class RDoc::Markup::AttributeManager

Manages changes of attributes in a block of text

Constants

NULL: The NUL character

Attributes

attributes[R]

The attributes enabled for this markup object.

exclusive_bitmap[R]

A bits of exclusive maps

html_tags[R]

This maps HTML tags to the corresponding attribute char

matching_word_pairs[R]

This maps delimiters that occur around words (such as bold or tt) where the start and end delimiters and the same. This lets us optimize the regexp

protectable[R]

A \ in front of a character that would normally be processed turns off processing. We do this by turning < into <#{PROTECT}

regexp_handlings[R]

And this maps _regexp handling_ sequences to a name. A regexp handling sequence is something like a WikiWord

word_pair_map[R]

And this is used when the delimiters aren’t the same. In this case the hash maps a pattern to the attribute character

Public Class Methods

new() click to toggle source

Creates a new attribute manager that understands bold, emphasized and teletype text.

# File rdoc/markup/attribute_manager.rb, line 69
def initialize
  @html_tags = {}
  @matching_word_pairs = {}
  @protectable = %w[<]
  @regexp_handlings = []
  @word_pair_map = {}
  @exclusive_bitmap = 0
  @attributes = RDoc::Markup::Attributes.new

  add_word_pair "*", "*", :BOLD, true
  add_word_pair "_", "_", :EM, true
  add_word_pair "+", "+", :TT, true

  add_html "em", :EM, true
  add_html "i",  :EM, true
  add_html "b",  :BOLD, true
  add_html "tt",   :TT, true
  add_html "code", :TT, true
end

Public Instance Methods

add_html(tag, name, exclusive = false) click to toggle source

Adds a markup class with name for words surrounded by HTML tag tag. To process emphasis tags:

am.add_html 'em', :EM

# File rdoc/markup/attribute_manager.rb, line 289
def add_html(tag, name, exclusive = false)
  bitmap = @attributes.bitmap_for name
  @html_tags[tag.downcase] = bitmap
  @exclusive_bitmap |= bitmap if exclusive
end

add_regexp_handling(pattern, name, exclusive = false) click to toggle source

Adds a regexp handling for pattern with name. A simple URL handler would be:

@am.add_regexp_handling(/((https?:)\S+\w)/, :HYPERLINK)

# File rdoc/markup/attribute_manager.rb, line 301
def add_regexp_handling pattern, name, exclusive = false
  bitmap = @attributes.bitmap_for(name)
  @regexp_handlings << [pattern, bitmap]
  @exclusive_bitmap |= bitmap if exclusive
end

add_word_pair(start, stop, name, exclusive = false) click to toggle source

Adds a markup class with name for words wrapped in the start and stop character. To make words wrapped with “*” bold:

am.add_word_pair '*', '*', :BOLD

# File rdoc/markup/attribute_manager.rb, line 264
def add_word_pair(start, stop, name, exclusive = false)
  raise ArgumentError, "Word flags may not start with '<'" if
    start[0,1] == '<'

  bitmap = @attributes.bitmap_for name

  if start == stop then
    @matching_word_pairs[start] = bitmap
  else
    pattern = /(#{Regexp.escape start})(\S+)(#{Regexp.escape stop})/
    @word_pair_map[pattern] = bitmap
  end

  @protectable << start[0,1]
  @protectable.uniq!

  @exclusive_bitmap |= bitmap if exclusive
end

attribute(turn_on, turn_off) click to toggle source

Return an attribute object with the given turn_on and turn_off bits set

# File rdoc/markup/attribute_manager.rb, line 92
def attribute(turn_on, turn_off)
  RDoc::Markup::AttrChanger.new turn_on, turn_off
end

change_attribute(current, new) click to toggle source

Changes the current attribute from current to new

# File rdoc/markup/attribute_manager.rb, line 99
def change_attribute current, new
  diff = current ^ new
  attribute(new & diff, current & diff)
end

changed_attribute_by_name(current_set, new_set) click to toggle source

Used by the tests to change attributes by name from current_set to new_set

# File rdoc/markup/attribute_manager.rb, line 108
def changed_attribute_by_name current_set, new_set
  current = new = 0
  current_set.each do |name|
    current |= @attributes.bitmap_for(name)
  end

  new_set.each do |name|
    new |= @attributes.bitmap_for(name)
  end

  change_attribute(current, new)
end

convert_attrs(str, attrs, exclusive = false) click to toggle source

Map attributes like textto the sequence 001002<char>001003<char>, where <char> is a per-attribute specific character

