Module:documentation: Difference between revisions

From Linguifex
Jump to navigation Jump to search
No edit summary
No edit summary
Line 1: Line 1:
local export = {}
local p = {}
local data = mw.loadData( 'Module:Interlinear/data' )
local gloss_override = {} -- for custom gloss abbreviations
local getArgs = require('Module:Arguments').getArgs
local yesno = require('Module:Yesno')
local lang_data = mw.loadData( 'Module:Lang/data' )


-- it is either here, or in [[Module:ugly hacks]], and it is not in ugly hacks.
--------------------------
function export.CONTENTMODEL()
-- Almost-global variables
return mw.title.getCurrentTitle().contentModel
--------------------------
end
local glossing_type, displaying_messages, free_translation, msg, buffer


local skins = {
-------------------
["common"     ] = "";
-- General settings
["vector"     ] = "Vector";
-------------------
["monobook"   ] = "Monobook";
local conf = { --settings
["cologneblue"] = "Cologne Blue";
WordSeparator = " \n\r\t", -- Don't replace with %s as this would include non-breaking spaces
["modern"     ] = "Modern";
GlossAbbrPattern = "^([Ø0-9A-Z]+)$", -- this isn't a full regex, but a Lua pattern
-- NOTE: The following characters must be formatted for use in a pattern set.
GlossAbbrBoundary = "-.,;:<>‹›/\\~+=%?%s%[%]()%_\127'",
GlossExcludeTable = {I = true,}, --strings not be treated as glossing abbreviations
GlossExcludePattern = '^[0-9][0-9]+$', -- excludes strings consisting entirely of digits
GlossSmallCapsExclude = "^[AOPS]$", -- glossing abbreviations matching this pattern will not be rendered in small caps
GlossingType = "label", -- if set to "label" gloss abbreviations are formatted as an <abbr> with the "label" appearing in a tooltip
-- if set to "wikilink" the abbreviation is formatted as a wikilink to the relevant wikipedia article
-- if set to "none" abbreviations aren't formatted at all
ErrorCategory = "[[Category:Pages with errors in interlinear text]]",
AmbiguousGlossCategory = "[[Category:Articles with ambiguous glossing abbreviations]]",
MessageGlossingError = "Error(s) in interlinear glossing",
combining_gender_numbers = "[0-9][0-9]?$", --e.g. G4 '4th gender' or CL7 'class 7'
combining_gender_prefixes = {G = "gender", CL = "class"},
combining_person = {["1"] = "first person", ["2"] = "second person", ["3"] = "third person"},
combining_number = {S = "singular", SG = "singular", P = "plural", PL = "plural", D = "dual", DU = "dual", TRI = "trial"},
combining_gender = {F = "feminine", M = "masculine", N = "neuter"},
LowerCaseGlosses = {["1sg"] = true, ["2sg"] = true, ["3sg"] = true, ["1du"] = true, ["2du"] = true, ["3du"] = true, ["1pl"] = true, ["2pl"] = true,
["3pl"] = true, ["Fsg"] = true, ["Fpl"] = true, ["Msg"] = true, ["Mpl"] = true,}, -- these are the non-all-upper-case strings that will be recognised as glossing abbreviations
ErrorHelpLocation = "Template:Interlinear",
}
}


local Array = require "Module:array"
---------------------
-- CSS styles and classes
---------------------
conf.style = { --CSS styles
WordDiv = "float: left; margin-bottom: 0.3em;",
WordMargin = "margin-right: 1em;",
WordP = "margin: 0px;", -- the style for the word <p> elements
GlossAbbr = "font-variant: small-caps; font-variant-numeric: oldstyle-nums; text-transform: lowercase; ", -- won't be applied to gloss abbreviations containing lower-case characters
HiddenText = "display: none;",
EndDiv = "clear: left; display: block;", -- style of the <div> element at the end of the interlinear display
ErrorMessage = "font-size: inherit",
}
conf.class = { --CSS classes
Interlinear = "interlinear",
GlossAbbr  = "gloss-abbr",
GlossAbbrAmb = "gloss-abbr-ambiguous",
GlossAbbrError = "gloss-abbr-error",
ErrorMessage = "error",
}
---------------------
-- Section transclusion
---------------------
local page_content = nil -- lazy initilization
local function get_section(frame, section_name)
if page_content == nil then
local current_title = mw.title.getCurrentTitle()
page_content = current_title:getContent()
end
if page_content then
if mw.ustring.find(page_content, section_name, 1, true) then
return frame:preprocess('{{#section:{{FULLPAGENAME}}|' .. section_name .. '}}')
end
end
return ''
end
---------------------
-- Sundry small functions
---------------------
local function normalise(str)
return mw.ustring.gsub(str,"[" .. conf.WordSeparator .. "]+"," ")
end


local function compare_pages(page1, page2, text)
local function tidyCss(str)
return "[" .. tostring(
str = mw.ustring.gsub(str, '^[\"\']*(.-)[\"\']*$', "%1") -- trims quotation marks
mw.uri.fullUrl("Special:ComparePages", { page1 = page1, page2 = page2 }))
if mw.ustring.sub(str, -1) ~= ";" then str = str .. ";" end -- appends ";" if missing
.. " " .. text .. "]"
return str
end
end


local function page_exists(title)
local function highlight(text)
local success, title_obj = pcall(mw.title.new, title)
if text then
return success and title_obj.exists
return '<span style="color:#C00;font-weight:bold;">' .. text .. '</span>'
else return "" end
end
end


-- Avoid transcluding [[Module:languages/cache]] everywhere.
local function tone_sup(str)
local lang_cache = setmetatable({}, { __index = function (self, k)
return mw.ustring.gsub(str, "([^%p%s0-9])([0-9]+)", "%1<sup>%2</sup>")
return require "Module:languages/cache"[k]
end
end })


local function zh_link(word)
local function is_empty(str) -- returns "false" if its argument is a string containing chars other than spaces &c.
return require("Module:links").full_link{
if not str then return true end
lang = lang_cache.zh,
if mw.ustring.find(str, "[^" .. conf.WordSeparator .. "]")
term = word
then return false
}
else return true end
end
end


local function make_Unicode_data_documentation(title, cats)
local function help_link (anchor)
local subpage, first_three_of_code_point
if anchor then
= title.fullText:match("^Module:Unicode data/([^/]+)/(%x%x%x)$")
return " ([[" .. conf.ErrorHelpLocation .. "#" .. anchor .. "|help]])"
if subpage == "names" or subpage == "images" then
else return "" end
local low, high =
tonumber(first_three_of_code_point .. "000", 16),
tonumber(first_three_of_code_point .. "FFF", 16)
return string.format(
"This data module contains the %s of " ..
"[[Appendix:Unicode|Unicode]] code points within the range U+%04X to U+%04X.",
subpage == "images" and "titles of images" or "names",
low, high)
end
end
end


-- This provides categories and documentation for various data modules, so that
-- the following is part of a trial implementation of automatic transliteration:
-- [[Category:Uncategorized modules]] isn’t unnecessarily cluttered.
local function transliterate (str, lang_from, lang_to,  scheme)
-- It is a list of tables, each of which have the following possible fields:
local lookup = {grc = {module = 'Module:Ancient Greek', funct = "transliterate", } }
if not lang_from then
msg:add("error", "Source language for transliteration is not set")
else
local t = lookup[lang_from]
if t then
local module = require(t.module)
return module[t.funct](str)
else msg:add("error", "Can't find transliterator for language '" .. lang_from .. "'")
end
end
return ""
end -- end of trial block


-- `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
-- The following two functions update the glossing settings based on the received
-- second, etc. (often used for creating the sortkey for the category). In addition, the captures are passed to the
-- template arguments. set_global_glossing_settings() updates the global settings
-- `process` function as the third and subsequent parameters.
-- that are valid for all gloss abbreviations. set_glossing_type()
-- returns the glossing type, which can vary between the different lines.
--------------------
local function set_global_glossing_settings(a)
local style = ""
if a.style then style = tidyCss(a.style) end
if a.underline == "no" then
style = style .. "text-decoration: none;" end
if a.small_caps == "no" then
style = style .. "font-variant:normal; text-transform: none;" end
if style ~= "" then conf.style.GlossAbbr = conf.style.GlossAbbr .. style end
end


-- `process` (optional): This may be a function or a string. If it is a function, it is called as follows:
local function set_glossing_type(glossing)
--    process(TITLE, CATS, CAPTURE1, CAPTURE2, ...)
if glossing then
-- where:
local GlossingType
--    * TITLE is a title object describing the module's title; see [https://www.mediawiki.org/wiki/Extension:Scribunto/Lua_reference_manual#Title_objects].
glossing = mw.ustring.lower(mw.text.trim(glossing))
--    * CATS is an array object (see [[Module:array]]) of categories that the module will be added to.
if mw.ustring.find(glossing, 'link') then
--    * CAPTURE1, CAPTURE2, ... contain any captures in the `regex` field.
GlossingType = "wikilink"
-- The return value of `process` should be a string and will be used as the module's documentation.
elseif mw.ustring.find(glossing, 'label')
-- If `process` is a string, it is the name of a submodule under 'Module:documentation/functions/' which returns a
or  mw.ustring.find(glossing, 'no link') then
-- function, of the same type as described above. If `process` is omitted entirely, the module will have no
GlossingType = 'label'
-- documentation.
elseif mw.ustring.find(glossing, 'no abbr') then
GlossingType = "no abbr"
elseif yesno(glossing) == false then
GlossingType = nil
elseif yesno(glossing) then
GlossingType = conf.GlossingType
else
msg:add('error', 'Glossing type "' .. glossing .. '" not recognised') end
return GlossingType
else error("set_glossing_type: 'glossing' is nil or false", 2)
end
end


