Wiktionary tgwiktionary https://tg.wiktionary.org/wiki/%D0%A1%D0%B0%D2%B3%D0%B8%D1%84%D0%B0%D0%B8_%D0%90%D1%81%D0%BB%D3%A3 MediaWiki 1.46.0-wmf.26 case-sensitive Медиа Вижа Баҳс Корбар Баҳси корбар Wiktionary Баҳси Wiktionary Акс Баҳси акс Медиавики Баҳси медиавики Шаблон Баҳси шаблон Роҳнамо Баҳси роҳнамо Гурӯҳ Баҳси гурӯҳ TimedText TimedText talk Модул Баҳси Модул Event Event talk 828 13966 123877 122088 2026-05-05T06:45:17Z Hiyuune 3448 123877 Scribunto text/plain local export = {} local array_module = "Module:array" local debug_track_module = "Module:debug/track" local frame_module = "Module:frame" local fun_is_callable_module = "Module:fun/isCallable" local languages_module = "Module:languages" local links_module = "Module:links" local load_module = "Module:load" local module_categorization_module = "Module:module categorization" local number_list_show_module = "Module:number list/show" local chemical_element_list_show_module = "Module:chemical element list/show" local pages_module = "Module:pages" local parameters_module = "Module:parameters" local scripts_module = "Module:scripts" local string_endswith_module = "Module:string/endswith" local string_gline_module = "Module:string/gline" local string_insert_module = "Module:string/insert" local string_startswith_module = "Module:string/startswith" local string_utilities_module = "Module:string utilities" local template_parser_module = "Module:template parser" local title_exists_module = "Module:title/exists" local title_new_title_module = "Module:title/newTitle" local concat = table.concat local error = error local full_url = mw.uri.fullUrl local get_current_title = mw.title.getCurrentTitle local insert = table.insert local ipairs = ipairs local list_to_text = mw.text.listToText local new_message = mw.message.new local pcall = pcall local require = require local tonumber = tonumber local tostring = tostring local type = type local unpack = unpack or table.unpack -- Lua 5.2 compatibility local function Array(...) Array = require(array_module) return Array(...) end local function categorize_module(...) categorize_module = require(module_categorization_module).categorize return categorize_module(...) end local function debug_track(...) debug_track = require(debug_track_module) return debug_track(...) end local function endswith(...) endswith = require(string_endswith_module) return endswith(...) end local function expand_template(...) expand_template = require(frame_module).expandTemplate return expand_template(...) end local function find_templates(...) find_templates = require(template_parser_module).find_templates return find_templates(...) end local function full_link(...) full_link = require(links_module).full_link return full_link(...) end local function get_lang(...) get_lang = require(languages_module).getByCode return get_lang(...) end local function get_pagetype(...) get_pagetype = require(pages_module).get_pagetype return get_pagetype(...) end local function get_script(...) get_script = require(scripts_module).getByCode return get_script(...) end local function gline(...) gline = require(string_gline_module) return gline(...) end local function is_callable(...) is_callable = require(fun_is_callable_module) return is_callable(...) end local function is_documentation(...) is_documentation = require(pages_module).is_documentation return is_documentation(...) end local function is_sandbox(...) is_sandbox = require(pages_module).is_sandbox return is_sandbox(...) end local function new_title(...) new_title = require(title_new_title_module) return new_title(...) end local function number_list_show_table(...) number_list_show_table = require(number_list_show_module).table return number_list_show_table(...) end local function chemical_element_list_show_table(...) chemical_element_list_show_table = require(chemical_element_list_show_module).table return chemical_element_list_show_table(...) end local function preprocess(...) preprocess = require(frame_module).preprocess return preprocess(...) end local function process_params(...) process_params = require(parameters_module).process return process_params(...) end local function safe_load_data(...) safe_load_data = require(load_module).safe_load_data return safe_load_data(...) end local function split(...) split = require(string_utilities_module).split return split(...) end local function startswith(...) startswith = require(string_startswith_module) return startswith(...) end local function string_insert(...) string_insert = require(string_insert_module) return string_insert(...) end local function title_exists(...) title_exists = require(title_exists_module) return title_exists(...) end local function ugsub(...) ugsub = require(string_utilities_module).gsub return ugsub(...) end local function umatch(...) umatch = require(string_utilities_module).match return umatch(...) end local skins = { ["common"] = "", ["vector"] = "Vector", ["monobook"] = "Monobook", ["cologneblue"] = "Cologne Blue", ["modern"] = "Modern", } local function track(page) debug_track("documentation/" .. page) return true end local function compare_pages(page1, page2, text) return "[" .. tostring( full_url("Special:ComparePages", { page1 = page1, page2 = page2 })) .. " " .. text .. "]" end -- Avoid transcluding [[Module:languages/cache]] everywhere. local lang_cache = setmetatable({}, { __index = function(self, k) return require("Module:languages/cache")[k] end }) local function zh_link(word) return full_link { lang = lang_cache.zh, term = word } end local function make_languages_data_documentation(title, cats, division) local doc_template, module_cat if endswith(division, "/extra") then division = division:sub(1, -7) doc_template = "language extradata documentation" module_cat = "Language extra data modules" else doc_template = "language data documentation" module_cat = "Language data modules" end local sort_key if division == "exceptional" then sort_key = "x" else sort_key = division:gsub("/", "") end cats:insert(module_cat .. "|" .. sort_key) return { title = doc_template } end local function make_Unicode_data_documentation(title, cats) local subpage, first_three_of_code_point = title.fullText:match("^Module:Unicode data/([^/]+)/(%x%x%x)$") if subpage == "names" or subpage == "images" or subpage == "emoji images" then local low, high = tonumber(first_three_of_code_point .. "000", 16), tonumber(first_three_of_code_point .. "FFF", 16) local text, text_type if subpage == "names" then text_type = "titles of images" elseif subpage == "images" then text_type = "titles of images" elseif subpage == "emoji images" then text_type = "emoji-style images" end text = string.format( "This data module contains the " .. text_type .. " of " .. "[[Appendix:Unicode|Unicode]] code points within the range U+%04X to U+%04X.", low, high) if subpage == "images" and safe_load_data("Module:Unicode data/emoji images/" .. first_three_of_code_point) then text = text .. " This list includes the text variants of emojis. For the list of emoji variants of those characters, see [[Module:Unicode data/emoji images/" .. first_three_of_code_point .. "]]." elseif subpage == "emoji images" then text = text .. " For text-style images, see [[Module:Unicode data/images/" .. first_three_of_code_point .. "]]." end return text end end local function insert_lang_data_module_cats(cats, langcode, overall_data_module_cat) local lang = lang_cache[langcode] if lang then local langname if lang._fullCode then langname = lang_cache[lang._fullCode]:getCanonicalName() else langname = lang:getCanonicalName() end cats:insert(overall_data_module_cat .. "|" .. langname) cats:insert(langname .. " modules") cats:insert(langname .. " data modules") return lang, langname end end --[=[ This provides categories and documentation for various data modules, so that [[Category:Uncategorized modules]] isn't unnecessarily cluttered. It is a list of tables, each of which have the following possible fields: `regex` (required): A Lua pattern to match the module's title. If it matches, the data in this entry will be used. Any captures in the pattern can by referenced in the `cat` field using %1 for the first capture, %2 for the second, etc. (often used for creating the sortkey for the category). In addition, the captures are passed to the `process` function as the third and subsequent parameters. `process` (optional): This may be a function or a string. If it is a function, it is called as follows: `process(TITLE, CATS, CAPTURE1, CAPTURE2, ...)` where: * TITLE is a title object describing the module's title; see [https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Title_objects]. * CATS is an array object (see [[Module:array]]) of categories that the module will be added to. * CAPTURE1, CAPTURE2, ... contain any captures in the `regex` field. The return value of `process` should either be a string (which will be used as the module's documentation), or a table specifying the name of a template to expand to get the documentation, along with the arguments to that template. In the latter format, the template name (bare, without the "Template:" prefix) should be in the `title` field, and any arguments should be in `args; in this case, the template name will be listed above the generated documentation as the source of the documentation, along with an edit button to edit the template's contents. If, however, the return value of the `process` function is a string, any template invocations will be expanded using frame:preprocess(), and [[Module:documentation]] will be listed as the source of the documentation. If `process` itself is a string rather than a function, it should name a submodule under [[Module:documentation/functions/]] which returns a function, of the same type as described above. This submodule will be specified as the source of the documentation (unless it returns a table naming a template to expand to get the documentation, as described above). If `process` is omitted entirely, the module will have no documentation. `cat` (optional): A string naming the category into which the module should be placed, or a list of such strings. Captures specified in `regex` may be referenced in this string using %1 for the first capture, %2 for the second, etc. It is also possible to add categories in the `process` function by inserting them into the passed-in CATS array (the second parameter). ]=] local module_regex = { { regex = "^Module:languages/data/(3/%l/extra)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/data/(3/%l)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/data/(2/extra)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/data/(2)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/data/(exceptional/extra)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/data/(exceptional)$", process = make_languages_data_documentation, }, { regex = "^Module:languages/.+$", cat = "Language and script modules", }, { regex = "^Module:scripts/.+$", cat = "Language and script modules", }, { regex = "^Module:data tables/data..?.?.?$", cat = "Reference module sharded data tables", }, { regex = "^Module:zh/data/dial%-pron/.+$", cat = "Chinese dialectal pronunciation data modules", process = "zh dial or syn", }, { regex = "^Module:zh/data/dial%-syn/.+$", cat = "Chinese dialect synonyms data modules", process = "zh dial or syn", }, { regex = "^Module:zh/data/glyph%-data/.+$", cat = "Chinese historical character forms data modules", process = function(title, cats) local character = title.fullText:match("^Module:zh/data/glyph%-data/(.+)") if character then return ("This module contains data on historical forms of the Chinese character %s.") :format(zh_link(character)) end end, }, { regex = "^Module:zh/data/ltc%-pron/(.+)$", cat = "Middle Chinese pronunciation data modules|%1", process = "zh data", }, { regex = "^Module:zh/data/och%-pron%-BS/(.+)$", cat = "Old Chinese (Baxter-Sagart) pronunciation data modules|%1", process = "zh data", }, { regex = "^Module:zh/data/och%-pron%-ZS/(.+)$", cat = "Old Chinese (Zhengzhang) pronunciation data modules|%1", process = "zh data", }, { -- capture rest of zh/data submodules regex = "^Module:zh/data/(.+)$", cat = "Chinese data modules|%1", }, { regex = "^Module:mul/guoxue%-data/cjk%-?(.*)$", process = "guoxue-data", }, { regex = "^Module:Unicode data/(.+)$", cat = "Unicode data modules|%1", process = make_Unicode_data_documentation, }, { regex = "^Module:number list/data/(.+)$", process = function(title, cats, lang_code) local lang = insert_lang_data_module_cats(cats, lang_code, "Number data modules") if lang then return ("This module contains data on various types of numbers in %s.\n%s") :format(lang:makeCategoryLink(), number_list_show_table() or "") end end, }, { regex = "^Module:chemical element list/data/(.+)$", process = function(title, cats, lang_code) local lang = insert_lang_data_module_cats(cats, lang_code, "Chemical element data modules") if lang then return ("This module contains data on chemical elements in %s.\n%s") :format(lang:makeCategoryLink(), chemical_element_list_show_table() or "") end end, }, { regex = "^Module:accel/(.+)$", process = function(title, cats) local lang_code = title.subpageText local lang = lang_cache[lang_code] if lang then cats:insert(lang:getCanonicalName() .. " modules|accel") cats:insert(("Accel submodules|%s"):format(lang:getCanonicalName())) return ("This module contains new entry creation rules for %s; see [[WT:ACCEL]] for an overview, and [[Module:accel]] for information on creating new rules.") :format(lang:makeCategoryLink()) end end, }, { regex = "^Module:inc%-ash/dial/data/(.+)$", cat = "Ashokan Prakrit modules|%1", process = function(title, cats) local word = title.fullText:match("^Module:inc%-ash/dial/data/(.+)$") if word then local lang = lang_cache["inc-ash"] return ("This module contains data on the pronunciation of %s in dialects of %s.") :format(full_link({ term = word, lang = lang }, "term"), lang:makeCategoryLink()) end end, }, { regex = "^.+%-translit$", process = "translit", }, { regex = "^Module:form of/lang%-data/(.+)$", process = function(title, cats, lang_code) local lang, langname = insert_lang_data_module_cats(cats, lang_code, "Language-specific form-of modules") if lang then -- FIXME, display more info. return "This module contains language-specific form-of data (tags, shortcuts, base lemma params. etc.) for " .. langname .. "." end end }, { regex = "^Module:labels/data/lang/(.+)$", process = function(title, cats, lang_code) local lang = insert_lang_data_module_cats(cats, lang_code, "Language-specific label data modules") if lang then return { title = "label language-specific data documentation", args = { [1] = lang_code }, } end end }, { regex = "^Module:category tree/lang/(.+)$", process = function(title, cats, lang_code) local lang, langname = insert_lang_data_module_cats(cats, lang_code, "Category tree data modules/lang") if lang then return "This module handles generating the descriptions and categorization for " .. langname .. " category pages " .. "of the format \"" .. langname .. " LABEL\" where LABEL can be any text. Examples are " .. "[[:Category:Bulgarian conjugation 2.1 verbs]] and [[:Category:Russian velar-stem neuter-form nouns]]. " .. "This module is part of the category tree system, which is a general framework for generating the " .. "descriptions and categorization of category pages.\n\n" .. "For more information, see [[Module:category tree/lang/documentation]].\n\n" .. "'''NOTE:''' If you add a new language-specific module, you must add the language code to the " .. "list at the top of [[Module:category tree/lang]] in order for the module to be recognized." end end }, { regex = "^Module:category tree/topic/(.+)$", process = function(title, cats, submodule) cats:insert("Category tree data modules/topic| ") return { title = "topic cat data submodule documentation" } end }, { regex = "^Module:category tree/(.+)$", process = function(title, cats, submodule) cats:insert("Category tree data modules| ") return { title = "category tree data submodule documentation" } end }, { regex = "^Module:ja/data/(.+)$", cat = "Japanese data modules|%1", }, { regex = "^Module:fi%-dialects/data/feature/Kettunen1940 ([0-9]+)$", cat = "Finnish dialectal data atlas modules|%1", process = function(title, cats, shard) return "This module contains shard " .. shard .. " of the online version of Lauri Kettunen's 1940 work " .. "''Suomen murteet III A. Murrekartasto'' (\"Finnish dialects III A: Dialect atlas\"). " .. "It was imported and converted from urn:nbn:fi:csc-kata20151130145346403821, published by the " .. "''Kotimaisten kielten keskus'' under the CC BY 4.0 license." end }, { regex = "^Module:fi%-dialects/data/feature/(.+)", cat = "Finnish dialectal data modules|%1", }, { regex = "^Module:fi%-dialects/data/word/(.+)", cat = "Finnish dialectal data modules|%1", }, { regex = "^Module:Swadesh/data/([%l-]+)$", process = function(title, cats, lang_code) local lang, langname = insert_lang_data_module_cats(cats, lang_code, "Swadesh modules") if lang then return "This module contains the [[Swadesh list]] of basic vocabulary in " .. langname .. "." end end }, { regex = "^Module:Swadesh/data/([%l-]+)/([^/]*)$", process = function(title, cats, lang_code, variety) local lang, langname = insert_lang_data_module_cats(cats, lang_code, "Swadesh modules") if lang then local prefix = "This module contains the [[Swadesh list]] of basic vocabulary in the " local etym_lang = get_lang(variety, nil, "allow etym") if etym_lang then return ("%s %s variety of %s."):format(prefix, etym_lang:getCanonicalName(), langname) end local script = get_script(variety) if script then return ("%s %s %s script."):format(prefix, langname, script:getCanonicalName()) end return ("%s %s variety of %s."):format(prefix, variety, langname) end end }, { regex = "^Module:typing%-aids", process = function(title, cats) local data_suffix = title.fullText:match("^Module:typing%-aids/data/(.+)$") local sortkey if data_suffix then if data_suffix:find "^[%l-]+$" then local lang = get_lang(data_suffix) if lang then sortkey = lang:getCanonicalName() cats:insert(sortkey .. " data modules") end elseif data_suffix:find "^%u%l%l%l$" then local script = get_script(data_suffix) if script then sortkey = script:getCanonicalName() cats:insert(script:getCategoryName()) end end cats:insert("Character insertion data modules|" .. (sortkey or data_suffix)) end end, }, { regex = "^Module:R:([%l-]+):(.+)$", process = function(title, cats, lang_code, refname) local lang = lang_cache[lang_code] if lang then cats:insert(lang:getCanonicalName() .. " modules|" .. refname) cats:insert(("Reference modules|%s"):format(lang:getCanonicalName())) return "This module implements the reference template {{temp|R:" .. lang_code .. ":" .. refname .. "}}." end end, }, { regex = "^Module:Quotations/([%l-]+)/?(.*)", process = "Quotation", }, { regex = "^Module:affix/lang%-data/([%l-]+)", process = "affix lang-data", }, { regex = "^Module:dialect synonyms/([%l-]+)$", process = function(title, cats, lang_code) local lang = lang_cache[lang_code] if lang then local langname = lang:getCanonicalName() cats:insert("Dialect synonyms data modules|" .. langname) cats:insert(langname .. " dialect synonyms data modules| ") return "This module contains data on specific varieties of " .. langname .. ", for use by " .. "{{tl|dialect synonyms}}. The actual synonyms themselves are contained in submodules.\n\n" .. "==== Language data module structure ====\n" .. "* <code>export.title</code> — optional; table title template (e.g. \"Regional synonyms of %s\").\n" .. "* <code>export.columns</code> — optional; list of column headers for location hierarchy (e.g. {\"Dialect group\", \"Dialect\", \"Location\"}).\n" .. "* <code>export.notes</code> — optional; table of note keys to text.\n" .. "* <code>export.sources</code> — optional; table of source keys to text.\n" .. "* <code>export.note_aliases</code> — optional; alias map for notes.\n" .. "* <code>export.varieties</code> — required; nested table of variety nodes. Each node must have <code>name</code>; array part holds children. Node keys can include <code>text_display</code>, <code>color</code>, <code>code</code>, <code>wikidata</code>, <code>lat</code>, <code>long</code>, and language-specific keys (e.g. <code>persian</code>, <code>armenian</code>, <code>chinese</code>).\n\n" .. expand_template({ title = 'dial syn', args = { lang_code, ["demo mode"] = "y" } }) end end, }, { regex = "^Module:dialect synonyms/([%l-]+)/([^/]+)$", process = function(title, cats, lang_code, term) local lang = lang_cache[lang_code] if lang then local langname = lang:getCanonicalName() cats:insert("Dialect synonyms data modules|" .. langname) cats:insert(langname .. " dialect synonyms data modules|" .. term) return ("%s\n\n%s"):format( "==== Term/sense module structure ====\n" .. "* <code>export.title</code> — optional; custom table title (e.g. \"Realization of 'strong R' between vowels\"). Overrides the language default.\n" .. "* <code>export.meaning</code> — optional; meaning/gloss (alternative to <code>gloss</code>).\n" .. "* <code>export.gloss</code> — optional; short meaning for the table.\n" .. "* <code>export.note</code> — optional; single note key or string, or list of note keys.\n" .. "* <code>export.notes</code> — optional; list of note keys.\n" .. "* <code>export.source</code> / <code>export.sources</code> — optional; source keys.\n" .. "* <code>export.last_column</code> — optional; label for the data column (default \"Words\"; e.g. \"Realization\").\n" .. "* <code>export.syns</code> — required; table mapping variety/location names (keys from the language data module) to a list of term entries. Each entry can be a string or a table (e.g. <code>{ ipa = \"[ɽ]\" }</code> or <code>{ term = \"word\" }</code>).\n\n" .. "Example (custom title and data column, IPA realizations):\n" .. "<pre>\nlocal export = {}\n\nexport.title = \"Realization of 'strong R' between vowels\"\n" .. "export.meaning = \"\"\nexport.note = \"realization of 'strong R' between vowels\"\n" .. "export.last_column = \"Realization\"\n\nexport.syns = {\n\t[\"ALERS-158\"] = { { ipa = \"[ɽ]\" } },\n\t[\"ALERS-175\"] = { { ipa = \"[x]\" } },\n}\n\nreturn export\n</pre>\n\n", expand_template({ title = 'dial syn', args = { lang_code, term } })) end end, }, { regex = "^Module:dialect synonyms/([%l-]+)/([^/]+)/([^/]+)$", process = function(title, cats, lang_code, term, id) local lang = lang_cache[lang_code] if lang then local langname = lang:getCanonicalName() cats:insert("Dialect synonyms data modules|" .. langname) cats:insert(langname .. " dialect synonyms data modules|" .. term) return ("%s\n\n%s"):format( "==== Term/sense module structure ====\n" .. "* <code>export.title</code> — optional; custom table title (e.g. \"Realization of 'strong R' between vowels\"). Overrides the language default.\n" .. "* <code>export.meaning</code> — optional; meaning/gloss (alternative to <code>gloss</code>).\n" .. "* <code>export.gloss</code> — optional; short meaning for the table.\n" .. "* <code>export.note</code> — optional; single note key or string, or list of note keys.\n" .. "* <code>export.notes</code> — optional; list of note keys.\n" .. "* <code>export.source</code> / <code>export.sources</code> — optional; source keys.\n" .. "* <code>export.last_column</code> — optional; label for the data column (default \"Words\"; e.g. \"Realization\").\n" .. "* <code>export.syns</code> — required; table mapping variety/location names (keys from the language data module) to a list of term entries. Each entry can be a string or a table (e.g. <code>{ ipa = \"[ɽ]\" }</code> or <code>{ term = \"word\" }</code>).