# File rdoc/markup/attribute_manager.rb, line 142
def convert_attrs(str, attrs, exclusive = false)
  convert_attrs_matching_word_pairs(str, attrs, exclusive)
  convert_attrs_word_pair_map(str, attrs, exclusive)
end

convert_attrs_matching_word_pairs(str, attrs, exclusive) click to toggle source

# File rdoc/markup/attribute_manager.rb, line 147
def convert_attrs_matching_word_pairs(str, attrs, exclusive)
  # first do matching ones
  tags = @matching_word_pairs.select { |start, bitmap|
    if exclusive && exclusive?(bitmap)
      true
    elsif !exclusive && !exclusive?(bitmap)
      true
    else
      false
    end
  }.keys
  return if tags.empty?
  all_tags = @matching_word_pairs.keys

  re = /(^|\W|[#{all_tags.join("")}])([#{tags.join("")}])(\2*[#\\]?[\w:.\/\[\]-]+?\S?)\2(?!\2)([#{all_tags.join("")}]|\W|$)/

  1 while str.gsub!(re) { |orig|
    attr = @matching_word_pairs[$2]
    attr_updated = attrs.set_attrs($`.length + $1.length + $2.length, $3.length, attr)
    if attr_updated
      $1 + NULL * $2.length + $3 + NULL * $2.length + $4
    else
      $1 + NON_PRINTING_START + $2 + NON_PRINTING_END + $3 + NON_PRINTING_START + $2 + NON_PRINTING_END + $4
    end
  }
  str.delete!(NON_PRINTING_START + NON_PRINTING_END)
end

convert_attrs_word_pair_map(str, attrs, exclusive) click to toggle source

# File rdoc/markup/attribute_manager.rb, line 175
def convert_attrs_word_pair_map(str, attrs, exclusive)
  # then non-matching
  unless @word_pair_map.empty? then
    @word_pair_map.each do |regexp, attr|
      if !exclusive
        next if exclusive?(attr)
      else
        next if !exclusive?(attr)
      end
      1 while str.gsub!(regexp) { |orig|
        updated = attrs.set_attrs($`.length + $1.length, $2.length, attr)
        if updated
          NULL * $1.length + $2 + NULL * $3.length
        else
          orig
        end
      }
    end
  end
end

convert_html(str, attrs, exclusive = false) click to toggle source

Converts HTML tags to RDoc attributes

# File rdoc/markup/attribute_manager.rb, line 199
def convert_html(str, attrs, exclusive = false)
  tags = @html_tags.select { |start, bitmap|
    if exclusive && exclusive?(bitmap)
      true
    elsif !exclusive && !exclusive?(bitmap)
      true
    else
      false
    end
  }.keys.join '|'

  1 while str.gsub!(/<(#{tags})>(.*?)<\/\1>/i) { |orig|
    attr = @html_tags[$1.downcase]
    html_length = $1.length + 2
    seq = NULL * html_length
    attrs.set_attrs($`.length + html_length, $2.length, attr)
    seq + $2 + seq + NULL
  }
end

convert_regexp_handlings(str, attrs, exclusive = false) click to toggle source

Converts regexp handling sequences to RDoc attributes

# File rdoc/markup/attribute_manager.rb, line 222
def convert_regexp_handlings str, attrs, exclusive = false
  @regexp_handlings.each do |regexp, attribute|
    if exclusive
      next if !exclusive?(attribute)
    else
      next if exclusive?(attribute)
    end
    str.scan(regexp) do
      capture = $~.size == 1 ? 0 : 1

