Difference between revisions of "Module:Delink/sandbox"
Jump to navigation
Jump to search
blackwiki>Mr. Stradivarius (try the version from fr:Module:Delink, written by User:Zebulon84) |
blackwiki>Mr. Stradivarius (start sanitizing this) |
||
| Line 1: | Line 1: | ||
-- This module de-links most wikitext. | -- This module de-links most wikitext. | ||
| + | |||
| + | local getArgs = require('Module:Arguments').getArgs | ||
| + | local yesno = require('Module:Yesno') | ||
p = {} | p = {} | ||
| − | local function | + | local function whitespaceYesno(val) |
| − | + | -- Like yesno, but trims whitespace from vals and removes blank strings. | |
| − | + | if type(val) == 'string' then | |
| − | + | val = val:match('^%s*(.-)%s*$') | |
| − | + | if val == '' then | |
| − | + | return nil | |
| + | end | ||
| + | end | ||
| + | return yesno(val) | ||
end | end | ||
| − | + | function p._delinkReversePipeTrick(s) | |
| + | if mw.ustring.match(s, "^%[%[|.*[|\n]") or s == "[[|]]" then -- Check for newlines or multiple pipes. | ||
| + | return s | ||
| + | else | ||
| + | return mw.ustring.match(s, "%[%[|(.*)%]%]") | ||
| + | end | ||
| + | end | ||
| + | |||
| + | function p._delinkPipeTrick(s) | ||
-- s the tile area, without bracket nor pipe | -- s the tile area, without bracket nor pipe | ||
| − | + | -- We need to deal with colons, brackets, and commas, per [[Help:Pipe trick]]. | |
| − | + | ||
| − | + | -- First, remove the text before the first colon, if any. | |
| − | + | s = mw.ustring.gsub(s, "^(.-:)", "") | |
| − | + | ||
| − | + | -- Next up, brackets and commas. | |
| − | + | if mw.ustring.match(s, "%(.-%)$") then -- Brackets trump commas. | |
| − | + | s = mw.ustring.match(s, "(.-) ?%(.-%)$") | |
| − | + | elseif mw.ustring.match(s, ",") then -- If there are no brackets, display only the text before the first comma. | |
| − | + | s = mw.ustring.match(s, "(.-),.*$") | |
| − | + | end | |
| − | + | return s | |
end | end | ||
| − | + | function p._delinkWikilink(s) | |
-- s should be a string starting by "[[" and ending by "]]". It should not contain other "]]" | -- s should be a string starting by "[[" and ending by "]]". It should not contain other "]]" | ||
| − | + | ||
-- Deal with nested links | -- Deal with nested links | ||
local nested | local nested | ||
| − | + | nested = "[[" .. mw.ustring.gsub( mw.ustring.sub(s, 3), "%[%[.-%]%]", p._delinkWikilink) | |
| − | + | if nested ~= s then | |
| − | + | return nested | |
| − | + | end | |
| − | + | ||
| − | + | -- Deal with the reverse pipe trick. | |
| − | + | if mw.ustring.match(s, "^%[%[|") then | |
| − | + | return p._delinkReversePipeTrick(s) | |
| − | + | end | |
| − | + | ||
| − | + | local decoded = mw.uri.decode(s, "PATH") -- decode percent-encoded entities. Leave underscores and plus signs. | |
| − | + | decoded = mw.text.decode(decoded, true) -- decode HTML entities. | |
| − | + | ||
| − | + | -- Check for bad titles. To do this we need to find the | |
| − | + | -- title area of the link, i.e. the part before any pipes. | |
| − | + | local titlearea, display = mw.ustring.match(decoded, "^%[%[([^|%]]*)|?(.*)%]%]") | |
| + | |||
| + | -- Check for bad characters. | ||
| + | if mw.ustring.match(titlearea, "[%[%]<>{}%%%c\n]") then | ||
| + | return s | ||
| + | end | ||
| − | + | -- Check for categories, interwikis, and files. | |
| − | + | local colonprefix = mw.ustring.match(titlearea, "^(.-):") 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 mw.ustring.sub(titlearea, 1, 1) == ":" then | ||
| + | titlearea = mw.ustring.sub(titlearea, 2) | ||
| + | end | ||
| + | |||
| + | -- Deal with links using the [[Help:Pipe trick]]. | ||
| + | if display == "" then | ||
| + | return p._delinkPipeTrick(titlearea ) | ||
| + | end | ||
| + | |||
| + | -- Find the display area of the wikilink | ||
| + | if not display then -- Find if we're dealing with a piped link. | ||