-- `cat` (optional): A string naming the category into which the module should be placed. Captures specified in `regex`
local function set_custom_glosses(list)
-- may be referenced in this string using %1 for the first capture, %2 for the second, etc. Currently, in order to put
local abbs = mw.text.split(list, '[;\n\t]')
-- the module into multiple categories, you need to write a `process` function that adds categories to the CATS
for _,v in pairs(abbs) do
-- parameter passed in.
local gloss = mw.text.split(v, ':')
local module_regex = {
local a = mw.text.trim(gloss[1])
{
if a and a ~= "" then
regex = "^Module:languages/.+$",
gloss_override[a] = {}
cat = "Language and script modules",
gloss_override[a].expansion = gloss[2]
},
gloss_override[a].wikipage = gloss[3]
{
regex = "^Module:scripts/.+$",
cat = "Language and script modules",
},
{
regex = "^Module:data tables/data..?.?.?$",
cat = "Reference module sharded data tables",
},
{
regex = "^Module:Quotations/...?.?.?.?.?.?$",
cat = "Quotation data modules",
},
{
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 dialectal 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 = "number list",
},
{
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/(.+)$")
if word then
local lang = lang_cache["inc-ash"]
return ("This module contains data on the pronunciation of %s in dialects of %s.")
:format(require("Module:links").full_link({ term = word, lang = lang }, "term"),
lang:makeCategoryLink())
end
end,
},
{
regex = "^Module:[%l-]+:Dialects$",
process = function(title, cats)
local content = title:getContent()
local has_aliases = content:find("aliases") ~= nil
return mw.getCurrentFrame():expandTemplate {
title = "dialectal data module",
args = { ["labels-aliases"] = has_aliases },
}
end,
},
{
regex = "^.+%-translit$",
process = "translit",
},
{
regex = "^Module:category tree/poscatboiler/data/lang%-specific/(.+)$",
process = function(title, cats, lang_code)
local lang = lang_cache[lang_code]
if lang then
local langname = lang:getCanonicalName()
cats:insert("Category tree data modules/poscatboiler|" .. langname)
cats:insert(langname .. " modules")
cats:insert(langname .. " data modules")
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 poscatboiler system, which is a general framework for generating the "
.. "descriptions and categorization of category pages.\n\n"
.. "For more information, see [[Module:category tree/poscatboiler/data/lang-specific/documentation]]."
end
end
end
},
end
{
end
regex = "^Module:category tree/poscatboiler/data/(.+)$",
cat = "Category tree data modules/poscatboiler|%1",
},
{
regex = "^Module:ja/data/(.+)$",
cat = "Japanese data modules|%1",
},
{
regex = "^Module:Swadesh/data/(.+)$",
cat = "Swadesh modules|%1",
},
{
regex = "^Module:typing%-aids",
process = function(title, cats)
local data_suffix = title.fullText:match("^Module:typing%-aids/data/(.+)$")
if data_suffix then
cats:insert("Data modules")
if data_suffix:find "^[%l-]+$" then
local lang = require "Module:languages".getByCode(data_suffix)
if lang then
cats:insert(lang:getCanonicalName() .. " modules")
end
elseif data_suffix:find "^%u%l%l%l$" then
local script = require "Module:scripts".getByCode(data_suffix)
if script then
cats:insert(script:getCategoryName())
end
end
end
end,
},
{
regex = "^Module:R:([a-z%-]+):(.+)$",
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 mw.getCurrentFrame():preprocess("This module implements the reference template {{temp|R:" ..
lang_code .. ":" .. refname .. "}}.")
end
end,
},
}


function export.show(frame)
---------------------
local params = {
-- The UserMessages object contains and processes error messages and warnings
["hr"] = {},
---------------------
["for"] = {},
local UserMessages = {errors = {}, warnings = {}, gloss_messages = {}}
["from"] = {},
function UserMessages:add(msgtype, text, gloss)
["notsubpage"] = { type = "boolean", default = false },
if msgtype == "gloss_message" then
["nodoc"] = { type = "boolean", default = false },
self.gloss_messages[gloss] = text
}
elseif msgtype == "warning" then
table.insert(self.warnings, text)
local args = require("Module:parameters").process(frame.args, params)
elseif msgtype == "non-repeating error" then
self.errors.nre = text
local output = Array('\n<div class="documentation" style="display:block; clear:both">\n')
elseif msgtype == "ambiguous gloss" then
local cats = Array()
self.if_ambiguous_glosses = true
elseif msgtype == "error" then
local nodoc = args.nodoc
table.insert(self.errors, text)
else return error("UserMessages:add(): unknown message type", 2)
if (not args.hr) or (args.hr == "above") then
output:insert("----\n")
end
end
end
local title = ((args['for'] ~= "") and args['for']) and mw.title.new(args['for'])
function UserMessages:print_errors()
or mw.title.getCurrentTitle()
local out = ""
local doc_title = mw.title.new((args['from'] ~= "") and args['from']
local namespace = mw.title.getCurrentTitle().namespace
or (title.fullText .. '/doc'))
if next(self.errors) or self.warnings[1] then
local contentModel = title.contentModel
local err_span = mw.html.create("span")
err_span:attr("style", conf.style.ErrorMessage)
local pagetype = mw.getContentLanguage():lcfirst(title.nsText) .. " page"
err_span:addClass(conf.class.ErrorMessage)
local preload, fallback_docs, doc_content, doc_content_module,
for _,v in pairs(self.errors) do
old_doc_title, user_name, skin_name, needs_doc
err_span:wikitext(" " .. v .. ";") end
if namespace % 2 == 0 and namespace ~= 2 -- non-talk namespaces, excluding user pages; if modifying please update the description on the category page
if contentModel == "javascript" then
then err_span:wikitext(conf.ErrorCategory)
pagetype = "script"
if title.nsText == 'MediaWiki' then
if title.text:find('Gadget-') then
preload = 'Template:documentation/preloadGadget'
else
preload = 'Template:documentation/preloadMediaWikiJavaScript'
end
else
preload  = 'Template:documentation/preloadTemplate' -- XXX
end
end
if title.nsText == 'User' then
out = tostring(err_span)
user_name = title.rootText
mw.addWarning(conf.MessageGlossingError)
end
if self.if_ambiguous_glosses then
if namespace == 0 -- article namespace
then out = out .. conf.AmbiguousGlossCategory -- this category will only track articles
end
end
elseif contentModel == "css" then
end
pagetype = "style sheet"
return out
preload  = 'Template:documentation/preloadTemplate' -- XXX
end
if title.nsText == 'User' then
function UserMessages:print_warnings()
user_name = title.rootText
local out = ""
-- Messages and warnings get displayed only if the page is being viewed in "preview" mode:
if displaying_messages and (next(self.gloss_messages) or next(self.warnings)) then
local div = mw.html.create("div")
div:addClass("interlinear-preview-warning")
:cssText('border: 1px solid #a2a9b1; background-color: #f8f9fa; width: 80%; padding: 0.2em;')
:wikitext("<i>This message box is shown only in preview:</i>")
:newline()
for _,v in ipairs(self.warnings) do
local p = div:tag("p")
p:addClass(conf.class.ErrorMessage)
p:attr("style", conf.style.ErrorMessage)
p:wikitext(v)
end
end
elseif contentModel == "Scribunto" then
if self.gloss_messages then
pagetype = "module"
div:wikitext("<p>  To change any of the following default expansions, see [[Template:Interlinear/doc#Custom abbreviations|the template's documentation]]:</p>")
user_name = title.rootText:match("^[Uu]ser:(.+)")
end
if user_name then
for _,v in pairs(self.gloss_messages) do
preload  = 'Template:documentation/preloadModuleSandbox'
div:wikitext("<p>" .. v .. "</p>")
else
preload  = 'Template:documentation/preloadModule'
end
end
elseif title.nsText == "Template" then
out = out .. "\n\n" .. tostring(div)
pagetype = "template"
preload  = 'Template:documentation/preloadTemplate'
elseif title.nsText == "Wiktionary" then
pagetype = "project page"
preload  = 'Template:documentation/preloadTemplate' -- XXX
end
if doc_title.isRedirect then
old_doc_title = doc_title
doc_title = mw.title.new(string.match(doc_title:getContent(),
"^#[Rr][Ee][Dd][Ii][Rr][Ee][Cc][Tt]%s*:?%s*%[%[([^%[%]]-)%]%]"))
end
end
return out
end


output:insert("<dl class=\"plainlinks\" style=\"font-size: smaller;\">")
---------------------
if doc_title.exists then
-- gloss_lookup() receives a gloss abbreviation and tries to uncover its meaning.
output:insert(
---------------------
"<dd><i style=\"font-size: larger;\">The following " ..
local function gloss_lookup(a, label, wikilink)
"[[wikt:Help:Documenting templates and modules|documentation]] is located at [[" ..
local _label, _wikilink, _lookup, source = nil, nil, nil, nil
doc_title.fullText .. "]]. " ..
if gloss_override[a] then
"<sup>[[" .. doc_title:fullUrl { action = 'edit' } .. " edit]]</sup> </i></dd>")
_lookup = gloss_override[a]
source = "local"
elseif data.abbreviations[a] then _lookup = data.abbreviations[a] end
if _lookup and _lookup.expansion ~= "" then
_label, _wikilink = _lookup.expansion, _lookup.wikipage
else
else
if contentModel == "Scribunto" then
local prefix = mw.ustring.sub(a,1,1)
local automatic_cats = nil
local suffix = mw.ustring.sub(a,2)
if user_name then
if conf.combining_person[prefix] then -- is it of the form 1PL or 3FS?
fallback_docs = "documentation/fallback/user module"
_label = conf.combining_person[prefix]
automatic_cats = {"User sandbox modules"}
local _suffix = conf.combining_number[suffix] or conf.combining_gender[suffix]
if _suffix then
_label = _label .. ", " .. _suffix
else
else
for _, data in ipairs(module_regex) do
local suffix1 = mw.ustring.sub(suffix,1,1)
        local captures = {mw.ustring.match(title.fullText, data.regex)}
local suffix2 = mw.ustring.sub(suffix,2)
        if #captures > 0 then
if conf.combining_gender[suffix1]
local cat
and  conf.combining_number[suffix2] then
local process_function
_label = _label .. ", " .. conf.combining_gender[suffix1] .. ", " .. conf.combining_number[suffix2]
if type(data.process) == "function" then
else _label = nil end
process_function = data.process
elseif type(data.process) == "string" then
doc_content_module = "Module:documentation/functions/" .. data.process
process_function = require(doc_content_module)
end
if process_function then
doc_content = process_function(title, cats, unpack(captures))
end
cat = data.cat
if cat then
if type(cat) == "string" then
cat = {cat}
end
for _, c in ipairs(cat) do
-- gsub() and Lua :gsub() return two arguments, which causes all sorts of problems.
-- Terrible design, there should have been a separate two-argument function.
local gsub_sucks = mw.ustring.gsub(title.fullText, data.regex, c)
table.insert(cats, gsub_sucks)
end
end
break
end
end
end
end
 
elseif mw.ustring.match(suffix,conf.combining_gender_numbers) then -- cases like G4 = gender 4
if not automatic_cats then
local _i,_j = mw.ustring.find(a, conf.combining_gender_numbers)
local success, auto_cats =
local _pre = mw.ustring.sub(a, 1, _i - 1)
pcall(require("Module:module categorization").categorize, frame, true)
local _suff = mw.ustring.sub(a, _i)
if success then
if conf.combining_gender_prefixes[_pre] then
for _, category in ipairs(auto_cats) do
_label = conf.combining_gender_prefixes[_pre] .. " " .. _suff
cats:insert(category)
end
end
elseif prefix == "N" then -- dealing with cases like NPST = non-past
end
local s = gloss_override[suffix] or data.abbreviations[suffix]
if s ~= nil and not s.ExcludeNegation then
_label = "non-" .. s.expansion
_wikilink = s.wikipage
end
end
s = nil
end
end
if _label == "" then _label = nil end
if _wikilink == "" then _wikilink = nil end
if not label then label = _label end
if not wikilink then wikilink = _wikilink end
return label, wikilink, source
end


