To answer the second part of your question: I don't think it is possible to go from a NM_ record (RefSeq mRNA) to chromosomal coordinates. RefSeq mRNAs represent the processed, mature mRNA, so the coordinates are relative to the processed sequence, after splicing.

To take a small example, NM_007102. Download in GenBank format and try the following code:

#!/usr/bin/perl -w
use strict;
use Bio::SeqIO;
use Data::Dumper;

my @exons;
my $file  = "NM_007102.gb";
my $seqio = Bio::SeqIO->new(-file => $file, -format => "genbank");
my $seq   = $seqio->next_seq;

for my $feat($seq->get_SeqFeatures) {
    if($feat->primary_tag eq "exon") {
    push(@exons, $feat->location);
    }
}

print Dumper @exons;

Result:

$VAR1 = bless( {
                 '_strand' => 1,
                 '_seqid' => 'NM_007102',
                 '_location_type' => 'EXACT',
                 '_start' => '1',
                 '_end' => '120',
                 '_root_verbose' => 0
               }, 'Bio::Location::Simple' );
$VAR2 = bless( {
                 '_strand' => 1,
                 '_seqid' => 'NM_007102',
                 '_location_type' => 'EXACT',
                 '_start' => '121',
                 '_end' => '307',
                 '_root_verbose' => 0
               }, 'Bio::Location::Simple' );
$VAR3 = bless( {
                 '_strand' => 1,
                 '_seqid' => 'NM_007102',
                 '_location_type' => 'EXACT',
                 '_start' => '308',
                 '_end' => '597',
                 '_root_verbose' => 0
               }, 'Bio::Location::Simple' );

You'll see that the exons are contiguous. Without information on intron coordinates, this cannot be mapped back to the chromosome.