add --experimental:vmopsDanger; add generic conversion for vmops

[timotheecour]

• this PR adds compiler/vmconv.nim with generic proc toLit*[T](a: T): PNode to convert a concrete type to a PNode (it works recursively). fromLit is also added for reverse direction. I only implemented the branches that were needed to support the vmops introduced, but it's easy to make it more complete in future PRs as needed. This avoids tedious manual conversion.

• it PR also adds --experimental:vmopsDanger to enable certain operations in VM that I and others need.
  Since this is hidden behind a flag, only code that needs this feature will be able to use it.

• can close staticExec removes trailing LF which interferes with binary output that happens to end with 0xA #13786 by providing a more flexible alternative

often requested

This was also requested in several occasions, eg:

• Times in the VM? - Nim forum : Times in the VM? - Nim forum

What I'm asking is, could there ever be a staticTime proc analog to times.getTime

• fix #13222: make relativePath more robust and flexible #13451 (comment)

We talked about adding getCurrentDir for bump's use-case. I would use this in nimph and testutils and golden and openapi, too. I guess I'm just a bad person.

I think you have to concede that if this isn't provided, a work-around will be developed, and that's clearly a worse option. Can you really champion NimScript as the right tool for writing build scripts when one cannot even find the current working directory in NimScript?

• nimterop also needs a lot of functions at CT (by design) and currently to resort to using staticExec which, I've argued in several places (eg https://forum.nim-lang.org/t/6058#37491) is often not the right tool (error prone / os specific shell commands, requires manual serialization/deserialization, inefficient, and many other drawbacks you encounter once you start using it more)

• staticExec is too limited (eg, no flags, and other issues), this PR provides a workaround that should close staticExec removes trailing LF which interferes with binary output that happens to end with 0xA #13786; IMO it's better to reuse stdlib's osproc API's than re-invent another proc with different semantics that keep tripping users; caching (as done in staticExec) can always be done via a separate / orthogonal API eg cache(execCmdEx(...), ... ) usable at both RT and CT

example after PR:

nim c -r --experimental:vmopsDanger t10460.nim works

import osproc, os, times
proc main()=
  {.push experimental: "vmopsDanger".} # not yet meaningful in localized context
  echo getCurrentDir()
  let t = getTime()
  echo t.toUnixFloat
  echo cpuTime()
  let (outp, errC) = execCmdEx("echo ok1 && echo ok2 && nonexistant", {poEchoCmd,poStdErrToStdOut})
  # let (outp, errC) = gorgeEx("echo ok1 && echo ok2")
    # removes LF and has other "surprising" behavior that differs from
    # `execCmdEx`; `execCmdEx` gives you more control when needed.
  echo (outp, errC)
  {.pop.}

static: main()
main()

links

• VM: 'nim e' etc, default to the VM's sandboxing behaviour #16943 (comment) --experimental:vmSandbox

[timotheecour]

ping @Araq