维基词典

模組:Scripts

下列說明文檔位於Module:Scripts/doc。[編輯]
相關連結：子頁面列表 • 链入 • 嵌入包含 • 測試用例 • 沙盒

This module is used to retrieve and manage Wiktionary's various writing systems and the information associated with them. See Wiktionary:Scripts for more information.

The information itself is stored in Module:scripts/data. The data module should not be used directly by any other module, the data should only be accessed through the functions provided by Module:scripts.

For functions that allow templates to use this module, see Module:scripts/templates.

Finding and retrieving scripts

The module exports a number of functions that are used to find scripts.

getByCode

getByCode(code)

Finds the script whose code matches the one provided. If it exists, it returns a Script object representing the script. Otherwise, it returns nil.

findBestScript

findBestScript(text, lang)

Given some text and a language object, this function iterates through the scripts of the given language and tries to find the script that best matches the text. It returns a Script object representing the script. If no match is found at all, it returns the None script object.

Script objects

A Script object is returned from one of the functions above. It is a Lua representation of a script and the data associated with it. It has a number of methods that can be called on it, using the : syntax. For example: 

local m_scripts = require("Module:scripts")
local sc = m_scripts.getByCode("Latn")
local name = sc:getCanonicalName()
-- "name" will now be "Latin"

Script:getCode

:getCode()

Returns the script code of the language. Example: "Cyrl" for Cyrillic.

Script:getCanonicalName

:getCanonicalName()

Returns the canonical name of the script. This is the name used to represent that script on Wiktionary. Example: "Cyrillic" for Cyrillic.

Script:getParent

:getParent()

Returns the parent of the script. Example: "Latn" for "Latinx" and "Arab" for "fa-Arab". It returns "top" for scripts without a parent, like "Latn", "Grek", etc.

Script:getWikipediaArticle

:getWikipediaArticle()

Returns the wikipedia_article item in the language's data file, or else calls Script:getCategoryName().

Script:countCharacters

:countCharacters(text)

Returns the number of characters in the text that are part of this script.

Note: You should never rely on text consisting entirely of the same script. Strings may contain spaces, punctuation and even wiki markup or HTML tags. HTML tags will skew the counts, as they contain Latin-script characters. So it's best to avoid them.

Script:getCharacters

:getCharacters()

Returns the regex defining the script's characters from the language's data file.

This can be used to search for words consisting only of this script, but see the warning above.

Script:getCategoryName

:getCategoryName()

Returns the name of the main category of that script. Example: "Cyrillic script" for Cyrillic, whose category is at Category:Cyrillic script.

Script:getDirection

:getDirection()

Returns the text direction, if any. Currently, left-to-right scripts are unmarked, while most right-to-left scripts have direction specified as "rtl" and Mongolian as "down".

Subpages

See also

local export = {}

local combining_classes_module = "Module:Unicode data/combining classes"
local data_module = "Module:scripts/data"
local debug_track_module = "Module:debug/track"
local json_module = "Module:JSON"
local language_like_module = "Module:language-like"
local languages_error_module = "Module:languages/error"
local scripts_by_name_module = "Module:scripts/by name"
local scripts_chartoscript_module = "Module:scripts/charToScript"
local string_utilities_module = "Module:string utilities"
local table_module = "Module:table"
local writing_systems_module = "Module:writing systems"

local concat = table.concat
local gmatch = string.gmatch
local insert = table.insert
local load_data = mw.loadData
local match = string.match
local select = select
local setmetatable = setmetatable
local toNFC = mw.ustring.toNFC
local toNFD = mw.ustring.toNFD
local toNFKC = mw.ustring.toNFKC
local toNFKD = mw.ustring.toNFKD
local type = type

