Module:Excerpt/sandbox: Difference between revisions

Revision as of 22:20, 28 April 2020

Revision as of 22:53, 28 April 2020

Line 37:

-- If template is wanted but produces an unwanted reference then return the string with |shortref or |ref removed

local norefnoRef = mw.ustring.gsub(t, "|%s*shortref%s*%f[|}]", "")

norefnoRef = mw.ustring.gsub(norefnoRef, "|%s*ref%s*%f[|}]", "")

-- If a wanted template has unwanted nested templates, purge them too

norefnoRef = mw.ustring.sub(norefnoRef, 1, 2) .. mw.ustring.gsub(mw.ustring.sub(norefnoRef, 3), "%b{}", stripTemplate)

-- Replace {{audio}} by its text parameter: {{Audio|Foo.ogg|Bar}} → Bar

norefnoRef = mw.ustring.gsub(norefnoRef, "^{{%s*[Aa]udio.-|.-|(.-)%f[|}].*", "%1")

-- Replace {{Nihongo foot}} by its text parameter: {{Nihongo foot|English|英語|eigo}} → English

norefnoRef = mw.ustring.gsub(norefnoRef, "^{{%s*[Nn]ihongo[ _]+foot%s*|(.-)%f[|}].*", "%1")

if norefnoRef ~= t then return norefnoRef end

return nil -- not an unwanted template: keep

Line 76:

end

local descfileDescription, rtitlefileTitle = getContent(page) -- get file description and title after following any redirect

if descfileDescription and descfileDescription ~= "" then -- found description on local wiki

if mw.ustring.match(descfileDescription, "[Nn]on%-free") then return false end

descfileDescription = mw.ustring.gsub(descfileDescription, "%b{}", stripTemplate) -- remove DEFAULTSORT etc. to avoid side effects of frame:preprocess

elseif not rtitlefileTitle then

return false

else

-- try commons

descfileDescription = "{{" .. rtitlefileTitle .. "}}"

end

frame = frame or mw.getCurrentFrame()

descfileDescription = frame:preprocess(descfileDescription)

return ( descfileDescription and descfileDescription ~= "" and not mw.ustring.match(descfileDescription, "[Nn]on%-free") ) and true or false -- hide non-free image

end

Line 106:

local function parseCaption(caption)

if not caption then return nil end

local lenlength = mw.ustring.len(caption)

local posposition = 1

while posposition <= lenlength do

local linkstartlinkStart, linkendlinkEnd = mw.ustring.find(caption, "%b[]", posposition)

linkstartlinkStart = linkstartlinkStart or lenlength + 1 -- avoid comparison with nil when no link

local templatestarttemplateStart, templateendtemplateEnd = mw.ustring.find(caption, "%b{}", posposition)

templatestarttemplateStart = templatestarttemplateStart or lenlength + 1 -- avoid comparison with nil when no template

local argendargEnd = mw.ustring.find(caption, "[|}]", posposition) or lenlength + 1

if linkstartlinkStart < templatestarttemplateStart and linkstartlinkStart < argendargEnd then

posposition = linkendlinkEnd + 1 -- skip wikilink

elseif templatestarttemplateStart < argendargEnd then

posposition = templateendtemplateEnd + 1 -- skip template

else -- argument ends before the next wikilink or template

return mw.ustring.sub(caption, 1, argendargEnd - 1)

end

end

Line 137:

local hasImages = false

local images = {}

local capture_fromcaptureFrom = 1

while capture_fromcaptureFrom < mw.ustring.len(text) do

local argname, position, image = mw.ustring.match(text, "|%s*([^=|]-[Ii][Mm][Aa][Gg][Ee][^=|]-)%s*=%s*()(.*)", capture_fromcaptureFrom)

if image then -- ImageCaption=, image_size=, image_upright=, etc. do not introduce an image

local lcArgname = mw.ustring.lower(argname)

