PDL fails on empty PDL case for version PDL-2.099, possible error in Slices.pm

[yelnats]

For PDL version 2.099 there seems to be a bug in PDL::Slices

for example:

use PDL;
my $pdl = sequence(4,3);
print "$pdl";

my  $idx        = whichND($pdl > 5);        # No problem as $idx is not empty.

my  $idx2       = whichND($pdl > 500);  # Will trigger PROBLEM as $idx2 is empty

my  $cpy = $pdl->copy;

# works for idx because it is not empty.
$cpy->indexND($idx) .= ($pdl + 5)->indexND($idx);

#fails for $idx2 because it is empty this time
$cpy->indexND($idx2) .= ($pdl + 5)->indexND($idx2);

#work around 1, evaluate and store the left side in the default variable and return that
($_=$cpy->indexND($idx2)) .= ($pdl + 5)->indexND($idx2);

#work around 2, if object is empty do not execute the code with the bug
$cpy->indexND($idx2)  .= ($pdl + 5)->indexND($idx2) unless $idx2->isempty;

Looking at Slices.pm, it seems one fix would be the following change to one line (455):

sub PDL::range :lvalue {
  my($source,$ind,$sz,$bound) = @_;
  # Convert to indx type up front (also handled in rangeb if necessary)
  my $index = (ref $ind && UNIVERSAL::isa($ind,'PDL') && $ind->type eq 'indx') ? $ind : indx($ind);
  my $size = defined($sz) ? PDL->pdl($sz) : undef;
  # Handle empty PDL case: return a properly constructed Empty.
  if($index->isempty) {
      my @sdims= $source->dims;
      splice(@sdims, 0, $index->dim(0) + ($index->dim(0)==0)); # added term is to treat Empty[0] like a single empty coordinate
      unshift(@sdims, $size->list) if(defined($size));
#     return PDL->new_from_specification(0 x ($index->ndims-1), @sdims);        # The original line that fails for empty PDL case
      return ($_=PDL->new_from_specification(0 x ($index->ndims-1), @sdims));   # Fix, Evaluate/Store the return value in $_ 
  }

Cheers

[mohawk2]

Thanks for the report! Please could you make a PR, especially with a test-case? The above construct of assigning to $_ isn't right, if imposing scalar context is what fixes it, then just use scalar.

[mohawk2]

Looking again at the code you're pointing at as the problem, I suspect what it wants is changing from 0 x $num to (0) x $num. Tests will light the way!

[mohawk2]

This was due to new_from_specification not itself being marked as lvalue, which is now fixed.