Module:headword/page
Documentation for this module may be created at Module:headword/page/doc
local require = require
local m_str_utils = require("Module:string utilities")
local mw = mw
local string = string
local table = table
local ustring = mw.ustring
local concat = table.concat
local decode_entities = m_str_utils.decode_entities
local encode_entities = m_str_utils.encode_entities
local find = string.find
local get_category = require("Module:maintenance category").get_category
local get_etym_lang = require("Module:etymology languages").getByCanonicalName
local insert = table.insert
local ipairs = ipairs
local list_to_set = require("Module:table").listToSet
local load_data = mw.loadData
local match = string.match
local new_title = mw.title.new
local remove_comments = m_str_utils.remove_comments
local split = m_str_utils.split
local string_sort = require("Module:collation").string_sort
local sub = string.sub
local toNFC = ustring.toNFC
local toNFD = ustring.toNFD
local type = type
local type_or_class = require("Module:parser").type_or_class
local u = m_str_utils.char
local ugsub = ustring.gsub
local uupper = m_str_utils.upper
local langnames = load_data("Module:languages/canonical names")
local etym_langnames = load_data("Module:etymology languages/canonical names")
local export = {}
-- Convert a numeric list of characters and ranges to the equivalent Lua pattern. WARNING: This destructively modifies
-- the contents of `ranges`.
local function char_ranges_to_pattern(ranges)
for j, range in ipairs(ranges) do
if type(range) == "table" then
for k, char in ipairs(range) do
range[k] = u(char)
end
ranges[j] = concat(range, "-")
else
ranges[j] = u(range)
end
end
return concat(ranges)
end
-- Combining character data used when categorising unusual characters. These resolve into two patterns, used to find
-- single combining characters (i.e. character + diacritic(s)) or double combining characters (i.e. character +
-- diacritic(s) + character).
local comb_chars = {
single = {
{0x0300, 0x034E},
-- Exclude combining grapheme joiner.
{0x0350, 0x035B},
{0x0363, 0x036F},
{0x0483, 0x0489},
{0x0591, 0x05BD},
0x05BF,
{0x05C1, 0x05C2},
{0x05C4, 0x05C5},
0x05C7,
{0x0610, 0x061A},
{0x064B, 0x065F},
0x0670,
{0x06D6, 0x06DC},
{0x06DF, 0x06E4},
{0x06E7, 0x06E8},
{0x06EA, 0x06ED},
0x0711,
{0x0730, 0x074A},
{0x07A6, 0x07B0},
{0x07EB, 0x07F3},
0x07FD,
{0x0816, 0x0819},
{0x081B, 0x0823},
{0x0825, 0x0827},
{0x0829, 0x082D},
{0x0859, 0x085B},
{0x0898, 0x089F},
{0x08CA, 0x08E1},
{0x08E3, 0x0903},
{0x093A, 0x093C},
{0x093E, 0x094F},
{0x0951, 0x0957},
{0x0962, 0x0963},
