Module:Lang#L-94--L-135
--[=[
Lua support for the {{lang}}, {{langx}}, {{lang-??}}, and {{transliteration}} 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 cfg = mw.loadData ('Module:Lang/configuration' .. (mw.getCurrentFrame():getTitle():match ('/sandbox') or '')); -- for internationalization
local is_latn_data = mw.loadData ('Module:Lang/data/is latn data');
local sizeof_ranges_t = is_latn_data.sizeof_ranges_t;
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
--[[--------------------------< M A K E _ E R R O R _ S P A N >--------------------------------------------------
]]
local function make_error_span (template, msg)
return table.concat ({'', cfg.misc_text_t.error, ': ', template, msg, ''})
end
--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
Substitutes $1, $2, etc in
]]
local function substitute (message, data_t)
return data_t and mw.message.newRawMessage (message, data_t):plain() or message;
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_t, template)
local category;
local text; -- handle the oddity that is {{langx}}
if cfg.templates_t.langxx == template then
text = args_t.text or args_t[1]; -- for {{lang-xx}}
else
text = args_t.text or args_t[2]; -- for {{lang}}, {{langx}}, and {{transliteration}}
end
if cfg.templates_t.transliteration == template then
category = cfg.make_error_msg_t.xlit_err_cat;
else
category = cfg.make_error_msg_t.lang_err_cat;
end
local category_link = ((0 == namespace or 10 == namespace) and not args_t.nocat) and substitute ('Category:$1', {category}) or '';
return substitute ('[$1] $2: {{$3}}: $4 ($6)$7', -- { prevents template name from being treated as a template call in certain situations
{
text or cfg.make_error_msg_t.undefined,
cfg.misc_text_t.error,
template,
msg,
category,
cfg.misc_text_t.help,
category_link
})
end
--[[--------------------------< P A R A M E T E R _ V A L I D A T E >------------------------------------------
]]
local function parameter_validate (args_t, template)
local err_msg = cfg.parameter_validate_t.invalid_param;
if cfg.templates_t.lang == template then -- for {{lang}}
for param, _ in pairs (args_t) do
if not cfg.known_params_t.params_lang_t[param] and -- unique {{lang}} parameters
not cfg.known_params_t.common_params_all_t[param] then -- common to all
return substitute (err_msg, {param}); -- not found so abandon
end
end
elseif cfg.templates_t.langx == template then -- for {{langx}}
for param, _ in pairs (args_t) do
if not cfg.known_params_t.params_langx_t[param] and -- unique {{langx}} parameters
not cfg.known_params_t.params_x_t[param] and -- common to {{langx}} and {{lang-xx}}
not cfg.known_params_t.common_params_all_t[param] then -- common to all
return substitute (err_msg, {param}); -- not found so abandon
end
end
elseif cfg.templates_t.langxx == template then -- for {{lang-xx}}
for param, _ in pairs (args_t) do
if not cfg.known_params_t.params_lang_xx_t[param] and -- unique {{lang-xx}} parameters
not cfg.known_params_t.params_x_t[param] and -- common to {{langx}} and {{lang-xx}}
not cfg.known_params_t.common_params_all_t[param] then -- common to all
return substitute (err_msg, {param}); -- not found so abandon
end
end
end
end
--[[--------------------------< 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[^\']", '''); -- 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 acceptable 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 unacceptable 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 xlit function to set args.italic according to the template's
defined default ('inherit' for {{lang}}, 'inherit' or 'italic' for {{lang-??}} depending on
the individual template's requirements, 'italic' for {{transliteration}}) or to the value appropriate to |script=, if set ({{lang}}
and {{lang-??}} 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 = {[cfg.keywords_t.affirmative] = 'italic', [cfg.keywords_t.negative] = 'normal', [cfg.keywords_t.unset] = 'inherit', [cfg.keywords_t.invert] = 'invert', [cfg.keywords_t.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, cfg.validate_italic_t.multi_italic;
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-??}} templates is to add categorization when the templates are used in main space.
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-??}} 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, cfg.get_ietf_parts_t.missing_lang_tag;
end
local pattern = { -- table of tables holding acceptable 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, substitute (cfg.get_ietf_parts_t.unrecog_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, substitute (cfg.get_ietf_parts_t.unrecog_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, substitute (cfg.get_ietf_parts_t.maint_promo_cat, {code}));
table.insert (maint_msgs, substitute (cfg.get_ietf_parts_t.maint_promo_msg, {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, cfg.get_ietf_parts_t.redundant_scr; -- 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, substitute (cfg.get_ietf_parts_t.unrecog_scr_code, {script, 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, substitute (cfg.get_ietf_parts_t.script_code, {script, 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, cfg.get_ietf_parts_t.redundant_reg; -- 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, substitute (cfg.get_ietf_parts_t.unrecog_reg_code, {region, code});
end
end
if is_set (variant) then
if is_set (args_variant) then
return code, nil, nil, nil, nil, cfg.get_ietf_parts_t.redundant_var; -- 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, substitute (cfg.get_ietf_parts_t.unrecog_var, {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, substitute (cfg.get_ietf_parts_t.unrecog_var_code_scr, {variant, 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, substitute (cfg.get_ietf_parts_t.unrecog_var_code_reg, {variant, 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, substitute (cfg.get_ietf_parts_t.unrecog_var_code, {variant, 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, substitute (cfg.get_ietf_parts_t.unrecog_pri, {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 _ 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 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 ({'', display, ''});
else
return table.concat ({'', link, ''});
end
else
return '';
end
end
--[[--------------------------< D I V _ M A R K U P _ A D D >--------------------------------------------------
Adds and tags to list-item text or to implied
..
text. Mixed not supported.]]
local function div_markup_add (text, style)
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%2'); -- 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]+', '
%1
'); -- insert p and italic markup tags at each implied p (two or more consecutive '\n\n' sequences)else
text = text:gsub ('[^\n]+', '
%1
'); -- insert p markup at each implied ptext = 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
HTML markup. This because the lang= attribute applies to the content of its enclosing 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
table.insert (wrapper_t, '\">'); -- end title attribute and close opening wrapper tag
table.insert (wrapper_t, content_text); -- add
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: or tags for inline content or
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:
casa
should be:
casa
or for
]]
local function make_text_html (code, text, tag, rtl, style, size, language)
local html_t = {};
local style_added = '';
local wrapper_tag = tag; --
if text:match ('^%*') then
table.insert (html_t, '*'); -- 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 tags
end
else -- must be div so go
text = div_markup_add (text, style); -- handle implied
, implied
with , and list markup (*;:#) with
end
table.insert (html_t, table.concat ({'<', tag})); -- open the , , or 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= 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' , , or if is_set (language) then -- create a local title_text; if 'zxx' == code then -- special case for this tag 'no linguistic content' title_text = substitute ('$1 $2', {language, cfg.make_text_html_t.zxx}); -- not a language so don't use 'language' in title text elseif mw.ustring.find (language, 'languages', 1, true) then title_text = substitute ('$1 $2', {language, cfg.make_text_html_t.collective}); -- for collective languages else title_text = substitute ('$1-$2', {language, cfg.make_text_html_t.individual}); -- 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, Category:Articles containing For English: Category:Articles containing explicitly cited English-language text For ISO 639-2 collective languages (and for 639-1 bh): Category:Articles with text in ]=] local function make_category (code, language_name, nocat, name_get) 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 substitute ('$1 $2', {cfg.make_category_t.collective_cat, language_name}); end if this_wiki_lang_tag == code then return substitute ('$1 $2 $3-$4', { -- unique category name for the local language cfg.make_category_t.cat_prefix, cfg.make_category_t.explicit_cat, language_name, cfg.make_category_t.cat_postfix, }); else return substitute ('$1 $2-$3', { -- category for individual languages cfg.make_category_t.cat_prefix, language_name, cfg.make_category_t.cat_postfix, }); end end --[[--------------------------< M A K E _ T R A N S L I T >---------------------------------------------------- Return translit ... where xx is the language code; else return empty string. The value |script= is not used in {{transliteration}} for this purpose; instead it uses |code. Because language scripts are listed in