User:Js/urldecoder.js: Difference between revisions

Revision as of 15:43, 25 April 2009

Latest revision as of 17:20, 18 September 2019

(17 intermediate revisions by 6 users not shown)

Line 1:

/*jshint strict: false, asi: true, eqeqeq: false, curly: false, laxbreak:true, forin:false, boss:true, funcscope:true, noempty: false */

//[[user:js/urldecoder]]

/*global mw, $, urlDecoderCustom, urlDecoderPrefixes, urlDecoderEngNames */

var btn = newToolbarBtn;

function urlDecoderButton(){

function addUrlDecoderButton(){

var tlb = document.getElementById('toolbar')

btn(

if (!tlb) return

'urlDecoder',

var bt = document.createElement('input')

urlDecoderRun,

bt.type = 'button'; bt.onclick = urlDecoderRun

'//upload.wikimedia.org/wikipedia/commons/9/91/Link_go.png',

bt.value = '→[\[]]'; bt.id = 'urlDecoder'

bt.title = 'Decode URL before cursor or all URLs in selected text',

window.urlDecoderKey

bt.style.cssText = 'background:#adbede; height:22px; vertical-align:top; padding:0'

);

tlb.appendChild(bt)

if (!window.urlDecoderKey) return

bt.accessKey = urlDecoderKey

bt.title += ' ['+urlDecoderKey+']'

updateTooltipAccessKeys([bt])

}

if (wgAction=='edit' || wgAction=='submit') addOnloadHook(urlDecoderButton)

mw.loader.using( [ 'user.options', 'jquery.textSelection', 'mediawiki.util' ], function () {

/* Check if view is in edit mode and that the required modules are available. Then, customize the toolbar . . . */

if ( $.inArray( mw.config.get( 'wgAction' ), [ 'edit', 'submit' ] ) !== -1 ) {

$( function () {

if ( mw.user.options.get('usebetatoolbar') ) {

mw.loader.using( 'ext.wikiEditor', addUrlDecoderButton );

} else {

btn = oldToolbarBtn;

addUrlDecoderButton();

}

} );

}

// Add the customizations to LiquidThreads' edit toolbar, if available

mw.hook( 'ext.lqt.textareaCreated' ).add( addUrlDecoderButton );

} );

function newToolbarBtn(bId, bFunc, bIcon, bTitle){

var msg = {}; msg[bId] = bTitle; mw.messages.set(msg) // mw.usability.addMessages(msg) doesn't work

$('#wpTextbox1').wikiEditor('addToToolbar', {

section:'main', group:'insert', tools: {

bId:{

type: 'button',

action: {type:'callback', execute: bFunc},

labelMsg: bId,

icon: bIcon

}}})

}

function oldToolbarBtn(bId, bFunc, bIcon, bTitle, bKey){

var btn = $('<img class=mw-toolbar-custombutton id="' + bId + '">')

.attr({ src: bIcon, title: bTitle, alt: bTitle.substr(0,3) })

.css({ height:'20px', 'background-color':'#bce', border:'1px outset #bce', margin:'0 1px', cursor:'pointer'})

.click(bFunc)

.appendTo('#toolbar')

if( bKey ){

btn.attr({ accesskey: bKey, title: bTitle + ' ['+bKey+']' })

.updateTooltipAccessKeys();

}

}

/**/

function urlDecoderRun(){ //main function

var httpRegExp = '(https?:\\/\\/[^\\]\\[\\n\\r<>" ]+)' // except []<>"

var beforeCursor = new RegExp('(\\[{0,2})'+httpRegExp+'( +[^\\]\n]+)?\\]{0,2}$', 'i')

var localPrefix = getPrefixes(wgServer.substring(7))

if (wgServer == 'https://secure.wikimedia.org')

localPrefix = getPrefixes(wgScript.split('/')[2]+'.'+wgScript.split('/')[1]+'.org')

var newText, linkSize, txtarea = document.editform.wpTextbox1

var isBeforeCursor = false

//WMF domains mess

if (document.selection) { //IE/Opera

//2nd-lvl domains; secure link: .../wikipedia/mediawiki, .../wikipedia/foundation