{0x0981, 0x0983},
0x09BC,
{0x09BE, 0x09C4},
{0x09C7, 0x09C8},
{0x09CB, 0x09CD},
0x09D7,
{0x09E2, 0x09E3},
0x09FE,
{0x0A01, 0x0A03},
0x0A3C,
{0x0A3E, 0x0A42},
{0x0A47, 0x0A48},
{0x0A4B, 0x0A4D},
0x0A51,
{0x0A70, 0x0A71},
0x0A75,
{0x0A81, 0x0A83},
0x0ABC,
{0x0ABE, 0x0AC5},
{0x0AC7, 0x0AC9},
{0x0ACB, 0x0ACD},
{0x0AE2, 0x0AE3},
{0x0AFA, 0x0AFF},
{0x0B01, 0x0B03},
0x0B3C,
{0x0B3E, 0x0B44},
{0x0B47, 0x0B48},
{0x0B4B, 0x0B4D},
{0x0B55, 0x0B57},
{0x0B62, 0x0B63},
0x0B82,
{0x0BBE, 0x0BC2},
{0x0BC6, 0x0BC8},
{0x0BCA, 0x0BCD},
0x0BD7,
{0x0C00, 0x0C04},
0x0C3C,
{0x0C3E, 0x0C44},
{0x0C46, 0x0C48},
{0x0C4A, 0x0C4D},
{0x0C55, 0x0C56},
{0x0C62, 0x0C63},
{0x0C81, 0x0C83},
0x0CBC,
{0x0CBE, 0x0CC4},
{0x0CC6, 0x0CC8},
{0x0CCA, 0x0CCD},
{0x0CD5, 0x0CD6},
{0x0CE2, 0x0CE3},
0x0CF3,
{0x0D00, 0x0D03},
{0x0D3B, 0x0D3C},
{0x0D3E, 0x0D44},
{0x0D46, 0x0D48},
{0x0D4A, 0x0D4D},
0x0D57,
{0x0D62, 0x0D63},
{0x0D81, 0x0D83},
0x0DCA,
{0x0DCF, 0x0DD4},
0x0DD6,
{0x0DD8, 0x0DDF},
{0x0DF2, 0x0DF3},
0x0E31,
{0x0E34, 0x0E3A},
{0x0E47, 0x0E4E},
0x0EB1,
{0x0EB4, 0x0EBC},
{0x0EC8, 0x0ECE},
{0x0F18, 0x0F19},
0x0F35,
0x0F37,
0x0F39,
{0x0F3E, 0x0F3F},
{0x0F71, 0x0F84},
{0x0F86, 0x0F87},
{0x0F8D, 0x0F97},
{0x0F99, 0x0FBC},
0x0FC6,
{0x102B, 0x103E},
{0x1056, 0x1059},
{0x105E, 0x1060},
{0x1062, 0x1064},
{0x1067, 0x106D},
{0x1071, 0x1074},
{0x1082, 0x108D},
0x108F,
{0x109A, 0x109D},
{0x135D, 0x135F},
{0x1712, 0x1715},
{0x1732, 0x1734},
{0x1752, 0x1753},
{0x1772, 0x1773},
{0x17B4, 0x17D3},
0x17DD,
-- Exclude Mongolian variation selectors.
{0x1885, 0x1886},
0x18A9,
{0x1920, 0x192B},
{0x1930, 0x193B},
{0x1A17, 0x1A1B},
{0x1A55, 0x1A5E},
{0x1A60, 0x1A7C},
0x1A7F,
{0x1AB0, 0x1ACE},
{0x1B00, 0x1B04},
{0x1B34, 0x1B44},
{0x1B6B, 0x1B73},
{0x1B80, 0x1B82},
{0x1BA1, 0x1BAD},
{0x1BE6, 0x1BF3},
{0x1C24, 0x1C37},
{0x1CD0, 0x1CD2},
{0x1CD4, 0x1CE8},
0x1CED,
0x1CF4,
{0x1CF7, 0x1CF9},
{0x1DC0, 0x1DCC},
{0x1DCE, 0x1DFB},
{0x1DFD, 0x1DFF},
{0x20D0, 0x20F0},
{0x2CEF, 0x2CF1},
0x2D7F,
{0x2DE0, 0x2DFF},
{0x302A, 0x302F},
{0x3099, 0x309A},
{0xA66F, 0xA672},
{0xA674, 0xA67D},
{0xA69E, 0xA69F},
{0xA6F0, 0xA6F1},
0xA802,
0xA806,
0xA80B,
{0xA823, 0xA827},
0xA82C,
{0xA880, 0xA881},
{0xA8B4, 0xA8C5},
{0xA8E0, 0xA8F1},
0xA8FF,
{0xA926, 0xA92D},
{0xA947, 0xA953},
{0xA980, 0xA983},
{0xA9B3, 0xA9C0},
0xA9E5,
{0xAA29, 0xAA36},
0xAA43,
{0xAA4C, 0xAA4D},
{0xAA7B, 0xAA7D},
0xAAB0,
{0xAAB2, 0xAAB4},
{0xAAB7, 0xAAB8},
{0xAABE, 0xAABF},
0xAAC1,
{0xAAEB, 0xAAEF},
{0xAAF5, 0xAAF6},
{0xABE3, 0xABEA},
{0xABEC, 0xABED},
0xFB1E,
{0xFE20, 0xFE2F},
0x101FD,
0x102E0,
{0x10376, 0x1037A},
{0x10A01, 0x10A03},
{0x10A05, 0x10A06},