\n\n" .. "Example (custom title and data column, IPA realizations):\n" .. "<pre>\nlocal export = {}\n\nexport.title = \"Realization of 'strong R' between vowels\"\n" .. "export.meaning = \"\"\nexport.note = \"realization of 'strong R' between vowels\"\n" .. "export.last_column = \"Realization\"\n\nexport.syns = {\n\t[\"ALERS-158\"] = { { ipa = \"[ɽ]\" } },\n\t[\"ALERS-175\"] = { { ipa = \"[x]\" } },\n}\n\nreturn export\n</pre>\n\n", expand_template({ title = 'dial syn', args = { lang_code, term, id = id } })) end end, }, { regex = "^Module:bibliography/data/([%l-]+)$", process = function(title, cats, lang_code) if lang_code == "preload" then return 'Used as a base model for other languages when the button "create new language submodule" is clicked.' end local page = require(title.fullText).bib_page if not page then page = lang_cache[lang_code]:getCanonicalName() if page then cats:insert(page .. " modules") end end cats:insert("Reference modules") return "This module holds bibliographical data for " .. page .. ". For the formatted bibliography see '''[[Appendix:Bibliography/" .. page .. "]]'''." end, }, } function export.show(frame) local boolean_default_false = { type = "boolean", default = false } local args = process_params(frame.args, { ["hr"] = true, ["for"] = true, ["from"] = true, ["allowondoc"] = boolean_default_false, -- Don't throw an error if used on a documentation subpage. ["notsubpage"] = boolean_default_false, ["nodoc"] = boolean_default_false, ["nolinks"] = boolean_default_false, -- suppress all "Useful links" ["nosandbox"] = boolean_default_false, -- supress sandbox }) local output = Array('\n<div class="documentation" style="display:block; clear:both">\n') local cats = Array() local nodoc = args.nodoc if (not args.hr) or (args.hr == "above") then output:insert("----\n") end local title = args["for"] and new_title(args["for"]) or get_current_title() local doc_title = args.from ~= "-" and new_title(args.from or title.fullText .. '/documentation') or nil local contentModel = title.contentModel local pagetype, is_script_or_stylesheet = get_pagetype(title) local preload, fallback_docs, doc_content, old_doc_title, user_name, skin_name, needs_doc local doc_content_source = "Module:documentation" local auto_generated_cat_source local cats_auto_generated = false if not args.allowondoc and is_documentation(title) then -- TODO: merge with {{documentation subpage}}, and choose behaviour based on the page type. error("This template should not be used on a documentation page. Please use [[Template:documentation subpage]].") elseif is_sandbox(title) then local sandbox_ns = title.nsText preload = ("Template:documentation/preload%s%sSandbox"):format( sandbox_ns == "Module" and sandbox_ns or "Template", title.rootText:match("^[Uu]ser:(.+)") and "User" or "" ) elseif pagetype:match("%f[%w]gadget%f[%W]") then preload = "Template:documentation/preloadGadget" elseif pagetype:match("%f[%w]script%f[%W]") then -- .js if title.nsText == "MediaWiki" then preload = "Template:documentation/preloadMediaWikiJavaScript" else preload = "Template:documentation/preloadTemplate" -- XXX if title.nsText == "User" then user_name = title.rootText end end is_script_or_stylesheet = true elseif pagetype:match("%f[%w]stylesheet%f[%W]") then -- .css preload = "Template:documentation/preloadTemplate" -- XXX if title.nsText == "User" then user_name = title.rootText end is_script_or_stylesheet = true elseif contentModel == "Scribunto" then -- Exclude pages in Module: which aren't Scribunto. preload = "Template:documentation/preloadModule" elseif pagetype:match("%f[%w]template%f[%W]") or pagetype:match("%f[%w]project%f[%W]") then preload = "Template:documentation/preloadTemplate" end if doc_title and doc_title.isRedirect then old_doc_title = doc_title doc_title = doc_title.redirectTarget end output:insert("<dl class=\"plainlinks\" style=\"font-size: smaller;\">") local function get_module_doc_and_cats(categories_only) cats_auto_generated = true local automatic_cats = nil if user_name then fallback_docs = "documentation/fallback/user module" automatic_cats = { "User sandbox modules" } else for _, data in ipairs(module_regex) do local captures = { umatch(title.fullText, data.regex) } if #captures > 0 then local cat, process_function if is_callable(data.process) then process_function = data.process elseif type(data.process) == "string" then doc_content_source = "Module:documentation/functions/" .. data.process process_function = require(doc_content_source) end if process_function then doc_content = process_function(title, cats, unpack(captures)) end if type(doc_content) == "table" then doc_content_source = doc_content.title and "Template:" .. doc_content.title or doc_content_source doc_content = expand_template(doc_content) elseif doc_content ~= nil then doc_content = preprocess(doc_content) end cat = data.cat if cat then if type(cat) == "string" then cat = { cat } end for _, c in ipairs(cat) do insert(cats, (ugsub(title.fullText, data.regex, c))) end end break end end end if title.subpageText == "templates" then cats:insert("Template interface modules") end if automatic_cats then for _, c in ipairs(automatic_cats) do cats:insert(c) end end if #cats == 0 then local auto_cats = categorize_module(frame, "return raw", "noerror") if #auto_cats > 0 then auto_generated_cat_source = "Module:module categorization" end for _, category in ipairs(auto_cats) do cats:insert(category) end end -- meaning module is not in user’s sandbox or one of many datamodule boring series needs_doc = not categories_only and not (automatic_cats or doc_content or fallback_docs) end -- Override automatic documentation, if present. if doc_title and doc_title.exists then local cats_auto_generated_text = "" if contentModel == "Scribunto" then local doc_page_content = doc_title.content -- Track then do nothing if there are uses of includeonly. The -- pattern is slightly too permissive, but any false-positives are -- obvious typos that should be corrected. if doc_page_content:lower():match("</?includeonly%f[%s/>][^>]*>") then track("module-includeonly") else -- Check for uses of {{module cat}}. find_templates treats the -- input as transcluded by default (i.e. it parses the wikitext -- which will be transcluded through to the module page). local module_cat for template in find_templates(doc_page_content) do if template:get_name() == "module cat" then module_cat = true break end end if not module_cat then get_module_doc_and_cats("categories only") auto_generated_cat_source = auto_generated_cat_source or doc_content_source cats_auto_generated_text = " Categories were auto-generated by [[" .. auto_generated_cat_source .. "]]. <sup>[[" .. new_title(auto_generated_cat_source):fullUrl { action = "edit" } .. " edit]]</sup>" end end end output:insert( "<dd><i style=\"font-size: larger;\">The following " .. "[[Help:Documenting templates and modules|documentation]] is located at [[" .. doc_title.fullText .. "]]. " .. "<sup>[[" .. doc_title:fullUrl { action = "edit" } .. " edit]]</sup>" .. cats_auto_generated_text .. "</i></dd>") else if contentModel == "Scribunto" then get_module_doc_and_cats(false) elseif title.nsText == "Template" then --cats:insert("Uncategorized templates") needs_doc = not (fallback_docs or nodoc) elseif user_name and is_script_or_stylesheet then skin_name = skins[title.text:sub(#title.rootText + 1):match("^/(%l+)%.[jc]ss?$")] if skin_name then fallback_docs = "documentation/fallback/user " .. contentModel end end if doc_content then output:insert( "<dd><i style=\"font-size: larger;\">The following " .. "[[Help:Documenting templates and modules|documentation]] is " .. "generated by [[" .. doc_content_source .. "]]. <sup>[[" .. new_title(doc_content_source):fullUrl { action = "edit" } .. " edit]]</sup> </i></dd>") elseif not nodoc then if doc_title then output:insert( "<dd><i style=\"font-size: larger;\">This " .. pagetype .. " lacks a [[Help:Documenting templates and modules|documentation subpage]]. " .. (fallback_docs and "You may " or "Please ") .. "[" .. doc_title:fullUrl { action = "edit", preload = preload } .. " create it].</i></dd>\n") else output:insert( "<dd><i style=\"font-size: larger; color: var(--wikt-palette-red-9,#FF0000);\">Unable to auto-generate " .. "documentation for this " .. pagetype .. ".</i></dd>\n") end end end if startswith(title.fullText, "MediaWiki:Gadget-") then local is_gadget = false for line in gline(new_title("MediaWiki:Gadgets-definition").content) do local gadget, items = line:match("^%*%s*(%a[%w_-]*)%[.-%]|(.+)$") if not gadget then gadget, items = line:match("^%*%s*(%a[%w_-]*)|(.+)$") end if gadget then items = Array(split(items, "|")) for i, item in ipairs(items) do if title.fullText == ("MediaWiki:Gadget-" .. item) then is_gadget = true output:insert("<dd> ''This script is a part of the <code>") output:insert(gadget) output:insert("</code> gadget ([") output:insert(tostring(full_url("MediaWiki:Gadgets-definition", { action = "edit" }))) output:insert(" edit definitions])'' <dl>") output:insert("<dd> ''Description ([") output:insert(tostring(full_url("MediaWiki:Gadget-" .. gadget, { action = "edit" }))) output:insert(" edit])'': ") output:insert(preprocess(new_message('Gadget-' .. gadget):plain())) output:insert(" </dd>") items:remove(i) if #items > 0 then for j, item in ipairs(items) do items[j] = '[[MediaWiki:Gadget-' .. item .. '|' .. item .. ']]' end output:insert("<dd> ''Other parts'': ") output:insert(list_to_text(items)) output:insert("</dd>") end output:insert("</dl></dd>") break end end end end if not is_gadget then output:insert("<dd> ''This script is not a part of any [") output:insert(tostring(full_url("Special:Gadgets", { uselang = "en" }))) output:insert(' gadget] ([') output:insert(tostring(full_url("MediaWiki:Gadgets-definition", { action = "edit" }))) output:insert(' edit definitions]).</dd>') -- else -- cats:insert("Wiktionary gadgets") end end if old_doc_title then output:insert("<dd> ''Redirected from'' [") output:insert(old_doc_title:fullUrl { redirect = "no" }) output:insert(" ") output:insert(old_doc_title.fullText) output:insert("] ([") output:insert(old_doc_title:fullUrl { action = "edit" }) output:insert(" edit]).</dd>\n") end if not args.nolinks then local links = Array() if title.isSubpage and not args.notsubpage then links:insert("[[:" .. title.nsText .. ":" .. title.rootText .. "|root page]]") links:insert("[[Special:PrefixIndex/" .. title.nsText .. ":" .. title.rootText .. "/|root page’s subpages]]") else links:insert("[[Special:PrefixIndex/" .. title.fullText .. "/|subpage list]]") end links:insert( "[" .. tostring(full_url("Special:WhatLinksHere/" .. title.fullText, { hidetrans = true, hideredirs = true })) .. " links]") if contentModel ~= "Scribunto" then links:insert( "[" .. tostring(full_url("Special:WhatLinksHere/" .. title.fullText, { hidelinks = true, hidetrans = true })) .. " redirects]") end if is_script_or_stylesheet then if user_name then links:insert("[[Special:MyPage" .. title.text:sub(#title.rootText + 1) .. "|your own]]") end else links:insert( "[" .. tostring(full_url("Special:WhatLinksHere/" .. title.fullText, { hidelinks = true, hideredirs = true })) .. " transclusions]") end if contentModel == "Scribunto" then local is_testcases = title.isSubpage and title.subpageText == "testcases" local without_subpage = title.nsText .. ":" .. title.baseText if is_testcases then links:insert("[[:" .. without_subpage .. "|tested module]]") else links:insert("[[" .. title.fullText .. "/testcases|testcases]]") end if user_name then links:insert("[[User:" .. user_name .. "|user page]]") links:insert("[[User talk:" .. user_name .. "|user talk page]]") links:insert("[[Special:PrefixIndex/User:" .. user_name .. "/|userspace]]") -- If sandbox module, add a link to the module that this is a sandbox of. -- Exclude user sandbox modules like [[User:Dine2016/sandbox]]. elseif title.text:find("^sandbox%d*/") or title.text:find("/sandbox%d*%f[/%z]") then cats:insert("Sandbox modules") -- Sandbox modules don’t really need documentation. needs_doc = false -- Don't track user sandbox modules. local text_title = new_title(title.text) if not (text_title and text_title.nsText == "User") then local diff local sandbox_of = title.text:match("^(.*)/sandbox%d*%f[/%z]") if sandbox_of then track("sandbox to be moved") else sandbox_of = title.text:match("^sandbox%d*/(.*)$") end if not sandbox_of then error(("Internal error: Something wrong, couldn't extract sandbox-of module from title '%s'") :format(title.text)) end sandbox_of = title.nsText .. ":" .. sandbox_of if title_exists(sandbox_of) then diff = " (" .. compare_pages(title.fullText, sandbox_of, "diff") .. ")" else track("no sandbox of") end links:insert("[[:" .. sandbox_of .. "|sandbox of]]" .. (diff or "")) end -- If not a sandbox module, add link to sandbox module. -- Sometimes there are multiple sandboxes for a single module: -- [[Module:sandbox/sa-pronunc]], [[Module:sandbox2/sa-pronunc]]. else local sandbox_title local user_prefix, user_rest = title.text:match("^(User:.-/)(.