var scrollTop = document.documentElement.scrollTop

var wmDomain = {

txtarea.focus()

mediawiki: 'mw',

range = document.selection.createRange()

wikimediafoundation: 'foundation' }

if (!range.moveStart) return

//2nd-lvl domains with multiple languages; secure link: wikinews/en

if (range.text){

var wmDomainM = {

newText = processSelText(range.text)

wikipedia:'w',

}else { //no selection

wikibooks:'b',

if (!(rr=range.duplicate())) return

wikinews:'n',

rr.moveStart('character', - 1000)

wikiquote:'q',

linkSize = processBeforeCursor(rr.text)

wikisource:'s',

if (!linkSize) return

wikiversity:'v',

range.moveStart('character', - linkSize) //select matched

wiktionary:'wikt'}

}

//3rd-lvl domains on .wikimedia.org; however secure link is wikipedia/*

//replace text

var wmSubDomains = /^(meta|commons|incubator|species|strategy)$/

if (newText != range.text){

//Exceptions:

range.text = newText

// https:/.../wikipedia/sources/wiki/Main_Page - not recognized by script

if (navigator.userAgent.indexOf('MSIE') != -1) newText = newText.replace(/\r/g,'') //for IE: do not count \r

range.moveStart('character', - newText.length)

range.select()

}

document.documentElement.scrollTop = scrollTop //restore window scroll position

}else if (txtarea.selectionStart || txtarea.selectionStart == '0') { // Mozilla

var scrollTop = txtarea.scrollTop, txt = txtarea.value

txtarea.focus()

var startPos = txtarea.selectionStart, endPos = txtarea.selectionEnd

if (startPos != endPos){

newText = processSelText(txt.substring(startPos, endPos))

}else{ //no selection

linkSize = processBeforeCursor(txt.substring((endPos-1000>0?endPos-1000:0), endPos))

if (!linkSize) return

startPos = endPos - linkSize //select matched

}

//replace text

if (newText != txt.substring(startPos, endPos)){

txtarea.value = txt.substring(0, startPos) + newText + txt.substring(endPos, txtarea.value.length)

txtarea.selectionEnd = startPos + newText.length

txtarea.selectionStart = startPos

}

txtarea.scrollTop = scrollTop

}//end of main function

var httpRegExp = '((?:https?:)?\\/\\/[^\\]\\[\\n\\r<>" ]+)' // any chars except []<>" and \n and spaces

var localPrefix = WMPrefixes( mw.config.get( 'wgServer' ).replace(/^\/\//,'http://') + mw.config.get( 'wgScript' ) )

var oldText, newText, isBeforeCursor, colonNS

var tbox = $('#wpTextbox1').focus()

oldText = tbox.textSelection( 'getSelection' )

var rx

if( oldText ){ //there was selection

rx = RegExp('(\\[{0,2})' + httpRegExp + '([^\\]\\[\\n\\r]*?\\]\\]?)?', 'ig')

newText = oldText.replace(rx, simplifyMatched)

if( window.urlDecoderIntLinks ){

var ut = '(' + mw.config.get( 'wgFormattedNamespaces' )[3].replace(/ /g,'_') + '|user_talk)' //both localized and canonical 'user_talk'

ut = RegExp ('\\[\\[' + ut.toLowerCase() + ':[^#]+$', 'i')

newText = newText.replace(/\[\[[^\]\|\n]+/g, function(lnk){

return ut.test(lnk) ? lnk : decodeAnchor(lnk) // skip user_talk, usually found in signatures

})

