fasttime / jscrewit Goto Github PK

Why are ['NODECONSTRUCTOR','IE_SRC'] not compatible?

Sorry for suggesting IE-exclusive features now.
AFAIK, IE11 on Win7 have some exclusive string related to Localised numbers. Such strings which would be useful for getting chars are listed below:

• Arabic Infinity ("+لا نهاية");
• Czech NaN ("Není číslo");
• Modern Greek Infinity ("Άπειρο");
• Modern Greek NaN ("μη αριθμός");
• Japanese Infinity ("+∞");
• Japanese NaN ("NaN (非数値)");
• Russian Infinity ("бесконечность");
• Simplified Chinese Infinity ("正无穷大");
• Simplified Chinese NaN ("非数字").
  But, IDK will language of IE affect how "zh" is rendered in Number.prototype.toLocaleString(). It's hard to test since we might need help from someone in Hong Kong/Macau/Taiwan.

Currently, there is a definition of "V" relies on document.createElement() would be active with ANY_DOCUMENT and supposed to work in Android Webview 4.0-4.3, but MDN says document.createElement() was added in Android Webview 4.4. should we make an independent feature for document.createElement()?
Additionally, []+document.createElement(false) will return "[object HTMLUnknownElement]" in most browsers, which would be useful for leters "M" and "L". but IDK how document.createElement(false) work in Android Webview 4.0-4.3 and would that be useful for letter "U".

I tried to get the encoding of ", but all failed

I pasted the output into the console and got the following error

Currently, (RP_3_WA + self)[11] only work as "W" when feature WINDOW available, i.e. outside web worker. if string representation of self in web worker is "[object WorkerGlobalScope]", you can add a new feature to make (RP_3_WA + self)[11] available as "W" in web worker. but, if string representation of self in web worker is "[object DedicatedWorkerGlobalScope]" etc., self may not be so useful for "W" in web worker.

if you use most valid Arabic variant as locale Parameters of Number.prototype.toLocaleString(), the output will be Arabic-Indic Numerals. usinglocale Parameters string for Arabic which have shorter code(the one with shortest generated code is all lower case version of Arabic(Saudi Arabia), ie. "ar-sa")can shorten generated code for single Arabic-Indic Numerals character. it works in new browsers and Node.JS 12+.
for example:
in all new browsers:
"١":2867 (current method)
(1).toLocaleString("ar-sa"):2764 (since converted via web, whether "String" optimized is unknown)
but converted length of code using Number.prototype.toLocaleString() is still longer than escape sequence it self(for"١", escaping sequence is "\u0661", which converted length is 1256 chars), so that method may not fit for strings escaped altogether.

There was time when [].filter is not defined, and jsfuck allowed 8 char types ! + ( ) [ ] { }.

1. Can I slightly modify JScrewIt to make it use 8ct and allow old environments like IE7?
2. Can I modify it so it allows 6ct as well as 8ct?

8ct Function == [][([]+![])[!![]+!![]+!![]]+({}+[])[+!![]]+([]+!![])[+!![]]+([]+!![])[+[]]][({}+{})[+!![]+!![]+[+[]]]+({}+{})[+!![]]+([][[]]+[])[+!![]]+([]+![])[!![]+!![]+!![]]+([]+!![])[+[]]+([]+!![])[+!![]]+([][[]]+[])[+[]]+({}+{})[+!![]+!![]+[+[]]]+([]+!![])[+[]]+({}+{})[+!![]]+([]+!![])[+!![]]]

I am a bit new to this , so forgive me if I sound like I dont understand this, I think the encoding is non-deterministic, for instance, there are multiple encodings that can be produced from a source String, does that make decoding impossible ?
Is there a decoder available somewhere ?

