whatpm/t/HTML-tokenizer.t

#!/usr/bin/perl
use strict;

my $DEBUG = $ENV{DEBUG};

my $dir_name;
my $test_dir_name;
BEGIN {
  $test_dir_name = 't/';
  $dir_name = 't/tokenizer/';
  my $skip = "You don't have JSON module";
  eval q{
         use JSON 1.07;
         $skip = "You don't have make command";
         system ("cd $test_dir_name; make tokenizer-files") == 0 or die
           unless -f $dir_name.'test1.test';
         $skip = '';
        };
  if ($skip) {
    print "1..1\n";
    print "ok 1 # $skip\n";
    exit;
  }
  $JSON::UnMapping = 1;
  $JSON::UTF8 = 1;
}

use Test;
BEGIN { plan tests => 477 }

use Data::Dumper;
$Data::Dumper::Useqq = 1;
sub Data::Dumper::qquote {
  my $s = shift;
  $s =~ s/([^\x20\x21-\x26\x28-\x5B\x5D-\x7E])/sprintf '\x{%02X}', ord $1/ge;
  return q<qq'> . $s . q<'>;
} # Data::Dumper::qquote

if ($DEBUG) {
  my $not_found = {%{$Whatpm::HTML::Debug::cp or {}}};

  $Whatpm::HTML::Debug::cp_pass = sub {
    my $id = shift;
    delete $not_found->{$id};
  };

  END {
    for my $id (sort {$a <=> $b || $a cmp $b} grep {!/^[ti]/}
                keys %$not_found) {
      print "# checkpoint $id is not reached\n";
    }
  }
}

use Whatpm::HTML;

for my $file_name (grep {$_} split /\s+/, qq[
                      ${dir_name}test1.test
                      ${dir_name}test2.test
                      ${dir_name}test3.test
                      ${dir_name}test4.test
                      ${dir_name}contentModelFlags.test
                      ${dir_name}escapeFlag.test
                      ${test_dir_name}tokenizer-test-1.test
                     ]) {
  open my $file, '<', $file_name
    or die "$0: $file_name: $!";
  local $/ = undef;
  my $js = <$file>;
  close $file;

  print "# $file_name\n";
  $js =~ s{\\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])
      \\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])}{
    ## NOTE: JSON::Parser does not decode surrogate pair escapes
    ## NOTE: In older version of JSON::Parser, utf8 string will be broken
    ## by parsing.  Use latest version!
    ## NOTE: Encode.pm is broken; it converts e.g. U+10FFFF to U+FFFD.
    my $c = 0x10000;
    $c += ((((hex $1) & 0b1111111111) << 10) | ((hex $2) & 0b1111111111));
    chr $c;
  }gex;
  my $tests = jsonToObj ($js)->{tests};
  TEST: for my $test (@$tests) {
    my $s = $test->{input};
    
    my $j = 1;
    while ($j < @{$test->{output}}) {
      if (ref $test->{output}->[$j - 1] and
          $test->{output}->[$j - 1]->[0] eq 'Character' and
          ref $test->{output}->[$j] and 
          $test->{output}->[$j]->[0] eq 'Character') {
        $test->{output}->[$j - 1]->[1]
          .= $test->{output}->[$j]->[1];
        splice @{$test->{output}}, $j, 1;
      }
      $j++;
    }

    my @cm = @{$test->{contentModelFlags} || ['PCDATA']};
    my $last_start_tag = $test->{lastStartTag};
    for my $cm (@cm) {
      my $p = Whatpm::HTML->new;
      my $i = 0;
      my @token;
      $p->{set_next_char} = sub {
        my $self = shift;

        pop @{$self->{prev_char}};
        unshift @{$self->{prev_char}}, $self->{next_char};

        $self->{next_char} = -1 and return if $i >= length $s;
        $self->{next_char} = ord substr $s, $i++, 1;

        if ($self->{next_char} == 0x000D) { # CR
          $i++ if substr ($s, $i, 1) eq "\x0A";
          $self->{next_char} = 0x000A; # LF # MUST
        } elsif ($self->{next_char} > 0x10FFFF) {
          $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
          push @token, 'ParseError';
        } elsif ($self->{next_char} == 0x0000) { # NULL
          $self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
          push @token, 'ParseError';
        }
      };
      $p->{prev_char} = [-1, -1, -1];
      $p->{next_char} = -1;
      
      $p->{parse_error} = sub {
        push @token, 'ParseError';
      };
      