}

if( newText == oldText) return

}else{ //process text before cursor

function processBeforeCursor(str){ //looks for http: inside string, returns it's distance to cursor, also sets newText var

isBeforeCursor = true

var pos = str.lastIndexOf('http://') //to cut out possible preceding http links

if (pos == -1) pos = str.lastIndexOf('https://')

if (pos == -1) return 0

else if (pos > 10) str = str.substring(pos-2) //cut chars before http, leaving some room for possible [

var ma = str.match(beforeCursor) // result: (whole string)' '[', 'http:...', ' name]'

if (!ma) return 0

if (ma[3]) //link with name: add brackets just in case

newText = decodeMatchedLink(ma[0], '[', ma[2], ma[3]+']')

else //just url: add closing bracket if there is leading bracket

newText = decodeMatchedLink(ma[0], ma[1], ma[2], ma[1]?']':'')

return ma[0].length

}

//move back enough characters

function processSelText(txt){

var caretPos = tbox.textSelection('getCaretPosition')

txt = txt.replace(RegExp('(\\[{0,2})' + httpRegExp + '([^\\]\\[\\n\\r]*?\\]\\]?)?', 'ig'),

var beginPos = caretPos - 2000

decodeMatchedLink)

if( beginPos < 0 ) beginPos = 0

if (window.urlDecoderIntLinks) txt = txt.replace(/\[\[[^\]\|\n]+/g, decodeWikiLink)

tbox.textSelection( 'setSelection', {start:beginPos, end:caretPos} )

return txt

oldText = tbox.textSelection( 'getSelection' )

tbox.textSelection( 'setSelection', {start:caretPos, end:caretPos} )

//try to find http in oldText

rx = new RegExp('(\\[{0,2})'+httpRegExp+'( +[^\\]\n]+)?\\]{0,2}$', 'i')

var ma = rx.exec( oldText ) // result: (whole string)' '[', 'http:...', ' name]'

if( !ma ) return

oldText = ma[0]

if( ma[3] ) //link with name: automatically add brackets

newText = simplifyMatched(ma[0], '[', ma[2], ma[3]+']')

else //just url: add closing bracket only if there is leading bracket

newText = simplifyMatched(ma[0], ma[1], ma[2], ma[1]?']':'')

if( oldText == newText ) return

tbox.textSelection( 'setSelection', {start: caretPos - oldText.length, end: caretPos} )

}

//replace text

tbox.textSelection( 'encapsulateSelection', {replace:true, peri:newText} )

//end of main code

return

//---FUNCTIONS

function simplifyMatched(str, bracket, url, rest){//arguments: (whole string), '[', url, ' name]'

if( !bracket ){//no brackets, just url

var trail = RegExp(

'['

+ ',;\\\\.:!\\?' //trailing punctuation, per Parser.php

+ ( /\(/.test(url) ? '' : '\\)' ) //also closing bracket without opening bracket

+ ']+$'

+ "|''+$" //or possible bold/italic at the end of url

)

.exec( url )

if( trail ){

url = url.substring( 0, url.length - trail[0].length ) //move these out of url

}

return decodeUrl(url) + str.substring(url.length)

}else if( rest ){ //both brackets and possibly name

return decodeUrl(url, rest.replace(/\]+$|^ +| +$/g,'')) //trim ending brackets and spaces in 'name]'

}else{

return str //probably broken wikicode in selected text

}

function decodeMatchedLink(str, bracket, url, rest){//gets: (whole string), '[', url, ' name]'

if (!bracket) return decodeLink(url) + str.substring(url.length) //just url

if (!rest) return str //no matching closing bracket

var name = rest.replace(/^ +|\]+$/g,'') //trim leading spaces and ending brackets

str = decodeLink(url, name)

if (str.charAt(0) != '[' ) str = '['+str+']' //not wikilink - restore single brackets

return str

}

function decodeLink(url, name){ //name is optional

var parts, endingDots, result

if (name) name = name.replace(/^ +| +$/g,'') //trim spaces

//secure Wikimedia link into non-secure

url = url.replace(/https:\/\/secure.wikimedia.org\/(\w+)\/(\w+)\/([^\]\|\n\r ]+)/, 'http://$2.$1.org/$3')

//bugzilla

if (parts = url.match(/^https:\/\/bugzilla\.wikimedia.org\/show_bug\.cgi\?id=(\d+)/))

return '[\[mediazilla:'+parts[1]+ (name?'|'+name:'') + ']]'

function decodeUrl(url, name){ //url -> %-decoded -> [[link|name]] (if possible); name is optional

//decode %

if (url.indexOf('%') != -1) try { url = decodeURI(url) } catch(e){}

var decodingFailed //need to skip some strange percent-encoded URIs

url = url.replace(/%(3B|2F|2C|3A)/g, function(s){return decodeURIComponent(s)}) // ;/,:

//encode back some chars not allowed by MediaWiki

//percent-decoding

url = url.replace(/[ <>"\[\]]/g, function(s){return encodeURIComponent(s)}) //"

if( url.indexOf('%') != -1 )

try {

if (isBeforeCursor)

url = decodeURI(url)

for (var n in window.urlDecoderEngNames) //to eng keywords

url = url.replace(/%(3B|2F|2C|3A)/g, decodeURIComponent) //decode ;/,:

url = url.replace(/[ <>"\[\]|]/g, encodeURIComponent) //" some disallowed chars, and pipe can screw template params

} catch(e){

decodingFailed = true

}

if( isBeforeCursor ) //user-defined conversion to eng keywords

for( var n in window.urlDecoderEngNames )

url = url.replace(RegExp('(title=|wiki\/)('+urlDecoderEngNames[n]+':)'), '$1' + n + ':')

//analyze

//try converting to internal link

parts = url.substring(7).split('/')

if( !decodingFailed && !/(\}\}|\|)$/.test(url) ) //trailing | or }} could mean a part of a template, skip to be safe

var ___domain = parts[0]

var linkPrefixlink = getPrefixestoWikilink(___domainurl)

//return unchanged if cannot convert into wiki link

//user-defined function

if (!linkPrefix || parts[1] != 'wiki' || url.indexOf('?')!=-1)

if( window.urlDecoderCustom ){

return name ? '[' + url + ' ' + name + ']' : url

url = urlDecoderCustom(url)

//convert to wiki link

if( ! /^((?:https?:)?\/\/|\{\{)/.test(url) ) link = url //was converted to internal link

var title = url.substring(parts[0].length + parts[1].length + 9) //get part after /wiki/

}

title = decodeWikiLink(title)

//return internal link

if( link ){

link = link.replace(/%(3f|26|22)/ig, decodeURIComponent) //decode ?&"

if( ( mw.config.get( 'wgNamespaceNumber' ) === 0 || mw.config.get( 'wgNamespaceNumber' ) === 14 ) && isBeforeCursor )

link = link.replace(/^:/,'') //probably user adding interwiki

return '[' + '[' + link + (name?'|'+name:'') + ']]'

}

//or return external link

if( typeof name == 'string' ){

if( isBeforeCursor ) url = url.replace(/''/g,'%27%27') //techically '' should stop URL, but more likely it's part of it

return '[' + url + (name?' '+name:'') + ']' //empty name

}else{

return url

}

}

function toWikilink(url){ // 'http://xx.wikipedia.org/wiki/YY' -> xx:YY

//add bugzilla to user-defined prefixes

window.urlDecoderPrefixes = $.extend( window.urlDecoderPrefixes,

{ 'https://bugzilla.wikimedia.org/show_bug.cgi?id=' : 'mediazilla' } )

//apply user-defined prefixes

for( var key in urlDecoderPrefixes )

if( url.toLowerCase().indexOf(key) != -1 )

return urlDecoderPrefixes[key] + ':' + url.substring( url.indexOf(key) + key.length )

//check if we can convert to internal link with WM prefixes

var ma = /^((?:https?:)?\/\/[^\/]+)\/wiki\/([^?]+)$/.exec( url )// 1:'http://___domain.org' 2:part after /wiki/

if( !ma ) return null

var linkPrefix = WMPrefixes( ma[1] )

if( !linkPrefix) return null

//convert to internal

var title = decodeAnchor( ma[2] )

var prefixes = ''

if ( linkPrefix[0] && (linkPrefix[0] != localPrefix[0]) ) prefixes += linkPrefix[0]

if ( linkPrefix[1] && (linkPrefix[1] != localPrefix[1]) ) prefixes += ':' + linkPrefix[1]

if ( prefixes || isColonNeeded(title) ) prefixes += ':' //colon after prefix or leading colon on cat/file link

return prefixes + title

if (!prefixes && isColonNeeded(title)) prefixes = ':' //leading colon for cat/file