{0x10A0C, 0x10A0F},
{0x10A38, 0x10A3A},
0x10A3F,
{0x10AE5, 0x10AE6},
{0x10D24, 0x10D27},
{0x10EAB, 0x10EAC},
{0x10EFD, 0x10EFF},
{0x10F46, 0x10F50},
{0x10F82, 0x10F85},
{0x11000, 0x11002},
{0x11038, 0x11046},
0x11070,
{0x11073, 0x11074},
{0x1107F, 0x11082},
{0x110B0, 0x110BA},
0x110C2,
{0x11100, 0x11102},
{0x11127, 0x11134},
{0x11145, 0x11146},
0x11173,
{0x11180, 0x11182},
{0x111B3, 0x111C0},
{0x111C9, 0x111CC},
{0x111CE, 0x111CF},
{0x1122C, 0x11237},
0x1123E,
0x11241,
{0x112DF, 0x112EA},
{0x11300, 0x11303},
{0x1133B, 0x1133C},
{0x1133E, 0x11344},
{0x11347, 0x11348},
{0x1134B, 0x1134D},
0x11357,
{0x11362, 0x11363},
{0x11366, 0x1136C},
{0x11370, 0x11374},
{0x11435, 0x11446},
0x1145E,
{0x114B0, 0x114C3},
{0x115AF, 0x115B5},
{0x115B8, 0x115C0},
{0x115DC, 0x115DD},
{0x11630, 0x11640},
{0x116AB, 0x116B7},
{0x1171D, 0x1172B},
{0x1182C, 0x1183A},
{0x11930, 0x11935},
{0x11937, 0x11938},
{0x1193B, 0x1193E},
0x11940,
{0x11942, 0x11943},
{0x119D1, 0x119D7},
{0x119DA, 0x119E0},
0x119E4,
{0x11A01, 0x11A0A},
{0x11A33, 0x11A39},
{0x11A3B, 0x11A3E},
0x11A47,
{0x11A51, 0x11A5B},
{0x11A8A, 0x11A99},
{0x11C2F, 0x11C36},
{0x11C38, 0x11C3F},
{0x11C92, 0x11CA7},
{0x11CA9, 0x11CB6},
{0x11D31, 0x11D36},
0x11D3A,
{0x11D3C, 0x11D3D},
{0x11D3F, 0x11D45},
0x11D47,
{0x11D8A, 0x11D8E},
{0x11D90, 0x11D91},
{0x11D93, 0x11D97},
{0x11EF3, 0x11EF6},
{0x11F00, 0x11F01},
0x11F03,
{0x11F34, 0x11F3A},
{0x11F3E, 0x11F42},
0x13440,
{0x13447, 0x13455},
{0x16AF0, 0x16AF4},
{0x16B30, 0x16B36},
0x16F4F,
{0x16F51, 0x16F87},
{0x16F8F, 0x16F92},
-- Exclude Khitan Small Script filler.
{0x16FF0, 0x16FF1},
{0x1BC9D, 0x1BC9E},
{0x1CF00, 0x1CF2D},
{0x1CF30, 0x1CF46},
{0x1D165, 0x1D169},
{0x1D16D, 0x1D172},
{0x1D17B, 0x1D182},
{0x1D185, 0x1D18B},
{0x1D1AA, 0x1D1AD},
{0x1D242, 0x1D244},
{0x1DA00, 0x1DA36},
{0x1DA3B, 0x1DA6C},
0x1DA75,
0x1DA84,
{0x1DA9B, 0x1DA9F},
{0x1DAA1, 0x1DAAF},
{0x1E000, 0x1E006},
{0x1E008, 0x1E018},
{0x1E01B, 0x1E021},
{0x1E023, 0x1E024},
{0x1E026, 0x1E02A},
0x1E08F,
{0x1E130, 0x1E136},
0x1E2AE,
{0x1E2EC, 0x1E2EF},
{0x1E4EC, 0x1E4EF},
{0x1E8D0, 0x1E8D6},
{0x1E944, 0x1E94A},
},
double = {
{0x035C, 0x0362},
0x1DCD,
0x1DFC,
},
vs = { -- variation selectors; separated out so that we don't get categories for them
{0xFE00, 0xFE0F},
{0xE0100, 0xE01EF},
}
}
for key, charset in pairs(comb_chars) do
comb_chars[key] = char_ranges_to_pattern(charset)
end
local comb_chars_all = comb_chars.single .. comb_chars.double .. comb_chars.vs
comb_chars = {
combined_single = "[^" .. comb_chars_all .. "][" .. comb_chars.single .. comb_chars.vs .. "]+%f[^" .. comb_chars_all .. "]",
