local p = {}
local Unicode_data = require 'Module:Unicode data/sandbox'
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 mw.logf(...)
return mw.log(string.format(...))
end
local output_mt = {}
function output_mt:insert(str)
self.n = self.n + 1
self[self.n] = str
end
-- also in [[Module:Unicode data/documentation functions]]
function output_mt:insert_format(...)
self:insert(string.format(...))
end
output_mt.join = table.concat
output_mt.__index = output_mt
local function Output()
return setmetatable({ n = 0 }, output_mt)
end
local Latn_pattern = table.concat {
'[',
'\n\32-\127',
'\194\160-\194\172',
'\195\128-\195\191',
'\196\128-\197\191',
'\198\128-\201\143',
'\225\184\128-\225\187\191',
'\226\177\160-\226\177\191',
'\234\156\160-\234\159\191',
'\234\172\176-\234\173\175',
'\239\172\128-\239\172\134',
'\239\188\129-\239\188\188',
'–',
'—',
'«', '»',
']',
};
local get_codepoint = mw.ustring.codepoint
local function expand_range(start, ending)
local lower, higher = get_codepoint(start), get_codepoint(ending)
if higher < lower then
return nil
end
local chars = {}
local i = 0
for codepoint = lower, higher do
i = i + 1
chars[i] = mw.ustring.char(codepoint)
end
return table.concat(chars)
end
local fun = require "Module:Fun"
local m_table = require "Module:Table"
local script_to_count_mt = {
__index = function (self, key)
self[key] = 0
return 0
end,
__call = function (self, ...)
return setmetatable({}, self)
end
}
setmetatable(script_to_count_mt, script_to_count_mt)
-- Uses an iterator (such as mw.ustring.gcodepoint) that generates a codepoint
-- each time it is called with an optional state and another value.
local function show_scripts(iterator, state, value)
local script_to_count = script_to_count_mt()
for codepoint in iterator, state, value do
local script = Unicode_data.lookup_script(codepoint)
script_to_count[script] = script_to_count[script] + 1
end
return table.concat(
fun.mapIter(
function (count, script)
return ("%s (%d)"):format(script, count)
end,
m_table.sortedPairs(
script_to_count,
function (script1, script2)
return script_to_count[script1] > script_to_count[script2]
end)),
", ")
end
local function get_chars_in_scripts(iterator, state, value)
local script_to_char_set = {}
for codepoint in iterator, state, value do
local script = Unicode_data.lookup_script(codepoint)
script_to_char_set[script] = script_to_char_set[script] or {}
script_to_char_set[script][codepoint] = true
end
return script_to_char_set
end
local function print_char_set_map(script_to_char_set, format, separator)
format = format or "%s: %s"
separator = separator or "\n"
return table.concat(
fun.mapIter(
function (char_set, script)
local char_list = fun.mapIter(
function (_, codepoint)
return mw.ustring.char(codepoint)
end,
m_table.sortedPairs(char_set))
return (format):format(script, mw.text.nowiki(table.concat(char_list)))
end,
m_table.sortedPairs(script_to_char_set)),
separator)
end
function p.show(frame)
local expanded_pattern = Latn_pattern
:gsub('%[(.-)%]', '%1')
:gsub( -- Find two UTF-8-encoded characters separated by hyphen-minus.
'([%z\1-\127\194-\244][\128-\191]*)%-([%z\1-\127\194-\244][\128-\191]*)',
function (char1, char2)
return expand_range(char1, char2)
end)
return ('* <div style="overflow-wrap: break-word;">%s</div><br>%s')
:format(expanded_pattern
:gsub('^%s*', ''), -- Remove initial '\n ' to avoid creating unwanted pre element.
