User:Js/urldecoder.js: Difference between revisions

Revision as of 21:35, 15 July 2010

Revision as of 04:55, 3 September 2011

Line 1:

if( wgAction=='edit' || wgAction=='submit' ) $(function(){

function urlDecoderButton(){

addFuncButton2('urlDecoder', urlDecoderRun, 'http://upload.wikimedia.org/wikipedia/commons/9/91/Link_go.png', 'Decode URL before cursor or all URLs in selected text', window.urlDecoderKey)

}

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

if( ! $.fn.textSelection ) mw.loader.load([ 'jquery.textSelection' ])

addFuncButton22(

'urlDecoder',

urlDecoderRun,

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

'Decode URL before cursor or all URLs in selected text',

window.urlDecoderKey)

})

/**/

function addFuncButton2(id, func, img, title, akey){

if (window.wgWikiEditorEnabledModules && wgWikiEditorEnabledModules.toolbar){ //new

var msg = {}; msg[id] = title; mw.usability.addMessages(msg)

function addFuncButton22(bId, bFunc, bIcon, bTitle, bKey){

// $j('#wpTextbox1').wikiEditor('addToToolbar', { section:'main', groups: {'ruwp':{}}})

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

// [[mw:Toolbar_customization]] is not helpful, plus a little delay onload is probably a good thing

section:'main', //groups: {'mytools':{}},

//setTimeout( $.fn.wikiEditor ? newToolbarBtn : oldToolbarBtn, 500 )

group:'insert', tools: {id:{

type:'button',

$.fn.wikiEditor ? newToolbarBtn() : oldToolbarBtn()

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

labelMsg:id,

function newToolbarBtn(){

icon:img

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

}}})

}else{ //old

}

var tlb_ = document.getElementById('toolbar')

if (!tlb_) return

var i=document.createElement('img')

function oldToolbarBtn(){

i.className='mw-toolbar-custombutton'; i.id=id

i.onclick=func; i.src=img; i.title=title; i.alt=title.substr(0,3)

var appendCSSbtn = $('<img. class=mw-toolbar-custombutton {height:20px;id="' background-color:#bce;+ border:1pxbId outset+ #bce; margin:0 1px; cursor:pointer}'">')

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

tlb_.appendChild(i)

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

if (akey){ i.accessKey = akey; i.title += ' ['+akey+']'; updateTooltipAccessKeys([i]) }

.click(bFunc)

}

.appendTo('#toolbar')

}

if( bKey ){

updateTooltipAccessKeys( btn[0] )

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

}

}

}

/**/

function urlDecoderRun(){ //main function

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

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

var localPrefix = WMPrefixes(unSecure(wgServer+wgScript))

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', - 1500)

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-1500>0?endPos-1500:0), 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( unSecure(wgServer+wgScript) )

var oldText, newText, isBeforeCursor, colonNS

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

oldText = tbox.textSelection( 'getSelection' )

if( oldText ){ //there was selection

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

newText = oldText.replace(rx, simplifyMatched)

if( window.urlDecoderIntLinks ){

var ut = '(' + 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){//finds http:.* in string, returns its length and also newText var

isBeforeCursor = true

var pos = str.lastIndexOf('http://')

//move back enough characters

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

var caretPos = tbox.textSelection('getCaretPosition')

if (pos == -1) return 0

var beginPos = caretPos - 2000

if (pos >= 2) str = str.substring(pos-2) //move left to include leading [s

if( beginPos < 0 ) beginPos = 0

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

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

if (!ma) return 0

oldText = tbox.textSelection( 'getSelection' )

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

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

//try to find http in oldText

var 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]?']':'')

return ma[0].length

}

if( oldText == newText ) return

function processSelText(txt){

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

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

simplifyMatched)

if (window.urlDecoderIntLinks){

var ut = 'user_talk' //skip user_talk, usually in sig

if (window.wgFormattedNamespaces) ut = wgFormattedNamespaces[3].replace(/ /g,'_')

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

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

return ut.test(lnk) ? lnk : decodeAnchor(lnk)

})

}

return txt

}

//replace text

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

//end of main code

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

return

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

var trail = url.match(RegExp('[,;\\\\\.:!\\?' //trailing punctuation, per Parser.php

+ (!/\(/.test(url) ? '\\)' : '') + ']+$' )) //trailing no-matching )

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

//---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

}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

}else{

return str //probably broken wikicode in selected text

}

}

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

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

url = unSecure(url)

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

//percent-decoding

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

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

url = url.replace(/[ <>"\[\]]/g, encodeURIComponent) //" disallowed chars

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 + ':')

var link

//try converting to internal link

if (!/(\}\}|\|)$/.test(url)) link = toWikilink(url) //trailing | or }} can be a part of template, skip to be safe

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

if (!link && window.urlDecoderCustom

var && (link = urlDecoderCustomtoWikilink(url)) && /^(https?:\/\/|\{\{)/.test(link))

