1#!../perl 2our $DEBUG = @ARGV; 3our (%ORD, %SEQ, $NTESTS); 4BEGIN { 5 if ($ENV{'PERL_CORE'}){ 6 chdir 't'; 7 unshift @INC, '../lib'; 8 } 9 require Config; import Config; 10 if ($Config{'extensions'} !~ /\bEncode\b/) { 11 print "1..0 # Skip: Encode was not built\n"; 12 exit 0; 13 } 14 if ($] <= 5.008 and !$Config{perl_patchlevel}){ 15 print "1..0 # Skip: Perl 5.8.1 or later required\n"; 16 exit 0; 17 } 18 # http://smontagu.damowmow.com/utf8test.html 19 # The numbers below, like 2.1.2 are test numbers on this web page 20 %ORD = ( 21 0x00000080 => 0, # 2.1.2 22 0x00000800 => 0, # 2.1.3 23 0x00010000 => 0, # 2.1.4 24 0x00200000 => 1, # 2.1.5 25 0x00400000 => 1, # 2.1.6 26 0x0000007F => 0, # 2.2.1 -- unmapped okay 27 0x000007FF => 0, # 2.2.2 28 0x0000FFFF => 1, # 2.2.3 29 0x001FFFFF => 1, # 2.2.4 30 0x03FFFFFF => 1, # 2.2.5 31 0x7FFFFFFF => 1, # 2.2.6 32 0x0000D800 => 1, # 5.1.1 33 0x0000DB7F => 1, # 5.1.2 34 0x0000D880 => 1, # 5.1.3 35 0x0000DBFF => 1, # 5.1.4 36 0x0000DC00 => 1, # 5.1.5 37 0x0000DF80 => 1, # 5.1.6 38 0x0000DFFF => 1, # 5.1.7 39 # 5.2 "Paird UTF-16 surrogates skipped 40 # because utf-8-strict raises exception at the first one 41 0x0000FFFF => 1, # 5.3.1 42 ); 43 $NTESTS += scalar keys %ORD; 44 if (ord('A') == 193) { 45 %SEQ = ( 46 qq/dd 64 73 73/ => 0, # 2.3.1 47 qq/dd 67 41 41/ => 0, # 2.3.2 48 qq/ee 42 73 73 71/ => 0, # 2.3.3 49 qq/f4 90 80 80/ => 1, # 2.3.4 -- out of range so NG 50 # EBCDIC TODO: "3 Malformed sequences" 51 # EBCDIC TODO: "4 Overlong sequences" 52 ); 53 } else { 54 %SEQ = ( 55 qq/ed 9f bf/ => 0, # 2.3.1 56 qq/ee 80 80/ => 0, # 2.3.2 57 qq/f4 8f bf bd/ => 0, # 2.3.3 58 qq/f4 90 80 80/ => 1, # 2.3.4 -- out of range so NG 59 qq/80/ => 1, # 3.1.1 60 qq/bf/ => 1, # 3.1.2 61 qq/80 bf/ => 1, # 3.1.3 62 qq/80 bf 80/ => 1, # 3.1.4 63 qq/80 bf 80 bf/ => 1, # 3.1.5 64 qq/80 bf 80 bf 80/ => 1, # 3.1.6 65 qq/80 bf 80 bf 80 bf/ => 1, # 3.1.7 66 qq/80 bf 80 bf 80 bf 80/ => 1, # 3.1.8 67 qq/80 81 82 83 84 85 86 87 88 89 8a 8b 8c 8d 8e 8f 90 91 92 93 94 95 96 97 98 99 9a 9b 9c 9d 9e 9f a0 a1 a2 a3 a4 a5 a6 a7 a8 a9 aa ab ac ad ae af b0 b1 b2 b3 b4 b5 b6 b7 b8 b9 ba bb bc bd be bf/ => 1, # 