whatpm/t/HTML-tokenizer.t

#!/usr/bin/perl
use strict;

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 => 416 }

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

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_input_character} = sub {
        my $self = shift;

        pop @{$self->{prev_input_character}};
        unshift @{$self->{prev_input_character}}, $self->{next_input_character};

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

        if ($self->{next_input_character} == 0x000D) { # CR
          $i++ if substr ($s, $i, 1) eq "\x0A";
          $self->{next_input_character} = 0x000A; # LF # MUST
        } elsif ($self->{next_input_character} > 0x10FFFF) {
          $self->{next_input_character} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
          push @token, 'ParseError';
        } elsif ($self->{next_input_character} == 0x0000) { # NULL
          $self->{next_input_character} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
          push @token, 'ParseError';
        }
      };
      $p->{prev_input_character} = [-1, -1, -1];
      $p->{next_input_character} = -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}}};
        } 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->{correct} ? 1 : 0;
        }

        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: 2008/03/02 14:32:27 $
1	wakaba	1.1	#!/usr/bin/perl
2			use strict;
3
4	wakaba	1.2	my $dir_name;
5	wakaba	1.4	my $test_dir_name;
6	wakaba	1.1	BEGIN {
7	wakaba	1.4	$test_dir_name = 't/';
8	wakaba	1.2	$dir_name = 't/tokenizer/';
9	wakaba	1.1	my $skip = "You don't have JSON module";
10			eval q{
11	wakaba	1.15	use JSON 1.07;
12	wakaba	1.1	$skip = "You don't have make command";
13	wakaba	1.3	system ("cd $test_dir_name; make tokenizer-files") == 0 or die
14	wakaba	1.2	unless -f $dir_name.'test1.test';
15	wakaba	1.1	$skip = '';
16			};
17			if ($skip) {
18			print "1..1\n";
19			print "ok 1 # $skip\n";
20			exit;
21			}
22			$JSON::UnMapping = 1;
23	wakaba	1.2	$JSON::UTF8 = 1;
24	wakaba	1.1	}
25
26			use Test;
27	wakaba	1.23	BEGIN { plan tests => 416 }
28	wakaba	1.2
29	wakaba	1.1	use Data::Dumper;
30	wakaba	1.2	$Data::Dumper::Useqq = 1;
31			sub Data::Dumper::qquote {
32			my $s = shift;
33			$s =~ s/([^\x20\x21-\x26\x28-\x5B\x5D-\x7E])/sprintf '\x{%02X}', ord $1/ge;
34			return q<qq'> . $s . q<'>;
35			} # Data::Dumper::qquote
36	wakaba	1.1
37	wakaba	1.5	use Whatpm::HTML;
38	wakaba	1.1
39	wakaba	1.4	for my $file_name (grep {$_} split /\s+/, qq[
40			${dir_name}test1.test
41			${dir_name}test2.test
42	wakaba	1.15	${dir_name}test3.test
43			${dir_name}test4.test
44	wakaba	1.4	${dir_name}contentModelFlags.test
45	wakaba	1.8	${dir_name}escapeFlag.test
46	wakaba	1.4	${test_dir_name}tokenizer-test-1.test
47	wakaba	1.1	]) {
48	wakaba	1.4	open my $file, '<', $file_name
49			or die "$0: $file_name: $!";
50	wakaba	1.1	local $/ = undef;
51			my $js = <$file>;
52			close $file;
53	wakaba	1.9
54			print "# $file_name\n";
55	wakaba	1.15	$js =~ s{\\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])
56			\\u[Dd]([89A-Fa-f][0-9A-Fa-f][0-9A-Fa-f])}{
57			## NOTE: JSON::Parser does not decode surrogate pair escapes
58			## NOTE: In older version of JSON::Parser, utf8 string will be broken
59			## by parsing. Use latest version!
60			## NOTE: Encode.pm is broken; it converts e.g. U+10FFFF to U+FFFD.
61			my $c = 0x10000;
62			$c += ((((hex $1) & 0b1111111111) << 10) \| ((hex $2) & 0b1111111111));
63			chr $c;
64			}gex;
65	wakaba	1.1	my $tests = jsonToObj ($js)->{tests};
66			TEST: for my $test (@$tests) {
67			my $s = $test->{input};
68
69			my $j = 1;
70			while ($j < @{$test->{output}}) {
71			if (ref $test->{output}->[$j - 1] and
72			$test->{output}->[$j - 1]->[0] eq 'Character' and