| + | -- Remove new lines from the display of multiline piped links, | ||
| + | -- where the pipe is before the first new line. | ||
| + | titlearea = mw.ustring.gsub(titlearea, "\n", "") | ||
| + | end | ||
| + | |||
| + | return display or titlearea | ||
end | end | ||
| − | + | function p._delinkURL(s) | |
| − | + | -- Assume we have already delinked internal wikilinks, and that | |
| − | + | -- we have been passed some text between two square brackets [foo]. | |
| − | + | ||
| − | + | -- If the text contains a line break it is not formatted as a URL, regardless of other content. | |
| − | + | if mw.ustring.match(s, "\n") then | |
| − | + | return s | |
| − | + | end | |
| − | + | ||
| − | + | -- Check if the text has a valid URL prefix and at least one valid URL character. | |
| − | + | local valid_url_prefixes = {"//", "http://", "https://", "ftp://", "gopher://", "mailto:", "news:", "irc://"} | |
| − | + | local url_prefix | |
| − | + | for i,v in ipairs(valid_url_prefixes) do | |
| − | + | if mw.ustring.match(s, '^%[' .. v ..'[^"%s].*%]' ) then | |
| − | + | url_prefix = v | |
| − | + | break | |
| − | + | end | |
| − | + | end | |
| − | + | ||
| − | + | -- Get display text | |
| − | + | if not url_prefix then | |
| − | + | -- Deal with nested links or send back original string. | |
| − | + | return "[" .. mw.ustring.gsub( mw.ustring.sub(s, 2), "%[.-%]", p._delinkURL ) | |
| − | + | end | |
| − | + | s = mw.ustring.match(s, "^%[" .. url_prefix .. "(.*)%]") -- Grab all of the text after the URL prefix and before the final square bracket. | |
| − | + | s = mw.ustring.match(s, '^.-(["<> [].*)') or "" -- Grab all of the text after the first URL separator character ("<> ). | |
| − | + | 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 mw.ustring.match(s_decoded, "%c") then | |
| − | + | return s | |
| − | + | else | |
| − | + | return s_decoded | |
| − | + | end | |
end | end | ||
| − | + | function p._delink(args) | |
| − | + | local text = args[1] or '' | |
| − | + | if whitespaceYesno(args.refs) then | |
| − | + | -- 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.ustring.gsub(text, 'UNIQ%w*%-ref%-%d*%-QINU', '') | |
| − | + | end | |
| − | + | if whitespaceYesno(args.comments) == false then | |
| − | + | text = mw.ustring.gsub(text, '<!%-%-.-%-%->', '') -- Remove html comments. | |
| − | + | end | |
| − | + | if whitespaceYesno(args.wikilinks) == false then | |
| − | + | text = mw.ustring.gsub(text, '%[%[.-%]%]', p._delinkWikilink) -- De-link wikilinks. | |
| − | + | end | |
| − | + | if whitespaceYesno(args.urls) == false then | |
| − | + | text = mw.ustring.gsub(text, '%[.-%]', p._delinkURL) -- De-link URLs. | |
| − | + | end | |
| − | + | if whitespaceYesno(args.whitespace) == false 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 = mw.ustring.gsub(text, '([^\n \t][ \t]*)\n([ \t]*[^\n \t])', '%1 %2') | |
| − | + | text = mw.ustring.gsub(text, '[ \t]+', ' ') -- Remove extra tabs and spaces. | |
| − | + | end | |
| − | + | return text | |
| − | |||
end | end | ||
function p.delink(frame) | function p.delink(frame) | ||
| − | + | local args = frame:getParent().args | |
| − | + | return p._delink(args) | |
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
end | end | ||
return p | return p | ||
Revision as of 15:15, 17 April 2014
Documentation for this module may be created at Module:Delink/sandbox/doc
-- This module de-links most wikitext.
local getArgs = require('Module:Arguments').getArgs
local yesno = require('Module:Yesno')
p = {}
local function whitespaceYesno(val)
-- Like yesno, but trims whitespace from vals and removes blank strings.
if type(val) == 'string' then
val = val:match('^%s*(.-)%s*$')
if val == '' then
return nil
end
end
return yesno(val)
end
function p._delinkReversePipeTrick(s)
if mw.ustring.match(s, "^%[%[|.*[|\n]") or s == "[[|]]" then -- Check for newlines or multiple pipes.
