wireshark/tools/make-manuf

#!/usr/bin/perl -w
#
# Make-manuf - Creates a file containing ethernet OUIs and their
# company IDs.  It merges the databases at IEEE and
# http://www.cavebear.com/archive/cavebear/Ethernet/Ethernet.txt
# with entries in our template file.
#
# The script reads the comments at the top of "manuf.tmpl" and writes
# them to "manuf".  It then joins the manufacturer listing in "manuf.tmpl"
# with the listing in "oui.txt", "iab.txt", etc, with the entries in 
# "manuf.tmpl" taking precedence.

# LWP is part of the standard Perl module libwww

# As of April 2012 the IEEE content is mostly UTF-8 encoded although some
# of the entries feature sequences listed at
# http://www.i18nqa.com/debug/utf8-debug.html

use Encode;
use open ':encoding(utf8)';

eval "require LWP::UserAgent;";
if( $@ ) {
  die "LWP isn't installed. It is part of the standard Perl\n" .
	" module libwww.  Bailing.\n";
}

$agent    = LWP::UserAgent->new;
$agent->env_proxy;
$agent->agent("Wireshark make-manuf");

$template = "manuf.tmpl";
$wkatmpl  = "wka.tmpl";
$outfile  = "manuf";
$inheader = 1;
$oui_url  = "http://standards.ieee.org/develop/regauth/oui/oui.txt";
$iab_url  = "http://standards.ieee.org/develop/regauth/iab/iab.txt";
$oui28_url = "http://standards.ieee.org/develop/regauth/oui28/oui28.txt";
$oui36_url = "http://standards.ieee.org/develop/regauth/oui36/oui36.txt";
$cb_url   = "http://www.cavebear.com/archive/cavebear/Ethernet/Ethernet.txt";
%oui_list = ();
$hp       = "[0-9a-fA-F]{2}";
$oui_re   = "$hp:$hp:$hp";
$cb_re    = "$hp$hp$hp";
$ieee_re  = "$hp-$hp-$hp";

$min_entries = 100;

$tmpl_added    = 0;
$cb_added      = 0;
$cb_skipped    = 0;
$cb_total      = 0;
$oui_added     = 0;
$oui_skipped   = 0;
$oui_total     = 0;
$iab_added     = 0;
$iab_skipped   = 0;
$iab_total     = 0;
$oui28_added   = 0;
$oui28_skipped = 0;
$oui28_total   = 0;
$oui36_added   = 0;
$oui36_skipped = 0;
$oui36_total   = 0;

sub shorten
{
  my $origmanuf = shift;
  my $manuf = " " . $origmanuf . " ";
  # Remove any punctuation
  $manuf =~ tr/',.()/    /;
  # & isn't needed when Standalone
  $manuf =~ s/ \& / /g;
  # Remove any "the", "inc", "plc" ...
  $manuf =~ s/\s(the|inc|incorporated|plc||systems|corp|corporation|s\/a|a\/s|ab|ag|kg|gmbh|co|company|limited|ltd|holding|spa)(?= )//gi;
  # Convert to consistent case
  $manuf =~ s/(\w+)/\u\L$1/g;
  # Remove all spaces
  $manuf =~ s/\s+//g;
  # Truncate all names to a reasonable length, say, 8 characters.
  # If the string contains UTF-8, this may be substantially more than 8 bytes.
  $manuf = substr($manuf, 0, 8);

  if ($manuf =~ /\Q$origmanuf\E/i) {
    return $manuf;
  } else {
    return sprintf("%-22s # %s", $manuf, $origmanuf);
  }
}

sub fetch
{
  my $url = shift;
  print "Fetching $url.\n";
  $request  = HTTP::Request->new(GET => $url);
  $result   = $agent->request($request);

  if (!$result->is_success) {
    die ("Error fetching $url: " . $result->status_line . "\n");
  }
  return decode("utf-8", $result->content);
}