*)$") if not user_prefix then user_prefix = "" user_rest = title.text end sandbox_title = title.nsText .. ":" .. user_prefix .. "sandbox/" .. user_rest local sandbox_link = "[[:" .. sandbox_title .. "|sandbox]]" local diff if title_exists(sandbox_title) then diff = " (" .. compare_pages(title.fullText, sandbox_title, "diff") .. ")" end links:insert(sandbox_link .. (diff or "")) end end if title.nsText == "Template" then -- Error search: all(any namespace), hastemplate (show pages using the template), insource (show source code), incategory (any/specific error) -- [[mw:Help:CirrusSearch]], [[w:Help:Searching/Regex]] -- apparently same with/without: &profile=advanced&fulltext=1 local errorq = 'searchengineselect=mediawiki&search=all: hastemplate:\"' .. title.rootText .. '\" insource:\"' .. title.rootText .. '\" incategory:' local eincategory = "Pages_with_module_errors|ParserFunction_errors|DisplayTitle_errors|Pages_with_ISBN_errors|Pages_with_ISSN_errors|Pages_with_reference_errors|Pages_with_syntax_highlighting_errors|Pages_with_TemplateStyles_errors" links:insert( '[' .. tostring(full_url('Special:Search', errorq .. eincategory)) .. ' errors]' .. ' (' .. '[' .. tostring(full_url('Special:Search', errorq .. 'ParserFunction_errors')) .. ' parser]' .. '/' .. '[' .. tostring(full_url('Special:Search', errorq .. 'Pages_with_module_errors')) .. ' module]' .. ')' ) if title.isSubpage and title.text:find("/sandbox%d*%f[/%z]") then -- This is a sandbox template. -- At the moment there are no user sandbox templates with subpage -- “/sandbox”. cats:insert("Sandbox templates") -- Sandbox templates don’t really need documentation. needs_doc = false -- Will behave badly if “/sandbox” occurs twice in title! local sandbox_of = title.fullText:gsub("/sandbox%d*%f[/%z]", "") local diff if title_exists(sandbox_of) then diff = " (" .. compare_pages(title.fullText, sandbox_of, "diff") .. ")" else track("no sandbox of") end links:insert("[[:" .. sandbox_of .. "|sandbox of]]" .. (diff or "")) -- This is a template that can have a sandbox. elseif not args.nosandbox then -- unless we tell it not to local sandbox_title = title.fullText .. "/sandbox" local diff if title_exists(sandbox_title) then diff = " (" .. compare_pages(title.fullText, sandbox_title, "diff") .. ")" end links:insert("[[:" .. sandbox_title .. "|sandbox]]" .. (diff or "")) end end if #links > 0 then output:insert("<dd> ''Useful links'': " .. links:concat(" • ") .. "</dd>") end end output:insert("</dl>\n") -- Show error from [[Module:category tree/topic cat/data]] on its submodules' -- documentation to, for instance, warn about duplicate labels. if startswith(title.fullText, "Module:category tree/topic/") then local ok, err = pcall(require, "Module:category tree/topic/data") if not ok then output:insert('<span class="error">' .. err .. '</span>\n\n') end end if doc_title and doc_title.exists then -- Override automatic documentation, if present. doc_content = expand_template { title = doc_title.fullText } elseif not doc_content and fallback_docs then doc_content = expand_template { title = fallback_docs, args = { ['user'] = user_name, ['page'] = title.fullText, ['skin name'] = skin_name, }, } end if doc_content then output:insert(doc_content) end output:insert(('\n<%s style="clear: both;" />'):format(args.hr == "below" and "hr" or "br")) if cats_auto_generated and not cats[1] and (not doc_content or not doc_content:find("%[%[Category:")) then if contentModel == "Scribunto" then cats:insert("Uncategorized modules") -- elseif title.nsText == "Template" then -- cats:insert("Uncategorized templates") end end if needs_doc then cats:insert("Templates and modules needing documentation") end for _, cat in ipairs(cats) do output:insert("[[Category:" .. cat .. "]]") end output:insert("</div>\n") return output:concat() end function export.module_auto_doc_table() local parts = {} local function ins(text) insert(parts, text) end ins('{|class="wikitable"') ins("! Regex !! Category !! Handling modules") for _, spec in ipairs(module_regex) do local cat_text local cats = spec.cat if cats then local cat_parts = {} if type(cats) == "string" then cats = { cats } end for _, cat in ipairs(cats) do insert(cat_parts, ("<code>%s</code>"):format((cat:gsub("|", "&#124;")))) end cat_text = concat(cat_parts, ", ") else cat_text = "''(unspecified)''" end ins("|-") ins(("| <code>%s</code> || %s || %s"):format(spec.regex, cat_text, is_callable(spec.process) and "''(handled internally)''" or type(spec.process) == "string" and ("[[Module:documentation/functions/%s]]"):format(spec.process) or "''(no documentation generator)''")) end ins("|}") return concat(parts, "\n") end -- Used by {{translit module documentation}}. function export.translitModuleLangList(frame) local pagename, subpage if frame.args[1] then pagename = frame.args[1] else local title = get_current_title() subpage = title.subpageText pagename = title.text if subpage ~= pagename then pagename = title.rootText end end local translitModule = pagename local languageObjects = require("Module:languages/byTranslitModule")(translitModule) local codeInPagename = pagename:match("^([%l-]+)%-.*translit$") local categories = Array() local codeInPagenameInList = false if codeInPagename then if languageObjects[1] and subpage ~= "documentation" then local agreement = languageObjects[2] and "s" or "" categories:insert("[[Category:Transliteration modules used by " .. #languageObjects .. " language" .. agreement .. "]]") end languageObjects = Array(languageObjects) :filter( function(lang) local result = lang:getCode() ~= codeInPagename codeInPagenameInList = codeInPagenameInList or result return result end) end if subpage ~= "documentation" then for script_code in pagename:gmatch("%f[^-%z]%u%l%l%l%f[-]") do local script = get_script(script_code) if script then categories:insert("[[Category:" .. script:getCategoryName() .. "]]") end end end if subpage ~= "documentation" and not title_exists("Module:" .. pagename .. "/testcases") then categories:insert("[[Category:Transliteration modules without a testcases subpage]]") end if not languageObjects[1] then return categories:concat() end local langs = Array(languageObjects) :sort( function(lang1, lang2) return lang1:getCode() < lang2:getCode() end) -- This will not error because languageObjects is not empty. :map(languageObjects[1].makeCategoryLink) :serialCommaJoin() return "It is " .. (codeInPagenameInList and "also" or "") .. " used to transliterate " .. langs .. "." .. categories:concat() end -- Used by {{strip diacritics module documentation}}. function export.stripDiacriticsModuleLangList(frame) local pagename, subpage if frame.args[1] then pagename = frame.args[1] else local title = get_current_title() subpage = title.subpageText pagename = title.text if subpage ~= pagename then pagename = title.rootText end end local stripDiacriticsModule = pagename local languageObjects = require("Module:languages/byStripDiacriticsModule")(stripDiacriticsModule) local codeInPagename = pagename:match("^([%l-]+)%-.*stripdiacritics$") local categories = Array() local codeInPagenameInList = false if codeInPagename then if languageObjects[1] and subpage ~= "documentation" then local agreement = languageObjects[2] and "s" or "" categories:insert("[[Category:Diacritic-stripping modules used by " .. #languageObjects .. " language" .. agreement .. "]]") end languageObjects = Array(languageObjects) :filter( function(lang) local result = lang:getCode() ~= codeInPagename codeInPagenameInList = codeInPagenameInList or result return result end) end if subpage ~= "documentation" then for script_code in pagename:gmatch("%f[^-%z]%u%l%l%l%f[-]") do local script = get_script(script_code) if script then categories:insert("[[Category:" .. script:getCategoryName() .. "]]") end end end if subpage ~= "documentation" and not title_exists("Module:" .. pagename .. "/testcases") then categories:insert("[[Category:Diacritic-stripping modules without a testcases subpage]]") end if not languageObjects[1] then return categories:concat() end local langs = Array(languageObjects) :sort( function(lang1, lang2) return lang1:getCode() < lang2:getCode() end) -- This will not error because languageObjects is not empty. :map(languageObjects[1].makeCategoryLink) :serialCommaJoin() return "It is " .. (codeInPagenameInList and "also" or "") .. " used to strip diacritics for " .. langs .. "." .. categories:concat() end -- Used by {{sortkey module documentation}}. function export.sortkeyModuleLangList(frame) local pagename, subpage if frame.args[1] then pagename = frame.args[1] else local title = get_current_title() subpage = title.subpageText pagename = title.text if subpage ~= pagename then pagename = title.rootText end end local sortkeyModule = pagename local languageObjects = require("Module:languages/bySortkeyModule")(sortkeyModule) local codeInPagename = pagename:match("^([%l-]+)%-.*sortkey$") local categories = Array() local codeInPagenameInList = false if codeInPagename then if languageObjects[1] and subpage ~= "documentation" then local agreement = languageObjects[2] and "s" or "" categories:insert("[[Category:Sortkey-generating modules used by " .. #languageObjects .. " language" .. agreement .. "]]") end languageObjects = Array(languageObjects) :filter( function(lang) local result = lang:getCode() ~= codeInPagename codeInPagenameInList = codeInPagenameInList or result return result end) end if subpage ~= "documentation" then for script_code in pagename:gmatch("%f[^-%z]%u%l%l%l%f[-]") do local script = get_script(script_code) if script then categories:insert("[[Category:" .. script:getCategoryName() .. "]]") end end end if subpage ~= "documentation" and not title_exists("Module:" .. pagename .. "/testcases") then categories:insert("[[Category:Sortkey-generating modules without a testcases subpage]]") end if not languageObjects[1] then return categories:concat() end local langs = Array(languageObjects) :sort( function(lang1, lang2) return lang1:getCode() < lang2:getCode() end) -- This will not error because languageObjects is not empty. :map(languageObjects[1].makeCategoryLink) :serialCommaJoin() return "It is " .. (codeInPagenameInList and "also" or "") .. " used to sort " .. langs .. "." .. categories:concat() end return export gs0i8pbsg53nl0m4qithir7evspynu2 828 39855 123878 122016 2026-05-05T06:45:51Z Hiyuune 3448 123878 Scribunto text/plain --[[ NOTE: This module works by using recursive backtracking to build a node tree, which can then be traversed as necessary. Because it is called by a number of high-use modules, it has been optimised for speed using a profiler, since it is used to scrape data from large numbers of pages very quickly. To that end, it rolls some of its own methods in cases where this is faster than using a function from one of the standard libraries. Please DO NOT "simplify" the code by removing these, since you are almost guaranteed to slow things down, which could seriously impact performance on pages which call this module hundreds or thousands of times. It has also been designed to emulate the native parser's behaviour as much as possible, which in some cases means replicating bugs or unintuitive behaviours in that code; these should not be "fixed", since it is important that the outputs are the same. Most of these originate from deficient regular expressions, which can't be used here, so the bugs have to be manually reintroduced as special cases (e.g. onlyinclude tags being case-sensitive and whitespace intolerant, unlike all other tags). If any of these are fixed, this module should also be updated accordingly. ]] local export = {} local data_module = "Module:template parser/data" local load_module = "Module:load" local magic_words_data_module = "Module:data/magic words" local pages_module = "Module:pages" local parser_extension_tags_data_module = "Module:data/parser extension tags" local parser_module = "Module:parser" local scribunto_module = "Module:Scribunto" local string_pattern_escape_module = "Module:string/patternEscape" local string_replacement_escape_module = "Module:string/replacementEscape" local string_utilities_module = "Module:string utilities" local table_length_module = "Module:table/length" local table_shallow_copy_module = "Module:table/shallowCopy" local table_sorted_pairs_module = "Module:table/sortedPairs" local title_is_title_module = "Module:title/isTitle" local title_make_title_module = "Module:title/makeTitle" local title_new_title_module = "Module:title/newTitle" local title_redirect_target_module = "Module:title/redirectTarget" local require = require local m_parser = require(parser_module) local mw = mw local mw_title = mw.title local mw_uri = mw.uri local string = string local table = table local anchor_encode = mw_uri.anchorEncode local build_template -- defined as export.buildTemplate below local class_else_type = m_parser.class_else_type local concat = table.concat local encode_uri = mw_uri.encode local find = string.find local format = string.format local gsub = string.gsub local html_create = mw.html.create local insert = table.insert local is_node = m_parser.is_node local lower = string.lower local match = string.match local next = next local pairs = pairs local parse -- defined as export.parse below local parse_template_name -- defined below local pcall = pcall local rep = string.rep local select = select local sub = string.sub local title_equals = mw_title.equals local tostring = m_parser.tostring local type = type local umatch = mw.ustring.match --[==[ Loaders for functions in other modules, which overwrite themselves with the target function when called. This ensures modules are only loaded when needed, retains the speed/convenience of locally-declared pre-loaded functions, and has no overhead after the first call, since the target functions are called directly in any subsequent calls.]