html5 0.1.0
Sign up to get free protection for your applications and to get access to all the features.
- data/History.txt +3 -0
- data/Manifest.txt +58 -0
- data/README +9 -0
- data/Rakefile.rb +17 -0
- data/lib/html5/constants.rb +818 -0
- data/lib/html5/filters/base.rb +10 -0
- data/lib/html5/filters/inject_meta_charset.rb +82 -0
- data/lib/html5/filters/optionaltags.rb +198 -0
- data/lib/html5/filters/sanitizer.rb +15 -0
- data/lib/html5/filters/whitespace.rb +36 -0
- data/lib/html5/html5parser/after_body_phase.rb +46 -0
- data/lib/html5/html5parser/after_frameset_phase.rb +34 -0
- data/lib/html5/html5parser/after_head_phase.rb +50 -0
- data/lib/html5/html5parser/before_head_phase.rb +41 -0
- data/lib/html5/html5parser/in_body_phase.rb +607 -0
- data/lib/html5/html5parser/in_caption_phase.rb +68 -0
- data/lib/html5/html5parser/in_cell_phase.rb +78 -0
- data/lib/html5/html5parser/in_column_group_phase.rb +55 -0
- data/lib/html5/html5parser/in_frameset_phase.rb +57 -0
- data/lib/html5/html5parser/in_head_phase.rb +138 -0
- data/lib/html5/html5parser/in_row_phase.rb +87 -0
- data/lib/html5/html5parser/in_select_phase.rb +84 -0
- data/lib/html5/html5parser/in_table_body_phase.rb +83 -0
- data/lib/html5/html5parser/in_table_phase.rb +110 -0
- data/lib/html5/html5parser/initial_phase.rb +134 -0
- data/lib/html5/html5parser/phase.rb +158 -0
- data/lib/html5/html5parser/root_element_phase.rb +42 -0
- data/lib/html5/html5parser/trailing_end_phase.rb +35 -0
- data/lib/html5/html5parser.rb +248 -0
- data/lib/html5/inputstream.rb +654 -0
- data/lib/html5/liberalxmlparser.rb +158 -0
- data/lib/html5/sanitizer.rb +188 -0
- data/lib/html5/serializer/htmlserializer.rb +180 -0
- data/lib/html5/serializer/xhtmlserializer.rb +20 -0
- data/lib/html5/serializer.rb +2 -0
- data/lib/html5/tokenizer.rb +968 -0
- data/lib/html5/treebuilders/base.rb +334 -0
- data/lib/html5/treebuilders/hpricot.rb +231 -0
- data/lib/html5/treebuilders/rexml.rb +208 -0
- data/lib/html5/treebuilders/simpletree.rb +185 -0
- data/lib/html5/treebuilders.rb +24 -0
- data/lib/html5/treewalkers/base.rb +154 -0
- data/lib/html5/treewalkers/hpricot.rb +48 -0
- data/lib/html5/treewalkers/rexml.rb +48 -0
- data/lib/html5/treewalkers/simpletree.rb +48 -0
- data/lib/html5/treewalkers.rb +26 -0
- data/lib/html5.rb +13 -0
- data/parse.rb +217 -0
- data/tests/preamble.rb +82 -0
- data/tests/test_encoding.rb +35 -0
- data/tests/test_lxp.rb +263 -0
- data/tests/test_parser.rb +68 -0
- data/tests/test_sanitizer.rb +142 -0
- data/tests/test_serializer.rb +68 -0
- data/tests/test_stream.rb +62 -0
- data/tests/test_tokenizer.rb +94 -0
- data/tests/test_treewalkers.rb +116 -0
- data/tests/tokenizer_test_parser.rb +63 -0
- metadata +120 -0
@@ -0,0 +1,134 @@
|
|
1
|
+
require 'html5/html5parser/phase'
|
2
|
+
|
3
|
+
module HTML5
|
4
|
+
class InitialPhase < Phase
|
5
|
+
|
6
|
+
# This phase deals with error handling as well which is currently not
|
7
|
+
# covered in the specification. The error handling is typically known as
|
8
|
+
# "quirks mode". It is expected that a future version of HTML5 will define this.
|
9
|
+
|
10
|
+
def process_eof
|
11
|
+
parse_error(_('Unexpected End of file. Expected DOCTYPE.'))