--[==[
Loaders for functions in other modules, which overwrite themselves with the target function when called. This ensures modules are only loaded when needed, retains the speed/convenience of locally-declared pre-loaded functions, and has no overhead after the first call, since the target functions are called directly in any subsequent calls.]==]
	local function deepcopy(...)
		deepcopy = require(table_module).deepcopy
		return deepcopy(...)
	end
	
	local function explode(...)
		explode = require(string_utilities_module).explode_utf8
		return explode(...)
	end
	
	local function get_writing_system(...)
		get_writing_system = require(writing_systems_module).getByCode
		return get_writing_system(...)
	end
	
	local function keys_to_list(...)
		keys_to_list = require(table_module).keysToList
		return keys_to_list(...)
	end
	
	local function languages_error(...)
		languages_error = require(languages_error_module)
		return languages_error(...)
	end
	
	local function split(...)
		split = require(string_utilities_module).split
		return split(...)
	end
	
	local function to_json(...)
		to_json = require(json_module).toJSON
		return to_json(...)
	end
	
	local function track(...)
		track = require(debug_track_module)
		return track(...)
	end
	
	local function ugsub(...)
		ugsub = require(string_utilities_module).gsub
		return ugsub(...)
	end
	
	local function umatch(...)
		umatch = require(string_utilities_module).match
		return umatch(...)
	end

--[==[
Loaders for objects, which load data (or some other object) into some variable, which can then be accessed as "foo or get_foo()", where the function get_foo sets the object to "foo" and then returns it. This ensures they are only loaded when needed, and avoids the need to check for the existence of the object each time, since once "foo" has been set, "get_foo" will not be called again.]==]
	local scripts_by_name
	local function get_scripts_by_name()
		scripts_by_name, get_scripts_by_name = load_data(scripts_by_name_module), nil
		return scripts_by_name
	end
	
	local scripts_data
	local function get_scripts_data()
		scripts_data, get_scripts_data = load_data(data_module), nil
		return scripts_data
	end

local Script = {}
Script.__index = Script

--[==[Returns the script code of the script. Example: {{lua|"Cyrl"}} for Cyrillic.]==]
function Script:getCode()
	return self._code
end

--[==[Returns the canonical name of the script. This is the name used to represent that script on Wiktionary. Example: {{lua|"Cyrillic"}} for Cyrillic.]==]
function Script:getCanonicalName()
	return self._rawData[1] or self._rawData.canonicalName
end

--[==[Returns the display form of the script. For scripts, this is the same as the value returned by <code>:getCategoryName("nocap")</code>, i.e. it reads "NAME script" (e.g. {{lua|"Arabic script"}}). The displayed text used in <code>:makeCategoryLink</code> is always the same as the display form.]==]
function Script:getDisplayForm()
	return self:getCategoryName("nocap")
end

function Script:getOtherNames(onlyOtherNames)
	Script.getOtherNames = require(language_like_module).getOtherNames
	return self:getOtherNames(onlyOtherNames)
end

function Script:getAliases()
	return self._rawData.aliases or {}
end

function Script:getVarieties(flatten)
	Script.getVarieties = require(language_like_module).getVarieties
	return self:getVarieties(flatten)
end

--[==[Returns the {{w|IETF language tag#Syntax of language tags|IETF subtag}} used for the script, which should always be a valid {{w|ISO 15924}} script code. This is used when constructing HTML {{code|html|lang{{=}}}} tags. The {{lua|ietf_subtag}} value from the script's data file is used, if present; otherwise, the script code is used. For script codes which contain a hyphen, only the part after the hyphen is used (e.g. {{lua|"fa-Arab"}} becomes {{lua|"Arab"}}).]==]
function Script:getIETFSubtag()
	local code = self._ietf_subtag
	if code == nil then
		code = self._rawData.ietf_subtag or match(self._code, "[^%-]+$")
		self._ietf_subtag = code
	end
	return code
end

--[==[Returns the parent of the script. Example: {{lua|"Arab"}} for {{lua|"fa-Arab"}}. It returns {{lua|"top"}} for scripts without a parent, like {{lua|"Latn"}}, {{lua|"Grek"}}, etc.]==]
function Script:getParent()
	return self._rawData.parent
end

function Script:getSystemCodes()
	if not self._systemCodes then
		local system_codes = self._rawData[3]
		if type(system_codes) == "table" then
			self._systemCodes = system_codes
		elseif type(system_codes) == "string" then
			self._systemCodes = split(system_codes, ",", true, true)
		else
			self._systemCodes = {}
		end
	end
	return self._systemCodes
end

function Script:getSystems()
	if not self._systemObjects then
		self._systemObjects = {}
		
		for _, system in ipairs(self:getSystemCodes()) do
			insert(self._systemObjects, get_writing_system(system))
		end
	end
	
	return self._systemObjects
end

--[==[Check whether the script is of type `system`, which can be a writing system code or object. If multiple systems are passed, return true if the script is any of the specified systems.]==]
function Script:isSystem(...)
	for _, system in ipairs{...} do
		if type(system) == "table" then
			system = system:getCode()
		end
		for _, s in ipairs(self:getSystemCodes()) do
			if system == s then
				return true
			end
		end
	end
	return false
end

--function Script:getAllNames()
--	return self._rawData.names
--end

--[==[Returns a table of types as a lookup table (with the types as keys).

