မေႃႇၵျူး:families

This module is used to retrieve and manage Wiktionary's various language families and the information associated with them. See Wiktionary:Families for more information.

This module provides access to other modules. To access the information from within a template, see Module:families/templates.

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

Finding and retrieving families

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

getByCode

getByCode(code)

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

getByCanonicalName

getByCanonicalName(name)

Looks for the family whose canonical name (the name used to represent that language on Wiktionary) matches the one provided. If it exists, it returns a Family object representing the family. Otherwise, it returns nil. The canonical name of families should always be unique (it is an error for two families on Wiktionary to share the same canonical name), so this is guaranteed to give at most one result.

Family objects

A Family object is returned from one of the functions above. It is a Lua representation of a family 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_families = require("Module:families")
local fam = m_families.getByCode("ine")
local name = fam:getCanonicalName()
-- "name" will now be "Indo-European"

Family:getCode

:getCode()

Returns the family code of the family. Example: "ine" for the Indo-European languages.

Family:getCanonicalName

:getCanonicalName()

Returns the canonical name of the family. This is the name used to represent that language family on Wiktionary, and is guaranteed to be unique to that family alone. Example: "Indo-European" for the Indo-European languages.

Family:getDisplayForm

:getDisplayForm()

Returns the display form of the family. For families, this is the same as the value returned by :getCategoryName("nocap"), i.e. it reads "NAME languages" (e.g. "Indo-Iranian languages"). For regular and etymology languages, this is the same as the canonical name, and for scripts, it reads "NAME script" (e.g. "Arabic script"). The displayed text used in :makeCategoryLink is always the same as the display form.

Family:getAllNames

:getAllNames()

Returns a table of all names that the family is known by, including the canonical name. The names are not guaranteed to be unique, sometimes more than one family is known by the same name. Example: {"Slavic", "Slavonic"} for the Slavic languages.

Family:getFamily

:getFamily()

Returns a Family object for the parent family that the family is a part of.

Family:getProtoLanguage

:getProtoLanguage()

Returns a Language object (see Module:languages) for the proto-language of this family, if one exists.

Family:getCategoryName

:getCategoryName(nocap)

Returns the name of the main category of that family. Example: "Germanic languages" for the Germanic languages, whose category is at Category:Germanic languages.

Unless optional argument nocap is given, the family name at the beginning of the returned value will be capitalized. This capitalization is correct for category names, but not if the family name is lowercase and the returned value of this function is used in the middle of a sentence. (For example, the pseudo-family with the code qfa-mix has the name "mixed", which should remain lowercase when used as part of the category name Category:Terms derived from mixed languages but should be capitalized in Category:Mixed languages.) If you are considering using getCategoryName("nocap"), use getDisplayForm() instead.

Family:getWikidataItem

:getWikidataItem()

Returns the Wikidata item of that family.

Family:getWikipediaArticle

:getWikipediaArticle()

Returns the Wikipedia article of that family, usually derived from :getWikidataItem().

---

local export = {}

local families_by_name_module = "Module:families/canonical names"
local families_data_module = "Module:families/data"
local families_etymology_data_module = "Module:families/data/etymology"
local json_module = "Module:JSON"
local language_like_module = "Module:language-like"
local languages_module = "Module:languages"
local load_module = "Module:load"
local table_module = "Module:table"

local get_by_code -- Defined below.
local gmatch = string.gmatch
local insert = table.insert
local ipairs = ipairs
local make_object -- Defined below.
local pairs = pairs
local require = require
local setmetatable = setmetatable
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 category_name_has_suffix(...)
		category_name_has_suffix = require(language_like_module).categoryNameHasSuffix
		return category_name_has_suffix(...)
	end

	local function category_name_to_code(...)
		category_name_to_code = require(language_like_module).categoryNameToCode
		return category_name_to_code(...)
	end

	local function deep_copy(...)
		deep_copy = require(table_module).deepCopy
		return deep_copy(...)
	end

	local function get_lang(...)
		get_lang = require(languages_module).getByCode
		return get_lang(...)
	end

	local function keys_to_list(...)
		keys_to_list = require(table_module).keysToList
		return keys_to_list(...)
	end

	local function load_data(...)
		load_data = require(load_module).load_data
		return load_data(...)
	end

	local function make_lang_object(...)
		make_lang_object = require(languages_module).makeObject
		return make_lang_object(...)
	end

	local function to_json(...)
		to_json = require(json_module).toJSON
		return to_json(...)
	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 families_by_name
	local function get_families_by_name()
		families_by_name, get_families_by_name = load_data(families_by_name_module), nil
		return families_by_name
	end
	