|
12
|
+
@parser.phase = @parser.phases[:rootElement]
|
13
|
+
@parser.phase.process_eof
|
14
|
+
end
|
15
|
+
|
16
|
+
def processComment(data)
|
17
|
+
@tree.insert_comment(data, @tree.document)
|
18
|
+
end
|
19
|
+
|
20
|
+
def processDoctype(name, publicId, systemId, correct)
|
21
|
+
if name.downcase != 'html' or publicId or systemId
|
22
|
+
parse_error(_('Erroneous DOCTYPE.'))
|
23
|
+
end
|
24
|
+
# XXX need to update DOCTYPE tokens
|
25
|
+
@tree.insertDoctype(name, publicId, systemId)
|
26
|
+
|
27
|
+
publicId = publicId.to_s.upcase
|
28
|
+
|
29
|
+
if name.downcase != 'html'
|
30
|
+
# XXX quirks mode
|
31
|
+
else
|
32
|
+
if ["+//silmaril//dtd html pro v0r11 19970101//en",
|
33
|
+
"-//advasoft ltd//dtd html 3.0 aswedit + extensions//en",
|
34
|
+
"-//as//dtd html 3.0 aswedit + extensions//en",
|
35
|
+
"-//ietf//dtd html 2.0 level 1//en",
|
36
|
+
"-//ietf//dtd html 2.0 level 2//en",
|
37
|
+
"-//ietf//dtd html 2.0 strict level 1//en",
|
38
|
+
"-//ietf//dtd html 2.0 strict level 2//en",
|
39
|
+
"-//ietf//dtd html 2.0 strict//en",
|
40
|
+
"-//ietf//dtd html 2.0//en",
|
41
|
+
"-//ietf//dtd html 2.1e//en",
|
42
|
+
"-//ietf//dtd html 3.0//en",
|
43
|
+
"-//ietf//dtd html 3.0//en//",
|
44
|
+
"-//ietf//dtd html 3.2 final//en",
|
45
|
+
"-//ietf//dtd html 3.2//en",
|
46
|
+
"-//ietf//dtd html 3//en",
|
47
|
+
"-//ietf//dtd html level 0//en",
|
48
|
+
"-//ietf//dtd html level 0//en//2.0",
|
49
|
+
"-//ietf//dtd html level 1//en",
|
50
|
+
"-//ietf//dtd html level 1//en//2.0",
|
51
|
+
"-//ietf//dtd html level 2//en",
|
52
|
+
"-//ietf//dtd html level 2//en//2.0",
|
53
|
+
"-//ietf//dtd html level 3//en",
|
54
|
+
"-//ietf//dtd html level 3//en//3.0",
|
55
|
+
"-//ietf//dtd html strict level 0//en",
|
56
|
+
"-//ietf//dtd html strict level 0//en//2.0",
|
57
|
+
"-//ietf//dtd html strict level 1//en",
|
58
|
+
"-//ietf//dtd html strict level 1//en//2.0",
|
59
|
+
"-//ietf//dtd html strict level 2//en",
|
60
|
+
"-//ietf//dtd html strict level 2//en//2.0",
|
61
|
+
"-//ietf//dtd html strict level 3//en",
|
62
|
+
"-//ietf//dtd html strict level 3//en//3.0",
|
63
|
+
"-//ietf//dtd html strict//en",
|
64
|
+
"-//ietf//dtd html strict//en//2.0",
|
65
|
+
"-//ietf//dtd html strict//en//3.0",
|
66
|
+
"-//ietf//dtd html//en",
|
67
|
+
"-//ietf//dtd html//en//2.0",
|
68
|
+
"-//ietf//dtd html//en//3.0",
|
69
|
+
"-//metrius//dtd metrius presentational//en",
|
70
|
+
"-//microsoft//dtd internet explorer 2.0 html strict//en",
|
71
|
+
"-//microsoft//dtd internet explorer 2.0 html//en",
|
72
|
+
"-//microsoft//dtd internet explorer 2.0 tables//en",
