# This constructs a small sample data set for what I am describing.
# This will be an HoHoA with 1 upper Key/value pair, 4 lower
# key/value pairs, and arrays of under 10 elements.  In reality,
# it would be 22,000 upper k/v pairs, 5ish lower k/v pairs, and
# up to hundreds of elements in each array.

# span is the variable used to determine the distance between 
# $upperlimit and $lowerlimit.  $num is the number of different
# patterns searched for.  Both of these are from <STDIN>

$span = 50;
$num = 4;

# These hashes are generated by storing m// matches for regex 
# using a different sub.  keys such as '>scaffold_1' are fasta_ids
# from text sequences in the file.  keys 0,1,2,3 represent the
# patterns searched for (these are the values of the @SortedKeys 
# array).  Values (arrays) are the positions of the matches to the 
# patterns in the file.

%{$matches{>scaffold_1}} = (
     0 => [0,15,100,120,250],
     1 => [25,76,149,211],
     2 => [20,80,115,172,225],
     3 => [40,90,125,180,220,240],
);

# in each case, $lowerlimit would be set to the first element of 
# array {0}, followed by its second element, all the way through, 
# then starting over again at {1}, etc.  $upperlimit is always
# $lowerlimit + $span.


# So, the desired output of the sub would be along these lines:

%{$sets{>scaffold_1}[0]} = (
    0 => (0,15)
    1 => (25)
    2 => (20)
    3 => (40)
    );
    
%{$sets{>scaffold_1}[1]} = (
    0 => (15)
    1 => (25)
    2 => (20)
    3 => (40)
    );
    
%{$sets{>scaffold_1}[2]} = (
    0 => (100,120)
    1 => (149)
    2 => (115)
    3 => (125)
    );
    
%{$sets{>scaffold_1}[3]} = (
    0 => (100,120)
    1 => (76)
    2 => (80,115)
    3 => (90,125)
    );
    
%{$sets{>scaffold_1}[4]} = (
    0 => (250)
    1 => (211)
    2 => (225)
    3 => (220,240)
    );
    
%{$sets{>scaffold_1}[5]} = (
    0 => (120)
    1 => (149)
    2 => (115)
    3 => (125)
    );