imported>Erutuon: function to print data module for "default ignorable" property

2019-09-16T09:39:40Z

function to print data module for "default ignorable" property

New page

local p = {}

function p.show(frame)
local page = frame.args[1] or "User:Erutuon/Unicode/DerivedCoreProperties.txt"
local text = assert(mw.title.new(page):getContent())
local defaultIgnorable = text
:match("Derived Property: Default_Ignorable_Code_Point.-(%f[^\n]%x%x%x%x.-)%s*\n# Total code points")
local singles, ranges = {}, {}
for codePoint1, codePoint2 in defaultIgnorable:gmatch("%f[^\n%z](%x+)%.?%.?(%x*)") do
codePoint1, codePoint2 = tonumber(codePoint1, 16), tonumber(codePoint2, 16)
local lastRange = ranges[#ranges]
if lastRange and lastRange[2] == codePoint1 - 1 then
lastRange[2] = codePoint2 or codePoint1
else
if not codePoint2 then
singles[codePoint1] = true
else
table.insert(ranges, { codePoint1, codePoint2 })
end
end
end

local template = [[
local data = {}

data.defaultIgnorable = {
singles = {
...
},

ranges = {
...
},
}

return data
]]

local Array = require "Module:array"
local printedRanges = Array()
for _, range in ipairs(ranges) do
local low, high, script_code = unpack(range)
printedRanges:insert(('\t\t{ 0x%05X, 0x%05X },'):format(low, high))
end

local printedSingles = Array()
for codepoint in require 'Module:TableTools'.sortedPairs(singles) do
printedSingles:insert(('\t\t[0x%05X] = true,'):format(codepoint))
end

local data = template
:gsub('%.%.%.', printedSingles:concat('\n'), 1)
:gsub('%.%.%.', printedRanges:concat('\n'), 1)

return data
end

local Unicode_data = require "Module:Unicode data/sandbox"
local fun = require "Module:fun"
local m_table = require "Module:TableTools"

local function errorf(level, ...)
if type(level) == "number" then
return error(string.format(...), level + 1)
else -- level is actually the format string.
return error(string.format(level, ...), 2)
end
end

function p.search_for_language_codes(frame)
local page_name = frame.args[1] or "English language"

local success, title_object = pcall(mw.title.new, page_name)
if not (success and title_object) then
mw.logf("Could not make title object for '%s'.", page_name)
return
end

local content = title_object:getContent()

local language_codes = {}
for lang_template in content:gmatch "{{lang[^}]+" do
local template_name = lang_template:match("{{([^|}]+)")
local language_code
if template_name == "lang" then
language_code = lang_template:match "{{lang|([^|}]+)"
elseif template_name:find "^lang-" then
language_code = lang_template:match "{{lang-([^|}]+)"
end
if language_code then
language_codes[language_code] = true
end
end

return table.concat(m_table.keysToList(language_codes), ", ")
end

local parsed_subtags_mt = {
__index = {
-- "error" is the error message.
-- "index" is the ordinal of the subtag in which the error was found.
throw = function (self, error, index)
self.error = self.error_messages[error]
self.invalid = table.concat(self.input, "-", index)
return self:remove_unnecessary_fields()
end,

remove_unnecessary_fields = function (self)
-- Only useful internally.
self.input = nil
self:pretty_print()
p.validate_lang_tag(self)
return self
end,

-- Regularize capitalization of language subtags:
-- ZH-LATN -> zh-Latn, FR-ca -> fr-CA
pretty_print = function (self)
for key, func in pairs(self.print_funcs) do
if self[key] then
self[key] = func(self[key])
end
end
return self
end,

-- Re-create the original tag from the parsed subtags.
get_tag = function (self)
if self.tag then return self.tag end

local tag = {}
for _, subtag_name in ipairs(self.subtag_order) do
if subtag_name == "private_use" then
table.insert(tag, "x")
end

if type(self[subtag_name]) == "table" then
for _, subtag in ipairs(self[subtag_name]) do
table.insert(tag, subtag)
end
else
table.insert(tag, self[subtag_name])
end
end

tag = table.concat(tag, "-")
self.tag = tag -- Cache the result.

return tag
end,

subtag_order = {
"language", "script", "region", "variant", "private_use"
},

error_messages = {
invalid_characters = "invalid characters",
no_language = "no language subtag",
invalid_subtag = "invalid subtag",
invalid_private_use = "length of private-use subtag out of range",
empty_private_use = "empty private-use subtag",
}
}
}
local function initial_caps_helper(initial, rest)
return string.upper(initial) .. string.lower(rest)
end
local function lower_or_map_lower(str)
if type(str) == "table" then
return fun.map(string.lower, str)
else
return string.lower(str)
end
end
parsed_subtags_mt.__index.print_funcs = {
language = string.lower,
script = function (script_code)
return (string.gsub(script_code, "^(%a)(%a%a%a)$", initial_caps_helper))
end,
region = string.upper,
variant = lower_or_map_lower,
private_use = lower_or_map_lower,
}

setmetatable(parsed_subtags_mt, {
__call = function (self, input)
return setmetatable({ input = input }, self)
end
})

