 
--[[--------------------------< _ L A N G _ X X _ I T A L I C >------------------------------------------------
 
Entry point ffrom another module.  Sets the initial style state to italic.
 
]]
 
local function _lang_xx_italic (args)
initial_style_state = 'italic';
return _lang_xx (args);
end
end


Line 990: Line 1,331:
]]
]]


function p.lang_xx_inherit (frame)
local function lang_xx_inherit (frame)
local args = lang_xx_args_get (frame);
 
initial_style_state = 'inherit';
return _lang_xx (args);
end
 
 
--[[--------------------------< _ L A N G _ X X _ I N H E R I T >----------------------------------------------
 
Entry point from another module.  Sets the initial style state to inherit.
 
]]
 
local function _lang_xx_inherit (args)
initial_style_state = 'inherit';
initial_style_state = 'inherit';
return _lang_xx (frame);
return _lang_xx (args);
end
 
 
--[[--------------------------< _ I S _ I E T F _ T A G >------------------------------------------------------
 
Returns true when a language name associated with IETF language tag exists; nil else.  IETF language tag must be valid.
 
All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.
 
Module entry point from another module
 
]]
 
local function _is_ietf_tag (tag) -- entry point when this module is require()d into another
local c, s, r, v, p, err; -- code, script, region, variant, private, error message
c, s, r, v, p, err = get_ietf_parts (tag); -- disassemble tag into constituent part and validate
 
return ((c and not err) and true) or nil; -- return true when code portion has a value without error message; nil else
end
 
 
--[[--------------------------< I S _ I E T F _ T A G >--------------------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]
 
local function is_ietf_tag (frame)
return _is_ietf_tag (getArgs (frame)[1]); -- args[1] is the ietf language tag to be tested; also get parent frame
end
 
 
--[[--------------------------< I S _ I E T F _ T A G _ F R A M E >--------------------------------------------
 
Module entry point from an {{#invoke:}}; same as is_ietf_tag() except does not get parameters from the parent
(template) frame.  This function not useful when called by {{lang|fn=is_ietf_tag_frame|<tag>}} because <tag>
is in the parent frame.
 
]]
 
local function is_ietf_tag_frame (frame)
return _is_ietf_tag (getArgs (frame, {frameOnly = true,})[1]); -- args[1] is the ietf language tag to be tested; do not get parent frame
end
end




--[[--------------------------< N A M E _ F R O M _ C O D E >--------------------------------------------------
--[[--------------------------< _ N A M E _ F R O M _ T A G >--------------------------------------------------


Returns language name associated with IETF language tag if valid; empty string else.
Returns language name associated with IETF language tag if valid; error message else.


All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.
All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.
Set invoke's |link= parameter to yes to get wikilinked version of the language name.
Module entry point from another module


]]
]]


function p.name_from_code (frame)
local function _name_from_tag (args)
local subtags = {}; -- IETF subtags script, region, variant, and private
local subtags = {}; -- IETF subtags script, region, variant, and private
local raw_code = (frame.args and frame.args[1]) or frame; -- save a copy of the input
local raw_code = args[1]; -- save a copy of the input IETF subtag
local link = 'yes' == args['link']; -- make a boolean
local label = args.label;
local code; -- the language code
local code; -- the language code
local msg; -- holds an error message (not used here) if IETF language tag is malformed or invalid
local msg; -- gets an error message if IETF language tag is malformed or invalid
local language_name = '';
local language_name = '';
 
code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (raw_code);
code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (raw_code);
if msg then
if msg then
return '';
local template = (args['template'] and table.concat ({'{{', args['template'], '}}: '})) or ''; -- make template name (if provided by the template)
return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
end
end


if lang_data.override[raw_code:lower()] then -- look for whole IETF tag in override table (force lower case)
raw_code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles; private omitted because private
language_name = lang_data.override[raw_code:lower()][1];
language_name = language_name_get (raw_code, code); -- get language name; try ietf tag first, then code w/o variant then code w/ variant
elseif lang_data.override[code] then -- not there so try basic language code in override table
 
language_name = lang_data.override[code][1];
language_name = language_name:gsub ('%s+%b()', ''); -- remove IANA parenthetical disambiguators or qualifiers from names that have them
elseif not is_set (subtags.variant) then
 
if lang_name_table.lang[code] then
if link then -- when |link=yes, wikilink the language name
language_name = lang_name_table.lang[code][1]; -- table entries sometimes have multiple names, always take the first one
if mw.ustring.find (language_name, 'languages', 1, true) then
language_name = make_wikilink (language_name, label); -- collective language name uses simple wikilink
elseif lang_data.article_name[code] then
language_name = make_wikilink (lang_data.article_name[code], label or language_name); -- language name with wikilink from override data
else
language_name = make_wikilink (language_name .. ' language', label or language_name); -- language name with wikilink
end
end
else -- TODO: is this the right thing to do: take language display name from variants table?
end
if lang_name_table.variant[subtags.variant] then -- TODO: there is some discussion at Template talk:Lang about having a label parameter for use when variant name is not desired among other things
 
language_name = lang_name_table.variant[subtags.variant]['descriptions'][1]; -- table entries sometimes have multiple names, always take the first one
return language_name;
end
 
 
--[[--------------------------< N A M E _ F R O M _ T A G >----------------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]
 
local function name_from_tag (frame) -- entry point from an {{#invoke:Lang|name_from_tag|<ietf tag>|link=<yes>|template=<template name>}}
return _name_from_tag (getArgs(frame)) -- pass-on the args table, nothing else; getArgs() so we also get parent frame
end
 
 
--[[--------------------------< _ T A G _ F R O M _ N A M E >--------------------------------------------------
 
Returns the ietf language tag associated with the language name.  Spelling of language name must be correct
according to the spelling in the source tables.  When a standard language name has a parenthetical disambiguator,
that disambiguator must be omitted (they are not present in the data name-to-tag tables).
 
Module entry point from another module
 
]]
 
local function _tag_from_name (args) -- entry point from another module
local msg;
 
if args[1] and '' ~= args[1] then
local data = mw.loadData ('Module:Lang/tag from name'); -- get the reversed data tables TODO: change when going live
local lang = args[1]:lower(); -- allow any-case for the language name (speeling must till be correct)
local tag = data.rev_override_table[lang] or data.rev_lang_table[lang] or data.rev_lang_dep_table[lang]; -- get the code; look first in the override then in the standard
 
if tag then
return tag, true; -- language name found so return tag and done; second return used by is_lang_name()
else
msg = 'language: ' .. args[1] .. ' not found' -- language name not found, error message
end
end
else
msg = 'missing language name' -- language name not provided, error message
end
local template = '';
if args.template and '' ~= args.template then
template = table.concat ({'{{', args['template'], '}}: '}); -- make template name (if provided by the template)
end
end
return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
end


language_name = language_name:gsub ('%s+%b()', ''); -- remove IANA parenthetical disambiguators or qualifiers from names that have them


return language_name;
--[[--------------------------< T A G _ F R O M _ N A M E >----------------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]
 
local function tag_from_name (frame) -- entry point from an {{#invoke:Lang|tag_from_name|<language name>|link=<yes>|template=<template name>}}
local result, _ = _tag_from_name (getArgs(frame)) -- pass-on the args table, nothing else; getArgs() so we also get parent frame; supress second return used by is_lang_name()
return result;
end
 
 
--[[--------------------------< I S _ L A N G _ N A M E >------------------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]


local function is_lang_name (frame)
local _, result = _tag_from_name (getArgs(frame)) -- pass-on the args table, nothing else; getArgs() so we also get parent frame; supress second return used by tag_from_name()
return result and true or nil;
end
end




--[[--------------------------< T R A N S L >------------------------------------------------------------------
--[[--------------------------< _ T R A N S L >----------------------------------------------------------------


Prospective replacement for the template {{transl}}
Module entry point from another module


]]
]]


function p.transl (frame)
local function _transl (args)
local args = getArgs(frame); -- no {{#invoke:}} parameters
local title_table = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
local title_table = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
local language_name; -- language name that matches language code; used for tool tip
local language_name; -- language name that matches language code; used for tool tip
local translit; -- translitterated text to display
local translit; -- transliterated text to display
local script; -- IANA script
local script; -- IANA script
local msg; -- for when called functions return an error message
local msg; -- for when called functions return an error message


maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
maint_msgs = {};
if is_set (args[3]) then -- [3] set when {{transl|code|standard|text}}
if is_set (args[3]) then -- [3] set when {{transl|code|standard|text}}
args.text = args[3]; -- get the transliterated text
args.text = args[3]; -- get the transliterated text
Line 1,056: Line 1,530:


if not title_table[args.translit_std] then
if not title_table[args.translit_std] then
return make_error_msg (table.concat ({'unrecognized transliteration standard: ', args.translit_std}), args, 'transl');
return make_error_msg (table.concat ({'unrecognized transliteration standard: ', args.translit_std}), args, 'Transl');
end
end
else
else
Line 1,062: Line 1,536:
args.text = args[2]; -- get the transliterated text
args.text = args[2]; -- get the transliterated text
else
else
if args[1] and args[1]:match ('^%a%a%a?%a?$') then -- args[2] missing; is args[1] a code or its it the transliterated text?
if args[1] and args[1]:match ('^%a%a%a?%a?$') then -- args[2] missing; is args[1] a code or is it the transliterated text?
return make_error_msg ('no text', args, 'transl'); -- args[1] is a code so we're missing text
return make_error_msg ('no text', args, 'Transl'); -- args[1] is a code so we're missing text
else
else
args.text = args[1]; -- args[1] is not a code so we're missing that; assign args.text for error message
args.text = args[1]; -- args[1] is not a code so we're missing that; assign args.text for error message
return make_error_msg ('missing language / script code', args, 'transl');
return make_error_msg ('missing language / script code', args, 'Transl');
end
end
end
end
Line 1,075: Line 1,549:
args.code = args[1]:lower(); -- use the language/script code; only (2, 3, or 4 alpha characters); lower case because table indexes are lower case
args.code = args[1]:lower(); -- use the language/script code; only (2, 3, or 4 alpha characters); lower case because table indexes are lower case
else
else
return make_error_msg (table.concat ({'unrecognized language / script code: ', args[1]}), args, 'transl'); -- invalid language / script code
return make_error_msg (table.concat ({'unrecognized language / script code: ', args[1]}), args, 'Transl'); -- invalid language / script code
end
end
else
else
return make_error_msg ('missing language / script code', args, 'transl'); -- missing language / script code so quit
return make_error_msg ('missing language / script code', args, 'Transl'); -- missing language / script code so quit
end
end


