Splitting string into variables

[Alan Bak]

I have a text file that I am reading in with perl line by line and breaking into variables for processing. The lines are being broken on white spaces using split.

Here are a couple of lines from the text file:

bl-render -render /renders/HOE/HD_RAW/REEL_5/reel_5.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-30408 max:hoe:reel_5
bl-render -render /renders/HOE/HD_RAW/REEL_6/reel_6.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-13472 max:hoe:reel_6

Question: is there a way to break each line into variables and keep the " " strings as 1 variable? Simply breaking on white spaces results in "Super 16mm" being broken into 2 variables "Super and 16mm" while I would like it to be one variable "Super 16mm".

Any ideas?


Thanks

[KevinADC]

What you have is the same as a comma delimited file but with spaces instead of commas. So we can use the code from the Perl Cookbook, parsing comma-seperated data, and substitute a space where the commas would be. (Not well tested):

my @data;
while (<DATA>) {
   push @data,parse_data($_);
}
print "$_\n" for @data;

sub parse_data {
    my $text = shift;
    my @new  = ();
    push(@new, $+) while $text =~ m{
        "([^\"\\]*(?:\\.[^\"\\]*)*)"\s?
           |  ([^ ]+)\s?
           | \s
       }gx;
       push(@new, undef) if substr($text, -1,1) eq ' ';
       return @new;
}  
__DATA__
bl-render -render /renders/HOE/HD_RAW/REEL_5/reel_5.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-30408 max:hoe:reel_5
bl-render -render /renders/HOE/HD_RAW/REEL_6/reel_6.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-13472 max:hoe:reel_6

It does assume there is a single space delimiting the data fields, if there can be more than one space the code would need to be changed a little bit. I have no idea how it will work if a field is blank.

Apply the parse_data() subroutine however is appropriate for your purposes.

[Alan Bak]

Thanks Kevin!

I will give this a try. I have access to the Perl Cookbook so I will look it up as well.

Cheers

Alan

What you have is the same as a comma delimited file but with spaces instead of commas. So we can use the code from the Perl Cookbook, parsing comma-seperated data, and substitute a space where the commas would be. (Not well tested):

my @data;
while (<DATA>) {
   push @data,parse_data($_);
}
print "$_\n" for @data;

sub parse_data {
    my $text = shift;
    my @new  = ();
    push(@new, $+) while $text =~ m{
        "([^\"\\]*(?:\\.[^\"\\]*)*)"\s?
           |  ([^ ]+)\s?
           | \s
       }gx;
       push(@new, undef) if substr($text, -1,1) eq ' ';
       return @new;
}  
__DATA__
bl-render -render /renders/HOE/HD_RAW/REEL_5/reel_5.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-30408 max:hoe:reel_5
bl-render -render /renders/HOE/HD_RAW/REEL_6/reel_6.%.6F.dpx -profile HD_bake -format "Super 16mm" -mask "Full" -range 0-13472 max:hoe:reel_6

It does assume there is a single space delimiting the data fields, if there can be more than one space the code would need to be changed a little bit. I have no idea how it will work if a field is blank.

Apply the parse_data() subroutine however is appropriate for your purposes.