Módulo:Webarchive
La decumentaçon pa este módulo puode ser criada na páigina Módulo:Webarchive/doc
--[[ ----------------------------------
Lua module implementing the {{webarchive}} template.
A merger of the functionality of three templates: {{wayback}}, {{webcite}} and {{cite archives}}
]]
local p = {}
--[[--------------------------< inlineError >-----------------------
Critical error. Render output completely in red. Add to tracking category.
]]
local function inlineError(arg, msg)
track["Catadorie:!Erros na predefeniçon Webarchive"] = 1
return '<span style="font-size:100%" class="error citation-comment">Erros na predefeniçon Webarchive: Berifique <code style="color:inherit; border:inherit; padding:inherit;">|' .. arg .. '=</code> valor. ' .. msg .. '</span>'
end
--[[--------------------------< inlineRed >-----------------------
Render a text fragment in red, such as a warning as part of the final output.
Add tracking category.
]]
local function inlineRed(msg, trackmsg)
if trackmsg == "warning" or trackmsg == "aviso" then
track["Catadorie:!Erros na predefeniçon Webarchive"] = 1
elseif trackmsg == "error" or trackmsg == "erro" then
track["Catadorie:!Erros na predefeniçon Webarchive"] = 1
end
return '<span style="font-size:100%" class="error citation-comment">' .. msg .. '</span>'
end
--[[--------------------------< trimArg >-----------------------
trimArg returns nil if arg is "" while trimArg2 returns 'true' if arg is ""
trimArg2 is for args that might accept an empty value, as an on/off switch like nolink=
]]
local function trimArg(arg)
if arg == "" or arg == nil then
return nil
else
return mw.text.trim(arg)
end
end
local function trimArg2(arg)
if arg == nil then
return nil
else
return mw.text.trim(arg)
end
end
--[[--------------------------< base62 >-----------------------
Convert base-62 to base-10
Credit: https://de.wikipedia.org/wiki/Modul:Expr
]]
local function base62( value )
local r = 1
if value:match( "^%w+$" ) then
local n = #value
local k = 1
local c
r = 0
for i = n, 1, -1 do
c = value:byte( i, i )
if c >= 48 and c <= 57 then
c = c - 48
elseif c >= 65 and c <= 90 then
c = c - 55
elseif c >= 97 and c <= 122 then
c = c - 61
else -- How comes?
r = 1
break -- for i
end
r = r + c * k
k = k * 62
end -- for i
end
return r
end
--[[--------------------------< tableLength >-----------------------
Given a 1-D table, return number of elements
]]
local function tableLength(T)
local count = 0
for _ in pairs(T) do count = count + 1 end
return count
end
--[[--------------------------< dateFormat >-----------------------
Given a date string, return its format: dmy, mdy, iso, ymd
If unable to determine return nil
]]
local function dateFormat(date)
local dt = {}
dt.split = {}
dt.split = mw.text.split(date, "-")
if tableLength(dt.split) == 3 then
if tonumber(dt.split[1]) > 1900 and tonumber(dt.split[1]) < 2200 and tonumber(dt.split[2]) and tonumber(dt.split[3]) then
return "iso"
else
return nil
end
end
dt.split = mw.text.split(date, " ")
if tableLength(dt.split) == 3 then
if tonumber(dt.split[3]) then
if tonumber(dt.split[3]) > 1900 and tonumber(dt.split[3]) < 2200 then
if tonumber(dt.split[1]) then
return "dmy"
else
return "mdy"
end
else
if tonumber(dt.split[1]) then
if tonumber(dt.split[1]) > 1900 and tonumber(dt.split[1]) < 2200 then
return "ymd"
end
end
end
end
end
return nil
end
--[[--------------------------< makeDate >-----------------------
Given a zero-padded 4-digit year, 2-digit month and 2-digit day, return a full date in df format
df = mdy, dmy, iso, ymd
]]
local function makeDate(year, month, day, df)
local month_names = {"janeiro", "febreiro", "márcio", "abril", "maio", "júnio", "júlio", "agosto", "setembre", "outubre", "nobembre", "dezembre"}
if not year or year == "" or not month or month == "" or not day or day == "" then
return nil
end
local zmonth = month -- month with leading 0
month = month:match("0*(%d+)") -- month without leading 0
if tonumber(month) < 1 or tonumber(month) > 12 then
return year
end
local nmonth = month_names[tonumber(month)]
-- local nmonth = os.date("%B", os.time{year=2000, month=month, day=1} ) -- month in name form
if not nmonth then
return year
end
local zday = day
day = zday:match("0*(%d+)")
if tonumber(day) < 1 or tonumber(day) > 31 then
if df == "mdy" or df == "dmy" then
return nmonth .. " " .. year
elseif df == "iso" then
return year .. "-" .. zmonth
elseif df == "ymd" then
return year .. " " .. nmonth
else
return nmonth .. " " .. year
end
end
if df == "mdy" then
return nmonth .. " " .. day .. ", " .. year -- setembre 1, 2016
elseif df == "dmy" then
return day .. " de " .. nmonth .. " de " .. year -- 1 de setembre de 2016
elseif df == "iso" then
return year .. "-" .. zmonth .. "-" .. zday -- 2016-09-01
elseif df == "ymd" then
return year .. " " .. nmonth .. " " .. cday -- 2016 setembre 1
else
return day .. " de " .. nmonth .. " de " .. year -- 1 de setembre de 2016
end
end
--[[--------------------------< decodeWebciteDate >-----------------------
Given a URI-path to Webcite (eg. /67xHmVFWP) return the encoded date in df format
]]
local function decodeWebciteDate(path, df)
local dt = {}
dt.split = {}
dt.split = mw.text.split(path, "/")
-- valid URL formats that are not base62
-- http://www.webcitation.org/query?id=1138911916587475
-- http://www.webcitation.org/1138911916587475
-- http://www.webcitation.org/cache/73e53dd1f16cf8c5da298418d2a6e452870cf50e
if dt.split[2] == "query" or dt.split[2] == "cache" or tonumber(dt.split[2]) then
return "query"
end
dt.full = os.date("%Y %m %d", string.sub(string.format("%d", base62(dt.split[2])),1,10) )
dt.split = mw.text.split(dt.full, " ")
dt.year = dt.split[1]
dt.month = dt.split[2]
dt.day = dt.split[3]
if not tonumber(dt.year) or not tonumber(dt.month) or not tonumber(dt.day) then
return inlineRed("[Erro data] (1)", "erro")
end
if tonumber(dt.month) > 12 or tonumber(dt.day) > 31 or tonumber(dt.month) < 1 then
return inlineRed("[Erro data] (2)", "erro")
end
if tonumber(dt.year) > tonumber(os.date("%Y")) or tonumber(dt.year) < 1900 then
return inlineRed("[Erro data] (3)", "erro")
end
fulldate = makeDate(dt.year, dt.month, dt.day, df)
if not fulldate then
return inlineRed("[Erro data] (4)", "erro")
else
return fulldate
end
end
--[[--------------------------< snapDateToString >-----------------------
Given a URI-path to Wayback (eg. /web/20160901010101/http://example.com )
return the formatted date eg. "September 1, 2016" in df format
Handle non-digits in snapshot ID such as "re_" and "-" and "*"
]]
local function decodeWaybackDate(path, df)
local snapdate, snapdatelong, currdate, fulldate
local safe = path
snapdate = string.gsub(safe, "^/w?e?b?/?", "") -- Remove leading "/web/" or "/"
safe = snapdate
local N = mw.text.split(safe, "/")
snapdate = N[1]
if snapdate == "*" then -- eg. /web/*/http..