if title.subpageText == "templates" then
---------------------
cats:insert("Template interface modules")
-- format_gloss() calls gloss_lookup() to find the meaning of a gloss
-- abbreviation, which it then proceeds to format
---------------------
local function format_gloss(gloss, label, wikilink)
if string.sub(gloss,1,3) == "000" then -- checks for a common component of exposed strip markers (see [[:mw:Strip marker]])
return gloss
end
local gloss2 = mw.ustring.gsub(gloss,"<.->","") -- remove any html fluff
gloss2 = mw.ustring.gsub(gloss2, "%'%'+", "") -- remove wiki bold/italic formatting
gloss2 = mw.text.trim(mw.ustring.upper(gloss2))
if not (label or wikilink)
or (not label and glossing_type == "label")
or (not wikilink  and glossing_type == "wikilink")
then
if glossing_type ~= "no abbr"
then label, wikilink, source = gloss_lookup(gloss2, label, wikilink)
end
end
 
end
if automatic_cats then
local gloss_node
for _, c in ipairs(automatic_cats) do
if glossing_type == "no abbr"
cats:insert(c)
then gloss_node = mw.html.create("span")
end
else gloss_node = mw.html.create("abbr") end
end
gloss_node:addClass(conf.class.GlossAbbr)
if label or wikilink then
-- meaning module is not in user’s sandbox or one of many datamodule boring series
if not mw.ustring.match(gloss, "%l") -- excluding glosses that contain lower-case characters
needs_doc = not (automatic_cats or doc_content or fallback_docs)
and not mw.ustring.match(gloss,conf.GlossSmallCapsExclude) -- and also excluding A, O etc. from rendering in small caps
elseif title.nsText == "Template" then
then gloss_node:attr("style", conf.style.GlossAbbr)
--cats:insert("Uncategorized templates")
end
needs_doc = not (fallback_docs or nodoc)
local abbr_label
elseif (contentModel == "css") or (contentModel == "javascript") then
if label then abbr_label = label
if user_name then
else abbr_label = wikilink end
skin_name = skins[title.text:sub(#title.rootText + 1):match("^/([a-z]+)%.[jc]ss?$")]
gloss_node:attr("title", abbr_label)
if skin_name then
if source ~= "local" and data.abbreviations[gloss2] then
fallback_docs = 'documentation/fallback/user ' .. contentModel
if data.abbreviations[gloss2].ambiguous then
gloss_node:addClass(conf.class.GlossAbbrAmb)
msg:add("ambiguous gloss")
end
end
end
if glossing_type == "wikilink" and wikilink
then gloss_node:wikitext("[[", wikilink, "|" , gloss, "]]")
else gloss_node:wikitext(gloss) end
if source ~= "local" and displaying_messages then -- logging gloss lookups:
local message = ""
if label then
message = "assuming " .. gloss2 .. " means \"" .. abbr_label .. "\";" end
if glossing_type == "wikilink" and wikilink then
message = message .. " linking to [[" .. wikilink .. "]];"
end
end
msg:add("gloss_message", message, gloss)
end
end
elseif glossing_type == "no abbr"
if doc_content then
then gloss_node
output:insert(
:attr("style", conf.style.GlossAbbr)
"<dd><i style=\"font-size: larger;\">The following " ..
:wikitext(gloss)
"[[wikt:Help:Documenting templates and modules|documentation]] is " ..
else
"generated by [[" .. (doc_content_module or "Module:documentation") .. "]]. <sup>[[" ..
if displaying_messages then
mw.title.new("Module:documentation"):fullUrl { action = 'edit' } ..
msg:add("warning", "Gloss abbreviation " .. highlight(gloss2) .. " not recognised" .. help_link("gloss abbr"))
" edit]]</sup> </i></dd>")
elseif not nodoc then
output:insert(
"<dd><i style=\"font-size: larger;\">This " .. pagetype ..
" lacks a [[wikt: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")
end
end
msg:add("non-repeating error", "Unknown glossing abbreviation(s)" .. help_link("gloss abbr"))
gloss_node
:addClass(conf.class.GlossAbbrError)
:addClass("error")
:css("font-size", "100%")
:attr("title", gloss2 .. ": glossing abbreviation not found")
:attr("style", conf.style.ErrorMessage)
:wikitext(gloss)
end
end
return tostring(gloss_node)
if title.fullText:match("^MediaWiki:Gadget%-") then
end
local is_gadget = false
local gadget_list = mw.title.new("MediaWiki:Gadgets-definition"):getContent()
for line in mw.text.gsplit(gadget_list, "\n") do
local gadget, opts, items = line:match("^%*%s*([A-Za-z][A-Za-z0-9_%-]*)%[(.-)%]|(.+)$") -- opts is unused
if not gadget then
gadget, items = line:match("^%*%s*([A-Za-z][A-Za-z0-9_%-]*)|(.+)$")
end
if gadget then
items = Array(mw.text.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)
-- find_gloss() parses a word into morphemes, and it calls format_gloss()
output:insert("</code> gadget ([")
-- for anything that looks like a glossing abbreviation.
output:insert(tostring(mw.uri.fullUrl('MediaWiki:Gadgets-definition', 'action=edit')))
---------------------
output:insert(" edit definitions])'' <dl>")
local function find_gloss(word)
local function scan_gloss(boundary, gloss_abbr) -- checks a morpheme if it is a gloss abbreviation
output:insert("<dd> ''Description ([")
if (mw.ustring.match(gloss_abbr, conf.GlossAbbrPattern)
output:insert(tostring(mw.uri.fullUrl('MediaWiki:Gadget-' .. gadget, 'action=edit')))
or conf.LowerCaseGlosses[gloss_abbr])
output:insert(" edit])'': ")
and not (conf.GlossExcludeTable[gloss_abbr]
or mw.ustring.match(gloss_abbr, conf.GlossExcludePattern))
local gadget_description = mw.message.new('Gadget-' .. gadget):plain()
then gloss_abbr = format_gloss(gloss_abbr)
gadget_description = frame:preprocess(gadget_description)
end
output:insert(gadget_description)
return boundary .. gloss_abbr
output:insert(" </dd>")
end
local word = mw.text.decode(word, true)
if word == "I" -- for the case of the English word "I", the 1SG pronoun
then return word end
local pattern = "([" .. conf.GlossAbbrBoundary .. "]?)([^" .. conf.GlossAbbrBoundary .. "]+)"
word = mw.ustring.gsub(word, pattern, scan_gloss) -- splits into morphemes
return word
end


items:remove(i)
---------------------
if #items > 0 then
-- The main purpose of the bletcherous parse() is to split a line into words and and then for each eligible word
for j, item in ipairs(items) do
-- to call find_gloss(). The parser outputs the individual words (with any gloss abbreviation formatting applied).
items[j] = '[[MediaWiki:Gadget-' .. item .. '|' .. item .. ']]'
-- The simple job of splitting at whitespaces has been made complicated by a) the fact that the input can contain
end
-- whitespaces inside the various html elements that are the result of the application of various formatting templates;
output:insert("<dd> ''Other parts'': ")
-- and b) the need to be able to recognise the output of the template that formats custom gloss abbreviations
output:insert(mw.text.listToText(items))
-- (and hence skip passing it on to find_gloss). See talk for a suggestion about its future.
output:insert("</dd>")
---------------------
end
local function parse(cline, i, tags_found,ifglossing)


output:insert("</dl></dd>")
local function issue_error(message, culprit)
 
UserMessages:add("error",  message .. ": ''" .. mw.ustring.sub(cline.whole, 1, i-1) .. "'''" .. culprit  .. "'''''")
break
end
end
if i > cline.length then return i end --this will only be triggered if the current line has less words than line 1
end
local next_step, j, _, chunk
local probe = mw.ustring.sub(cline.whole,i,i)
if mw.ustring.match(probe,"[" .. conf.WordSeparator .. "]") and tags_found == 0
then next_step =  i-1
elseif probe == "[" then --Wikilink?
if mw.ustring.sub(cline.whole,i+1,i+1) == "[" then
_,j,chunk = mw.ustring.find(cline.whole,"(%[%[.-%]%])", i)
else chunk = "["; j = i end --not a wikilink then
buffer = buffer .. chunk
next_step =  parse(cline, j+1,tags_found,ifglossing)
elseif probe == "{"  and tags_found == 0 then --curly brackets enclose a sequence of words to be treated as a single unit
_,j,chunk = mw.ustring.find(cline.whole,"(.-)(})", i+1)
if not chunk then
issue_error("Unclosed curly bracket", "{")
chunk = highlight("{"); j = i
elseif ifglossing==true then
chunk = find_gloss(chunk)
else
if cline.tone_sup then chunk = tone_sup(chunk) end
end
buffer = buffer .. chunk
next_step =  parse(cline, j+1,tags_found,ifglossing)
elseif probe == "<" then -- We've encountered an HTML tag. What do we do now?
local _,j,chunk = mw.ustring.find(cline.whole,"(<.->)",i)
if not chunk then
issue_error("Unclosed angle bracket", "<")
chunk = highlight("<"); j = i
elseif mw.ustring.sub(cline.whole,i,i+1) == "</" then -- It's a CLOSING tag
if cline.glossing
and ifglossing==false
and mw.ustring.match(chunk,"</abbr>")
then ifglossing=true end
tags_found = tags_found - 1
elseif not mw.ustring.match(chunk, "/>$") -- It's an OPENING tag, unless it opens a self-closing element (in which case the element is ignored)
then if ifglossing == true -- the following checks for the output of {{ggl}}:
and mw.ustring.find(chunk, conf.class.GlossAbbr, 1, true) -- it's important that the "find" function uses literal strings and not patterns
then ifglossing = false end
tags_found = tags_found + 1
end
buffer = buffer .. chunk
next_step = parse(cline, j+1,tags_found,ifglossing)
else -- No HTML tags, so we only need to find where the word ends
local _,k,chunk = mw.ustring.find(cline.whole,"(..-)([ <[])",i)
if k then --ordinary text
if ifglossing==true then
buffer = buffer .. find_gloss(chunk)
else
if cline.tone_sup then chunk = tone_sup(chunk) end
buffer = buffer .. chunk
end
next_step = parse(cline, k, tags_found, ifglossing)
else -- reached end of string
if ifglossing == true then
chunk = find_gloss(mw.ustring.sub(cline.whole,i))
else
chunk = mw.ustring.sub(cline.whole,i)
if cline.tone_sup then chunk = tone_sup(chunk) end
end
end
end
buffer = buffer .. chunk
next_step = cline.length
if not is_gadget then
output:insert("<dd> ''This script is not a part of any [")
output:insert(tostring(mw.uri.fullUrl('Special:Gadgets', 'uselang=en')))
output:insert(' gadget] ([')
output:insert(tostring(mw.uri.fullUrl('MediaWiki:Gadgets-definition', 'action=edit')))
output:insert(' edit definitions]).</dd>')
-- else
-- cats:insert("Wiktionary gadgets")
end
end
end
end
return next_step
if old_doc_title then
end
output:insert("<dd> ''Redirected from'' [")
--------------------
output:insert(old_doc_title:fullUrl { redirect = 'no' })
-- The following function is called by Template:gcl and is used for formatting an individual glossing abbreviation
output:insert(" ")
--------------------
output:insert(old_doc_title.fullText)
function p.gcl(frame)
output:insert("] ([")
local args = getArgs(frame,{
output:insert(old_doc_title:fullUrl { action = 'edit' })
trim = true,
output:insert(" edit]).</dd>\n")
removeBlanks = false,
parentOnly = true,
wrappers = {'Template:Gcl'},
})
msg = UserMessages
set_global_glossing_settings{style = args.style, underline = args.underline, small_caps = args['small-caps']}
if not args.glossing then
glossing_type = conf.GlossingType -- a global variable
else glossing_type = set_glossing_type(args.glossing)
end
end
local gloss, label, wikilink = args[1], args[2], args[3]
local links = Array()
if not gloss then UserMessages:add("error", "No gloss supplied")
return UserMessages:print() end
if title.isSubpage and not args.notsubpage then
if wikilink and not args.glossing then -- if a wikilink is supplied and glossing isn't set to 'label'...
links:insert("[[:" .. title.nsText .. ":" .. title.rootText .. "|root page]]")
glossing_type = 'wikilink' end --    .. then the wikilink will be formatted as such
links:insert("[[Special:PrefixIndex/" .. title.nsText .. ":" .. title.rootText .. "/|root page’s subpages]]")
if label == "" then label = nil end
else
if wikilink == "" then wikilink = nil end
links:insert("[[Special:PrefixIndex/" .. title.fullText .. "/|subpage list]]")
local result = format_gloss(gloss, label, wikilink)
return result
end
 