Line 151:

hasImages = true

images[position] = image

capture_fromcaptureFrom = position

else

capture_fromcaptureFrom = mw.ustring.len(text)

end

end

capture_fromcaptureFrom = 1

while capture_fromcaptureFrom < mw.ustring.len(text) do

local position, image = mw.ustring.match(text, "|%s*[^=|]-[Pp][Hh][Oo][Tt][Oo][^=|]-%s*=%s*()(.*)", capture_fromcaptureFrom)

if image then

hasImages = true

images[position] = image

capture_fromcaptureFrom = position

else

capture_fromcaptureFrom = mw.ustring.len(text)

end

end

capture_fromcaptureFrom = 1

while capture_fromcaptureFrom < mw.ustring.len(text) do

local position, image = mw.ustring.match(text, "|%s*[^=|{}]-%s*=%s*()%[?%[?([^|{}]*%.%a%a%a%a?)%s*%f[|}]", capture_fromcaptureFrom)

if image then

hasImages = true

Line 175:

images[position] = image

end

capture_fromcaptureFrom = position

else

capture_fromcaptureFrom = mw.ustring.len(text)

end

end

Line 185:

-- find all captions

local captions = {}

capture_fromcaptureFrom = 1

while capture_fromcaptureFrom < mw.ustring.len(text) do

local position, caption = matchAny(text, "|%s*", d.captionParams, "%s*=%s*()([^\n]+)", capture_fromcaptureFrom)

if caption then

-- extend caption to parse "| caption = Foo {{Template\n on\n multiple lines}} Bar\n"

Line 207:

end

end

capture_fromcaptureFrom = position

else

capture_fromcaptureFrom = mw.ustring.len(text)

end

end

Line 223:

mw.ustring.match(altText, ".*%[%b[]%]()") or 1)

local lenlength = mw.ustring.len(altText)

local aftertext = math.min( -- find position after whichever comes first: end of string, }} or |

mw.ustring.match(altText, "()}}", lookfrom) or lenlength+1,

mw.ustring.match(altText, "()|", lookfrom) or lenlength+1)

altText = mw.ustring.sub(altText, 1, aftertext-1) -- chop off |... or }}... which is not part of [[...]] or {{...}}

Line 333:

}

local function modifyImage(image, fileargsfileArgs)

if fileargsfileArgs then

for _, filearg in pairs(mw.text.split(fileargsfileArgs, "|")) do -- handle fileargsfileArgs=left|border etc.

local fa = mw.ustring.gsub(filearg, "=.*", "") -- "upright=0.75" → "upright"

local group = {fa} -- group of "border" is ["border"]...

Line 362:

-- @param filesOnly : If set, only return the files and not the prose

local function parse(text, options, filesOnly)

local allparasallParagraphs = true -- keep all paragraphs?

if options.paraflags then

if type(options.paraflags) ~= "table" then options.paraflags = numberFlags(options.paraflags) end

for _, v in pairs(options.paraflags) do

if v then allparasallParagraphs = false end -- if any para specifically requested, don't keep all

end

end

if filesOnly then

allparasallParagraphs = false

options.paraflags = {}

end

Line 382:

end

local fileargsfileArgs = options.fileargs and mw.text.trim(options.fileargs)

if fileargsfileArgs == '' then fileargsfileArgs = nil end

local leadstartleadStart = nil -- have we found some text yet?

local t = "" -- the stripped down output text

local filetextfileText = "" -- output text with concatenated [[File:Foo|...]]\n entries

local files = 0 -- how many images so far

local paras = 0 -- how many paragraphs so far

Line 401:

f = "[[File:" .. f .. "]]"

f = modifyImage(f, "thumb")

f = modifyImage(f, fileargsfileArgs)

if checkImage(f) then filetextfileText = filetextfileText .. f .. "\n" end

end

repeat -- loop around parsing a template, image or paragraph

local token = mw.ustring.match(text, "^%b{}%s*") or false -- {{Template}} or {| Table |}

if not leadstartleadStart and not token then token = mw.ustring.match(text, "^%b<>%s*%b{}%s*") end -- allow <tag>{{template}} before lead has started

local line = mw.ustring.match(text, "[^\n]*")