return s
else
return mw.ustring.match(s, "%[%[|(.*)%]%]")
end
end
function p._delinkPipeTrick(s)
-- 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 = mw.ustring.gsub(s, "^(.-:)", "")
-- Next up, brackets and commas.
if mw.ustring.match(s, "%(.-%)$") then -- Brackets trump commas.
s = mw.ustring.match(s, "(.-) ?%(.-%)$")
elseif mw.ustring.match(s, ",") then -- If there are no brackets, display only the text before the first comma.
s = mw.ustring.match(s, "(.-),.*$")
end
return s
end
function p._delinkWikilink(s)
-- s should be a string starting by "[[" and ending by "]]". It should not contain other "]]"
-- Deal with nested links
local nested
nested = "[[" .. mw.ustring.gsub( mw.ustring.sub(s, 3), "%[%[.-%]%]", p._delinkWikilink)
if nested ~= s then
return nested
end
-- Deal with the reverse pipe trick.
if mw.ustring.match(s, "^%[%[|") then
return p._delinkReversePipeTrick(s)
end
local decoded = mw.uri.decode(s, "PATH") -- decode percent-encoded entities. Leave underscores and plus signs.
decoded = mw.text.decode(decoded, true) -- decode HTML entities.
-- Check for bad titles. To do this we need to find the
-- title area of the link, i.e. the part before any pipes.
local titlearea, display = mw.ustring.match(decoded, "^%[%[([^|%]]*)|?(.*)%]%]")
-- Check for bad characters.
if mw.ustring.match(titlearea, "[%[%]<>{}%%%c\n]") then
return s
end
-- Check for categories, interwikis, and files.
local colonprefix = mw.ustring.match(titlearea, "^(.-):") 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 mw.ustring.sub(titlearea, 1, 1) == ":" then
titlearea = mw.ustring.sub(titlearea, 2)
end
-- Deal with links using the [[Help:Pipe trick]].
if display == "" then
return p._delinkPipeTrick(titlearea )
end
-- Find the display area of the wikilink
if not display then -- Find if we're dealing with a piped link.
-- Remove new lines from the display of multiline piped links,
-- where the pipe is before the first new line.
titlearea = mw.ustring.gsub(titlearea, "\n", "")
end
return display or titlearea
end
function p._delinkURL(s)
-- Assume we have already delinked internal wikilinks, and that
-- we have been passed some text between two square brackets [foo].
-- If the text contains a line break it is not formatted as a URL, regardless of other content.
if mw.ustring.match(s, "\n") then
return s
end
-- Check if the text has a valid URL prefix and at least one valid URL character.
local valid_url_prefixes = {"//", "http://", "https://", "ftp://", "gopher://", "mailto:", "news:", "irc://"}
local url_prefix
for i,v in ipairs(valid_url_prefixes) do
if mw.ustring.match(s, '^%[' .. v ..'[^"%s].*%]' ) then
url_prefix = v
break
end
end
-- Get display text
if not url_prefix then
-- Deal with nested links or send back original string.
return "[" .. mw.ustring.gsub( mw.ustring.sub(s, 2), "%[.-%]", p._delinkURL )
end
s = mw.ustring.match(s, "^%[" .. url_prefix .. "(.*)%]") -- Grab all of the text after the URL prefix and before the final square bracket.
s = mw.ustring.match(s, '^.-(["<> [].*)') or "" -- Grab all of the text after the first URL separator character ("<> ).
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 mw.ustring.match(s_decoded, "%c") then
return s
else
return s_decoded
end
end
function p._delink(args)
local text = args[1] or ''
if whitespaceYesno(args.refs) then
-- 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.ustring.gsub(text, 'UNIQ%w*%-ref%-%d*%-QINU', '')
end
if whitespaceYesno(args.comments) == false then
text = mw.ustring.gsub(text, '<!%-%-.-%-%->', '') -- Remove html comments.
end
if whitespaceYesno(args.wikilinks) == false then
text = mw.ustring.gsub(text, '%[%[.-%]%]', p._delinkWikilink) -- De-link wikilinks.
end
if whitespaceYesno(args.urls) == false then
text = mw.ustring.gsub(text, '%[.-%]', p._delinkURL) -- De-link URLs.
end
if whitespaceYesno(args.whitespace) == false 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 = mw.ustring.gsub(text, '([^\n \t][ \t]*)\n([ \t]*[^\n \t])', '%1 %2')
text = mw.ustring.gsub(text, '[ \t]+', ' ') -- Remove extra tabs and spaces.
end
return text
end
function p.delink(frame)
local args = frame:getParent().args
return p._delink(args)
end
return p