|
73
|
+
"-//microsoft//dtd internet explorer 3.0 html strict//en",
|
74
|
+
"-//microsoft//dtd internet explorer 3.0 html//en",
|
75
|
+
"-//microsoft//dtd internet explorer 3.0 tables//en",
|
76
|
+
"-//netscape comm. corp.//dtd html//en",
|
77
|
+
"-//netscape comm. corp.//dtd strict html//en",
|
78
|
+
"-//o'reilly and associates//dtd html 2.0//en",
|
79
|
+
"-//o'reilly and associates//dtd html extended 1.0//en",
|
80
|
+
"-//spyglass//dtd html 2.0 extended//en",
|
81
|
+
"-//sq//dtd html 2.0 hotmetal + extensions//en",
|
82
|
+
"-//sun microsystems corp.//dtd hotjava html//en",
|
83
|
+
"-//sun microsystems corp.//dtd hotjava strict html//en",
|
84
|
+
"-//w3c//dtd html 3 1995-03-24//en",
|
85
|
+
"-//w3c//dtd html 3.2 draft//en",
|
86
|
+
"-//w3c//dtd html 3.2 final//en",
|
87
|
+
"-//w3c//dtd html 3.2//en",
|
88
|
+
"-//w3c//dtd html 3.2s draft//en",
|
89
|
+
"-//w3c//dtd html 4.0 frameset//en",
|
90
|
+
"-//w3c//dtd html 4.0 transitional//en",
|
91
|
+
"-//w3c//dtd html experimental 19960712//en",
|
92
|
+
"-//w3c//dtd html experimental 970421//en",
|
93
|
+
"-//w3c//dtd w3 html//en",
|
94
|
+
"-//w3o//dtd w3 html 3.0//en",
|
95
|
+
"-//w3o//dtd w3 html 3.0//en//",
|
96
|
+
"-//w3o//dtd w3 html strict 3.0//en//",
|
97
|
+
"-//webtechs//dtd mozilla html 2.0//en",
|
98
|
+
"-//webtechs//dtd mozilla html//en",
|
99
|
+
"-/w3c/dtd html 4.0 transitional/en",
|
100
|
+
"html"].include?(publicId) or
|
101
|
+
(systemId == nil and
|
102
|
+
["-//w3c//dtd html 4.01 frameset//EN",
|
103
|
+
"-//w3c//dtd html 4.01 transitional//EN"].include?(publicId)) or
|
104
|
+
(systemId == "http://www.ibm.com/data/dtd/v11/ibmxhtml1-transitional.dtd")
|
105
|
+
#XXX quirks mode
|
106
|
+
end
|
107
|
+
end
|
108
|
+
|
109
|
+
@parser.phase = @parser.phases[:rootElement]
|
110
|
+
end
|
111
|
+
|
112
|
+
def processSpaceCharacters(data)
|
113
|
+
end
|
114
|
+
|
115
|
+
def processCharacters(data)
|
116
|
+
parse_error(_('Unexpected non-space characters. Expected DOCTYPE.'))
|
117
|
+
@parser.phase = @parser.phases[:rootElement]
|
118
|
+
@parser.phase.processCharacters(data)
|
119
|
+
end
|
120
|
+
|
121
|
+
def processStartTag(name, attributes)
|
122
|
+
parse_error(_("Unexpected start tag (#{name}). Expected DOCTYPE."))
|
123
|
+
@parser.phase = @parser.phases[:rootElement]
|
124
|
+
@parser.phase.processStartTag(name, attributes)
|
125
|
+
end
|
126
|
+
|
127
|
+
def processEndTag(name)
|
128
|
+
parse_error(_("Unexpected end tag (#{name}). Expected DOCTYPE."))
|
129
|
+
@parser.phase = @parser.phases[:rootElement]
|
130
|
+
@parser.phase.processEndTag(name)