Currently, only comments after short code with only one simple object (such as Boolean, number and string literal, one-element array or one-parameter functions) would be deleted, but otherwise code will be transformed with comment. I think the best way handling comments is unconditional deletion, but you can also create an option for toggling comments deletion just in case.

currently, supplementary plane characters like emojis are escaped into 2 surrogate characters. But current JS have alternate methods to treat supplementary plane characters as single character, i.e. String.fromCodePoint(), decodeURI() (like unescape() but treat standard UTF8 better) and \u{XXXXXX}(alternate Escape sequence). when will those be supported?

afaik, call Object.prototype.toString on object Location (which equals to window.location) in Edge will return "[object Location]", and Function("return ({}).toString.call(location)") only cost 5066 char in non-IE browser, and currently 'L' cost 6181 char in non-IE browser when web worker supported. since Location is web-worker-safe, definition of 'L' can be optimized with Location. But, Firefox and Safari not tested, so idk how practical it is.

Let's say we encoding 'hhhhh'. I directly type 'hhhhh' and get 16679 chars. Even if I select 'All new browsers', 6014 chars are generated. However 1889567..toString(18) also be 'hhhhh' and it's length is 3532 chars(All env.) and 1611 chars(new browser),

Change the def of 'D' to escape and see its being directly connect the six letters while d-input generates 15610201..toString(32)+'e'

for IE9
'\' 5120
([].filter.constructor('return/false/')().constructor('\n')+[])[1] 2438
([].filter.constructor('return/false/')().constructor([].filter)+[])[1] 2428
For some environment(e.g.Chrome*) it returns '\n', but for IE9 it's '\'

' ' in All env. turns into (+(+!![]+[+([][(![]+[])[+[]]+([![]]+[][[]])[+!![]+[+[]]]+(![]+[])[!![]+!![]]+(!![]+[])[+[]]+(!![]+[])[!![]+!![]+!![]]+(!![]+[])[+!![]]]+[])[+[]]])+[][(![]+[])[+[]]+([![]]+[][[]])[+!![]+[+[]]]+(![]+[])[!![]+!![]]+(!![]+[])[+[]]+(!![]+[])[!![]+!![]+!![]]+(!![]+[])[+!![]]])[+!![]+[+!![]]] where a simple NaN is encoded as (+(+!![]+[+([][(![]+[])[+[]]+([![]]+[][[]])[+!![]+[+[]]]+(![]+[])[!![]+!![]]+(!![]+[])[+[]]+(!![]+[])[!![]+!![]+!![]]+(!![]+[])[+!![]]]+[])[+[]]])) (+'1NaN').
I temporary locally fixed it by changing defineFHCharAt('ANY_FUNCTION', 8), to define('(NaN + FILTER)[11]'), and get space to (+[![]]+[][(![]+[])[+[]]+([![]]+[][[]])[+!![]+[+[]]]+(![]+[])[!![]+!![]]+(!![]+[])[+[]]+(!![]+[])[!![]+!![]+!![]]+(!![]+[])[+!![]]])[+!![]+[+!![]]], but that just skip the problem

Currently, there is only 1 triplet has batch atob() definition: "mCh" == atob("bUNo"). But there are more triplets which batch atob() definition are practical:
"#!/" (common in URL hashbang grammar) currently costs 1989 chars, while atob("IyEv") only costs 1942 chars;
"{b:" (sometimes can be seen in object definitions) currently costs 1121 chars, while atob("e2I6") only costs 961 chars;
"www" (well-known part of domain name) currently costs 990 chars, while atob("d3d3") only costs 900 chars.
But batch atob() can go further since string segment made up by triplets mentioned before can be created by only one atob(), which is shorter by three-by-three conversion. additionally, some triplets are impractical with atob() as triplet, but their corresponding base64 string have code shorter than original triplet, hence can be combined into single atob() with suitable context. For example, "://" (common in URL header) currently costs 1311 chars, which base64 string "Oi8v" only costs 723 chars, but atob("Oi8v") costs 1544 chars.
Such usage of batch atob() conversion would be challenging since you need an algorithm to clip out most suitable segment form original string for batch atob() conversion.

1. [Said]Array slice on the two(or more if the middle part has exactly 1 char) adjacent chars
2. JScrewIt.encode('"0,00,0"',{features:['FILL','NO_IE_SRC']}).length = 766
JScrewIt.encode('[0].concat(0)+[0].concat(0)',{features:['FILL','NO_IE_SRC']}).length = 763
3. In output.txt 391 | 1 | FILL, IE_SRC but when I run in it's 394. [Update]Ok I know that's because WITHOUT +[] ITS 391

JScrewIt turns "undefined" into "[][[]]" but undefined may be 3 while [][[]] may be 4

Microsoft officially stopped IE support in Feb 14 and trying to remove IE from Windows 11. When will you drop IE support?