combined_double = "[^" .. comb_chars_all .. "][" .. comb_chars.single .. comb_chars.vs .. "]*[" .. comb_chars.double .. "]+[" .. comb_chars_all .. "]*.[" .. comb_chars.single .. comb_chars.vs .. "]*",
diacritics_single = "[" .. comb_chars.single .. "]",
diacritics_double = "[" .. comb_chars.double .. "]",
diacritics_all = "[" .. comb_chars_all .. "]"
}
-- From https://unicode.org/Public/emoji/15.1/emoji-sequences.txt
local emoji_chars = {
{0x231A, 0x231B}, -- watch..hourglass done # E0.6 [2] (⌚..⌛)
{0x23E9, 0x23EC}, -- fast-forward button..fast down button # E0.6 [4] (⏩..⏬)
0x23F0, -- alarm clock # E0.6 [1] (⏰)
0x23F3, -- hourglass not done # E0.6 [1] (⏳)
{0x25FD, 0x25FE}, -- white medium-small square..black medium-small square # E0.6 [2] (◽..◾)
{0x2614, 0x2615}, -- umbrella with rain drops..hot beverage # E0.6 [2] (☔..☕)
{0x2648, 0x2653}, -- Aries..Pisces # E0.6 [12] (♈..♓)
0x267F, -- wheelchair symbol # E0.6 [1] (♿)
0x2693, -- anchor # E0.6 [1] (⚓)
0x26A1, -- high voltage # E0.6 [1] (⚡)
{0x26AA, 0x26AB}, -- white circle..black circle # E0.6 [2] (⚪..⚫)
{0x26BD, 0x26BE}, -- soccer ball..baseball # E0.6 [2] (⚽..⚾)
{0x26C4, 0x26C5}, -- snowman without snow..sun behind cloud # E0.6 [2] (⛄..⛅)
0x26CE, -- Ophiuchus # E0.6 [1] (⛎)
0x26D4, -- no entry # E0.6 [1] (⛔)
0x26EA, -- church # E0.6 [1] (⛪)
{0x26F2, 0x26F3}, -- fountain..flag in hole # E0.6 [2] (⛲..⛳)
0x26F5, -- sailboat # E0.6 [1] (⛵)
0x26FA, -- tent # E0.6 [1] (⛺)
0x26FD, -- fuel pump # E0.6 [1] (⛽)
0x2705, -- check mark button # E0.6 [1] (✅)
{0x270A, 0x270B}, -- raised fist..raised hand # E0.6 [2] (✊..✋)
0x2728, -- sparkles # E0.6 [1] (✨)
0x274C, -- cross mark # E0.6 [1] (❌)
0x274E, -- cross mark button # E0.6 [1] (❎)
{0x2753, 0x2755}, -- red question mark..white exclamation mark # E0.6 [3] (❓..❕)
0x2757, -- red exclamation mark # E0.6 [1] (❗)
{0x2795, 0x2797}, -- plus..divide # E0.6 [3] (➕..➗)
0x27B0, -- curly loop # E0.6 [1] (➰)
0x27BF, -- double curly loop # E1.0 [1] (➿)
{0x2B1B, 0x2B1C}, -- black large square..white large square # E0.6 [2] (⬛..⬜)
0x2B50, -- star # E0.6 [1] (⭐)
0x2B55, -- hollow red circle # E0.6 [1] (⭕)
{0x1F300, 0x1FAFF}, -- emoji in Plane 1
-- NOTE: There are lots more emoji sequences involving non-emoji Plane 0 symbols followed by 0xFE0F, which we don't
-- (yet?) handle.
}
emoji_chars = char_ranges_to_pattern(emoji_chars)
local unsupported_characters = {}
for k, v in pairs(require("Module:links/data").unsupported_characters) do
unsupported_characters[v] = k
end
-- Get the list of unsupported titles and invert it (so the keys are pagenames and values are canonical titles).
