#!/usr/bin/perl
# Convert CSV to TAB delimited
# 2002-07-10 JPV
# $ver = "v2.0"; # 2002-10-16 JPV Had to re-write, as the MRE RegEx was not
# 	correctly parsing escaped quotes ( "" ) within quoted strings.
#	This is about 5 times slower than the RegEx, but it parses right.
#$ver = "v2.1"; # 2002-12-23 JPV BUGFIXS for OUTFILE and separatores.
#$ver = "v3.0"; # 2002-12-30 JPV Changed back to regex (MRE2) from quotewords.
#$ver = "v3.1"; # 2003-01-02 JPV Fixed extra trailing tab on each line
#$ver = "v3.2"; # 2003-03-21 JPV Added -d
$ver = "v3.2a"; # 2003-04-01 JPV Annoying bugfix for -h checking


# Original RegEx from _Mastering_Regular_Expressions_ and
# Eric Soulliard <Eric.SOULLIARD@gemplus.com> 2002-04-01
# Text::ParseWords/quotewords is from _Perl_Cookbook_ 1.15 pg 31
# Other code from csv2html.pl (Eric Soulliard)

# See sub parse_csv_mre2 for details about the regex to parse CSV.


##########################################################################
(($myname = $0) =~ s/^.*(\/|\\)//ig); # remove up to last "\" or "/"
$Greeting =  ("$myname $ver Copyright 2002 JP Vossen (http://www.jpsdomain.org/)\n");
$Greeting .= ("    Licensed under the GNU GENERAL PUBLIC LICENSE:\n");
$Greeting .= ("    See http://www.gnu.org/copyleft/gpl.html for full text and details.\n");

if (("@ARGV" =~ /\?/) || ("@ARGV" =~ / -h/) || "@ARGV" =~ /--help/) {
    print STDERR ("\n$Greeting\n");
    print STDERR <<"EoN";    # Various usage notes
	Usage: $myname (-i {infile}) (-o {outfile}) (-q)

   -i {infile}    = Use infile as the input file, otherwise use STDIN.
   -o {outfile}   = Use outfile as the output file, otherwise use STDOUT.
   -d {delimiter} = Use the specificed delimiter instead of TAB.
   -q             = Be quiet about it.

EoN
    die ("\n");
}

###use Text::ParseWords;           # Use this to parse CSV input
use Getopt::Std;                # Use Perl5 built-in program argument handler
getopts('i:o:d:q');             # Define possible args.

if (! $opt_i) { $opt_i = "-"; } # If no input file specified, use STDIN
if (! $opt_o) { $opt_o = "-"; } # If no output file specified, use STDOUT

open (INFILE,   "$opt_i") or die ("$myname: error opening $opt_i for input: $!\n");
open (OUTFILE, ">$opt_o") or die ("$myname: error opening $opt_o for output: $!\n");

$delimiter=$opt_d||"\t";        # Use the specified delimiter or TAB

if (! $opt_q) { print STDERR ("\n$Greeting\n"); }

while ($aline = <INFILE>) {
    chomp($aline);
    ### @arecord = quotewords(",", $KeepSep, $aline);
    @arecord = &parse_csv_mre2 ($aline);
    $outline = join ($delimiter, @arecord);
    print OUTFILE "$outline\n";
}

if (! $opt_q) { print STDERR ("\n\a$myname finished in ",time()-$^T," seconds.\n"); }


#+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
sub parse_csv_mre2 {

# Called like: @arecord = &parse_csv_mre2 ($aline);

# Regex to parse CSV from _Mastering_Regular_Expressions,_Second_Edition_;
# page 271. See http://regex.info/ esp. http://regex.info/dlisting.cgi?id=1253)

    #if (@_[0] eq undef()) { warn ("$myname: empty variable passed to parse_csv_mre2!\n"); }
    if (@_[0] eq undef()) { return(); }
    my $line = @_[0];
    my @parsedline = ();
    my $field = '';

    # See top for details about the regex
    while ($line =~ m{
              \G(?:^|,)
              (?:
                 # Either a double-quoted field (with "" for each ")...
                 " # field's opening quote
                  ( (?> [^"]* ) (?> "" [^"]* )*  )
                 " # field's closing quote
               # ..or...         
               |
                 # ... some non-quote/non-comma text....
                 ( [^",]* )
              )
          }gx)
    {                         # OK, done with regex, NOW what...
       if (defined $2) {      # Got some non-quote/non-comma text
           $field = $2;
       } else {               # Got escaped quotes and stuff
           $field = $1;
           $field =~ s/""/"/g;
       }
    push (@parsedline, $field);
    } # end of while block

    return (@parsedline);
} # end of sub parse_csv_mre2