Module:Lang: Difference between revisions

No edit summary
 
m 1 revision imported
 
(6 intermediate revisions by 3 users not shown)
Line 8: Line 8:


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


Line 22: Line 22:


local synonym_table = mw.loadData ('Module:Lang/ISO 639 synonyms'); -- ISO 639-2/639-2T code translation to 639-1 code
local synonym_table = mw.loadData ('Module:Lang/ISO 639 synonyms'); -- ISO 639-2/639-2T code translation to 639-1 code
local cfg = mw.loadData ('Module:Lang/configuration' .. (mw.getCurrentFrame():getTitle():match ('/sandbox') or '')); -- for internationalization


local namespace = mw.title.getCurrentTitle().namespace; -- used for categorization
local namespace = mw.title.getCurrentTitle().namespace; -- used for categorization
Line 35: Line 37:




--[[--------------------------< I S _ S E T >------------------------------------------------------------------
--[[--------------------------< M A K E _ E R R O R _ S P A N >--------------------------------------------------
 
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 )
local function make_error_span (template, msg)
return not (var == nil or var == '');
return table.concat ({'<span style="color:#d33">', cfg.misc_text_t.error, ': ', template, msg, '</span>'})
end
end




--[[--------------------------< I N V E R T _ I T A L I C S >-------------------------------------------------
--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
 
Substitutes $1, $2, etc in <message> with data from <data_t>. Returns plain-text substituted string when
<data_t> not nil; returns <message> else.
 
]]
 
local function substitute (message, data_t)
return data_t and mw.message.newRawMessage (message, data_t):plain() or message;
end


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,
--[[--------------------------< M A K E _ E R R O R _ M S G >--------------------------------------------------
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.
Assembles an error message from template name, message text, help link, and error category.


]]
]]


local function invert_italics (source)
local function make_error_msg (msg, args_t, template)
local invert_pattern_table = { -- leading/trailing markup add/remove/replace patterns
local category;
[""]="\'\'", -- empty string becomes italic markup
local text; -- handle the oddity that is {{langx}}
["\'\'"]="", -- italic markup becomes empty string
if cfg.templates_t.langxx == template then
["\'\'\'"]="\'\'\'\'\'", -- bold becomes bold italic
text = args_t.text or args_t[1]; -- for {{lang-xx}}
["\'\'\'\'\'"]="\'\'\'", -- bold italic become bold
else
};
text = args_t.text or args_t[2]; -- for {{lang}}, {{langx}}, and {{transliteration}}
local seg = {};
end
 
source = source:gsub ("%f[\']\'%f[^\']", '&#39;'); -- protect single quote marks from being interpreted as bold markup


seg[1] = source:match ('^(\'\'+%f[^\']).+') or ''; -- get leading markup, if any; ignore single quote
if cfg.templates_t.transliteration == template then
seg[3] = source:match ('.+(%f[\']\'\'+)$') or ''; -- get trailing markup, if any; ignore single quote
category = cfg.make_error_msg_t.xlit_err_cat;
 
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
else
seg[2] = source -- when there is no markup
category = cfg.make_error_msg_t.lang_err_cat;
end
end
local category_link = ((0 == namespace or 10 == namespace) and not args_t.nocat) and substitute ('[[Category:$1]]', {category}) or '';
return substitute ('[$1] <span style="color:#d33">$2: {{$3}}: $4 ([[:Category:$5|$6]])</span>$7',
{
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 >------------------------------------------
]]


seg[1] = invert_pattern_table[seg[1]] or seg[1]; -- replace leading markup according to pattern table
local function parameter_validate (args_t, template)
seg[3] = invert_pattern_table[seg[3]] or seg[3]; -- replace leading markup according to pattern table
local err_msg = cfg.parameter_validate_t.invalid_param;


return table.concat (seg); -- put it all back together and done
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}); -- <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}); -- <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}); -- <param> not found so abandon
end
end
end
end
end




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


validates |italic= or |italics= assigned values.
Returns true if argument is set; false otherwise. Argument is 'set' when it exists (not nil) or when it is not
an empty string.


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


When |italic= is not set, or has an unacceptible assigned value, return nil and a nil error message.
local function is_set (var)
return not (var == nil or var == '');
end


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


The return value nil causes the calling lang, lang_xx, or transl function to set args.italic according to the template's
--[[--------------------------< I N V E R T  _ I T A L I C 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:
This function attempts to invert the italic markup a args.text by adding/removing leading/trailing italic markup
nil - when |italic= absent or not set; returns nil
in args.text.  Like |italic=unset, |italic=invert disables automatic italic markup.  Individual leading/trailing
default - for completeness, should rarely if ever be used; returns nil
apostrophes are converted to their html numeric entity equivalent so that the new italic markup doesn't become
yes - force args.text to be rendered in italic font; returns 'italic'
bold markup inadvertently.
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'
Leading and trailing wiki markup is extracted from args.text into separate table elements.  Addition, removal,
invert - disables font control so that font-style applied to text is dictated by markup outside or inverted inside the template; returns 'invert'
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 validate_italic (args)
local function invert_italics (source)
local properties = {['yes'] = 'italic', ['no'] = 'normal', ['unset'] = 'inherit', ['invert'] = 'invert', ['default'] = nil};
local invert_pattern_table = { -- leading/trailing markup add/remove/replace patterns
[""]="\'\'", -- empty string becomes italic markup
local count = 0
["\'\'"]="", -- italic markup becomes empty string
for _, arg in pairs {'italic', 'italics', 'i'} do
["\'\'\'"]="\'\'\'\'\'", -- bold becomes bold italic
if args[arg] then
["\'\'\'\'\'"]="\'\'\'", -- bold italic become bold
count = count + 1
};
end
local seg = {};
 
source = source:gsub ("%f[\']\'%f[^\']", '&#39;'); -- protect single quote marks from being interpreted as bold markup
 
seg[1] = source:match ('^(\'\'+%f[^\']).+') or ''; -- get leading markup, if any; ignore single quote
seg[3] = source:match ('.+(%f[\']\'\'+)$') or ''; -- get trailing markup, if any; ignore single quote
 
if '' ~= seg[1] and '' ~= seg[3] then -- extract the 'text'
seg[2] = source:match ('^\'\'+%f[^\'](.+)%f[\']\'\'+$') -- from between leading and trailing markup
elseif '' ~= seg[1] then
seg[2] = source:match ('^\'\'+%f[^\'](.+)') -- following leading markup
elseif '' ~= seg[3] then
seg[2] = source:match ('(.+)%f[\']\'\'+$') -- preceding trailing markup
else
seg[2] = source -- when there is no markup
end
end


if count > 1 then -- return nil and an error message if more than one is set
seg[1] = invert_pattern_table[seg[1]] or seg[1]; -- replace leading markup according to pattern table
return nil, 'only one of |italic=, |italics=, or |i= can be specified';
seg[3] = invert_pattern_table[seg[3]] or seg[3]; -- replace leading markup according to pattern table
end
 
return table.concat (seg); -- put it all back together and done
return properties[args.italic or args.italics or args.i], nil; -- return an appropriate value and a nil error message
end
end




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


Default behavior of the {{lang}} and {{lang-??}} templates is to add categorization when the templates are used in mainspace.
Validates |italic= or |italics= assigned values.
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
When |italic= is set and has an acceptible assigned value, return the matching css font-style property value or,
values for |nocat= and only negative values for |cat=; in both cases the 'other' sense (and non-sense) is not accepted and the
for the special case 'default', return nil.
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.
When |italic= is not set, or has an unacceptible assigned value, return nil and a nil error message.


Accepted values for |nocat= are the text strings:
When both |italic= and |italics= are set, returns nil and a 'conflicting' error message.
'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


]=]
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-??}} depending on
the individual template's requirements, 'italic' for {{transliteration}}) or to the value appropriate to |script=, if set ({{lang}}
and {{lang-??}} only).


local function validate_cat_args (args)
Accepted values and the values that this function returns are are:
if not (args.nocat or args.cat) then -- both are nil, so categorize
nil - when |italic= absent or not set; returns nil
return;
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
end
if false == yesno (args.cat) or true == yesno (args.nocat) then
return properties[args.italic or args.italics or args.i], nil; -- return an appropriate value and a nil error message
args.nocat = true; -- set to true when args.nocat is affirmative; nil else (as if the parameter were not set in the template)
else -- args.nocat is the parameter actually used.
args.nocat = nil;
end
end
end




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


Whether needle is in haystack
Default behavior of the {{lang}} and {{lang-??}} 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.


local function in_array ( needle, haystack )
Sets args.nocat to true if categorization is to be turned off; to nil if the default behavior should apply.
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


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


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


prettify ietf tags to use recommended subtag formats:
local function validate_cat_args (args)
code: lower case
if not (args.nocat or args.cat) then -- both are nil, so categorize
script: sentence case
return;
region: upper case
end
variant: lower case
private: lower case prefixed with -x-
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 format_ietf_tag (code, script, region, variant, private)
local function in_array (needle, haystack)
local out = {};
if needle == nil then
 
return false;
if is_set (private) then
return table.concat ({code:lower(), 'x', private:lower()}, '-'); -- if private, all other tags ignored
end
end
for n,v in ipairs (haystack) do
table.insert (out, code:lower());
if v == needle then
if is_set (script) then
return n;
script = script:lower():gsub ('^%a', string.upper);
end
table.insert (out, script);
end
end
return false;
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


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


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


extracts and returns IETF language tag parts:
local function format_ietf_tag (code, script, region, variant, private)
primary language subtag (required) - 2 or 3 character IANA language code
local out = {};
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
if is_set (private) then
lang lang-variant
return table.concat ({code:lower(), 'x', private:lower()}, '-'); -- if private, all other tags ignored
lang-script lang-script-variant
end
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
table.insert (out, code:lower());
if is_set (script) then
script = script:lower():gsub ('^%a', string.upper);
table.insert (out, script);
end


Languages with both two- and three-character code synonyms are promoted to the two-character synonym because
if is_set (region) then
the IANA registry file omits the synonymous three-character code; we cannot depend on browsers understanding
table.insert (out, region:upper());
the synonymous three-character codes in the lang= attribute.
end
 
