Module:links/data

local u = require("Module:string utilities").char

local data = {}

data.phonetic_extraction = {
	 = "Module:th",
	 = "Module:km",
}

data.ignored_prefixes = {
	 = true,
	 = true,
	 = true,
	 = true
}

data.pos_tags = {
	 = "adjective",
	 = "adverb",
	 = "interjection",
	 = "noun",
	 = "pronoun",
	 = "verb",
	 = "intransitive verb",
	 = "transitive verb",
	 = "transitive and intransitive verb",
}

-- Scheme for using unsupported characters in titles.
data.unsupported_characters = {
	 = "`num`",
	 = "`percnt`", -- only escaped in percent encoding
	 = "`amp`", -- only escaped in HTML entities
	 = "`period`", -- only escaped in dot-slash notation
	 = "`lt`",
	 = "`gt`",
	 = "`lsqb`",
	"] = "`rsqb`",
	 = "`lowbar`",
	 = "`grave`", -- used to enclose unsupported characters in the scheme, so a raw use in an unsupported title must be escaped to prevent interference
	 = "`lcub`",
	 = "`vert`",
	 = "`rcub`",
	 = "`tilde`", -- only escaped when 3 or more are consecutive
	 = "`repl`" -- replacement character U+FFFD, which can't be typed directly here due to an abuse filter
}

-- Manually specified unsupported titles. Only put titles here if there is a different reason why they are unsupported, and not just because they contain one of the unsupported characters above.
data.unsupported_titles = {
	 = "Space",
	 = "`amp`amp;",
	 = "Ancient Greek dish",
	 = "Thai name of Bangkok",
	 = "Ogham space",
	 = "Ideographic space"
}

return data