|
131
|
+
end
|
132
|
+
|
133
|
+
end
|
134
|
+
end
|
@@ -0,0 +1,158 @@
|
|
1
|
+
module HTML5
|
2
|
+
# Base class for helper objects that implement each phase of processing.
|
3
|
+
#
|
4
|
+
# Handler methods should be in the following order (they can be omitted):
|
5
|
+
#
|
6
|
+
# * EOF
|
7
|
+
# * Comment
|
8
|
+
# * Doctype
|
9
|
+
# * SpaceCharacters
|
10
|
+
# * Characters
|
11
|
+
# * StartTag
|
12
|
+
# - startTag* methods
|
13
|
+
# * EndTag
|
14
|
+
# - endTag* methods
|
15
|
+
#
|
16
|
+
class Phase
|
17
|
+
|
18
|
+
extend Forwardable
|
19
|
+
def_delegators :@parser, :parse_error
|
20
|
+
|
21
|
+
# The following example call:
|
22
|
+
#
|
23
|
+
# tag_handlers('startTag', 'html', %w( base link meta ), %w( li dt dd ) => 'ListItem')
|
24
|
+
#
|
25
|
+
# ...would return a hash equal to this:
|
26
|
+
#
|
27
|
+
# { 'html' => 'startTagHtml',
|
28
|
+
# 'base' => 'startTagBaseLinkMeta',
|
29
|
+
# 'link' => 'startTagBaseLinkMeta',
|
30
|
+
# 'meta' => 'startTagBaseLinkMeta',
|
31
|
+
# 'li' => 'startTagListItem',
|
32
|
+
# 'dt' => 'startTagListItem',
|
33
|
+
# 'dd' => 'startTagListItem' }
|
34
|
+
#
|
35
|
+
def self.tag_handlers(prefix, *tags)
|
36
|
+
mapping = {}
|
37
|
+
if tags.last.is_a?(Hash)
|
38
|
+
tags.pop.each do |names, handler_method_suffix|
|
39
|
+
handler_method = prefix + handler_method_suffix
|
40
|
+
Array(names).each {|name| mapping[name] = handler_method }
|
41
|
+
end
|
42
|
+
end
|
43
|
+
tags.each do |names|
|
44
|
+
names = Array(names)
|
45
|
+
handler_method = prefix + names.map {|name| name.capitalize }.join
|
46
|
+
names.each {|name| mapping[name] = handler_method }
|
47
|
+
end
|
48
|
+
mapping
|
49
|
+
end
|
50
|
+
|
51
|
+
def self.start_tag_handlers
|
52
|
+
@start_tag_handlers ||= Hash.new('startTagOther')
|
53
|
+
end
|
54
|
+
|
55
|
+
# Declare what start tags this Phase handles. Can be called more than once.
|
56
|
+
#
|
57
|
+
# Example usage:
|
58
|
+
#
|
59
|
+
# handle_start 'html'
|
60
|
+
# # html start tags will be handled by a method named 'startTagHtml'
|
61
|
+
#
|
62
|
+
# handle_start %( base link meta )
|
63
|
+
# # base, link and meta start tags will be handled by a method named 'startTagBaseLinkMeta'
|
64
|
+
#
|
65
|
+
# handle_start %( li dt dd ) => 'ListItem'
|
66
|
+
# # li, dt, and dd start tags will be handled by a method named 'startTagListItem'
|
67
|
+
#
|
68
|
+
def self.handle_start(*tags)
|
69
|
+
start_tag_handlers.update tag_handlers('startTag', *tags)
|
70
|
+
end
|
71
|
+
|
72
|
+
def self.end_tag_handlers
|
73
|
+
@end_tag_handlers ||= Hash.new('endTagOther')
|
74
|
+
end
|
75
|
+
|
76
|
+
# Declare what end tags this Phase handles. Behaves like handle_start.
|
77
|
+
#
|
78
|
+
def self.handle_end(*tags)
|
79
|
+
end_tag_handlers.update tag_handlers('endTag', *tags)
|
80
|
+
end
|
81
|
+
|
82
|
+
def initialize(parser, tree)
|
83
|
+
@parser, @tree = parser, tree
|
84
|
+
end
|
85
|
+
|
86
|
+
def process_eof
|
87
|
+
@tree.generateImpliedEndTags
|
88
|
+
|
89
|
+
if @tree.open_elements.length > 2
|
90
|
+
parse_error(_('Unexpected end of file. Missing closing tags.'))
|
91
|
+
elsif @tree.open_elements.length == 2 and @tree.open_elements[1].name != 'body'
|
92
|
+
# This happens for framesets or something?
|
93
|
+
parse_error(_("Unexpected end of file. Expected end tag (#{@tree.open_elements[1].name}) first."))
|
94
|
+
elsif @parser.inner_html and @tree.open_elements.length > 1
|
95
|
+
# XXX This is not what the specification says. Not sure what to do here.
|
96
|
+
parse_error(_('XXX inner_html EOF'))
|
97
|
+
end
|
98
|
+
# Betting ends.
|
99
|
+
end
|
100
|
+
|
101
|
+
def processComment(data)
|
102
|
+
# For most phases the following is correct. Where it's not it will be
|
103
|
+
# overridden.
|
104
|
+
@tree.insert_comment(data, @tree.open_elements.last)
|
105
|
+
end
|
106
|
+
|
107
|
+
def processDoctype(name, publicId, systemId, correct)
|
108
|
+
parse_error(_('Unexpected DOCTYPE. Ignored.'))