==] local function decode_entities(...) decode_entities = require(string_utilities_module).decode_entities return decode_entities(...) end local function encode_entities(...) encode_entities = require(string_utilities_module).encode_entities return encode_entities(...) end local function get_link_target(...) get_link_target = require(pages_module).get_link_target return get_link_target(...) end local function is_title(...) is_title = require(title_is_title_module) return is_title(...) end local function load_data(...) load_data = require(load_module).load_data return load_data(...) end local function make_title(...) make_title = require(title_make_title_module) return make_title(...) end local function new_title(...) new_title = require(title_new_title_module) return new_title(...) end local function pattern_escape(...) pattern_escape = require(string_pattern_escape_module) return pattern_escape(...) end local function php_htmlspecialchars(...) php_htmlspecialchars = require(scribunto_module).php_htmlspecialchars return php_htmlspecialchars(...) end local function php_ltrim(...) php_ltrim = require(scribunto_module).php_ltrim return php_ltrim(...) end local function php_trim(...) php_trim = require(scribunto_module).php_trim return php_trim(...) end local function redirect_target(...) redirect_target = require(title_redirect_target_module) return redirect_target(...) end local function replacement_escape(...) replacement_escape = require(string_replacement_escape_module) return replacement_escape(...) end local function scribunto_parameter_key(...) scribunto_parameter_key = require(scribunto_module).scribunto_parameter_key return scribunto_parameter_key(...) end local function shallow_copy(...) shallow_copy = require(table_shallow_copy_module) return shallow_copy(...) end local function sorted_pairs(...) sorted_pairs = require(table_sorted_pairs_module) return sorted_pairs(...) end local function split(...) split = require(string_utilities_module).split return split(...) end local function table_len(...) table_len = require(table_length_module) return table_len(...) end local function uupper(...) uupper = require(string_utilities_module).upper return uupper(...) end --[==[ Loaders for objects, which load data (or some other object) into some variable, which can then be accessed as "foo or get_foo()", where the function get_foo sets the object to "foo" and then returns it. This ensures they are only loaded when needed, and avoids the need to check for the existence of the object each time, since once "foo" has been set, "get_foo" will not be called again.]==] local data local function get_data() data, get_data = load_data(data_module), nil return data end local frame local function get_frame() frame, get_frame = mw.getCurrentFrame(), nil return frame end local magic_words local function get_magic_words() magic_words, get_magic_words = load_data(magic_words_data_module), nil return magic_words end local parser_extension_tags local function get_parser_extension_tags() parser_extension_tags, get_parser_extension_tags = load_data(parser_extension_tags_data_module), nil return parser_extension_tags end ------------------------------------------------------------------------------------ -- -- Nodes -- ------------------------------------------------------------------------------------ local Node = m_parser.node() local new_node = Node.new local function expand(obj, frame_args) return is_node(obj) and obj:expand(frame_args) or obj end export.expand = expand function Node:expand(frame_args) local output = {} for i = 1, #self do output[i] = expand(self[i], frame_args) end return concat(output) end local Wikitext = Node:new_class("wikitext") -- force_node ensures the output will always be a Wikitext node. function Wikitext:new(this, force_node) if type(this) ~= "table" then return force_node and new_node(self, {this}) or this elseif #this == 1 then local this1 = this[1] return force_node and class_else_type(this1) ~= "wikitext" and new_node(self, this) or this1 end local success, str = pcall(concat, this) if success then return force_node and new_node(self, {str}) or str end return new_node(self, this) end -- First value is the parameter name. -- Second value is the parameter's default value. -- Any additional values are ignored: e.g. "{{{a|b|c}}}" is parameter "a" with default value "b" (*not* "b|c"). local Parameter = Node:new_class("parameter") function Parameter:new(this) local this2 = this[2] if class_else_type(this2) == "argument" then insert(this2, 2, "=") this2 = Wikitext:new(this2) end if this[3] == nil then this[2] = this2 else this = {this[1], this2} end return new_node(self, this) end function Parameter:__tostring() local output = {} for i = 1, #self do output[i] = tostring(self[i]) end return "{{{" .. concat(output, "|") .. "}}}" end function Parameter:get_name(frame_args) return scribunto_parameter_key(expand(self[1], frame_args)) end function Parameter:get_default(frame_args) local default = self[2] if default ~= nil then return expand(default, frame_args) end return "{{{" .. expand(self[1], frame_args) .. "}}}" end function Parameter:expand(frame_args) if frame_args == nil then return self:get_default() end local name = expand(self[1], frame_args) local val = frame_args[scribunto_parameter_key(name)] -- Parameter in use. if val ~= nil then return val end val = self[2] -- Default. if val ~= nil then return expand(val, frame_args) end return "{{{" .. name .. "}}}" end local Argument = Node:new_class("argument") function Argument:new(this) local key = this._parse_data.key this = Wikitext:new(this) if key == nil then return this end return new_node(self, {Wikitext:new(key), this}) end function Argument:__tostring() return tostring(self[1]) .. "=" .. tostring(self[2]) end function Argument:expand(frame_args) return expand(self[1], frame_args) .. "=" .. expand(self[2], frame_args) end local Template = Node:new_class("template") function Template:__tostring() local output = {} for i = 1, #self do output[i] = tostring(self[i]) end return "{{" .. concat(output, "|") .. "}}" end -- Normalize the template name, check it's a valid template, then memoize results (using false for invalid titles). -- Parser functions (e.g. {{#IF:a|b|c}}) need to have the first argument extracted from the title, as it comes after the colon. Because of this, the parser function and first argument are memoized as a table. -- FIXME: Some parser functions have special argument handling (e.g. {{#SWITCH:}}). do local templates, parser_variables, parser_functions = {}, {}, {} local function retrieve_magic_word_data(chunk) local mgw_data = (magic_words or get_magic_words())[chunk] if mgw_data then return mgw_data end local normalized = uupper(chunk) mgw_data = magic_words[normalized] if mgw_data and not mgw_data.case_sensitive then return mgw_data end end -- Returns the name required to transclude the title object `title` using -- template {{ }} syntax. If the `shortcut` flag is set, then any calls -- which require a namespace prefix will use the abbreviated form where one -- exists (e.g. "Template:PAGENAME" becomes "T:PAGENAME"). local function get_template_invocation_name(title, shortcut) if not (is_title(title) and not title.isExternal) then error("Template invocations require a valid page title, which cannot contain an interwiki prefix.") end local namespace = title.namespace -- If not in the template namespace, include the prefix (or ":" if -- mainspace). if namespace ~= 10 then return get_link_target(title, shortcut) end -- If in the template namespace and it shares a name with a magic word, -- it needs the prefix "Template:". local text, fragment = title.text, title.fragment if fragment and fragment ~= "" then text = text .. "#" .. fragment end local colon = find(text, ":", nil, true) if not colon then local mgw_data = retrieve_magic_word_data(text) return mgw_data and mgw_data.parser_variable and get_link_target(title, shortcut) or text end local mgw_data = retrieve_magic_word_data(sub(text, 1, colon - 1)) if mgw_data and (mgw_data.parser_function or mgw_data.transclusion_modifier) then return get_link_target(title, shortcut) end -- Also if "Template:" is necessary for disambiguation (e.g. -- "Template:Category:Foo" can't be called with "Category:Foo"). local check = new_title(text, namespace) return check and title_equals(title, check) and text or get_link_target(title, shortcut) end export.getTemplateInvocationName = get_template_invocation_name function parse_template_name(name, has_args, fragment, force_transclusion) local chunks, colon, start, n, p = {}, find(name, ":", nil, true), 1, 0, 0 while colon do local mgw_data = retrieve_magic_word_data(php_ltrim(sub(name, start, colon - 1))) if not mgw_data then break end local priority = mgw_data.priority if not (priority and priority > p) then local pf = mgw_data.parser_function and mgw_data.name or nil if pf then n = n + 1 chunks[n] = pf .. ":" return chunks, "parser function", sub(name, colon + 1) end break end n = n + 1 chunks[n] = mgw_data.name .. ":" start, p = colon + 1, priority colon = find(name, ":", start, true) end if start > 1 then name = sub(name, start) end name = php_trim(name) -- Parser variables can only take SUBST:/SAFESUBST: as modifiers. if not has_args and p <= 1 then local mgw_data = retrieve_magic_word_data(name) local pv = mgw_data and mgw_data.parser_variable and mgw_data.name or nil if pv then n = n + 1 chunks[n] = pv return chunks, "parser variable" end end -- Get the template title with the custom new_title() function in -- [[Module:title/newTitle]], with `allowOnlyFragment` set to false -- (e.g. "{{#foo}}" is invalid) and `allowRelative` set to true, for -- relative links for namespaces with subpages (e.g. "{{/foo}}"). local title = new_title(name, 10, false, true) if not (title and not title.isExternal) then return nil end if force_transclusion ~= "no_redirect" then -- Resolve any redirects. If the redirect target is an interwiki link, -- the template won't fail, but the redirect does not get resolved (i.e. -- the redirect page itself gets transcluded, so the template name -- should not be normalized to the target). local redirect = redirect_target(title, force_transclusion) if redirect and not redirect.isExternal then title = redirect end end -- If `fragment` is not true, unset it from the title object to prevent -- it from being included by get_template_invocation_name. if not fragment then title.fragment = "" end chunks[n + 1] = get_template_invocation_name(title) return chunks, "template" end -- Note: `force_transclusion` avoids incrementing the expensive parser function count by forcing transclusion -- instead. This should only be used when there is a real risk that the expensive parser function limit of -- 500 will be hit. If `force_transclusion` has the value "no_redirect", redirect checking is turned off -- entirely. This should only be used in very limited circumstances when it is otherwise impossible to avoid -- hitting