|
require('Module:No globals');
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 lang_name_table = mw.loadData ('Module:Language/name/data');
local namespace = mw.title.getCurrentTitle().namespace; -- used for categorization
local this_wiki_lang = mw.language.getContentLanguage().code; -- get this wiki's language
local maint_cats = {}; -- maintenance categories go here
--[[--------------------------< I N V E R T S _ I TL A LT I C SN >----------------------------------------------------------------
Returns true if all of text argument is written using Latn script for letters, numbers and punctuationset; false else.
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.
For the purposes of this function, Latn script is characters less control characters from these Unicode 10.0 Character Code Charts:
Leading and trailing wiki markup is extracted from args.text into separate table elements. Addition, removal,
[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
replacement of wiki markup is handled by a string.gsub() replacement table operating only on these separate elements.
[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)
In the string.gsub() matching pattern, '.*' matches empty string as well as the three expected wiki markup patterns.
[http://www.unicode.org/charts/PDF/U0100.pdf Latin Extended-A] U+0100–U+017F (C4 80 - C5 BF)
[http://www.unicode.org/charts/PDF/U0180.pdf Latin Extended-B] U+0180–U+024F (C6 80 - C9 8F)
[http://www.unicode.org/charts/PDF/U1E00.pdf Latin Extended Additional] U+1E00-U+1EFF (E1 B8 80 - E1 BB BF)
[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)
[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:
This function expects that markup in args.text is complete and correct; if it is not, oddness may result.
[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
poem stripmarkers. These have the form:
?'"`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.
]]
local function invert_italicsis_latn (sourcetext)
local latn = table.concat (
local invert_pattern_table = { -- leading/trailing markup add/remove/replace patterns
{
[""]="\'\'", -- empty string becomes italic markup
["\'\['"]="", -- italicthis is a set markupso becomesinclude emptyopening stringbracket
'\n\32-\127', -- C0 Controls and Basic Latin U+0020–U+007E (20 - 7E) + (U+0010 and U+007F <poem>...</poem> support)
["\'\'\'"]="\'\'\'\'\'", -- bold becomes bold italic
["\'\'194\'160-\'194\172'"]="\'\'\'", -- boldC1 italicControls and Latin-1 Supplement U+00A0-U+00AC (C2 A0 - becomeC2 boldAC)
'\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)
local seg = {};
'\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)
']', -- close the set
});
text = mw.text.decode (text, true); -- replace numeric and named html entities with their unicode characters
source = source:gsub ("%f[\']\'%f[^\']", '''); -- protect single quote marks from being interpreted as bold markup
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
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
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 (argsitalic, italics)
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 countitalic >and 1italics then -- return nil and an error message if more than oneboth isare set
return nil, 'only one ofconflicting: |italic=, and |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
]]
local function in_array ( needle, haystack )
if needle == nil then
return false;
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.