|
109
|
+
end
|
110
|
+
|
111
|
+
def processSpaceCharacters(data)
|
112
|
+
@tree.insertText(data)
|
113
|
+
end
|
114
|
+
|
115
|
+
def processStartTag(name, attributes)
|
116
|
+
send self.class.start_tag_handlers[name], name, attributes
|
117
|
+
end
|
118
|
+
|
119
|
+
def startTagHtml(name, attributes)
|
120
|
+
if @parser.first_start_tag == false and name == 'html'
|
121
|
+
parse_error(_('html needs to be the first start tag.'))
|
122
|
+
end
|
123
|
+
# XXX Need a check here to see if the first start tag token emitted is
|
124
|
+
# this token... If it's not, invoke parse_error.
|
125
|
+
attributes.each do |attr, value|
|
126
|
+
unless @tree.open_elements.first.attributes.has_key?(attr)
|
127
|
+
@tree.open_elements.first.attributes[attr] = value
|
128
|
+
end
|
129
|
+
end
|
130
|
+
@parser.first_start_tag = false
|
131
|
+
end
|
132
|
+
|
133
|
+
def processEndTag(name)
|
134
|
+
send self.class.end_tag_handlers[name], name
|
135
|
+
end
|
136
|
+
|
137
|
+
def _(string)
|
138
|
+
string
|
139
|
+
end
|
140
|
+
|
141
|
+
def assert(value)
|
142
|
+
throw AssertionError.new unless value
|
143
|
+
end
|
144
|
+
|
145
|
+
def in_scope?(*args)
|
146
|
+
@tree.elementInScope(*args)
|
147
|
+
end
|
148
|
+
|
149
|
+
def remove_open_elements_until(name=nil)
|
150
|
+
finished = false
|
151
|
+
until finished
|
152
|
+
element = @tree.open_elements.pop
|
153
|
+
finished = name.nil? ? yield(element) : element.name == name
|
154
|
+
end
|
155
|
+
return element
|
156
|
+
end
|
157
|
+
end
|
158
|
+
end
|
@@ -0,0 +1,42 @@
|
|
1
|
+
require 'html5/html5parser/phase'
|
2
|
+
|
3
|
+
module HTML5
|
4
|
+
class RootElementPhase < Phase
|
5
|
+
|
6
|
+
def process_eof
|
7
|
+
insert_html_element
|
8
|
+
@parser.phase.process_eof
|
9
|
+
end
|
10
|
+
|
11
|
+
def processComment(data)
|
12
|
+
@tree.insert_comment(data, @tree.document)
|
13
|
+
end
|
14
|
+
|
15
|
+
def processSpaceCharacters(data)
|
16
|
+
end
|
17
|
+
|
18
|
+
def processCharacters(data)
|
19
|
+
insert_html_element
|
20
|
+
@parser.phase.processCharacters(data)
|
21
|
+
end
|
22
|
+
|
23
|
+
def processStartTag(name, attributes)
|
24
|
+
@parser.first_start_tag = true if name == 'html'
|
25
|
+
insert_html_element
|
26
|
+
@parser.phase.processStartTag(name, attributes)
|
27
|
+
end
|
28
|
+
|
29
|
+
def processEndTag(name)
|
30
|
+
insert_html_element
|
31
|
+
@parser.phase.processEndTag(name)
|
32
|
+
end
|
33
|
+
|
34
|
+
def insert_html_element
|
35
|
+
element = @tree.createElement('html', {})
|
36
|
+
@tree.open_elements.push(element)
|
37
|
+
@tree.document.appendChild(element)
|
38
|
+
@parser.phase = @parser.phases[:beforeHead]
|
39
|
+
end
|
40
|
+
|
41
|
+
end
|
42
|
+
end
|
@@ -0,0 +1,35 @@
|
|
1
|
+
require 'html5/html5parser/phase'
|
2
|
+
|
3
|
+
module HTML5
|
4
|
+
class TrailingEndPhase < Phase
|
5
|
+
|
6
|
+
def process_eof
|
7
|
+
end
|
8
|
+
|
9
|
+
def processComment(data)
|
10
|
+
@tree.insert_comment(data, @tree.document)
|
11
|
+
end
|
12
|
+
|
13
|
+
def processSpaceCharacters(data)
|
14
|
+
@parser.last_phase.processSpaceCharacters(data)
|
15
|
+
end
|
16
|
+
|
17
|
+
def processCharacters(data)
|
18
|
+
parse_error(_('Unexpected non-space characters. Expected end of file.'))