For {{lang-??}} templates, the parameters |script=, |region=, and |variant= are supported (not supported in {{lang}}
if is_set (variant) then
because those parameters are superfluous to the IETF subtags in |code=)
table.insert (out, variant:lower());
end
return table.concat (out, '-');
end


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
--[[--------------------------< 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


local function get_ietf_parts (source, args_script, args_region, args_variant)
in any one of these forms
local code, script, region, variant, private; -- ietf tag parts
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.


if not is_set (source) then
Languages with both two- and three-character code synonyms are promoted to the two-character synonym because
return nil, nil, nil, nil, nil, 'missing language tag';
the IANA registry file omits the synonymous three-character code; we cannot depend on browsers understanding
end
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 acceptibe ietf tag patterns and short names of the ietf part captured by the pattern
local pattern = { -- table of tables holding acceptibe ietf tag patterns and short names of the ietf part captured by the pattern
Line 314: Line 407:


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


Line 320: Line 413:
if not (override_table[code] or lang_table[code] or synonym_table[code] or lang_dep_table[code]) then
if not (override_table[code] or lang_table[code] or synonym_table[code] or lang_dep_table[code]) then
return nil, nil, nil, nil, nil, table.concat ({'unrecognized language code: ', code}); -- invalid language code, don't know about the others (don't care?)
return nil, nil, nil, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_code, {code}); -- invalid language code, don't know about the others (don't care?)
end
end
 
if synonym_table[code] then -- if 639-2/639-2T code has a 639-1 synonym
if synonym_table[code] then -- if 639-2/639-2T code has a 639-1 synonym
table.insert (maint_cats, table.concat ({'Lang and lang-xx code promoted to ISO 639-1|', code}));
table.insert (maint_cats, substitute (cfg.get_ietf_parts_t.maint_promo_cat, {code}));
table.insert (maint_msgs, table.concat ({'code: ', code, ' promoted to code: ', synonym_table[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
code = synonym_table[code]; -- use the synonym
end
end
Line 331: Line 424:
if is_set (script) then
if is_set (script) then
if is_set (args_script) then
if is_set (args_script) then
return code, nil, nil, nil, nil, 'redundant script tag'; -- both code with script and |script= not allowed
return code, nil, nil, nil, nil, cfg.get_ietf_parts_t.redundant_scr; -- both code with script and |script= not allowed
end
end
else
else
Line 340: Line 433:
script = script:lower(); -- ensure that we use and return lower case version of this
script = script:lower(); -- ensure that we use and return lower case version of this
if not script_table[script] then
if not script_table[script] then
return code, nil, nil, nil, nil, table.concat ({'unrecognized script: ', script, ' for code: ', code}); -- language code ok, invalid script, don't know about the others (don't care?)
return code, nil, nil, nil, nil, 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
end
end
if suppressed_table[script] then -- ensure that code-script does not use a suppressed script
if suppressed_table[script] then -- ensure that code-script does not use a suppressed script
if in_array (code, suppressed_table[script]) then
if in_array (code, suppressed_table[script]) then
return code, nil, nil, nil, nil, table.concat ({'script: ', script, ' not supported for code: ', code}); -- language code ok, script is suppressed for this code
return code, nil, nil, nil, nil, substitute (cfg.get_ietf_parts_t.script_code, {script, code}); -- language code ok, script is suppressed for this code
end
end
end
end
Line 351: Line 444:
if is_set (region) then
if is_set (region) then
if is_set (args_region) then
if is_set (args_region) then
return code, nil, nil, nil, nil, 'redundant region tag'; -- both code with region and |region= not allowed
return code, nil, nil, nil, nil, cfg.get_ietf_parts_t.redundant_reg; -- both code with region and |region= not allowed
end
end
else
else
Line 360: Line 453:
region = region:lower(); -- ensure that we use and return lower case version of this
region = region:lower(); -- ensure that we use and return lower case version of this
if not region_table[region] then
if not region_table[region] then
return code, script, nil, nil, nil, table.concat ({'unrecognized region: ', region, ' for code: ', code});
return code, script, nil, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_reg_code, {region, code});
end
end
end
end
Line 366: Line 459:
if is_set (variant) then
if is_set (variant) then
if is_set (args_variant) then
if is_set (args_variant) then
return code, nil, nil, nil, nil, 'redundant variant tag'; -- both code with variant and |variant= not allowed
return code, nil, nil, nil, nil, cfg.get_ietf_parts_t.redundant_var; -- both code with variant and |variant= not allowed
end
end
else
else
Line 375: Line 468:
variant = variant:lower(); -- ensure that we use and return lower case version of this
variant = variant:lower(); -- ensure that we use and return lower case version of this
if not variant_table[variant] then -- make sure variant is valid
if not variant_table[variant] then -- make sure variant is valid
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant});
return code, script, region, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_var, {variant});
end -- does this duplicate/replace tests in lang() and lang_xx()?
end -- does this duplicate/replace tests in lang() and lang_xx()?
if is_set (script) then -- if script set it must be part of the 'prefix'
if is_set (script) then -- if script set it must be part of the 'prefix'
if not in_array (table.concat ({code, '-', script}), variant_table[variant]['prefixes']) then
if not in_array (table.concat ({code, '-', script}), variant_table[variant]['prefixes']) then
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code-script pair: ', code, '-', script});
return code, script, region, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_var_code_scr, {variant, code, script});
end
end
elseif is_set (region) then -- if region set, there are some prefixes that require lang code and region (en-CA-newfound)
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 (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)
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});
return code, script, region, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_var_code_reg, {variant, code, region});
end
end
end
end
else -- cheap way to determine if there are prefixes; fonipa and others don't have prefixes; # operator always returns 0
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
if variant_table[variant]['prefixes'][1] and not in_array (code, variant_table[variant]['prefixes']) then
return code, script, region, nil, nil, table.concat ({'unrecognized variant: ', variant, ' for code: ', code});
return code, script, region, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_var_code, {variant, code});
end
end
end
end
Line 397: Line 490:
private = private:lower(); -- ensure that we use and return lower case version of this
private = private:lower(); -- ensure that we use and return lower case version of this
if not override_table[table.concat ({code, '-x-', private})] then -- make sure private tag is valid; note that index  
if not override_table[table.concat ({code, '-x-', private})] then -- make sure private tag is valid; note that index  
return code, script, region, nil, nil, table.concat ({'unrecognized private tag: ', private});
return code, script, region, nil, nil, substitute (cfg.get_ietf_parts_t.unrecog_pri, {private});
end
end
end
end
Line 404: Line 497:




--[[--------------------------< S U B S T I T U T E >----------------------------------------------------------
--[=[-------------------------< M A K E _ W I K I L I N K >----------------------------------------------------


substitutes $1, $2, etc in <message> with data from <data_t>.  Returns plain-text substituted string when
Makes a wikilink; when both link and display text is provided, returns a wikilink in the form [[L|D]]; if only
<data_t> not nil; returns <message> else
link is provided, returns a wikilink in the form [[L]]; if neither are provided or link is omitted, returns an
empty string.


]]
]=]


local function substitute (message, data_t)
local function make_wikilink (link, display)
return data_t and mw.message.newRawMessage (message, data_t):plain() or message;
if is_set (link) then
if is_set (display) then
return table.concat ({'[[', link, '|', display, ']]'});
else
return table.concat ({'[[', link, ']]'});
end
else
return '';
end
end
end




--[[--------------------------< M A K E _ E R R O R _ M S G >--------------------------------------------------
--[[--------------------------< D I V _ M A R K U P _ A D D >--------------------------------------------------


assembles an error message from template name, message text, help link, and error category.
Adds <i> and </i> tags to list-item text or to implied <p>..</p> text. Mixed not supported.


]]
]]


local function make_error_msg (msg, args_t, template)
local function div_markup_add (text, style)
local category;
if text:find ('^\n[%*:;#]') then -- look for list markup; list markup must begin at start of text
if 'italic' == style then
if 'Transliteration' == template then
return text:gsub ('(\n[%*:;#]+)([^\n]+)', '%1<i>%2</i>'); -- insert italic markup at each list item
category = 'Transliteration';
else
else
return text;
category = 'Lang and lang-xx'
end
end
end
local category_link = ((0 == namespace or 10 == namespace) and not args_t.nocat) and table.concat ({'[[Category:', category, ' template errors]]'}) or '';
return substitute ('[$1] <span style="color:#d33">Error: {{$2}}: $3 ([[:Category:$4 template errors|help]])</span>$5',
{
args_t.text or 'undefined',
template,
msg,
category,
category_link
})
end


--[=[-------------------------< M A K E _ W I K I L I N K >----------------------------------------------------
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
Makes a wikilink; when both link and display text is provided, returns a wikilink in the form [[L|D]]; if only
 
link is provided, returns a wikilink in the form [[L]]; if neither are provided or link is omitted, returns an
if 'italic' == style then
empty string.
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)
 
]=]
 
local function make_wikilink (link, display)
if is_set (link) then
if is_set (display) then
return table.concat ({'[[', link, '|', display, ']]'});
else
else
return table.concat ({'[[', link, ']]'});
text = text:gsub ('[^\n]+', '<p>%1</p>'); -- insert p markup at each impled p
text = text:gsub ('\n', ''); -- strip newline characters
end
end
else
return '';
end
end
return text;
end
end




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


adds <i> and </i> tags to list-item text or to implied <p>..</p> text. mixed not supported
<tag> holds a string 'div' or 'span' used to choose the correct wrapping tag.


