Module:User name utils

Lua implementation of the getCanonical() and isValid() functions from https://gerrit.wikimedia.org/r/plugins/gitiles/mediawiki/core/+/refs/heads/master/includes/User/UserNameUtils.php

Originally created to validate usernames before creating a User info card, since the {{#uic:}} parser function adds pages to the category Category:Pages with an IP as user info card target if getCanonical() returns false.

Usage

{{#invoke:User name utils|getCanonical|name}} Returns a normalized version of name if valid, and an empty string if not valid. Normalization will, among other things, capitalize the first letter, remove the "User:" namespace prefix if present, replace underscores with spaces, and decode HTML entities.

{{#invoke:User name utils|isValid|name}} Returns "1" if name is a valid username (including leading capitalization and without the namespace), and an empty string if not valid. Note that "example" would be considered invalid because it does not have a leading capital letter, and "User:Example" would be considered invalid because the namespace is specified.

The p._getCanonical(name) and p._isValid(name) functions can be accessed directly from other Lua modules and behave similarly, except _getCanonical returns boolean false instead of an empty string for invalid names, and _isValid returns boolean true or false.

---

-- Lua implementation of mediawiki/core/+/refs/heads/master/includes/User/UserNameUtils.php
-- with the default RIGOR_VALID option

local p = {}

local IPV4_ADDRESS = "%d%d?%d?%.%d%d?%d?%.%d%d?%d?%.%d%d?%d?"
local lang = mw.getContentLanguage()

function p._getCanonical(name)
	if (name or '') == '' then
		return false
	end
	
	-- force usernames to capital
	name = lang:ucfirst(name)
	
	-- Reject names containing '#'; these will be cleaned up
    -- with title normalisation, but then it's too late to
    -- check elsewhere
	if mw.ustring.match(name, "#") then
		return false
	end

	-- Clean up name according to title rules
	local title = mw.title.new(name,2)

	-- Check for invalid titles
	if title == nil or title.namespace ~= 2 or title.isExternal then
		return false
	end

	name = title.text

	return p._isValid(name) and name or false
end

function p._isValid(name)
	if (name or '') == '' or
		require('Module:IPAddress')._isIpOrRange(name) ~= "" or
		mw.ustring.match(name, "^" .. IPV4_ADDRESS .. "$") or -- anyIPv4, even if invalid
		mw.ustring.match(name, "^" .. IPV4_ADDRESS .. "-" .. IPV4_ADDRESS .. "$") or -- isLikeIPv4DashRange
		mw.ustring.match(name, "/") or
		name ~= lang:ucfirst(name) then
		return false
	end

	-- Ensure that the name can't be misresolved as a different title,
    -- such as with extra namespace keys at the start.
	local title = mw.title.new(name)
	if not title or
		title.namespace ~= 0 or
		name ~= title.text then
		return false
	end

	return true	
end

function p.getCanonical(frame)
	local name = p._getCanonical(frame.args[1])
	return name and name or ""
end

function p.isValid(frame)
	return p._isValid(frame.args[1]) and "1" or ""
end

return p