--------------------
-- The following is the function called by Template:Interlinear.
-- It processes the template arguments, then calls parse() to split the input lines into words
-- and it then builds the output html.
--------------------
function p.interlinearise(frame)
---------------------
-- Prepare arguments
---------------------
local if_auto_translit = false
local args = getArgs(frame, { -- configuration for Module:Arguments
trim = true,
removeBlanks = false,
parentFirst = true,
wrappers = {'Template:Interlinear', 'Template:Fs interlinear'},
})
local template_name = frame:getParent():getTitle()
if template_name == 'Template:Fs interlinear' then
args.italics1 = args.italics1 or "no"
args.italics2 = args.italics2 or "yes"
args.glossing3 = args.glossing3 or "yes"
if args.lang and not args.lang2 then args.lang2 = args.lang .."-Latn" end
if args.transl and not args.transl2 then args.transl2 = args.transl end
if_auto_translit = true
 
end
local revid = frame:preprocess( "{{REVISIONID}}" )
if  revid == "" then
if not args['display-messages'] or yesno(args['display-messages']) then
displaying_messages = true end-- messages will be displayed only in preview mode
end
end
msg = UserMessages
links:insert(
local line = {}
'[' .. tostring(mw.uri.fullUrl('Special:WhatLinksHere/' .. title.fullText,
'hidetrans=1&hideredirs=1')) .. ' links]')


if contentModel ~= "Scribunto" then
local function set_italics(n)
links:insert(
line[n].attr.style = line[n].attr.style .. "font-style: italic;"
'[' .. tostring(mw.uri.fullUrl('Special:WhatLinksHere/' .. title.fullText,
line[n].tone_sup = true -- single digits are assumed to be tone markers and will hence be superscripted
'hidelinks=1&hidetrans=1')) .. ' redirects]')
if args['tone-superscripting'] and not yesno(args['tone-superscripting'])
then line[n].tone_sup = false end
end
end


if (contentModel == "javascript") or (contentModel == "css") then
if args.glossing then -- the glossing= parameter sets the default glossing type
if user_name then
local _gl = set_glossing_type(args.glossing)
links:insert("[[Special:MyPage" .. title.text:sub(#title.rootText + 1) .. "|your own]]")
if _gl then conf.GlossingType = _gl end
end
end
else
--this looks for a list of glossing abbreviations on the page that transcludes the template:
links:insert(
local _ablist_section = get_section(frame, 'list-of-glossing-abbreviations')
'[' .. tostring(mw.uri.fullUrl('Special:WhatLinksHere/' .. title.fullText,
if _ablist_section and _ablist_section ~= "" then
'hidelinks=1&hideredirs=1')) .. ' transclusions]')
local _a = mw.ustring.gsub(_ablist_section, '</?div [^\n]*>', '') -- strips off the div tags
set_custom_glosses(_a)
end
--and this looks looks for a list of abbreviations set within the template:
local _ablist = args.abbreviations
if _ablist and _ablist ~= ""
then set_custom_glosses(_ablist) end
local _ablist = args.ablist
if _ablist and _ablist ~= ""
then set_custom_glosses(_ablist) end
 
local _spacing = tonumber(args.spacing)
if _spacing and _spacing <= 20
then conf.style.WordDiv = conf.style.WordDiv .. 'margin-right: ' .. _spacing .. 'em;'
else conf.style.WordDiv = conf.style.WordDiv .. conf.style.WordMargin
end
end
 
if contentModel == "Scribunto" then
local offset, last_line = 0, 0
local is_testcases = title.isSubpage and title.subpageText == "testcases"
for j,v in ipairs(args) do -- iterates over the unnamed parameters from the template
local without_subpage = title.nsText .. ":" .. title.baseText
last_line = last_line +1
if is_testcases then
if is_empty(v)
links:insert("[[:" .. without_subpage .. "|tested module]]")
then offset = offset + 1
else
else
links:insert("[[" .. title.fullText .. "/testcases|testcases]]")
local i = j - offset
line[i] = {}
v = normalise(v)
 
-- the following is part of a trial implementation of automatic transliteration:
if if_auto_translit and v == "auto" and i > 1 then
local source_line = line[i-1]
local src_lang = source_line.lang
if not src_lang then src_lang = args.lang end
if src_lang then
v = transliterate(source_line.whole, src_lang)
else v = ""; msg:add("error", "No language specified for automatic transliteration")
end
end  -- end of trial block
 
line[i].whole = v
line[i].length = mw.ustring.len(v)
 
local _c = args["c" .. i]
if _c and _c ~= "" then
line.hasComments = true
line[i].c = _c
end
end
 
if user_name then
---prepare style arguments----
links:insert("[[User:" .. user_name .. "|user page]]")
line[i].class = ""
links:insert("[[User talk:" .. user_name .. "|user talk page]]")
local _style = args["style" .. i]
links:insert("[[Special:PrefixIndex/User:" .. user_name .. "/|userspace]]")
if not _style then _style = ""
else
else _style = tidyCss(_style) end
-- If sandbox module, add a link to the module that this is a sandbox of.
--line[i].attr holds the attributes for the <p> elements that enclose the words in line i
-- Exclude user sandbox modules like [[User:Dine2016/sandbox]].
line[i].attr = {style = conf.style.WordP .. _style}
if title.text:find("/sandbox%d*%f[/%z]") then
 
cats:insert("Sandbox modules")
local _lang = args["lang" .. i]
if _lang and #_lang > 1 then
-- Sandbox modules don’t really need documentation.
line[i].lang = _lang
needs_doc = false
else _lang = args.lang
if _lang and #_lang > 1 and i == 1 then -- if a lang= parameter is supplied, it's assumed to apply to line 1
-- Will behave badly if “/sandbox” occurs twice in title!
line[i].lang = _lang
local sandbox_of = title.fullText:gsub("/sandbox%d*%f[/%z]", "")
end
end
local diff
line[i].attr.lang = line[i].lang
if page_exists(sandbox_of) then
--the following emulates the behaviour of {{Bo-textonly}} (see Template talk:Fs interlinear#Tibetan):
diff = " (" .. compare_pages(title.fullText, sandbox_of, "diff") .. ")"
if template_name == 'Template:Fs interlinear' then
if _lang == "bo" and i == 1 then
line[1].class = line[1].class .. " uchen"
line[1].attr.style = line[1].attr.style .. "font-size:1.25em; word-wrap:break-word;"
end
end
 
if yesno(args["italics" .. i]) then
set_italics(i)
end
 
local _transl = args["transl" .. i]
if _transl and #_transl > 1 then
_transl = mw.ustring.lower(_transl)
local _lookup = lang_data.translit_title_table[_transl]
if _lookup then
if _lang and  _lookup[_lang] then
_transl = _lookup[_lang]
else _transl = _lookup.default
end
end
if _transl then
links:insert("[[:" .. sandbox_of .. "|sandbox of]]" .. (diff or ""))
line[i].attr.title = _transl
-- If not a sandbox module, add link to sandbox module.
-- Sometimes there are multiple sandboxes for a single module:
-- [[Module:sa-pronunc/sandbox]],  [[Module:sa-pronunc/sandbox2]].
-- Occasionally sandbox modules have their own subpages that are also
-- sandboxes: [[Module:grc-decl/sandbox/decl]].
else
local sandbox_title
if title.fullText:find("^Module:grc%-decl/") then
sandbox_title = title.fullText:gsub("^Module:grc%-decl/", "Module:grc-decl/sandbox/")
elseif is_testcases then
sandbox_title = title.fullText:gsub("/testcases", "/sandbox/testcases")
else
sandbox_title = title.fullText .. "/sandbox"
end
end
local sandbox_link = "[[:" .. sandbox_title .. "|sandbox]]"
else  msg:add("error", "Transliteration scheme '" .. _transl .. "' not recognised")
end
local diff
end
if page_exists(sandbox_title) then
 
diff = " (" .. compare_pages(title.fullText, sandbox_title, "diff") .. ")"
local _glossing = args["glossing" .. i]
end
if _glossing then
line[i].glossing = set_glossing_type(_glossing)
links:insert(sandbox_link .. (diff or ""))
-- Do not treat default glossing settings as custom.
if not ((i == 1 and not yesno(_glossing)) or (i == 2 and yesno(_glossing))) then
line.HasCustomGlossing = true
end
end
end
end
local _ipa = args['ipa' .. i]
if yesno(_ipa) then
line[i].class = "IPA"
end
local _class = args['class' .. i]
if _class then
line[i].class = line[i].class .. " " .. _class
end
if line[i].class == ""
then line[i].class = nil end
end -- ends the first if-statement in the loop
end -- ends the FOR cycle
local line_count = #line
if line_count == 0 then
msg:add("error", template_name .. ": no lines supplied.")
return msg:print_errors()
end
end
 