      $p->_initialize_tokenizer;
      $p->{content_model} = {
        CDATA => Whatpm::HTML::CDATA_CONTENT_MODEL (),
        RCDATA => Whatpm::HTML::RCDATA_CONTENT_MODEL (),
        PCDATA => Whatpm::HTML::PCDATA_CONTENT_MODEL (),
        PLAINTEXT => Whatpm::HTML::PLAINTEXT_CONTENT_MODEL (),
      }->{$cm};
      $p->{last_emitted_start_tag_name} = $last_start_tag;

      while (1) {
        my $token = $p->_get_next_token;
        last if $token->{type} == Whatpm::HTML::END_OF_FILE_TOKEN ();
        
        my $test_token = [
         {
          Whatpm::HTML::DOCTYPE_TOKEN () => 'DOCTYPE',
          Whatpm::HTML::START_TAG_TOKEN () => 'StartTag',
          Whatpm::HTML::END_TAG_TOKEN () => 'EndTag',
          Whatpm::HTML::COMMENT_TOKEN () => 'Comment',
          Whatpm::HTML::CHARACTER_TOKEN () => 'Character',
         }->{$token->{type}} || $token->{type},
        ];
        $test_token->[1] = $token->{tag_name} if defined $token->{tag_name};
        $test_token->[1] = $token->{data} if defined $token->{data};
        if ($token->{type} == Whatpm::HTML::START_TAG_TOKEN ()) {
          $test_token->[2] = {map {$_->{name} => $_->{value}} values %{$token->{attributes}}};
          if ({
               ## NOTE: Permitted slash
               br => 1, link => 1,
              }->{$token->{tag_name}}) {
            delete $p->{self_closing};
          }
        } elsif ($token->{type} == Whatpm::HTML::DOCTYPE_TOKEN ()) {
          $test_token->[1] = $token->{name};
          $test_token->[2] = $token->{public_identifier};
          $test_token->[3] = $token->{system_identifier};
          $test_token->[4] = $token->{quirks} ? 0 : 1;
        }

        if (@token and ref $token[-1] and $token[-1]->[0] eq 'Character' and
            $test_token->[0] eq 'Character') {
          $token[-1]->[1] .= $test_token->[1];
        } else {
          push @token, $test_token;
        }
      }
      
      my $expected_dump = Dumper ($test->{output});
      my $parser_dump = Dumper (\@token);
      ok $parser_dump, $expected_dump,
        $test->{description} . ': ' . Data::Dumper::qquote ($test->{input});
    }
  }
}