{url = link; link = null} //still external

//user-defined function

if (link){

if( window.urlDecoderCustom ){

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

url = urlDecoderCustom(url)

if ((wgNamespaceNumber==0 || wgNamespaceNumber==14)

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

&& isBeforeCursor)

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

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

}

if (isBeforeCursor || typeof name == 'string')

//return internal link

url = url.replace(/''/g,'%27%27')//techically '' means the end of URL, but more likely it's part of it

if( link ){

if (typeof name == 'string') return '[' + url + (name?' '+name:'') + ']' //empty name

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

else return url

if( (wgNamespaceNumber==0 || 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){//url -> wikilink, otherwise null

//try bugzilla and user-defined prefixes

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

if (!window.urlDecoderPrefixes) urlDecoderPrefixes = {}

urlDecoderPrefixes['https://bugzilla.wikimedia.org/show_bug.cgi?id=']='mediazilla'

//add bugzilla to user-defined prefixes

for (var key in urlDecoderPrefixes)

urlDecoderPrefixes = $.extend( window.urlDecoderPrefixes,

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

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

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

//try WM prefixes

//apply user-defined prefixes

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

for( var key in urlDecoderPrefixes )

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

var linkPrefix = WMPrefixesif( url.toLowerCase().indexOf(key), prefixes != ''-1 )

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

if (!linkPrefix) return null

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

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

title = decodeAnchor(title)

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

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

if( !ma ) return null

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

var linkPrefix = WMPrefixes( ma[1] )

if (prefixes || isColonNeeded(title)) prefixes += ':' //dividing colon or cat/file leading colon

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

}

Line 173 ⟶ 254:

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 = []

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

Line 186 ⟶ 267:

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

var ch = deChar(hhh)

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

})

//unhide IPs and return

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

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

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

function deChar(ss){

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

return ss

}

}

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

var dd = url.substring(7).split('/')[0].split('.') // -> ['en','wikipedia','org']

if (dd.pop() != 'org') return null

var proj='', lang = '', part = dd.pop()

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

if (proj = {'mediawiki':'mw','wikimediafoundation':'foundation'}[part]);

else if (proj = {'wikipedia':'w','wikibooks':'b','wikinews':'n','wikiquote':'q',

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

'wikisource':'s','wikiversity':'v','wiktionary':'wikt'}[part]){

if( !dd ) return null

lang = dd.pop()

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

if (!lang || lang=='www') lang = ''

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

else if (lang=='test') {lang=''; proj='testwiki'}

}else if (part == 'wikimedia'){

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

if (!part ||subdomain part== 'www' ) projsubdomain = 'foundation'

else if (/^(meta|commons|incubator|species|strategy)$/.test(part)) proj = part

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

else return null

if( !subdomain )

}else return null

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 unSecure(url){

return url.replace(/https:\/\/secure\.wikimedia\.org\/(\w+)\/(\w+)\/([^\]\|\n\r ]+)/,

var mm = /https:\/\/secure\.wikimedia\.org\/(\w+)\/(\w+)\/([^\]\|\n\r ]+)/i.exec( url )

'http://$2.$1.org/$3')

if( !mm) return url

}

var ___domain = mm[1].toLowerCase(), sub = mm[2].toLowerCase()

if( ! wmDomainM[___domain] ) return url //___domain not recognized

function isColonNeeded(pg){

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

if( ___domain == 'wikipedia' ) //handle some special cases

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

switch( sub ){

case 'mediawiki': sub = 'www'; ___domain = 'mediawiki'; break

case 'foundation': sub = ''; ___domain = 'wikimediafoundation'; break

case 'sources': sub = ''; ___domain = 'wikisource'; break

default:

if( wmSubDomains.test(sub) ) ___domain = 'wikimedia' // .../wikipedia/meta -> meta.wikimedia.org

//otherwise: consider it language: .../wikipedia/en

}

return 'http://' + (sub ? sub + '.' : '') + ___domain + '.org/' + mm[3]

}

function getDecoderNSisColonNeeded(pg){

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

if (!window.urlDecoderNS) var urlDecoderNS = {}

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

urlDecoderNS.en = 'image'

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

var ns = urlDecoderNS[wgContentLanguage]

//if( !window.wgNamespaceIds ) return alert('Warning: wgNamespaceIds not defined, old MediaWiki version?');

if (typeof ns == 'string') return ns //user-defined list

if for(!window. var name in wgNamespaceIds) {)

alert if('Warning: (wgNamespaceIds[name]==6 not|| defined,wgNamespaceIds[name]==14) old&& MediaWiki$.inArray(name, version?'list); return== ''-1 )

list.push(name)

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

}

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

ns = ''

for (var name in wgNamespaceIds)

if (wgNamespaceIds[name]==6 || wgNamespaceIds[name]==14)

ns += '|' + name

ns = ns.substring(1)

urlDecoderNS[wgContentLanguage] = ns

return ns

}

}