===================================================================
RCS file: /cvs/pta/pta_import.pl,v
retrieving revision 1.25
retrieving revision 1.28
diff -u -p -r1.25 -r1.28
--- pta/pta_import.pl	2020/11/29 14:00:50	1.25
+++ pta/pta_import.pl	2020/12/11 19:28:12	1.28
@@ -32,36 +32,15 @@ my %banks = (
     "wellsfargo"            => \&import_wellsfargo,
 );
 
+my %date_formats = (
+    'MM/DD/YYYY' => 's#(\d+)/(\d+)/(\d+)#$3$1$2#',
+    'MM/DD/YY'   => 's#(\d+)/(\d+)/(\d+)#20$3$1$2#',
+    'YYYY-MM-DD' => 's#(\d+)-(\d+)-(\d+)#$1$2$3#',
+    'DD.MM.YY'   => 's#(\d+)\.(\d+)\.(\d+)#20$3$2$1#',
+);
+
 # === SUBROUTINES  =====================================================
-sub get_accounts_by_amount_sign ($$$) {
-	my ($amount, $account, $csv_account) = @_;
-	my ($debit, $credit);
-	if ($amount <= 0) {
-		$amount = substr $amount, 1;
-		$credit = $csv_account;
-		$debit = $account;
-	} else {
-		$debit = $csv_account;
-		$credit = $account;
-	}
-	return ($amount, $debit, $credit);
-}
 
-sub get_accounts_by_csv_col ($$$$) {
-	my ($account, $csv_account, $csv_debit, $csv_credit) = @_;
-	my ($amount, $debit, $credit);
-	if ($csv_debit eq "") {
-		$amount = $csv_credit;
-		$credit = $account;
-		$debit = $csv_account;
-	} else {
-		$amount = $csv_debit;
-		$credit = $csv_account;
-		$debit = $account;
-	}
-	return ($amount, $debit, $credit);
-}
-
 sub usage () {
 	printf STDERR "usage: %s -I accountname csvfilename\n", $0;
 	exit 1;
@@ -80,57 +59,76 @@ $banks{$account_name} or die "unknown accountname: $ac
 # Parse the configuration file.
 my $fn = "import/" . $account_name . ".txt";
 open my $in, '<', $fn or die "$fn: $!";
-my ($cost_center_field, $csv_account, $date_field, $date_format,
+my ($cost_center_field, $csv_account, $date_field, $date_regex,
     $delim, $header, $quantity_field, $quote, @amount_fields,
     @compiled, @description_fields, @ignored);
 while (<$in>) {
 	chomp;
+	s/\s+$//;
 	next if /^(?:#|$)/;
 	my $line = $_;
-	if (/^ACCOUNT\s+(\S+)$/) {
-		$csv_account and die "duplicate ACCOUNT line: $1";
+	if (s/^ACCOUNT\s+//) {
+		$csv_account and die "duplicate ACCOUNT line: $_";
+		/^(\d+)$/ or die "ACCOUNT parse error: $_";
 		$csv_account = $1;
 		next;
 	}
 	if (s/^AMOUNT\s+//) {
+		@amount_fields and die "duplicate AMOUNT line: $_";
 		push @amount_fields, $1 - 1 while s/(\d+)\s*//;
 		$_ eq '' or die "trailing garbage: AMOUNT ... $_";
 		next;
 	}
-	if (/^COSTCENTER\s+(\d+)/) {
+	if (s/^COSTCENTER\s+//) {
+		$cost_center_field and
+		    die "duplicate COSTCENTER line: $_";
+		/^(\d+)$/ or die "COSTCENTER parse error: $_";
 		$cost_center_field = $1 - 1;
 		next;
 	}
-	if (/^DATE\s+(\d+)\s+(\S+)/) {
-		$date_field = $1 - 1;
-		$date_format = $2;
+	if (s/^DATE\s+//) {
+		$date_field || $date_regex and
+		    die "duplicate DATE line: $_";
+		s/^(\d+)\s+(\S+)$//;
+		$date_field = $1 or
+		    die "DATE date_field parse error: $_";
+		$date_field -= 1;
+		$date_regex = $date_formats{$2}
+		    or die "unknown date format: $2";
 		next;
 	}
-	if (/^DELIM\s+(\S)$/) {