|
19
|
+
@parser.phase = @parser.last_phase
|
20
|
+
@parser.phase.processCharacters(data)
|
21
|
+
end
|
22
|
+
|
23
|
+
def processStartTag(name, attributes)
|
24
|
+
parse_error(_('Unexpected start tag (#{name}). Expected end of file.'))
|
25
|
+
@parser.phase = @parser.last_phase
|
26
|
+
@parser.phase.processStartTag(name, attributes)
|
27
|
+
end
|
28
|
+
|
29
|
+
def processEndTag(name)
|
30
|
+
parse_error(_('Unexpected end tag (#{name}). Expected end of file.'))
|
31
|
+
@parser.phase = @parser.last_phase
|
32
|
+
@parser.phase.processEndTag(name)
|
33
|
+
end
|
34
|
+
end
|
35
|
+
end
|
@@ -0,0 +1,248 @@
|
|
1
|
+
require 'html5/constants'
|
2
|
+
require 'html5/tokenizer'
|
3
|
+
require 'html5/treebuilders/rexml'
|
4
|
+
|
5
|
+
Dir.glob(File.join(File.dirname(__FILE__), 'html5parser', '*_phase.rb')).each do |path|
|
6
|
+
require 'html5/html5parser/' + File.basename(path)
|
7
|
+
end
|
8
|
+
|
9
|
+
module HTML5
|
10
|
+
|
11
|
+
# Error in parsed document
|
12
|
+
class ParseError < Exception; end
|
13
|
+
class AssertionError < Exception; end
|
14
|
+
|
15
|
+
# HTML parser. Generates a tree structure from a stream of (possibly malformed) HTML
|
16
|
+
#
|
17
|
+
class HTMLParser
|
18
|
+
|
19
|
+
attr_accessor :phase, :first_start_tag, :inner_html, :last_phase, :insert_from_table
|
20
|
+
|
21
|
+
attr_reader :phases, :tokenizer, :tree, :errors
|
22
|
+
|
23
|
+
def self.parse(stream, options = {})
|
24
|
+
encoding = options.delete(:encoding)
|
25
|
+
new(options).parse(stream,encoding)
|
26
|
+
end
|
27
|
+
|
28
|
+
def self.parse_fragment(stream, options = {})
|
29
|
+
container = options.delete(:container) || 'div'
|
30
|
+
encoding = options.delete(:encoding)
|
31
|
+
new(options).parse_fragment(stream, container, encoding)
|
32
|
+
end
|
33
|
+
|
34
|
+
@@phases = %w( initial rootElement beforeHead inHead afterHead inBody inTable inCaption
|
35
|
+
inColumnGroup inTableBody inRow inCell inSelect afterBody inFrameset afterFrameset trailingEnd )
|
36
|
+
|
37
|
+
# :strict - raise an exception when a parse error is encountered
|
38
|
+
# :tree - a treebuilder class controlling the type of tree that will be
|
39
|
+
# returned. Built in treebuilders can be accessed through
|
40
|
+
# HTML5::TreeBuilders[treeType]
|
41
|
+
def initialize(options = {})
|
42
|
+
@strict = false
|
43
|
+
@errors = []
|
44
|
+
|
45
|
+
@tokenizer = HTMLTokenizer
|
46
|
+
@tree = TreeBuilders::REXML::TreeBuilder
|
47
|
+
|
48
|
+
options.each {|name, value| instance_variable_set("@#{name}", value) }
|
49
|
+
@lowercase_attr_name = nil unless instance_variable_defined?(:@lowercase_attr_name)
|
50
|
+
@lowercase_element_name = nil unless instance_variable_defined?(:@lowercase_element_name)
|
51
|
+
|
52
|
+
@tree = @tree.new
|
53
|
+
|
54
|
+
@phases = @@phases.inject({}) do |phases, phase_name|
|
55
|
+
phase_class_name = phase_name.sub(/(.)/) { $1.upcase } + 'Phase'
|
56
|
+
phases[phase_name.to_sym] = HTML5.const_get(phase_class_name).new(self, @tree)
|
57
|
+
phases
|
58
|
+
end
|
59
|
+
end
|
60
|
+
|
61
|
+
def _parse(stream, inner_html, encoding, container = 'div')
|
62
|
+
@tree.reset
|
63
|
+
@first_start_tag = false
|
64
|
+
@errors = []
|
65
|
+
|
66
|
+
@tokenizer = @tokenizer.class unless Class === @tokenizer