]]
]]


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


if text:find ('^\n[%*:;#]') then -- look for list markup; list markup must begin at start of text
return table.concat (wrapper_t); -- make a big string and done
if 'italic' == style then
end
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
--[[--------------------------< M A K E _ T E X T _ H T M L >--------------------------------------------------
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;
Add the html markup to text according to the type of content that it is: <span> or <i> tags for inline content or
end
<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>


--[[--------------------------< T I T L E _ W R A P P E R _ M A K E >------------------------------------------
makes a <span title="<title text>"><content_text></span> or <div title="<title text>"><content_text></div> where
<title text> is in the tool-tip in the wiki's local language and <content_text> is non-local-language text in
html markup.  This because the lang= attibute applies to the content of its enclosing tag.
<tag> holds a string 'div' or 'span' used to choose the correct wrapping tag
]]
]]


local function title_wrapper_make (title_text, content_text, tag)
local function make_text_html (code, text, tag, rtl, style, size, language)
local wrapper_t = {};
local html_t = {};
local style_added = '';
table.insert (wrapper_t, table.concat ({'<', tag})); -- open opening wrapper tag
local wrapper_tag = tag; -- <tag> gets modified so save a copy for use when/if we create a wrapper span or div
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
if text:match ('^%*') then
end
table.insert (html_t, '&#42;'); -- move proto language text prefix outside of italic markup if any; use numeric entity because plain splat confuses MediaWiki
text = text:gsub ('^%*', ''); -- remove the splat from the text
end


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


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


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


The lang= attribute also applies to the content of the tag where it is placed so this is wrong because 'Spanish
if 'normal' == style then -- when |italic=no
language text' is English:
table.insert (html_t, ' style=\"font-style: normal;'); -- override external markup, if any
<i lang="es" title="Spanish language text">casa</i>
style_added = '\"'; -- remember that style attribute added and is not yet closed
should be:
end
<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>


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


local function make_text_html (code, text, tag, rtl, style, size, language)
table.insert (html_t, table.concat ({style_added, '>'})); -- close the opening html tag
local html_t = {};
table.insert (html_t, text); -- insert the text
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
table.insert (html_t, table.concat ({'</', tag, '>'})); -- close the 'text' <i>, <span>, or <div> html tag


if text:match ('^%*') then
if is_set (language) then -- create a <title_text> string for the title= attribute in a wrapper span or div
table.insert (html_t, '&#42;'); -- move proto language text prefix outside of italic markup if any; use numeric entity because plain splat confuses MediaWiki
local title_text;
text = text:gsub ('^%*', ''); -- remove the splat from the text
if 'zxx' == code then -- special case for this tag 'no linguistic content'
end
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


if 'span' == tag then -- default html tag for inline content
return title_wrapper_make (title_text, table.concat (html_t), wrapper_tag);
if 'italic' == style then -- but if italic
else
tag = 'i'; -- change to <i> tags
return table.concat (html_t);
end
else -- must be div so go
text = div_markup_add (text, style); -- handle implied <p>, implied <p> with <i>, and list markup (*;:#) with <i>
end
end
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
--[=[-------------------------< M A K E _ C A T E G O R Y >----------------------------------------------------
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
For individual language, <language>, returns:
table.insert (html_t, ' dir="ltr"'); -- add direction attribute for left-to-right languages
[[Category:Articles containing <language>-language text]]
end
 
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 <language> languages]]
 
]=]


if 'normal' == style then -- when |italic=no
local function make_category (code, language_name, nocat, name_get)
table.insert (html_t, ' style=\"font-style: normal;'); -- override external markup, if any
if ((0 ~= namespace) or nocat) and not name_get then -- only categorize in article space
style_added = '\"'; -- remember that style attribute added and is not yet closed
return ''; -- return empty string for concatenation
end
end


if is_set (size) then -- when |size=<something>
if mw.ustring.find (language_name, 'languages', 1, true) then
if is_set (style_added) then
return substitute ('[[$1 $2]]', {cfg.make_category_t.collective_cat, language_name});
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
end
 
table.insert (html_t, table.concat ({style_added, '>'})); -- close the opening html tag
if this_wiki_lang_tag == code then
table.insert (html_t, text); -- insert the text
return substitute ('[[$1 $2 $3-$4]]', { -- unique category name for the local language
 
cfg.make_category_t.cat_prefix,
table.insert (html_t, table.concat ({'</', tag, '>'})); -- close the 'text' <i>, <span>, or <div> html tag
cfg.make_category_t.explicit_cat,
 
language_name,
if is_set (language) then -- create a <title_text> string for the title= attribute in a wrapper span or div
cfg.make_category_t.cat_postfix,
local title_text;
});
if 'zxx' == code then -- special case for this tag 'no linguistic content'
title_text = table.concat ({language, ' text'}); -- not a language so don't use 'language' in title text
elseif mw.ustring.find (language, 'languages', 1, true) then
title_text = table.concat ({language, ' collective text'}); -- for collective languages
else
title_text = table.concat ({language, '-language text'}); -- for individual languages
end
 
return title_wrapper_make (title_text, table.concat (html_t), wrapper_tag);
else
else
return table.concat (html_t);
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
end
end




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


For individual language, <language>, returns:
Return translit <i lang=xx-Latn>...</i> where xx is the language code; else return empty string.
[[Category:Articles containing <language>-language text]]


for English:
The value |script= is not used in {{transliteration}} for this purpose; instead it uses |code.  Because language scripts
[[Category:Articles containing explicitly cited English-language text]]
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.
for ISO 639-2 collective languages (and for 639-1 bh):
[[Category:Articles with text in <language> languages]]


]=]
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).


local function make_category (code, language_name, nocat, name_get)
This function is used by both lang_xx() and transl()
local cat = {};
lang_xx() always provides code, language_name, and translit; may provide tscript; never provides style
local retval;
transl() always provides language_name, translit, and one of code or tscript, never both; always provides style


if ((0 ~= namespace) or nocat) and not name_get then -- only categorize in article space
For {{transliteration}}, style only applies when a language code is provided.
return ''; -- return empty string for concatenation
end


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


if this_wiki_lang_tag == code then
local function make_translit (code, language_name, translit, std, tscript, style, engvar)
table.insert (cat, 'explicitly cited ' .. language_name); -- unique category name for the local language
local title_t = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
else
local title_text = ''; -- tool tip text for title= attribute
table.insert (cat, language_name);
end
std = std and std:lower(); -- lower case for table indexing
table.insert (cat, '-language text]]');
 
return table.concat (cat);
end
 
 
--[[--------------------------< M A K E _ T R A N S L I T >----------------------------------------------------
 
return translit <i lang=xx-Latn>...</i> where xx is the language code; else return empty string
 
The value |script= is not used in {{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 transl()
lang_xx() always provides code, language_name, and translit; may provide tscript; never provides style
transl() always provides language_name, translit, and one of code or tscript, never both; always provides style
 
For {{transliteration}}, style only applies when a language code is provided
]]
 
local function make_translit (code, language_name, translit, std, tscript, style, engvar)
local title;
local out_t = {};
local title_t = lang_data.translit_title_table; -- table of transliteration standards and the language codes and scripts that apply to those standards
local title_text = ''; -- tool tip text for title= attribute
std = std and std:lower(); -- lower case for table indexing
if not is_set (std) and not is_set (tscript) then -- when neither standard nor script specified
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
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)
if not mw.ustring.find (language_name, 'languages', 1, true) then -- collective language names (plural 'languages' is part of the name)
title_text = title_text .. '-language'; -- skip this text (individual and macro languages only)
title_text = substitute ('$1-$2', {title_text, cfg.misc_text_t.language}); -- skip this text (individual and macro languages only)
end
end
title_text = substitute ('$1 $2', {title_text, lang_data.engvar_t[engvar]['romanisz_lc']}); -- finish the tool tip; use romanization when neither script nor standard supplied
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
elseif is_set (std) and is_set (tscript) then -- when both are specified
Line 674: Line 726:
if title_t[std][tscript] then -- and if script for that standard is legitimate
if title_t[std][tscript] then -- and if script for that standard is legitimate
if script_table[tscript] then
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
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
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?
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?
Line 689: Line 747:
if title_t[std][code] then -- if language code is in the table (transl may not provide a language code)
if title_t[std][code] then -- if language code is in the table (transl may not provide a language code)
title_text = title_text .. table.concat ({title_t[std][code:lower()], ' (', language_name, ' language) transliteration'}); -- add the appropriate text to the tool tip
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
else -- code doesn't match
title_text = title_text .. title_t[std]['default']; -- so use the standard's default
title_text = title_text .. title_t[std]['default']; -- so use the standard's default
Line 701: Line 765:
else
else
if is_set (tscript) then
if is_set (tscript) then
title_text = title_text .. table.concat ({language_name, '-script transliteration'}); -- write a script tool tip
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
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)
if not mw.ustring.find (language_name, 'languages', 1, true) then -- collective language names (plural 'languages' is part of the name)
title_text = title_text .. '-language'; -- skip this text (individual and macro languages only)
title_text = substitute ('$1-$2', { -- skip this text (individual and macro languages only)
title_text,
cfg.misc_text_t.language,
});
end
end
title_text = title_text .. ' transliteration'; -- finish the tool tip
title_text = substitute ('$1 $2', { -- finish the tool tip
title_text,
cfg.make_translit_t.transliteration,
});
else
else
title_text = title_text .. ' transliteration'; -- generic tool tip (can we ever get here?)
title_text = substitute ('$1 $2', { -- generic tool tip (can we ever get here?)
title_text,
cfg.make_translit_t.transliteration,
});
end
end
end
end
end
end


local close_tag;
local tag
if is_set (code) then -- when a language code is provided (always with {{lang-??}} templates, not always with {{transliteration}})
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
code = code:match ('^(%a%a%a?)'); -- strip all subtags leaving only the language subtag
if not style then -- nil for the default italic style
if not style then -- nil for the default italic style
table.insert (out_t, "<i lang=\""); -- so use <i> tag
tag = '<i lang="%s-Latn">%s</i>'
close_tag = '</i>'; -- tag to be used when closing
else
else
table.insert (out_t, table.concat ({'<span style=\"font-style: ', style, '\" lang=\"'})); -- non-standard style, construct a span tag for it
tag = '<span style="font-style: %s" lang="%s-Latn">%s</span>' -- non-standard style, construct a span tag for it
close_tag = '</span>'; -- tag to be used when closing
tag = string.format(tag, style, "%s", "%s")
end
end
table.insert (out_t, code);
tag = string.format(tag, code, "%s")
table.insert (out_t, "-Latn\">"); -- transliterations are always Latin script
else
else
table.insert (out_t, "<span>"); -- when no language code: no lang= attribute, not italic ({{transliteration}} only)
tag = '<span>%s</span>' -- when no language code: no lang= attribute, not italic ({{transliteration}} only)
close_tag = '</span>';
end
end
 
table.insert (out_t, translit); -- add the translit text
tag = string.format(tag, 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
if '' == title_text then -- when there is no need for a tool-tip
return table.concat (out_t); -- make a string and done
return tag; -- done
else
else
title_text = 'gb_t' == engvar and title_text:gsub ('([Rr]omani)z', '%1s') or title_text; -- gb eng when engvar specifies gb eng; us eng 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, table.concat (out_t), 'span'); -- wrap with a tool-tip span and done
return title_wrapper_make (title_text, tag, 'span'); -- wrap with a tool-tip span and done
end
end
end
end
Line 746: Line 819:


This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
This function checks the content of args.text and returns empty string if nothing is amiss else it returns an
error message. The tests are for empty or missing text and for improper or disallowed use of apostrophe markup.
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
Italic rendering is controlled by the |italic= template parameter so italic markup should never appear in args.text
Line 755: Line 828:
local function validate_text (template, args)
local function validate_text (template, args)
if not is_set (args.text) then
if not is_set (args.text) then
return make_error_msg ('no text', args, template);
return make_error_msg (cfg.validate_text_t.no_text, args, template);
end
end


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


local style = args.italic;
local style = args.italic;


if ('unset' ~= style) and ('invert' ~=style) then
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
if args.text:find ("%f[\']\'\'%f[^\']") or args.text:find ("%f[\']\'\'\'\'\'%f[^\']") then -- italic but not bold, or bold italic
return make_error_msg ('text has italic markup', args, template);
return make_error_msg (cfg.validate_text_t.italic_markup, args, template);
end
end
end
end
Line 774: Line 847:
--[[--------------------------< R E N D E R _ M A I N T >------------------------------------------------------
--[[--------------------------< R E N D E R _ M A I N T >------------------------------------------------------


render mainenance messages and categories
Render mainenance messages and categories.


]]
]]


local function render_maint(nocat)
local function render_maint (nocat)
local maint = {};
local maint = {};
Line 801: Line 874:
--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------
--[[--------------------------< P R O T O _ P R E F I X >------------------------------------------------------


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


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


]]
]]
Line 823: Line 896:
--[[--------------------------< H A S _ P O E M _ T A G >------------------------------------------------------
--[[--------------------------< H A S _ P O E M _ T A G >------------------------------------------------------


looks for a poem strip marker in text; returns true when found; false else
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
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.
the content that will replace the stripmarker.


Line 880: Line 953:
--[[--------------------------< V A L I D A T E _ P R O T O >--------------------------------------------------
--[[--------------------------< 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,
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
empty string (or parameter omitted) returns as nil; any other value returns nil with <proto_param> as second
true indicating that some other value has been assigned to |proto=
return value for use in error message.


]]
]]


local function validate_proto (proto_param)
local function validate_proto (proto_param)
if 'yes' == proto_param then
if cfg.keywords_t.affirmative == proto_param then
return true;
return true;
elseif 'no' == proto_param then
elseif cfg.keywords_t.negative == proto_param then
return false;
return false;
elseif is_set (proto_param) then
elseif is_set (proto_param) then
return nil, true; -- |proto= something other than 'yes' or 'no'
return nil, proto_param; -- |proto= something other than 'yes' or 'no'
else
else
return nil; -- missing or empty
return nil; -- missing or empty
Line 901: Line 974:
--[[--------------------------< L A N G U A G E _ N A M E _ G E T >--------------------------------------------
--[[--------------------------< 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
Common function to return language name from the data set according to IETF tag.


returns language name if found in data tables; nil else
Returns language name if found in data tables; nil else.


]]
]]
Line 923: Line 996:


if lang_dep_table[code] and cat then -- because deprecated code may have been overridden to en.wiki preferred name
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_cats, substitute (cfg.language_name_get_t.deprecated_cat, {code}));
table.insert (maint_msgs, table.concat ({'code: ', code, ' is deprecated'}));
table.insert (maint_msgs, substitute (cfg.language_name_get_t.deprecated_msg, {code}));
end
end


Line 931: Line 1,004:




--[[--------------------------< _ L A N G >--------------------------------------------------------------------
--[[--------------------------< 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 <text>.  This module does not attempt to know all scripts and
what they look like.  It does know what Latn script looks like so when <text> is written using other than the Latn
script, the IETF script subtag, if present, should not be Latn.


entry point for {{lang}}
Conversely, when <text> is written using the Latn script, the IETF script subtag, if present, should be Latn.


there should be no reason to set parameters in the {{lang}} {{#invoke:}}
For the purposes of this test, Latf (Fraktur) and Latg (Gaelic) are considered to be equivalent to Latn because
<includeonly>{{#invoke:lang|lang}}</includeonly>
unicode treats these two scripts as font-specific variants of Latn.


parameters are received from the template's frame (parent frame)
Returns an error message when mismatch detected; nil else.


]]
]]


local function _lang (args)
local function text_script_match_test (script, is_latn_text, pos)
local out = {};
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
local language_name; -- used to make category names
if is_set (script) then -- don't bother with the rest of this if <script> is nil or empty string
local category_name; -- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
script = script:lower(); -- lower case to index into <scripts_t>
local subtags = {}; -- IETF subtags script, region, variant, and private
if is_latn_text then -- when text is wholly Latn script
local code; -- the language code
if not scripts_t[script] then -- but a non-Latn script is specified
local msg; -- for error messages
return cfg.text_script_match_test_t.latn_txt_mismatch; -- emit an error message
local tag = 'span'; -- initial value for make_text_html()
end
local template = args.template or 'Lang';
else -- when text is not wholly Latn script
if scripts_t[script] then -- but a Latn script is specified
return substitute (cfg.text_script_match_test_t.latn_scr_mismatch, {pos}); -- emit an error message with position of first offending character
end
end
end
end
 
 
--[[--------------------------< I S _ L A T I N >--------------------------------------------------------------
 
Created because unicode does not have Latin theta 'θ' character.  The Greek 'θ' is, apparently, commonly used
in certain romanizations. This function attempts to discover if <text> script is Latn with Greek 'θ' or some
other script with Greek 'θ'.


maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
when <text> is or has Greek 'θ':
maint_msgs = {};
for {{lang}}, {{langx}}, and {{lang-??}}, returns boolean true when <text>:
has theta and all other characters are Latn
has theta as the only character, <script> is Latn
validate_cat_args (args); -- determine if categorization should be suppressed
for {{transliteration}}, returns boolean true when <text>:
has theta and all other characters are Latn
has theta as the only character
]]


if args[1] and args.code then
local function is_latin (text, script, transl)
return make_error_msg ('conflicting: {{{1}}} and |code=', args, template);
text = mw.text.decode (text); -- convert html entities to characters
else
script = (script and script:lower()) or nil; -- ensure lowercase for comparisons
args.code = args[1] or args.code; -- prefer args.code
local whitelist = '[ʻʼʽʾʿΔαβγδθλσφχϑъьӾӿᾱῑ῾上入去平阳阴]'
 
if not transl then -- testing {{lang}} or {{langx}} <text>
if '' == mw.ustring.gsub (text, whitelist, '') then -- when <text> uses only whitelisted characters
return ('latn' == script); -- true when latn script; false when other script or no script
end
end
end


if args[2] and args.text then
if mw.ustring.find (text, whitelist) then -- does <text> contain characters from the whitelist?
return make_error_msg ('conflicting: {{{2}}} and |text=', args, template);
text = mw.ustring.gsub (text, whitelist, 'x'); -- replace whitelisted characters from <text> with known Latn-script char 'x'
else
-- if 0 == text:len() then -- will be zero if theta was the only character in <text>
args.text = args[2] or args.text; -- prefer args.text
if '' == text then -- will be empty string if the only character(s) in <text> are in <whitelist>
if transl then -- not nil for {{transliteration}}; assume Latn because this is 'transliteration' template
return true;
end
return (script and ('latn' == script)) or false; -- {{lang}}, {{langx}}, and {{lang-??}}; true when script is Latn; false else
end
return unicode.is_Latin (text); -- return true when all characters in modified <text> are Latn script; false else
end
end
msg = validate_text (template, args); -- ensure that |text= is set
if is_set (msg) then -- msg is an already-formatted error message
return msg;
end
args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly


args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
return unicode.is_Latin (text); -- return true when all characters in <text> are Latn script; false else
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


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


if msg then
Entry point for {{lang}}.
return make_error_msg ( msg, args, template);
end


args.italic, msg = validate_italic (args);
There should be no reason to set parameters in the {{lang}} {{#invoke:}}
if msg then
<includeonly>{{#invoke:lang|lang}}</includeonly>
return make_error_msg (msg, args, template);
end


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


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


if 'invert' == args.italic and 'span' == tag then -- invert only supported for in-line content
local function _lang (args)
args.text = invert_italics (args.text)
local out = {};
end
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
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat
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 cfg.templates_t.lang;


table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, language_name));
maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
table.insert (out, make_category (code, language_name, args.nocat));
maint_msgs = {};
table.insert (out, render_maint(args.nocat)); -- maintenance messages and categories
validate_cat_args (args); -- determine if categorization should be suppressed


return table.concat (out); -- put it all together and done
if args[1] and args.code then
end
return make_error_msg (substitute (cfg.lang_t.conflict_n_param, {'1', cfg.lang_t.conflict_n_param_types.code}), args, template);
else
args.code = args[1] or args.code; -- prefer args.code
end


if args[2] and args.text then
return make_error_msg (substitute (cfg.lang_t.conflict_n_param, {'2', cfg.lang_t.conflict_n_param_types.text}), args, template);
else
args.text = args[2] or args.text; -- prefer args.text
end


--[[--------------------------< L A N G >----------------------------------------------------------------------
msg = validate_text (template, args); -- ensure that |text= is set
 
if is_set (msg) then -- msg is an already-formatted error message
entry point for {{lang}}
return msg;
end
args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly


there should be no reason to set parameters in the {{lang}} {{#invoke:}}
args.rtl = args.rtl == cfg.keywords_t.affirmative; -- convert to boolean: 'yes' -> true, other values -> false
<includeonly>{{#invoke:lang|lang}}</includeonly>


parameters are received from the template's frame (parent frame)
args.proto, msg = validate_proto (args.proto); -- return boolean, or nil, or nil and error message flag
if msg then
return make_error_msg (substitute (cfg.lang_t.invalid_proto, {msg}), 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}})


local function lang (frame)
if msg then
local args = getArgs (frame, { -- this code so that we can detect and handle wiki list markup in text
return make_error_msg (msg, args, template);
valueFunc = function (key, value)
end
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);
local is_latn_text, pos = is_latin (args.text, subtags.script); -- make a boolean
end


msg = text_script_match_test (subtags.script, is_latn_text, pos)
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args, template);
end