show_scripts(mw.ustring.gcodepoint(expanded_pattern)))
end
local function get_block_info_from_arg(args, arg)
local block_name = args[1]
or errorf("Parameter %s is required", tostring(arg))
local block_info = Unicode_data.get_block_info(block_name)
or errorf("The block '%s' could be found", block_name)
return block_info
end
local function get_boolean_from_arg(args, arg)
return args[arg] and require "Module:Yesno" (args[arg])
end
function p.scripts_in_block(frame)
local block_info = get_block_info_from_arg(frame.args, 1)
local show_block_name = get_boolean_from_arg(frame.args, 2)
local script_list = show_scripts(fun.range(block_info[1], block_info[2]))
if show_block_name then
return ("%s: %s"):format(block_info[3], script_list)
else
return script_list
end
end
local function link_block_name(block_name)
if block_name:find ' ' then
return ("[[%s]]"):format(block_name)
else
return ("[[%s (Unicode block)|%s]]"):format(block_name, block_name)
end
end
function p.scripts_in_blocks(frame)
local output = Output()
local start = frame.args[1] and tonumber(frame.args[1], 16) or 0
local ending = frame.args[2] and tonumber(frame.args[2], 16) or 0x4000
local script_data = mw.loadData "Module:Unicode data/scripts"
local singles = script_data.singles
local ranges = script_data.ranges
local function clear (self)
for _, key in ipairs(m_table.keysToList(self, false)) do
self[key] = nil
end
end
local counts = {}
setmetatable(counts, {
__index = {
increment = function(self, script_code, amount)
self[script_code] = (self[script_code] or 0) + (amount or 1)
end,
clear = clear,
}
})
local codepoints_per_script = {}
setmetatable(codepoints_per_script, {
__index = {
add = function(self, script_code, codepoint)
self[script_code] = self[script_code] or { n = 0 }
if self[script_code].n <= 0x20
and not (codepoint <= 0x9F and (codepoint >= 0x80
or codepoint <= 0x1F)) then
if self[script_code].n == 0x20 then
local period = ('.'):byte()
for _ = 1, 3 do
self[script_code].n = self[script_code].n + 1
self[script_code][self[script_code].n] = period
end
else
if script_code == "Zinh" then -- probably combining character
self[script_code].n = self[script_code].n + 1
self[script_code][self[script_code].n] = 0x25CC
end
self[script_code].n = self[script_code].n + 1
self[script_code][self[script_code].n] = codepoint
end
end
end,
clear = clear,
}
})
output:insert [[
{| class="wikitable"
|+ Scripts in each Unicode block
! block !! codepoints !! scripts
]]
for _, block in pairs(mw.loadData "Module:Unicode data/blocks") do
local codepoint = block[1]
if codepoint > ending then break end
if codepoint >= start then
while codepoint <= block[2] do
local script = singles[codepoint]
local count
if script then -- Codepoint is in "singles" map.
counts:increment(script)
codepoints_per_script:add(script, codepoint)
codepoint = codepoint + 1
count = 1 -- for potential future use
else
local range, index = Unicode_data.binary_range_search(codepoint, ranges)
if range then -- Codepoint is in "ranges" array.
count = 0
script = range[3]
while codepoint <= range[2] and codepoint <= block[2] do
count = count + 1
codepoints_per_script:add(script, codepoint)
codepoint = codepoint + 1
end
counts:increment(script, count)
else -- Codepoint doesn't have data; it's Zzzz.
-- Get range immediately above codepoint.