args.italic, msg = validate_italic (args.italic, args.italics);
args.italic, msg = validate_italic (args);
if msg then
if msg then
return make_error_msg (msg, args, 'transl');
return make_error_msg (msg, args, 'Transl');
end
end
Line 1,090: Line 1,564:
end
end


if lang_data.override[args.code] then -- is code a language code defined in the override table?
if override_table[args.code] then -- is code a language code defined in the override table?
language_name = lang_data.override[args.code][1];
language_name = override_table[args.code];
elseif lang_name_table.lang[args.code] then -- is code a language code defined in the standard language code tables?
elseif lang_table[args.code] then -- is code a language code defined in the standard language code tables?
language_name = lang_name_table.lang[args.code][1];
language_name = lang_table[args.code];
elseif lang_name_table.script[args.code] then -- if here, code is not a language code; is it a script code?
elseif lang_dep_table[args.code] then -- is code a language code defined in the deprecated language code tables?
language_name = lang_name_table.script[args.code][1];
language_name = lang_dep_table[args.code];
elseif script_table[args.code] then -- if here, code is not a language code; is it a script code?
language_name = script_table[args.code];
script = args.code; -- code was an ISO 15924 script so use that instead
script = args.code; -- code was an ISO 15924 script so use that instead
args.code = ''; -- unset because not a language code
args.code = ''; -- unset because not a language code
else
else
return make_error_msg (table.concat ({'unrecognized language / script code: ', args.code}), args, 'transl'); -- invalid language / script code
return make_error_msg (table.concat ({'unrecognized language / script code: ', args.code}), args, 'Transl'); -- invalid language / script code
end
end
-- here only when all parameters passed to make_translit() are valid
-- here only when all parameters passed to make_translit() are valid
Line 1,106: Line 1,582:




return p;
--[[--------------------------< T R A N S L >------------------------------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]
 
local function transl (frame)
return _transl (getArgs(frame));
end
 
 
--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------
 
Returns category name associated with IETF language tag if valid; error message else
 
All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.
 
Module entry point from another module
 
]]
 
local function _category_from_tag (args_t)
local subtags = {}; -- IETF subtags script, region, variant, and private
local raw_code = args_t[1]; -- save a copy of the input IETF subtag
local link = 'yes' == args_t.link; -- make a boolean
local label = args_t.label;
local code; -- the language code
local msg; -- gets an error message if IETF language tag is malformed or invalid
local category_name = '';
code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (raw_code);
if msg then
local template = (args_t.template and table.concat ({'{{', args_t.template, '}}: '})) or ''; -- make template name (if provided by the template)
return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
end
 
raw_code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles; private omitted because private
category_name = language_name_get (raw_code, code); -- get language name; try ietf tag first, then code w/o variant then code w/ variant
category_name = make_category (code, category_name, nil, true):gsub ('[%[%]]', '');
 
if link then
return table.concat ({'[[:', category_name, ']]'});
else
return category_name;
end
end
 
 
--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------
 
Module entry point from an {{#invoke:}}
 
]]
 
local function category_from_tag (frame) -- entry point from an {{#invoke:Lang|category_from_tag|<ietf tag>|template=<template name>}}
return _category_from_tag (getArgs (frame)); -- pass-on the args table, nothing else; getArgs() so we also get parent frame
end
 
 
--[[--------------------------< E X P O R T E D  F U N C T I O N S >------------------------------------------
]]
 
return {
category_from_tag = category_from_tag,
lang = lang, -- entry point for {{lang}}
lang_xx_inherit = lang_xx_inherit, -- entry points for {{lang-??}}
lang_xx_italic = lang_xx_italic,
is_ietf_tag = is_ietf_tag,
is_ietf_tag_frame = is_ietf_tag_frame,
is_lang_name = is_lang_name,
tag_from_name = tag_from_name, -- returns ietf tag associated with language name
name_from_tag = name_from_tag, -- used for template documentation; possible use in ISO 639 name from code templates
transl = transl, -- entry point for {{transl}}
_category_from_tag = _category_from_tag, -- entry points when this module is require()d into other modules
_lang = _lang,
_lang_xx_inherit = _lang_xx_inherit,
_lang_xx_italic = _lang_xx_italic,
_is_ietf_tag = _is_ietf_tag,
get_ietf_parts = get_ietf_parts,
_tag_from_name = _tag_from_name,
_name_from_tag = _name_from_tag,
_transl = _transl,
_translation_make = translation_make,
};

Latest revision as of 09:18, 13 December 2022

Documentation for this module may be created at Module:Lang/doc

--[=[

Lua support for the {{lang}}, {{lang-xx}}, and {{transl}} templates and replacement of various supporting templates. 

]=]

-- require('strict');

local getArgs = require ('Module:Arguments').getArgs;
local unicode = require ("Module:Unicode data");								-- for is_latin() and is_rtl()
local yesno = require ('Module:Yesno');

local lang_data =  mw.loadData ('Module:Lang/data');							-- language name override and transliteration tool-tip tables
local lang_name_table = lang_data.lang_name_table;								-- language codes, names, regions, scripts, suppressed scripts
local lang_table = lang_data.lang_name_table.lang;
local lang_dep_table = lang_data.lang_name_table.lang_dep;
local script_table = lang_data.lang_name_table.script;
local region_table = lang_data.lang_name_table.region;
local variant_table = lang_data.lang_name_table.variant;
local suppressed_table = lang_data.lang_name_table.suppressed;
local override_table = lang_data.override;

local synonym_table = mw.loadData ('Module:Lang/ISO 639 synonyms');				-- ISO 639-2/639-2T code translation to 639-1 code

local namespace = mw.title.getCurrentTitle().namespace;							-- used for categorization

local content_lang = mw.language.getContentLanguage();
local this_wiki_lang_tag = content_lang.code;									-- get this wiki's language tag
local this_wiki_lang_dir = content_lang:getDir();								-- get this wiki's language direction

local initial_style_state;														-- set by lang_xx_normal() and lang_xx_italic()

local maint_cats = {};															-- maintenance categories go here
local maint_msgs = {};															-- and their messages go here


--[[--------------------------< I S _ S E T >------------------------------------------------------------------

Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not an empty string.

]]

local function is_set( var )
	return not (var == nil or var == '');
end


--[[--------------------------< I N V E R T  _ I T A L I C S >-------------------------------------------------

This function attempts to invert the italic markup a args.text by adding/removing leading/trailing italic markup
in args.text.  Like |italic=unset, |italic=invert disables automatic italic markup.  Individual leading/trailing
apostrophes are converted to their html numeric entity equivalent so that the new italic markup doesn't become
bold markup inadvertently.

Leading and trailing wiki markup is extracted from args.text into separate table elements.  Addition, removal,
replacement of wiki markup is handled by a string.gsub() replacement table operating only on these separate elements.
In the string.gsub() matching pattern, '.*' matches empty string as well as the three expected wiki markup patterns.

This function expects that markup in args.text is complete and correct; if it is not, oddness may result.

]]

local function invert_italics (source)
	local invert_pattern_table = {												-- leading/trailing markup add/remove/replace patterns
		[""]="\'\'",															-- empty string becomes italic markup
		["\'\'"]="",															-- italic markup becomes empty string
		["\'\'\'"]="\'\'\'\'\'",												-- bold becomes bold italic
		["\'\'\'\'\'"]="\'\'\'",												-- bold italic become bold
		};
	local seg = {};

	source = source:gsub ("%f[\']\'%f[^\']", '&#39;');							-- protect single quote marks from being interpreted as bold markup

	seg[1] = source:match ('^(\'\'+%f[^\']).+') or '';							-- get leading markup, if any; ignore single quote
	seg[3] = source:match ('.+(%f[\']\'\'+)$') or '';							-- get trailing markup, if any; ignore single quote

	if '' ~= seg[1] and '' ~= seg[3] then										-- extract the 'text' 
		seg[2] = source:match ('^\'\'+%f[^\'](.+)%f[\']\'\'+$')					-- from between leading and trailing markup
	elseif '' ~= seg[1] then
		seg[2] = source:match ('^\'\'+%f[^\'](.+)')								-- following leading markup
	elseif '' ~= seg[3] then
		seg[2] = source:match ('(.+)%f[\']\'\'+$')								-- preceding trailing markup
	else
		seg[2] = source															-- when there is no markup
	end

	seg[1] = invert_pattern_table[seg[1]] or seg[1];							-- replace leading markup according to pattern table
	seg[3] = invert_pattern_table[seg[3]] or seg[3];							-- replace leading markup according to pattern table

	return table.concat (seg);													-- put it all back together and done
end


--[[--------------------------< V A L I D A T E _ I T A L I C >------------------------------------------------

validates |italic= or |italics= assigned values.

When |italic= is set and has an acceptible assigned value, return the matching css font-style property value or,
for the special case 'default', return nil.

When |italic= is not set, or has an unacceptible assigned value, return nil and a nil error message.

When both |italic= and |italics= are set, returns nil and a 'conflicting' error message.

The return value nil causes the calling lang, lang_xx, or transl function to set args.italic according to the template's
defined default ('inherit' for {{lang}}, 'inherit' or 'italic' for {{lang-xx}} depending on
the individual template's requirements, 'italic' for {{transl}}) or to the value appropriate to |script=, if set ({{lang}}
and {{lang-xx}} only).