	local families_data
	local function get_families_data()
		families_data, get_families_data = load_data(families_data_module), nil
		return families_data
	end
	
	local families_etymology_data
	local function get_families_etymology_data()
		families_etymology_data, get_families_etymology_data = load_data(families_etymology_data_module), nil
		return families_etymology_data
	end
	
	local families_suffixes
	local function get_families_suffixes()
		families_suffixes, get_families_suffixes = {
			"languages",
			"lects"
		}, nil
		return families_suffixes
	end

local Family = {}
Family.__index = Family

--[==[
Return the family code of the family, e.g. {"ine"} for the Indo-European languages.
]==]
function Family:getCode()
	return self._code
end

--[==[
Return the canonical name of the family. This is the name used to represent that language family on Wiktionary,
and is guaranteed to be unique to that family alone. Example: {"Indo-European"} for the Indo-European languages.
]==]
function Family:getCanonicalName()
	local name = self._name
	if name == nil then
		name = self._data[1]
		self._name = name
	end
	return name
end

--[==[
Return the display form of the family. For families, this is the same as the value returned by
{getCategoryName("nocap")}, i.e. it reads <code>"<var>name</var> languages"</code> (e.g.
{"Indo-Iranian languages"}). For full and etymology-only languages, this is the same as the canonical name, and
for scripts, it reads <code>"<var>name</var> script"</code> (e.g. {"Arabic script"}). The displayed text used in
{makeCategoryLink()} is always the same as the display form.
]==]
function Family:getDisplayForm()
	return self:getCategoryName("nocap")
end

function Family:getAliases()
	Family.getAliases = require(language_like_module).getAliases
	return self:getAliases()
end

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

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

function Family:getAllNames()
	Family.getAllNames = require(language_like_module).getAllNames
	return self:getAllNames()
end

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

The possible types are
* {family}: This object is a family.
* {full}: This object is a "full" family. This includes all families but a couple of etymology-only
		  families for Old and Middle Iranian languages.
* {etymology-only}: This object is an etymology-only family, similar to etymology-only languages. There
					are currently only two such families, for Old Iranian languages and Middle Iranian
					languages (which do not represent proper clades and have no proto-languages, hence
					cannot be full families).
]==]
function Family:getTypes()
	local types = self._types
	if types == nil then
		types = {family = true}
		if self:getFullCode() == self:getCode() then
			types.full = true
		else
			types["etymology-only"] = true
		end
		local rawtypes = self._data.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 family has all of them.]==]
function Family:hasType(...)
	Family.hasType = require(language_like_module).hasType
	return self:hasType(...)
end

--[==[Returns a {Family} object for the superfamily that the family belongs to.]==]
function Family:getFamily()
	if self._familyObject == nil then
		local familyCode = self:getFamilyCode()
		if familyCode then
			self._familyObject = get_by_code(familyCode)
		else
			self._familyObject = false
		end
	end
	return self._familyObject or nil
end

--[==[Returns the code of the family's superfamily.]==]
function Family:getFamilyCode()
	if not self._familyCode then
		self._familyCode = self._data[3]
	end
	return self._familyCode
end

--[==[Returns the canonical name of the family's superfamily.]==]
function Family:getFamilyName()
	if self._familyName == nil then
		local family = self:getFamily()
		if family then
			self._familyName = family:getCanonicalName()
		else
			self._familyName = false
		end
	end
	return self._familyName or nil
end

--[==[Check whether the family belongs to {superfamily} (which can be a family code or object), and returns a boolean. If more than one is given, returns {true} if the family belongs to any of them. A family is '''not''' considered to belong to itself.]==]
function Family:inFamily(...)
	for _, superfamily in ipairs{...} do
		if type(superfamily) == "table" then
			superfamily = superfamily:getCode()
		end
		local family, code = self:getFamily()
		while true do
			if not family then
				return false
			end
			code = family:getCode()
			family = family:getFamily()
			-- If family is parent to itself, return false.
			if family and family:getCode() == code then
				return false
			elseif code == superfamily then
				return true
			end
		end
	end
end

function Family:getParent()
	if self._parentObject == nil then
		local parentCode = self:getParentCode()
		if parentCode then
			self._parentObject = get_lang(parentCode, nil, true, true)
		else
			self._parentObject = false
		end
	end
	return self._parentObject or nil
end

function Family:getParentCode()
	if not self._parentCode then
		self._parentCode = self._data.parent
	end
	return self._parentCode
end

