[Pdx-pm] array question

[Tkil]

>>>>> "Annett" == aspen  <108" <aspen_108 at hotmail.com>> writes:

Annett> Here're some parts of the code where I'm running into the
Annett> problem.  Be forewarned, I'm kind of a perl newbie...

That's ok, we're here to educate.  Occasionally with a stick with
spiky bits on it, but really, we're just using tough love...

Annett> Here's the part where I'm reading the data into these arrays:

Annett>     if ($get_temp == 1 && $inp_line !~ /Temperature/) {
Annett>         @temp_array = split ' ', $inp_line;
Annett>         for $temp_data (@temp_array) {
Annett>             $temp[$seg][$lay] = $temp_data;
Annett>             printf "%s%10.1f%10.1f%8.2f%8.2f%8.1f\n", $reg_date, $seg,
Annett>               $lay, $temp[$seg][$lay], $tot_seg, $#{$temp[$seg]};
Annett>             $lay++;
Annett>         }
Annett>     }

Without knowing where $seg is getting set (and is $lay getting reset
somewhere?), it's hard to figure out which values are being duplicated
or not.

Consider:

| #!/usr/bin/perl -w
| 
| use strict;
| 
| use Data::Dumper qw( Dumper );
| $Data::Dumper::Indent = 1;
| 
| my @temp;
| 
| while ( my $line = <DATA> )
| {
|     next unless $line =~ /^Temperature/;
| 
|     my @vals = split ' ', $line;
|     shift @vals; # get rid of label
| 
|     push @temp, \@vals;
| }
| 
| print Dumper \@temp;
| 
| exit 0;
| 
| __DATA__
| Temperature: a b c
| Pressure:    1 2 3
| Temperature: alpha beta gamma

Which prints:

| $ ./annett2.plx
| $VAR1 = [
|   [
|     'a',
|     'b',
|     'c'
|   ],
|   [
|     'alpha',
|     'beta',
|     'gamma'
|   ]
| ];

Annett> Here's one example of how I'm trying to clear the array:

Annett>         for ($seg = $#temp; $seg > 0; $seg--) {
Annett>             for ($lay = $#{$temp[$seg]}; $lay > 0; $lay--) {
Annett>                 undef $temp[$seg][$lay];
Annett>             }
Annett>         }

You really should be able to just do:

   undef @temp;

And this will blow away everything.

Remember, Perl is reference-counting and garbage-collecting; you
don't have to "undef" anything when you're done using it.  (Exception:
circular structures.)

Annett> That print statement with the $#{$temp[$seg]} lets me look at
Annett> the index as it runs, and the array indices are not being
Annett> cleared with the undef.  I gave your one suggestion
Annett> "@some_array = ();" a quick try, but that didn't clear the
Annett> indices either... not sure if there's a different syntax for a
Annett> 2-d array though.

I'm not quite clear what you mean by "clear the indices".  If you
undef an array, or assign the empty list into it, then $#array should
be -1, and "scalar @array" should be 0.  Consider:

| #!/usr/bin/perl -w
| 
| use strict;
| 
| my @a;
| 
| @a = qw( a b c );
| print "orig:         \@a = [ @a ], \$#a = $#a\n";
| 
| undef @a;
| print "after undef:  \@a = [ @a ], \$#a = $#a\n";
| 
| @a = qw( a b c );
| print "refilled:     \@a = [ @a ], \$#a = $#a\n";
| 
| @a = ();
| print "after assign: \@a = [ @a ], \$#a = $#a\n";
| 
| 0;

Which prints:

| $ ./annett1.plx
| orig:         @a = [ a b c ], $#a = 2
| after undef:  @a = [  ], $#a = -1
| refilled:     @a = [ a b c ], $#a = 2
| after assign: @a = [  ], $#a = -1

So either "undef @array" or "@array = ()" will get rid of the indices
as advertised.  Which is why I think there's a conversational gap
here.

Offhand, I'd guess that $seg and $lay aren't getting (re)set quite
right.  Also, remember that Perl arrays are (almost) always counted
from 0, so if you set $array[100], then all the indexes from 0 to 99
inclusive are created and set to undef.  If you want sparse arrays,
use hashes.  :)

t.