local unsupported_titles = {}
for k, v in pairs(require("Module:links/data").unsupported_titles) do
unsupported_titles[v] = k
end
--[==[
Given a pagename (or {nil} for the current page), create and return a data structure describing the page. The returned
object includes the following fields:
* `comb_chars`: A table containing various Lua character class patterns for different types of combined characters
(those that decompose into multiple characters in the NFD decomposition). The patterns are meant to be used with
{mw.ustring.find()}. The keys are:
** `single`: Single combining characters (character + diacritic), without surrounding brackets;
** `double`: Double combining characters (character + diacritic + character), without surrounding brackets;
** `vs`: Variation selectors, without surrounding brackets;
** `all`: Concatenation of `single` + `double` + `vs`, without surrounding brackets;
** `diacritics_single`: Like `single` but with surrounding brackets;
** `diacritics_double`: Like `double` but with surrounding brackets;
** `diacritics_all`: Like `all` but with surrounding brackets;
** `combined_single`: Lua pattern for matching a spacing character followed by one or more single combining characters;
** `combined_double`: Lua pattern for matching a combination of two spacing characters separated by one or more double
combining characters, possibly also with single combining characters;
* `emoji_pattern`: A Lua character class pattern (including surrounding brackets) that matches emojis. Meant to be used
with {mw.ustring.find()}.
* `L2_list`: Ordered list of L2 headings on the page, with the extra key `n` that gives the length of the list.
* `L2_sections`: Lookup table of L2 headings on the page, where the key is the section number assigned by the preprocessor, and the value is the L2 heading name. Once an invocation has got its actual section number from get_current_section in [[Module:utilities]], it can use this table to determine its parent L2. TODO: We could expand this to include subsections, to check POS headings are correct etc.
* `unsupported_titles`: Map from pagenames to canonical titles for unsupported-title pages.
* `namespace`: Namespace of the pagename.
* `ns`: Namespace table for the page from mw.site.namespaces (TODO: merge with `namespace` above).
* `full_raw_pagename`: Full version of the '''RAW''' pagename (i.e. unsupported-title pages aren't canonicalized);
including the namespace and the root (portion before the slash).
* `pagename`: Canonicalized subpage portion of the pagename (unsupported-title pages are canonicalized).
* `decompose_pagename`: Equivalent of `pagename` in NFD decomposition.
* `pagename_len`: Length of `pagename` in Unicode chars, where combinations of spacing character + decomposed diacritic
are treated as single characters.
* `explode_pagename`: Set of characters found in `pagename`. The keys are characters (where combinations of spacing
character + decomposed diacritic are treated as single characters).
* `encoded_pagename`: FIXME: Document me.
* `pagename_defaultsort`: FIXME: Document me.
* `raw_defaultsort`: FIXME: Document me.
* `wikitext_topic_cat`: FIXME: Document me.
* `wikitext_langname_cat`: FIXME: Document me.
]==]
function export.process_page(pagename)
local data = {
comb_chars = comb_chars,
emoji_pattern = "[" .. emoji_chars .. "]",
unsupported_titles = unsupported_titles
}
local cats = {}
data.cats = cats
-- We cannot store `raw_title` in `data` because it contains a metatable.
local raw_title
local function bad_pagename()
if not pagename then
error("Internal error: Something wrong, `data.pagename` not specified but current title containg illegal characters")
else
error(("Bad value for `data.pagename`: '%s', which must not contain illegal characters"):format(pagename))
end
end
if pagename then -- for testing, doc pages, etc.
raw_title = new_title(pagename)
if not raw_title then
bad_pagename()
end
else
raw_title = mw.title.getCurrentTitle()
end
data.namespace = raw_title.nsText
data.ns = mw.site.namespaces[raw_title.namespace]
data.full_raw_pagename = raw_title.fullText
local frame = mw.getCurrentFrame()
-- WARNING: `content` May be nil, e.g. if we're substing a template like {{ja-new}} on a not-yet-created page
-- or if the module specifies the subpage as `data.pagename` (which many modules do) and we're in an Appendix
-- or other non-mainspace page. We used to make the latter an error but there are too many modules that do it,
-- and substing on a nonexistent page is totally legit, and we don't actually need to be able to access the
-- content of the page.
local content = raw_title:getContent()
local content_lang = mw.getContentLanguage()
--Get the pagename.
pagename = raw_title.subpageText
:gsub("^Unsupported titles/(.*)", function(m)
insert(cats, "Unsupported titles")
return unsupported_titles[m] or (m:gsub("`.-`", unsupported_characters))
end)
-- Save pagename, as local variable will be destructively modified.
data.pagename = pagename
-- Decompose the pagename in Unicode normalization form D.
data.decompose_pagename = toNFD(pagename)
-- Explode the current page name into a character table, taking decomposed combining characters into account.