if title.nsText == "Template" then
if line_count > 1 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]]
local _italics = args.italics
-- apparently same with/without: &profile=advanced&fulltext=1
local n = tonumber(_italics)
local errorq = 'searchengineselect=mediawiki&search=all: hastemplate:\"'..title.rootText..'\" insource:\"'..title.rootText..'\" incategory:'
if n and n > 0 then
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"
set_italics(n)
elseif not (_italics and not yesno(_italics)) and not (args["italics1"] and not yesno(args["italics1"])) then
links:insert(
set_italics(1) -- by default, the first line will get italicised, unless italics=no or italics1=no
'[' .. tostring(mw.uri.fullUrl('Special:Search', errorq..eincategory )) .. ' errors]'
.. ' (' ..
'[' .. tostring(mw.uri.fullUrl('Special:Search', errorq..'ParserFunction_errors' )) .. ' parser]'
.. '/' ..
'[' .. tostring(mw.uri.fullUrl('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 page_exists(sandbox_of) then
diff = " (" .. compare_pages(title.fullText, sandbox_of, "diff") .. ")"
end
links:insert("[[:" .. sandbox_of .. "|sandbox of]]" .. (diff or ""))
else -- This is a template that can have a sandbox.
local sandbox_title = title.fullText .. "/sandbox"
local diff
if page_exists(sandbox_title) then
diff = " (" .. compare_pages(title.fullText, sandbox_title, "diff") .. ")"
end
links:insert("[[:" .. sandbox_title .. "|sandbox]]" .. (diff or ""))
end
end
-- the last unnamed parameter is assumed to be the free translation:
free_translation = args[last_line]
if not is_empty(free_translation) then
line [line_count] = nil  end  --... and is thus excluded from interlinearising
end
end
 
if #links > 0 then
-- If glossing isn't specified for any line, then it's chosen by default to occur
output:insert("<dd> ''Useful links'': " .. links:concat(" • ") .. "</dd>")
-- in the second line, unless only a single line has been supplied, in which case
-- the assumption is that it is the one containing grammatical glosses
if yesno(args.glossing) == false then
line.HasCustomGlossing = true
end
end
if not line.HasCustomGlossing then
output:insert("</dl>\n")
if line_count == 1 then
line[1].glossing = conf.GlossingType
-- Show error from [[Module:category tree/topic cat/data]] on its submodules'
elseif line[2] then
-- documentation to, for instance, warn about duplicate labels.
line[2].glossing = conf.GlossingType
if title.fullText:find("Module:category tree/topic cat/data", 1, true) == 1 then
local ok, err = pcall(require, "Module:category tree/topic cat/data")
if not ok then
output:insert('<span class="error">' .. err .. '</span>\n\n')
end
end
end
end
set_global_glossing_settings{style = args['glossing-style'], underline = args.underline, small_caps = args['small-caps']}
if doc_title.exists then
-- Override automatic documentation, if present.
doc_content = frame:expandTemplate { title = doc_title.fullText }
elseif not doc_content and fallback_docs then
doc_content = frame:expandTemplate {
title = fallback_docs;
args = {
['user'] = user_name;
['page'] = title.fullText;
['skin name'] = skin_name;
}
}
end


if doc_content then
---------------------
output:insert(doc_content)
-- Segment lines into words
---------------------
for i,v in ipairs(line) do
local ifglossing = false
if line[i].glossing then
ifglossing = true -- if true the parser will attempt to format gloss abbreviations in the current line
glossing_type = line[i].glossing -- neccessarily a global variable
end
local wc, n = 1, 1
line[i].words = {}
while n <= line[i].length do
buffer = ""
n = parse(line[i], n, 0, ifglossing)+2
line[i].words[wc] = buffer
wc = wc + 1
end
end
end


output:insert(('\n<%s style="clear: both;" />'):format(args.hr == "below" and "hr" or "br"))
----Check for mismatches in number of words across lines----
local number_of_words, mismatch_found = 0, false
if not cats[1] and not doc_content then
for i,v in ipairs(line) do -- find the maximum number of words in any line
if contentModel == "Scribunto" then
local wc = #line[i].words
cats:insert("Uncategorized modules")
if wc ~= number_of_words then
-- elseif title.nsText == "Template" then
if i ~= 1 and wc ~= 0 then
-- cats:insert("Uncategorized templates")
mismatch_found = true
end
if wc > number_of_words then
number_of_words = wc
end
end
end
end
end
----Deal with mismatches---
if needs_doc then
if mismatch_found then
cats:insert("Templates and modules needing documentation")
local error_text = "Mismatch in the number of words between lines: "
for i,v in ipairs(line) do
local wc = #line[i].words
error_text = error_text .. wc .. " word(s) in line " .. i .. ", "
if wc ~= number_of_words then
for current_word = wc+1, number_of_words do
line[i].words[current_word] = "&nbsp;"
end
end
end
if string.sub(error_text, -2) == ", "
then error_text = string.sub(error_text, 1, #error_text - 2) .. " "
end
error_text = error_text .. help_link("mismatch")
UserMessages:add("error", error_text)
end
end
 
for _, cat in ipairs(cats) do
---------------------
output:insert("[[Category:" .. cat .. "]]")
-- Build the HTML
---------------------
---- If just a single line was supplied, format it as inline text
if line_count == 1 then
local span = mw.html.create('span')
span:attr(line[1].attr)
for wi = 1, number_of_words do
local space
if wi < number_of_words then space = " " else space = "" end
span:wikitext(line[1].words[wi] .. space)
end
return tostring(span)
end
end
output:insert("</div>\n")


return output:concat()
---- More than one line supplied, so we'll produce interlinear display
end
local div = mw.html.create("div")
div:addClass(conf.class.Interlinear)


-- Used by {{translit module documentation}}.
-- For stuff to be displayed in the left margin, like example numbering
function export.translitModuleLangList(frame)
local number, indent = nil, nil
local pagename, subpage
if args.number and args.number ~= ""
then number = args.number end
if frame.args[1] then
if args.indent and args.indent ~=""
pagename = frame.args[1]
then indent = args.indent end
else
if indent or number then
local title = mw.title.getCurrentTitle()
if not indent then indent = "4" end --default value
subpage = title.subpageText
div:css("margin-left", indent .. 'em')
pagename = title.text
if number then
div:tag("div")
if subpage ~= pagename then
:css("position", "absolute")
pagename = title.rootText
:css("left", "1em")
:wikitext(args.number)
end
end
end
end
 
local translitModule = pagename
if args.box and args.box ~= "" then
div:css("background-color", "#f8f9fa")
local languageObjects = require("Module:languages/byTranslitModule")(translitModule)
:css("border", "1px solid #eaecf0")
local codeInPagename = pagename:match("^([%l-]+)%-.*translit$")
:css("padding", "1em") end
if args.top and args.top ~= "" then --lines to display above the interlinear block
local categories = Array()
div:tag("div")
local codeInPagenameInList = false
:wikitext(args.top)
if codeInPagename then
end
if languageObjects[1] and subpage ~= "documentation" then
 
local agreement = languageObjects[2] and "s" or ""
-- Producing the interlinear block
categories:insert("[[Category:Transliteration modules used by " ..
for wi = 1, number_of_words do
#languageObjects .. " language" .. agreement .. "]]")
local div2 = div:tag("div")
:attr("style", conf.style.WordDiv)
for i,_ in ipairs (line) do
if line[i].whole ~= "" then -- skipping empty lines
local p = div2:tag("p")
p:attr(line[i].attr)
if line[i].class then
p:addClass(line[i].class)
end
local _text = line[i].words[wi]
if _text == "" or _text == " "
then _text = "&nbsp;" end -- <p> elements without content mess up the interlinear display
p:wikitext(_text)
end
end
end
languageObjects = Array(languageObjects)
:filter(
function (lang)
local result = lang:getCode() ~= codeInPagename
codeInPagenameInList = codeInPagenameInList or result
return result
end)
end
end
 
if subpage ~= "documentation" then
--- If any "comments" have been specified, add them at the end of each line
for script_code in pagename:gmatch("%f[^-%z]%u%l%l%l%f[-]") do
if line.hasComments then
local script = require "Module:scripts".getByCode(script_code)
local divc = div:tag("div")
if script then
:attr("style", conf.style.WordDiv)
categories:insert("[[Category:" .. script:getCategoryName() .. "]]")
for i,_ in ipairs (line) do
local p = divc:tag("p")
p:attr("style", conf.style.WordP)
if line[i].c then
p:wikitext(line[i].c)
else p:wikitext("&nbsp;")
end
end
end
end
end
end
 
if subpage ~= "documentation" and not page_exists("Module:" .. pagename .. "/testcases") then
--Add hidden lines containing the content of each line of interlinear text: this is for accessibility
categories:insert("[[Category:Transliteration modules without a testcases subpage]]")
for i,v in ipairs(line) do
local hidden_line = div:tag("p")
hidden_line:attr("style", conf.style.HiddenText)
:wikitext(v.whole)
end
 
-- Format the free translation
local ft_line = div:tag("p")
if free_translation and free_translation ~= "" then
ft_line:attr("style", "clear: left;")
ft_line:wikitext(free_translation)
end
if args.bottom and args.bottom ~= ""
then local bottom = div:tag('p')
bottom:css('margin-top', '0')
bottom:wikitext(args.bottom)
end
ft_line:node(msg:print_errors()) -- for error messages
 
local end_div = div:tag("div")
end_div:attr("style", conf.style.EndDiv)
div:newline()
local temp_track = ""
if last_line == 2
then temp_track = "[[Category:Pages with interlinear glosses using two unnamed parameters]]"
end
end
if last_line > 3 and template_name ~= 'Template:Fs interlinear'
if not languageObjects[1] then
then  temp_track = "[[Category:Pages with interlinear glosses using more than three unnamed parameters]]"
return categories:concat()
end
end
return tostring(div) .. temp_track .. msg:print_warnings()
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)
:serial_comma_join()
return "It is " .. ( codeInPagenameInList and "also" or "" ) ..
" used to transliterate " .. langs .. "." .. categories:concat()
end
end


return export
return p

Revision as of 18:34, 5 April 2023



local p = {}
local data = mw.loadData( 'Module:Interlinear/data' )
local gloss_override = {} -- for custom gloss abbreviations
local getArgs = require('Module:Arguments').getArgs
local yesno = require('Module:Yesno')
local lang_data = mw.loadData( 'Module:Lang/data' )

--------------------------
-- Almost-global variables
--------------------------
local glossing_type, displaying_messages, free_translation, msg, buffer