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
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)$', 's', 'r', 'v'}, -- 1 - ll-Ssss-RR-variant (where variant is 4 digits);
{ elseif source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)%-(%d%d%d%d)$',) 's', 'r', 'v'},then -- 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)%-(%ad%ad%d)%-(%wd%wd%wd%w%w%w?%w?%w?d)$', 's', 'r', 'v'}, -- 3 - ll-Ssss-RR-variant (where variant is 5-8 alnum characters);
{ elseif source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%da%da%-[%a%d)][%-(a%wd][%wa%wd][%wa%wd][%w?a%w?%w?)d]+$',) 's', 'r', 'v'},then -- 4 - ll-Ssss-DDDRR-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]?)$');
{ elseif source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d%-[%a%d)][%a%d][%a%d][%a%d][%a%d]+$',) 's', 'v'},then -- 5 - ll-Ssss-DDD-variant (where variantregion is 43 digits; variant is 5-8 alnum characters)
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)
c1code, c2region, c3, c4variant = source:match (pattern[i][1]'^(%a%a%a?); %-(%d%d%d)%- one or more captures set if source matches pattern[i](%d%d%d%d)$');
elseif source:match ('^%a%a%a?%-%a%a%-[%a%d][%a%d][%a%d][%a%d][%a%d]+$') 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]?)$');
[pattern[i][2] or 'x'] = c2, -- fill the table of captures with the rest of the captures
[pattern[i][3] or 'x'] = c3, -- take index names from pattern table and assign sequential captures
[pattern[i][4] or 'x'] = c4, -- index name may be nil in pattern[i] table so "or 'x'" spoofs a name for this index in this table
};
script = t.s or ''; -- translate table contents to named variables;
region = t.r or ''; -- absent table entries are nil so set named ietf parts to empty string for concatenation
variant= t.v or '';
private = t.p or '';
break; -- and done
end
end
elseif source:match ('^%a%a%a?%-%d%d%d%d$') then -- ll-variant (where variant is 4 digits)
if not code then
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)
code, variant = source:match ('^(%a%a%a?)%-([%a%d][%a%d][%a%d][%a%d][%a%d][%a%d]?[%a%d]?[%a%d]?)$');
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%a%a$') then -- ll-Ssss-RR
code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%a%a)$');
elseif source:match ('^%a%a%a?%-%a%a%a%a%-%d%d%d$') then -- ll-Ssss-DDD (region is 3 digits)
code, script, region = source:match ('^(%a%a%a?)%-(%a%a%a%a)%-(%d%d%d)$');
elseif source:match ('^%a%a%a?%-%a%a%a%a$') 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
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language tag: ', source}); -- don't know what we got but it is malformed
end
table.insert (out, '</span>');
if (0 == namespace) and not is_set (args.nocat) then -- only categorize in article space
table.insert (out, table.concat ({'[[Category:', category, ' template errors]]'}));
end
--[[--------------------------< DM IA VK E _ MT AE RX KT U_ S P _ A D DN >--------------------------------------------------
TODO: add support for block: div tags instead of span tags; would need some sort of proper parameter to control the switch
adds <i> and </i> tags to list-item text or to implied <p>..</p> text. mixed not supported
For italic style, can't do ''{{lang|xx|text}}'' without using <span/> tags when text is italic because of -Latn, |italic=yes,
]]
or auto-italics because the wrapping wikimarkup produces this:
<i><i lang="xx">text</i></i>
local function div_markup_add (text, style)
which is later reduced to this:
local implied_p = {};
<i>text</i>
This reduction happens in some sort of cleanup process outside the scope of this template/module.
if text:find ('^\n[%*:;#]') then -- look for list markup; list markup must begin at start of text
if 'italic' == style then
return mw.ustring.gsub (text, '(\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
--[[--------------------------< M A K E _ T E X T _ H T M L >--------------------------------------------------
Until or unless this is fixed italic text must be:
Add the html markup to text according to the type of content that it is: <span> or <i> tags for inline content or
<i><span lang="xx">text</span></i>
<div> tags for block content
]]
local function make_text_htmlmake_text_span (code, text, tag, rtl, style, size, language)
local htmlspan = {};
local style_added = '';
if text:match ('^%*') then
table.insert (htmlspan, '*'); -- move proto language text prefix outside of italic markup if any; use numeric entity because plainplan splat confuses MediaWiki
text = text:gsub ('^%*', ''); -- remove the splatspat from the text
end
if 'italic' == style then
if 'span' == tag then -- default html tag for inline content
iftable.insert (span, 'italic<i>' == style then); -- butopen ifitalic italicstyle tag
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 (span, table.concat ({'<span lang="'})); -- open <span> tag
table.insert (span, table.concat ({code, '\"'})); -- add language attribute
if rtl then
table.insert (html, table.concat ({'<', tag})); -- open the <i>, <span>, or <div> html tag
table.insert (htmlspan, table.concat ({' langdir="', code, '\rtl"'})); -- add languagedirection attribute for right to left languages
if rtl or unicode.is_rtl(text) then
table.insert (html, ' dir="rtl"'); -- add direction attribute for right to left languages
end
if 'normal' == style then -- when |italic=no
