Class: MaRuKu::In::Markdown::SpanLevelParser::HTMLHelper

Inherits:
Object
  • Object
show all
Defined in:
lib/maruku/input/html_helper.rb

Overview

This class helps me read and sanitize HTML blocks

Constant Summary collapse

Tag = 
%r{^<(/)?(\w+)\s*([^>]*?)>}m
PartialTag = 
%r{^<.*}m
CData = 
%r{^\s*<!\[CDATA\[}m
CDataEnd = 
%r{\]\]>}m
EverythingElse = 
%r{^[^<]+}m
CommentStart = 
%r{^<!--}x
CommentEnd = 
%r{-->}
TO_SANITIZE = 
['img', 'hr', 'br']

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ HTMLHelper

Returns a new instance of HTMLHelper.



17
18
19
20
21
22
23
 
# File 'lib/maruku/input/html_helper.rb', line 17

def initialize
  @rest = ""
  @tag_stack = []
  @m = nil
  @already = ""
  self.state = :inside_element
end

Instance Attribute Details

#first_tag ⇒ Object (readonly)

Returns the value of attribute first_tag.



15
16
17
 
# File 'lib/maruku/input/html_helper.rb', line 15

def first_tag
  @first_tag
end

#rest ⇒ Object (readonly)

Returns the value of attribute rest.



15
16
17
 
# File 'lib/maruku/input/html_helper.rb', line 15

def rest
  @rest
end

#state ⇒ Object

:inside_element, :inside_tag, :inside_comment, :inside_cdata



25
26
27
 
# File 'lib/maruku/input/html_helper.rb', line 25

def state
  @state
end

Instance Method Details

#eat_this(line) ⇒ Object 



27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
 
# File 'lib/maruku/input/html_helper.rb', line 27

def eat_this(line)
  @rest = line + @rest
  things_read = 0
  until @rest.empty?
    case self.state
    when :inside_comment
      if @m = CommentEnd.match(@rest)
        debug_state 'Comment End'
        # Workaround for https://bugs.ruby-lang.org/issues/9277 and another bug in 1.9.2 where even a
        # single dash in a comment will cause REXML to error.
        @already << @m.pre_match.gsub(/-(?![^\-])/, '- ') << @m.to_s
        @rest = @m.post_match
        self.state = :inside_element
      else
        @already << @rest.gsub(/-(?![^\-])/, '- ') # Workaround for https://bugs.ruby-lang.org/issues/9277
        @rest = ""
        self.state = :inside_comment
      end
    when :inside_element
      if @m = CommentStart.match(@rest)
        debug_state 'Comment'
        things_read += 1
        @already << @m.pre_match << @m.to_s
        @rest = @m.post_match
        self.state = :inside_comment
      elsif @m = Tag.match(@rest)
        debug_state 'Tag'
        things_read += 1
        self.state = :inside_element
        handle_tag
      elsif @m = CData.match(@rest)
        debug_state 'CDATA'
        @already << @m.pre_match
        close_script_style if script_style?
        @already << @m.to_s
        @rest = @m.post_match
        self.state = :inside_cdata
      elsif @m = PartialTag.match(@rest)
        debug_state 'PartialTag'
        @already << @m.pre_match
        @rest = @m.post_match
        @partial_tag = @m.to_s
        self.state = :inside_tag
      elsif @m = EverythingElse.match(@rest)
        debug_state 'EverythingElse'
        @already << @m.pre_match << @m.to_s
        @rest = @m.post_match
        self.state = :inside_element
      else
        error "Malformed HTML: not complete: #{@rest.inspect}"
      end
    when :inside_tag
      if @m = /^[^>]*>/.match(@rest)
        @partial_tag << @m.to_s
        @rest = @partial_tag + @m.post_match
        @partial_tag = nil
        self.state = :inside_element
        if @m = Tag.match(@rest)
          things_read += 1
          handle_tag
        end
      else
        @partial_tag << @rest
        @rest = ""
        self.state = :inside_tag
      end
    when :inside_cdata
      if @m = CDataEnd.match(@rest)
        self.state = :inside_element
        @already << @m.pre_match << @m.to_s
        @rest = @m.post_match
        start_script_style if script_style?
      else
        @already << @rest
        @rest = ""
        self.state = :inside_cdata
      end
    else
      raise "Bug bug: state = #{self.state.inspect}"
    end

    break if is_finished? && things_read > 0
  end
end

#handle_tag ⇒ Object 



112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
 
# File 'lib/maruku/input/html_helper.rb', line 112

def handle_tag
  @already << @m.pre_match
  @rest = @m.post_match

  is_closing = !!@m[1]
  tag = @m[2]
  @first_tag ||= tag
  attributes = @m[3].to_s

  is_single = false
  if attributes[-1, 1] == '/'
    attributes = attributes[0, attributes.size - 1]
    is_single = true
  end

  if TO_SANITIZE.include? tag
    attributes.strip!
    if attributes.size > 0
      @already << '<%s %s />' % [tag, attributes]
    else
      @already << '<%s />' % [tag]
    end
  elsif is_closing
    if @tag_stack.empty?
      error "Malformed: closing tag #{tag.inspect} in empty list"
    elsif @tag_stack.last != tag
      error "Malformed: tag <#{tag}> closes <#{@tag_stack.last}>"
    end

    close_script_style if script_style?

    @already << @m.to_s
    @tag_stack.pop
  else
    @already << @m.to_s
    @tag_stack.push(tag) unless is_single

    start_script_style if script_style?
  end
end

#is_finished? ⇒ Boolean

Returns:

  • (Boolean) 


157
158
159
 
# File 'lib/maruku/input/html_helper.rb', line 157

def is_finished?
  self.state == :inside_element && @tag_stack.empty?
end

#stuff_you_read ⇒ Object 



153
154
155
 
# File 'lib/maruku/input/html_helper.rb', line 153

def stuff_you_read
  @already
end