73			ref $test->{output}->[$j] and
74			$test->{output}->[$j]->[0] eq 'Character') {
75			$test->{output}->[$j - 1]->[1]
76			.= $test->{output}->[$j]->[1];
77			splice @{$test->{output}}, $j, 1;
78			}
79			$j++;
80			}
81
82	wakaba	1.2	my @cm = @{$test->{contentModelFlags} \|\| ['PCDATA']};
83			my $last_start_tag = $test->{lastStartTag};
84	wakaba	1.1	for my $cm (@cm) {
85	wakaba	1.5	my $p = Whatpm::HTML->new;
86	wakaba	1.1	my $i = 0;
87	wakaba	1.9	my @token;
88	wakaba	1.1	$p->{set_next_input_character} = sub {
89			my $self = shift;
90	wakaba	1.10
91			pop @{$self->{prev_input_character}};
92			unshift @{$self->{prev_input_character}}, $self->{next_input_character};
93
94	wakaba	1.1	$self->{next_input_character} = -1 and return if $i >= length $s;
95			$self->{next_input_character} = ord substr $s, $i++, 1;
96	wakaba	1.2
97			if ($self->{next_input_character} == 0x000D) { # CR
98	wakaba	1.15	$i++ if substr ($s, $i, 1) eq "\x0A";
99	wakaba	1.2	$self->{next_input_character} = 0x000A; # LF # MUST
100			} elsif ($self->{next_input_character} > 0x10FFFF) {
101			$self->{next_input_character} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
102	wakaba	1.9	push @token, 'ParseError';
103	wakaba	1.2	} elsif ($self->{next_input_character} == 0x0000) { # NULL
104			$self->{next_input_character} = 0xFFFD; # REPLACEMENT CHARACTER # MUST
105	wakaba	1.9	push @token, 'ParseError';
106	wakaba	1.2	}
107	wakaba	1.1	};
108	wakaba	1.10	$p->{prev_input_character} = [-1, -1, -1];
109			$p->{next_input_character} = -1;
110	wakaba	1.1
111			$p->{parse_error} = sub {
112			push @token, 'ParseError';
113			};
114
115			$p->_initialize_tokenizer;
116	wakaba	1.18	$p->{content_model} = {
117			CDATA => Whatpm::HTML::CDATA_CONTENT_MODEL (),
118			RCDATA => Whatpm::HTML::RCDATA_CONTENT_MODEL (),
119			PCDATA => Whatpm::HTML::PCDATA_CONTENT_MODEL (),
120			PLAINTEXT => Whatpm::HTML::PLAINTEXT_CONTENT_MODEL (),
121			}->{$cm};
122	wakaba	1.2	$p->{last_emitted_start_tag_name} = $last_start_tag;
123	wakaba	1.1
124			while (1) {
125			my $token = $p->_get_next_token;
126	wakaba	1.19	last if $token->{type} == Whatpm::HTML::END_OF_FILE_TOKEN ();
127	wakaba	1.1
128			my $test_token = [
129			{
130	wakaba	1.19	Whatpm::HTML::DOCTYPE_TOKEN () => 'DOCTYPE',
131			Whatpm::HTML::START_TAG_TOKEN () => 'StartTag',
132			Whatpm::HTML::END_TAG_TOKEN () => 'EndTag',
133			Whatpm::HTML::COMMENT_TOKEN () => 'Comment',
134			Whatpm::HTML::CHARACTER_TOKEN () => 'Character',
135	wakaba	1.1	}->{$token->{type}} \|\| $token->{type},
136			];
137			$test_token->[1] = $token->{tag_name} if defined $token->{tag_name};
138			$test_token->[1] = $token->{data} if defined $token->{data};
139	wakaba	1.19	if ($token->{type} == Whatpm::HTML::START_TAG_TOKEN ()) {
140	wakaba	1.11	$test_token->[2] = {map {$_->{name} => $_->{value}} values %{$token->{attributes}}};
141	wakaba	1.19	} elsif ($token->{type} == Whatpm::HTML::DOCTYPE_TOKEN ()) {
142	wakaba	1.11	$test_token->[1] = $token->{name};
143			$test_token->[2] = $token->{public_identifier};
144			$test_token->[3] = $token->{system_identifier};
145			$test_token->[4] = $token->{correct} ? 1 : 0;
146			}
147
148	wakaba	1.1	if (@token and ref $token[-1] and $token[-1]->[0] eq 'Character' and
149			$test_token->[0] eq 'Character') {
150			$token[-1]->[1] .= $test_token->[1];
151			} else {
152			push @token, $test_token;
153			}
154			}
155
156			my $expected_dump = Dumper ($test->{output});
157			my $parser_dump = Dumper (\@token);
158	wakaba	1.2	ok $parser_dump, $expected_dump,
159	wakaba	1.17	$test->{description} . ': ' . Data::Dumper::qquote ($test->{input});
160	wakaba	1.1	}
161			}
162			}
163
164	wakaba	1.19	## License: Public Domain.
165	wakaba	1.23	## $Date: 2008/03/02 14:32:27 $