function Family:getParentName()
	if self._parentName == nil then
		local parent = self:getParent()
		if parent then
			self._parentName = parent:getCanonicalName()
		else
			self._parentName = false
		end
	end
	return self._parentName or nil
end

function Family:getParentChain()
	if not self._parentChain then
		self._parentChain = {}
		local parent = self:getParent()
		while parent do
			insert(self._parentChain, parent)
			parent = parent:getParent()
		end
	end
	return self._parentChain
end

function Family:hasParent(...)
	--checkObject("family", nil, ...)
	for _, other_family in ipairs{...} do
		for _, parent in ipairs(self:getParentChain()) do
			if type(other_family) == "string" then
				if other_family == parent:getCode() then return true end
			else
				if other_family:getCode() == parent:getCode() then return true end
			end
		end
	end
	return false
end

--[==[
If the family is etymology-only, this iterates through its parents until a full family is found, and the
corresponding object is returned. If the family is a full family, then it simply returns itself.
]==]
function Family:getFull()
	if not self._fullObject then
		local fullCode = self:getFullCode()
		if fullCode ~= self:getCode() then
			self._fullObject = get_lang(fullCode, nil, nil, true)
		else
			self._fullObject = self
		end
	end
	return self._fullObject
end

--[==[
If the family is etymology-only, this iterates through its parents until a full family is found, and the
corresponding code is returned. If the family is a full family, then it simply returns the family code.
]==]
function Family:getFullCode()
	return self._fullCode or self:getCode()
end

--[==[
If the family is etymology-only, this iterates through its parents until a full family is found, and the
corresponding canonical name is returned. If the family is a full family, then it simply returns the canonical name
of the family.
]==]
function Family:getFullName()
	if self._fullName == nil then
		local full = self:getFull()
		if full then
			self._fullName = full:getCanonicalName()
		else
			self._fullName = false
		end
	end
	return self._fullName or nil
end

--[==[
Return a {Language} object (see [[Module:languages]]) for the proto-language of this family, if one exists.
Otherwise, return {nil}.
]==]
function Family:getProtoLanguage()
	if self._protoLanguageObject == nil then
		self._protoLanguageObject = get_lang(self._data.protoLanguage or self:getCode() .. "-pro", nil, true) or false
	end
	return self._protoLanguageObject or nil
end

function Family:getProtoLanguageCode()
	if self._protoLanguageCode == nil then
		local protoLanguage = self:getProtoLanguage()
		self._protoLanguageCode = protoLanguage and protoLanguage:getCode() or false
	end
	return self._protoLanguageCode or nil
end

function Family:getProtoLanguageName()
	if not self._protoLanguageName then
		self._protoLanguageName = self:getProtoLanguage():getCanonicalName()
	end
	return self._protoLanguageName
end

function Family:hasAncestor(...)
	-- Go up the family tree until a protolanguage is found.
	local family = self
	local protolang = family:getProtoLanguage()
	while not protolang do
		family = family:getFamily()
		protolang = family:getProtoLanguage()
		-- Return false if the family is its own family, to avoid an infinite loop.
		if family:getFamilyCode() == family:getCode() then
			return false
		end
	end
	-- If the protolanguage is not in the family, it must therefore be ancestral to it. Check if it is a match.
	for _, otherlang in ipairs{...} do
		if (
			type(otherlang) == "string" and protolang:getCode() == otherlang or
			type(otherlang) == "table" and protolang:getCode() == otherlang:getCode()
		) and not protolang:inFamily(self) then
			return true
		end
	end
	-- If not, check the protolanguage's ancestry.
	return protolang:hasAncestor(...)
end

local function fetch_descendants(self, format)
	local languages = require("Module:languages/code to canonical name")
	local etymology_languages = require("Module:etymology languages/code to canonical name")
	local families = require("Module:families/code to canonical name")
	local descendants = {}
	-- Iterate over all three datasets.
	for _, data in ipairs{languages, etymology_languages, families} do
		for code in pairs(data) do
			local lang = get_lang(code, nil, true, true)
			if lang:inFamily(self) then
				if format == "object" then
					insert(descendants, lang)
				elseif format == "code" then
					insert(descendants, code)
				elseif format == "name" then
					insert(descendants, lang:getCanonicalName())
				end
			end
		end
	end
	return descendants
end

function Family:getDescendants()
	if not self._descendantObjects then
		self._descendantObjects = fetch_descendants(self, "object")
	end
	return self._descendantObjects
end

function Family:getDescendantCodes()
	if not self._descendantCodes then
		self._descendantCodes = fetch_descendants(self, "code")
	end
	return self._descendantCodes
end

function Family:getDescendantNames()
	if not self._descendantNames then
		self._descendantNames = fetch_descendants(self, "name")
	end
	return self._descendantNames
end

function Family:hasDescendant(...)
	for _, lang in ipairs{...} do
		if type(lang) == "string" then
			lang = get_lang(lang, nil, true)
		end
		if lang:inFamily(self) then
			return true
		end
	end
	return false
end

--[==[
Return the name of the main category of that family. Example: {"Germanic languages"} for the Germanic languages,
whose category is at [[:Category:Germanic languages]].

