#!/opt/bin/perl

use dinfo;

for (qw(branche hausnr name ort plz strasse vorname vorwahl zusatz1 zusatz2 zusatz3)) {
   open my $fh, "<", "data/row/$_" or die "$_: $!";
   push @fh, $fh;
}

open NUMMER, "<", "data/dump/nummer" or die "nummer: $!";
open ROW, ">", "data/col/row.txt" or die "row.txt: $!";

$/ = "\0";

$typ{""} = 1;

my $count;

for(;;) {
   for (0 .. $#fh) {
      sysread $fh[$_], my $buf, 4*1024*256;
      $data[$_] = [unpack "V*", $buf];
   }

   last unless @{$data[0]};

   $count += @{$data[0]};
   print "$count\n";
   for my $i (0 .. $#{$data[0]}) {
      $_ = <NUMMER>;
      /^((?:[a-zA-Z][-.\/a-zA-Z0-9]*)?) (?:\s*\+)? \s* ([0-9\ ]+)\0$/x or do {
         warn "ERR: unparseable telnr. '$_'";
         next;
      };

      my $typ = ($typ{$1} ||= 1 + ++$typnr) - 1;
      $_ = $2;

      if (/^0/) {
         s/^01\d\d//;
         s/^06033//;
         s/^0940//;
      }

      11 < length $_ and warn "WARN: truncated <$data[6][$i]:$_>\n";

      $_ = dinfo::nummer2str $_;

      s/([\\\x0a\x09])/\\$1/g; s/\x00/\\0/g;

      print ROW "$data[0][$i]\t$data[1][$i]\t$data[2][$i]\t$data[3][$i]\t$data[4][$i]\t$data[5][$i]\t$data[6][$i]\t$_\t$typ\t$data[7][$i]\t$data[8][$i]\t$data[9][$i]\t$data[10][$i]\n";
   }
}

open TYP, ">", "data/col/typ.txt" or die "typ.txt: $!";
   while (my ($k, $v) = each %typ) {
      $v--;
      print TYP "$v\t$k\n";
   }
close TYP;