I'd say add a header line and use DBD::CSV. Of course you can do it with Text::CSV or Text::CSV_XS.

I'd use DBD::CSV if the dataset isn't too large (the whole set fits easily in memory).

I'd use Text::CSV_XS' csv function with on_in that lets you change fields on the fly.

$ cat test.pl
#!/pro/bin/perl

use 5.20.0;
use warnings;

use Text::CSV_XS "csv";

my %state = qw(
    0H  Ohio
    0R  Oregon

    C   CONDENSED
    F   Final
    );

csv (in => "test.csv", on_in => sub {
    $_[1][0] = $state{$_[1][0]} || $_[1][0];
    $_[1][4] = $state{$_[1][4]} || $_[1][4];
    });
$ cat test.csv
C,003187995,0H047013N,20141212,0H,2050,EX,,
C,013139784,0H0430000,20141110,0H,2014,EX,,
C,153188023,0H0020000,20150105,0H,2015,BA,,
C,173167740,MT007015G,20141202,0R,2015,BA,,
$ perl test.pl
CONDENSED,003187995,0H047013N,20141212,Ohio,2050,EX,,
CONDENSED,013139784,0H0430000,20141110,Ohio,2014,EX,,
CONDENSED,153188023,0H0020000,20150105,Ohio,2015,BA,,
CONDENSED,173167740,MT007015G,20141202,Oregon,2015,BA,,
$
[download]

update: As of version 1.17, you can also use filter (if you prefer that):

csv (in => "test.csv", filter => {
    1 => sub { $_ = $state{$_} || $_ },
    5 => sub { $_ = $state{$_} || $_ },
    });
[download]