--[[--------------------------< T R A N S L A T I O N _ M A K E >----------------------------------------------
args.italic, msg = validate_italic (args);
if msg then
return make_error_msg (msg, args, template);
end


stand-alone function to create literal translation of main text
if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
 
if ('latn' == subtags.script) or -- script is latn
Also used by {{lang-x2}}
(this_wiki_lang_tag ~= code and not is_set (subtags.script) and not has_poem_tag (args.text) and is_latn_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
]]
 
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
else
table.insert (translation_t, make_wikilink ('Literal translation', 'lit.')); -- linked form
args.italic = 'inherit'; -- italic not set; script not latn; inherit current style
end
end
if is_set (subtags.script) then -- if script set, override rtl setting
if in_array (subtags.script, lang_data.rtl_scripts) then
args.rtl = true; -- script is an rtl script
else
args.rtl = false; -- script is not an rtl script
end
end
table.insert (translation_t, "&thinsp;</small>"); -- close the <small> html tag
end
end
table.insert (translation_t, table.concat ({'&#39;', args_t.translation, '&#39;'})); -- use html entities to avoid wiki markup confusion
return table.concat (translation_t); -- make a big string and done
end


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


--[[--------------------------< _ L A N G _ X X >--------------------------------------------------------------
if cfg.keywords_t.invert == args.italic and 'span' == tag then -- invert only supported for in-line content
args.text = invert_italics (args.text)
end


For the {{lang-??}} templates, the only parameter required to be set in the template is the language code. All
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat
other parameters can, usually should, be written in the template call. For {{lang-??}} templates for languages
that can have multiple writing systems, it may be appropriate to set |script= as well.


For each {{lang-??}} template choose the appropriate entry-point function so that this function knows the default
table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, language_name));
styling that should be applied to text.
table.insert (out, make_category (code, language_name, args.nocat));
table.insert (out, render_maint (args.nocat)); -- maintenance messages and categories


For normal, upright style:
return table.concat (out); -- put it all together and done
<includeonly>{{#invoke:lang|lang_xx_inherit|code=xx}}</includeonly>
end
For italic style:
<includeonly>{{#invoke:lang|lang_xx_italic|code=xx}}</includeonly>


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


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


For {{lang-??}}, the positional parameters are:
Parameters are received from the template's frame (parent frame).
{{{1}}} text
{{{2}}} transliterated text
{{{3}}} literal translation text
no other positional parameters are allowed


]]
]]


local function _lang_xx (args, base_template) -- base_template will be either of 'Langx' or 'Lang-xx'
local function lang (frame)
local out = {};
local args_t = getArgs (frame, { -- this code so that we can detect and handle wiki list markup in text
local language_name; -- used to make display text, article links
valueFunc = function (key, value)
local category_name; -- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
if 2 == key or 'text' == key then -- the 'text' parameter; do not trim wite space
local subtags = {}; -- IETF subtags script, region, and variant
return value; -- return untrimmed 'text'
local code; -- the language code
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
});


local translit_script_name; -- name associated with IANA (ISO 15924) script code
args_t.fn = nil; -- unset because not supported but this function might have been called by {{lang|fn=lang|...}}
local translit;
local translit_title;
local msg; -- for error messages
local tag = 'span'; -- initial value for make_text_html()
local template = args.template or base_template;


maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
local msg = parameter_validate (args_t, cfg.templates_t.lang); -- verify that all supplied parameters are supported by {{lang-??}}
maint_msgs = {};
if msg then
return make_error_msg (msg, args_t, cfg.templates_t.lang); -- when template has unsupported params, abandon with error message
local text_idx = ('Langx' == base_template) and 2 or 1; -- for {{langx}} 'text' positional parameter is '2'
end
local translit_idx = ('Langx' == base_template) and 3 or 2;
local xlate_idx = ('Langx' == base_template) and 4 or 3;


if args[text_idx] and args.text then
return _lang (args_t);
return make_error_msg ('conflicting: {{{' .. text_idx .. '}}} and |text=', args, template);
end
else
args.text = args[text_idx] or args.text; -- prefer positional 'text' parameter
end


msg = validate_text (template, args); -- ensure that |text= is set, does not contain italic markup and is protected from improper bolding
if is_set (msg) then
return msg;
end


args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly
--[[--------------------------< T R A N S L A T I O N _ M A K E >----------------------------------------------


if args[translit_idx] and args.translit then
Stand-alone function to create literal translation of main text.
return make_error_msg ('conflicting: {{{'.. translit_idx .. '}}} and |translit=', args, template);
else
args.translit = args[translit_idx] or args.translit -- prefer positional 'translit' parameter
end


args.engvar = lang_data.engvar_sel_t[args.engvar] or 'us_t'; -- either 'gb_t' or 'us_t' when |engvar= valid; 'us_t' else
Also used by {{lang-x2}}.


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


if args.links and args.link then
local function translation_make (args_t)
return make_error_msg ('conflicting: |links= and |link=', args, template);
local translation_t = {', '}; -- initialize output
else
 
args.link = args.link or args.links; -- prefer args.link
if 'none' ~= args_t.label then -- if we want a label
table.insert (translation_t, '<small>'); -- open the <small> html tag
if cfg.keywords_t.negative == args_t.link then
table.insert (translation_t, substitute ('<abbr title="$1">$2</abbr>', {cfg.translation_make_t.lit_xlation, cfg.translation_make_t.lit_abbr})); -- unlinked form
else
table.insert (translation_t, make_wikilink (cfg.translation_make_t.lit_xlation, cfg.translation_make_t.lit_abbr)); -- linked form
end
table.insert (translation_t, "&thinsp;</small>"); -- close the <small> html tag
end
end
table.insert (translation_t, table.concat ({'&#39;', args_t.translation, '&#39;'})); -- use html entities to avoid wiki markup confusion
return table.concat (translation_t); -- make a big string and done
end


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


args.rtl = args.rtl == 'yes'; -- convert to boolean: 'yes' -> true, other values -> false
--[[--------------------------< _ L A N G _ X X >--------------------------------------------------------------


code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code, args.script, args.region, args.variant); -- private omitted because private
For the {{lang-??}} templates, the only parameter required to be set in the template is the language code. All
other parameters can, usually should, be written in the template call. For {{lang-??}} templates for languages
that can have multiple writing systems, it may be appropriate to set |script= as well.


if msg then -- if an error detected then there is an error message
For each {{lang-??}} template choose the appropriate entry-point function so that this function knows the default
return make_error_msg (msg, args, template);
styling that should be applied to text.
end
args.italic, msg = validate_italic (args);
if msg then
return make_error_msg (msg, args, template);
end
-- TODO: this appears to work so we might in future dispense inherit_t in Module:Lang/langx
-- retain existing system until {{langx}} has replaced {{lang-??}}
-- if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
-- if ('latn' == subtags.script) or -- script is latn
-- (this_wiki_lang_tag ~= code and not is_set (subtags.script) and unicode.is_Latin (args.text)) then -- text is not this wiki's language, no script specified and is wholly latn script (auto-italics)
-- args.italic = 'italic'; -- set font-style:italic
-- else
-- args.italic = 'inherit'; -- italic not set; script not latn; inherit current style
-- end
-- end


-- TODO: retain this system until {{langx}} has replaced {{lang-??}}
For normal, upright style:
if nil == args.italic then -- args.italic controls
<includeonly>{{#invoke:lang|lang_xx_inherit|code=xx}}</includeonly>
if is_set (subtags.script) then
For italic style:
if 'latn' == subtags.script then
<includeonly>{{#invoke:lang|lang_xx_italic|code=xx}}</includeonly>
args.italic = 'italic'; -- |script=Latn; set for font-style:italic
 
else
All other parameters should be received from the template's frame (parent frame)
args.italic = initial_style_state; -- italic not set; script is not latn; set for font-style:<initial_style_state>
 
end
Supported parameters are:
else
|code = (required) the IANA language code
args.italic = initial_style_state; -- here when |italic= and |script= not set; set for font-style:<initial_style_state>
|script = IANA script code; especially for use with languages that use multiple writing systems
end
|region = IANA region code
|variant = IANA variant code
|text = (required) the displayed text in language specified by code
|link = boolean false ('no') does not link code-spcified language name to associated language article
|rtl = boolean true ('yes') identifies the language specified by code as a right-to-left language
|nocat = boolean true ('yes') inhibits normal categorization; error categories are not affected
|cat = boolian false ('no') opposite form of |nocat=
|italic = boolean true ('yes') renders displayed text in italic font; boolean false ('no') renders displayed text in normal font; not set renders according to initial_style_state
|lit = text that is a literal translation of text
|label = 'none' to suppress all labeling (language name, 'translit.', 'lit.')
any other text replaces language-name label - automatic wikilinking disabled
for those {{lang-??}} templates that support transliteration (those templates where |text= is not entirely latn script):
|translit = text that is a transliteration of text
|translit-std = the standard that applies to the transliteration
|translit-script = ISO 15924 script name; falls back to code
 
For {{lang-??}}, the positional parameters are:
{{{1}}} text
{{{2}}} transliterated text
{{{3}}} literal translation text
no other positional parameters are allowed
 
