Skip to content

Dirty fix for broken Youtube this evening #1

@SpyroTEQ

Description

@SpyroTEQ

Hi,
This evening, the Savetube script (2022.07.11) stopped working on Youtube for Firefox
Digging in the console, we get:

Uncaught (in promise) SyntaxError: unterminated regular expression literal
    ytGetUnscrambleParamFunc moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:799
    SaveTube moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:919
    window["__f__ljqiosax.rql"]/</< moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:1308
    window["__f__ljqiosax.rql"]/< moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:1322
    St https://www.youtube.com/watch?v=ZU74ATbAhns:9
    window["__f__ljqiosax.rql"]/< moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:1
    "__f__ljqiosax.rql" moz-extension://437a936d-5257-43bf-a2f7-372a6e0a7b5b/userscripts/SaveTube.user.js?id=5095000f-66d1-4199-ade0-8aace8f6b18a:1
    St https://www.youtube.com/watch?v=ZU74ATbAhns:9
    s https://www.youtube.com/watch?v=ZU74ATbAhns:72
    <anonymous> https://www.youtube.com/watch?v=ZU74ATbAhns:75
    g https://www.youtube.com/watch?v=ZU74ATbAhns:69

The erroneous line at 799 is

ytMainFuncBody = 'try {' + ytMainFuncBody + '} catch(e) {return null}';

Adding some console logs, it seems the body of the parsed ytMainFuncBody function is:

var b=a.split(""),c=[-265090265,751968325,-1951798130,160288705,function(d,e,f,h,l,m,n,p){return e(f,h,l,m,n,p)},
1941204574,/()\/"()(;)[;,'

which looks indeed truncated
From the original script https://www.youtube.com/s/player/1dda5629/player_ias.vflset/fr_FR/base.js we see a regex /()\/"()(;)[;,'};]();/ which holds a dandling "}", that probably breaks the parser

As a quick and dirty fix, I made the "parseMyContent" function as below:

function parseMyContent(content, pattern) {
	var parse, response;
	content = content.replace(/(\r\n|\n|\r)/gm, '');
    content = content.replace("/()\\/\"()(;)[;,'};]();/","/####/"); // Manual dirty patch
	parse = content.match(pattern);
	if (parse) {
		response = (/g$/.test(pattern)) ? parse : parse[1];
        response = response.replace("/####/", "/()\\/\"()(;)[;,'};]();/"); // Manual dirty patch
	}
	return response;
}

By replacing the regex with dandling "}" before running the parsing Regex of line 797 ( ytMainFuncBody = getMyContent(ytScriptUrl, new RegExp('(?:^|;)' + ytMainFuncName.replace(/\$/, '\\$') + '\\s*=\\s*function\\s*' + '\\s*\\(\\w+\\)\\s*\\{(.*?)\\};')); ) I get a properly parsed youtube script and SaveTube works again

I'm not opening a pull request here, because this is clearly not a nice "fix", but I wanted to share my root cause analysis of this issue so you may find a proper fix quicker and easier :)

(Note: since I edited this parseMyContent, i'm not 100% of my line numbers in this issue)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions