Module:List and Module:Delink: Difference between pages
(Difference between pages)
en>Mr. Stradivarius (add support for horizontal numbered lists) |
en>Mr. Stradivarius (html comments are probably mostly removed by the parser anyway - testing) |
||
Line 1: | Line 1: | ||
-- This module | -- This module de-links most wikitext. | ||
p = {} | |||
local function | local function delinkReversePipeTrick(s) | ||
if mw.ustring.match(s, "^%[%[|.*|") then -- Check for multiple pipes. | |||
return s | |||
else | |||
return mw.ustring.match(s, "%[%[|(.*)%]%]") | |||
end | |||
end | end | ||
local function | local function delinkPipeTrick(s) | ||
local linkarea, display = "", "" | |||
-- We need to deal with colons, brackets, and commas, per [[Help:Pipe trick]]. | |||
-- First, remove the text before the first colon, if any. | |||
if mw.ustring.match(s, ":") then | |||
s = mw.ustring.match(s, "%[%[.-:(.*)|%]%]") | |||
-- If there are no colons, grab all of the text apart from the square brackets and the pipe. | |||
else | |||
s = mw.ustring.match(s, "%[%[(.*)|%]%]") | |||
end | |||
-- 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, "(.-),.*$") | |||
end | |||
return s | |||
end | end | ||
local function | local function delinkWikilink(s) | ||
-- Deal with the reverse pipe trick. | |||
if mw.ustring.match(s, "%[%[|") then | |||
return delinkReversePipeTrick(s) | |||
end | |||
-- Check for bad titles. To do this we need to find the | |||
-- title area of the link, i.e. the part before any pipes. | |||
local titlearea | |||
if mw.ustring.match(s, "|") then -- Find if we're dealing with a piped link. | |||
titlearea = mw.ustring.match(s, "^%[%[(.-)|.*%]%]") | |||
else | |||
titlearea = mw.ustring.match(s, "^%[%[(.-)%]%]") | |||
end | |||
-- If the title area is not a valid title, return the whole string. | |||
-- Use pcall in case we're over the expensive functions limit. | |||
local goodcall, title = pcall(mw.title.new, titlearea, "") | |||
if not (goodcall and title) then | |||
return s | |||
end | |||
-- Check for characters that are allowed in titles but not in wikilinks. | |||
local other_invalid_link_strings = { '�' } | |||
for i,v in ipairs(other_invalid_link_strings) do | |||
if mw.ustring.match(titlearea, v) then | |||
return s | |||
end | |||
end | |||
-- Check for categories, interwikis, and files. | |||
local colonprefix = mw.ustring.match(s, "%[%[(.-):.*%]%]") or "" -- Get the text before the first colon. | |||
if mw.language.isKnownLanguageTag(colonprefix) | |||
or mw.ustring.match(colonprefix, "^[Cc]ategory$") | |||
or mw.ustring.match(colonprefix, "^[Ff]ile$") | |||
or mw.ustring.match(colonprefix, "^[Ii]mage$") then | |||
return "" | |||
end | |||
-- Remove the colon if the link is using the [[Help:Colon trick]]. | |||
if mw.ustring.match(s, "%[%[:") then | |||
s = "[[" .. mw.ustring.match(s, "%[%[:(.*%]%])") | |||
end | |||
-- Deal with links using the [[Help:Pipe trick]]. | |||
if mw.ustring.match(s, "^%[%[[^|]*|%]%]") then | |||
return delinkPipeTrick(s) | |||
end | |||
-- Find the display area of the wikilink | |||
local display | |||
if mw.ustring.match(s, "|") then -- Find if we're dealing with a piped link. | |||
display = mw.ustring.match(s, "^%[%[.-|(.+)%]%]") | |||
else | |||
display = mw.ustring.match(s, "^%[%[(.-)%]%]") | |||
end | |||
return display | |||
end | end | ||
local function | local function delinkURL(s) | ||
-- Assume we have already delinked internal wikilinks, and that | |||
-- we have been passed some text between two square brackets [foo]. | |||
-- 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 | |||
return s | |||
else | |||
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. | |||
return s | |||
end | |||
end | end | ||
function | local function delinkLinkClass(s, pattern, delinkFunction) | ||
if not type(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 mw.ustring.len(s) > 0 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. | |||
result = result .. mw.ustring.sub(s, 1, 1) | |||
s = mw.ustring.sub(s, 2, -1) | |||
end | |||
return result | |||
end | end | ||
local function | local function _delink(args) | ||
local text = args[1] or "" | |||
-- text = mw.ustring.gsub(text, "<!%-%-.-%-%->", "") -- Remove html comments. | |||
text = delinkLinkClass(text, "^%[%[.-%]%]", delinkWikilink) -- De-link wikilinks. | |||
text = delinkLinkClass(text, "^%[.-%]", delinkURL) -- De-link URLs. | |||
text = mw.ustring.gsub(text, "%s+", " ") -- Remove extra whitespace. | |||
return text | |||
end | end | ||
local | function p.delink(frame) | ||
local args | |||
for | if frame == mw.getCurrentFrame() then | ||
-- We're being called via #invoke. If the invoking template passed any args, use | |||
-- them. Otherwise, use the args that were passed into the template. | |||
args = frame:getParent().args | |||
for k, v in pairs(frame.args) do | |||
args = frame.args | |||
break | |||
end | |||
else | |||
-- We're being called from another module or from the debug console, so assume | |||
-- the args are passed in directly. | |||
args = frame | |||
end | |||
return _delink(args) | |||
end | end | ||
return p | return p |
Revision as of 07:50, 5 April 2013
Documentation for this module may be created at Module:Delink/doc
-- This module de-links most wikitext.
p = {}
local function delinkReversePipeTrick(s)
if mw.ustring.match(s, "^%[%[|.*|") then -- Check for multiple pipes.
return s
else
return mw.ustring.match(s, "%[%[|(.*)%]%]")
end
end
local function delinkPipeTrick(s)
local linkarea, display = "", ""
-- We need to deal with colons, brackets, and commas, per [[Help:Pipe trick]].