the expensive parser limit and we are willing to handle the possible redirects ourselves. local function process_name(self, frame_args, force_transclusion) local name = expand(self[1], frame_args) local has_args, norm = #self > 1 if not has_args then norm = parser_variables[name] if norm then return norm, "parser variable" end end norm = templates[name] if norm then local pf_arg1 = parser_functions[name] return norm, pf_arg1 and "parser function" or "template", pf_arg1 elseif norm == false then return nil end local chunks, subclass, pf_arg1 = parse_template_name(name, has_args, nil, force_transclusion) -- Fail if invalid. if not chunks then templates[name] = false return nil end local chunk1 = chunks[1] -- Fail on SUBST:. if chunk1 == "SUBST:" then templates[name] = false return nil -- Any modifiers are ignored. elseif subclass == "parser function" then local pf = chunks[#chunks] templates[name] = pf parser_functions[name] = pf_arg1 return pf, "parser function", pf_arg1 end -- Ignore SAFESUBST:, and treat MSGNW: as a parser function with the pagename as its first argument (ignoring any RAW: that comes after). if chunks[chunk1 == "SAFESUBST:" and 2 or 1] == "MSGNW:" then pf_arg1 = chunks[#chunks] local pf = "MSGNW:" templates[name] = pf parser_functions[name] = pf_arg1 return pf, "parser function", pf_arg1 end -- Ignore any remaining modifiers, as they've done their job. local output = chunks[#chunks] if subclass == "parser variable" then parser_variables[name] = output else templates[name] = output end return output, subclass end function Template:get_name(frame_args, force_transclusion) -- Only return the first return value. return (process_name(self, frame_args, force_transclusion)) end function Template:get_arguments(frame_args) local name, subclass, pf_arg1 = process_name(self, frame_args) if name == nil then return nil elseif subclass == "parser variable" then return {} end local template_args = {} if subclass == "parser function" then template_args[1] = pf_arg1 for i = 2, #self do template_args[i] = expand(self[i], frame_args) -- Not trimmed. end return template_args end local implicit = 0 for i = 2, #self do local arg = self[i] if class_else_type(arg) == "argument" then template_args[scribunto_parameter_key(expand(arg[1], frame_args))] = php_trim((expand(arg[2], frame_args))) else implicit = implicit + 1 template_args[implicit] = expand(arg, frame_args) -- Not trimmed. end end return template_args end -- BIG TODO: manual template expansion. function Template:expand(frame_args) local name, subclass, pf_arg1 = process_name(self, frame_args) if name == nil then local output = {} for i = 1, #self do output[i] = expand(self[i], frame_args) end return "{{" .. concat(output, "|") .. "}}" elseif subclass == "parser variable" then return (frame or get_frame()):preprocess("{{" .. name .. "}}") elseif subclass == "parser function" then local f = frame or get_frame() if frame_args ~= nil then local success, new_f = pcall(f.newChild, f, {args = frame_args}) if success then f = new_f end end return f:preprocess(tostring(self)) end local output = {} for i = 1, #self do output[i] = expand(self[i], frame_args) end return (frame or get_frame()):preprocess("{{" .. concat(output, "|") .. "}}") end end local Tag = Node:new_class("tag") function Tag:__tostring() local open_tag, attributes, n = {"<", self.name}, self:get_attributes(), 2 for attr, value in next, attributes do n = n + 1 open_tag[n] = " " .. php_htmlspecialchars(attr) .. "=\"" .. php_htmlspecialchars(value, "compat") .. "\"" end if self.self_closing then return concat(open_tag) .. "/>" end return concat(open_tag) .. ">" .. concat(self) .. "</" .. self.name .. ">" end do local valid_attribute_name local function get_valid_attribute_name() valid_attribute_name, get_valid_attribute_name = (data or get_data()).valid_attribute_name, nil return valid_attribute_name end function Tag:get_attributes() local raw = self.attributes if not raw then self.attributes = {} return self.attributes elseif type(raw) == "table" then return raw end if sub(raw, -1) == "/" then raw = sub(raw, 1, -2) end local attributes, head = {}, 1 -- Semi-manual implementation of the native regex. while true do local name, loc = match(raw, "([^\t\n\f\r />][^\t\n\f\r /=>]*)()", head) if not name then break end head = loc local value loc = match(raw, "^[\t\n\f\r ]*=[\t\n\f\r ]*()", head) if loc then head = loc -- Either "", '' or the value ends on a space/at the end. Missing -- end quotes are repaired by closing the value at the end. value, loc = match(raw, "^\"([^\"]*)\"?()", head) if not value then value, loc = match(raw, "^'([^']*)'?()", head) if not value then value, loc = match(raw, "^([^\t\n\f\r ]*)()", head) end end head = loc end -- valid_attribute_name is a pattern matching a valid attribute name. -- Defined in the data due to its length - see there for more info. if umatch(name, valid_attribute_name or get_valid_attribute_name()) then -- Sanitizer applies PHP strtolower (ASCII-only). attributes[lower(name)] = value and decode_entities( php_trim((gsub(value, "[\t\n\r ]+", " "))) ) or "" end end self.attributes = attributes return attributes end end function Tag:expand() return (frame or get_frame()):preprocess(tostring(self)) end local Heading = Node:new_class("heading") function Heading:new(this) if #this > 1 then local success, str = pcall(concat, this) if success then return new_node(self, { str, level = this.level, section = this.section, index = this.index }) end end return new_node(self, this) end do local node_tostring = Node.__tostring function Heading:__tostring() local eq = rep("=", self.level) return eq .. node_tostring(self) .. eq end end do local expand_node = Node.expand -- Expanded heading names can contain "\n" (e.g. inside nowiki tags), which -- causes any heading containing them to fail. However, in such cases, the -- native parser still treats it as a heading for the purpose of section -- numbers. local function validate_name(self, frame_args) local name = expand_node(self, frame_args) if find(name, "\n", nil, true) then return nil end return name end function Heading:get_name(frame_args) local name = validate_name(self, frame_args) return name ~= nil and php_trim(name) or nil end -- FIXME: account for anchor disambiguation. function Heading:get_anchor(frame_args) local name = validate_name(self, frame_args) return name ~= nil and decode_entities(anchor_encode(name)) or nil end function Heading:expand(frame_args) local eq = rep("=", self.level) return eq .. expand_node(self, frame_args) .. eq end end ------------------------------------------------------------------------------------ -- -- Parser -- ------------------------------------------------------------------------------------ local Parser = m_parser.string_parser() -- Template or parameter. -- Parsed by matching the opening braces innermost-to-outermost (ignoring lone closing braces). Parameters {{{ }}} take priority over templates {{ }} where possible, but a double closing brace will always result in a closure, even if there are 3+ opening braces. -- For example, "{{{{foo}}}}" (4) is parsed as a parameter enclosed by single braces, and "{{{{{foo}}}}}" (5) is a parameter inside a template. However, "{{{{{foo }} }}}" is a template inside a parameter, due to "}}" forcing the closure of the inner node. do -- Handlers. local handle_name local handle_argument local handle_value local function do_template_or_parameter(self, inner_node) self:push_sublayer(handle_name) self:set_pattern("[\n<[{|}]") -- If a node has already been parsed, nest it at the start of the new -- outer node (e.g. when parsing"{{{{foo}}bar}}", the template "{{foo}}" -- is parsed first, since it's the innermost, and becomes the first -- node of the outer template. if inner_node then self:emit(inner_node) end end local function pipe(self) self:emit(Wikitext:new(self:pop_sublayer())) self:push_sublayer(handle_argument) self:set_pattern("[\n<=[{|}]") end local function rbrace(self, this) if self:read(1) == "}" then self:emit(Wikitext:new(self:pop_sublayer())) return self:pop() end self:emit(this) end function handle_name(self, ...) handle_name = self:switch(handle_name, { ["\n"] = Parser.heading_block, ["<"] = Parser.tag, ["["] = Parser.wikilink_block, ["{"] = Parser.braces, ["|"] = pipe, ["}"] = rbrace, [""] = Parser.fail_route, [false] = Parser.emit }) return handle_name(self, ...) end function handle_argument(self, ...) handle_argument = self:switch(handle_argument, { ["\n"] = function(self, this) return self:heading_block(this, "==") end, ["<"] = Parser.tag, ["="] = function(self) local key = self:pop_sublayer() self:push_sublayer(handle_value) self:set_pattern("[\n<[{|}]") self.current_layer._parse_data.key = key end, ["["] = Parser.wikilink_block, ["{"] = Parser.braces, ["|"] = pipe, ["}"] = rbrace, [""] = Parser.fail_route, [false] = Parser.emit }) return handle_argument(self, ...) end function handle_value(self, ...) handle_value = self:switch(handle_value, { ["\n"] = Parser.heading_block, ["<"] = Parser.tag, ["["] = Parser.wikilink_block, ["{"] = Parser.braces, ["|"] = function(self) self:emit(Argument:new(self:pop_sublayer())) self:push_sublayer(handle_argument) self:set_pattern("[\n<=[{|}]") end, ["}"] = function(self, this) if self:read(1) == "}" then self:emit(Argument:new(self:pop_sublayer())) return self:pop() end self:emit(this) end, [""] = Parser.fail_route, [false] = Parser.emit }) return handle_value(self, ...) end function Parser:template_or_parameter() local text, head, node_to_emit, failed = self.text, self.head -- Comments/tags interrupt the brace count. local braces = match(text, "^{+()", head) - head self:advance(braces) while true do local success, node = self:try(do_template_or_parameter, node_to_emit) -- Fail means no "}}" or "}}}" was found, so emit any remaining -- unmatched opening braces before any templates/parameters that -- were found. if not success then self:emit(rep("{", braces)) failed = true break -- If there are 3+ opening and closing braces, it's a parameter. elseif braces >= 3 and self:read(2) == "}" then self:advance(3) braces = braces - 3 node = Parameter:new(node) -- Otherwise, it's a template. else self:advance(2) braces = braces - 2 node = Template:new(node) end local index = head + braces node.index = index node.raw = sub(text, index, self.head - 1) node_to_emit = node -- Terminate once not enough braces remain for further matches. if braces == 0 then break -- Emit any stray opening brace before any matched nodes. elseif braces == 1 then self:emit("{") break end end if node_to_emit then self:emit(node_to_emit) end return braces, failed end end -- Tag. do local end_tags local function get_end_tags() end_tags, get_end_tags = (data or get_data()).end_tags, nil return end_tags end -- Handlers. local handle_start local handle_tag local function do_tag(self) local layer = self.current_layer layer._parse_data.handler, layer.index = handle_start, self.head self:set_pattern("[%s/>]") self:advance() end local function is_ignored_tag(self, this) if self.transcluded then return this == "includeonly" end return this == "noinclude" or this == "onlyinclude" end local function ignored_tag(self, text, head) local loc = find(text, ">", head, true) if not loc then return self:fail_route() end self:jump(loc) local tag = self:pop() tag.ignored = true return tag end function handle_start(self, this) if this == "/" then local text, head = self.text, self.head + 1 local this = match(text, "^[^%s/>]+", head) if this and is_ignored_tag(self, lower(this)) then head = head + #this if not match(text, "^/[^>]", head) then return ignored_tag(self, text, head) end end return self:fail_route() elseif this == "" then return self:fail_route() end -- Tags are only case-insensitive with ASCII characters. local raw_name = this this = lower(this) local end_tag_pattern = (end_tags or get_end_tags())[this] if not end_tag_pattern then -- Validity check. return self:fail_route() end local layer = self.current_layer local pdata = layer._parse_data local text, head = self.text, self.head + pdata.step if match(text, "^/[^>]", head) then return self:fail_route() elseif is_ignored_tag(self, this) then return ignored_tag(self, text, head) -- If an onlyinclude tag is not ignored (and cannot be active since it -- would have triggered special handling earlier), it must be plaintext. elseif this == "onlyinclude" then return self:fail_route() elseif this == "noinclude" or this == "includeonly" then layer.ignored = true -- Ignored block. layer.raw_name = raw_name end layer.name, pdata.handler, pdata.end_tag_pattern = this, handle_tag, end_tag_pattern self:set_pattern(">") end function handle_tag(self, this) if this == "" then return self:fail_route() end local layer = self.current_layer if this ~= ">" then layer.attributes = this return elseif self:read(-1) == "/" then layer.self_closing = true return self:pop() end local text, head = self.text, self.head + 1 local loc1, loc2 = find(text, layer._parse_data.end_tag_pattern, head) if loc1 then if loc1 > head then self:emit(sub(text, head, loc1 - 1)) end self:jump(loc2) return self:pop() -- noinclude and includeonly will tolerate having no closing tag, but -- only if given in lowercase. This is due to a preprocessor bug, as -- it uses a regex with the /i (case-insensitive) flag to check for -- end tags, but a simple array lookup with lowercase tag names when -- looking up which tags should tolerate no closing tag (exact match -- only, so case-sensitive). elseif layer.ignored then local raw_name = layer.raw_name if raw_name == "noinclude" or raw_name == "includeonly" then self:jump(#text) return self:pop() end end return self:fail_route() end function Parser:tag() -- HTML comment. if self:read(1, 3) == "!