Currently, the only possible type is {script}.]==]
function Script:getTypes()
	local types = self._types
	if types == nil then
		types = {script = true}
		local rawtypes = self._rawData.type
		if rawtypes then
			for t in gmatch(rawtypes, "[^,]+") do
				types[t] = true
			end
		end
		self._types = types
	end
	return types
end

--[==[Given a list of types as strings, returns true if the script has all of them.

Use {{lua|hasType("script")}} to determine if an object that may be a language, family or script is a script.]==]	
function Script:hasType(...)
	local args, types = {...}, self:getTypes()
	for i = 1, #args do
		if not types[args[i]] then
			return false
		end
	end
	return true
end

--[==[Returns the name of the main category of that script. Example: {{lua|"Cyrillic script"}} for Cyrillic, whose category is at [[:Category:Cyrillic script]].
Unless optional argument <code>nocap</code> is given, the script name at the beginning of the returned value will be capitalized. This capitalization is correct for category names, but not if the script name is lowercase and the returned value of this function is used in the middle of a sentence. (For example, the script with the code <code>Semap</code> has the name <code>"flag semaphore"</code>, which should remain lowercase when used as part of the category name [[:Category:Translingual letters in flag semaphore]] but should be capitalized in [[:Category:Flag semaphore templates]].) If you are considering using <code>getCategoryName("nocap")</code>, use <code>getDisplayForm()</code> instead.]==]
function Script:getCategoryName(nocap)
	local name = self:getCanonicalName()
	-- If the name already has "script", "code" or "semaphore" at the end, don't add it.
	-- if not ( -- L10N
	-- 	name:find("[ %-][Ss]cript$") or
	-- 	name:find("[ %-][Cc]ode$") or
	-- 	name:find("[ %-][Ss]emaphore$")
	-- ) then
	-- 	name = name .. " script"
	-- end
	if not nocap then
		name = mw.getContentLanguage():ucfirst(name)
	end
	return name
end

function Script:makeCategoryLink()
	return "[[:Category:" .. self:getCategoryName() .. "|" .. self:getDisplayForm() .. "]]"
end

--[==[Returns the Wikidata item id for the script or <code>nil</code>. This corresponds to the the second field in the data modules.]==]
function Script:getWikidataItem()
	Script.getWikidataItem = require(language_like_module).getWikidataItem
	return self:getWikidataItem()
end

--[==[
Returns the name of the Wikipedia article for the script. `project` specifies the language and project to retrieve
the article from, defaulting to {"enwiki"} for the English Wikipedia. Normally if specified it should be the project
code for a specific-language Wikipedia e.g. "zhwiki" for the Chinese Wikipedia, but it can be any project, including
non-Wikipedia ones. If the project is the English Wikipedia and the property {wikipedia_article} is present in the data
module it will be used first. In all other cases, a sitelink will be generated from {:getWikidataItem} (if set). The
resulting value (or lack of value) is cached so that subsequent calls are fast. If no value could be determined, and
`noCategoryFallback` is {false}, {:getCategoryName} is used as fallback; otherwise, {nil} is returned. Note that if
`noCategoryFallback` is {nil} or omitted, it defaults to {false} if the project is the English Wikipedia, otherwise
to {true}. In other words, under normal circumstances, if the English Wikipedia article couldn't be retrieved, the
return value will fall back to a link to the script's category, but this won't normally happen for any other project.
]==]
function Script:getWikipediaArticle(noCategoryFallback, project)
	Script.getWikipediaArticle = require(language_like_module).getWikipediaArticle
	return self:getWikipediaArticle(noCategoryFallback, project)
end

--[==[Returns the name of the Wikimedia Commons category page for the script.]==]
function Script:getCommonsCategory()
	Script.getCommonsCategory = require(language_like_module).getCommonsCategory
	return self:getCommonsCategory()
end