-- First, remove the text before the first colon, if any.
if mw.ustring.match(s, ":") then
s = mw.ustring.match(s, "%[%[.-:(.*)|%]%]")
-- If there are no colons, grab all of the text apart from the square brackets and the pipe.
else
s = mw.ustring.match(s, "%[%[(.*)|%]%]")
end
-- 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
local function delinkWikilink(s)
-- Deal with the reverse pipe trick.
if mw.ustring.match(s, "%[%[|") then
return delinkReversePipeTrick(s)
end
-- 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
if mw.ustring.match(s, "|") then -- Find if we're dealing with a piped link.
titlearea = mw.ustring.match(s, "^%[%[(.-)|.*%]%]")
else
titlearea = mw.ustring.match(s, "^%[%[(.-)%]%]")
end
-- If the title area is not a valid title, return the whole string.
-- Use pcall in case we're over the expensive functions limit.
local goodcall, title = pcall(mw.title.new, titlearea, "")
if not (goodcall and title) then
return s
end
-- Check for characters that are allowed in titles but not in wikilinks.
local other_invalid_link_strings = { '�' }
for i,v in ipairs(other_invalid_link_strings) do
if mw.ustring.match(titlearea, v) then
return s
end
end
-- Check for categories, interwikis, and files.
local colonprefix = mw.ustring.match(s, "%[%[(.-):.*%]%]") or "" -- Get the text before the first colon.
if mw.language.isKnownLanguageTag(colonprefix)
or mw.ustring.match(colonprefix, "^[Cc]ategory$")
or mw.ustring.match(colonprefix, "^[Ff]ile$")
or mw.ustring.match(colonprefix, "^[Ii]mage$") then
return ""
end
-- Remove the colon if the link is using the [[Help:Colon trick]].
if mw.ustring.match(s, "%[%[:") then
s = "[[" .. mw.ustring.match(s, "%[%[:(.*%]%])")
end
-- Deal with links using the [[Help:Pipe trick]].
if mw.ustring.match(s, "^%[%[[^|]*|%]%]") then
return delinkPipeTrick(s)
end
-- Find the display area of the wikilink
local display
if mw.ustring.match(s, "|") then -- Find if we're dealing with a piped link.
display = mw.ustring.match(s, "^%[%[.-|(.+)%]%]")
else
display = mw.ustring.match(s, "^%[%[(.-)%]%]")
end
return display
end
local function delinkURL(s)
-- Assume we have already delinked internal wikilinks, and that
-- we have been passed some text between two square brackets [foo].
-- Check if the text has a valid URL prefix and at least one valid URL character.
local valid_url_prefixes = {"//", "http://", "https://", "ftp://", "gopher://", "mailto:", "news:", "irc://"}
local url_prefix
for i,v in ipairs(valid_url_prefixes) do
if mw.ustring.match(s, '^%[' .. v ..'[^"%s].*%]' ) then
url_prefix = v
break
end
end
-- Get display text
if not url_prefix then
return s
else
s = mw.ustring.match(s, "^%[" .. url_prefix .. "(.*)%]") -- Grab all of the text after the URL prefix and before the final square bracket.
s = mw.ustring.match(s, '^.-(["<> ].*)') or "" -- Grab all of the text after the first URL separator character ("<> ).
s = mw.ustring.match(s, "^%s*(%S.*)$") or "" -- If the separating character was a space, trim it off.
return s
end
end
local function delinkLinkClass(s, pattern, delinkFunction)
if not type(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 mw.ustring.len(s) > 0 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.
result = result .. mw.ustring.sub(s, 1, 1)
s = mw.ustring.sub(s, 2, -1)
end
return result
end
local function _delink(args)
local text = args[1] or ""
-- text = mw.ustring.gsub(text, "<!%-%-.-%-%->", "") -- Remove html comments.
text = delinkLinkClass(text, "^%[%[.-%]%]", delinkWikilink) -- De-link wikilinks.
text = delinkLinkClass(text, "^%[.-%]", delinkURL) -- De-link URLs.
text = mw.ustring.gsub(text, "%s+", " ") -- Remove extra whitespace.
return text
end
function p.delink(frame)
local args
if frame == mw.getCurrentFrame() then
-- We're being called via #invoke. If the invoking template passed any args, use
-- them. Otherwise, use the args that were passed into the template.
args = frame:getParent().args
for k, v in pairs(frame.args) do
args = frame.args
break
end
else
-- We're being called from another module or from the debug console, so assume
-- the args are passed in directly.
args = frame
end
return _delink(args)
end
return p