return "index"
end
safe = snapdate
snapdate = string.gsub(safe, "[a-z][a-z]_[0-9]?$", "") -- Remove any trailing "re_" from date
safe = snapdate
snapdate = string.gsub(safe, "[-]", "") -- Remove dashes from date eg. 2015-01-01
safe = snapdate
snapdate = string.gsub(safe, "[*]$", "") -- Remove trailing "*"
if not tonumber(snapdate) then
return inlineRed("[Erro data] (2)", "erro")
end
local dlen = string.len(snapdate)
if dlen < 4 then
return inlineRed("[Erro data] (3)", "erro")
end
if dlen < 14 then
snapdatelong = snapdate .. string.rep("0", 14 - dlen)
else
snapdatelong = snapdate
end
local year = string.sub(snapdatelong, 1, 4)
local month = string.sub(snapdatelong, 5, 6)
local day = string.sub(snapdatelong, 7, 8)
if not tonumber(year) or not tonumber(month) or not tonumber(day) then
return inlineRed("[Erro data] (4)", "erro")
end
if tonumber(month) > 12 or tonumber(day) > 31 or tonumber(month) < 1 then
return inlineRed("[Erro data] (5)", "erro")
end
currdate = os.date("%Y")
if tonumber(year) > tonumber(currdate) or tonumber(year) < 1900 then
return inlineRed("[Erro data] (6)", "erro")
end
fulldate = makeDate(year, month, day, df)
if not fulldate then
return inlineRed("[Erro data] (7)", "erro")
else
return fulldate
end
end
--[[--------------------------< serviceName >-----------------------
Given a domain extracted by mw.uri.new() (eg. web.archive.org) set tail string and service ID
]]
local function serviceName(host, nolink)
local tracking = "Catadorie:!Predefeniçon Webarchive outros arquibos"
local bracketopen = "[["
local bracketclose = "]]"
if nolink then
bracketopen = ""
bracketclose = ""
end
ulx.url1.service = "outro"
ulx.url1.tail = " ne l " .. ulx.url1.host
if mw.ustring.find( host, "archive.org", 1, plain ) then
ulx.url1.service = "wayback"
ulx.url1.tail = " ne l " .. bracketopen .. "Wayback Machine" .. bracketclose
tracking = "Catadorie:!Predefeniçon Webarchive wayback links"
elseif mw.ustring.find( host, "webcitation.org", 1, plain ) then
ulx.url1.service = "webcite"
ulx.url1.tail = " ne l " .. bracketopen .. "WebCite" .. bracketclose
tracking = "Catadorie:!Predefeniçon Webarchive webcite links"
elseif mw.ustring.find( host, "archive.is", 1, plain ) then
ulx.url1.service = "archiveis"
ulx.url1.tail = " ne l " .. bracketopen .. "Archive.is" .. bracketclose
tracking = "Catadorie:!Predefeniçon Webarchive archiveis links"
elseif mw.ustring.find( host, "archive.today", 1, plain ) then
ulx.url1.service = "archiveis"
ulx.url1.tail = " ne l " .. bracketopen .. "Archive.is" .. bracketclose
tracking = "Catadorie:!Predefeniçon Webarchive archiveis links"
elseif mw.ustring.find( host, "archive-it", 1, plain ) then
ulx.url1.service = "archiveit"
ulx.url1.tail = " ne l " .. bracketopen .. "Archive-It" .. bracketclose
elseif mw.ustring.find( host, "arquibo.pt", 1, plain) then
ulx.url1.tail = " ne l " .. "Arquibo Web Mirandés"
elseif mw.ustring.find( host, "loc.gov", 1, plain ) then
ulx.url1.tail = " na " .. bracketopen .. "Biblioteca de l Cungresso" .. bracketclose
elseif mw.ustring.find( host, "webharvest.gov", 1, plain ) then
ulx.url1.tail = " ne ls " .. bracketopen .. "Arquibos Nacionales i Admenistraçon de Decumientos de ls Stados Ounidos" .. bracketclose
elseif mw.ustring.find( host, "bibalex.org", 1, plain ) then
ulx.url1.tail = " na " .. "[[Bibliotheca_Alexandrina]]"
elseif mw.ustring.find( host, "collectionscanada", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo Web de l Gobierno Canadiano"
elseif mw.ustring.find( host, "haw.nsk", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo Web Croata (HAW)"
elseif mw.ustring.find( host, "nlib.ee", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo web Stoniano"
elseif mw.ustring.find( host, "vefsafn.is", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo Web Eislandés"
elseif mw.ustring.find( host, "proni.gov", 1, plain ) then
ulx.url1.tail = " ne l " .. bracketopen .. "Cunserbatória de Registos Públicos de la Irlanda de l Norte" .. bracketclose
elseif mw.ustring.find( host, "uni-lj.si", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo Web Slobeno"
elseif mw.ustring.find( host, "stanford.edu", 1, plain ) then
ulx.url1.tail = " nas " .. "Bibliotecas de la Ounibersidade de Stanford"
elseif mw.ustring.find( host, "nationalarchives.gov.uk", 1, plain ) then
ulx.url1.tail = " ne l " .. bracketopen .. "Arquibo Web de l Gobierno de l Reino Ounido" .. bracketclose
elseif mw.ustring.find( host, "parliament.uk", 1, plain ) then
ulx.url1.tail = " ne l " .. bracketopen .. "Arquibo Web de l Parlamiento de l Reino Ounido" .. bracketclose
elseif mw.ustring.find( host, "webarchive.org.uk", 1, plain ) then
ulx.url1.tail = " ne l " .. bracketopen .. "Arquibo Web de l Reino Ounido" .. bracketclose
elseif mw.ustring.find( host, "nlb.gov.sg", 1, plain ) then
ulx.url1.tail = " ne l " .. "Arquibo Web de la Singapura"
elseif mw.ustring.find( host, "screenshots.com", 1, plain ) then
ulx.url1.tail = " ne l " .. "Screenshots"
elseif mw.ustring.find( host, "wikiwix.com", 1, plain ) then
ulx.url1.tail = " ne l " .. "Wikiwix"
elseif mw.ustring.find( host, "perma.cc", 1, plain ) then
ulx.url1.tail = " ne l " .. bracketopen .. "Perma.cc" .. bracketclose
else
tracking = "Catadorie:!Predefeniçones Webarchive arquibos çconhecidos"
end
track[tracking] = 1
end
--[[--------------------------< parseExtraArgs >-----------------------
Parse numbered arguments starting at 2, such as url2..url10, date2..date10, title2..title10
For example: {{webarchive |url=.. |url4=.. |url7=..}}
Three url arguments not in numeric sequence (1..4..7).