Line 421:

if token then -- found a template which is not the prefix to a line of text

if leadstartleadStart then -- lead has already started, so keep the template within the text, unless it's a whole line (navbox etc.)

if not filesOnly and not startLine then t = t .. token end

Line 441:

if options.fileflags and options.fileflags[files] then -- if displaying this image

image = modifyImage(image, "thumb")

image = modifyImage(image, fileargsfileArgs)

filetextfileText = filetextfileText .. image

end

end

Line 454:

if options.fileflags and options.fileflags[files] then

local image = token -- copy token for manipulation by adding |right etc. without changing the original

image = modifyImage(image, fileargsfileArgs)

filetextfileText = filetextfileText .. image

end

end

else -- got a paragraph, which ends at a file, image, blank line or end of text

local afterendafterEnd = mw.ustring.len(text) + 1

local blankposblankPosition = mw.ustring.find(text, "\n%s*\n") or afterendafterEnd -- position of next paragraph delimiter (or end of text)

local endposendPosition = math.min( -- find position of whichever comes first: [[File:, [[Image: or paragraph delimiter

mw.ustring.find(text, "%[%[%s*[Ff]ile%s*:") or afterendafterEnd,

mw.ustring.find(text, "%[%[%s*[Ii]mage%s*:") or afterendafterEnd,

blankposblankPosition)

token = mw.ustring.sub(text, 1, endposendPosition-1)

if blankposblankPosition < afterendafterEnd and blankposblankPosition == endposendPosition then -- paragraph ends with a blank line

token = token .. mw.ustring.match(text, "\n%s*\n", blankposblankPosition)

end

local isHatnote = not(leadstartleadStart) and mw.ustring.sub(token, 1, 1) == ':'

if not isHatnote then

leadstartleadStart = leadstartleadStart or mw.ustring.len(t) + 1 -- we got a paragraph, so mark the start of the lead section

paras = paras + 1

if allparasallParagraphs or (options.paraflags and options.paraflags[paras]) then t = t .. token end -- add if this paragraph wanted

end

end -- of "else got a paragraph"

Line 483:

text = mw.ustring.gsub(t, "\n+$", "") -- remove trailing line feeds, so "{{Transclude text excerpt|Foo}} more" flows on one line

return filetextfileText, text

end

Line 531:

-- Parse a ==Section== from a page

local function getSection(text, section, mainonlymainOnly)

local escapedSection = mw.ustring.gsub(mw.uri.decode(section), "([%^%$%(%)%%%.%[%]%*%+%-%?])", "%%%1") -- %26 → & etc, then ^ → %^ etc.

local level, content = mw.ustring.match(text .. "\n", "\n(==+)%s*" .. escapedSection .. "%s*==.-\n(.*)")

if not content then return nil end -- no such section

local nextsectionnextSection

if mainonlymainOnly then

nextsectionnextSection = "\n==.*" -- Main part of section terminates at any level of header

else

nextsectionnextSection = "\n==" .. mw.ustring.rep("=?", #level - 2) .. "[^=].*" -- "===" → "\n===?[^=].*", matching "==" or "===" but not "===="

end

content = mw.ustring.gsub(content, nextsectionnextSection, "") -- remove later sections with headings at this level or higher

return content

end

Line 547:

-- Remove unmatched <tag> or </tag> tags

local function fixTags(text, tag)

local startcountstartCount = 0

for i in mw.ustring.gmatch(text, "<%s*" .. tag .. "%f[^%w_].->") do startcountstartCount = startcountstartCount + 1 end

local endcountendCount = 0

for i in mw.ustring.gmatch(text, "<%s*/" .. tag .. "%f[^%w_].->") do endcountendCount = endcountendCount + 1 end

if startcountstartCount > endcountendCount then -- more <tag> than </tag>: remove the last few <tag>s

local i = 0

text = mw.ustring.gsub(text, "<%s*" .. tag .. "%f[^%w_].->", function(t)

i = i + 1

if i > endcountendCount then return "" else return nil end

end) -- "end" here terminates the anonymous replacement function(t) passed to gsub

elseif endcountendCount > startcountstartCount then -- more </tag> than <tag>: remove the first few </tag>s

text = mw.ustring.gsub(text, "<%s*/" .. tag .. "%f[^%w_].->", "", endcountendCount - startcountstartCount)

end

return text

Line 566:

-- Main function returns a string value: text of the lead of a page

local function main(pagenamespageNames, options)

if not pagenamespageNames or #pagenamespageNames < 1 then return err("No page names given") end

local pagenamepageName

local text

local pagecountpageCount = #pagenamespageNames

local firstpagefirstPage = pagenamespageNames[1] or "(nil)" -- save for error message, as it the name will be deleted

local gotoptgotOptions

local pageoptstrpageOptionsString

local section

-- read the page, or a random one if multiple pages were provided

if pagecountpageCount > 1 then math.randomseed(os.time()) end

while not text and pagecountpageCount > 0 do

local pagenumpageNumber = 1

if pagecountpageCount > 1 then pagenumpageNumber = math.random(pagecountpageCount) end -- pick a random title

pagenamepageName = pagenamespageNames[pagenumpageNumber]

if pagenamepageName and pagenamepageName ~= "" then

-- We have page or [[page]] or [[page|text]], possibly followed by |opt1|opt2...

local pn

pn, gotoptgotOptions, pageoptstrpageOptionsString = mw.ustring.match(pagenamepageName, "^%s*(%[%b[]%])%s*(|?)(.*)")

if pn then

pagenamepageName = mw.ustring.match(pn, "%[%[([^|%]]*)") -- turn [[page|text]] into page, discarding text

else -- we have page or page|opt...

pagenamepageName, gotoptgotOptions, pageoptstrpageOptionsString = mw.ustring.match(pagenamepageName, "%s*([^|]*[^|%s])%s*(|?)(.*)")

end

if pagenamepageName and pagenamepageName ~= "" then

local pn

pn, section = mw.ustring.match(pagenamepageName, "(.-)#(.*)")

pagenamepageName = pn or pagenamepageName

text, normalisedPagenamenormalisedPageName = getContent(pagenamepageName)

if not normalisedPagenamenormalisedPageName then

return err("No title for page name " .. pagenamepageName)

else

pagenamepageName = normalisedPagenamenormalisedPageName

end

if text and options.nostubs then

Line 607:

end

if not section then

section = mw.ustring.match(pagenamepageName, ".-#(.*)") -- parse redirect to Page#Section

end

if text and section and section ~= "" then text = getSection(text, section) end

end

end

if not text then table.remove(pagenamespageNames, pagenumpageNumber) end -- this one didn't work; try another

pagecountpageCount = pagecountpageCount - 1 -- ensure that we exit the loop after at most #pagenamespageNames iterations

end

if not text then return err("Cannot read a valid page: first name is " .. firstpagefirstPage) end

text = cleanupText(text, options)

local pageoptspageOptions = {} -- pageoptspageOptions (even if value is "") have priority over global options

for k, v in pairs(options) do pageoptspageOptions[k] = v end

if gotoptgotOptions and gotoptgotOptions ~= "" then

for _, t in pairs(mw.text.split(pageoptstrpageOptionsString, "|")) do

local k, v = mw.ustring.match(t, "%s*([^=]-)%s*=(.-)%s*$")

pageoptspageOptions[k] = v

end

pageoptspageOptions.paraflags = numberFlags(pageoptspageOptions["paragraphs"] or "") -- parse paragraphs, e.g. "1,3-5" → {"1","3-5"}

pageoptspageOptions.fileflags = numberFlags(pageoptspageOptions["files"] or "") -- parse file numbers

if pageoptspageOptions.more and pageoptspageOptions.more == "" then pageoptspageOptions.more = "Read more..." end -- more= is short for this default text

end

local filetextfileText

filetextfileText, text = parse(text, pageoptspageOptions)