## License: Public Domain.
## $Date: 2008/03/28 14:23:09 $
1	#!/usr/bin/perl
2	use strict;
3
4	my $DEBUG = $ENV{DEBUG};
5
6	my $dir_name;
7	my $test_dir_name;
8	BEGIN {
9	$test_dir_name = 't/';
10	$dir_name = 't/tokenizer/';
11	my $skip = "You don't have JSON module";
12	eval q{
13	use JSON 1.07;
14	$skip = "You don't have make command";
15	system ("cd $test_dir_name; make tokenizer-files") == 0 or die
16	unless -f $dir_name.'test1.test';
17	$skip = '';
18	};
19	if ($skip) {
20	print "1..1\n";
21	print "ok 1 # $skip\n";
22	exit;
23	}
24	$JSON::UnMapping = 1;
25	$JSON::UTF8 = 1;
26	}
27
28	use Test;
29	BEGIN { plan tests => 477 }
30
31	use Data::Dumper;
32	$Data::Dumper::Useqq = 1;
33	sub Data::Dumper::qquote {
34	my $s = shift;
35	$s =~ s/([^\x20\x21-\x26\x28-\x5B\x5D-\x7E])/sprintf '\x{%02X}', ord $1/ge;
36	return q<qq'> . $s . q<'>;
37	} # Data::Dumper::qquote
38
39	if ($DEBUG) {
40	my $not_found = {%{$Whatpm::HTML::Debug::cp or {}}};
41
42	$Whatpm::HTML::Debug::cp_pass = sub {
43	my $id = shift;
44	delete $not_found->{$id};
45	};
46
47	END {
48	for my $id (sort {$a <=> $b \|\| $a cmp $b} grep {!/^[ti]/}
49	keys %$not_found) {
50	print "# checkpoint $id is not reached\n";
51	}
52	}
53	}
54
55	use Whatpm::HTML;
56
57	for my $file_name (grep {$_} split /\s+/, qq[
58	${dir_name}test1.test
59	${dir_name}test2.test
60	${dir_name}test3.test
61	${dir_name}test4.test
62	${dir_name}contentModelFlags.test
63	${dir_name}escapeFlag.test
64	${test_dir_name}tokenizer-test-1.test
65	]) {
66	open my $file, '<', $file_name
67	or die "$0: $file_name: $!";
68	local $/ = undef;
69	my $js = <$file>;
70	close $file;
71
72	print "# $file_name\n";
73	$js =~ s{\\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])
74	\\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])}{
75	## NOTE: JSON::Parser does not decode surrogate pair escapes
76	## NOTE: In older version of JSON::Parser, utf8 string will be broken
77	## by parsing. Use latest version!
78	## NOTE: Encode.pm is broken; it converts e.g. U+10FFFF to U+FFFD.
79	my $c = 0x10000;
80	$c += ((((hex $1) & 0b1111111111) << 10) \| ((hex $2) & 0b1111111111));
81	chr $c;
82	}gex;
83	my $tests = jsonToObj ($js)->{tests};
84	TEST: for my $test (@$tests) {
85	my $s = $test->{input};
86
87	my $j = 1;
88	while ($j < @{$test->{output}}) {
89	if (ref $test->{output}->[$j - 1] and
90	$test->{output}->[$j - 1]->[0] eq 'Character' and
91	ref $test->{output}->[$j] and
92	$test->{output}->[$j]->[0] eq 'Character') {
93	$test->{output}->[$j - 1]->[1]
94	.= $test->{output}->[$j]->[1];
95	splice @{$test->{output}}, $j, 1;
96	}
97	$j++;
98	}
99
100	my @cm = @{$test->{contentModelFlags} \|\| ['PCDATA']};
101	my $last_start_tag = $test->{lastStartTag};
102	for my $cm (@cm) {
103	my $p = Whatpm::HTML->new;
104	my $i = 0;
105	my @token;
106	$p->{set_next_char} = sub {
107	my $self = shift;
108
109	pop @{$self->{prev_char}};
110	unshift @{$self->{prev_char}}, $self->{next_char};
111
112	$self->{next_char} = -1 and return if $i >= length $s;
113	$self->{next_char} = ord substr $s, $i++, 1;
114
115	if ($self->{next_char} == 0x000D) { # CR
116	$i++ if substr ($s, $i, 1) eq "\x0A";
117	$self->{next_char} = 0x000A; # LF # MUST
118	} elsif ($self->{next_char} > 0x10FFFF) {
119	$self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
120	push @token, 'ParseError';
121	} elsif ($self->{next_char} == 0x0000) { # NULL
122	$self->{next_char} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
123	push @token, 'ParseError';
124	}
125	};
126	$p->{prev_char} = [-1, -1, -1];
127	$p->{next_char} = -1;
128
129	$p->{parse_error} = sub {
130	push @token, 'ParseError';
131	};
132
133	$p->_initialize_tokenizer;
134	$p->{content_model} = {
135	CDATA => Whatpm::HTML::CDATA_CONTENT_MODEL (),
136	RCDATA => Whatpm::HTML::RCDATA_CONTENT_MODEL (),
137	PCDATA => Whatpm::HTML::PCDATA_CONTENT_MODEL (),
138	PLAINTEXT => Whatpm::HTML::PLAINTEXT_CONTENT_MODEL (),
139	}->{$cm};
140	$p->{last_emitted_start_tag_name} = $last_start_tag;
141
142	while (1) {
143	my $token = $p->_get_next_token;
144	last if $token->{type} == Whatpm::HTML::END_OF_FILE_TOKEN ();
145
146	my $test_token = [
147	{
148	Whatpm::HTML::DOCTYPE_TOKEN () => 'DOCTYPE',
149	Whatpm::HTML::START_TAG_TOKEN () => 'StartTag',
150	Whatpm::HTML::END_TAG_TOKEN () => 'EndTag',
151	Whatpm::HTML::COMMENT_TOKEN () => 'Comment',
152	Whatpm::HTML::CHARACTER_TOKEN () => 'Character',
153	}->{$token->{type}} \|\| $token->{type},
154	];
155	$test_token->[1] = $token->{tag_name} if defined $token->{tag_name};
156	$test_token->[1] = $token->{data} if defined $token->{data};
157	if ($token->{type} == Whatpm::HTML::START_TAG_TOKEN ()) {
158	$test_token->[2] = {map {$_->{name} => $_->{value}} values %{$token->{attributes}}};
159	if ({
160	## NOTE: Permitted slash
161	br => 1, link => 1,
162	}->{$token->{tag_name}}) {
163	delete $p->{self_closing};
164	}
165	} elsif ($token->{type} == Whatpm::HTML::DOCTYPE_TOKEN ()) {
166	$test_token->[1] = $token->{name};
167	$test_token->[2] = $token->{public_identifier};
168	$test_token->[3] = $token->{system_identifier};
169	$test_token->[4] = $token->{quirks} ? 0 : 1;
170	}
171
172	if (@token and ref $token[-1] and $token[-1]->[0] eq 'Character' and
173	$test_token->[0] eq 'Character') {
174	$token[-1]->[1] .= $test_token->[1];
175	} else {
176	push @token, $test_token;
177	}
178	}
179
180	my $expected_dump = Dumper ($test->{output});
181	my $parser_dump = Dumper (\@token);
182	ok $parser_dump, $expected_dump,
183	$test->{description} . ': ' . Data::Dumper::qquote ($test->{input});
184	}
185	}
186	}
187
188	## License: Public Domain.
189	## $Date: 2008/03/28 14:23:09 $