table.insert (htmlspan, ' 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 (htmlspan, table.concat ({' font-size:', size, ';'})); -- add when style attribute already inserted
else
table.insert (htmlspan, 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 (span, table.concat ({style_added, '>'})); -- close the style attribute and close opening span tag
if is_set (language) then
table.insert (htmlspan, table.concat ({style_added, ' title=\"', language})text); --start theinsert titlethe text
if language:find ('languages') then
table.insert (htmlspan, ' collective text\"</span>'); -- forclose the collective<span> languagestag
if 'italic' == style then
else
table.insert (htmlspan, ' language text\"</i>'); -- forclose italic individualstyle languagestag
end
table.insert (html, '>'); -- close the opening html tag
else
table.insert (html, table.concat ({style_added, '>'})); -- close the style attribute and close opening html tag
end
table.insert (html, text); -- insert the text
if rtl then -- legacy; shouldn't be necessary because all of the rtl text is wrapped in <span dir="rtl">text</span>
table.insert (htmlspan, table.concat ({'</‎', tag, '>'})); -- closemake sure the <i>,browser <span>,knows orthat <div>we're htmlat the end of the tagrtl
if rtl then -- legacy; shouldn't be necessary because all of the rtl text is wrapped inside an html tag with dir="rtl" attribute
table.insert (html, '‎'); -- make sure the browser knows that we're at the end of the rtl
end
return table.concat (htmlspan); -- put it all together and done
end
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, '-language') -- skip this text (individual and macro languages only)
end
table.insert (tout, ' romanizationtransliteration'); -- 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_table[std] then -- and if standard is legitimate
table.insert (tout, '">');
table.insert (tout, translit);
if is_set (code) and not style then -- when a language code is provided (always with {{lang-xx}} templates, not always with {{transl}})
table.insert (tout, "</i>"); -- close the italic tag
else
--[=[--------------------------< 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
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)
end
if 'lang-xx' == template then -- for the time being, this error checking does not apply to {{lang}}
local style = args.italic;
if ('unset' ~= args.italic) and ('unset' ~= args.italics) then -- allow italic markup when |italic=unset or |italics=unset
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
return make_error_msg ('text has italic markup', args, template);
if args.text:find ("%f[\']\'\'%f[^\']") or args.text:find ("%f[\']\'\'\'\'\'%f[^\']") then -- italic but not bold, or bold italic
end
return make_error_msg ('text has italic markup', args, template);
end
end
end
if (0 < #maint_cats) and (0 == namespace) and not is_set (nocat) then -- when there are maintenancemainenance categories; article namespace only
for _, cat in ipairs (maint_cats) do
table.insert (maint, table.concat ({'[[Category:', cat, ']]'})); -- format and add the categories
--[[--------------------------< 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_htmlmake_text_span() 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 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 text:gsubtable.concat ({'^%**', ''text}); -- return text without splat prefix regardless ofproto language nametext orwith existinga 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;
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
<includeonly>{{#invoke:lang|lang}}</includeonly>
parameters are receivedrecieved from the template's frame (parent frame)
]]
local function _langp.lang (argsframe)
local args = getArgs(frame);
local out = {};
local language_name; -- used to make category names
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';
if args[1] and args.code then
return make_error_msg ('conflicting: {{{1}}} and |code=', args, template'lang');
else
args.code = args[1] or args.code; -- prefer args.code
if args[2] and args.text then
return make_error_msg ('conflicting: {{{2}}} and |text=', args, template'lang');
else
args.text = args[2] or args.text; -- prefer args.text
end
msg = validate_text (template'lang', args); -- ensure that |text= is set (italic test disabled for the time being)
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
validate_cat_args (args); -- determine if categorization should be suppressed
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'lang');
end
args.italic, msg = validate_italic (args.italic, args.italics);
if msg then
return make_error_msg (msg, args, template'lang');
end
if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
if ('latn' == subtags.script) or (is_latn (args.text) and 'en' ~= code) then -- script set to latn or text is wholly latn script but not rendering English
args.italic = 'italic'; -- DEFAULT for {{lang}} templates is upright; but if latn script set for font-style:italic
(this_wiki_lang ~= 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
args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant); -- format to recommended subtag styles; private omitted because private
subtags.privateif =is_set (subtags.private) and lang_data.override[table.concat ({code, '-x-', subtags.private})] or nil; then -- assemblelook a completefor private ietfuse subtagtags; args.codedone doesthis notway getbecause private subtag...