Accepted values and the values that this function returns are are:
	nil		-	when |italic= absent or not set; returns nil
	default	-	for completeness, should rarely if ever be used; returns nil
	yes		-	force args.text to be rendered in italic font; returns 'italic'
	no		-	force args.text to be rendered in normal font; returns 'normal'
	unset	-	disables font control so that font-style applied to text is dictated by markup inside or outside the template; returns 'inherit'
	invert	-	disables font control so that font-style applied to text is dictated by markup outside or inverted inside the template; returns 'invert'

]]

local function validate_italic (args)
	local properties = {['yes'] = 'italic', ['no'] = 'normal', ['unset'] = 'inherit', ['invert'] = 'invert', ['default'] = nil};
	
	local count = 0
	for _, arg in pairs {'italic', 'italics', 'i'} do
		if args[arg] then
			count = count + 1
		end
	end

	if count > 1 then															-- return nil and an error message if more than one is set 
		return nil, 'only one of |italic=, |italics=, or |i= can be specified';
	end
	
	return properties[args.italic or args.italics or args.i], nil;				-- return an appropriate value and a nil error message
end


--[=[--------------------------< V A L I D A T E _ C A T _ A R G S >----------------------------------------------------------

Default behavior of the {{lang}} and {{lang-xx}} templates is to add categorization when the templates are used in mainspace.
This default functionality may be suppressed by setting |nocat=yes or |cat=no.  This function selects one of these two parameters
to control categorization.

Because having two parameters with 'opposite' names and 'opposite' values is confusing, this function accepts only affirmative
values for |nocat= and only negative values for |cat=; in both cases the 'other' sense (and non-sense) is not accepted and the
parameter is treated as if it were not set in the template.

Sets args.nocat to true if categorization is to be turned off; to nil if the default behavior should apply.

Accepted values for |nocat= are the text strings:
	'yes', 'y', 'true', 't', on, '1'		-- [[Module:Yesno]] returns logical true for all of these; false or nil else
for |cat=
	'no', 'n', 'false', 'f', 'off', '0'		-- [[Module:Yesno]] returns logical false for all of these; true or nil else

]=]

local function validate_cat_args (args)
	if not (args.nocat or args.cat) then										-- both are nil, so categorize
		return;
	end
	
	if false == yesno (args.cat) or true == yesno (args.nocat) then
		args.nocat = true;														-- set to true when args.nocat is affirmative; nil else (as if the parameter were not set in the template)
	else																		-- args.nocat is the parameter actually used.
		args.nocat = nil;
	end
end


--[[--------------------------< I N _ A R R A Y >--------------------------------------------------------------

Whether needle is in haystack

]]

local function in_array ( needle, haystack )
	if needle == nil then
		return false;
	end
	for n,v in ipairs( haystack ) do
		if v == needle then
			return n;
		end
	end
	return false;
end


--[[--------------------------< F O R M A T _ I E T F _ T A G >------------------------------------------------

prettify ietf tags to use recommended subtag formats:
	code: lower case
	script: sentence case
	region: upper case
	variant: lower case
	private: lower case prefixed with -x-
	
]]

local function format_ietf_tag (code, script, region, variant, private)
	local out = {};

	if is_set (private) then
		return table.concat ({code:lower(), 'x', private:lower()}, '-');		-- if private, all other tags ignored
	end
	
	table.insert (out, code:lower());
	if is_set (script) then
		script = script:lower():gsub ('^%a', string.upper);
		table.insert (out, script);
	end

	if is_set (region) then
		table.insert (out, region:upper());
	end
	
	if is_set (variant) then
		table.insert (out, variant:lower());
	end
	
	return table.concat (out, '-');
end


--[[--------------------------< G E T _ I E T F _ P A R T S >--------------------------------------------------

extracts and returns IETF language tag parts:
	primary language subtag (required) - 2 or 3 character IANA language code
	script subtag - four character IANA script code
	region subtag - two-letter or three digit IANA region code
	variant subtag - four digit or 5-8 alnum variant code; only one variant subtag supported
	private subtag - x- followed by 1-8 alnum private code; only supported with the primary language tag

in any one of these forms
	lang					lang-variant
	lang-script				lang-script-variant
	lang-region				lang-region-variant
	lang-script-region		lang-script-region-variant
	lang-x-private	
	
each of lang, script, region, variant, and private, when used, must be valid

Languages with both two- and three-character code synonyms are promoted to the two-character synonym because
the IANA registry file omits the synonymous three-character code; we cannot depend on browsers understanding
the synonymous three-character codes in the lang= attribute.

For {{lang-xx}} templates, the parameters |script=, |region=, and |variant= are supported (not supported in {{lang}}
because those parameters are superfluous to the IETF subtags in |code=)

returns six  values; all lower case.  Valid parts are returned as themselves; omitted parts are returned as empty strings, invalid
parts are returned as nil; the sixth returned item is an error message (if an error detected) or nil.

see http://www.rfc-editor.org/rfc/bcp/bcp47.txt section 2.1

]]

local function get_ietf_parts (source, args_script, args_region, args_variant)
	local code, script, region, variant, private;								-- ietf tag parts

	if not is_set (source) then
		return nil, nil, nil, nil, nil, 'missing language tag';
	end

	local pattern = {															-- table of tables holding acceptibe ietf tag patterns and short names of the ietf part captured by the pattern
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%d%d%d%d)$', 's', 'r', 'v'}, 				-- 1 -  ll-Ssss-RR-variant (where variant is 4 digits)
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%d%d%d%d)$', 's', 'r', 'v'},				-- 2 -  ll-Ssss-DDD-variant (where region is 3 digits; variant is 4 digits)
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'r', 'v'},		-- 3 -  ll-Ssss-RR-variant (where variant is 5-8 alnum characters)
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'r', 'v'},	-- 4 -  ll-Ssss-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
		
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d%d)$', 's', 'v'},						-- 5 -  ll-Ssss-variant (where variant is 4 digits)
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 's', 'v'},			-- 6 -  ll-Ssss-variant (where variant is 5-8 alnum characters)
		
		{'^(%a%a%a?)%-(%a%a)%-(%d%d%d%d)$', 'r', 'v'},							-- 7 -  ll-RR-variant (where variant is 4 digits)
		{'^(%a%a%a?)%-(%d%d%d)%-(%d%d%d%d)$', 'r', 'v'},						-- 8 -  ll-DDD-variant (where region is 3 digits; variant is 4 digits)
		{'^(%a%a%a?)%-(%a%a)%-(%w%w%w%w%w%w?%w?%w?)$', 'r', 'v'},				-- 9 -  ll-RR-variant (where variant is 5-8 alnum characters)
		{'^(%a%a%a?)%-(%d%d%d)%-(%w%w%w%w%w%w?%w?%w?)$', 'r', 'v'},				-- 10 - ll-DDD-variant (where region is 3 digits; variant is 5-8 alnum characters)
		
		{'^(%a%a%a?)%-(%d%d%d%d)$', 'v'},										-- 11 - ll-variant (where variant is 4 digits)
		{'^(%a%a%a?)%-(%w%w%w%w%w%w?%w?%w?)$', 'v'},							-- 12 - ll-variant (where variant is 5-8 alnum characters)
		
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$', 's', 'r'},							-- 13 - ll-Ssss-RR
		{'^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$', 's', 'r'},						-- 14 - ll-Ssss-DDD (region is 3 digits)
		
		{'^(%a%a%a?)%-(%a%a%a%a)$', 's'},										-- 15 - ll-Ssss
		
		{'^(%a%a%a?)%-(%a%a)$', 'r'},											-- 16 - ll-RR
		{'^(%a%a%a?)%-(%d%d%d)$', 'r'},											-- 17 - ll-DDD (region is 3 digits)
		
		{'^(%a%a%a?)$'},														-- 18 - ll
		
		{'^(%a%a%a?)%-x%-(%w%w?%w?%w?%w?%w?%w?%w?)$', 'p'},						-- 19 - ll-x-pppppppp (private is 1-8 alnum characters)
		}

	local t = {};																-- table of captures; serves as a translator between captured ietf tag parts and named variables

	for i, v in ipairs (pattern) do												-- spin through the pattern table looking for a match
		local c1, c2, c3, c4;													-- captures in the 'pattern' from the pattern table go here
	
		c1, c2, c3, c4 = source:match (pattern[i][1]);							-- one or more captures set if source matches pattern[i])
			if c1 then															-- c1 always set on match
				code = c1;														-- first capture is always code
				t = {
					[pattern[i][2] or 'x'] = c2,								-- fill the table of captures with the rest of the captures
					[pattern[i][3] or 'x'] = c3,								-- take index names from pattern table and assign sequential captures
					[pattern[i][4] or 'x'] = c4,								-- index name may be nil in pattern[i] table so "or 'x'" spoofs a name for this index in this table
					};
				script = t.s or '';												-- translate table contents to named variables;
				region = t.r or '';												-- absent table entries are nil so set named ietf parts to empty string for concatenation
				variant= t.v or '';
				private = t.p or '';
				break;															-- and done
			end
	end

	if not code then
		return nil, nil, nil, nil, nil, table.concat ({'unrecognized language tag: ', source});		-- don't know what we got but it is malformed
	end

	code = code:lower();														-- ensure that we use and return lower case version of this
	
	if not (override_table[code] or lang_table[code] or synonym_table[code] or lang_dep_table[code]) then
		return nil, nil, nil, nil, nil, table.concat ({'unrecognized language code: ', code});		-- invalid language code, don't know about the others (don't care?)
	end
	
	if synonym_table[code] then													-- if 639-2/639-2T code has a 639-1 synonym
		table.insert (maint_cats, table.concat ({'Lang and lang-xx code promoted to ISO 639-1|', code}));
		table.insert (maint_msgs, table.concat ({'code: ', code, ' promoted to code: ', synonym_table[code]}));
		code = synonym_table[code];												-- use the synonym
	end