while ranges[index][2] < codepoint do
index = index + 1
end
count = 0
script = "Zzzz"
local range = ranges[index]
while codepoint < range[1] and codepoint <= block[2]
and not singles[codepoint] do
count = count + 1
codepoint = codepoint + 1
end
counts:increment(script, count)
end
end
end
output:insert_format([[
|-
| %s
| U+%04X–U+%04X
| %s
]], link_block_name(block[3]), block[1], block[2],
table.concat(
fun.map(
function (count, script)
return ('<abbr title="%s">%s</abbr> (<span title="%s">%d</span>)')
:format(
script_data.aliases[script], script,
codepoints_per_script[script]
and mw.text.nowiki(mw.ustring.char(
unpack(codepoints_per_script[script])))
or "",
count)
end,
m_table.sortedPairs(
counts,
function (script1, script2)
return counts[script1] > counts[script2]
end)),
", "))
end
-- mw.logObject(codepoints_per_script, block[3])
counts:clear()
codepoints_per_script:clear()
end
output:insert "|}"
return output:join()
end
function p.chars_in_scripts_in_block(frame)
local block_info = get_block_info_from_arg(frame.args, 1)
local show_block_name = get_boolean_from_arg(frame.args, 2)
local script_char_set_map = print_char_set_map(
get_chars_in_scripts(fun.range(block_info[1], block_info[2])))
if show_block_name then
return ("%s: %s"):format(block_info[3], script_char_set_map)
else
return script_char_set_map
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
-- A previous draft, in [[Module:Lang/sandbox]]:
-- https://en.wikipedia.org/w/index.php?oldid=812819217
function p.parse_IETF(tag)
if not tag or tag == "" or type(tag) ~= "string" then
return nil
end
-- This contains the special fields "matched_count" and "invalid".
-- "matched_count" tracks the number of subtags, "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"
-- "invalid" is the portion of the tag after the last valid subtag (minus a
-- hyphen).
local parsed_subtags = { matched_count = 0 }
if not tag:find '^[A-Za-z0-9-]+$' then
parsed_subtags.error = "invalid characters"
parsed_subtags.invalid = tag
return parsed_subtags
end
local subtags = mw.text.split(tag, "-")
-- An array of patterns for each subtag, and a "type" field for the name
-- of the subtag.
-- The patterns are checked in order, and any of the subtags can be skipped.
-- So, for example, the "language" subtag must precede the "script"
-- subtag, but a tag may contain a "language" subtag, no "script" subtag
-- and then a "region" subtag.
-- If the full list of subtags has been iterated over, the remaining subtags
-- must match the pattern for a private-use subtag, or the tag is invalid.
local subtag_info = { -- can be put in data module
{ "%a%a%a?", "1%a+", type = "language" }, -- ll or lll; special case
-- include extlang?
{ "%a%a%a%a", type = "script" }, -- Ssss
{ "%a%a", "%d%d%d", type = "region" }, -- rr, DDD
{
"%d%d%d%d", -- 4 digits
"%w%w%w%w%w%w?%w?%w?", -- 5-8 alnum characters
type = "variant"
}
}
local index = 1
local last_matched_subtag_i = 0
for subtag_i, subtag in ipairs(subtags) do
local type
local matched = false
while not matched do
-- Check each pattern for the subtag type at "index" in "subtag_info".
for _, pattern in ipairs(subtag_info[index]) do
if subtag:find("^" .. pattern .. "$") then
type = subtag_info[index].type
matched = true
break
end
end
if not matched then -- Go to next item in subtag_info.
index = index + 1
if not subtag_info[index] then
break
end
end
end
if type then
parsed_subtags[type] = subtag
last_matched_subtag_i = subtag_i
parsed_subtags.matched_count = parsed_subtags.matched_count + 1
elseif not subtag_info[index] then
break
end
end
if #subtags > parsed_subtags.matched_count then
-- Not all subtags were matched. The unmatched tail end of the tag
-- (after the subtag at the index last_matched_subtag_i) is a
-- private-use subtag if it starts with "x". Otherwise, the tag is
-- invalid.
local suffix = table.concat(subtags, "-", last_matched_subtag_i + 1)
if suffix:sub(1, 1) == "x" then
parsed_subtags.private_use = suffix
parsed_subtags.matched_count = parsed_subtags.matched_count + 1
else
parsed_subtags.invalid = suffix
parsed_subtags.error = "invalid subtag"
end
end
if not parsed_subtags.language then
parsed_subtags.error = "no language"
end
return parsed_subtags
end
return p
