Re: Will this work?

#!/usr/bin/perl

use warnings;
use strict;
use diagnostics;

use Convert::Translit;
use Text::CSV::Simple;

use Gedcom;

use Data::Dumper;

/*
	The following is just for getting the CSV:
*/

my $kiprequest = shift;

my $kipdata = "/Volumes/terra/Genealogy/FT + KB/kipdata.txt";
my $kippath = "/Volumes/terra/Genealogy/FT + KB/KIP/";

open (KIPDATA, $kipdata) || die "$!: $kipdata";
my @kipdata;
while (my $a = <KIPDATA>) {
	chomp $a;
	push @kipdata, $a;
}
close(KIPDATA);

my @kipmatches = grep { $_ =~ $kiprequest } @kipdata;

if (@kipmatches != 1) {
	print @kipmatches." matches for '$kiprequest' found:\n";
	foreach (@kipmatches) {
		print "  $_\n";
	}
	print "More accuracy needed.\n\n";
	exit(0);
}

my ($kipamt, $kipherred, $kipsogn, $kipyear, $kipfile);

($kipamt, $kipherred, $kipsogn, $kipyear, $kipfile) = split(";", $kipmatches[0]);
my $file = $kippath.substr($kipfile,0,1)."/".$kipfile.".csv";

my $translator = new Convert::Translit("latin1", "mac");

open(HEADERS, "$file") || die "$!: $file";
my $headers = <HEADERS>;
close(HEADERS);
$headers =~ s/[\r\n]+//ig;
my @fields = split(";",$translator->transliterate($headers));

#print Dumper \@fields;

#$/ = "\r\n";

my $csv = Text::CSV::Simple->new({
                                  sep_char => ";", 
                                  binary => 1,
                                  eol => "\r\n"
                                 });
$csv->field_map(@fields);
my @data = $csv->read_file($file);

#fix character map
for (my $i = 0; $i < @data; $i++) {
	foreach my $key (keys %{$data[$i]}) {
		$data[$i]->{$key} = $translator->transliterate($data[$i]->{$key});
	}
}

shift @data;

#print Dumper \@data;

/*
	Actual parsing starts here:
*/

my $ged = Gedcom->new;

my %fam;
my %houses;

#ADD PERSONS

foreach my $person (@data) {
	my $indi = $ged->add_individual;
	
	print "Adding... ".$person->{"Kildenavn"}."\n";
	
	#name
	my $gedname = $person->{"Kildenavn"};
	$gedname =~ s|(\s)([^\s]+)$|$1/$2/|; #put // around surname as per gedcom standard
	$indi->add("name", $gedname);
	
	#gender (parsing is based on patronyms if no gender is in CSV)
	my $gedgender = "U";
	if ($person->{"Køn"} eq "M"
	 || $person->{"Kildenavn"} =~ /(sen|son|søn)$/i
	 || $person->{"Stilling i husstanden"} =~ /bonde/i) {
		$gedgender = "M";
	} 
	if ($person->{"Køn"} eq "F"
	 || $person->{"Køn"} eq "K"
	 || $person->{"Kildenavn"} =~ /datter$/i
	 || $person->{"Kildenavn"} =~ /enke$/i
	 || $person->{"Stilling i husstanden"} =~ /moder/i
	 || $person->{"Stilling i husstanden"} =~ /pige/i
	 || $person->{"Stilling i husstanden"} =~ /kone/i
	 || $person->{"Stilling i husstanden"} =~ /datter/i
	 || $person->{"Stilling i husstanden"} =~ /faster/i
	 || $person->{"Stilling i husstanden"} =~ /moster/i
	 || $person->{"Stilling i husstanden"} =~ /ske$/i) {
		$gedgender = "F";
	}
	$indi->add("sex", $gedgender);
	
	#birth
	my $gedbirthdate;
	if ($person->{"Født kildedato"}) {
		$gedbirthdate = $person->{"Født kildedato"}; #TODO: convert date
	} elsif ($person->{"Fødeår"}) {
		$gedbirthdate = $person->{"Fødeår"};
	} elsif ($person->{"Alder Kbhv."}) {
		$gedbirthdate = "ABT ".($kipyear - $person->{"Alder Kbhv."}); #guess year from age
	} elsif ($person->{"Alder"}) {
		$gedbirthdate = "ABT ".($kipyear - $person->{"Alder"}); #guess year from age
	} elsif ($person->{"Kildealder"}) {
		$gedbirthdate = "ABT ".($kipyear - $person->{"Kildealder"}); #guess year from age
	} else {
		$gedbirthdate = "ABT $kipyear";
	}
	$indi->add("birth date", $gedbirthdate);
	if ($person->{"Kildefødested"}) {
		$indi->add("birth place", $person->{"Kildefødested"});
	} elsif ($person->{"Fødested kun Frb. og Kbhv."}) {
		$indi->add("birth place", $person->{"Fødested kun Frb. og Kbhv."});
	}
	