	if is_set (script) then
		if is_set (args_script) then
			return code, nil, nil, nil, nil, 'redundant script tag';			-- both code with script and |script= not allowed
		end
	else
		script = args_script or '';												-- use args.script if provided
	end 

	if is_set (script) then
		script = script:lower();												-- ensure that we use and return lower case version of this
		if not script_table[script] then
			return code, nil, nil, nil, nil, table.concat ({'unrecognized script: ', script, ' for code: ', code});	-- language code ok, invalid script, don't know about the others (don't care?)
		end
	end
	if suppressed_table[script] then											-- ensure that code-script does not use a suppressed script
		if in_array (code, suppressed_table[script]) then
			return code, nil, nil, nil, nil, table.concat ({'script: ', script, ' not supported for code: ', code});	-- language code ok, script is suppressed for this code
		end
	end

	if is_set (region) then
		if is_set (args_region) then
			return code, nil, nil, nil, nil, 'redundant region tag';			-- both code with region and |region= not allowed
		end
	else
		region = args_region or '';												-- use args.region if provided
	end 

	if is_set (region) then
		region = region:lower();												-- ensure that we use and return lower case version of this
		if not region_table[region] then
			return code, script, nil, nil, nil, table.concat ({'unrecognized region: ', region, ' for code: ', code});
		end
	end
	
	if is_set (variant) then
		if is_set (args_variant) then
			return code, nil, nil, nil, nil, 'redundant variant tag';			-- both code with variant and |variant= not allowed
		end
	else
		variant = args_variant or '';											-- use args.variant if provided
	end 

	if is_set (variant) then
		variant = variant:lower();												-- ensure that we use and return lower case version of this
		if not variant_table[variant] then										-- make sure variant is valid
			return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant});
		end																		-- does this duplicate/replace tests in lang() and lang_xx()?
		if is_set (script) then													-- if script set it must be part of the 'prefix'
			if not in_array (table.concat ({code, '-', script}), variant_table[variant]['prefixes']) then
				return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-script pair: ', code, '-', script});
			end
		elseif is_set (region) then												-- if region set, there are some prefixes that require lang code and region (en-CA-newfound)
			if not in_array (code, variant_table[variant]['prefixes']) then		-- first see if lang code is all that's required (en-oxendict though en-GB-oxendict is preferred)
				if not in_array (table.concat ({code, '-', region}), variant_table[variant]['prefixes']) then		-- now try for lang code and region (en-CA-newfound)
					return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-region pair: ', code, '-', region});
				end
			end
		else																	-- cheap way to determine if there are prefixes; fonipa and others don't have prefixes; # operator always returns 0
			if variant_table[variant]['prefixes'][1] and not in_array (code, variant_table[variant]['prefixes']) then
				return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code: ', code});
			end
		end
	end

	if is_set (private) then
		private = private:lower();												-- ensure that we use and return lower case version of this
		if not override_table[table.concat ({code, '-x-', private})] then		-- make sure private tag is valid; note that index 
			return code, script, region, nil, nil, table.concat ({'unrecognized private tag: ', private});
		end
	end
	return code, script, region, variant, private, nil;							-- return the good bits; make sure that msg is nil
end


--[[--------------------------< M A K E _ E R R O R _ M S G >--------------------------------------------------

assembles an error message from template name, message text, help link, and error category.

]]

local function make_error_msg (msg, args, template)
	local out = {};
	local category;
	
	if 'Transl' == template then
		category = 'Transl';
	else
		category = 'Lang and lang-xx'
	end
	
	table.insert (out, table.concat ({'[', args.text or 'undefined', '] '}));	-- for error messages output args.text if available
	table.insert (out, table.concat ({'<span style=\"color:#d33\">Error: {{', template, '}}: '}));
	table.insert (out, msg);
	table.insert (out, table.concat ({' ([[:Category:', category, ' template errors|help]])'}));
	table.insert (out, '</span>');
	
	if (0 == namespace or 10 == namespace) and not args.nocat then				-- categorize in article space (and template space to take care of broken usages)
		table.insert (out, table.concat ({''}));
	end

	return table.concat (out);
end
	

--[=[-------------------------< M A K E _ W I K I L I N K >----------------------------------------------------

Makes a wikilink; when both link and display text is provided, returns a wikilink in the form [[L|D]]; if only
link is provided, returns a wikilink in the form [[L]]; if neither are provided or link is omitted, returns an
empty string.

]=]

local function make_wikilink (link, display)
	if is_set (link) then
		if is_set (display) then
			return table.concat ({'[[', link, '|', display, ']]'});
		else
			return table.concat ({'[[', link, ']]'});
		end
	else
		return '';
	end
end


--[[--------------------------< D I V _ M A R K U P _ A D D >--------------------------------------------------

adds <i> and </i> tags to list-item text or to implied <p>..</p> text.  mixed not supported

]]

local function div_markup_add (text, style)
local implied_p = {};

	if text:find ('^\n[%*:;#]') then											-- look for list markup; list markup must begin at start of text
		if 'italic' == style then
			return text:gsub ('(\n[%*:;#]+)([^\n]+)', '%1<i>%2</i>');			-- insert italic markup at each list item
		else
			return text;
		end
	end

	if text:find ('\n+') then													-- look for any number of \n characters in text
		text = text:gsub ('([^\n])\n([^\n])', '%1 %2');							-- replace single newline characters with a space character which mimics mediawiki

		if 'italic' == style then
			text = text:gsub('[^\n]+', '<p><i>%1</i></p>');						-- insert p and italic markup tags at each impled p (two or more consecutive '\n\n' sequences)
		else
			text = text:gsub ('[^\n]+', '<p>%1</p>');							-- insert p markup at each impled p
			text = text:gsub ('\n', '');										-- strip newline characters
		end
	end

	return text;
end


--[[--------------------------< T I T L E _ W R A P P E R _ M A K E >------------------------------------------

makes a <span title="<title text>"><content_text></span> or <div title="<title text>"><content_text></div> where
<title text> is in the tool-tip in the wiki's local language and <content_text> is non-local-language text in
html markup.  This because the lang= attibute applies to the content of its enclosing tag.

<tag> holds a string 'div' or 'span' used to choose the correct wrapping tag
]]

local function title_wrapper_make (title_text, content_text, tag)
	local wrapper_t = {};
	
	table.insert (wrapper_t, table.concat ({'<', tag}));						-- open opening wrapper tag
	table.insert (wrapper_t, ' title=\"');										-- begin title attribute
	table.insert (wrapper_t, title_text);										-- add <title_text>
	table.insert (wrapper_t, '\">');											-- end title attribute and close opening wrapper tag
	table.insert (wrapper_t, content_text);										-- add <content_text>
	table.insert (wrapper_t, table.concat ({'</', tag, '>'}));					-- add closing wrapper tag

	return table.concat (wrapper_t);											-- make a big string and done
end


--[[--------------------------< M A K E _ T E X T _ H T M L >--------------------------------------------------

Add the html markup to text according to the type of content that it is: <span> or <i> tags for inline content or
<div> tags for block content

The lang= attribute also applies to the content of the tag where it is placed so this is wrong because 'Spanish
language text' is English:
	<i lang="es" title="Spanish language text">casa</i>
should be:
	<span title="Spanish language text"><i lang="es">casa</i></span>
or for <div>...</div>:
	<div title="Spanish language text"><div lang="es"><spanish-language-text></div></div>

]]

local function make_text_html (code, text, tag, rtl, style, size, language)
	local html_t = {};
	local style_added = '';
	local wrapper_tag = tag;													-- <tag> gets modified so save a copy for use when/if we create a wrapper span or div

	if text:match ('^%*') then
		table.insert (html_t, '&#42;');											-- move proto language text prefix outside of italic markup if any; use numeric entity because plain splat confuses MediaWiki
		text = text:gsub ('^%*', '');											-- remove the splat from the text
	end

	if 'span' == tag then														-- default html tag for inline content
		if 'italic' == style then												-- but if italic
			tag = 'i';															-- change to <i> tags
		end
	else																		-- must be div so go
		text = div_markup_add (text, style);									-- handle implied <p>, implied <p> with <i>, and list markup (*;:#) with <i>
	end

	table.insert (html_t, table.concat ({'<', tag}));							-- open the <i>, <span>, or <div> html tag
	code = code:gsub ('%-x%-.*', '');											-- strip private use subtag from code tag because meaningless outside of wikipedia
	table.insert (html_t, table.concat ({' lang="', code, '\"'}));				-- add language attribute

	if (rtl or unicode.is_rtl(text)) and ('ltr' == this_wiki_lang_dir) then		-- text is right-to-left on a left-to-right wiki
		table.insert (html_t, ' dir="rtl"');									-- add direction attribute for right-to-left languages
	elseif not (rtl or unicode.is_rtl(text)) and ('rtl' == this_wiki_lang_dir) then	-- text is left-to-right on a right-to-left wiki
		table.insert (html_t, ' dir="ltr"');									-- add direction attribute for left-to-right languages
	end

	if 'normal' == style then													-- when |italic=no
		table.insert (html_t, ' style=\"font-style: normal;');					-- override external markup, if any
		style_added = '\"';														-- remember that style attribute added and is not yet closed
	end

	if is_set (size) then														-- when |size=<something>
		if is_set (style_added) then
			table.insert (html_t, table.concat ({' font-size: ', size, ';'}));	-- add when style attribute already inserted
		else
			table.insert (html_t, table.concat ({' style=\"font-size: ', size, ';'}));	-- create style attribute
			style_added = '\"';													-- remember that style attribute added and is not yet closed
		end
	end

	table.insert (html_t, table.concat ({style_added, '>'}));					-- close the opening html tag
	table.insert (html_t, text);												-- insert the text

	table.insert (html_t, table.concat ({'</', tag, '>'}));						-- close the 'text' <i>, <span>, or <div> html tag

	if is_set (language) then													-- create a <title_text> string for the title= attribute in a wrapper span or div
		local title_text;
			if 'zxx' == code then												-- special case for this tag 'no linguistic content'
				title_text = table.concat ({language, ' text'});				-- not a language so don't use 'language' in title text
			elseif mw.ustring.find (language, 'languages', 1, true) then
				title_text = table.concat ({language, ' collective text'});		-- for collective languages
			else
				title_text = table.concat ({language, '-language text'});		-- for individual languages
			end

		return title_wrapper_make (title_text, table.concat (html_t), wrapper_tag);
	else
		return table.concat (html_t);
	end
end