3.1.9 68 qq/c0 20 c1 20 c2 20 c3 20 c4 20 c5 20 c6 20 c7 20 c8 20 c9 20 ca 20 cb 20 cc 20 cd 20 ce 20 cf 20 d0 20 d1 20 d2 20 d3 20 d4 20 d5 20 d6 20 d7 20 d8 20 d9 20 da 20 db 20 dc 20 dd 20 de 20 df 20/ => 1, # 3.2.1 69 qq/e0 20 e1 20 e2 20 e3 20 e4 20 e5 20 e6 20 e7 20 e8 20 e9 20 ea 20 eb 20 ec 20 ed 20 ee 20 ef 20/ => 1, # 3.2.2 70 qq/f0 20 f1 20 f2 20 f3 20 f4 20 f5 20 f6 20 f7 20/ => 1, # 3.2.3 71 qq/f8 20 f9 20 fa 20 fb 20/ => 1, # 3.2.4 72 qq/fc 20 fd 20/ => 1, # 3.2.5 73 qq/c0/ => 1, # 3.3.1 74 qq/e0 80/ => 1, # 3.3.2 75 qq/f0 80 80/ => 1, # 3.3.3 76 qq/f8 80 80 80/ => 1, # 3.3.4 77 qq/fc 80 80 80 80/ => 1, # 3.3.5 78 qq/df/ => 1, # 3.3.6 79 qq/ef bf/ => 1, # 3.3.7 80 qq/f7 bf bf/ => 1, # 3.3.8 81 qq/fb bf bf bf/ => 1, # 3.3.9 82 qq/fd bf bf bf bf/ => 1, # 3.3.10 83 qq/c0 e0 80 f0 80 80 f8 80 80 80 fc 80 80 80 80 df ef bf f7 bf bf fb bf bf bf fd bf bf bf bf/ => 1, # 3.4.1 84 qq/fe/ => 1, # 3.5.1 85 qq/ff/ => 1, # 3.5.2 86 qq/fe fe ff ff/ => 1, # 3.5.3 87 qq/c0 af/ => 1, # 4.1.1 88 qq/e0 80 af/ => 1, # 4.1.2 89 qq/f0 80 80 af/ => 1, # 4.1.3 90 qq/f8 80 80 80 af/ => 1, # 4.1.4 91 qq/fc 80 80 80 80 af/ => 1, # 4.1.5 92 qq/c1 bf/ => 1, # 4.2.1 93 qq/e0 9f bf/ => 1, # 4.2.2 94 qq/f0 8f bf bf/ => 1, # 4.2.3 95 qq/f8 87 bf bf bf/ => 1, # 4.2.4 96 qq/fc 83 bf bf bf bf/ => 1, # 4.2.5 97 qq/c0 80/ => 1, # 4.3.1 98 qq/e0 80 80/ => 1, # 4.3.2 99 qq/f0 80 80 80/ => 1, # 4.3.3 100 qq/f8 80 80 80 80/ => 1, # 4.3.4 101 qq/fc 80 80 80 80 80/ => 1, # 4.3.5 102 ); 103 } 104 $NTESTS += scalar keys %SEQ; 105} 106use strict; 107use Encode; 108use utf8; 109use Test::More tests => $NTESTS; 110 111local($SIG{__WARN__}) = sub { $DEBUG and $@ and print STDERR $@ }; 112 113my $d = find_encoding("utf-8-strict"); 114for my $u (sort keys %ORD){ 115 my $c = chr($u); 116 eval { $d->encode($c,1) }; 117 $DEBUG and $@ and warn $@; 118 my $t = $@ ? 1 : 0; 119 is($t, $ORD{$u}, sprintf "U+%04X", $u); 120} 121for my $s (sort keys %SEQ){ 122 my $o = pack "C*" => map {hex} split /\s+/, $s; 123 eval { $d->decode($o,1) }; 124 $DEBUG and $@ and warn $@; 125 my $t = $@ ? 1 : 0; 126 is($t, $SEQ{$s}, "sequence: $s"); 127} 128 129__END__ 130 131 132