--" then local text = self.text self:jump(select(2, find(text, "-->", self.head + 4, true)) or #text) -- onlyinclude tags (which must be lowercase with no whitespace). elseif self.onlyinclude and self:read(1, 13) == "/onlyinclude>" then local text = self.text self:jump(select(2, find(text, "<onlyinclude>", self.head + 14, true)) or #text) else local success, tag = self:try(do_tag) if not success then self:emit("<") elseif not tag.ignored then self:emit(Tag:new(tag)) end end end end -- Heading. -- The preparser assigns each heading a number, which is used for things like section edit links. The preparser will only do this for heading blocks which aren't nested inside templates, parameters and parser tags. In some cases (e.g. when template blocks contain untrimmed newlines), a preparsed heading may not be treated as a heading in the final output. That does not affect the preparser, however, which will always count sections based on the preparser heading count, since it can't know what a template's final output will be. do -- Handlers. local handle_start local handle_body local handle_possible_end local function do_heading(self) local layer, head = self.current_layer, self.head layer._parse_data.handler, layer.index = handle_start, head self:set_pattern("[\t\n ]") -- Comments/tags interrupt the equals count. local eq = match(self.text, "^=+()", head) - head layer.level = eq self:advance(eq) end local function do_heading_possible_end(self) self.current_layer._parse_data.handler = handle_possible_end self:set_pattern("[\n<]") end function handle_start(self, ...) -- ===== is "=" as an L2; ======== is "==" as an L3 etc. local function newline(self) local layer = self.current_layer local eq = layer.level if eq <= 2 then return self:fail_route() end -- Calculate which equals signs determine the heading level. local level_eq = eq - (2 - eq % 2) level_eq = level_eq > 12 and 12 or level_eq -- Emit the excess. self:emit(rep("=", eq - level_eq)) layer.level = level_eq / 2 return self:pop() end local function whitespace(self) local success, possible_end = self:try(do_heading_possible_end) if success then self:emit(Wikitext:new(possible_end)) self.current_layer._parse_data.handler = handle_body self:set_pattern("[\n<=[{]") return self:consume() end return newline(self) end handle_start = self:switch(handle_start, { ["\t"] = whitespace, ["\n"] = newline, [" "] = whitespace, [""] = newline, [false] = function(self) -- Emit any excess = signs once we know it's a conventional heading. Up till now, we couldn't know if the heading is just a string of = signs (e.g. ========), so it wasn't guaranteed that the heading text starts after the 6th. local layer = self.current_layer local eq = layer.level if eq > 6 then self:emit(1, rep("=", eq - 6)) layer.level = 6 end layer._parse_data.handler = handle_body self:set_pattern("[\n<=[{]") return self:consume() end }) return handle_start(self, ...) end function handle_body(self, ...) handle_body = self:switch(handle_body, { ["\n"] = Parser.fail_route, ["<"] = Parser.tag, ["="] = function(self) -- Comments/tags interrupt the equals count. local eq = match(self.text, "^=+", self.head) local eq_len = #eq self:advance(eq_len) local success, possible_end = self:try(do_heading_possible_end) if success then self:emit(eq) self:emit(Wikitext:new(possible_end)) return self:consume() end local layer = self.current_layer local level = layer.level if eq_len > level then self:emit(rep("=", eq_len - level)) elseif level > eq_len then layer.level = eq_len self:emit(1, rep("=", level - eq_len)) end return self:pop() end, ["["] = Parser.wikilink_block, ["{"] = function(self, this) return self:braces(this, true) end, [""] = Parser.fail_route, [false] = Parser.emit }) return handle_body(self, ...) end function handle_possible_end(self, ...) handle_possible_end = self:switch(handle_possible_end, { ["\n"] = Parser.fail_route, ["<"] = function(self) if self:read(1, 3) ~= "!--" then return self:pop() end local head = select(2, find(self.text, "-->", self.head + 4, true)) if not head then return self:pop() end self:jump(head) end, [""] = Parser.fail_route, [false] = function(self, this) if not match(this, "^[\t ]+()$") then return self:pop() end self:emit(this) end }) return handle_possible_end(self, ...) end function Parser:heading() local success, heading = self:try(do_heading) if success then local section = self.section + 1 heading.section = section self.section = section self:emit(Heading:new(heading)) return self:consume() else self:emit("=") end end end ------------------------------------------------------------------------------------ -- -- Block handlers -- ------------------------------------------------------------------------------------ -- Block handlers. -- These are blocks which can affect template/parameter parsing, since they're also parsed by Parsoid at the same time (even though they aren't processed until later). -- All blocks (including templates/parameters) can nest inside each other, but an inner block must be closed before the outer block which contains it. This is why, for example, the wikitext "{{template| [[ }}" will result in an unprocessed template, since the inner "[[" is treated as the opening of a wikilink block, which prevents "}}" from being treated as the closure of the template block. On the other hand, "{{template| [[ ]] }}" will process correctly, since the wikilink block is closed before the template closure. It makes no difference whether the block will be treated as valid or not when it's processed later on, so "{{template| [[ }} ]] }}" would also work, even though "[[ }} ]]" is not a valid wikilink. -- Note that nesting also affects pipes and equals signs, in addition to block closures. -- These blocks can be nested to any degree, so "{{template| [[ [[ [[ ]] }}" will not work, since only one of the three wikilink blocks has been closed. On the other hand, "{{template| [[ [[ [[ ]] ]] ]] }}" will work. -- All blocks are implicitly closed by the end of the text, since their validity is irrelevant at this stage. -- Language conversion block. -- Opens with "-{" and closes with "}-". However, templates/parameters take priority, so "-{{" is parsed as "-" followed by the opening of a template/parameter block (depending on what comes after). -- Note: Language conversion blocks aren't actually enabled on the English Wiktionary, but Parsoid still parses them at this stage, so they can affect the closure of outer blocks: e.g. "[[ -{ ]]" is not a valid wikilink block, since the "]]" falls inside the new language conversion block. do --Handler. local handle_language_conversion_block local function do_language_conversion_block(self) self.current_layer._parse_data.handler = handle_language_conversion_block self:set_pattern("[\n<[{}]") end function handle_language_conversion_block(self, ...) handle_language_conversion_block = self:switch(handle_language_conversion_block, { ["\n"] = Parser.heading_block, ["<"] = Parser.tag, ["["] = Parser.wikilink_block, ["{"] = Parser.braces, ["}"] = function(self, this) if self:read(1) == "-" then self:emit("}-") self:advance() return self:pop() end self:emit(this) end, [""] = Parser.pop, [false] = Parser.emit }) return handle_language_conversion_block(self, ...) end function Parser:braces(this, fail_on_unclosed_braces) local language_conversion_block = self:read(-1) == "-" if self:read(1) == "{" then local braces, failed = self:template_or_parameter() -- Headings will fail if they contain an unclosed brace block. if failed and fail_on_unclosed_braces then return self:fail_route() -- Language conversion blocks cannot begin "-{{", but can begin -- "-{{{" iff parsed as "-{" + "{{". elseif not (language_conversion_block and braces == 1) then return self:consume() end else self:emit(this) if not language_conversion_block then return end self:advance() end self:emit(Wikitext:new(self:get(do_language_conversion_block))) end end --[==[ Headings Opens with "\n=" (or "=" at the start of the text), and closes with "\n" or the end of the text. Note that it doesn't matter whether the heading will fail to process due to a premature newline (e.g. if there are no closing signs), so at this stage the only thing that matters for closure is the newline or end of text. Note: Heading blocks are only parsed like this if they occur inside a template, since they do not iterate the preparser's heading count (i.e. they aren't proper headings). Note 2: if directly inside a template argument with no previous equals signs, a newline followed by a single equals sign is parsed as an argument equals sign, not the opening of a new L1 heading block. This does not apply to any other heading levels. As such, {{template|key\n=}}, {{template|key\n=value}} or even {{template|\n=}} will successfully close, but {{template|key\n==}}, {{template|key=value\n=more value}}, {{template\n=}} etc. will not, since in the latter cases the "}}" would fall inside the new heading block. ]==] do --Handler. local handle_heading_block local function do_heading_block(self) self.current_layer._parse_data.handler = handle_heading_block self:set_pattern("[\n<[{]") end function handle_heading_block(self, ...) handle_heading_block = self:switch(handle_heading_block, { ["\n"] = function(self) self:newline() return self:pop() end, ["<"] = Parser.tag, ["["] = Parser.wikilink_block, ["{"] = Parser.braces, [""] = Parser.pop, [false] = Parser.emit }) return handle_heading_block(self, ...) end function Parser:heading_block(this, nxt) self:newline() this = this .. (nxt or "=") local loc = #this - 1 while self:read(0, loc) == this do self:advance() self:emit(Wikitext:new(self:get(do_heading_block))) end end end -- Wikilink block. -- Opens with "[[" and closes with "]]". do -- Handler. local handle_wikilink_block local function do_wikilink_block(self) self.current_layer._parse_data.handler = handle_wikilink_block self:set_pattern("[\n<[%]{]") end function handle_wikilink_block(self, ...) handle_wikilink_block = self:switch(handle_wikilink_block, { ["\n"] = Parser.heading_block, ["<"] = Parser.tag, ["["] = Parser.wikilink_block, ["]"] = function(self, this) if self:read(1) == "]" then self:emit("]]") self:advance() return self:pop() end self:emit(this) end, ["{"] = Parser.braces, [""] = Parser.pop, [false] = Parser.emit }) return handle_wikilink_block(self, ...) end function Parser:wikilink_block() if self:read(1) == "[" then self:emit("[[") self:advance(2) self:emit(Wikitext:new(self:get(do_wikilink_block))) else self:emit("[") end end end -- Lines which only contain comments, " " and "\t" are eaten, so long as -- they're bookended by "\n" (i.e. not the first or last line). function Parser:newline() local text, head = self.text, self.head while true do repeat local loc = match(text, "^[\t ]*<!