]]
 
local function _lang_xx (args, base_template) -- base_template will be either of 'Langx' or 'Lang-xx'
local out = {};
local language_name; -- used to make display text, article links
local category_name; -- same as language_name except that it retains any parenthetical disambiguators (if any) from the data set
local subtags = {}; -- IETF subtags script, region, and variant
local code; -- the language code
 
local translit_script_name; -- name associated with IANA (ISO 15924) script code
local translit;
local msg; -- for error messages
local tag = 'span'; -- initial value for make_text_html()
local template = args.template or base_template;
 
maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
maint_msgs = {};
local text_idx = (cfg.templates_t.langx == base_template) and 2 or 1; -- for {{langx}} 'text' positional parameter is '2'
local translit_idx = (cfg.templates_t.langx == base_template) and 3 or 2;
local xlate_idx = (cfg.templates_t.langx == base_template) and 4 or 3;
 
if args[text_idx] and args.text then
return make_error_msg (substitute (cfg.lang_t.conflict_n_param, {translit_idx, cfg.lang_t.conflict_n_param_types.text}), args, template);
else
args.text = args[text_idx] or args.text; -- prefer positional 'text' parameter
end
 
msg = validate_text (template, args); -- ensure that |text= is set, does not contain italic markup and is protected from improper bolding
if is_set (msg) then
return msg;
end
 
args.text, tag = html_tag_select (args.text); -- inspects text; returns appropriate html tag with text trimmed accordingly
 
if args[translit_idx] and args.translit then
return make_error_msg (substitute (cfg.lang_t.conflict_n_param, {translit_idx, cfg.lang_t.conflict_n_param_types.translit}), args, template);
 
else
args.translit = args[translit_idx] or args.translit -- prefer positional 'translit' parameter
end
 
if args.translit then
local latn, pos = is_latin (args.translit, nil, true);
if not latn then
return make_error_msg (substitute (cfg.lang_xx_t.translit_nonlatn, {pos}), args, template);
end
end
 
args.engvar = cfg.engvar_sel_t[args.engvar] or cfg.default_engvar; -- |engvar= when valid; cfg.default_engvar else
 
if args[xlate_idx] and (args.translation or args.lit) then
return make_error_msg (substitute (cfg.lang_xx_t.conflict_n_lit, {translit_idx}), args, template);
elseif args.translation and args.lit then
return make_error_msg (cfg.lang_xx_t.conflict_lit, args, template);
else
args.translation = args[xlate_idx] or args.translation or args.lit; -- prefer positional 'translation' parameter
end
 
if args.links and args.link then
return make_error_msg (cfg.lang_xx_t.conflict_link, args, template);
else
args.link = args.link or args.links; -- prefer args.link
end
 
validate_cat_args (args); -- determine if categorization should be suppressed
 
args.rtl = args.rtl == cfg.keywords_t.affirmative; -- convert to boolean: 'yes' -> true, other values -> false
 
code, subtags.script, subtags.region, subtags.variant, subtags.private, msg = get_ietf_parts (args.code, args.script, args.region, args.variant); -- private omitted because private
 
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args, template);
end
local is_latn_text, pos = is_latin (args.text, subtags.script); -- make a boolean
 
msg = text_script_match_test (subtags.script, is_latn_text, pos)
if msg then -- if an error detected then there is an error message
return make_error_msg (msg, args, template);
end
 
args.italic, msg = validate_italic (args);
if msg then
return make_error_msg (msg, args, template);
end
if cfg.templates_t.langx == base_template then -- auto-italics  for {{langx}} templates; adapted from {{lang}} (no support for poem tag)
if nil == args.italic then -- nil when |italic= absent or not set or |italic=default; args.italic controls
if ('latn' == subtags.script) or -- script is latn
(this_wiki_lang_tag ~= code and not is_set (subtags.script) and is_latn_text) then -- text is not this wiki's language, no script specified and is wholly latn script (auto-italics)
args.italic = 'italic'; -- set font-style:italic
else
args.italic = 'inherit'; -- italic not set; script not latn; inherit current style
end
end
else -- {{lang-xx}} does not do auto italics; retained for those wikis that don't support {{langx}}
if nil == args.italic then -- args.italic controls
if is_set (subtags.script) then
if 'latn' == subtags.script then
args.italic = 'italic'; -- |script=Latn; set for font-style:italic
else
args.italic = initial_style_state; -- italic not set; script is not latn; set for font-style:<initial_style_state>
end
else
args.italic = initial_style_state; -- here when |italic= and |script= not set; set for font-style:<initial_style_state>
end
end
end
if is_set (subtags.script) then -- if script set override rtl setting
if in_array (subtags.script, lang_data.rtl_scripts) then
args.rtl = true; -- script is an rtl script
else
args.rtl = false; -- script is not an rtl script
end
end
 
args.proto, msg = validate_proto (args.proto); -- return boolean, or nil, or nil and error message flag
if msg then
return make_error_msg (substitute (cfg.lang_t.invalid_proto, {args.proto}), args, template);
end
 
args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles
language_name = language_name_get (args.code, code, true); -- get language name; try ietf tag first, then code w/o variant then code w/ variant
 
category_name = language_name; -- category names retain IANA parenthetical diambiguators (if any)
language_name = language_name:gsub ('%s+%b()', ''); -- remove IANA parenthetical disambiguators or qualifiers from names that have them
 
if args.label then
if 'none' ~= args.label then
table.insert (out, table.concat ({args.label, ': '})); -- custom label
end
else
if cfg.keywords_t.negative == args.link then
table.insert (out, language_name); -- language name without wikilink
else
if mw.ustring.find (language_name, 'languages', 1, true) then
table.insert (out, make_wikilink (language_name)); -- collective language name uses simple wikilink
elseif lang_data.article_name[args.code:lower()] then -- is ietf tag in article name over ride?
table.insert (out, make_wikilink (lang_data.article_name[args.code:lower()], language_name)); -- language name with wikilink from override data
elseif lang_data.article_name[code] then -- is language tag in article override
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 (substitute ('$1 $2', {language_name, cfg.misc_text_t.language}), language_name)); -- language name with wikilink
end
end
table.insert (out, ': '); -- separator
end
 
if cfg.keywords_t.invert == args.italic then
args.text = invert_italics (args.text)
end
end
if is_set (subtags.script) then -- if script set override rtl setting
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat
if in_array (subtags.script, lang_data.rtl_scripts) then
 
args.rtl = true; -- script is an rtl script
table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, ('none' == args.label) and language_name or nil))
else
args.rtl = false; -- script is not an rtl script
end
end


args.proto, msg = validate_proto (args.proto); -- return boolean, or nil, or nil and error message flag
if is_set (args.translit) and not is_latn_text then -- transliteration; not supported when args.text is wholly latn text (this is an imperfect test)
if msg then
table.insert (out, ', '); -- comma to separate text from translit
return make_error_msg (table.concat ({'invalid |proto=: ', args.proto}), args, template);
if 'none' ~= args.label then
end
table.insert (out, '<small>');
if script_table[args['translit-script']] then -- when |translit-script= is set, try to use the script's name
translit_script_name = script_table[args['translit-script']];
else
translit_script_name = language_name; -- fall back on language name
end


args.code = format_ietf_tag (code, subtags.script, subtags.region, subtags.variant, subtags.private); -- format to recommended subtag styles
local translit_title_obj = mw.title.makeTitle (0, substitute ('$1 $2',{cfg.lang_xx_t.romanization, translit_script_name})); -- make a title object; no engvar, article titles use US spelling
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 translit_title_obj.exists and (cfg.keywords_t.negative ~= args.link) then
table.insert (out, make_wikilink (substitute ('$1 $2', {cfg.lang_xx_t.romanization, translit_script_name or language_name}),
substitute ('$1', {cfg.engvar_t[args.engvar]['romanisz_pt']})) .. ':'); -- make a wikilink if there is an article to link to; engvar the display text


category_name = language_name; -- category names retain IANA parenthetical diambiguators (if any)
else
language_name = language_name:gsub ('%s+%b()', ''); -- remove IANA parenthetical disambiguators or qualifiers from names that have them
table.insert (out, substitute ('$1:', {cfg.engvar_t[args.engvar]['romanisz_pt']})); -- else plain text per engvar


if args.label then
end
if 'none' ~= args.label then
table.insert (out, '&nbsp;</small>'); -- close the small tag
table.insert (out, table.concat ({args.label, ': '})); -- custom label
end
end
else
if 'no' == args.link then
translit = make_translit (args.code, language_name, args.translit, args['translit-std'], args['translit-script'], nil, args.engvar)
table.insert (out, language_name); -- language name without wikilink
if is_set (translit) then
table.insert (out, translit);
else
else
if mw.ustring.find (language_name, 'languages', 1, true) then
return make_error_msg (substitute ('$1: $2', {cfg.lang_xx_t.invalid_xlit_std, args['translit-std'] or '[missing]'}), args, template);
table.insert (out, make_wikilink (language_name)); -- collective language name uses simple wikilink
elseif lang_data.article_name[args.code:lower()] then -- is ietf tag in article name over ride?
table.insert (out, make_wikilink (lang_data.article_name[args.code:lower()], language_name)); -- language name with wikilink from override data
elseif lang_data.article_name[code] then -- is language tag in article override
table.insert (out, make_wikilink (lang_data.article_name[code], language_name)); -- language name with wikilink from override data
else
table.insert (out, make_wikilink (language_name .. ' language', language_name)); -- language name with wikilink
end
end
end
table.insert (out, ': '); -- separator
elseif is_set (args.translit) and is_latn_text then -- when creating a transliteration of Latn script <text>
return make_error_msg (cfg.lang_xx_t.xlit_of_latn, args, template); -- emit error message and abandon
end
end


if 'invert' == args.italic then
if is_set (args.translation) then -- translation (not supported in {{lang}})
args.text = invert_italics (args.text)
table.insert (out, translation_make (args));
end
args.text = proto_prefix (args.text, language_name, args.proto); -- prefix proto-language text with a splat
 
table.insert (out, make_text_html (args.code, args.text, tag, args.rtl, args.italic, args.size, ('none' == args.label) and language_name or nil))
 
if is_set (args.translit) and not unicode.is_Latin (args.text) then -- transliteration (not supported in {{lang}}); not supported when args.text is wholly latn text (this is an imperfect test)
table.insert (out, ', '); -- comma to separate text from translit
if 'none' ~= args.label then
table.insert (out, '<small>');
if script_table[args['translit-script']] then -- when |translit-script= is set, try to use the script's name
translit_script_name = script_table[args['translit-script']];
else
translit_script_name = language_name; -- fall back on language name
end
translit_title = mw.title.makeTitle (0, table.concat ({'Romanization of ', translit_script_name})); -- make a title object; no engvar, article titles use US spelling
if translit_title.exists and ('no' ~= args.link) then
table.insert (out, make_wikilink ('Romanization of ' .. translit_script_name or language_name,
substitute ('$1', {lang_data.engvar_t[args.engvar]['romanisz_pt']})) .. ':'); -- make a wikilink if there is an article to link to; engvar the display text
 
else
table.insert (out, substitute ('$1:', {lang_data.engvar_t[args.engvar]['romanisz_pt']})); -- else plain text per engvar
 
end
table.insert (out, '&nbsp;</small>'); -- close the small tag
end
translit = make_translit (args.code, language_name, args.translit, args['translit-std'], args['translit-script'], nil, args.engvar)
if is_set (translit) then
table.insert (out, translit);
else
return make_error_msg (table.concat ({'invalid translit-std: \'', args['translit-std'] or '[missing]'}), args, template);
end
end
if is_set (args.translation) then -- translation (not supported in {{lang}})
table.insert (out, translation_make (args));
end
end
Line 1,311: Line 1,494:
--[[--------------------------< L A N G _ X X _ A R G S _ G E T >----------------------------------------------
--[[--------------------------< L A N G _ X X _ A R G S _ G E T >----------------------------------------------


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


returns table of args
Returns table of args.


text positional parameters are not trimmed here but are selectively trimmed at html_tag_select()
Text positional parameters are not trimmed here but are selectively trimmed at html_tag_select().