      s, e = $~.offset capture

      attrs.set_attrs s, e - s, attribute | @attributes.regexp_handling
    end
  end
end

copy_string(start_pos, end_pos) click to toggle source

Copies start_pos to end_pos from the current string

# File rdoc/markup/attribute_manager.rb, line 124
def copy_string(start_pos, end_pos)
  res = @str[start_pos...end_pos]
  res.gsub!(/\000/, '')
  res
end

display_attributes() click to toggle source

Debug method that prints a string along with its attributes

# File rdoc/markup/attribute_manager.rb, line 332
def display_attributes
  puts
  puts @str.tr(NULL, "!")
  bit = 1
  16.times do |bno|
    line = ""
    @str.length.times do |i|
      if (@attrs[i] & bit) == 0
        line << " "
      else
        if bno.zero?
          line << "S"
        else
          line << ("%d" % (bno+1))
        end
      end
    end
    puts(line) unless line =~ /^ *$/
    bit <<= 1
  end
end

exclusive?(attr) click to toggle source

# File rdoc/markup/attribute_manager.rb, line 130
def exclusive?(attr)
  (attr & @exclusive_bitmap) != 0
end

flow(str) click to toggle source

Processes str converting attributes, HTML and regexp handlings

# File rdoc/markup/attribute_manager.rb, line 310
def flow str
  @str = str.dup

  mask_protected_sequences

  @attrs = RDoc::Markup::AttrSpan.new @str.length, @exclusive_bitmap

  convert_attrs            @str, @attrs, true
  convert_html             @str, @attrs, true
  convert_regexp_handlings @str, @attrs, true
  convert_attrs            @str, @attrs
  convert_html             @str, @attrs
  convert_regexp_handlings @str, @attrs

  unmask_protected_sequences

  split_into_flow
end

mask_protected_sequences() click to toggle source

Escapes regexp handling sequences of text to prevent conversion to RDoc

# File rdoc/markup/attribute_manager.rb, line 242
def mask_protected_sequences
  # protect __send__, __FILE__, etc.
  @str.gsub!(/__([a-z]+)__/i,
    "_#{PROTECT_ATTR}_#{PROTECT_ATTR}\\1_#{PROTECT_ATTR}_#{PROTECT_ATTR}")
  @str.gsub!(/(\A|[^\\])\\([#{Regexp.escape @protectable.join}])/m,
             "\\1\\2#{PROTECT_ATTR}")
  @str.gsub!(/\\(\\[#{Regexp.escape @protectable.join}])/m, "\\1")
end

split_into_flow() click to toggle source

Splits the string into chunks by attribute change

# File rdoc/markup/attribute_manager.rb, line 357
def split_into_flow
  res = []
  current_attr = 0

  str_len = @str.length

  # skip leading invisible text
  i = 0
  i += 1 while i < str_len and @str[i].chr == "\0"
  start_pos = i

  # then scan the string, chunking it on attribute changes
  while i < str_len
    new_attr = @attrs[i]
    if new_attr != current_attr
      if i > start_pos
        res << copy_string(start_pos, i)
        start_pos = i
      end

      res << change_attribute(current_attr, new_attr)
      current_attr = new_attr

      if (current_attr & @attributes.regexp_handling) != 0 then
        i += 1 while
          i < str_len and (@attrs[i] & @attributes.regexp_handling) != 0

        res << RDoc::Markup::RegexpHandling.new(current_attr,
                                                copy_string(start_pos, i))
        start_pos = i
        next
      end
    end

    # move on, skipping any invisible characters
    begin
      i += 1
    end while i < str_len and @str[i].chr == "\0"
  end

  # tidy up trailing text
  if start_pos < str_len
    res << copy_string(start_pos, str_len)
  end

  # and reset to all attributes off
  res << change_attribute(current_attr, 0) if current_attr != 0

  res
end

unmask_protected_sequences() click to toggle source

Unescapes regexp handling sequences of text

# File rdoc/markup/attribute_manager.rb, line 254
def unmask_protected_sequences
  @str.gsub!(/(.)#{PROTECT_ATTR}/, "\\1\000")
end