Skip to content

parent.frame(n) produces different result from sys.frame(sys.parent(n))

3 messages · brodie gaslam, Taras Zakharko

Original
Taras Zakharko
# 
Dear all, 

The documentation states that parent.frame(n) is equivalent to sys.frame(sys.parent(n)) but I have discovered a case where they produce different results. Before I submit a bug report I thought it would be good to run it by the R community in case it?s (somehow?) expected behaviour. Consider the following MRE (this is R 4.2.1 running on Apple M1):

  f1 <- function() {
    f2()
  }
  
  f2 <- function() {
    f3()
  }
  
  f3 <- function() {
    evalq(check_parents(), parent.frame())
  }
  
  check_parents <- function() {
    print(vctrs::data_frame(
      call = as.list(sys.calls()),
      frame = as.list(sys.frames()),
      parent = as.list(sys.parents())
    ))
  
    print(parent.frame(2L))
    print(sys.frame(sys.parent(2L)))
  }
  
  f1()

This produces

                                    call                          frame                                           parent
1                                   f1()                        <environment: 0x10785d408>      0
2                                   f2()                        <environment: 0x107898830>      1
3                                   f3()                        <environment: 0x107898788>      2
4 evalq(check_parents(), parent.frame())  <environment: 0x1078a1f30>       3
5 evalq(check_parents(), parent.frame())  <environment: 0x107898830>      4
6                        check_parents()               <environment: 0x1078a1b08>      2
<environment: 0x1078a1f30>  # parent.frame(2L)
<environment: 0x10785d408> # sys.frame(sys.parent(2L))

It seems like parent.frame(2L) resolves to frame 4 which is not part of the call stack of frame 6 at all. I haven?t yet looked at the C code. 

Best, 

Taras
brodie gaslam
# 
Hi Taras,

I have not looked in detail at your examples here, but the
use of evalq and sys.parent makes me think these issues:

https://bugs.r-project.org/show_bug.cgi?id=17849
https://bugs.r-project.org/show_bug.cgi?id=15531

are possibly related.

Best,

B.
On Wednesday, September 28, 2022 at 05:29:08 AM EDT, Taras Zakharko <taras.zakharko at uzh.ch> wrote:
Dear all, 

The documentation states that parent.frame(n) is equivalent to sys.frame(sys.parent(n)) but I have discovered a case where they produce different results. Before I submit a bug report I thought it would be good to run it by the R community in case it?s (somehow?) expected behaviour. Consider the following MRE (this is R 4.2.1 running on Apple M1):

? f1 <- function() {
? ? f2()
? }
? 
? f2 <- function() {
? ? f3()
? }
? 
? f3 <- function() {
? ? evalq(check_parents(), parent.frame())
? }
? 
? check_parents <- function() {
? ? print(vctrs::data_frame(
? ? ? call = as.list(sys.calls()),
? ? ? frame = as.list(sys.frames()),
? ? ? parent = as.list(sys.parents())
? ? ))
? 
? ? print(parent.frame(2L))
? ? print(sys.frame(sys.parent(2L)))
? }
? 
? f1()

This produces

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? call? ? ? ? ? ? ? ? ? ? ? ? ? frame? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? parent
1? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? f1()? ? ? ? ? ? ? ? ? ? ? ? <environment: 0x10785d408>? ? ? 0
2? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? f2()? ? ? ? ? ? ? ? ? ? ? ? <environment: 0x107898830>? ? ? 1
3? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? f3()? ? ? ? ? ? ? ? ? ? ? ? <environment: 0x107898788>? ? ? 2
4 evalq(check_parents(), parent.frame())? <environment: 0x1078a1f30>? ? ? 3
5 evalq(check_parents(), parent.frame())? <environment: 0x107898830>? ? ? 4
6? ? ? ? ? ? ? ? ? ? ? ? check_parents()? ? ? ? ? ? ? <environment: 0x1078a1b08>? ? ? 2
<environment: 0x1078a1f30>? # parent.frame(2L)
<environment: 0x10785d408> # sys.frame(sys.parent(2L))

It seems like parent.frame(2L) resolves to frame 4 which is not part of the call stack of frame 6 at all. I haven?t yet looked at the C code. 

Best, 

Taras 

______________________________________________
R-devel at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
Taras Zakharko
# 
Thanks Brodie, this certainly seems like the same issue! I?ll add some comments to the issue tracker and hope that this can finally be fixed. 

Best, 

Taras