%-%-()", head + 1) if not loc then break end loc = select(2, find(text, "-->", loc, true)) head = loc or head until not loc -- Fail if no comments found. if head == self.head then break end head = match(text, "^[\t ]*()\n", head + 1) if not head then break end self:jump(head) end self:emit("\n") end do -- Handlers. local handle_start local main_handler -- If `transcluded` is true, then the text is checked for a pair of -- onlyinclude tags. If these are found (even if they're in the wrong -- order), then the start of the page is treated as though it is preceded -- by a closing onlyinclude tag. -- Note 1: unlike other parser extension tags, onlyinclude tags are case- -- sensitive and cannot contain whitespace. -- Note 2: onlyinclude tags *can* be implicitly closed by the end of the -- text, but the hard requirement above means this can only happen if -- either the tags are in the wrong order or there are multiple onlyinclude -- blocks. local function do_parse(self, transcluded) self.current_layer._parse_data.handler = handle_start self:set_pattern(".") self.section = 0 if not transcluded then return end self.transcluded = true local text = self.text if find(text, "</onlyinclude>", nil, true) then local head = find(text, "<onlyinclude>", nil, true) if head then self.onlyinclude = true self:jump(head + 13) end end end -- If the first character is "=", try parsing it as a heading. function handle_start(self, this) self.current_layer._parse_data.handler = main_handler self:set_pattern("[\n<{]") if this == "=" then return self:heading() end return self:consume() end function main_handler(self, ...) main_handler = self:switch(main_handler, { ["\n"] = function(self) self:newline() if self:read(1) == "=" then self:advance() return self:heading() end end, ["<"] = Parser.tag, ["{"] = function(self, this) if self:read(1) == "{" then self:template_or_parameter() return self:consume() end self:emit(this) end, [""] = Parser.pop, [false] = Parser.emit }) return main_handler(self, ...) end function export.parse(text, transcluded) local text_type = type(text) return (select(2, Parser:parse{ text = text_type == "string" and text or text_type == "number" and tostring(text) or error("bad argument #1 (string expected, got " .. text_type .. ")"), node = {Wikitext, true}, route = {do_parse, transcluded} })) end parse = export.parse end function export.find_templates(text, not_transcluded) return parse(text, not not_transcluded):iterate_nodes("template") end do local link_parameter_1, link_parameter_2 local function get_link_parameter_1() link_parameter_1, get_link_parameter_1 = (data or get_data()).template_link_param_1, nil return link_parameter_1 end local function get_link_parameter_2() link_parameter_2, get_link_parameter_2 = (data or get_data()).template_link_param_2, nil return link_parameter_2 end -- Generate a link. If the target title doesn't have a fragment, use "#top" -- (which is an implicit anchor at the top of every page), as this ensures -- self-links still display as links, since bold display is distracting and -- unintuitive for template links. local function link_page(title, display) local fragment = title.fragment if fragment == "" then fragment = "top" end return format( "[[:%s|%s]]", encode_uri(title.prefixedText .. "#" .. fragment, "WIKI"), display ) end -- pf_arg1 or pf_arg2 may need to be linked if a given parser function -- treats them as a pagename. If a key exists in `namespace`, the value is -- the namespace for the page: if not 0, then the namespace prefix will -- always be added to the input (e.g. {{#invoke:}} can only target the -- Module: namespace, so inputting "Template:foo" gives -- "Module:Template:foo", and "Module:foo" gives "Module:Module:foo"). -- However, this isn't possible with mainspace (namespace 0), so prefixes -- are respected. make_title() handles all of this automatically. local function finalize_arg(pagename, namespace) if namespace == nil then return pagename end local title = make_title(namespace, pagename) return title and not title.isExternal and link_page(title, pagename) or pagename end local function render_title(name, args) -- parse_template_name returns a table of transclusion modifiers plus -- the normalized template/magic word name, which will be used as link -- targets. The third return value pf_arg1 is the first argument of a -- a parser function, which comes after the colon (e.g. "foo" in -- "{{#IF:foo|bar|baz}}"). This means args[1] (i.e. the first argument -- that comes after a pipe is actually argument 2, and so on. Note: the -- second parameter of parse_template_name checks if there are any -- arguments, since parser variables cannot take arguments (e.g. -- {{CURRENTYEAR}} is a parser variable, but {{CURRENTYEAR|foo}} -- transcludes "Template:CURRENTYEAR"). In such cases, the returned -- table explicitly includes the "Template:" prefix in the template -- name. The third parameter instructs it to retain any fragment in the -- template name in the returned table, if present. local chunks, subclass, pf_arg1 = parse_template_name( name, args and pairs(args)(args) ~= nil, true ) if chunks == nil then return name, args end local chunks_len = #chunks -- Additionally, generate the corresponding table `rawchunks`, which -- is a list of colon-separated chunks in the raw input. This is used -- to retrieve the display forms for each chunk. local rawchunks = split(name, ":") for i = 1, chunks_len - 1 do chunks[i] = format( "[[%s|%s]]", encode_uri((magic_words or get_magic_words())[sub(chunks[i], 1, -2)].transclusion_modifier, "WIKI"), rawchunks[i] ) end local chunk = chunks[chunks_len] -- If it's a template, return a link to it with link_page, concatenating -- the remaining chunks in `rawchunks` to form the display text. -- Use new_title with the default namespace 10 (Template:) to generate -- a target title, which is the same setting used for retrieving -- templates (including those in other namespaces, as prefixes override -- the default). if subclass == "template" then chunks[chunks_len] = link_page( new_title(chunk, 10), concat(rawchunks, "&#58;", chunks_len) -- : ) return concat(chunks, "&#58;"), args -- : elseif subclass == "parser variable" then chunks[chunks_len] = format( "[[%s|%s]]", encode_uri((magic_words or get_magic_words())[chunk].parser_variable, "WIKI"), rawchunks[chunks_len] ) return concat(chunks, "&#58;"), args -- : end -- Otherwise, it must be a parser function. local mgw_data = (magic_words or get_magic_words())[sub(chunk, 1, -2)] local link = mgw_data.parser_function or mgw_data.transclusion_modifier local pf_arg2 = args and args[1] or nil -- Some magic words have different links, depending on whether argument -- 2 is specified (e.g. "baz" in {{foo:bar|baz}}). if type(link) == "table" then link = pf_arg2 and link[2] or link[1] end chunks[chunks_len] = format( "[[%s|%s]]", encode_uri(link, "WIKI"), rawchunks[chunks_len] ) -- #TAG: has special handling, because documentation links for parser -- extension tags come from [[Module:data/parser extension tags]]. if chunk == "#TAG:" then -- Tags are only case-insensitive with ASCII characters. local tag = (parser_extension_tags or get_parser_extension_tags())[lower(php_trim(pf_arg1))] if tag then pf_arg1 = format( "[[%s|%s]]", encode_uri(tag, "WIKI"), pf_arg1 ) end -- Otherwise, finalize pf_arg1 and add it to `chunks`. else pf_arg1 = finalize_arg(pf_arg1, (link_parameter_1 or get_link_parameter_1())[chunk]) end chunks[chunks_len + 1] = pf_arg1 -- Finalize pf_arg2 (if applicable), then return. if pf_arg2 then args = shallow_copy(args) -- Avoid destructively modifying args. args[1] = finalize_arg(pf_arg2, (link_parameter_2 or get_link_parameter_2())[chunk]) end return concat(chunks, "&#58;"), args -- : end function export.buildTemplate(title, args) local output = {title} if not args then return output end -- Iterate over all numbered parameters in order, followed by any -- remaining parameters in codepoint order. Implicit parameters are -- used wherever possible, even if explicit numbers are interpolated -- between them (e.g. 0 would go before any implicit parameters, and -- 2.5 between 2 and 3). -- TODO: handle "=" and "|" in params/values. local implicit for k, v in sorted_pairs(args) do if type(k) == "number" and k >= 1 and k % 1 == 0 then if implicit == nil then implicit = table_len(args) end insert(output, k <= implicit and v or k .. "=" .. v) else insert(output, k .. "=" .. v) end end return output end build_template = export.buildTemplate function export.templateLink(title, args, no_link) if not no_link then title, args = render_title(title, args) end local output = build_template(title, args) for i = 1, #output do output[i] = encode_entities(output[i], "={}", true, true) end return tostring(html_create("code") :css("white-space", "pre-wrap") :wikitext("&#123;&#123;" .. concat(output, "&#124;") .. "&#125;&#125;") -- {{ | }} ) end end do function export.find_parameters(text, not_transcluded) return parse(text, not not_transcluded):iterate_nodes("parameter") end function export.displayParameter(name, default) return tostring(html_create("code") :css("white-space", "pre-wrap") :wikitext("&#123;&#123;&#123;" .. concat({name, default}, "&#124;") .. "&#125;&#125;&#125;") -- {{{ | }}} ) end end do local function check_level(level) if type(level) ~= "number" then error("Heading levels must be numbers.") elseif level < 1 or level > 6 or level % 1 ~= 0 then error("Heading levels must be integers between 1 and 6.") end return level end -- FIXME: should headings which contain "\n" be returned? This may depend -- on variable factors, like template expansion. They iterate the heading -- count number, but fail on rendering. However, in some cases a different -- heading might still be rendered due to intermediate equals signs; it -- may even be of a different heading level: e.g., this is parsed as an -- L2 heading with a newline (due to the wikilink block), but renders as the -- L1 heading "=foo[[". Section edit links are sometimes (but not always) -- present in such cases. -- ==[[= -- ]]== -- TODO: section numbers for edit links seem to also include headings -- nested inside templates and parameters (but apparently not those in -- parser extension tags - need to test this more). If we ever want to add -- section edit links manually, this will need to be accounted for. function export.find_headings(text, i, j) local parsed = parse(text) if i == nil and j == nil then return parse(text):iterate_nodes("heading") end i = i and check_level(i) or 1 j = j and check_level(j) or 6 return parsed:iterate(function(v) if class_else_type(v) == "heading" then local level = v.level return level >= i and level <= j end end) end end do local function make_tag(tag) return tostring(html_create("code") :css("white-space", "pre-wrap") :wikitext("&lt;" .. tag .. "&gt;") ) end -- Note: invalid tags are returned without links. function export.wikitagLink(tag) -- ">" can't appear in tags (including attributes) since the parser -- unconditionally treats ">" as the end of a tag. if find(tag, ">", nil, true) then return make_tag(tag) end -- Tags must start "<tagname..." or "</tagname...", with no whitespace -- after "<" or "</". local slash, tagname, remainder = match(tag, "^(/?)([^/%s]+)(.*)$") if not tagname then return make_tag(tag) end -- Tags are only case-insensitive with ASCII characters. local link = lower(tagname) if ( -- onlyinclude tags must be lowercase and are whitespace intolerant. link == "onlyinclude" and (link ~= tagname or remainder ~= "") or -- Closing wikitags (except onlyinclude) can only have whitespace -- after the tag name. slash == "/" and not match(remainder, "^%s*()$") or -- Tagnames cannot be followed immediately by "/", unless it comes -- at the end (e.g. "<nowiki/>", but not "<nowiki/ >"). remainder ~= "/" and sub(remainder, 1, 1) == "/" ) then -- Output with no link. return make_tag(tag) end -- Partial transclusion tags aren't in the table of parser extension -- tags. if link == "noinclude" or link == "includeonly" or link == "onlyinclude" then link = "mw:Transclusion#Partial transclusion" else link = (parser_extension_tags or get_parser_extension_tags())[link] end if link then tag = gsub(tag, pattern_escape(tagname), "[[" .. replacement_escape(encode_uri(link, "WIKI")) .. "|%0]]", 1) end return make_tag(tag) end end -- For convenience. export.class_else_type = class_else_type return export g7883esvq9rqr796jeq8jsmk6y4r8t1