--[==[Returns the charset defining the script's characters from the script's data file.
This can be used to search for words consisting only of this script, but see the warning above.]==]
function Script:getCharacters()
	return self.characters or nil
end

--[==[Returns the number of characters in the text that are part of this script.
'''Note:''' You should never assume that text consists entirely of the same script. Strings may contain spaces, punctuation and even wiki markup or HTML tags. HTML tags will skew the counts, as they contain Latin-script characters. So it's best to avoid them.]==]
function Script:countCharacters(text)
	local charset = self._rawData.characters
	if charset == nil then
		return 0
	end
	return select(2, ugsub(text, "[" .. charset .. "]", ""))
end

function Script:hasCapitalization()
	return not not self._rawData.capitalized
end

function Script:hasSpaces()
	return self._rawData.spaces ~= false
end

function Script:isTransliterated()
	return self._rawData.translit ~= false
end

--[==[Returns true if the script is (sometimes) sorted by scraping page content, meaning that it is sensitive to changes in capitalization during sorting.]==]
function Script:sortByScraping()
	return not not self._rawData.sort_by_scraping
end

--[==[Returns the text direction. Horizontal scripts return {{lua|"ltr"}} (left-to-right) or {{lua|"rtl"}} (right-to-left), while vertical scripts return {{lua|"vertical-ltr"}} (vertical left-to-right) or {{lua|"vertical-rtl"}} (vertical right-to-left).]==]
function Script:getDirection()
	return self._rawData.direction or "ltr"
end

function Script:getRawData()
	return self._rawData
end

--[==[Returns {{lua|true}} if the script contains characters that require fixes to Unicode normalization under certain circumstances, {{lua|false}} if it doesn't.]==]
function Script:hasNormalizationFixes()
	return not not self._rawData.normalizationFixes
end

--[==[Corrects discouraged sequences of Unicode characters to the encouraged equivalents.]==]
function Script:fixDiscouragedSequences(text)
	if self:hasNormalizationFixes() then
		local norm_fixes = self._rawData.normalizationFixes
		local to = norm_fixes.to
		if to then
			for i, v in ipairs(norm_fixes.from) do
				text = ugsub(text, v, to[i] or "")
			end
		end
	end
	return text
end

do
	local combining_classes
	
	-- Obtain the list of default combining classes.
	local function get_combining_classes()
		combining_classes, get_combining_classes = load_data(combining_classes_module), nil
		return combining_classes
	end
	
	-- Implements a modified form of Unicode normalization for instances where there are identified deficiencies in the default Unicode combining classes.
	local function fixNormalization(text, self)
		if not self:hasNormalizationFixes() then
			return text
		end
		local norm_fixes = self._rawData.normalizationFixes
		local new_classes = norm_fixes.combiningClasses
		if not (new_classes and umatch(text, "[" .. norm_fixes.combiningClassCharacters .. "]")) then
			return text
		end
		text = explode(text)
		-- Manual sort based on new combining classes.
		-- We can't use table.sort, as it compares the first/last values in an array as a shortcut, which messes things up.
		for i = 2, #text do
			local char = text[i]
			local class = new_classes[char] or (combining_classes or get_combining_classes())[char]
			if class then
				repeat
					i = i - 1
					local prev = text[i]
					if (new_classes[prev] or (combining_classes or get_combining_classes())[prev] or 0) < class then
						break
					end
					text[i], text[i + 1] = char, prev
				until i == 1
			end
		end
		return concat(text)
	end
	
	function Script:toFixedNFC(text)
		return fixNormalization(toNFC(text), self)
	end
	
	function Script:toFixedNFD(text)
		return fixNormalization(toNFD(text), self)
	end
	
	function Script:toFixedNFKC(text)
		return fixNormalization(toNFKC(text), self)
	end
	
	function Script:toFixedNFKD(text)
		return fixNormalization(toNFKD(text), self)
	end
end

function Script:toJSON(returnTable)
	-- Use `deepcopy` when returning a table, so that there are no editing restrictions imposed by `mw.loadData`.
	return (returnTable and deepcopy or to_json){
		canonicalName = self:getCanonicalName(),
		categoryName = self:getCategoryName("nocap"),
		code = self._code,
		otherNames = self:getOtherNames(true),
		aliases = self:getAliases(),
		varieties = self:getVarieties(),
		type = keys_to_list(self:getTypes()),
		direction = self:getDirection(),
		characters = self:getCharacters(),
		parent = self:getParent(),
		systems = self:getSystemCodes(),
		wikipediaArticle = self._rawData.wikipedia_article,
	}
end
	
function export.makeObject(code, data)
	local data_type = type(data)
	if data_type ~= "table" then
		error(("bad argument #2 to 'makeObject' (table expected, got %s)"):format(data_type))
	end
	return setmetatable({
		_rawData = data,
		_code = code,
		characters = data.characters
	}, Script)
end