+	if (s/^DELIM\s+//) {
 		$delim and die "duplicate DELIM line: $1";
+		/^([^|\^\$\*\+\?\(\)\[\]\{\}\\])$/ or
+		    die "DELIM parse error: $_";
 		$delim = $1;
 		next;
 	}
 	if (s/^DESCRIPTION\s+//) {
-	    push @description_fields, $1 - 1 while s/(\d+)\s*//;
-	    $_ eq '' or die "trailing garbage: DESCRIPTION ... $_";
-	    next;
+		@description_fields and die "duplicate DESCRIPTION line: $_";
+		push @description_fields, $1 - 1 while s/(\d+)\s*//;
+		$_ eq '' or die "trailing garbage: DESCRIPTION ... $_";
+		next;
 	}
-	if (/^HEADER\s+(.*)$/) {
+	if (s/^HEADER\s+//) {
 		$header and die "duplicate HEADER line: $1";
-		$header = $1;
+		$header = $_;
 		next;
 	}
-	if (/^IGNORE\s+(.*)/) {
-		push @ignored, qr/$1/;
+	if (s/^IGNORE\s+//) {
+		push @ignored, qr/$_/;
 		next;
 	}
-	if (/^QUANTITY\s+(\d+)/) {
+	if (s/^QUANTITY\s+//) {
+		$quantity_field and die "duplicate QUANTITY line: $1";
+		/^(\d+)$/ or die "QUANTITY parse error: $_";
 		$quantity_field = $1 - 1;
 		next;
 	}
-	if (/^QUOTE\s+(\S)$/) {
+	if (s/^QUOTE\s+//) {
 		$quote and die "duplicate QUOTE line: $1";
+		/^([^|\^\$\*\+\?\(\)\[\]\{\}\\])$/ or
+		die "QUOTE parse error: $_";
 		$quote = $1;
 		next;
 	}
@@ -152,6 +150,7 @@ if (@ARGV) {
 }
 LINE: while (<STDIN>) {
 	chomp;
+	s/\s*$//;
 	next if (/^$/);
 	foreach my $ignore (@ignored) {
 		next LINE if /$ignore/;
@@ -188,17 +187,7 @@ LINE: while (<STDIN>) {
 	$matches or die "unmatched CSV line: $line";
 	my $date = $fields[$date_field] or
 	    die "date parse error: $line";
-	if ($date_format eq "MM/DD/YYYY") {
-		$date =~ s#(\d+)/(\d+)/(\d+)#$3$1$2#;
-	} elsif ($date_format eq "MM/DD/YY") {
-		$date =~ s#(\d+)/(\d+)/(\d+)#20$3$1$2#;
-	} elsif ($date_format eq "YYYY-MM-DD") {
-		$date =~ s#(\d+)-(\d+)-(\d+)#$1$2$3#;
-	} elsif ($date_format eq "DD.MM.YY") {
-		$date =~ s#(\d+)\.(\d+)\.(\d+)#20$3$2$1#;
-	} else {
-		die "unhandled date format: $line";
-	}
+	eval '$date =~ ' . $date_regex;
 	foreach my $i (@amount_fields) {
 		if (defined($fields[$i])) {
 			$fields[$i] =~ s/,/./;
@@ -206,19 +195,16 @@ LINE: while (<STDIN>) {
 			$fields[$i] = "-$1" if $fields[$i] =~ /^\((\d+\.\d+)\)/;
 		}
 	}
-	my $description = join ' ', @fields[@description_fields];
-	my ($amount, $credit, $debit);
-	if ($#amount_fields) {
-		($amount, $debit, $credit) =
-		    get_accounts_by_csv_col($account,
-		    $csv_account, $fields[$amount_fields[0]],
-		    $fields[$amount_fields[1]]);
-	} else {
-		($amount, $debit, $credit) =
-		    get_accounts_by_amount_sign(
-		        $fields[$amount_fields[0]],
-		        $account, $csv_account);
+	my $debit = $csv_account;
+	my $credit = $account;
+	my $amount = $fields[$amount_fields[-1]] || -$fields[$amount_fields[0]];
+	if ($amount < 0) {
+		$amount *= -1;
+		$credit = $csv_account;
+		$debit = $account;
 	}
+	$amount = sprintf "%.2f", $amount;
+	my $description = join ' ', @fields[@description_fields];
 	$date && $amount && $debit && $credit && $description
 	    or die "import parse error: $line";
 	$description =~ s/#//g;