On server side, there is a new JS runtime: Deno. When will you have support of Deno?

Currently, only Number.prototype.toString() is used in the optimizer, but it has a disadvantage: it would be limited by MAX_SAFE_INTEGER. BigInt.prototype.toString() is not affected by that. To get number in BigInt type, you can just call "return "+number written in string+"n" in Function().

Currently, when fromCharCode or fromCodePoint is used, the whole string is converted into decimal codepoint even if the string begin or ended with simple characters (numbers, 'N','a','d','e','f','i','l','n','r','s','t','u'). Using "+" to attach those simple characters on head or end of string genetated by fromCharCode can shorten generated code.

alert(1) is [].filter.constructor('return alert')()(1)
/./ is [].filter.constructor('return eval')()('/./')
Why not [].filter.constructor('return/./')()?
({}) is [].filter.constructor('return eval')()('({})'),
not [].filter.constructor('return{}')()

There should only be XXXeYYY and (AAA)(.BBB) -- since point costs 72 chars far larger than '9' 47, it's never cheap to use both.
With x being a positive finite number, turning x to XXXeYYY style uses x.toExponential().replace(/^(\d)\.?(\d*)e(.*)$/,function(x,a,b,c){return a+b+'e'+(c-b.length)}) and to (AAA)(.BBB) use Number.prototype.forceFixed=function(){var t=this.toString().split('e');if(t.length==1)return t.replace(/^0/,'');t[0]=t[0].replace(/\./,'');return t[1]>0?t[0]+Array(2- -t[1]-t[0].length).join(0):'.'+Array(-t[1]).join('0')+t[0]}? (code not tested fully)

Currently, the tool only have major browser (i.e. IE, Firefox, Chromium-based browsers, Safari and Android browser) and Node.JS support. Should we add support for minor and local browser (e.g. Konqueror, UC, Yandex) and non-browser JS engine (e.g. Nashorn, lv, IronJS) now?

'\040' 28903 bytes as Function("return '\040'")() instead of directly describe char32(space)?

when I tried to convert "Infinitywhitenight" in Edge, a 1201-chars-code is returned and evals into "infinit". why would this happen? will this be fixed?

When I try to compile const test = require("test") and then I run it with NodeJS v8.1.2 (which uses V8 v5.8.283.41) I get

ReferenceError: require is not defined
    at eval (eval at <anonymous> (C:\Users\H\Desktop\game\dist\server\index.js:1
:2288), <anonymous>:2:14)
    at eval (<anonymous>)
    at Object.<anonymous> (C:\Users\H\Desktop\game\dist\server\index.js:1:2288)
    at Module._compile (module.js:569:30)
    at Object.Module._extensions..js (module.js:580:10)
    at Module.load (module.js:503:32)
    at tryModuleLoad (module.js:466:12)
    at Function.Module._load (module.js:458:3)
    at Function.Module.runMain (module.js:605:10)
    at startup (bootstrap_node.js:158:16)

When running it directly (without compilation) I don't get any errors.
I'm building with node and JScrewIt.Feature.AUTO

Have I done something wrong or is this actually a bug?

Currently, "fromCharCo" is defined as "from3har3o".split(3).join("C"). but I tried "from0har0o".split(0).join("C"), i's shorter and works will in Chromium. What did you think at that time?

[All Supported Env.]
'!!!!!' 26486
unescape(unescape("false2521false2521false2521false2521false2521".split(false).join("%"))) 22547
unescape("false21false21false21false21false21".split(false).join("%")) 17058, but with 'b' or 'c' escaped that'd be too long

(though I don't know what code generated by the present way

currently, "browser usage" is treated as synonym of "HTML related usage". But, XML and SVG also support JS and can be opened by browsers too. Should we add support for XML and SVG?