--[=[-------------------------< M A K E _ C A T E G O R Y >----------------------------------------------------

For individual language, <language>, returns:
	

for English:
	
	
for ISO 639-2 collective languages (and for 639-1 bh):
	

]=]

local function make_category (code, language_name, nocat, name_get)
	local cat = {};
	local retval;

	if ((0 ~= namespace) or nocat) and not name_get then						-- only categorize in article space
		return '';																-- return empty string for concatenation
	end

	if mw.ustring.find (language_name, 'languages', 1, true) then
		return table.concat ({''});
	end
		
	table.insert (cat, '[[Category:Articles containing ');

	if this_wiki_lang_tag == code then
		table.insert (cat, 'explicitly cited ' .. language_name);				-- unique category name for the local language
	else
		table.insert (cat, language_name);
	end
	
	table.insert (cat, '-language text]]');

	return table.concat (cat);
end


--[[--------------------------< M A K E _ T R A N S L I T >----------------------------------------------------

return translit <i lang=xx-Latn>...</i> where xx is the language code; else return empty string

The value |script= is not used in {{transl}} for this purpose; instead it uses |code.  Because language scripts
are listed in the {{transl}} switches they are included in the data tables.  The script parameter is introduced
at {{Language with name and transliteration}}.  If |script= is set, this function uses it in preference to code.

To avoid confusion, in this module and the templates that use it, the transliteration script parameter is renamed
to be |translit-script= (in this function, tscript)

This function is used by both lang_xx() and transl()
	lang_xx() always provides code, language_name, and translit; may provide tscript; never provides style
	transl() always provides language_name, translit, and one of code or tscript, never both; always provides style

For {{transl}}, style only applies when a language code is provided
]]

local function make_translit (code, language_name, translit, std, tscript, style)
	local title;
	local out_t = {};
	local title_t = lang_data.translit_title_table;								-- table of transliteration standards and the language codes and scripts that apply to those standards
	local title_text = '';														-- tool tip text for title= attribute
	
	std = std and std:lower();													-- lower case for table indexing
	
	if not is_set (std) and not is_set (tscript) then							-- when neither standard nor script specified
		title_text = language_name;												-- write a generic tool tip
		if not mw.ustring.find (language_name, 'languages', 1, true) then		-- collective language names (plural 'languages' is part of the name)
			title_text = title_text .. '-language';								-- skip this text (individual and macro languages only)
		end
		title_text = title_text .. ' romanization';								-- finish the tool tip; use romanization when neither script nor standard supplied

	elseif is_set (std) and is_set (tscript) then								-- when both are specified
		if title_t[std] then													-- and if standard is legitimate
			if title_t[std][tscript] then										-- and if script for that standard is legitimate
				if script_table[tscript] then
					title_text = title_text .. table.concat ({title_t[std][tscript:lower()], ' (', script_table[tscript], ' script) transliteration'});	-- add the appropriate text to the tool tip
				else
					title_text = title_text .. title_t[std]['default'];			-- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?
				end
			else
				title_text = title_text .. title_t[std]['default'];				-- use the default if script not in std table; TODO: maint cat? error message because script not found for this standard?
			end
		else
			return '';															-- invalid standard, setup for error message
		end

	elseif is_set (std) then													-- translit-script not set, use language code
		if not title_t[std] then return ''; end									-- invalid standard, setup for error message
		
		if title_t[std][code] then												-- if language code is in the table (transl may not provide a language code)
			title_text = title_text .. table.concat ({title_t[std][code:lower()], ' (', language_name, ' language) transliteration'});	-- add the appropriate text to the tool tip
		else																	-- code doesn't match
			title_text = title_text .. title_t[std]['default'];					-- so use the standard's default
		end

	else																		-- here if translit-script set but translit-std not set
		if title_t['no_std'][tscript] then
			title_text = title_text .. title_t['no_std'][tscript];				-- use translit-script if set
		elseif title_t['no_std'][code] then
			title_text = title_text .. title_t['no_std'][code];					-- use language code
		else
			if is_set (tscript) then
				title_text = title_text .. table.concat ({language_name, '-script transliteration'});	-- write a script tool tip
			elseif is_set (code) then
				if not mw.ustring.find (language_name, 'languages', 1, true) then	-- collective language names (plural 'languages' is part of the name)
					title_text = title_text .. '-language';						-- skip this text (individual and macro languages only)
				end
				title_text = title_text .. ' transliteration';					-- finish the tool tip
			else
				title_text = title_text .. ' transliteration';					-- generic tool tip (can we ever get here?)
			end
		end
	end

	local close_tag;
	if is_set (code) then														-- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})
		if not style then														-- nil for the default italic style
			table.insert (out_t, "<i lang=\"");									-- so use <i> tag
			close_tag = '</i>';													-- tag to be used when closing
		else
			table.insert (out_t, table.concat ({'<span style=\"font-style: ', style, '\" lang=\"'}));	-- non-standard style, construct a span tag for it
			close_tag = '</span>';												-- tag to be used when closing
		end
		table.insert (out_t, code);
		table.insert (out_t, "-Latn\">");										-- transliterations are always Latin script
	else
		table.insert (out_t, "<span>");											-- when no language code: no lang= attribute, not italic ({{transl}} only)
		close_tag = '</span>';
	end

	table.insert (out_t, translit);												-- add the translit text
	table.insert (out_t, close_tag);											-- and add the appropriate </i> or </span>

	if '' == title_text then													-- when there is no need for a tool-tip
		return table.concat (out_t);											-- make a string and done
	else
		return title_wrapper_make (title_text, table.concat (out_t), 'span');	-- wrap with a tool-tip span and don
	end
end


--[[--------------------------< V A L I D A T E _ T E X T >----------------------------------------------------

This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
error message.  The tests are for empty or missing text and for improper or disallowed use of apostrophe markup.

Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
either as ''itself'' or as '''''bold italic''''' unless |italic=unset or |italic=invert.

]]

local function validate_text (template, args)
	if not is_set (args.text) then
		return make_error_msg ('no text', args, template);
	end

	if args.text:find ("%f[\']\'\'\'\'%f[^\']") or args.text:find ("\'\'\'\'\'[\']+") then	-- because we're looking, look for 4 appostrophes or 6+ appostrophes
		return make_error_msg ('text has malformed markup', args, template);
	end

	local style = args.italic;

	if ('unset' ~= style) and ('invert' ~=style) then
		if args.text:find ("%f[\']\'\'%f[^\']") or args.text:find ("%f[\']\'\'\'\'\'%f[^\']") then	-- italic but not bold, or bold italic
			return make_error_msg ('text has italic markup', args, template);
		end
	end
end


--[[--------------------------< R E N D E R _ M A I N T >------------------------------------------------------

render mainenance messages and categories

]]

local function render_maint(nocat)
	local maint = {};
	
	if 0 < #maint_msgs then														-- when there are maintenance messages
		table.insert (maint, table.concat ({'<span class="lang-comment" style="font-style: normal; display: none; color: #33aa33; margin-left: 0.3em;">'}));	-- opening <span> tag
		for _, msg in ipairs (maint_msgs) do
			table.insert (maint, table.concat ({msg, ' '}));					-- add message strings
		end
		table.insert (maint, '</span>');										-- close the span
	end
	
	if (0 < #maint_cats) and (0 == namespace) and not nocat then				-- when there are maintenance categories; article namespace only
		for _, cat in ipairs (maint_cats) do
			table.insert (maint, table.concat ({''}));	-- format and add the categories
		end
	end
	
	return table.concat (maint);
end


--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------

for proto languages, text is prefixed with a splat.  We do that here as a flag for make_text_html() so that a splat
will be rendered outside of italic markup (if used).  If the first character in text here is already a splat, we
do nothing

proto_param is boolean or nil; true adds splat prefix regardless of language name; false removes and / or inhibits
regardless of language name; nil does nothing; presumes that the value in text is correct but removes extra splac

]]

local function proto_prefix (text, language_name, proto_param)
	if false == proto_param then												-- when forced by |proto=no
		return text:gsub ('^%**', '');											-- return text without splat prefix regardless of language name or existing splat prefix in text
	elseif (language_name:find ('^Proto%-') or (true == proto_param)) then		-- language is a proto or forced by |proto=yes
		return text:gsub ('^%**', '*');											-- prefix proto-language text with a splat; also removes duplicate prefixing splats
	end
	
	return text:gsub ('^%*+', '*');												-- return text unmolested except multiple splats reduced to one splat
end


--[[--------------------------< H A S _ P O E M _ T A G >------------------------------------------------------

looks for a poem strip marker in text; returns true when found; false else

auto-italic detection disabled when text has poem stripmarker because it is not possible for this code to know
the content that will replace the stripmarker.

]]

local function has_poem_tag (text)
	return text:find ('\127[^\127]*UNIQ%-%-poem%-[%a%d]+%-QINU[^\127]*\127') and true or false;
end


--[[--------------------------< H T M L _ T A G _ S E L E C T >------------------------------------------------

Inspects content of and selectively trims text.  Returns text and the name of an appropriate html tag for text.