]]
]]
Line 1,324: Line 1,507:
parentFirst= true, -- parameters in the template override parameters set in the {{#invoke:}}
parentFirst= true, -- parameters in the template override parameters set in the {{#invoke:}}
valueFunc = function (key, value)
valueFunc = function (key, value)
if (('Langx' == base_template) and 2 or 1) == key then -- the 'text' positional parameter; 1 for {{lang-??}}, 2 for {{langx}}; do not trim wite space
if ((cfg.templates_t.langx == base_template) and 2 or 1) == key then -- the 'text' positional parameter; 1 for {{lang-??}}, 2 for {{langx}}; do not trim wite space
return value; -- return untrimmed 'text' positional parameter
return value; -- return untrimmed 'text' positional parameter
elseif value then -- all other values: if the value is not nil
elseif value then -- all other values: if the value is not nil
Line 1,342: Line 1,525:
--[[--------------------------< L A N G _ X X _ I T A L I C >--------------------------------------------------
--[[--------------------------< L A N G _ X X _ I T A L I C >--------------------------------------------------


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


]]
]]


local function lang_xx_italic (frame)
local function lang_xx_italic (frame)
local args = lang_xx_args_get (frame, 'lang-xx');
local args = lang_xx_args_get (frame, cfg.templates_t.langxx);
args.fn = nil; -- unset because not supported but this function might have been called by {{lang|fn=lang_xx_italic|...}}
local msg = parameter_validate (args, cfg.templates_t.langxx); -- verify that all supplied parameters are supported by {{lang-??}}
if msg then
return make_error_msg (msg, args, cfg.templates_t.langxx); -- when template has unsupported params, abandon with error message
end
initial_style_state = 'italic';
initial_style_state = 'italic';
return _lang_xx (args, 'Lang-xx') .. '[[Category:Pages using Lang-xx templates]]'; -- temporary category
return _lang_xx (args, cfg.templates_t.langxx) .. '[[Category:Pages using Lang-xx templates]]'; -- temporary category
end
end


Line 1,356: Line 1,545:
--[[--------------------------< _ L A N G _ X X _ I T A L I C >------------------------------------------------
--[[--------------------------< _ L A N G _ X X _ I T A L I C >------------------------------------------------


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


]]
]]
Line 1,362: Line 1,551:
local function _lang_xx_italic (args)
local function _lang_xx_italic (args)
initial_style_state = 'italic';
initial_style_state = 'italic';
return _lang_xx (args, 'Lang-xx');
return _lang_xx (args, cfg.templates_t.langxx);
end
end


Line 1,368: Line 1,557:
--[[--------------------------< L A N G _ X X _ I N H E R I T >------------------------------------------------
--[[--------------------------< L A N G _ X X _ I N H E R I T >------------------------------------------------


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


]]
]]


local function lang_xx_inherit (frame)
local function lang_xx_inherit (frame)
local args = lang_xx_args_get (frame, 'lang-xx');
local args = lang_xx_args_get (frame, cfg.templates_t.langxx);
args.fn = nil; -- unset because not supported but this function might have been called by {{lang|fn=lang_xx_inherit|...}}
 
local msg = parameter_validate (args, cfg.templates_t.langxx); -- verify that all supplied parameters are supported by {{lang-??}}
if msg then
return make_error_msg (msg, args, cfg.templates_t.langxx); -- when template has unsupported params, abandon with error message
end


initial_style_state = 'inherit';
initial_style_state = 'inherit';
return _lang_xx (args, 'Lang-xx') .. '[[Category:Pages using Lang-xx templates]]'; -- temporary category
return _lang_xx (args, cfg.templates_t.langxx) .. '[[Category:Pages using Lang-xx templates]]'; -- temporary category
end
end


Line 1,382: Line 1,577:
--[[--------------------------< _ L A N G _ X X _ I N H E R I T >----------------------------------------------
--[[--------------------------< _ 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.
Entry point from another module. Sets the initial style state to inherit.


]]
]]
Line 1,388: Line 1,583:
local function _lang_xx_inherit (args)
local function _lang_xx_inherit (args)
initial_style_state = 'inherit';
initial_style_state = 'inherit';
return _lang_xx (args, 'Lang-xx');
return _lang_xx (args, cfg.templates_t.langxx);
end
end


Line 1,400: Line 1,595:
local function _langx (args_t)
local function _langx (args_t)
local langx_data = mw.loadData ('Module:Lang/langx'); -- get necessary data  
local langx_data = mw.loadData ('Module:Lang/langx'); -- get necessary data  
local inherit_t = langx_data.inherit_t; -- get list of language tags extracted from {{lang-??}} template names for languages that are rendered in upright font
local rtl_t = langx_data.rtl_t; -- get list of language tags for languages that are rendered right-to-left
local rtl_t = langx_data.rtl_t; -- get list of language tags for languages that are rendered right-to-left
local script_t = langx_data.script_t; -- get list of language tags for {{lang-??}} templates that set |script=<something>
local link_t = langx_data.link_t; -- get list of language tags for {{lang-??}} templates that set |link=<something>
local link_t = langx_data.link_t; -- get list of language tags for {{lang-??}} templates that set |link=<something>
local size_t = langx_data.size_t; -- get list of language tags for {{lang-??}} templates that set |size=<something>
local size_t = langx_data.size_t; -- get list of language tags for {{lang-??}} templates that set |size=<something>
local msg = parameter_validate (args_t, cfg.templates_t.langx);
if msg then
return make_error_msg (msg, args_t, cfg.templates_t.langx);
end


args_t.code = args_t[1] or args_t.code; -- get the language tag; must be {{{1}}} or |code=
args_t.code = args_t[1] or args_t.code; -- get the language tag; must be {{{1}}} or |code=
if not args_t.code then
if not args_t.code then
return make_error_msg ('missing language tag', args_t, 'Langx');
return make_error_msg (cfg.get_ietf_parts_t.missing_lang_tag, args_t, cfg.templates_t.langx);
end
end
args_t.rtl = args_t.rtl or (rtl_t[args_t.code] and 'yes'); -- prefer |rtl= in template call, use rtl_t else
args_t.rtl = args_t.rtl or (rtl_t[args_t.code] and cfg.keywords_t.affirmative); -- prefer |rtl= in template call, use rtl_t else
args_t.script = args_t.script or script_t[args_t.code]; -- prefer |script= in template call, use script_t else
args_t.link = args_t.link or link_t[args_t.code]; -- prefer |link= in template call, use link_t felse
args_t.link = args_t.link or link_t[args_t.code]; -- prefer |link= in template call, use link_t felse
args_t.size = args_t.size or size_t[args_t.code]; -- prefer |size= in template call, use size_t else
args_t.size = args_t.size or size_t[args_t.code]; -- prefer |size= in template call, use size_t else
Line 1,419: Line 1,616:


local lang_subtag = args_t.code; -- use only the base language subtag for unsupported tag test; some args_t.code are modified by |script= etc
local lang_subtag = args_t.code; -- use only the base language subtag for unsupported tag test; some args_t.code are modified by |script= etc
initial_style_state = inherit_t[args_t.code:lower()] and 'inherit' or 'italic'; -- if listed in inherit_t, set as 'inherit'; 'italic' else
return _lang_xx (args_t, cfg.templates_t.langx);
return _lang_xx (args_t, 'Langx') .. ((langx_data.unsupported_t[lang_subtag:lower()] and '[[Category:Langx uses unsupported language tag|'.. lang_subtag .. ']]') or ''); -- temporary category for unsupported language tags
-- return _lang_xx (args_t, 'Langx');
end
end


Line 1,429: Line 1,624:
Entry point for {{langx}}.
Entry point for {{langx}}.


this function calls _lang_xx() to render non-English text. The {{lang-??}} templates have three positional paramters
this function calls _lang_xx() to render non-English text. The {{lang-??}} templates have three positional paramters
but {{langx}} has four:
but {{langx}} has four:


|  1 |  2 |  3 |  4
|  1 |  2 |  3 |  4
{{lang-xx |<text> |<transl> |<xlate> }}
{{lang-xx |<text> |<transl> |<xlate> }}
{{langx |<tag> |<text> |<transl> |<xlate> }}
{{langx |<tag> |<text> |<transl> |<xlate> }}


The calls to lang_xx_args_get() and _lang_xx() use 'Langx' as a flag for those functions to select the proper
The calls to lang_xx_args_get() and _lang_xx() use 'Langx' as a flag for those functions to select the proper
positional parameters.
positional parameters.


{{lang-??}} depends on the calling template to select 'inherit' or 'italic' as the default renderer.
{{lang-??}} depends on the calling template to select 'inherit' or 'italic' to establish the default rendering.
{{langx}} can't do that so, intead, relies on the list of tags scraped from the {{lang-??}} templates that call
 
lang_xx_inherit().
{{langx}} can't do that.  The initial version of {{langx}} relied on a list of language tags (inherit_t in ~/langx)
scraped from those {{lang-??}} templates that call lang_xx_inherit() to render text in upright font. Langx now
uses auto-italics code adapted from {{lang}} (doesn't support poem tags).