Function only processes arguments numbered 2 or greater (in this case 4 and 7)
It creates numeric sequenced table entries like:
urlx.url2.url = <argument value for url4>
urlx.url3.url = <argument value for url7>
Returns the number of URL arguments found numbered 2 or greater (in this case returns "2")
]]
local function parseExtraArgs()
local i, j, argurl, argurl2, argdate, argtitle
j = 2
for i = 2, maxurls do
argurl = "url" .. i
if trimArg(args[argurl]) then
argurl2 = "url" .. j
ulx[argurl2] = {}
ulx[argurl2]["url"] = args[argurl]
argdate = "data" .. j
if trimArg(args[argdate]) then
ulx[argurl2]["date"] = args[argdate]
else
ulx[argurl2]["date"] = inlineRed("[Falta data]", "aviso")
end
argtitle = "título" .. j
if trimArg(args[argtitle]) then
ulx[argurl2]["title"] = args[argtitle]
else
ulx[argurl2]["title"] = nil
end
j = j + 1
end
end
if j == 2 then
return 0
else
return j - 2
end
end
--[[--------------------------< comma >-----------------------
Given a date string, return "," if it's MDY
]]
local function comma(date)
local N = mw.text.split(date, " ")
local O = mw.text.split(N[1], "-") -- for ISO
if O[1] == "index" then return "" end
if not tonumber(O[1]) then
return ","
else
return ""
end
end
--[[--------------------------< createTracking >-----------------------
Return data in track[] ie. tracking categories
]]
local function createTracking()
local sand = ""
if tableLength(track) > 0 then
for key,_ in pairs(track) do
sand = sand .. "[[" .. key .. "]]"
end
end
return sand
end
--[[--------------------------< createRendering >-----------------------
Return a rendering of the data in ulx[][]
]]
local function createRendering()
local sand, displayheader, displayfield
local period1 = "" -- For backwards compat with {{wayback}}
local period2 = "."
local indexstr = "arquibado"
if ulx.url1.date == "index" then
indexstr = "arquibo"
end
-- For {{wayback}}, {{webcite}}
if ulx.url1.format == "none" then
if not ulx.url1.title and not ulx.url1.date then -- No title. No date
sand = "[" .. ulx.url1.url .. " Cópia de l'arquibo]" .. ulx.url1.tail
elseif not ulx.url1.title and ulx.url1.date then -- No title. Date.
if ulx.url1.service == "wayback" then
period1 = "."
period2 = ""
end
sand = "[" .. ulx.url1.url .. " Arquibado an] " .. ulx.url1.date .. comma(ulx.url1.date) .. ulx.url1.tail .. period1
elseif ulx.url1.title and not ulx.url1.date then -- Title. No date.
sand = "[" .. ulx.url1.url .. " " .. ulx.url1.title .. "]" .. ulx.url1.tail
elseif ulx.url1.title and ulx.url1.date then -- Title. Date.
sand = "[" .. ulx.url1.url .. " " .. ulx.url1.title .. "]" .. ulx.url1.tail .. " (" .. indexstr .. " " .. ulx.url1.date .. ")"
else
return nil
end
if ulx.url1.extraurls > 0 then -- For multiple archive URLs
local tot = ulx.url1.extraurls + 1
sand = sand .. period2 .. " Arquibos adicionales: "
for i=2,tot do
local indx = "url" .. i
if ulx[indx]["title"] then
displayfield = "title"
else
displayfield = "date"
end
sand = sand .. "[" .. ulx[indx]["url"] .. " " .. ulx[indx][displayfield] .. "]"
if i == tot then
sand = sand .. "."