|
67
|
+
@tokenizer = @tokenizer.new(stream, :encoding => encoding,
|
68
|
+
:parseMeta => !inner_html, :lowercase_attr_name => @lowercase_attr_name, :lowercase_element_name => @lowercase_element_name)
|
69
|
+
|
70
|
+
if inner_html
|
71
|
+
case @inner_html = container.downcase
|
72
|
+
when 'title', 'textarea'
|
73
|
+
@tokenizer.content_model_flag = :RCDATA
|
74
|
+
when 'style', 'script', 'xmp', 'iframe', 'noembed', 'noframes', 'noscript'
|
75
|
+
@tokenizer.content_model_flag = :CDATA
|
76
|
+
when 'plaintext'
|
77
|
+
@tokenizer.content_model_flag = :PLAINTEXT
|
78
|
+
else
|
79
|
+
# content_model_flag already is PCDATA
|
80
|
+
#@tokenizer.content_model_flag = :PCDATA
|
81
|
+
end
|
82
|
+
|
83
|
+
@phase = @phases[:rootElement]
|
84
|
+
@phase.insert_html_element
|
85
|
+
reset_insertion_mode
|
86
|
+
else
|
87
|
+
@inner_html = false
|
88
|
+
@phase = @phases[:initial]
|
89
|
+
end
|
90
|
+
|
91
|
+
# We only seem to have InBodyPhase testcases where the following is
|
92
|
+
# relevant ... need others too
|
93
|
+
@last_phase = nil
|
94
|
+
|
95
|
+
# XXX This is temporary for the moment so there isn't any other
|
96
|
+
# changes needed for the parser to work with the iterable tokenizer
|
97
|
+
@tokenizer.each do |token|
|
98
|
+
token = normalize_token(token)
|
99
|
+
|
100
|
+
method = 'process%s' % token[:type]
|
101
|
+
|
102
|
+
case token[:type]
|
103
|
+
when :Characters, :SpaceCharacters, :Comment
|
104
|
+
@phase.send method, token[:data]
|
105
|
+
when :StartTag
|
106
|
+
@phase.send method, token[:name], token[:data]
|
107
|
+
when :EndTag
|
108
|
+
@phase.send method, token[:name]
|
109
|
+
when :Doctype
|
110
|
+
@phase.send method, token[:name], token[:publicId],
|
111
|
+
token[:systemId], token[:correct]
|
112
|
+
else
|
113
|
+
parse_error(token[:data])
|
114
|
+
end
|
115
|
+
end
|
116
|
+
|
117
|
+
# When the loop finishes it's EOF
|
118
|
+
@phase.process_eof
|
119
|
+
end
|
120
|
+
|
121
|
+
# Parse a HTML document into a well-formed tree
|
122
|
+
#
|
123
|
+
# stream - a filelike object or string containing the HTML to be parsed
|
124
|
+
#
|
125
|
+
# The optional encoding parameter must be a string that indicates
|
126
|
+
# the encoding. If specified, that encoding will be used,
|
127
|
+
# regardless of any BOM or later declaration (such as in a meta
|
128
|
+
# element)
|
129
|
+
def parse(stream, encoding=nil)
|
130
|
+
_parse(stream, false, encoding)
|
131
|
+
@tree.get_document
|
132
|
+
end
|
133
|
+
|
134
|
+
# Parse a HTML fragment into a well-formed tree fragment
|
135
|
+
|
136
|
+
# container - name of the element we're setting the inner_html property
|
137
|
+
# if set to nil, default to 'div'
|
138
|
+
#
|
139
|
+
# stream - a filelike object or string containing the HTML to be parsed
|
140
|
+
#
|
141
|
+
# The optional encoding parameter must be a string that indicates
|
142
|
+
# the encoding. If specified, that encoding will be used,
|
143
|
+
# regardless of any BOM or later declaration (such as in a meta
|
144
|
+
# element)
|
145
|
+
def parse_fragment(stream, container='div', encoding=nil)
|
146
|
+
_parse(stream, true, encoding, container)
|
147
|
+
@tree.get_fragment
|
148
|
+
end
|
149
|
+
|
150
|
+
def parse_error(data = 'XXX ERROR MESSAGE NEEDED')