If text contains:
	\n\n	text has implied <p>..</p> tags - trim leading and trailing whitespace and return 
If text begins with list markup:
	\n*		unordered
	\n;		definition
	\n:		definition
	\n#		ordered
trim all leading whitespace except  \n and trim all trailing whitespace

If text contains <poem>...</poem> stripmarker, return text unmodified and choose <div>..</div> tags because 
the stripmarker is replaced with text wrapped in <div>..</div> tags.

]]

local function html_tag_select (text)
	local tag;
	
	if has_poem_tag (text) then													-- contains poem stripmarker (we can't know the content of that)
		tag = 'div';															-- poem replacement is in div tags so lang must use div tags
	elseif mw.text.trim (text):find ('\n\n+') then								-- contains implied p tags
		text = mw.text.trim (text);												-- trim leading and trailing whitespace characters
		tag = 'div';															-- must be div because span may not contain p tags (added later by MediaWiki); poem replacement is in div tags
	elseif text:find ('\n[%*:;%#]') then										-- if text has list markup
		text = text:gsub ('^[\t\r\f ]*', ''):gsub ('%s*$', '');					-- trim all whitespace except leading newline character '\n'
		tag = 'div';															-- must be div because span may not contain ul, dd, dl, ol tags (added later by MediaWiki)
	else
		text = mw.text.trim (text);												-- plain text
		tag = 'span';															-- so span is fine
	end
	
	return text, tag;
end


--[[--------------------------< V A L I D A T E _ P R O T O >--------------------------------------------------

validates value assigned to |proto=; permitted values are yes and no; yes returns as true, no returns as false,
empty string (or parameter omitted) returns as nil; any other value returns as nil with a second return value of
true indicating that some other value has been assigned to |proto=

]]

local function validate_proto (proto_param)
	if 'yes' == proto_param then
		return true;
	elseif 'no' == proto_param then
		return false;
	elseif is_set (proto_param) then
		return nil, true;														-- |proto= something other than 'yes' or 'no'
	else
		return nil;																-- missing or empty
	end
end


--[[--------------------------< L A N G U A G E _ N A M E _ G E T >--------------------------------------------

common function to return language name from the data set according to IETF tag

returns language name if found in data tables; nil else

]]

local function language_name_get (ietf, code, cat)
	ietf = ietf:lower();														-- ietf:lower() because format_ietf_tag() returns mixed case
	
	local name;																	-- remains nil if not found
	
	if override_table[ietf] then												-- look for whole IETF tag in override table
		name = override_table[ietf];
	elseif override_table[code] then											-- not there so try basic language tag
		name = override_table[code];
	elseif lang_table[code] then												-- shift to iana active tag/name table
		name = lang_table[code];
	elseif lang_dep_table[code] then											-- try the iana deprecated tag/name table
		name = lang_dep_table[code];
	end

	if lang_dep_table[code] and cat then										-- because deprecated code may have been overridden to en.wiki preferred name
		table.insert (maint_cats, table.concat ({'Lang and lang-xx using deprecated ISO 639 codes|', code}));
		table.insert (maint_msgs, table.concat ({'code: ', code, ' is deprecated'}));
	end

	return name;																-- name from data tables or nil
end


--[[--------------------------< _ L A N G >--------------------------------------------------------------------

entry point for {{lang}}

there should be no reason to set parameters in the {{lang}} {{#invoke:}}
	<includeonly>{{#invoke:lang|lang}}</includeonly>

parameters are received from the template's frame (parent frame)

]]

local function _lang (args)
	local out = {};
	local language_name;														-- used to make category names
	local category_name;														-- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
	local subtags = {};															-- IETF subtags script, region, variant, and private
	local code;																	-- the language code
	local msg;																	-- for error messages
	local tag = 'span';															-- initial value for make_text_html()
	local template = args.template or 'Lang';

	maint_cats = {};															-- initialize because when this module required into another module, these only declared once so only initialzed once
	maint_msgs = {};
	
	validate_cat_args (args);													-- determine if categorization should be suppressed

	if args[1] and args.code then
		return make_error_msg ('conflicting: {{{1}}} and |code=', args, template);
	else
		args.code = args[1] or args.code;										-- prefer args.code
	end

	if args[2] and args.text then
		return make_error_msg ('conflicting: {{{2}}} and |text=', args, template);
	else
		args.text = args[2] or args.text;										-- prefer args.text
	end
	
	msg = validate_text (template, args);										-- ensure that |text= is set
	if is_set (msg) then														-- msg is an already-formatted error message
		return msg;
	end
	
	args.text, tag = html_tag_select (args.text);								-- inspects text; returns appropriate html tag with text trimmed accordingly

	args.rtl = args.rtl == 'yes';												-- convert to boolean: 'yes' -> true, other values -> false

	args.proto, msg = validate_proto (args.proto);								-- return boolean, or nil, or nil and error message flag
	if msg then
		return make_error_msg (table.concat ({'invalid |proto=: ', args.proto}), args, template);
	end

	code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code);	-- |script=, |region=, |variant= not supported because they should be part of args.code ({{{1}}} in {{lang}})

	if msg then
		return make_error_msg ( msg, args, template);
	end

	args.italic, msg = validate_italic (args);
	if msg then
		return make_error_msg (msg, args, template);
	end

	if nil == args.italic then													-- nil when |italic= absent or not set or |italic=default; args.italic controls
		if ('latn' == subtags.script) or								 		-- script is latn
			(this_wiki_lang_tag ~= code and not is_set (subtags.script) and not has_poem_tag (args.text) and unicode.is_Latin (args.text)) then -- text not this wiki's language, no script specified and not in poem markup but is wholly latn script (auto-italics)
				args.italic = 'italic';											-- DEFAULT for {{lang}} templates is upright; but if latn script set for font-style:italic
		else
			args.italic = 'inherit';											-- italic not set; script not latn; inherit current style
		end
	end
	
	if is_set (subtags.script) then												-- if script set, override rtl setting
		if in_array (subtags.script, lang_data.rtl_scripts) then
			args.rtl = true;													-- script is an rtl script
		else
			args.rtl = false;													-- script is not an rtl script
		end
	end

	args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private);	-- format to recommended subtag styles
	language_name = language_name_get (args.code, code, true);					-- get language name; try ietf tag first, then code w/o variant then code w/ variant

	if 'invert' == args.italic and 'span' == tag then							-- invert only supported for in-line content
		args.text = invert_italics (args.text)
	end

	args.text = proto_prefix (args.text, language_name, args.proto);			-- prefix proto-language text with a splat

	table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, language_name));
	table.insert (out, make_category (code, language_name, args.nocat));
	table.insert (out, render_maint(args.nocat));								-- maintenance messages and categories

	return table.concat (out);													-- put it all together and done
end


--[[--------------------------< L A N G >----------------------------------------------------------------------

entry point for {{lang}}

there should be no reason to set parameters in the {{lang}} {{#invoke:}}
	<includeonly>{{#invoke:lang|lang}}</includeonly>

parameters are received from the template's frame (parent frame)

]]

local function lang (frame)
	local args = getArgs (frame, {												-- this code so that we can detect and handle wiki list markup in text
		valueFunc = function (key, value)
			if 2 == key or 'text' == key then									-- the 'text' parameter; do not trim wite space
				return value;													-- return untrimmed 'text'
			elseif value then													-- all other values: if the value is not nil
				value = mw.text.trim (value);									-- trim whitespace
				if '' ~= value then												-- empty string when value was only whitespace
					return value;
				end
			end
			return nil;															-- value was empty or contained only whitespace
		end																		-- end of valueFunc
		});

	return _lang (args);
end


--[[--------------------------< T R A N S L A T I O N _ M A K E >----------------------------------------------

stand-alone function to create literal translation of main text

Also used by {{lang-x2}}

]]

local function translation_make (args_t)
	local translation_t = {', '};												-- initialize output

	if 'none' ~= args_t.label then												-- if we want a label
		table.insert (translation_t, '<small>');								-- open the <small> html tag
		if 'no' == args_t.link then
			table.insert (translation_t, '<abbr title="literal translation">lit.</abbr>');	-- unlinked form
		else
			table.insert (translation_t, make_wikilink ('Literal translation', 'lit.'));	-- linked form
		end
		table.insert (translation_t, "&thinsp;</small>");						-- close the <small> html tag
	end
	table.insert (translation_t, table.concat ({'&#39;', args_t.translation, '&#39;'}));	-- use html entities to avoid wiki markup confusion
	return table.concat (translation_t);										-- make a big string and done
end


--[[--------------------------< L A N G _ X X >----------------------------------------------------------------

For the {{lang-xx}} templates, the only parameter required to be set in the template is the language code.  All
other parameters can, usually should, be written in the template call.  For {{lang-xx}} templates for languages
that can have multiple writing systems, it may be appropriate to set |script= as well.

For each {{lang-xx}} template choose the appropriate entry-point function so that this function knows the default
styling that should be applied to text.