language_name = lang_data.override[table.concat ({code, '-x-', subtags.private})][1]; -- ... args.code does not get private subtag
if is_set (subtags.private) andelseif lang_data.override[subtags.privatecode] then -- get the language name for categorization
language_name = lang_data.override[subtags.privatecode][1]; -- firstprefer looklanguage fornames privatetaken usefrom tagthe languageoverride nametable
elseif lang_data.override[code] then
language_name = lang_data.override[code][1] -- then 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
args.text = proto_prefix (args.text, language_name); -- prefix proto-language text with a splat
if 'invert' == args.italic and 'span' == tag then -- invert only supported for in-line content
args.text = invert_italics (args.text)
end
table.insert (out, make_text_span (args.code, args.text, args.rtl, args.italic, args.size));
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
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 knowsknow the default
styling that should be applied to text.
Supported parameters are:
|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]
|region = IANA region code
|variant = IANA variant code
|text = (required) the displayed text in language specified by code
|link = boolean false ('no') doesunlinks notlanguage linkspecified by 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 templatestemplate where |text= is not entirely latn script):
|translit = text that is a transliteration of text
|translit-std = the standard that applies to the transliteration
]]
local function _lang_xx (argsframe)
local args = getArgs(frame, {parentFirst= true}); -- parameters in the template override parameters set in the {{#invoke:}}
local out = {};
local language_name; -- used to make display text, article links
local translit_title;
local msg; -- for error messages
local tag = 'span'; -- initial value for make_text_html()
local template = args.template or 'lang-xx';
if args[1] and args.text then
return make_error_msg ('conflicting: {{{1}}} and |text=', args, template'lang-xx');
else
args.text = args[1] or args.text; -- prefer args.text
end
msg = validate_text (template'lang-xx', 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'lang-xx');
else
args.translit = args[2] or args.translit -- prefer args.translit
if args[3] and (args.translation or args.lit) then
return make_error_msg ('conflicting: {{{3}}} and |lit= or |translation=', args, template'lang-xx');
elseif args.translation and args.lit then
return make_error_msg ('conflicting: |lit= and |translation=', args, template'lang-xx');
else
args.translation = args[3] or args.translation or args.lit; -- prefer args.translation
if args.links and args.link then
return make_error_msg ('conflicting: |links= and |link=', args, template'lang-xx');
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
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args, template'lang-xx');
end
args.italic, msg = validate_italic (args.italic, args.italics);
if msg then
return make_error_msg (msg, args, template'lang-xx');
end
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); -- format to recommended subtag styles
subtags.privateif =is_set (subtags.private) and lang_data.override[table.concat ({code, '-x-', subtags.private})] or nil; then -- assemblelook a completefor private ietfuse subtagtags; args.codedone doesthis notway getbecause private subtag...
