Question

I have a file that contains parameters using this syntax

RANGE {<value> | <value>-<value>} [ , ...]

where value s are numbers.

for example, all these are valid syntax

RANGE 34
RANGE 45, 234
RANGE 2-99
RANGE 3-7, 15, 16, 2, 54

How can I parse the values to an array in Perl?

For example for the last example, I want my array to have 3, 4, 5, 6, 7, 15, 16, 2, 54. The ordering of elements does not matter.

---

The most basic way is to check for a - symbol to determine whether there is a range or not, parse the range using a loop and then parse the rest of the elements

my @arr;
my $fh, "<", "file.txt" or die (...);
while (<$fh>) {
    if ($_ =~ /RANGE/) {
        if ($_ =~ /-/) { # parse the range
            < how do I parse the lower and upper limits? >
            for($lower..$upper) {
                $arr[++$#arr] = $_;
            }
        } else { # parse the first value
            < how do I parse the first value? >
        }

        # parse the rest of the values after the comma
        < how do I parse the values after the comma? >
    }
}

• I need help parsing the numbers. For parsing, one way I can think of is to use successive splits (on -, , and ). Is there some better (clean and elegant, using regex maybe?) way?

• Also, comments/suggestions on the overall program structure are welcome.

Answer 1

I would suggest parsing the line into a separate variable, as $_ tends to get clobbered by other function calls. You can remove the trailing newline at the same time, with chomp.

while (<$fh)>
{
    chomp (my $line = $_);
    # ...
}

Next, you need to detect the ‘RANGE’ indicator, and extract the numbers that follow. If there is no such indicator, you can just skip to the next line:

next if $line !~ /^RANGE (.*)$/;

Now, you can start extracting the numbers, splitting on the comma delimiter:

my @ranges = split /, /, $1;

Now you can extract the dashes and translate those into ranges. This is the tricky part — if the value has a dash in it, get the first and second numbers, and turn them into a range with the .. operator; otherwise, leave the number alone:

@ranges = map { /(\d+)-(\d+)/ ? ($1 .. $2) : $_ } @ranges;

Putting all that together, and combining expressions, gives us:

my @numbers;
while (<$fh)>
{
    chomp (my $line = $_);
    next if $line !~ /^RANGE (.*)$/;

    push @numbers, map { /(\d+)-(\d+)/ ? ($1 .. $2) : $_ } (split /, /, $1);
}