Global: Do not use reflection for init

[rPraml]

See #1983
use lambda expressions for init

[rPraml]

I'm still struggling with a few CodeQL errors... Does anyone have a tip?

/edit: It happens, because we do not invoke the methods with reflection - and CodeQL "reaches" the methods now. There are some possible dangerous methods like desreialize(obj,'/etc/passwd') (as same as dangerous as new Packages.java.io.File('/etc/passwd').unlink() in liveConnect)

[gbrail]

Huh -- this is the first time CodeQL has "gone off" and it noticed some 20-year-old code ;-)

Seriously, I do think that we should add some path-sanitization logic in there, but in the meantime I'll see if I can figure out how to move on with this change...

[rPraml]

I'm totally puzzled now

• This PR is green now. @gbrail did you retrigger or reconfigure some bulds?
• CodeQl goes completely crazy. In my test branch Just testing CodeQL #1994 I tried to strip down Global.java. Now I have 47 violations scattered across the entire code base: https://github.com/mozilla/rhino/pull/1994/checks?check_run_id=46706261459

this is the first time CodeQL has "gone off"

I think I've discovered a bug in CodeQL. I assume. that some weird code in Global.java will block CodeQL from properly working.

[rPraml]

@gbrail I've converted this to a draft. I would like to tidy up/refactor the Global class first and replace the 20 year old code by some modern versions