4,919
edits
(synch from sandbox;) |
m (1 revision imported) |
||
(32 intermediate revisions by 8 users not shown) | |||
Line 1:
--[=[
Lua support for the {{lang}}, {{lang-xx}}, and {{
]=]
require('
local getArgs = require ('Module:Arguments').getArgs;
local
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
Line 34 ⟶ 46:
--[[--------------------------< I
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
local invert_pattern_table = { -- leading/trailing markup add/remove/replace patterns
[""]="\'\'", -- empty string becomes italic markup
["\'\'\'"]="\'\'\'\'\'", -- bold becomes bold italic
["\'\
};
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
Line 99 ⟶ 105:
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 {{
and {{lang-xx}} only).
Line 108 ⟶ 114:
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 (
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
return nil, '
end
return properties[args.italic or args.italics or args.i], nil;
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
Line 128 ⟶ 174:
]]
local function in_array ( needle, haystack )
if needle == nil then
return false;
Line 148 ⟶ 194:
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
Line 178 ⟶ 229:
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
Line 197 ⟶ 248:
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.
Line 205 ⟶ 256:
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
Line 215 ⟶ 262:
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)%-(%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
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
Line 268 ⟶ 319:
code = code:lower(); -- ensure that we use and return lower case version of this
if not (
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
Line 288 ⟶ 339:
if is_set (script) then
script = script:lower(); -- ensure that we use and return lower case version of this
if not
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
if in_array (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
Line 308 ⟶ 359:
if is_set (region) then
region = region:lower(); -- ensure that we use and return lower case version of this
if not
return code, script, nil, nil, nil, table.concat ({'unrecognized region: ', region, ' for code: ', code});
end
Line 323 ⟶ 374:
if is_set (variant) then
variant = variant:lower(); -- ensure that we use and return lower case version of this
if not
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}),
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,
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
return code, script, region, nil, nil, table.concat ({'unrecognized private tag: ', private});
end
Line 357 ⟶ 414:
local category;
if '
category = '
else
category = '
end
table.insert (out, table.concat ({'
table.insert (out, table.concat ({'<span style=\"
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
table.insert (out, table.concat ({'[[Category:', category, ' template errors]]'}));
end
Line 398 ⟶ 455:
--[[--------------------------<
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
local
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 (
text = text:gsub ('^%*', ''); -- remove the
end
if 'span' == tag then -- default html tag for inline content
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 (
style_added = '\"'; -- remember that style attribute added and is not yet closed
end
Line 441 ⟶ 560:
if is_set (size) then -- when |size=<something>
if is_set (style_added) then
table.insert (
else
table.insert (
style_added = '\"'; -- remember that style attribute added and is not yet closed
end
end
table.insert (
table.insert (
table.insert (
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
Line 472 ⟶ 597:
[[Category:Articles containing explicitly cited English-language text]]
for ISO 639-2 collective languages (and for 639-1 bh):
[[Category:Articles with text
]=]
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
return table.concat ({'[[Category:Articles with text
end
table.insert (cat, '[[Category:Articles containing ');
if
table.insert (cat, 'explicitly cited
else
table.insert (cat, language_name);
Line 503 ⟶ 624:
table.insert (cat, '-language text]]');
return table.concat (cat);
end
Line 511 ⟶ 632:
return translit <i lang=xx-Latn>...</i> where xx is the language code; else return empty string
The value |script= is not used in {{
are listed in the {{
at {{Language with name and transliteration}}. If |script= is set, this function uses it in preference to code.
Line 522 ⟶ 643:
transl() always provides language_name, translit, and one of code or tscript, never both; always provides style
For {{
]]
local function make_translit (code, language_name, translit, std, tscript, style)
local title;
local
local
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
if not
end
elseif is_set (std) and is_set (tscript) then -- when both are specified
if
if
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
end
else
Line 562 ⟶ 677:
elseif is_set (std) then -- translit-script not set, use language code
if not
if
else -- code doesn't match
end
else -- here if translit-script set but translit-std not set
if
elseif
else
if is_set (tscript) then
elseif is_set (code) then
if not
end
else
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 {{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
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 ({{transliteration}} 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.
end
end
--[
This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
Line 605 ⟶ 739:
Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
either as ''itself
]
local function validate_text (template, args)
Line 618 ⟶ 752:
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
Line 638 ⟶ 772:
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
Line 645 ⟶ 779:
end
if (0 < #maint_cats) and (0 == namespace) and not
for _, cat in ipairs (maint_cats) do
table.insert (maint, table.concat ({'[[Category:', cat, ']]'})); -- format and add the categories
Line 657 ⟶ 791:
--[[--------------------------< 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
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
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
--[[--------------------------<
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.