]]
]]


local function langx (frame)
local function langx (frame)
local args_t = lang_xx_args_get (frame, 'Langx'); -- get the arguments; 'Langx' is the <base_template> used to decide which positional param is 'text', 'translit', 'lit'
local args_t = lang_xx_args_get (frame, cfg.templates_t.langx); -- get the arguments; 'Langx' is the <base_template> used to decide which positional param is 'text', 'translit', 'lit'
return _langx (args_t);
return _langx (args_t);
Line 1,458: Line 1,655:
All code combinations supported by {{lang}} and the {{lang-??}} templates are supported by this function.
All code combinations supported by {{lang}} and the {{lang-??}} templates are supported by this function.


Module entry point from another module
Module entry point from another module.


]]
]]
Line 1,472: Line 1,669:
--[[--------------------------< I S _ I E T F _ T A G >--------------------------------------------------------
--[[--------------------------< I S _ I E T F _ T A G >--------------------------------------------------------


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


]]
]]
Line 1,502: Line 1,699:
Set invoke's |link= parameter to yes to get wikilinked version of the language name.
Set invoke's |link= parameter to yes to get wikilinked version of the language name.


Module entry point from another module
Module entry point from another module.


]]
]]
Line 1,509: Line 1,706:
local subtags = {}; -- IETF subtags script, region, variant, and private
local subtags = {}; -- IETF subtags script, region, variant, and private
local raw_code = args[1]; -- save a copy of the input IETF subtag
local raw_code = args[1]; -- save a copy of the input IETF subtag
local link = 'yes' == args['link']; -- make a boolean
local link = cfg.keywords_t.affirmative == args['link']; -- make a boolean
local label = args.label;
local label = args.label;
local code; -- the language code
local code; -- the language code
Line 1,518: Line 1,715:
if msg then
if msg then
local template = (args['template'] and table.concat ({'{{', args['template'], '}}: '})) or ''; -- make template name (if provided by the template)
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>'});
return make_error_span (template, msg);
end
end


Line 1,524: Line 1,721:
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_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 'yes' ~= args.raw then
language_name = language_name:gsub ('%s+%b()', ''); -- remove IANA parenthetical disambiguators or qualifiers from names that have them
end


if link then -- when |link=yes, wikilink the language name
if link then -- when |link=yes, wikilink the language name
Line 1,534: Line 1,733:
language_name = make_wikilink (lang_data.article_name[code], label or language_name); -- language name with wikilink from override data
language_name = make_wikilink (lang_data.article_name[code], label or language_name); -- language name with wikilink from override data
else
else
language_name = make_wikilink (language_name .. ' language', label or language_name); -- language name with wikilink
language_name = make_wikilink (substitute ('$1 $2', {language_name, cfg.misc_text_t.language}), label or language_name); -- language name with wikilink
end
end
end
end
Line 1,544: Line 1,743:
--[[--------------------------< N A M E _ F R O M _ T A G >----------------------------------------------------
--[[--------------------------< N A M E _ F R O M _ T A G >----------------------------------------------------


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


]]
]]
Line 1,559: Line 1,758:
that disambiguator must be omitted (they are not present in the data name-to-tag tables).
that disambiguator must be omitted (they are not present in the data name-to-tag tables).


Module entry point from another module
Module entry point from another module.


]]
]]
Line 1,574: Line 1,773:
return tag, true; -- language name found so return tag and done; second return used by is_lang_name()
return tag, true; -- language name found so return tag and done; second return used by is_lang_name()
else
else
msg = 'language: ' .. args[1] .. ' not found' -- language name not found, error message
msg = substitute (cfg.tag_from_name_t.lang_not_found, {args[1]}); -- language name not found, error message
end
end
else
else
msg = 'missing language name' -- language name not provided, error message
msg = cfg.tag_from_name_t.missing_lang_name; -- language name not found, error message
end
end


Line 1,584: Line 1,783:
template = table.concat ({'{{', args['template'], '}}: '}); -- make template name (if provided by the template)
template = table.concat ({'{{', args['template'], '}}: '}); -- make template name (if provided by the template)
end
end
return table.concat ({'<span style=\"color:#d33\">Error: ', template, msg, '</span>'});
return make_error_span (template, msg);
 
end
end


Line 1,590: Line 1,790:
--[[--------------------------< T A G _ F R O M _ N A M E >----------------------------------------------------
--[[--------------------------< T A G _ F R O M _ N A M E >----------------------------------------------------


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


]]
]]
Line 1,602: Line 1,802:
--[[--------------------------< I S _ L A N G _ N A M E >------------------------------------------------------
--[[--------------------------< I S _ L A N G _ N A M E >------------------------------------------------------


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


]]
]]
Line 1,614: Line 1,814:
--[[--------------------------< _ T R A N S L >----------------------------------------------------------------
--[[--------------------------< _ T R A N S L >----------------------------------------------------------------


Module entry point from another module
Module entry point from another module.


]]
]]
Line 1,624: Line 1,824:
local script; -- IANA script
local script; -- IANA script
local msg; -- for when called functions return an error message
local msg; -- for when called functions return an error message
local template = args.template or cfg.templates_t.transliteration;


maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
maint_cats = {}; -- initialize because when this module required into another module, these only declared once so only initialzed once
Line 1,633: Line 1,834:


if not title_table[args.translit_std] then
if not title_table[args.translit_std] then
return make_error_msg (table.concat ({'unrecognized transliteration standard: ', args.translit_std}), args, 'Transliteration');
return make_error_msg (substitute (cfg.transl_t.unrecog_xlit_std, {args.translit_std}), args, template);
end
end
else
else
Line 1,641: Line 1,842:
if args[1] and (args[1]:match ('^%a%a%a?%a?$') or -- args[2] missing; is args[1] a language or script tag or is it the transliterated text?
if args[1] and (args[1]:match ('^%a%a%a?%a?$') or -- args[2] missing; is args[1] a language or script tag or is it the transliterated text?
args[1]:match ('^%a%a%a?%-x%-')) then -- or is args[1] a private-use tag
args[1]:match ('^%a%a%a?%-x%-')) then -- or is args[1] a private-use tag
return make_error_msg ('no text', args, 'Transliteration'); -- args[1] is a code so we're missing text
return make_error_msg (cfg.transl_t.no_text, args, template); -- args[1] is a code so we're missing text
else
else
args.text = args[1]; -- args[1] is not a code so we're missing that; assign args.text for error message
args.text = args[1]; -- args[1] is not a code so we're missing that; assign args.text for error message
return make_error_msg ('missing language / script code', args, 'Transliteration');
return make_error_msg (cfg.transl_t.missing_lang_scr, args, template);
end
end
end
end
Line 1,653: Line 1,854:
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
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
else
return make_error_msg (table.concat ({'unrecognized language / script code: ', args[1]}), args, 'Transliteration'); -- invalid language / script code
return make_error_msg (substitute (cfg.transl_t.unrecog_lang_scr, {args[1]}), args, template); -- invalid language / script code
end
end
else
else
return make_error_msg ('missing language / script code', args, 'Transliteration'); -- missing language / script code so quit
return make_error_msg (cfg.transl_t.missing_lang_scr, args, template); -- missing language / script code so quit
end
end
local latn, pos = is_latin (args.text, nil, true);
if not latn then -- text is not latn
return make_error_msg (substitute (cfg.lang_xx_t.translit_nonlatn, {pos}), args, template);
end


args.italic, msg = validate_italic (args);
args.italic, msg = validate_italic (args);
if msg then
if msg then
return make_error_msg (msg, args, 'Transliteration');
return make_error_msg (msg, args, template);
end
end
Line 1,668: Line 1,875:
end
end


args.engvar = lang_data.engvar_sel_t[args.engvar] or 'us_t'; -- either 'gb_t' or 'us_t' when |engvar= valid; 'us_t' else
args.engvar = cfg.engvar_sel_t[args.engvar] or cfg.default_engvar; -- |engvar= when valid; cfg.default_engvar else


if override_table[args.code] then -- is code a language code defined in the override table?
if override_table[args.code] then -- is code a language code defined in the override table?
Line 1,682: Line 1,889:
args.code = ''; -- unset because not a language code
args.code = ''; -- unset because not a language code
else
else
return make_error_msg (table.concat ({'unrecognized language / script code: ', args.code}), args, 'Transliteration'); -- invalid language / script code
return make_error_msg (substitute (cfg.transl_t.unrecog_lang_scr, {args.code}), args, template); -- invalid language / script code
end
end
-- here only when all parameters passed to make_translit() are valid
-- here only when all parameters passed to make_translit() are valid
Line 1,691: Line 1,898:
--[[--------------------------< T R A N S L >------------------------------------------------------------------
--[[--------------------------< T R A N S L >------------------------------------------------------------------


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


]]
]]
Line 1,702: Line 1,909:
--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------
--[[--------------------------< 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
Returns category name associated with IETF language tag if valid; error message else.


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


Module entry point from another module
Module entry point from another module.


]]
]]
Line 1,713: Line 1,920:
local subtags = {}; -- IETF subtags script, region, variant, and private
local subtags = {}; -- IETF subtags script, region, variant, and private
local raw_code = args_t[1]; -- save a copy of the input IETF subtag
local raw_code = args_t[1]; -- save a copy of the input IETF subtag
local link = 'yes' == args_t.link; -- make a boolean
local link = cfg.keywords_t.affirmative == args_t.link; -- make a boolean
local label = args_t.label;
local label = args_t.label;
local code; -- the language code
local code; -- the language code
Line 1,722: Line 1,929:
if msg then
if msg then
local template = (args_t.template and table.concat ({'{{', args_t.template, '}}: '})) or ''; -- make template name (if provided by the template)
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>'});
return make_error_span (template, msg);
end
end


Line 1,739: Line 1,946:
--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------
--[[--------------------------< C A T E G O R Y _ F R O M _ T A G >--------------------------------------------


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


]]
]]