# Write out the header and populate the OUI list with our entries.

open (TMPL, "< $template") ||
  die "Couldn't open template file for reading ($template)\n";

while ($line = <TMPL>) {
  chomp($line);
  if ($line !~ /^$oui_re\s+\S/ && $inheader) {
    $header .= "$line\n";
  } elsif (($oui, $manuf) = ($line =~ /^($oui_re)\s+(\S.*)$/)) {
    $inheader = 0;
    # Ensure OUI is all upper-case
    $oui =~ tr/a-f/A-F/;
    # $oui_list{$oui} = &shorten($manuf);
    $oui_list{$oui} = $manuf;
    $tmpl_added++;
  }
}

# Add IEEE entries for IABs

$ieee_list = fetch($iab_url);

foreach $line (split(/\n/, $ieee_list)) {
  # determine the OUI used for IAB (currently only 00-50-C2)
  if (($iab_tmp, $manuf) = ($line =~ /^\s*($ieee_re)\s+\(hex\)\s+(\S.*)$/)) {
      $iab_base = $iab_tmp;
  }
  # determine next two bytes
  if (($iab4, $iab5, $manuf) = ($line =~ /^\s*($hp)($hp)$hp-$hp$hp$hp\s+\(base\s16\)\s+(\S.*)$/)) {
    $iab = "$iab_base:$iab4:$iab5:00/36";
    $iab =~ tr /-/:/; # The IEEE bytes are separated by dashes.
    # Ensure IAB is all upper-case
    $iab =~ tr/a-f/A-F/;
    if (exists $oui_list{$iab}) {
      printf "$iab - Skipping IEEE \"$manuf\" in favor of \"$oui_list{$iab}\"\n";
      $iab_skipped++;
    } else {
      $oui_list{$iab} = &shorten($manuf);
      $iab_added++;
    }
  }
  $iab_total++;
}

if ($iab_total < $min_entries) { die "Too few IAB entries ($iab_total)\n"; }

# Add IEEE entries for OUI-28

$ieee_list = fetch($oui28_url);

foreach $line (split(/\n/, $ieee_list)) {
  # determine the OUI used for OUI-28
  if (($oui28_tmp, $manuf) = ($line =~ /^\s*($ieee_re)\s+\(hex\)\s+(\S.*)$/)) {
      $oui28_base = $oui28_tmp;
  }
  # determine next two bytes
  if (($oui28_4, $oui28_5, $manuf) = ($line =~ /^\s*($hp)($hp)$hp-$hp$hp$hp\s+\(base\s16\)\s+(\S.*)$/)) {
    $oui28 = "$oui28_base:$oui28_4:$oui28_5:00/28";
    $oui28 =~ tr /-/:/; # The IEEE bytes are separated by dashes.
    # Ensure OUI-28 is all upper-case
    $oui28 =~ tr/a-f/A-F/;
    if (exists $oui_list{$oui28}) {
      printf "$oui28 - Skipping IEEE \"$manuf\" in favor of \"$oui_list{$oui28}\"\n";
      $oui28_skipped++;
    } else {
      $oui_list{$oui28} = &shorten($manuf);
      $oui28_added++;
    }
  }
  $oui28_total++;
}

if ($oui28_total < $min_entries) { die "Too few OUI-28 entries ($oui28_total)\n"; }

# Add IEEE entries for OUI-36

$ieee_list = fetch($oui36_url);

foreach $line (split(/\n/, $ieee_list)) {
  # determine the OUI used for OUI-36 (currently only 00-1B-C5)
  if (($oui36_tmp, $manuf) = ($line =~ /^\s*($ieee_re)\s+\(hex\)\s+(\S.*)$/)) {
      $oui36_base = $oui36_tmp;
  }
  # determine next two bytes
  if (($oui36_4, $oui36_5, $manuf) = ($line =~ /^\s*($hp)($hp)$hp-$hp$hp$hp\s+\(base\s16\)\s+(\S.*)$/)) {
    $oui36 = "$oui36_base:$oui36_4:$oui36_5:00/36";
    $oui36 =~ tr /-/:/; # The IEEE bytes are separated by dashes.
    # Ensure OUI-36 is all upper-case
    $oui36 =~ tr/a-f/A-F/;
    if (exists $oui_list{$oui36}) {
      printf "$oui36 - Skipping IEEE \"$manuf\" in favor of \"$oui_list{$oui36}\"\n";
      $oui36_skipped++;
    } else {
      $oui_list{$oui36} = &shorten($manuf);
      $oui36_added++;
    }
  }
  $oui36_total++;
}

if ($oui36_total < $min_entries) { die "Too few OUI-36 entries ($oui36_total)\n"; }