language_name = lang_data.override[table.concat ({code, '-x-', subtags.private})][1]; -- ... args.code does not get private subtag
if is_set (subtags.private) and lang_data.override[subtags.private] then -- get the language name for categorization
language_name = lang_data.override[subtags.private][1]; -- first look for private use tag language name
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
if language_name:find ('languages') 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][1], 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
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_htmlmake_text_span (args.code, args.text, tag, args.rtl, args.italic, args.size))
if is_set (args.translit) and not unicode.is_Latinis_latn (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
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, 'romanizedtranslit.') .. ':'); -- make a wikilink if there is an article to link to
else
table.insert (out, 'romanized:<abbr title="transliteration">translit.</abbr>'); -- else plaindefine the textabbreviation
end
table.insert (out, ' </small>'); -- close the small tag
table.insert (out, translit);
else
return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args, template'lang-xx');
end
end
table.insert (out, " </small>");
end
table.insert (out, table.concat ({"'''", args.translation, "'''"})); -- use html entities to avoid wiki markup confusion
end
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
]]
local function p.lang_xx_italic (frame)
local args = lang_xx_args_get (frame);
initial_style_state = 'italic';
return _lang_xx (argsframe);
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
]]
local function p.lang_xx_inherit (frame)
local args = lang_xx_args_get (frame);
initial_style_state = 'inherit';
return _lang_xx (argsframe);
end
--[[--------------------------< _ LN A NM GE _ XF XR O M _ IC NO HD 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.
The purpose of this function is to replace {{#ifexist:Template:ISO 639 name xx|<exists>|<does not exist>}} in
templates that are better served by using {{#invoke:lang|name_from_tag|<code>}}
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, 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) -- entry point from an {{#invoke:Lang|is_ietf_tag|<ietf tag>}}
return _is_ietf_tag (frame.args[1]); -- frame.args[1] is the ietf language tag
end
--[[--------------------------< _ N A M E _ F R O M _ C O D E >------------------------------------------------
Returns language name associated with IETF language tag if valid; empty string 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_tagp.name_from_code (argsframe)
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 IETF subtag
local link = 'yes' == args['link']; -- make a boolean
local code; -- the language code
local msg; -- getsholds an error message (not used here) 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
return '';
local template = (args['template'] and table.concat ({'{{', args['template'], '}}: '})) or ''; -- make template name (if provided by the template)
return table.concat ({'<span style=\"font-size:100%; font-style:normal;\" class=\"error\">error: ', template, msg, '</span>'});
end
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 language_name:find ('languages') then
return make_wikilink (language_name); -- collective language name uses simple wikilink
elseif lang_data.article_name[code] then
return make_wikilink (lang_data.article_name[code][1], language_name); -- language name with wikilink from override data
else
return make_wikilink (language_name .. ' language', language_name); -- language name with wikilink
end
end
return language_name;
--[[--------------------------< N A M E _ FT R OA M _ C ON DS EL >------------------------------------------------------------------
Prospective replacement for the template {{transl}}
Module entry point from an {{#invoke:}}
]]
function p.transl (frame)
local function name_from_tag (frame) -- entry point from an {{#invoke:Lang|name_from_tag|<ietf tag>|link=<yes>|template=<template name>}}
returnlocal _name_from_tagargs = getArgs(frame.args); -- pass-on the args table,no nothing{{#invoke:}} elseparameters
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/name to tag'); -- get the reversed data tables
local lang = args[1]:lower(); -- allow any-case for the language name (speeling must till be correct)
local tag = data.rev_lang_data[lang] or data.rev_lang_name_table[lang];-- get the code; look first in the override then in the standard
if tag then
return tag; -- language name found so return tag and done
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=\"font-size:100%; font-style:normal;\" class=\"error\">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>}}
return _tag_from_name (frame.args); -- pass-on the args table, nothing else
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
end
args.italic, msg = validate_italic (args.italic, args.italics);
if msg then
return make_error_msg (msg, args, 'transl');
return p;
--[[--------------------------< T R A N S L >------------------------------------------------------------------
Module entry point from an {{#invoke:}}
]]
local function transl (frame)
local args = getArgs(frame); -- no {{#invoke:}} parameters
return _transl (args);
end
--[[--------------------------< E X P O R T E D F U N C T I O N S >------------------------------------------
]]
return {
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,
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}}
_lang = _lang, -- entry points when this module is require()d into other modules
_lang_xx_inherit = _lang_xx_inherit,
_lang_xx_italic = _lang_xx_italic,
_is_ietf_tag = _is_ietf_tag,
_tag_from_name = _tag_from_name,
_name_from_tag = _name_from_tag,
_transl = _transl,
};
| 