//return result

result = '[\[' + prefixes + title

if (name) result += '|' + name + ']]'

else result = (result + ']]').replace(/(\.+)\]\]/,']]$1') //url. -> [[w]]. not [[w.]]

return result

}

function decodeWikiLink(link){

link = link.replace(/_/g, ' ').replace(/^ +| +$/g, '')

function decodeAnchor(link){//simplify internal link: replace %20 and _ then decode anchor

link = link.replace(/(_|%20)/g, ' ').replace(/^ +| +$/g, '')

var parts = link.split('#')

if ( parts.length != 2 ) return link //no anchor

var anchor = parts[1], hidIdx = -1, hidden = []

link = parts[0]

anchor = anchor.replace( //hide IPs

/(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)/,

function(s){ hidden[++hidIdx] = s; return '\x01' + hidIdx + '\x02' }

)

//decode 4, 3 and 2-byte: http://en.wikipedia.org/wiki/UTF-8

anchor = anchor.replace(/\.F[0-4]\.[89AB][\dA-F]\.[89AB][\dA-F]\.[89AB][\dA-F]/g, deChar)

anchor = anchor.replace(/\.E[\dA-F]\.[89AB][\dA-F]\.[89AB][\dA-F]/g, deChar)

anchor = anchor.replace(/\.[CD][\dA-F]\.[89AB][\dA-F]/g, deChar)

anchor = anchor.replace( //hide IPs

//decode reserved 1-byte chars: http://meta.wikimedia.org/wiki/Url , http://en.wikipedia.org/wiki/Percent_encoding

/(?:^|[^0-9A-F\.])(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)/,

anchor = anchor.replace(/\.[2-5][0-9A-F]/g, function(str){

function(s){ hidden[++hidIdx] = s; return '\x01' + hidIdx + '\x02' }

var ch = deChar(str) // ;:@&=+$,/?%#[]<>

)

if ((ch != ':') &&

//decode 1-byte chars: all symbols except -.:_ and []{} prohibited in links

((encodeURIComponent(ch).replace(/%/,'.') == str) || ('!*\'()'.indexOf(ch) != -1)))

anchor = anchor.replace(/\.[2-7][0-9A-F]/g, function(hhh){

return ch

var ch = deChar(hhh)

else

if( '!"#$%&\'()*+,/;<=>?@\\^`~'.indexOf(ch) >= 0 ) return ch

return str //do not decode e.g. %35 which is '5' and not supposed to be encoded

else return hhh

})

//unhide IPs and return

function deChar(str){ return tryDecode(str.replace(/\.([0-9A-F][0-9A-F])/g, '%$1')) }

for( var i=hidIdx; i>=0; i-- ) anchor = anchor.replace('\x01'+i+'\x02', hidden[i])

function tryDecode(s){ try {s = decodeURIComponent(s)} catch(e){}; return s }

if( anchor.indexOf("''") != -1 ) return link //cannot have double '' in link

//unhide IPs

else return parts[0] + '#' + anchor

for (var i=hidIdx; i>=0; i--) anchor = anchor.replace('\x01'+i+'\x02', hidden[i])

//return

return link + '#' + anchor

}

function deChar(ss){

function getPrefixes(___domain){ //en.wikipedia.org -> [ 'w', 'en'] , using projPrefix{}

try{ss = decodeURIComponent(ss.replace(/\.([0-9A-F][0-9A-F])/g, '%$1'))} catch(e){}

var projPrefix = {

return ss

'test.wikipedia.org':'testwiki',

}

'wikipedia.org':'w',

'wikibooks.org':'b',

'wikinews.org':'n',

'wikiquote.org':'q',

'wikisource.org':'s',

'wikiversity.org':'v',

'wiktionary.org':'wikt',

'mediawiki.org':'mw',

'www.wikimedia.org':'foundation',

'wikimedia.org':'*',

'wikimediafoundation.org':'foundation'

}

___domain = ___domain.toLowerCase()

var proj, isFound

for (proj in projPrefix) if (___domain.indexOf(proj) != -1) {isFound=true; break}

if (!isFound) return null

var prefix = projPrefix[proj], lang = ''

if (prefix == '*') prefix = ___domain.split('.')[0] // .wikimedia.org project

else if ((___domain=___domain.replace(proj, '')) && (___domain != 'www.')) lang = ___domain.split('.')[0] //multi-lang project

//else if (prefix.charAt(0) == ':'){ //multi-lang project if (pp.length == 3) lang = pp[0]

return [prefix, lang]

}

function isColonNeeded(pg){

if (pg.indexOf(':')==-1) return false

urlDecoderNS = window.urlDecoderNS || {}

urlDecoderNS.en = 'image'

var ns = urlDecoderNS[wgContentLanguage]

if (typeof ns != 'string') ns = requestColonNS()

return RegExp('^('+ns+'|file|category) *:','i').test(pg)

}

function requestColonNS(){

var whatsthis = ' &nbsp; &nbsp; <a href="http://en.wikipedia.org/wiki/user:js/urldecoder#Localization" target=_blank>(?)</a>'

showMsg('Requesting namespaces...'+whatsthis)

var aj = sajax_init_object(), q = null, ns = []

aj.open('GET', '/w/api.php?format=json&action=query&meta=siteinfo&siprop=namespaces|namespacealiases', false)

aj.send(null)

try { eval('q='+aj.responseText); q = q.query

} catch(e){return null}

ns.push(q.namespaces[6]['*']); ns.push(q.namespaces[14]['*'])

for (var k in q.namespacealiases)

if (q.namespacealiases[k].id==6 || q.namespacealiases[k].id==14)

ns.push(q.namespacealiases[k]['*'])

ns = ns.join('|').toLowerCase()

urlDecoderNS[wgContentLanguage] = ns

showMsg("<code>urlDecoderNS = {'"+wgContentLanguage+"':'"+ns+"'}<code>"+whatsthis)

return ns

}

function showMsg(htm){

var dv = document.getElementById('edit-msg')

if (!dv){

dv = document.createElement('div')

dv.id = 'edit-msg'; dv.style.cssText = 'margin:5px; border:1px solid gray'

var tbox = document.getElementById('wpTextbox1')

tbox.parentNode.insertBefore(dv, tbox)

}

dv.innerHTML = htm; dv.style.display = ''

setTimeout(hideMsg, 15000)

}

function WMPrefixes(url){ // http: //en.wikipedia.org/wiki/... -> [ 'w', 'en']

function hideMsg(){

var dv = document.getElementById('edit-msg')

var dd = /^(?:https?:)?\/\/([a-z\.]+)\.org/.exec( url.toLowerCase() )

dv.style.display = 'none'

if( !dd ) return null

dd = dd[1].split('.') //domains, e.g. ['en','wikipedia']

if( dd.length > 2 ) return null //too many subdomains, possibly mobile site XX.m.wikipedia.org/

var lang = '', proj = '', ___domain = dd.pop(), subdomain = dd.pop()

if( subdomain == 'www' ) subdomain = ''

if( ___domain == 'wikimedia' ){ // *.wikimedia.org

if( !subdomain )

proj = 'foundation'

else if( wmSubDomains.test(subdomain) )

proj = subdomain

else

return null

}else if( (proj = wmDomain[___domain]) && !subdomain ){ // mediawiki.org & wikimediafoundation.org

//done: proj is set

}else if( proj = wmDomainM[___domain] ){ //multi-lang domains

if( !subdomain );

//done: e.g. 'wikisource.org'

else if( proj == 'w' && subdomain == 'test' )

proj = 'testwiki'

else if( subdomain.length >= 2 )

lang = subdomain

else

return null

}else return null //unrecognized ___domain

return [proj, lang]

}

function isColonNeeded(pg){

if( ! /:/.test(pg) ) return false

if( ! colonNS ){ //define list of all possible category and file namespaces

var list = ['file', 'category'] //canonical aliases

var namespaces = mw.config.get( 'wgNamespaceIds' );

for( var name in namespaces )

if( (namespaces[name]==6 || namespaces[name]==14) && $.inArray(name, list) == -1 )

list.push(name)

colonNS = RegExp( '^(' + list.join('|') + ') *:', 'i')

}

return colonNS.test( $.trim(pg) )

}

}