-- An array of patterns for each subtag, and a "type" field for the name
-- of the subtag.
-- The patterns are checked in order, and any of the subtags can be skipped.
-- So, for example, the "language" subtag must precede the "script"
-- subtag, but a tag may contain a "language" subtag, no "script" subtag
-- and then a "region" subtag.
-- If the full list of subtags has been iterated over, the remaining subtags
-- must match the pattern for a private-use subtag, or the tag is invalid.
local subtag_info = { -- can be put in data module
{ "%a%a%a?", "1%a+", type = "language" }, -- ll or lll; special case
-- include extlang?
{ "%a%a%a%a", type = "script" }, -- Ssss
{ "%a%a", "%d%d%d", type = "region" }, -- rr, DDD
{
"%d%d%d%d", -- 4 digits
"%w%w%w%w%w%w?%w?%w?", -- 5-8 alnum characters
type = "variant",
repeatable = true, -- There can be multiple variants.
}
}

-- A previous draft, in [[Module:Lang/sandbox]]:
-- https://en.wikipedia.org/w/index.php?oldid=812819217

-- Based on https://www.w3.org/International/articles/language-tags/.

-- Parse a language tag.
-- Returns nil if tag is not a string or empty.
-- Else returns a table with a map of subtag type to subtag for all subtags that
-- were parsed.
-- If there was an error, returns an "error" field with a description of the
-- error, and an "invalid" field with the suffix of the tag starting at the
-- index where the error occurred.

-- Does not recognize "extension" tags, such as those introduced by "u", as they
-- are not needed on Wikipedia. Does not recognize "grandfathered" tags.
-- Does not recognize extended language subtags, such as "zh-yue".
-- https://www.rfc-editor.org/rfc/rfc6067.txt, https://tools.ietf.org/html/bcp47

-- Only checks that the syntax is correct, not that the values are valid. For
-- instance, will accept non-existent language codes, like "zz".
function p.parse_IETF(tag)
if type(tag) ~= "string" or tag == "" then
return nil
end

-- This may contain the special fields "invalid", "error".
-- "error" indicates why the
-- tag is invalid (if applicable).
-- All other fields are subtags, and they appear in the tag in the following
-- order:
-- "language", "script", "region", "variant", "private_use", "invalid"
-- All these subtags can be strings or nil, while "variant" can also be an
-- array of strings if more than one variant subtag was found.
-- "invalid" is the portion of the tag after the last valid subtag (minus a
-- hyphen).
local segments = mw.text.split(tag, "-")
local parsed_subtags = parsed_subtags_mt(segments)

-- Language tags probably only contain ASCII alphabetic and numerical
-- characters and hyphen-minus.
if not tag:find "^[A-Za-z0-9-]+$" then
return parsed_subtags:throw(
"invalid_characters",
fun.indexOf(
function (tag)
return tag:find "[^A-Za-z0-9-]"
end,
segments))
end

local subtag_i = 1 -- Index of current item in subtag_info.
local segment_i = 1 -- Index of current segment.
while segments[segment_i] and subtag_info[subtag_i] do
local segment = segments[segment_i]
local subtag_type
while not subtag_type and subtag_info[subtag_i] do
-- Check each pattern for the subtag type at "subtag_i" in "subtag_info".
local cur_subtag = subtag_info[subtag_i]
for _, pattern in ipairs(cur_subtag) do
if segment:find("^" .. pattern .. "$") then
subtag_type = cur_subtag.type
-- There can be multiple "variant" subtags (and "extension"
-- subtags, if those are added).
if not cur_subtag.repeatable then
subtag_i = subtag_i + 1
end
break
end
end

if not subtag_type then -- No match; try next subtag.
subtag_i = subtag_i + 1
end
end

-- If language subtag has not been found, or the current segment has not
-- been matched as a subtag, break the loop and check for
-- a private-use subtag.
if segment_i == 1 and subtag_type ~= "language" or not subtag_type then
break
else
if parsed_subtags[subtag_type] then -- Create an array.
if type(parsed_subtags[subtag_type]) == "string" then
parsed_subtags[subtag_type] = { parsed_subtags[subtag_type] }
end -- else table
table.insert(parsed_subtags[subtag_type], segment)
else
parsed_subtags[subtag_type] = segment
end
last_matched_segment_i = segment_i
end

segment_i = segment_i + 1
end

if segments[segment_i] then -- More segments to scan?
-- Not all potential subtags were matched. Check for private-use subtags.
-- https://tools.ietf.org/html/bcp47#section-2.2.7
-- Private-use subtags consist of one or more sequences of 1 to 8
-- alphanumeric characters preceded by "x-".
-- Alphanumericity has already been checked.