local explode_pagename = {}
local pagename_len = 0
local function explode(char)
explode_pagename[char] = true
pagename_len = pagename_len + 1
return ""
end
pagename = ugsub(pagename, comb_chars.combined_double, explode)
pagename = ugsub(pagename, comb_chars.combined_single, explode)
:gsub(".[\128-\191]*", explode)
data.explode_pagename = explode_pagename
data.pagename_len = pagename_len
-- Generate DEFAULTSORT.
data.encoded_pagename = encode_entities(data.pagename)
data.pagename_defaultsort = require("Module:languages").getByCode("mul"):makeSortKey(data.encoded_pagename)
frame:callParserFunction(
"DEFAULTSORT",
data.pagename_defaultsort
)
data.raw_defaultsort = raw_title.text:uupper()
-- Make `L2_list` and `L2_sections`, note raw wikitext use of {{DEFAULTSORT:}} and {{DISPLAYTITLE:}}, then add categories if any unwanted L1 headings are found, the L2 headings are in the wrong order, or they don't match a canonical language name.
-- Note: HTML comments shouldn't be removed from `content` until after this step, as they can affect the result.
do
local L2_list, L2_list_len, L2_sections, sort_cache, prev = {}, 0, {}, {}
local defaultsort, displaytitle, page_has_L1, L2_wrong_order, L2_nonstandard, page_has_arg
local function get_weight(L2)
if L2 == "Translingual" then
return "\1"
elseif L2 == "English" then
return "\2"
elseif match(L2, "^[%z\1-\b\14-!#-&(-,.-\127]+$") then
return L2
end
local weight = sort_cache[L2]
if weight then
return weight
end
weight = toNFC(ugsub(ugsub(toNFD(L2), "[" .. comb_chars_all .. "'\"ʻʼ]+", ""), "[%s%-]+", " "))
sort_cache[L2] = weight
return weight
end
local function handle_heading(node)
local level = node.level
if level > 2 then
return
end
local name = node:get_name()
-- Check there are no newline characters in the heading, which might appear after preprocessing (e.g. from an expanded template). In such cases, the preprocessor section count still increments (since it's calculated pre-expansion), but the heading will fail, so we shouldn't increment the L2 count.
if find(name, "\n", 1, true) then
return
end
L2_list_len = L2_list_len + 1
L2_list[L2_list_len] = name
L2_sections[node.section] = name
-- We also add any L1s, since they terminate the preceding L2, but add a maintenance category since it's probably a mistake.
if level == 1 then
page_has_L1 = true
end
-- Check the heading is in the right order.
-- FIXME: we need a more sophisticated sorting method which handles non-diacritic special characters (e.g. Magɨ).
if prev and not (
L2_wrong_order or
string_sort(get_weight(prev), get_weight(name))
) then
L2_wrong_order = true
end
-- Check it's a canonical language name.
if not langnames[name] then
L2_nonstandard = true
end
prev = name
end
local function handle_template(node)
local name = node:get_name()
if name == "DEFAULTSORT:" then
defaultsort = true
elseif name == "DISPLAYTITLE:" then
displaytitle = true
end
end
if content then
for node in require("Module:template parser").parse(content):__pairs("next_node") do
local node_type = type_or_class(node)
if node_type == "heading" then
handle_heading(node)
elseif node_type == "template" and not (defaultsort and displaytitle) then
handle_template(node)
elseif node_type == "argument" then
page_has_arg = true
end
end
end
L2_list.n = L2_list_len
data.L2_list = L2_list
data.L2_sections = L2_sections
if defaultsort then
insert(cats, get_category("Pages with DEFAULTSORT conflicts"))
end
if displaytitle then
insert(cats, get_category("Pages with DISPLAYTITLE conflicts"))
end
if page_has_L1 then
insert(cats, get_category("Pages with unwanted L1 headings"))
end
if L2_wrong_order then
insert(cats, get_category("Pages with language headings in the wrong order"))
end
if L2_nonstandard then
insert(cats, get_category("Pages with nonstandard language headings"))
end
if page_has_arg then
insert(cats, get_category("Pages with raw triple-brace template arguments"))
end
end
------ 4. Parse page for maintenance categories. ------
-- Use of tab characters.
if content and find(content, "\t", 1, true) then
insert(cats, get_category("Pages with tab characters"))
end
-- Unencoded character(s) in title.
local IDS = list_to_set{"⿰", "⿱", "⿲", "⿳", "⿴", "⿵", "⿶", "⿷", "⿸", "⿹", "⿺", "⿻", "⿼", "⿽", "⿾", "⿿", "㇯"}
for char in pairs(explode_pagename) do
if IDS[char] and char ~= data.pagename then
insert(cats, "Terms containing unencoded characters")
break
end
end
-- Raw wikitext use of a topic or langname category. Also check if any raw sortkeys have been used.
do
local wikitext_topic_cat = {}
local wikitext_langname_cat = {}
local raw_sortkey
-- If a raw sortkey has been found, add it to the relevant table.