	#census record
	$indi->add("census date", $kipyear);
	if ($person->{"Kildestednavn"}) {
		$indi->add("census place", "Husstand ".$person->{'Husstands/familienr.'}.", ".$person->{"Kildestednavn"}.", $kipsogn sogn, $kipherred herred, $kipamt amt, Denmark");
	} else {
		$indi->add("census place", "Husstand ".$person->{'Husstands/familienr.'}.", $kipsogn sogn, $kipherred herred, $kipamt amt, Denmark");
	}
	if ($person->{"Alder"}) {
		$indi->add("census age", $person->{"Alder"});
	} elsif ($person->{"Kildealder"}) {
		$indi->add("census age", $person->{"Kildealder"});
	}
	
	#family key
	my $famnum = $person->{'Kildestednavn'}."-".$person->{'Husstands/familienr.'};
	my %p = (
				'xref' => $indi->xref,
				'marital' => $person->{"Civilstand"},
				'position' => $person->{"Stilling i husstanden"},
				'sex' => $gedgender
			);
	push @{$houses{$famnum}}, \%p;
	
	#occupation
	if ($person->{"Kildeerhverv"}) {
		$indi->add("occupation", $person->{"Kildeerhverv"});
	} elsif ($person->{"Stilling i husstanden"}) {
		$indi->add("occupation", $person->{"Stilling i husstanden"});
	}
}

/*
	In the above, I've pushed the individuals into an array,
	keyed by a "family key" (see line 157).
	
	I run through that and create families after all the individuals
	have been created.
*/

foreach my $h (keys %houses) {
	# Works on the assumption that families are ordered (father first, then mother, then children)
	
	my $fam = $ged->add_family;
	
	#TODO: mandens faster, konens moder, moster faster etc.
	#TODO: foster children
	
	foreach my $i (@{$houses{$h}}) {
		if ($i->{'marital'} =~ /(^Gift|enke)/i) {
			if ($i->{'sex'} eq "M") {
				if ($fam->husband) { $fam = $ged->add_family; }
				$fam->add_husband($ged->resolve_xref($i->{'xref'}));
			} elsif ($i->{'sex'} eq "F") {
				if ($fam->wife) { $fam = $ged->add_family; }
				$fam->add_wife($ged->resolve_xref($i->{'xref'}));
			}
		} elsif ($i->{'position'} =~ /(b[aø]rn|datter|søn)/i) {
			$fam->add_child($ged->resolve_xref($i->{'xref'}));
		}
	}
}

foreach my $family ($ged->families) {
	print "Verifying family: ".$family->xref."\n";
	
	my @husbands = $family->husband;
	my @wives = $family->wife;
	
	if (@husbands > 1) {
		print "  too many husbands (".@husbands.")\n";
		
	}
	if (@wives > 1) {
		print "  too many wives (".@wives.")\n";
	}
}

#print Dumper \%houses;

#WRITE OUTPUT

print "\nWriting output...";

$ged->order($ged->sort_sub);

$ged->write(substr($kipfile,0,5).".ged", 0);

print "Done.\n";

__END__

__DATA__

Some notes:

The colums that were in the original CSV:

	'KIPnr'							# census transciption ID
	'Løbenr i indtastning'			# serial
	'Husstands/familienr.'			# household number
	'Antal familier/hus'			# number of families per household

*	'Kildenavn'						# name of individual

*	'Køn'							# gender

*	'Fødeår'						# birth year
*	'Født kildedato'				# birth date
*	'Alder Kbhv.'					# age (copenhagen census)
*	'Alder'							# age
*	'Kildefødested'					# birth place
*	'Fødested kun Frb. og Kbhv.'	# birth place (copenhagen census)

*	'Kildestednavn'					# place
	'Adresse 1916'
	'Adresse'
	'Flyttet til kommunen'
	'Matr.nr./Adresse'
	'Matrikel 1916'
	'Matrikel'
	'Sidste bopæl/oph.sted'
	'Midlertidig opholdssted'
	'Midlertidigt nærværende'

*	'Civilstand'					# civil position (married, widow, etc)
	'Antal ægteskab'
	'Hvornår æf. død'
	'Hvornår gift'
	'Nr. ægteskab'
	'Stilling i husstanden'
	'Antal levende børn'
	'Antal døde børn'

	'Kilde erhvervskommune'
	'Kilde erhvervssted'
	'Kildeerhverv'
	'Arbejdsplads'
	'Arbejdssted'
	'Indkomst'
	'Formue'
	'Stats skat'
	'Kommune skat'

	'Trossamfund'

	'Kommentar'

	'Handicaps'
	'Kildehenvisning'
	'Transportmiddel'
	'Boligtælling'