-------------------
-- General settings
-------------------
local conf = { --settings
	WordSeparator = " \n\r\t", -- Don't replace with %s as this would include non-breaking spaces
	GlossAbbrPattern = "^([Ø0-9A-Z]+)$", -- this isn't a full regex, but a Lua pattern
	-- NOTE: The following characters must be formatted for use in a pattern set.
	GlossAbbrBoundary = "-.,;:<>‹›/\\~+=%?%s%[%]()%_\127'",
	GlossExcludeTable = {I = true,}, --strings not be treated as glossing abbreviations
	GlossExcludePattern = '^[0-9][0-9]+$', -- excludes strings consisting entirely of digits
	GlossSmallCapsExclude = "^[AOPS]$", -- glossing abbreviations matching this pattern will not be rendered in small caps
	GlossingType = "label", -- if set to "label" gloss abbreviations are formatted as an <abbr> with the "label" appearing in a tooltip
						-- if set to "wikilink" the abbreviation is formatted as a wikilink to the relevant wikipedia article
						-- if set to "none" abbreviations aren't formatted at all
	ErrorCategory = "[[Category:Pages with errors in interlinear text]]",
	AmbiguousGlossCategory = "[[Category:Articles with ambiguous glossing abbreviations]]",
	MessageGlossingError = "Error(s) in interlinear glossing",
	combining_gender_numbers = "[0-9][0-9]?$", --e.g. G4 '4th gender' or CL7 'class 7'
	combining_gender_prefixes = {G = "gender", CL = "class"},
	combining_person = {["1"] = "first person", ["2"] = "second person", ["3"] = "third person"},
	combining_number = {S = "singular", SG = "singular", P = "plural", PL = "plural", D = "dual", DU = "dual", TRI = "trial"},
	combining_gender = {F = "feminine", M = "masculine", N = "neuter"},
	LowerCaseGlosses = {["1sg"] = true, ["2sg"] = true, ["3sg"] = true, ["1du"] = true, ["2du"] = true, ["3du"] = true, ["1pl"] = true, ["2pl"] = true,
		["3pl"] = true, ["Fsg"] = true, ["Fpl"] = true, ["Msg"] = true, ["Mpl"] = true,}, -- these are the non-all-upper-case strings that will be recognised as glossing abbreviations
	ErrorHelpLocation = "Template:Interlinear",
}

---------------------
-- CSS styles and classes
---------------------
conf.style = { --CSS styles
	WordDiv = "float: left; margin-bottom: 0.3em;",
	WordMargin = "margin-right: 1em;",
	WordP = "margin: 0px;", -- the style for the word <p> elements
	GlossAbbr = "font-variant: small-caps; font-variant-numeric: oldstyle-nums; text-transform: lowercase; ", -- won't be applied to gloss abbreviations containing lower-case characters
	HiddenText = "display: none;",
	EndDiv = "clear: left; display: block;", -- style of the <div> element at the end of the interlinear display
	ErrorMessage = "font-size: inherit",
}
conf.class = { --CSS classes
	Interlinear = "interlinear",
	GlossAbbr  = "gloss-abbr",
	GlossAbbrAmb = "gloss-abbr-ambiguous",
	GlossAbbrError = "gloss-abbr-error",
	ErrorMessage = "error",
}
---------------------
-- Section transclusion
---------------------
local page_content = nil -- lazy initilization
local function get_section(frame, section_name)
	if page_content == nil then
		local current_title = mw.title.getCurrentTitle()
		page_content = current_title:getContent()
	end
	if page_content then
		if mw.ustring.find(page_content, section_name, 1, true) then
			return frame:preprocess('{{#section:{{FULLPAGENAME}}|' .. section_name .. '}}')
		end
	end
	return ''
end
---------------------
-- Sundry small functions
---------------------
local function normalise(str)
	return mw.ustring.gsub(str,"[" .. conf.WordSeparator .. "]+"," ")
end

local function tidyCss(str)
	str = mw.ustring.gsub(str, '^[\"\']*(.-)[\"\']*$', "%1") -- trims quotation marks
	if mw.ustring.sub(str, -1) ~= ";" then str = str .. ";" end -- appends ";" if missing
	return str
end

local function highlight(text)
	if text then
		return '<span style="color:#C00;font-weight:bold;">' .. text .. '</span>'
	else return "" end
end

local function tone_sup(str)
	return mw.ustring.gsub(str, "([^%p%s0-9])([0-9]+)", "%1<sup>%2</sup>")
end

local function is_empty(str) -- returns "false" if its argument is a string containing chars other than spaces &c.
	if not str then return true end
	if mw.ustring.find(str, "[^" .. conf.WordSeparator .. "]")
		then return false
	else return true end
end

local function help_link (anchor)
	if anchor then
		return " ([[" .. conf.ErrorHelpLocation .. "#" .. anchor .. "|help]])"
	else return "" end
end

-- the following is part of a trial implementation of automatic transliteration:
local function transliterate (str, lang_from, lang_to,  scheme)
	local lookup = {grc = {module = 'Module:Ancient Greek', funct = "transliterate", } }
	if not lang_from then
		msg:add("error", "Source language for transliteration is not set")
	else
		local t = lookup[lang_from]
		if t then
			local module = require(t.module)
			return module[t.funct](str)
		else msg:add("error", "Can't find transliterator for language '" .. lang_from .. "'")
		end
	end
	return ""
end -- end of trial block

--------------------
-- The following two functions update the glossing settings based on the received
-- template arguments. set_global_glossing_settings() updates the global settings
-- that are valid for all gloss abbreviations. set_glossing_type()
-- returns the glossing type, which can vary between the different lines.
--------------------
local function set_global_glossing_settings(a)
	local style = ""
	if a.style then style = tidyCss(a.style) end
	if a.underline == "no" then
		style = style .. "text-decoration: none;" end
	if a.small_caps == "no" then
		style = style .. "font-variant:normal; text-transform: none;" end
	if style ~= "" then conf.style.GlossAbbr = conf.style.GlossAbbr .. style end
end

local function set_glossing_type(glossing)
	if glossing then
		local GlossingType
		glossing = mw.ustring.lower(mw.text.trim(glossing))
		if mw.ustring.find(glossing, 'link') then
			GlossingType = "wikilink"
		elseif mw.ustring.find(glossing, 'label')
			or  mw.ustring.find(glossing, 'no link') then
			GlossingType = 'label'
		elseif mw.ustring.find(glossing, 'no abbr') then
			GlossingType = "no abbr"
		elseif yesno(glossing) == false then
			GlossingType = nil
		elseif yesno(glossing) then
			GlossingType = conf.GlossingType
		else
			msg:add('error', 'Glossing type "' .. glossing .. '" not recognised') end
		return GlossingType
	else error("set_glossing_type: 'glossing' is nil or false", 2)
	end
end

local function set_custom_glosses(list)
	local abbs = mw.text.split(list, '[;\n\t]')
	for _,v in pairs(abbs) do
		local gloss = mw.text.split(v, ':')
		local a = mw.text.trim(gloss[1])
		if a and a ~= "" then
			gloss_override[a] = {}
			gloss_override[a].expansion = gloss[2]
			gloss_override[a].wikipage = gloss[3]
		end
	end
end

---------------------
-- The UserMessages object contains and processes error messages and warnings
---------------------
local UserMessages = {errors = {}, warnings = {}, gloss_messages = {}}
function UserMessages:add(msgtype, text, gloss)
	if msgtype == "gloss_message" then
		self.gloss_messages[gloss] = text
	elseif msgtype == "warning" then
		table.insert(self.warnings, text)
	elseif msgtype == "non-repeating error" then
		self.errors.nre = text
	elseif msgtype == "ambiguous gloss" then
		self.if_ambiguous_glosses = true
	elseif msgtype == "error" then
		table.insert(self.errors, text)
	else return error("UserMessages:add(): unknown message type", 2)
	end
end
function UserMessages:print_errors()
	local out = ""
	local namespace = mw.title.getCurrentTitle().namespace
	if next(self.errors) or self.warnings[1] then
		local err_span = mw.html.create("span")
		err_span:attr("style", conf.style.ErrorMessage)
		err_span:addClass(conf.class.ErrorMessage)
		for _,v in pairs(self.errors) do
			err_span:wikitext(" " .. v .. ";") end
		if namespace % 2 == 0 and namespace ~= 2 -- non-talk namespaces, excluding user pages; if modifying please update the description on the category page
			then err_span:wikitext(conf.ErrorCategory)
		end
		out = tostring(err_span)
		mw.addWarning(conf.MessageGlossingError)
	end
	if self.if_ambiguous_glosses then
		if namespace == 0 -- article namespace
			then out = out .. conf.AmbiguousGlossCategory -- this category will only track articles
		end
	end
	return out
end
function UserMessages:print_warnings()
	local out = ""
	-- Messages and warnings get displayed only if the page is being viewed in "preview" mode:
	if displaying_messages and (next(self.gloss_messages) or next(self.warnings)) then
		local div = mw.html.create("div")
		div:addClass("interlinear-preview-warning")
			:cssText('border: 1px solid #a2a9b1; background-color: #f8f9fa; width: 80%; padding: 0.2em;')
			:wikitext("<i>This message box is shown only in preview:</i>")
			:newline()
		for _,v in ipairs(self.warnings) do
			local p = div:tag("p")
			p:addClass(conf.class.ErrorMessage)
			p:attr("style", conf.style.ErrorMessage)
			p:wikitext(v)
		end
		if self.gloss_messages then
			div:wikitext("<p>  To change any of the following default expansions, see [[Template:Interlinear/doc#Custom abbreviations|the template's documentation]]:</p>")
			end
		for _,v in pairs(self.gloss_messages) do
			div:wikitext("<p>" .. v .. "</p>")
		end
		out = out .. "\n\n" .. tostring(div)
	end
	return out
end

---------------------
-- gloss_lookup() receives a gloss abbreviation and tries to uncover its meaning.
---------------------
local function gloss_lookup(a, label, wikilink)
	local _label, _wikilink, _lookup, source = nil, nil, nil, nil
	if gloss_override[a] then
		_lookup = gloss_override[a]
		source = "local"
	elseif data.abbreviations[a] then _lookup = data.abbreviations[a] end
	if _lookup and _lookup.expansion ~= "" then
		_label, _wikilink = _lookup.expansion, _lookup.wikipage
	else
		local prefix = mw.ustring.sub(a,1,1)
		local suffix = mw.ustring.sub(a,2)
		if conf.combining_person[prefix] then -- is it of the form 1PL or 3FS?
			_label = conf.combining_person[prefix]
		local _suffix = conf.combining_number[suffix] or conf.combining_gender[suffix]
			if _suffix then
				_label = _label .. ", " .. _suffix
			else
				local suffix1 = mw.ustring.sub(suffix,1,1)
				local suffix2 = mw.ustring.sub(suffix,2)
					if conf.combining_gender[suffix1]
					and  conf.combining_number[suffix2] then
						_label = _label .. ", " .. conf.combining_gender[suffix1] .. ", " .. conf.combining_number[suffix2]
					else _label = nil end
			end
	elseif mw.ustring.match(suffix,conf.combining_gender_numbers) then -- cases like G4 = gender 4
		local _i,_j = mw.ustring.find(a, conf.combining_gender_numbers)
		local _pre = mw.ustring.sub(a, 1, _i - 1)
		local _suff = mw.ustring.sub(a, _i)
		if conf.combining_gender_prefixes[_pre] then
			_label = conf.combining_gender_prefixes[_pre] .. " " .. _suff
		end
	elseif prefix == "N" then -- dealing with cases like NPST = non-past
		local s = gloss_override[suffix] or data.abbreviations[suffix]
			if s ~= nil and not s.ExcludeNegation then
				_label = "non-" .. s.expansion
				_wikilink = s.wikipage
			end
			s = nil
		end
	end
	if _label == "" then _label = nil end
	if _wikilink == "" then _wikilink = nil end
	if not label then label = _label end
	if not wikilink then wikilink = _wikilink end
	return label, wikilink, source
end

