Module:zlw-lch-IPA/testcases/driver

local export = {}

--[=[
This module contains driver code for running a given set of testcases for ]. The testcases are
split into multiple subsets for different lects, and also to allow for splitting within a given lect if there end up
being too many to include in a single module (because of time-expired errors).

To create a new subset, copy an existing subset and modify the examples, and set up the corresponding documentation page
as appropriate. For example, to create a subset specifically for English-origin words, copy an existing page such as
] to e.g. ]. Also copy the documentation
page ] to e.g.
], modifying the module invocation reference in the
documentation page as appropriate to point to your subset module. Then modify the examples in e.g.
] according to the following format.

Each line of the example text is either a high-level header beginning with ###, a subheader beginning with ##, a comment
beginning with #, a blank line or an example. Examples consist of three tab-separated fields, followed by an optional
comment to be shown along with the example (delimited by a # preceded by whitespace). The first field is the actual
spelling of the term in question. The second field is the respelling, or # if the respelling is the same as the
spelling. The third field is the expected IPA pronunciation (without slashes or brackets).
]=]

local m_zlw_lch_IPA = require("Module:zlw-lch-IPA")
local m_links = require("Module:links")
-- FIXME
local lang = require("Module:languages").getByCode("pl")

local rsplit = mw.text.split
local rmatch = mw.ustring.match

local function tag_IPA(IPA)
	return '<span class="IPA">' .. IPA .. "</span>"
end

local function link(text)
	return m_links.full_link{ term = text, lang = lang }
end

local options = { display = tag_IPA }

function export.check_ipa(self, lang, lect, spelling, respelling, expected, comment)
	if respelling == "#" then
		respelling = spelling
	end
	local pron_list =
		m_zlw_lch_IPA.get_lect_pron_info({{respelling = respelling}}, spelling, "", lang, lect).pron_list
	local phonemic = {}
	for _, pron in ipairs(pron_list) do
		local syldiv_pron = pron.pron_with_syldivs
		syldiv_pron = syldiv_pron:gsub("^/(.*)/$", "%1")
		table.insert(phonemic, syldiv_pron)
	end
	phonemic = table.concat(phonemic, ",")
	options.comment = comment or ""
	self:equals(
		link(spelling) .. (respelling == spelling and "" or ", respelled " .. respelling),
		phonemic,
		expected,
		options
	)
end

function export.parse(examples)
	-- The following is a list of parsed examples where each element is a four-element list of
	-- {SPELLING, RESPELLING, EXPECTED, COMMENT}. SPELLING is the actual spelling of the term; RESPELLING is the
	-- respelling; EXPECTED is the phonemic IPA; and COMMENT is an optional comment or nil.
	local parsed_examples = {}
	-- Snarf each line.
	for line in examples:gmatch "+" do
		-- Trim whitespace at beginning and end.
		line = line:gsub("^%s*(.-)%s*$", "%1")
		local function err(msg)
			error(msg .. ": " .. line)
		end
		if line == "" then
			-- Skip blank lines.
		elseif line:find("^###") then
			-- Line beginning with ### is a high-level header.
			line = line:gsub("^###%s*", "")
			table.insert(parsed_examples, "#h1:" .. line)
		elseif line:find("^##") then
			-- Line beginning with ## is a subheader.
			line = line:gsub("^##%s*", "")
			table.insert(parsed_examples, line)
		elseif line:find("^#") then
			-- Line beginning with # is a comment; ignore.
		else
			local line_no_comment, _, _, _, comment = rmatch(line, "^((*)\t(*)\t(-))%s+#%s*(.*)$")
			line_no_comment = line_no_comment or line
			local parts = rsplit(line_no_comment, "\t")
			if #parts ~= 3 then
				err("Expected 3 tab-separated components in example (not including any comment)")
			end
			table.insert(parts, comment)
			table.insert(parsed_examples, parts)
		end
	end
	return parsed_examples
end

return export