For normal, upright style:
	<includeonly>{{#invoke:lang|lang_xx_inherit|code=xx}}</includeonly>
For italic style:
	<includeonly>{{#invoke:lang|lang_xx_italic|code=xx}}</includeonly>

All other parameters should be received from the template's frame (parent frame)

Supported parameters are:
	|code = (required) the IANA language code
	|script = IANA script code; especially for use with languages that use multiple writing systems
	|region = IANA region code
	|variant = IANA variant code
	|text = (required) the displayed text in language specified by code
	|link = boolean false ('no') does not link code-spcified language name to associated language article
	|rtl = boolean true ('yes') identifies the language specified by code as a right-to-left language
	|nocat = boolean true ('yes') inhibits normal categorization; error categories are not affected
	|cat = boolian false ('no') opposite form of |nocat=
	|italic = boolean true ('yes') renders displayed text in italic font; boolean false ('no') renders displayed text in normal font; not set renders according to initial_style_state
	|lit = text that is a literal translation of text
	|label = 'none' to suppress all labeling (language name, 'translit.', 'lit.')
				any other text replaces language-name label - automatic wikilinking disabled
	
	for those {{lang-xx}} templates that support transliteration (those templates where |text= is not entirely latn script):
	|translit = text that is a transliteration of text
	|translit-std = the standard that applies to the transliteration
	|translit-script = ISO 15924 script name; falls back to code

For {{lang-xx}}, the positional parameters are:
	{{{1}}}	text
	{{{2}}}	transliterated text
	{{{3}}}	literal translation text
no other positional parameters are allowed

]]

local function _lang_xx (args)
	local out = {};
	local language_name;														-- used to make display text, article links
	local category_name;														-- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
	local subtags = {};															-- IETF subtags script, region, and variant
	local code;																	-- the language code

	local translit_script_name;													-- name associated with IANA (ISO 15924) script code
	local translit;
	local translit_title;
	local msg;																	-- for error messages
	local tag = 'span';															-- initial value for make_text_html()
	local template = args.template or 'Lang-xx';

	maint_cats = {};															-- initialize because when this module required into another module, these only declared once so only initialzed once
	maint_msgs = {};
	
	if args[1] and args.text then
		return make_error_msg ('conflicting: {{{1}}} and |text=', args, template);
	else
		args.text = args[1] or args.text;										-- prefer args.text
	end
	
	msg = validate_text (template, args);										-- ensure that |text= is set, does not contain italic markup and is protected from improper bolding
	if is_set (msg) then
		return msg;
	end

	args.text, tag = html_tag_select (args.text);								-- inspects text; returns appropriate html tag with text trimmed accordingly

	if args[2] and args.translit then
		return make_error_msg ('conflicting: {{{2}}} and |translit=', args, template);
	else
		args.translit = args[2] or args.translit								-- prefer args.translit
	end
	
	if args[3] and (args.translation or args.lit) then
		return make_error_msg ('conflicting: {{{3}}} and |lit= or |translation=', args, template);
	elseif args.translation and args.lit then
		return make_error_msg ('conflicting: |lit= and |translation=', args, template);
	else
		args.translation = args[3] or args.translation or args.lit;				-- prefer args.translation
	end

	if args.links and args.link then
		return make_error_msg ('conflicting: |links= and |link=', args, template);
	else
		args.link = args.link or args.links;									-- prefer args.link
	end

	validate_cat_args (args);													-- determine if categorization should be suppressed

	args.rtl = args.rtl == 'yes';												-- convert to boolean: 'yes' -> true, other values -> false

	code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code, args.script, args.region, args.variant);	-- private omitted because private

	if msg then																	-- if an error detected then there is an error message
		return make_error_msg (msg, args, template);
	end
	
	args.italic, msg = validate_italic (args);
	if msg then
		return make_error_msg (msg, args, template);
	end

	if nil == args.italic then													-- args.italic controls
		if is_set (subtags.script) then
			if 'latn' == subtags.script then
				args.italic = 'italic';											-- |script=Latn; set for font-style:italic
			else
				args.italic = initial_style_state;								-- italic not set; script is not latn; set for font-style:<initial_style_state>
			end
		else
			args.italic = initial_style_state;									-- here when |italic= and |script= not set; set for font-style:<initial_style_state>
		end
	end
	
	if is_set (subtags.script) then												-- if script set override rtl setting
		if in_array (subtags.script, lang_data.rtl_scripts) then
			args.rtl = true;													-- script is an rtl script
		else
			args.rtl = false;													-- script is not an rtl script
		end
	end

	args.proto, msg = validate_proto (args.proto);								-- return boolean, or nil, or nil and error message flag
	if msg then
		return make_error_msg (table.concat ({'invalid |proto=: ', args.proto}), args, template);
	end

	args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private);	-- format to recommended subtag styles
	language_name = language_name_get (args.code, code, true);					-- get language name; try ietf tag first, then code w/o variant then code w/ variant

	category_name = language_name;												-- category names retain IANA parenthetical diambiguators (if any)
	language_name = language_name:gsub ('%s+%b()', '');							-- remove IANA parenthetical disambiguators or qualifiers from names that have them

	if args.label then
		if 'none' ~= args.label then
			table.insert (out, table.concat ({args.label, ': '}));				-- custom label
		end
	else
		if 'no' == args.link then
			table.insert (out, language_name);									-- language name without wikilink
		else
			if mw.ustring.find (language_name, 'languages', 1, true) then
				table.insert (out, make_wikilink (language_name));				-- collective language name uses simple wikilink
			elseif lang_data.article_name[code] then
				table.insert (out, make_wikilink (lang_data.article_name[code], language_name));	-- language name with wikilink from override data
			else
				table.insert (out, make_wikilink (language_name .. ' language', language_name));	-- language name with wikilink
			end
		end
		table.insert (out, ': ');												-- separator
	end

	if 'invert' == args.italic then
		args.text = invert_italics (args.text)
	end
	
	args.text = proto_prefix (args.text, language_name, args.proto);			-- prefix proto-language text with a splat

	table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, ('none' == args.label) and language_name or nil))

	if is_set (args.translit) and not unicode.is_Latin (args.text) then			-- transliteration (not supported in {{lang}}); not supported when args.text is wholly latn text (this is an imperfect test)
		table.insert (out, ', ');												-- comma to separate text from translit
		if 'none' ~= args.label then
			table.insert (out, '<small>');
			if script_table[args['translit-script']] then						-- when |translit-script= is set, try to use the script's name
				translit_script_name = script_table[args['translit-script']];
			else
				translit_script_name = language_name;							-- fall back on language name
			end
			translit_title = mw.title.makeTitle (0, table.concat ({'Romanization of ', translit_script_name}));		-- make a title object
			if translit_title.exists and ('no' ~= args.link) then
				table.insert (out, make_wikilink ('Romanization of ' .. translit_script_name or language_name, 'romanized') .. ':');	-- make a wikilink if there is an article to link to
			else
				table.insert (out, 'romanized:');								-- else plain text
			end
			table.insert (out, '&nbsp;</small>');								-- close the small tag
		end
		
		translit = make_translit (args.code, language_name, args.translit, args['translit-std'], args['translit-script'])
		if is_set (translit) then
			table.insert (out, translit);
		else
			return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args, template);
		end
	end
	
	if is_set (args.translation) then											-- translation (not supported in {{lang}})
		table.insert (out, translation_make (args));
	end
	
	table.insert (out, make_category (code, category_name, args.nocat));
	table.insert (out, render_maint(args.nocat));								-- maintenance messages and categories

	return table.concat (out);													-- put it all together and done
end


--[[--------------------------< L A N G _ X X _ A R G S _ G E T >----------------------------------------------

common function to get args table from {{lang-??}} templates

returns table of args

]]

local function lang_xx_args_get (frame)
	local args = getArgs(frame,
		{
		parentFirst= true,														-- parameters in the template override parameters set in the {{#invoke:}}
		valueFunc = function (key, value)
			if 1 == key then													-- the 'text' parameter; do not trim wite space
				return value;													-- return untrimmed 'text'
			elseif value then													-- all other values: if the value is not nil
				value = mw.text.trim (value);									-- trim whitespace
				if '' ~= value then												-- empty string when value was only whitespace
					return value;
				end
			end
			return nil;															-- value was empty or contained only whitespace
		end																		-- end of valueFunc
	});

	return args;
end


--[[--------------------------< L A N G _ X X _ I T A L I C >--------------------------------------------------

Entry point for those {{lang-xx}} templates that call lang_xx_italic().  Sets the initial style state to italic.

]]

local function lang_xx_italic (frame)
	local args = lang_xx_args_get (frame);
	
	initial_style_state = 'italic';
	return _lang_xx (args);
end


--[[--------------------------< _ L A N G _ X X _ I T A L I C >------------------------------------------------

Entry point ffrom another module.  Sets the initial style state to italic.

]]

local function _lang_xx_italic (args)
	initial_style_state = 'italic';
	return _lang_xx (args);
end


--[[--------------------------< L A N G _ X X _ I N H E R I T >------------------------------------------------

Entry point for those {{lang-xx}} templates that call lang_xx_inherit().  Sets the initial style state to inherit.

]]

local function lang_xx_inherit (frame)
	local args = lang_xx_args_get (frame);

	initial_style_state = 'inherit';
	return _lang_xx (args);
end


--[[--------------------------< _ L A N G _ X X _ I N H E R I T >----------------------------------------------

Entry point from another module.  Sets the initial style state to inherit.

]]

local function _lang_xx_inherit (args)
	initial_style_state = 'inherit';
	return _lang_xx (args);
end


--[[--------------------------< _ I S _ I E T F _ T A G >------------------------------------------------------

Returns true when a language name associated with IETF language tag exists; nil else.  IETF language tag must be valid.

All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.

Module entry point from another module

]]

local function _is_ietf_tag (tag)												-- entry point when this module is require()d into another
	local c, s, r, v, p, err;													-- code, script, region, variant, private, error message
	c, s, r, v, p, err = get_ietf_parts (tag);									-- disassemble tag into constituent part and validate

	return ((c and not err) and true) or nil;									-- return true when code portion has a value without error message; nil else
end


--[[--------------------------< I S _ I E T F _ T A G >--------------------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function is_ietf_tag (frame)
	return _is_ietf_tag (getArgs (frame)[1]);									-- args[1] is the ietf language tag to be tested; also get parent frame
end


--[[--------------------------< I S _ I E T F _ T A G _ F R A M E >--------------------------------------------

Module entry point from an {{#invoke:}}; same as is_ietf_tag() except does not get parameters from the parent
(template) frame.  This function not useful when called by {{lang|fn=is_ietf_tag_frame|<tag>}} because <tag>
is in the parent frame.

]]

local function is_ietf_tag_frame (frame)
	return _is_ietf_tag (getArgs (frame, {frameOnly = true,})[1]);				-- args[1] is the ietf language tag to be tested; do not get parent frame
end


--[[--------------------------< _ N A M E _ F R O M _ T A G >--------------------------------------------------

Returns language name associated with IETF language tag if valid; error message else.