the {{transliteration}} 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 xlit() lang_xx() always provides code, language_name, and translit; may provide tscript; never provides style xlit() always provides language_name, translit, and one of code or tscript, never both; always provides style For {{transliteration}}, style only applies when a language code is provided. ]] local function make_translit (code, language_name, translit, std, tscript, style, engvar) 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 = substitute ('$1-$2', {title_text, cfg.misc_text_t.language}); -- skip this text (individual and macro languages only) end title_text = substitute ('$1 $2', {title_text, mw.ustring.lower (cfg.engvar_t[engvar]['romanisz_lc'])}); -- 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 = substitute ('$1$2 ($3 $4) $5', { -- add the appropriate text to the tool tip title_text, title_t[std][tscript:lower()], script_table[tscript], cfg.make_translit_t.script, cfg.make_translit_t.transliteration, }); 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 (xlit may not provide a language code) title_text = substitute ('$1$2 ($3 $4) $5', { -- add the appropriate text to the tool tip title_text, title_t[std][code:lower()], language_name, cfg.misc_text_t.language, cfg.make_translit_t.transliteration, }); 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 = substitute ('$1$2-$3 $4', { -- write a script tool tip title_text, language_name, cfg.make_translit_t.script, cfg.make_translit_t.transliteration, }); 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 = substitute ('$1-$2', { -- skip this text (individual and macro languages only) title_text, cfg.misc_text_t.language, }); end title_text = substitute ('$1 $2', { -- finish the tool tip title_text, cfg.make_translit_t.transliteration, }); else title_text = substitute ('$1 $2', { -- generic tool tip (can we ever get here?) title_text, cfg.make_translit_t.transliteration, }); end end end local tag if is_set (code) then -- when a language code is provided (always with {{lang-??}} templates, not always with {{transliteration}}) code = code:match ('^(%a%a%a?)'); -- strip all subtags leaving only the language subtag if not style then -- nil for the default italic style tag = '%s' else tag = '%s' -- non-standard style, construct a span tag for it tag = string.format(tag, style, "%s", "%s") end tag = string.format(tag, code, "%s") else tag = '%s' -- when no language code: no lang= attribute, not italic ({{transliteration}} only) end tag = string.format(tag, translit) -- add the translit text if '' == title_text then -- when there is no need for a tool-tip return tag; -- done else title_text = cfg.engvar_sel_t.gb == engvar and title_text:gsub ('([Rr]omani)z', '%1s') or title_text; -- gb eng when engvar specifies gb eng; us eng else return title_wrapper_make (title_text, tag, 'span'); -- wrap with a tool-tip span and done 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 (cfg.validate_text_t.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 (cfg.validate_text_t.malformed_markup, args, template); end local style = args.italic; if (cfg.keywords_t.unset ~= style) and (cfg.keywords_t.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 (cfg.validate_text_t.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 ({' for _, msg in ipairs (maint_msgs) do table.insert (maint, table.concat ({msg, ' '})); -- add message strings end table.insert (maint, ''); -- 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 ({'Category:', cat, ''})); -- 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 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 .. 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 the stripmarker is replaced with text wrapped in If the text contains any actual tags are used to wrap it, to prevent div/span inversion. ]] 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 text:find (' tag = 'div'; 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 nil with return value for use in error message. ]] local function validate_proto (proto_param) if cfg.keywords_t.affirmative == proto_param then return true; elseif cfg.keywords_t.negative == proto_param then return false; elseif is_set (proto_param) then return nil, proto_param; -- |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, substitute (cfg.language_name_get_t.deprecated_cat, {code})); table.insert (maint_msgs, substitute (cfg.language_name_get_t.deprecated_msg, {code})); end return name; -- name from data tables or nil end --[[--------------------------< T E X T _ S C R I P T _ M A T C H _ T E S T >---------------------------------- IETF script subtag should match the script of the what they look like. It does know what Latn script looks like so when script, the IETF script subtag, if present, should not be Latn. Conversely, when For the purposes of this test, Latf (Fraktur) and Latg (Gaelic) are considered to be equivalent to Latn because unicode treats these two scripts as font-specific variants of Latn. Returns an error message when mismatch detected; nil else. ]] local function text_script_match_test (script, is_latn_text, pos, char) local scripts_t = {['latf'] = true, ['latg'] = true, ['latn'] = true}; -- unicode 'latn' scripts; 'latf' and 'latg' are font variants so there are no Fraktur or Gaelic codepoints if is_set (script) then -- don't bother with the rest of this if