else
sand = sand .. ", "
end
end
else
return sand
end
return sand
-- For {{cite archives}}
else
if ulx.url1.format == "addlarchives" then -- Multiple archive services
displayheader = "Arquibos adicionales: "
else -- Multiple pages from the same archive
displayheader = "Páginas arquivadas em " .. ulx.url1.date .. ": "
end
local tot = 1 + ulx.url1.extraurls
local sand = displayheader
for i=1,tot do
local indx = "url" .. i
displayfield = ulx[indx]["title"]
if ulx.url1.format == "addlarchives" then
if not displayfield then
displayfield = ulx[indx]["date"]
end
else
if not displayfield then
displayfield = "Página " .. i
end
end
sand = sand .. "[" .. ulx[indx]["url"] .. " " .. displayfield .. "]"
if i == tot then
sand = sand .. "."
else
sand = sand .. ", "
end
end
return sand
end
end
function p.webarchive(frame)
args = frame.args
if (args[1]==nil) and (args["url"]==nil) then -- if no argument provided than check parent template/module args
args = frame:getParent().args
end
local tname = "Webarchive" -- name of calling template. Change if template rename.
ulx = {} -- Associative array to hold template data
track = {} -- Associative array to hold tracking categories
maxurls = 10 -- Max number of URLs allowed.
local verifydates = "yes" -- See documentation. Set "no" to disable.
-- URL argument (first)
local url1 = trimArg(args.url) or trimArg(args.url1)
if not url1 then
return inlineError("url", "Vazio.") .. createTracking()
end
if mw.ustring.find( url1, "https://web.http", 1, plain ) then -- track bug
track["Catadorie:!Erros na predefeniçon Webarchive"] = 1
return inlineError("url", "https://web.http") .. createTracking()
end
ulx.url1 = {}
ulx.url1.url = url1
local uri1 = mw.uri.new(ulx.url1.url)
ulx.url1.host = uri1.host
ulx.url1.extraurls = parseExtraArgs()
-- Nolink argument
local nolink = trimArg2(args.nolink)
serviceName(uri1.host, nolink)
-- Date argument
local date = trimArg(args.date) or trimArg(args.date1)
if date == "*" and ulx.url1.service == "wayback" then
date = "index"
elseif date and ulx.url1.service == "wayback" and verifydates == "yes" then
local ldf = dateFormat(date)
if ldf then
local udate = decodeWaybackDate( uri1.path, ldf )
if udate ~= date then
date = udate .. inlineRed("<sup>[Erro data trocada]</sup>", "aviso")
end
end
elseif date and ulx.url1.service == "webcite" and verifydates == "yes" then
local ldf = dateFormat(date)
if ldf then
local udate = decodeWebciteDate( uri1.path, ldf )
if udate == "query" then -- skip
elseif udate ~= date then
date = udate .. inlineRed("<sup>[Erro data trocada]</sup>", "aviso")
end
end
elseif not date and ulx.url1.service == "wayback" then
date = decodeWaybackDate( uri1.path, "iso" )
if not date then
date = inlineRed("[Erro data] (1)", "erro")
end
elseif not date and ulx.url1.service == "webcite" then
date = decodeWebciteDate( uri1.path, "iso" )
if date == "query" then
date = inlineRed("[Falta a data]", "aviso")
elseif not date then
date = inlineRed("[Erro data] (1)", "erro")
end
elseif not date then
date = inlineRed("[Falta la data]", "abiso")
end
ulx.url1.date = date
-- Format argument
local format = trimArg(args.format)
if not format then
format = "none"
else
if format == "addlpages" then
if not ulx.url1.date then
format = "none"
end
elseif format == "addlarchives" then
format = "addlarchives"
else
format = "none"
end
end
ulx.url1.format = format
-- Title argument
local title = trimArg(args.title) or trimArg(args.title1)
ulx.url1.title = title
local rend = createRendering()
if not rend then
rend = '<span style="font-size:100%" class="error citation-comment">Erro na [[:Predefeniçon:' .. tname .. ']]: Porblema çconhecido. Por fabor relate na páigina de çcusson desta predefeniçon.</span>'
track["Catadorie:!Predefeniçones cun erros Webarchive"] = 1
end
return rend .. createTracking()
end
return p