If the text contains any actual <div>...</div> tags, then it's again returned unmodified and <div>...</div>
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 ('<div') then -- reductive; if the text contains a div tag, we must use div tags
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 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}}
Line 679 ⟶ 928:
<includeonly>{{#invoke:lang|lang}}</includeonly>
parameters are
]]
local function
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
else
args.code = args[1] or args.code; -- prefer args.code
Line 698 ⟶ 954:
if args[2] and args.text then
return make_error_msg ('conflicting: {{{2}}} and
else
args.text = args[2] or args.text; -- prefer args.text
end
msg = validate_text (
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,
end
args.italic, msg = validate_italic (args
if msg then
return make_error_msg (msg, args,
end
if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
if ('latn' == subtags.script) or
(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
Line 737 ⟶ 1,001:
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);
table.insert (out,
table.insert (out, make_category (code, language_name, args.nocat));
table.insert (out, render_maint(args.nocat));
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, " </small>"); -- close the <small> html tag
end
table.insert (translation_t, table.concat ({''', args_t.translation, '''})); -- use html entities to avoid wiki markup confusion
return table.concat (translation_t); -- make a big string and done
end
Line 763 ⟶ 1,079:
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
styling that should be applied to text.
Line 775 ⟶ 1,091:
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')
|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 =
any other text replaces language-name label - automatic wikilinking disabled
for those {{lang-xx}} templates that support transliteration (those
|translit = text that is a transliteration of text
|translit-std = the standard that applies to the transliteration
Line 800 ⟶ 1,117:
]]
local function _lang_xx (
local out = {};
local language_name; -- used to make display text, article links
Line 812 ⟶ 1,128:
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
else
args.text = args[1] or args.text; -- prefer args.text
end
msg = validate_text (
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
else
args.translit = args[2] or args.translit -- prefer args.translit
Line 831 ⟶ 1,154:
if args[3] and (args.translation or args.lit) then
return make_error_msg ('conflicting: {{{3}}} and
elseif args.translation and args.lit then
return make_error_msg ('conflicting:
else
args.translation = args[3] or args.translation or args.lit; -- prefer args.translation
Line 839 ⟶ 1,162:
if args.links and args.link then
return make_error_msg ('conflicting:
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
Line 849 ⟶ 1,174:
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args,
end
args.italic, msg = validate_italic (args
if msg then
return make_error_msg (msg, args,
end
Line 877 ⟶ 1,202:
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)
Line 906 ⟶ 1,221:
table.insert (out, language_name); -- language name without wikilink
else
if
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
Line 915 ⟶ 1,232:
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,
if is_set (args.translit) and not
table.insert (out, ', '); -- comma to separate text from translit
if 'none' ~= args.label then
table.insert (out, '<small>');
if
translit_script_name =
else
translit_script_name = language_name; -- fall back on language name
Line 930 ⟶ 1,251:
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, '
else
table.insert (out, '
end
table.insert (out, ' </small>'); -- close the small tag
Line 941 ⟶ 1,262:
table.insert (out, translit);
else
return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args,
end
end
if is_set (args.translation) then -- translation (not supported in {{lang}})
table.insert (out,
end
Line 963 ⟶ 1,274:
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
Line 972 ⟶ 1,312:
]]
local function
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 (
end
Line 984 ⟶ 1,338:
]]
local function
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 (
end
--[[--------------------------<
Returns true when a language name associated with IETF language tag
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 =
local link = 'yes' == args['link']; -- make a boolean
local label = args.label;
local code; -- the language code
local msg; --
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
--[[--------------------------<
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
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; --
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 {{transliteration|code|standard|text}}
args.text = args[3]; -- get the transliterated text
args.translit_std = args[2] and args[2]:lower();
if not title_table[args.translit_std] then
return make_error_msg (table.concat ({'unrecognized transliteration standard: ', args.translit_std}), args, '
end
else
if is_set (args[2]) then -- [2] set when {{
args.text = args[2]; -- get the transliterated text
else
if args[1] and (args[1]:match ('^%a%a%a?%a?$')
return make_error_msg ('no text', args, 'Transliteration'); -- args[1] is a code so we're missing text
else
args.text = args[1]; -- args[1] is not a code so we're missing that; assign args.text for error message
return make_error_msg ('missing language / script code', args, '
end
end
Line 1,066 ⟶ 1,554:
if is_set (args[1]) then -- IANA language code used for html lang= attribute; or ISO 15924 script code
if args[1]:match ('^%a%a%a?%a?$') or args[1]:match ('^%a%a%a?%-x%-') then -- args[1] has correct form?
args.code = args[1]:lower(); -- use the language/script code; only (2, 3, or 4 alpha characters) or private-use; lower case because table indexes are lower case
else
return make_error_msg (table.concat ({'unrecognized language / script code: ', args[1]}), args, 'Transliteration'); -- invalid language / script code
end
else
return make_error_msg ('missing language / script code', args, '
end
args.italic, msg = validate_italic (args
if msg then
return make_error_msg (msg, args, '
end
Line 1,080 ⟶ 1,572:
end
if
language_name =
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, '
end
-- here only when all parameters passed to make_translit() are valid
Line 1,096 ⟶ 1,591:
--[[--------------------------< 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 {{transliteration}}
_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,
};
|