-- If there's no table (or the index is just `true`), create one first.
local function add_cat_table(t, lang, sortkey)
local t_lang = t[lang]
if not sortkey then
if not t_lang then
t[lang] = true
end
return
elseif t_lang == true or not t_lang then
t_lang = {}
t[lang] = t_lang
end
t_lang[uupper(decode_entities(sortkey))] = true
end
local function do_iteration(name, sortkey, wikitext_langname_cat)
if langnames[name] then
return add_cat_table(wikitext_langname_cat, name, sortkey)
end
name = etym_langnames[name] and name or content_lang:lcfirst(name)
if etym_langnames[name] then
name = get_etym_lang(name):getFullName()
return add_cat_table(wikitext_langname_cat, name, sortkey)
end
end
local function process_category(content, cat, colon, nxt)
local pipe = find(cat, "|", colon + 1, true)
-- Categories cannot end "|]]".
if pipe == #cat then
return
end
local title = new_title(pipe and sub(cat, 1, pipe - 1) or cat)
if not (title and title.namespace == 14) then
return
end
-- Get the sortkey (if any), then canonicalize category title.
local sortkey = pipe and sub(cat, pipe + 1) or nil
cat = title.text
if sortkey then
raw_sortkey = true
-- If the sortkey contains "[", the first "]" of a final "]]]" is treated as part of the sortkey.
if find(sortkey, "[", 1, true) and sub(content, nxt, nxt) == "]" then
sortkey = sortkey .. "]"
end
end
local code = match(cat, "^([%w%-.]+):")
if code then
return add_cat_table(wikitext_topic_cat, code, sortkey)
end
-- Split by word.
cat = split(cat, " ", true, true)
-- Iterate over the category name, starting with the longest possible name and shaving off the first word until we find one. We do it this way because:
-- (a) Going from shortest to longest risks falsely matching (e.g.) German Low German categories as German.
-- (b) Checking the start of category names first risks falsely match (e.g.) Alsatian French as Alsatian (a variety of Alemannic German), not French.
-- If no matches are found, then check the start of the category name, shaving off the last word each iteration.
local cat_len, n, name, done = #cat, 1
repeat
name = concat(cat, " ", n, cat_len)
done = do_iteration(name, sortkey, wikitext_langname_cat)
if done then
return
end
n = n + 1
until n > cat_len
n = cat_len - 1
if n <= 0 then
return
end
repeat
name = concat(cat, " ", 1, n)
done = do_iteration(name, sortkey, wikitext_langname_cat)
if done then
return
end
n = n - 1
until n == 0
end
if content then
-- Remove comments, then iterate over category links.
content = remove_comments(content, "BOTH")
local head = find(content, "[[", 1, true)
while head do
local close = find(content, "]]", head + 2, true)
if not close then
break
end
-- Make sure there are no intervening "[[" between head and close.
local open = find(content, "[[", head + 2, true)
while open and open < close do
head = open
open = find(content, "[[", head + 2, true)
end
local cat = sub(content, head + 2, close - 1)
-- Locate the colon, and weed out most unwanted links. "[ _\128-\244]*" catches valid whitespace, and ensures any category links using the colon trick are ignored. We match all non-ASCII characters, as there could be multibyte spaces, and mw.title.new will filter out any remaining false-positives; this is a lot faster than running mw.title.new on every link.
local colon = match(cat, "^[ _\128-\244]*[Cc][Aa][Tt][EeGgOoRrYy _\128-\244]*():")
if colon then
process_category(content, cat, colon, close + 2)
end
head = open
end
end
data.wikitext_topic_cat = wikitext_topic_cat
data.wikitext_langname_cat = wikitext_langname_cat
if raw_sortkey then
insert(cats, get_category("Pages with raw sortkeys"))
end
end
return data
end
return export