-- A tag must start with either a language subtag or a private-use subtag.
-- If next segment is not "x", introducing a private-use subtag, there
-- is no private-use subtag.
if segments[segment_i] and segments[segment_i]:lower() ~= "x" then
if not parsed_subtags.language then
return parsed_subtags:throw("no_language", 1)
else
return parsed_subtags:throw("invalid_subtag",
segment_i)
end
elseif not segments[segment_i + 1] then
return parsed_subtags:throw("empty_private_use",
segment_i)
end

-- Check length of all segments after "x".
for i = segment_i + 1, #segments do
local length = #segments[i]

if not (1 <= length and length <= 8) then
return parsed_subtags
:throw("invalid_private_use", segment_i)
end
end

if not segments[last_matched_segment_i + 3] then -- There is only one private-use subtag.
parsed_subtags.private_use = segments[segment_i + 1]
else
parsed_subtags.private_use = {}
for i = segment_i + 1, #segments do
table.insert(parsed_subtags.private_use, segments[i])
end
end
end

return parsed_subtags:remove_unnecessary_fields()
end

local lang_name_table = mw.loadData "Module:Language/name/data"
local synonym_table = mw.loadData "Module:Lang/ISO 639 synonyms"
local lang_data = mw.loadData "Module:Lang/data"

function p.validate_lang_tag(parsed_subtags)
-- Already checked that the tag starts with a language subtag or a private-use subtag.
-- Script code is initially capitalized, region code is uppercase,
-- everything else is lowercase.

-- Check existence of language tag.
if parsed_subtags.language and
not (lang_data.override[parsed_subtags.language]
or lang_name_table.lang[parsed_subtags.language]) then
mw.log("Invalid language code", parsed_subtags.language, "in", parsed_subtags:get_tag())
end

-- Check existence of script tag.
if parsed_subtags.script then
local lower_script = parsed_subtags.script:lower()
if not lang_name_table.script[lower_script] then
mw.log("Invalid script code", parsed_subtags.script, "in", parsed_subtags:get_tag())
end

-- Check that script tag is not marked as superfluous (because the
-- it is considered the default one for the language).
if lang_name_table.suppressed[lower_script]
and parsed_subtags.language
and m_table.inArray(
lang_name_table.suppressed[lower_script],
parsed_subtags.language:lower()) then
mw.log(parsed_subtags.script, "is suppressed with",
parsed_subtags.language, "in", parsed_subtags:get_tag())
end
end

-- Check existence of region code..
if parsed_subtags.region and not lang_name_table.region[parsed_subtags.region:lower()] then
mw.log("Invalid region code", parsed_subtags.region, "in", parsed_subtags:get_tag())
end

-- Check that variant code is valid, and that it can validly be used with the
-- given combination of language, script, region, and variant.
-- Check for duplicate variant subtags?
if parsed_subtags.variant then
local lower_tag = parsed_subtags:get_tag():lower()

for _, variant in ipairs(type(parsed_subtags.variant) == "table"
and parsed_subtags.variant or { parsed_subtags.variant }) do
if not lang_name_table.variant[variant] then
mw.log("Invalid variant code", variant, "in", parsed_subtags:get_tag())
else
local prefix = parsed_subtags:get_tag():lower():match("^(.-)%-" .. variant)

-- Check that at least one of the prefixes is found at the
-- beginning of lower_tag.
if not fun.some(function (prefix)
return lower_tag:find(prefix, 1, true) == 1
end,
lang_name_table.variant[variant].prefixes) then
mw.log("Variant tag", variant, "does not belong with prefix",
prefix, "in", parsed_subtags:get_tag())
end
end
end
end

-- Check that the private-use subtag is actually used by Wikipedia.
if parsed_subtags.private_use and not lang_data.override[parsed_subtags.tag] then
mw.log("Invalid private-use subtag in", parsed_subtags:get_tag())
end
end

function p.show_COinS(frame)
local ref = frame.args[1]

local tag = ref:match('<span [^>]*class="Z3988"[^>]*>')
local data = tag:match('title="(.-)"')

local vals = {}

for item in mw.text.gsplit(data, "&") do
local key, value = item:match("(.-)=(.*)")
vals[key] = mw.uri.decode(value)
end

return ref .. "\n\n" .. table.concat(
require "Module:fun".mapIter(
function (value, key)
return ("%s: %s"):format(key, value)
end,
m_table.sortedPairs(
vals)),
", ")
end

return p

Module:Sandbox/Erutuon - Revision history

imported>Erutuon: function to print data module for "default ignorable" property