---------------------
-- format_gloss() calls gloss_lookup() to find the meaning of a gloss
-- abbreviation, which it then proceeds to format
---------------------
local function format_gloss(gloss, label, wikilink)
	if string.sub(gloss,1,3) == "000" then -- checks for a common component of exposed strip markers (see [[:mw:Strip marker]])
		return gloss
	end
	local gloss2 = mw.ustring.gsub(gloss,"<.->","") -- remove any html fluff
	gloss2 = mw.ustring.gsub(gloss2, "%'%'+", "") -- remove wiki bold/italic formatting
	gloss2 = mw.text.trim(mw.ustring.upper(gloss2))
	if not (label or wikilink)
		or (not label and glossing_type == "label")
		or (not wikilink  and glossing_type == "wikilink")
		then
			if glossing_type ~= "no abbr"
				then label, wikilink, source = gloss_lookup(gloss2, label, wikilink)
			end
	end
	local gloss_node
	if glossing_type == "no abbr"
		then gloss_node = mw.html.create("span")
	else gloss_node = mw.html.create("abbr") end
	gloss_node:addClass(conf.class.GlossAbbr)
	if label or wikilink then
		if not mw.ustring.match(gloss, "%l") -- excluding glosses that contain lower-case characters
			and not mw.ustring.match(gloss,conf.GlossSmallCapsExclude) -- and also excluding A, O etc. from rendering in small caps
			then gloss_node:attr("style", conf.style.GlossAbbr)
		end
		local abbr_label
		if label then abbr_label = label
			else abbr_label = wikilink end
		gloss_node:attr("title", abbr_label)
		if source ~= "local" and data.abbreviations[gloss2] then
			if data.abbreviations[gloss2].ambiguous then
				gloss_node:addClass(conf.class.GlossAbbrAmb)
					msg:add("ambiguous gloss")
				end
		end
		if glossing_type == "wikilink" and wikilink
			then gloss_node:wikitext("[[", wikilink, "|" , gloss, "]]")
			else gloss_node:wikitext(gloss) end
		if source ~= "local" and displaying_messages then -- logging gloss lookups:
			local message = ""
			if label then
				message = "assuming " .. gloss2 .. " means \"" .. abbr_label .. "\";" end
			if glossing_type == "wikilink" and wikilink then
				message = message .. " linking to [[" .. wikilink .. "]];"
			end
			msg:add("gloss_message", message, gloss)
		end
	elseif glossing_type == "no abbr"
		then gloss_node
				:attr("style", conf.style.GlossAbbr)
				:wikitext(gloss)
	else
		if displaying_messages then
			msg:add("warning", "Gloss abbreviation " .. highlight(gloss2) .. "  not recognised" .. help_link("gloss abbr"))
		end
		msg:add("non-repeating error", "Unknown glossing abbreviation(s)" .. help_link("gloss abbr"))
		gloss_node
			:addClass(conf.class.GlossAbbrError)
			:addClass("error")
			:css("font-size", "100%")
			:attr("title", gloss2 .. ": glossing abbreviation not found")
			:attr("style", conf.style.ErrorMessage)
			:wikitext(gloss)
	end
	return tostring(gloss_node)
end

---------------------
-- find_gloss() parses a word into morphemes, and it calls format_gloss()
-- for anything that looks like a glossing abbreviation.
---------------------
local function find_gloss(word)
	local function scan_gloss(boundary, gloss_abbr) -- checks a morpheme if it is a gloss abbreviation
		if (mw.ustring.match(gloss_abbr, conf.GlossAbbrPattern)
			or conf.LowerCaseGlosses[gloss_abbr])
			and not (conf.GlossExcludeTable[gloss_abbr]
				or mw.ustring.match(gloss_abbr, conf.GlossExcludePattern))
			then gloss_abbr = format_gloss(gloss_abbr)
		end
		return boundary .. gloss_abbr
	end
	local word = mw.text.decode(word, true)
	if word == "I" -- for the case of the English word "I", the 1SG pronoun
		then return word end
	local pattern = "([" .. conf.GlossAbbrBoundary .. "]?)([^" .. conf.GlossAbbrBoundary .. "]+)"
	word = mw.ustring.gsub(word, pattern, scan_gloss) -- splits into morphemes
	return word
end

---------------------
-- The main purpose of the bletcherous parse() is to split a line into words and and then for each eligible word
-- to call find_gloss(). The parser outputs the individual words (with any gloss abbreviation formatting applied).
-- The simple job of splitting at whitespaces has been made complicated by a) the fact that the input can contain
-- whitespaces inside the various html elements that are the result of the application of various formatting templates;
-- and b) the need to be able to recognise the output of the template that formats custom gloss abbreviations
-- (and hence skip passing it on to find_gloss). See talk for a suggestion about its future.
---------------------
local function parse(cline, i, tags_found,ifglossing)

	local function issue_error(message, culprit)
		UserMessages:add("error",  message .. ": ''" .. mw.ustring.sub(cline.whole, 1, i-1) .. "'''" .. culprit  .. "'''''")
	end
	if i > cline.length then return i end --this will only be triggered if the current line has less words than line 1
	local next_step, j, _, chunk
	local probe = mw.ustring.sub(cline.whole,i,i)
	if mw.ustring.match(probe,"[" .. conf.WordSeparator .. "]") and tags_found == 0
		then next_step =  i-1
	elseif probe == "[" then --Wikilink?
		if mw.ustring.sub(cline.whole,i+1,i+1) == "[" then
			_,j,chunk = mw.ustring.find(cline.whole,"(%[%[.-%]%])", i)
		else chunk = "["; j = i end --not a wikilink then
		buffer = buffer .. chunk
		next_step =  parse(cline, j+1,tags_found,ifglossing)
	elseif probe == "{"  and tags_found == 0 then --curly brackets enclose a sequence of words to be treated as a single unit
		_,j,chunk = mw.ustring.find(cline.whole,"(.-)(})", i+1)
		if not chunk then
			issue_error("Unclosed curly bracket", "{")
			chunk = highlight("{"); j = i
		elseif ifglossing==true then
			chunk = find_gloss(chunk)
		else
			if cline.tone_sup then chunk = tone_sup(chunk) end
		end
		buffer = buffer .. chunk
		next_step =  parse(cline, j+1,tags_found,ifglossing)
	elseif probe == "<" then -- We've encountered an HTML tag. What do we do now?
		local _,j,chunk = mw.ustring.find(cline.whole,"(<.->)",i)
		if not chunk then
			issue_error("Unclosed angle bracket", "<")
			chunk = highlight("<"); j = i
		elseif mw.ustring.sub(cline.whole,i,i+1) == "</" then -- It's a CLOSING tag
			if cline.glossing
				and ifglossing==false
				and mw.ustring.match(chunk,"</abbr>")
				then ifglossing=true end
			tags_found = tags_found - 1
		elseif not mw.ustring.match(chunk, "/>$") -- It's an OPENING tag, unless it opens a self-closing element (in which case the element is ignored)
			then if ifglossing == true -- the following checks for the output of {{ggl}}:
					and mw.ustring.find(chunk, conf.class.GlossAbbr, 1, true) -- it's important that the "find" function uses literal strings and not patterns
						then ifglossing = false end
			tags_found = tags_found + 1
		end
		buffer = buffer .. chunk
		next_step = parse(cline, j+1,tags_found,ifglossing)
	else -- No HTML tags, so we only need to find where the word ends
		local _,k,chunk = mw.ustring.find(cline.whole,"(..-)([ <[])",i)
		if k then --ordinary text
			if ifglossing==true then
				buffer = buffer .. find_gloss(chunk)
			else
				if cline.tone_sup then chunk = tone_sup(chunk) end
				buffer = buffer .. chunk
			end
			next_step = parse(cline, k, tags_found, ifglossing)
		else -- reached end of string
			if ifglossing == true then
				chunk = find_gloss(mw.ustring.sub(cline.whole,i))
			else
				chunk = mw.ustring.sub(cline.whole,i)
				if cline.tone_sup then chunk = tone_sup(chunk) end
			end
			buffer = buffer .. chunk
			next_step = cline.length
		end
	end
	return next_step
end
--------------------
-- The following function is called by Template:gcl and is used for formatting an individual glossing abbreviation
--------------------
function p.gcl(frame)
	local args = getArgs(frame,{
		trim = true,
		removeBlanks = false,
		parentOnly = true,
		wrappers = {'Template:Gcl'},
	})
	msg = UserMessages
	set_global_glossing_settings{style = args.style, underline = args.underline, small_caps = args['small-caps']}
	if not args.glossing then
		glossing_type = conf.GlossingType -- a global variable
	else glossing_type = set_glossing_type(args.glossing)
	end
	local gloss, label, wikilink = args[1], args[2], args[3]
	if not gloss then UserMessages:add("error", "No gloss supplied")
		return UserMessages:print() end
	if wikilink and not args.glossing then -- if a wikilink is supplied and glossing isn't set to 'label'...
		glossing_type = 'wikilink' end --     .. then the wikilink will be formatted as such
	if label == "" then label = nil end
	if wikilink == "" then wikilink = nil end
	local result = format_gloss(gloss, label, wikilink)
	return result
end