# Add IEEE entries for OUIs not yet known.

$ieee_list = fetch($oui_url);

foreach $line (split(/\n/, $ieee_list)) {
  if (($oui, $manuf) = ($line =~ /^\s*($ieee_re)\s+\(hex\)\s+(\S.*)$/)) {
    $oui =~ tr /-/:/;  # The IEEE bytes are separated by dashes.
    # Ensure OUI is all upper-case
    $oui =~ tr/a-f/A-F/;
    if (exists $oui_list{$oui}) {
      printf "$oui - Skipping IEEE \"$manuf\" in favor of \"$oui_list{$oui}\"\n";
      $oui_skipped++;
    } else {
      $oui_list{$oui} = &shorten($manuf);
      $oui_added++;
    }
  }
  $oui_total++;
}

if ($oui_total < $min_entries) { die "Too few OUI entries ($oui_total)\n"; }

# Add CaveBear entries for OUIs not yet known.

$cb_list = fetch($cb_url);

foreach $line (split(/\n/, $cb_list)) {
  if (($oui, $manuf) = ($line =~ /^($cb_re)\s+(\S.*)$/)) {
    ($h1, $h2, $h3) = ($oui =~ /($hp)($hp)($hp)/);  # The CaveBear bytes have no separators
    $oui = "$h1:$h2:$h3";
    # Ensure OUI is all upper-case
    $oui =~ tr/a-f/A-F/;
    if (exists $oui_list{$oui}) {
      # printf "$oui - Skipping CaveBear \"$manuf\" in favor of \"$oui_list{$oui}\"\n";
      $cb_skipped++;
    } else {
      printf "$oui - adding \"$manuf\" from CaveBear\n";
      $oui_list{$oui} = &shorten($manuf);
      $cb_added++;
    }
  }
  $cb_total++;
}

if ($cb_total < $min_entries) { die "Too few CaveBear entries ($cb_total)\n"; }

# Write output file

open (OUT, "> $outfile") ||
  die "Couldn't open output file for writing ($outfile)\n";

print(OUT "# This file was generated by running ./make-manuf.\n");
print(OUT "# Don't change it directly, change manuf.tmpl and wka.tmpl instead.\n#\n");
print(OUT "$header");

foreach $oui (sort(keys %oui_list)) {
  print(OUT "$oui\t$oui_list{$oui}\n");
}

# Write out a blank line separating the OUIs from the well-known
# addresses, and then read the well-known address template file
# and write it to the manuf file.

open (WKATMPL, "< $wkatmpl") ||
  die "Couldn't open well-known address template file for reading ($wkatmpl)\n";

# XXX - it'd be nice to get this from the Cavebear file, but inferring
# the address mask from entries in that file involves some work.
#
print(OUT "\n");
while ($line = <WKATMPL>) {
  chomp($line);
  print(OUT "$line\n");
}

$total_added = $tmpl_added + $cb_added + $oui_added + $iab_added;
print <<"Fin"
Original entries : $tmpl_added
IEEE OUI added   : $oui_added
IEEE IAB added   : $iab_added
IEEE OUI28 added : $oui28_added
IEEE OUI36 added : $oui36_added
CaveBear added   : $cb_added
Total            : $total_added

IEEE OUI skipped : $oui_skipped
IEEE IAB skipped : $iab_skipped
IEEE OUI28 skipd : $oui28_skipped
IEEE OUI36 skipd : $oui36_skipped
CaveBear skipped : $cb_skipped
Fin