Module:Delink/sandbox: Difference between revisions

Revision as of 10:18, 2 May 2017

Revision as of 08:52, 5 January 2020

Line 8:

local function delinkReversePipeTrick(s)

if s:match("^%[%[|.*[|\n]") or s == "[[|]]" then -- Check for newlines or multiple pipes.

return s

else

Line 16:

local function delinkPipeTrick(s)

local linkarea, display = "", ""

-- s the tile area, without bracket nor pipe

-- We need to deal with colons, brackets, and commas, per [[Help:Pipe trick]].

-- First, remove the text before the first colon, if any.

s =if s:gsubmatch("^(.-:")", "")then

local titlearea, display s = s:match("^%[%[(.-)|:(.*)|%]%]$")

-- If there are no colons, grab all of the text apart from the square brackets and the pipe.

endelse

s = s:match("%[%[(.*)|%]%]")

end

-- Next up, brackets and commas.

Line 32 ⟶ 37:

local function delinkWikilink(s)

local result = s

-- s is a string starting with "[[" and ending with "]]". It does not contain any other "]]" strings.

-- Deal with nested links

local nested = "[[" .. s:sub(3):gsub("%[%[.-%]%]$", delinkWikilink)

if nested ~= s then

return nested

end

-- Deal with the reverse pipe trick.

if sresult:match("^%[%[|") then

return delinkReversePipeTrick(sresult)

end

titlearearesult = mw.uri.decode(titlearearesult, "PATH") -- decode percent-encoded entities. Leave underscores and plus signs.

-- Check for bad titles. To do this we need to match the

titlearearesult = mw.text.decode(titlearearesult, true) -- decode HTML entities.

-- title area of the link, i.e. the part before any pipes.

local titlearea, display = s:match("^%[%[(.-)|(.*)%]%]$")

titlearea = titlearea or s:sub(3, -3)

titlearea = mw.uri.decode(titlearea, "PATH") -- decode percent-encoded entities. Leave underscores and plus signs.

titlearea = mw.text.decode(titlearea, true) -- decode HTML entities.

local temptitlearea, fragment = titlearea:match("^(.-)#(.*)$")

temptitlearea = temptitlearea or titlearea

fragment = fragment or ""

-- Check for bad titles. To do this we need to matchfind the

-- title area of the link, i.e. the part before any pipes.

local titlearea

if result:match("|") then -- Find if we're dealing with a piped link.

titlearea = result:match("^%[%[(.-)|.*%]%]")

else

titlearea = result:match("^%[%[(.-)%]%]")

end

-- Check for bad characters.

if not if mw.ustring.isutf8match(titlearea, "[%[%]<>{}%%%c\n]") then

return s

or temptitlearea:match("[%[%]<>|{}%z%c\n]")

or temptitlearea:match("%%%x%x")

or temptitlearea:match("&..-;")

or fragment:match("[%[%]{}%c\n]")

then

return s

end

-- Check for categories, interwikis, and files.

local colonprefix = titlearearesult:match("^%[%[(.-):.*%]%]") or "" -- Get the text before the first colon.

local ns = mw.site.namespaces[colonprefix] -- see if this is a known namespace

if mw.language.isKnownLanguageTag(colonprefix)

or ( ns and ( ns.canonicalName == "File" or ns.canonicalName == "Category" ) ) then

return ""

end

-- Remove the colon if the link is using the [[Help:Colon trick]].

if titlearearesult:submatch(1, 1) == "%[%[:") then

titlearea result = titlearea"[[" .. result:match("%[%[:sub(2.*%]%])")

end

-- Deal with links using the [[Help:Pipe trick]].

if display ==mw.ustring.match(result, "^%[%[[^|]*|%]%]") then

return delinkPipeTrick(titlearea result)

end

-- matchFind the display area of the wikilink

if not displayresult:match("|") then -- matchFind if we're dealing with a piped link.

result = result:match("^%[%[.-|(.+)%]%]")

-- Remove new lines from the display of multiline piped links,

-- where the pipe is before the first new line.

titlearearesult = titlearearesult:gsub("\n", "")

else

result = result:match("^%[%[(.-)%]%]")

end

return display or titlearea result

end

Line 109 ⟶ 103:

local url_prefix

for i,v in ipairs(valid_url_prefixes) do

if s:findmw.ustring.match(s, '^%[' .. v ..'[^"%s].*%]' ) then

url_prefix = v

break

Line 117 ⟶ 111:

-- Get display text

if not url_prefix then

return s

-- Deal with nested links or send back original string.

return "[" .. mw.ustring.gsub( mw.ustring.sub(s, 2), "%[.-%]", delinkURL )

end

s = s:match("^%[" .. url_prefix .. "(.*)%]") -- Grab all of the text after the URL prefix and before the final square bracket.

s = s:match('^.-(["<> [].*)') or "" -- Grab all of the text after the first URL separator character ("<> ).

s = s:mw.ustring.match(s, "^%s*(%S.*)$") or "" -- If the separating character was a space, trim it off.

local s_decoded = mw.text.decode(s, true)

if s_decoded:mw.ustring.match(s_decoded, "%c") then

return s

else

Line 132 ⟶ 125:

end

local function delinkLinkClass(s, pattern, delinkFunction)

if nestednot type(s) ~== s"string" then

error("Attempt to de-link non-string input.", 2)

end

if not ( type(pattern) == "string" and mw.ustring.sub(pattern, 1, 1) == "^" ) then

error('Invalid pattern detected. Patterns must begin with "^".', 2)

end

-- Iterate over the text string, and replace any matched text. using the

-- delink function. We need to iterate character by character rather

-- than just use gsub, otherwise nested links aren't detected properly.

local result = ""

while s ~= '' do

-- Replace text using one iteration of gsub.

s = mw.ustring.gsub(s, pattern, delinkFunction, 1)

-- Append the left-most character to the result string.

returnresult "["= ..result mw.ustring.gsub( mw.ustring.sub(s, 2)1, "%[.-%]", delinkURL 1)

s = mw.ustring.sub(s, 2, -1)

end

return nestedresult

end

function p._delink(args)

Line 138 ⟶ 151:

-- Remove any [[Help:Strip markers]] representing ref tags. In most situations

-- this is not a good idea - only use it if you know what you are doing!

text = mw.textustring.unstripgsub(text, "UNIQ%w*%-ref%-%d*%-QINU", "")

end

if not (args.comments == "no") then

Line 144 ⟶ 157:

end

if not (args.wikilinks == "no") then

text = text:gsubdelinkLinkClass(text, "^%[%[.-%]%]", delinkWikilink) -- De-link wikilinks.

end

if not (args.urls == "no") then

text = text:gsubdelinkLinkClass(text, "^%[.-%]", delinkURL) -- De-link URLs.

end

if not (args.whitespace == "no") then

-- Replace single new lines with a single space, but leave double new lines

-- and new lines only containing spaces or tabs before a second new line.

text = text:mw.ustring.gsub(text, "([^\n \t][ \t]*)\n([ \t]*[^\n \t])", "%1 %2")

--text = text:gsub("\n\t+", "\n") -- Remove leading tab.

text = text:gsub("[ \t]+", " ") -- Remove extra tabs and spaces.

end

if not (args.nowiki == "no") then

text = mw.text.unstripNoWiki(text)

end

return text