--------------------
-- The following is the function called by Template:Interlinear.
-- It processes the template arguments, then calls parse() to split the input lines into words
-- and it then builds the output html.
--------------------
function p.interlinearise(frame)
---------------------
-- Prepare arguments
---------------------
	local if_auto_translit = false
	local args = getArgs(frame, { -- configuration for Module:Arguments
		trim = true,
		removeBlanks = false,
		parentFirst = true,
		wrappers = {'Template:Interlinear', 'Template:Fs interlinear'},
	})
	local template_name = frame:getParent():getTitle()
	if template_name == 'Template:Fs interlinear' then
		args.italics1 = args.italics1 or "no"
		args.italics2 = args.italics2 or "yes"
		args.glossing3 = args.glossing3 or "yes"
		if args.lang and not args.lang2 then args.lang2 = args.lang .."-Latn" end
		if args.transl and not args.transl2 then args.transl2 = args.transl end
		if_auto_translit = true

	end
	local revid = frame:preprocess( "{{REVISIONID}}" )
	if  revid == "" then
		if not args['display-messages'] or yesno(args['display-messages']) then
		displaying_messages = true end-- messages will be displayed only in preview mode
	end
	msg = UserMessages
	local line = {}

	local function set_italics(n)
		line[n].attr.style = line[n].attr.style .. "font-style: italic;"
		line[n].tone_sup = true -- single digits are assumed to be tone markers and will hence be superscripted
		if args['tone-superscripting'] and not yesno(args['tone-superscripting'])
			then line[n].tone_sup = false end
	end

	if args.glossing then -- the glossing= parameter sets the default glossing type
		local _gl = set_glossing_type(args.glossing)
		if _gl then conf.GlossingType = _gl end
	end
	--this looks for a list of glossing abbreviations on the page that transcludes the template:
	local _ablist_section = get_section(frame, 'list-of-glossing-abbreviations')
	if _ablist_section and _ablist_section ~= "" then
		local _a = mw.ustring.gsub(_ablist_section, '</?div [^\n]*>', '') -- strips off the div tags
		set_custom_glosses(_a)
	end
	--and this looks looks for a list of abbreviations set within the template:
	local _ablist = args.abbreviations
	if _ablist and _ablist ~= ""
		then set_custom_glosses(_ablist) end
	local _ablist = args.ablist
	if _ablist and _ablist ~= ""
		then set_custom_glosses(_ablist) end

	local _spacing = tonumber(args.spacing)
	if _spacing and _spacing <= 20
		then conf.style.WordDiv = conf.style.WordDiv .. 'margin-right: ' .. _spacing .. 'em;'
	else conf.style.WordDiv = conf.style.WordDiv .. conf.style.WordMargin
	end

	local offset, last_line = 0, 0
	for j,v in ipairs(args) do -- iterates over the unnamed parameters from the template
		last_line = last_line +1
		if is_empty(v)
			then offset = offset + 1
		else
		local i = j - offset
		line[i] = {}
		v = normalise(v)

		-- the following is part of a trial implementation of automatic transliteration:
		if if_auto_translit and v == "auto" and i > 1 then
			local source_line = line[i-1]
			local src_lang = source_line.lang
			if not src_lang then src_lang = args.lang end
			if src_lang then
					v = transliterate(source_line.whole, src_lang)
			else v = ""; msg:add("error", "No language specified for automatic transliteration")
			end
		end  -- end of trial block

		line[i].whole = v
		line[i].length = mw.ustring.len(v)

		local _c = args["c" .. i]
		if _c and _c ~= "" then
			line.hasComments = true
			line[i].c = _c
		end

		---prepare style arguments----
		line[i].class = ""
		local _style = args["style" .. i]
		if not _style then _style = ""
		else _style = tidyCss(_style) end
		--line[i].attr holds the attributes for the <p> elements that enclose the words in line i
		line[i].attr = {style = conf.style.WordP .. _style}

		local _lang = args["lang" .. i]
		if _lang and #_lang > 1 then
			line[i].lang = _lang
		else _lang = args.lang
			if _lang and #_lang > 1 and i == 1 then -- if a lang= parameter is supplied, it's assumed to apply to line 1
				line[i].lang = _lang
			end
		end
		line[i].attr.lang = line[i].lang
		--the following emulates the behaviour of {{Bo-textonly}} (see Template talk:Fs interlinear#Tibetan):
		if template_name == 'Template:Fs interlinear' then
			if _lang == "bo" and i == 1 then
				line[1].class = line[1].class .. " uchen"
				line[1].attr.style = line[1].attr.style .. "font-size:1.25em; word-wrap:break-word;"
			end
		end

		if yesno(args["italics" .. i]) then
			set_italics(i)
		end

		local _transl = args["transl" .. i]
		if _transl and #_transl > 1 then
			_transl = mw.ustring.lower(_transl)
			local _lookup = lang_data.translit_title_table[_transl]
			if _lookup then
				if _lang and  _lookup[_lang] then
					_transl = _lookup[_lang]
				else _transl = _lookup.default
				end
				if _transl then
					line[i].attr.title = _transl
				end
			else  msg:add("error", "Transliteration scheme '" .. _transl .. "' not recognised")
			end
		end

		local _glossing = args["glossing" .. i]
		if _glossing then
			line[i].glossing = set_glossing_type(_glossing)
			-- Do not treat default glossing settings as custom.
			if not ((i == 1 and not yesno(_glossing)) or (i == 2 and yesno(_glossing))) then
				line.HasCustomGlossing = true
			end
		end

		local _ipa = args['ipa' .. i]
		if yesno(_ipa) then
			line[i].class = "IPA"
		end

		local _class = args['class' .. i]
		if _class then
			line[i].class = line[i].class .. " " .. _class
		end

		if line[i].class == ""
			then line[i].class = nil end
		end -- ends the first if-statement in the loop
	end -- ends the FOR cycle

	local line_count = #line
	if line_count == 0 then
		msg:add("error", template_name .. ": no lines supplied.")
		return msg:print_errors()
	end

	if line_count > 1 then
		local _italics = args.italics
		local n = tonumber(_italics)
		if n and n > 0 then
			set_italics(n)
		elseif not (_italics and not yesno(_italics)) and not (args["italics1"] and not yesno(args["italics1"])) then
			set_italics(1) -- by default, the first line will get italicised, unless italics=no or italics1=no
		end
		-- the last unnamed parameter is assumed to be the free translation:
		free_translation = args[last_line]
		if not is_empty(free_translation) then
			line [line_count] = nil   end  --... and is thus excluded from interlinearising
	end

-- If glossing isn't specified for any line, then it's chosen by default to occur
-- in the second line, unless only a single line has been supplied, in which case
-- the assumption is that it is the one containing grammatical glosses
	if yesno(args.glossing) == false then
		line.HasCustomGlossing = true
	end
	if not line.HasCustomGlossing then
		if line_count == 1 then
			line[1].glossing = conf.GlossingType
		elseif line[2] then
			line[2].glossing = conf.GlossingType
		end
	end
	set_global_glossing_settings{style = args['glossing-style'], underline = args.underline, small_caps = args['small-caps']}

---------------------
-- Segment lines into words
---------------------
	for i,v in ipairs(line) do
		local ifglossing = false
		if line[i].glossing then
			ifglossing = true -- if true the parser will attempt to format gloss abbreviations in the current line
			glossing_type = line[i].glossing -- neccessarily a global variable
		end
		local wc, n = 1, 1
		line[i].words = {}
		while n <= line[i].length do
			buffer = ""
			n = parse(line[i], n, 0, ifglossing)+2
			line[i].words[wc] = buffer
			wc = wc + 1
		end
	end

	----Check for mismatches in number of words across lines----
	local number_of_words, mismatch_found = 0, false
	for i,v in ipairs(line) do -- find the maximum number of words in any line
		local wc = #line[i].words
		if wc ~= number_of_words then
			if i ~= 1 and wc ~= 0 then
				mismatch_found = true
			end
			if wc > number_of_words then
				number_of_words = wc
			end
		end
	end
	----Deal with mismatches---
	if mismatch_found then
		local error_text = "Mismatch in the number of words between lines: "
		for i,v in ipairs(line) do
			local wc = #line[i].words
			error_text = error_text .. wc .. " word(s) in line " .. i .. ", "
			if wc ~= number_of_words then
				for current_word = wc+1, number_of_words do
					line[i].words[current_word] = "&nbsp;"
				end
			end
		end
		if string.sub(error_text, -2) == ", "
			then error_text = string.sub(error_text, 1, #error_text - 2) .. " "
		end
		error_text = error_text .. help_link("mismatch")
		UserMessages:add("error", error_text)
	end

---------------------
-- Build the HTML
---------------------
	---- If just a single line was supplied, format it as inline text
	if line_count == 1 then
		local span = mw.html.create('span')
		span:attr(line[1].attr)
		for wi = 1, number_of_words do
			local space
			if wi < number_of_words then space = " " else space = "" end
			span:wikitext(line[1].words[wi] .. space)
		end
		return tostring(span)
	end

	---- More than one line supplied, so we'll produce interlinear display
	local div = mw.html.create("div")
	div:addClass(conf.class.Interlinear)

	-- For stuff to be displayed in the left margin, like example numbering
	local number, indent = nil, nil
	if args.number and args.number ~= ""
		then number = args.number end
	if args.indent and args.indent ~=""
		then indent = args.indent end
	if indent or number then
		if not indent then indent = "4" end --default value
		div:css("margin-left", indent .. 'em')
		if number then
			div:tag("div")
				:css("position", "absolute")
				:css("left", "1em")
				:wikitext(args.number)
		end
	end

	if args.box and args.box ~= "" then
		div:css("background-color", "#f8f9fa")
			:css("border", "1px solid #eaecf0")
			:css("padding", "1em") end
	if args.top and args.top ~= "" then --lines to display above the interlinear block
		div:tag("div")
			:wikitext(args.top)
	end

	-- Producing the interlinear block
	for wi = 1, number_of_words do
		local div2 = div:tag("div")
					:attr("style", conf.style.WordDiv)
		for i,_ in ipairs (line) do
			if line[i].whole ~= "" then -- skipping empty lines
				local p = div2:tag("p")
				p:attr(line[i].attr)
				if line[i].class then
					p:addClass(line[i].class)
				end
				local _text = line[i].words[wi]
				if _text == "" or _text == " "
					then _text = "&nbsp;" end -- <p> elements without content mess up the interlinear display
				p:wikitext(_text)
			end
		end
	end

	--- If any "comments" have been specified, add them at the end of each line
	if line.hasComments then
		local divc = div:tag("div")
					:attr("style", conf.style.WordDiv)
		for i,_ in ipairs (line) do
			local p = divc:tag("p")
			p:attr("style", conf.style.WordP)
			if line[i].c then
				p:wikitext(line[i].c)
			else p:wikitext("&nbsp;")
			end
		end
	end

	--Add hidden lines containing the content of each line of interlinear text: this is for accessibility
	for i,v in ipairs(line) do
		local hidden_line = div:tag("p")
		hidden_line:attr("style", conf.style.HiddenText)
					:wikitext(v.whole)
	end

	-- Format the free translation
	local ft_line = div:tag("p")
	if free_translation and free_translation ~= "" then
		ft_line:attr("style", "clear: left;")
		ft_line:wikitext(free_translation)
	end
	if args.bottom and args.bottom ~= ""
		then local bottom = div:tag('p')
		bottom:css('margin-top', '0')
		bottom:wikitext(args.bottom)
	end
	ft_line:node(msg:print_errors()) -- for error messages

	local end_div = div:tag("div")
		end_div:attr("style", conf.style.EndDiv)
	div:newline()
	local temp_track = ""
	if last_line == 2
		then temp_track = "[[Category:Pages with interlinear glosses using two unnamed parameters]]"
	end
	if last_line > 3 and template_name ~= 'Template:Fs interlinear'
		then  temp_track = "[[Category:Pages with interlinear glosses using more than three unnamed parameters]]"
	end
	return tostring(div) .. temp_track .. msg:print_warnings()
end

return p