Unless optional argument `nocap` is given, the family name at the beginning of the returned value will be
capitalized. This capitalization is correct for category names, but not if the family name is lowercase and
the returned value of this function is used in the middle of a sentence. (For example, the pseudo-family with
the code {qfa-mix} has the name {"mixed"}, which should remain lowercase when used as part of the category name
[[:Category:Terms derived from mixed languages]] but should be capitalized in [[:Category:Mixed languages]].)
If you are considering using {getCategoryName("nocap")}, use {getDisplayForm()} instead.
]==]
function Family:getCategoryName(nocap)
	local name = self._data[1]
	if category_name_has_suffix(name, families_suffixes or get_families_suffixes()) then
		name = name .. " languages"
	end
	if not nocap then
		name = mw.getContentLanguage():ucfirst(name)
	end
	return name
end

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

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

--[==[
Returns the name of the Wikipedia article for the family. `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 family's category, but this won't normally happen for any other project.
]==]
function Family:getWikipediaArticle(noCategoryFallback, project)
	Family.getWikipediaArticle = require(language_like_module).getWikipediaArticle
	return self:getWikipediaArticle(noCategoryFallback, project)
end

function Family:makeWikipediaLink()
	return "[[w:" .. self:getWikipediaArticle() .. "|" .. self:getCanonicalName() .. "]]"
end

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

function Family:toJSON(opts)
	local ret = {
		canonicalName = self:getCanonicalName(),
		categoryName = self:getCategoryName("nocap"),
		code = self:getCode(),
		parent = self:getParentCode(),
		full = self:getFullCode(),
		family = self:getFamilyCode(),
		protoLanguage = self:getProtoLanguageCode(),
		aliases = self:getAliases(),
		varieties = self:getVarieties(),
		otherNames = self:getOtherNames(),
		type = keys_to_list(self:getTypes()),
		wikidataItem = self:getWikidataItem(),
		wikipediaArticle = self:getWikipediaArticle(true),
	}
	-- Use `deep_copy` when returning a table, so that there are no editing restrictions imposed by `mw.loadData`.
	return opts and opts.lua_table and deep_copy(ret) or to_json(ret, opts)
end

function Family:getData()
	return self._data
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({_data = data, _code = code, _fullCode = code}, Family)
end
make_object = export.makeObject

--[==[
Finds the family whose code matches the one provided. If it exists, it returns a {Family} object representing the
family. Otherwise, it returns {nil}.]==]
function export.getByCode(code)
	local data = (families_data or get_families_data())[code]
	if data ~= nil then
		return make_object(code, data)
	end
	data = (families_etymology_data or get_families_etymology_data())[code]
	return data ~= nil and make_lang_object(code, data) or nil
end
get_by_code = export.getByCode

--[==[
Look for the family whose canonical name (the name used to represent that family on Wiktionary) matches the one
provided. If it exists, it returns a {Family} object representing the family. Otherwise, it returns {nil}. The
canonical name of families should always be unique (it is an error for two families on Wiktionary to share the same
canonical name), so this is guaranteed to give at most one result.]==]
function export.getByCanonicalName(name)
	if name == nil then
		return nil
	end
	local code = (families_by_name or get_families_by_name())[name]
	if code == nil then
		return nil
	end
	return get_by_code(code)
end

--[==[
Look for the family whose category name (the name used in categories for that family) matches the one provided.
If it exists, it returns a {Family} object representing the family. Otherwise, it returns {nil}. In almost all cases,
the category name for a family is its canonical name plus the word "languages", e.g. "Indo-European" has the category
name "Indo-European languages". Where a canonical name ends with "languages" or "lects", the category name is identical
to the canonical name.]==]
function export.getByCategoryName(name)
	if name == nil then
		return nil
	end
	local code = category_name_to_code(
		name,
		" languages",
		families_by_name or get_families_by_name(),
		families_suffixes or get_families_suffixes()
	)
	if code == nil then
		return nil
	end
	return get_by_code(code)
end

return export