|
151
|
+
# XXX The idea is to make data mandatory.
|
152
|
+
@errors.push([@tokenizer.stream.position, data])
|
153
|
+
raise ParseError if @strict
|
154
|
+
end
|
155
|
+
|
156
|
+
# HTML5 specific normalizations to the token stream
|
157
|
+
def normalize_token(token)
|
158
|
+
|
159
|
+
if token[:type] == :EmptyTag
|
160
|
+
# When a solidus (/) is encountered within a tag name what happens
|
161
|
+
# depends on whether the current tag name matches that of a void
|
162
|
+
# element. If it matches a void element atheists did the wrong
|
163
|
+
# thing and if it doesn't it's wrong for everyone.
|
164
|
+
|
165
|
+
unless VOID_ELEMENTS.include?(token[:name])
|
166
|
+
parse_error(_('Solidus (/) incorrectly placed in tag.'))
|
167
|
+
end
|
168
|
+
|
169
|
+
token[:type] = :StartTag
|
170
|
+
end
|
171
|
+
|
172
|
+
if token[:type] == :StartTag
|
173
|
+
token[:name] = token[:name].downcase
|
174
|
+
|
175
|
+
# We need to remove the duplicate attributes and convert attributes
|
176
|
+
# to a dict so that [["x", "y"], ["x", "z"]] becomes {"x": "y"}
|
177
|
+
|
178
|
+
unless token[:data].empty?
|
179
|
+
data = token[:data].reverse.map {|attr, value| [attr.downcase, value] }
|
180
|
+
token[:data] = Hash[*data.flatten]
|
181
|
+
end
|
182
|
+
|
183
|
+
elsif token[:type] == :EndTag
|
184
|
+
parse_error(_('End tag contains unexpected attributes.')) unless token[:data].empty?
|
185
|
+
token[:name] = token[:name].downcase
|
186
|
+
end
|
187
|
+
|
188
|
+
token
|
189
|
+
end
|
190
|
+
|
191
|
+
@@new_modes = {
|
192
|
+
'select' => :inSelect,
|
193
|
+
'td' => :inCell,
|
194
|
+
'th' => :inCell,
|
195
|
+
'tr' => :inRow,
|
196
|
+
'tbody' => :inTableBody,
|
197
|
+
'thead' => :inTableBody,
|
198
|
+
'tfoot' => :inTableBody,
|
199
|
+
'caption' => :inCaption,
|
200
|
+
'colgroup' => :inColumnGroup,
|
201
|
+
'table' => :inTable,
|
202
|
+
'head' => :inBody,
|
203
|
+
'body' => :inBody,
|
204
|
+
'frameset' => :inFrameset
|
205
|
+
}
|
206
|
+
|
207
|
+
def reset_insertion_mode
|
208
|
+
# The name of this method is mostly historical. (It's also used in the
|
209
|
+
# specification.)
|
210
|
+
last = false
|
211
|
+
|
212
|
+
@tree.open_elements.reverse.each do |node|
|
213
|
+
node_name = node.name
|
214
|
+
|
215
|
+
if node == @tree.open_elements.first
|
216
|
+
last = true
|
217
|
+
unless ['td', 'th'].include?(node_name)
|
218
|
+
# XXX
|
219
|
+
# assert @inner_html
|
220
|
+
node_name = @inner_html
|
221
|
+
end
|
222
|
+
end
|
223
|
+
|
224
|
+
# Check for conditions that should only happen in the inner_html
|
225
|
+
# case
|
226
|
+
if ['select', 'colgroup', 'head', 'frameset'].include?(node_name)
|
227
|
+
# XXX
|
228
|
+
# assert @inner_html
|
229
|
+
end
|
230
|
+
|
231
|
+
if @@new_modes.has_key?(node_name)
|
232
|
+
@phase = @phases[@@new_modes[node_name]]
|
233
|
+
elsif node_name == 'html'
|
234
|
+
@phase = @phases[@tree.head_pointer.nil?? :beforeHead : :afterHead]
|
235
|
+
elsif last
|
236
|
+
@phase = @phases[:inBody]
|
237
|
+
else
|
238
|
+
next
|
239
|
+
end
|
240
|
+
|
241
|
+
break
|
242
|
+
end
|
243
|
+
end
|
244
|
+
|
245
|
+
def _(string); string; end
|
246
|
+
end
|
247
|
+
|
248
|
+
end
|