All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.

Set invoke's |link= parameter to yes to get wikilinked version of the language name.

Module entry point from another module

]]

local function _name_from_tag (args)
	local subtags = {};															-- IETF subtags script, region, variant, and private
	local raw_code = args[1];													-- save a copy of the input IETF subtag
	local link = 'yes' == args['link'];											-- make a boolean
	local label = args.label;
	local code;																	-- the language code
	local msg;																	-- gets an error message if IETF language tag is malformed or invalid
	local language_name = '';
	
	code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (raw_code);
	if msg then
		local template = (args['template'] and table.concat ({'{{', args['template'], '}}: '})) or '';	-- make template name (if provided by the template)
		return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
	end

	raw_code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private);	-- format to recommended subtag styles; private omitted because private
	language_name = language_name_get (raw_code, code);							-- get language name; try ietf tag first, then code w/o variant then code w/ variant

	language_name = language_name:gsub ('%s+%b()', '');							-- remove IANA parenthetical disambiguators or qualifiers from names that have them

	if link then																-- when |link=yes, wikilink the language name
		if mw.ustring.find (language_name, 'languages', 1, true) then
			language_name = make_wikilink (language_name, label);				-- collective language name uses simple wikilink
		elseif lang_data.article_name[code] then
			language_name = make_wikilink (lang_data.article_name[code], label or language_name);	-- language name with wikilink from override data
		else
			language_name = make_wikilink (language_name .. ' language', label or language_name);	-- language name with wikilink
		end
	end

	return language_name;
end


--[[--------------------------< N A M E _ F R O M _ T A G >----------------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function name_from_tag (frame)											-- entry point from an {{#invoke:Lang|name_from_tag|<ietf tag>|link=<yes>|template=<template name>}}
	return _name_from_tag (getArgs(frame))										-- pass-on the args table, nothing else; getArgs() so we also get parent frame
end


--[[--------------------------< _ T A G _ F R O M _ N A M E >--------------------------------------------------

Returns the ietf language tag associated with the language name.  Spelling of language name must be correct
according to the spelling in the source tables.  When a standard language name has a parenthetical disambiguator,
that disambiguator must be omitted (they are not present in the data name-to-tag tables).

Module entry point from another module

]]

local function _tag_from_name (args)											-- entry point from another module
	local msg;

	if args[1] and '' ~= args[1] then
		local data = mw.loadData ('Module:Lang/tag from name');					-- get the reversed data tables TODO: change when going live
		local lang = args[1]:lower();											-- allow any-case for the language name (speeling must till be correct)
		local tag = data.rev_override_table[lang] or data.rev_lang_table[lang] or data.rev_lang_dep_table[lang];	-- get the code; look first in the override then in the standard

		if tag then
			return tag, true;													-- language name found so return tag and done; second return used by is_lang_name()
		else
			msg = 'language: ' .. args[1] .. ' not found'						-- language name not found, error message
		end
	else
		msg = 'missing language name'											-- language name not provided, error message
	end

	local template = '';
	if args.template and '' ~= args.template then
		template = table.concat ({'{{', args['template'], '}}: '});				-- make template name (if provided by the template)
	end
	return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
end


--[[--------------------------< T A G _ F R O M _ N A M E >----------------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function tag_from_name (frame)											-- entry point from an {{#invoke:Lang|tag_from_name|<language name>|link=<yes>|template=<template name>}}
	local result, _ = _tag_from_name (getArgs(frame))							-- pass-on the args table, nothing else; getArgs() so we also get parent frame; supress second return used by is_lang_name()
	return result;		
end


--[[--------------------------< I S _ L A N G _ N A M E >------------------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function is_lang_name (frame)
	local _, result = _tag_from_name (getArgs(frame))							-- pass-on the args table, nothing else; getArgs() so we also get parent frame; supress second return used by tag_from_name()
	return result and true or nil;		
end


--[[--------------------------< _ T R A N S L >----------------------------------------------------------------

Module entry point from another module

]]

local function _transl (args)
	local title_table = lang_data.translit_title_table;							-- table of transliteration standards and the language codes and scripts that apply to those standards
	local language_name;														-- language name that matches language code; used for tool tip
	local translit;																-- transliterated text to display
	local script;																-- IANA script
	local msg;																	-- for when called functions return an error message

	maint_cats = {};															-- initialize because when this module required into another module, these only declared once so only initialzed once
	maint_msgs = {};
	
	if is_set (args[3]) then													-- [3] set when {{transl|code|standard|text}}
		args.text = args[3];													-- get the transliterated text
		args.translit_std = args[2] and args[2]:lower();						-- get the standard; lower case for table indexing

		if not title_table[args.translit_std] then
			return make_error_msg (table.concat ({'unrecognized transliteration standard: ', args.translit_std}), args, 'Transl');
		end
	else
		if is_set (args[2]) then												-- [2] set when {{transl|code|text}}
			args.text = args[2];												-- get the transliterated text
		else
			if args[1] and args[1]:match ('^%a%a%a?%a?$') then					-- args[2] missing; is args[1] a code or is it the transliterated text?
				return make_error_msg ('no text', args, 'Transl');				-- args[1] is a code so we're missing text
			else
				args.text = args[1];											-- args[1] is not a code so we're missing that; assign args.text for error message
				return make_error_msg ('missing language / script code', args, 'Transl');
			end
		end
	end

	if is_set (args[1]) then													-- IANA language code used for html lang= attribute; or ISO 15924 script code
		if args[1]:match ('^%a%a%a?%a?$') then									-- args[1] has correct form?
			args.code = args[1]:lower();										-- use the language/script code; only (2, 3, or 4 alpha characters); lower case because table indexes are lower case
		else
			return make_error_msg (table.concat ({'unrecognized language / script code: ', args[1]}), args, 'Transl');	-- invalid language / script code
		end
	else
		return make_error_msg ('missing language / script code', args, 'Transl');	-- missing language / script code so quit
	end

	args.italic, msg = validate_italic (args);
	if msg then
		return make_error_msg (msg, args, 'Transl');
	end
	
	if 'italic' == args.italic then												-- 'italic' when |italic=yes; because that is same as absent or not set and |italic=default
		args.italic = nil;														-- set to nil; 
	end

	if override_table[args.code] then											-- is code a language code defined in the override table?
		language_name = override_table[args.code];
	elseif lang_table[args.code] then											-- is code a language code defined in the standard language code tables?
		language_name = lang_table[args.code];
	elseif lang_dep_table[args.code] then										-- is code a language code defined in the deprecated language code tables?
		language_name = lang_dep_table[args.code];
	elseif script_table[args.code] then											-- if here, code is not a language code; is it a script code?
		language_name = script_table[args.code];
		script = args.code;														-- code was an ISO 15924 script so use that instead
		args.code = '';															-- unset because not a language code
	else
		return make_error_msg (table.concat ({'unrecognized language / script code: ', args.code}), args, 'Transl');	-- invalid language / script code
	end
																				-- here only when all parameters passed to make_translit() are valid
	return make_translit (args.code, language_name, args.text, args.translit_std, script, args.italic);
end


--[[--------------------------< T R A N S L >------------------------------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function transl (frame)
	return _transl (getArgs(frame));
end


--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------

Returns category name associated with IETF language tag if valid; error message else

All code combinations supported by {{lang}} and the {{lang-xx}} templates are supported by this function.

Module entry point from another module

]]

local function _category_from_tag (args_t)
	local subtags = {};															-- IETF subtags script, region, variant, and private
	local raw_code = args_t[1];													-- save a copy of the input IETF subtag
	local link = 'yes' == args_t.link;											-- make a boolean
	local label = args_t.label;
	local code;																	-- the language code
	local msg;																	-- gets an error message if IETF language tag is malformed or invalid
	local category_name = '';
	
	code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (raw_code);
	if msg then
		local template = (args_t.template and table.concat ({'{{', args_t.template, '}}: '})) or '';	-- make template name (if provided by the template)
		return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
	end

	raw_code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private);	-- format to recommended subtag styles; private omitted because private
	category_name = language_name_get (raw_code, code);							-- get language name; try ietf tag first, then code w/o variant then code w/ variant
	category_name = make_category (code, category_name, nil, true):gsub ('[%[%]]', '');

	if link then
		return table.concat ({'[[:', category_name, ']]'});
	else
		return category_name;
	end
end


--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------

Module entry point from an {{#invoke:}}

]]

local function category_from_tag (frame)										-- entry point from an {{#invoke:Lang|category_from_tag|<ietf tag>|template=<template name>}}
	return _category_from_tag (getArgs (frame));								-- pass-on the args table, nothing else; getArgs() so we also get parent frame
end


--[[--------------------------< E X P O R T E D   F U N C T I O N S >------------------------------------------
]]

return {
	category_from_tag = category_from_tag,
	lang = lang,																-- entry point for {{lang}}
	lang_xx_inherit = lang_xx_inherit,											-- entry points for {{lang-??}}
	lang_xx_italic = lang_xx_italic,
	is_ietf_tag = is_ietf_tag,
	is_ietf_tag_frame = is_ietf_tag_frame,
	is_lang_name = is_lang_name,
	tag_from_name = tag_from_name,												-- returns ietf tag associated with language name
	name_from_tag = name_from_tag,												-- used for template documentation; possible use in ISO 639 name from code templates
	transl = transl,															-- entry point for {{transl}}
	
	_category_from_tag = _category_from_tag,									-- entry points when this module is require()d into other modules
	_lang = _lang,
	_lang_xx_inherit = _lang_xx_inherit,
	_lang_xx_italic = _lang_xx_italic,
	_is_ietf_tag = _is_ietf_tag,
	get_ietf_parts = get_ietf_parts,
	_tag_from_name = _tag_from_name, 
	_name_from_tag = _name_from_tag,
	_transl = _transl,
	_translation_make = translation_make,
	};