-- replace the bold title or synonym near the start of the article by a wikilink to the article

local lang = mw.language.getContentLanguage()

local pos = mw.ustring.find(text, "'''" .. lang:ucfirst(pagenamepageName) .. "'''", 1, true) -- look for "'''Foo''' is..." (uc) or "A '''foo''' is..." (lc)

or mw.ustring.find(text, "'''" .. lang:lcfirst(pagenamepageName) .. "'''", 1, true) -- plain search: special characters in pagenamepageName represent themselves

if pos then

local len = mw.ustring.len(pagenamepageName)

text = mw.ustring.sub(text, 1, pos + 2) .. "[[" .. mw.ustring.sub(text, pos + 3, pos + len + 2) .. "]]" .. mw.ustring.sub(text, pos + len + 3, -1) -- link it

else -- look for anything unlinked in bold, assumed to be a synonym of the title (e.g. a person's birth name)

text = mw.ustring.gsub(text, "()'''(.-'*)'''", function(a, b)

if a < 100 and not mw.ustring.find(b, "%[") then ---if early in article and not wikilinked

return "'''[[" .. pagenamepageName .. "|" .. b .. "]]'''" -- replace '''Foo''' by '''[[pagenamepageName|Foo]]'''

else

return nil -- instruct gsub to make no change

Line 652:

-- remove '''bold text''' if requested

if is(pageoptspageOptions.nobold) then text = mw.ustring.gsub(text, "'''", "") end

text = filetextfileText .. text

-- Seek and destroy unterminated templates and wikilinks

Line 674:

text = fixTags(text, "div")

if pageoptspageOptions.more then text = text .. " '''[[" .. pagenamepageName .. "|" .. pageoptspageOptions.more .. "]]'''" end -- wikilink to article for more info

if pageoptspageOptions.list and not pageoptspageOptions.showall then -- add a collapsed list of pages which might appear

local listtext = pageoptspageOptions.list

if listtext == "" then listtext = "Other articles" end

text = text .. "{{collapse top|title={{resize|85%|" ..listtext .. "}}|bg=fff}}{{hlist"

for _, p in pairs(pagenamespageNames) do

if mw.ustring.match(p, "%S") then text = text .. "|[[" .. mw.text.trim(p) .. "]]" end

end

Line 702:

end

local pagenamespageNames = {}

if func == "lead" then

pagenamespageNames = { args[1] }

elseif func == "linked" or func == "listitem" then

-- Read named page and find its wikilinks

Line 721:

text = mw.ustring.gsub(text, "{{%s*[Aa]nnotated[ _]link%s*|%s*(.-)%s*}}", "[[%1]]")

if func == "linked" then

for p in mw.ustring.gmatch(text, "%[%[%s*([^%]|\n]*)") do table.insert(pagenamespageNames, p) end

else -- listitem: first wikilink on a line beginning *, :#, etc. except in "See also" or later section

text = mw.ustring.gsub(text, "\n== *See also.*", "")

for p in mw.ustring.gmatch(text, "\n:*[%*#][^\n]-%[%[%s*([^%]|\n]*)") do table.insert(pagenamespageNames, p) end

end

elseif func == "random" then

-- accept any number of page names. If more than one, we'll pick one randomly

for i, p in pairs(args) do

if p and type(i) == 'number' then table.insert(pagenamespageNames, p) end

end

elseif func == "selected" then

Line 737:

if articlekey == 0 then articlekey = articlecount end

end

pagenamespageNames = { args[articlekey] }

end

Line 748:

if options.showall then

local separator = ""

for _, p in pairs(pagenamespageNames) do

local t = main({ p }, options)

if t ~= "" then

Line 757:

end

else

text = main(pagenamespageNames, options)

end

Line 811:

if options.more and options.more == "" then options.more = "Read more..." end -- more= is short for this default text

local pagenamespageNames = { (article .. '#' .. (section or '')) }

local text = main(pagenamespageNames, options)

if text == "" and d.brokenCategory and d.brokenCategory ~= "" and mw.title.getCurrentTitle().isContentPage then

Line 847:

function p.parseimage(text, start) return parseImage(text, start) end

function p.cleanupText(text, options) return cleanupText(text, options) end

function p.main(pagenamespageNames, options) return main(pagenamespageNames, options) end

function p.numberflags(str) return numberFlags(str) end