--[==[Finds the script whose code matches the one provided. If it exists, it returns a {{lua|Script}} object representing the script. Otherwise, it returns {{lua|nil}}, unless <span class="n">paramForError</span> is given, in which case an error is generated. If <code class="n">paramForError</code> is {{lua|true}}, a generic error message mentioning the bad code is generated; otherwise <code class="n">paramForError</code> should be a string or number specifying the parameter that the code came from, and this parameter will be mentioned in the error message along with the bad code.]==]
function export.getByCode(code, paramForError, disallowNil)
	-- Track uses of paramForError and disallowNil, ultimately so they can be removed, as error-handling should be done by [[Module:parameters]], not here.
	if paramForError ~= nil then
		track("scripts/paramForError")
	end
	if disallowNil ~= nil then
		track("scripts/disallowNil")
	end
	
	if code == nil then
		-- If paramForError is set, a nil code should always throw an error (i.e. disallowNil should be mandatory, matching [[Module:languages]]), so track instances where it happens so that they can be fixed in the calling module.
		if paramForError then
			track("scripts/nil with paramForError")
		end
		if not disallowNil then
			return nil
		end
	end
	
	local data = (scripts_data or get_scripts_data())[code]
	if data ~= nil then
		return export.makeObject(code, data)
	elseif paramForError then
		languages_error(code, paramForError, "script code", nil, "not real lang")
	end
	return nil
end

function export.getByCanonicalName(name)
	return export.getByCode((scripts_by_name or get_scripts_by_name())[name])
end

--[==[
	Takes a codepoint or a character and finds the script code (if any) that is
	appropriate for it based on the codepoint, using the data module
	[[Module:scripts/recognition data]]. The data module was generated from the
	patterns in [[Module:scripts/data]] using [[Module:User:Erutuon/script recognition]].

	Converts the character to a codepoint. Returns a script code if the codepoint
	is in the list of individual characters, or if it is in one of the defined
	ranges in the 4096-character block that it belongs to, else returns "None".
]==]
function export.charToScript(char)
	export.charToScript = require(scripts_chartoscript_module).charToScript
	return export.charToScript(char)
end

--[==[
Returns the code for the script that has the greatest number of characters in `text`. Useful for script tagging text
that is unspecified for language. Uses [[Module:scripts/recognition data]] to determine a script code for a character
language-agnostically. Specifically, it works as follows:
	
Convert each character to a codepoint. Increment the counter for the script code if the codepoint is in the list
of individual characters, or if it is in one of the defined ranges in the 4096-character block that it belongs to.
	
Each script has a two-part counter, for primary and secondary matches. Primary matches are when the script is the
first one listed; otherwise, it's a secondary match. When comparing scripts, first the total of both are compared
(i.e. the overall number of matches). If these are the same, the number of primary and then secondary matches are
used as tiebreakers. For example, this is used to ensure that `Grek` takes priority over `Polyt` if no characters
which exclusively match `Polyt` are found, as `Grek` is a subset of `Polyt`.
	
If `none_is_last_resort_only` is specified, this will never return {"None"} if any characters in `text` belong to a
script. Otherwise, it will return {"None"} if there are more characters that don't belong to a script than belong to
any individual script. (FIXME: This behavior is probably wrong, and `none_is_last_resort_only` should probably
become the default.)
]==]
function export.findBestScriptWithoutLang(text, none_is_last_resort_only)
	export.findBestScriptWithoutLang = require(scripts_chartoscript_module).findBestScriptWithoutLang
	return export.findBestScriptWithoutLang(text, none_is_last_resort_only)
end

return export
来自“https://zh.wiktionary.org/w/index.php?title=Module:Scripts&oldid=8985286”