Class: Parser::Source::Buffer

24
25
26
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
111
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
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242

# File 'lib/parser/source/buffer.rb', line 24

class Buffer
  attr_reader :name, :first_line

  ##
  # @api private
  #
  ENCODING_RE =
    /\#.*coding\s*[:=]\s*
      (
        # Special-case: there's a UTF8-MAC encoding.
        (utf8-mac)
      |
        # Chew the suffix; it's there for emacs compat.
        ([A-Za-z0-9_-]+?)(-unix|-dos|-mac)
      |
        ([A-Za-z0-9_-]+)
      )
    /x

  ##
  # Try to recognize encoding of `string` as Ruby would, i.e. by looking for
  # magic encoding comment or UTF-8 BOM. `string` can be in any encoding.
  #
  # @param [String]  string
  # @return [String|nil] encoding name, if recognized
  #
  def self.recognize_encoding(string)
    return if string.empty?

    # extract the first two lines in an efficient way
    string =~ /\A(.*)\n?(.*\n)?/
    first_line, second_line = $1, $2

    if first_line =~ /\A\xef\xbb\xbf/ # BOM
      return Encoding::UTF_8
    elsif first_line[0, 2] == '#!'
      encoding_line = second_line
    else
      encoding_line = first_line
    end

    if (result = ENCODING_RE.match(encoding_line))
      Encoding.find(result[2] || result[3] || result[5])
    else
      nil
    end
  end

  ##
  # Recognize encoding of `input` and process it so it could be lexed.
  #
  #  * If `input` does not contain BOM or magic encoding comment, it is
  #    kept in the original encoding.
  #  * If the detected encoding is binary, `input` is kept in binary.
  #  * Otherwise, `input` is re-encoded into UTF-8 and returned as a
  #    new string.
  #
  # This method mutates the encoding of `input`, but not its content.
  #
  # @param  [String] input
  # @return [String]
  # @raise  [EncodingError]
  #
  def self.reencode_string(input)
    original_encoding = input.encoding
    detected_encoding = recognize_encoding(input.force_encoding(Encoding::BINARY))

    if detected_encoding.nil?
      input.force_encoding(original_encoding)
    elsif detected_encoding == Encoding::BINARY
      input
    else
      input.
        force_encoding(detected_encoding).
        encode(Encoding::UTF_8)
    end
  end

  def initialize(name, first_line = 1)
    @name        = name
    @source      = nil
    @first_line  = first_line

    @lines       = nil
    @line_begins = nil
  end

  ##
  # Populate this buffer from correspondingly named file.
  #
  # @example
  #  Parser::Source::Buffer.new('foo/bar.rb').read
  #
  # @return [Buffer] self
  # @raise  [ArgumentError] if already populated
  #
  def read
    File.open(@name, 'rb') do |io|
      self.source = io.read
    end

    self
  end

  ##
  # Source code contained in this buffer.
  #
  # @return [String] source code
  # @raise  [RuntimeError] if buffer is not populated yet
  #
  def source
    if @source.nil?
      raise RuntimeError, 'Cannot extract source from uninitialized Source::Buffer'
    end

    @source
  end

  ##
  # Populate this buffer from a string with encoding autodetection.
  # `input` is mutated if not frozen.
  #
  # @param [String] input
  # @raise [ArgumentError] if already populated
  # @raise [EncodingError] if `input` includes invalid byte sequence for the encoding
  # @return [String]
  #
  def source=(input)
    if defined?(Encoding)
      input = input.dup if input.frozen?
      input = self.class.reencode_string(input)

      unless input.valid_encoding?
        raise EncodingError, "invalid byte sequence in #{input.encoding.name}"
      end
    end

    self.raw_source = input
  end

  ##
  # Populate this buffer from a string without encoding autodetection.
  #
  # @param [String] input
  # @raise [ArgumentError] if already populated
  # @return [String]
  #
  def raw_source=(input)
    if @source
      raise ArgumentError, 'Source::Buffer is immutable'
    end

    @source = input.gsub("\r\n", "\n").freeze
  end

  ##
  # Convert a character index into the source to a `[line, column]` tuple.
  #
  # @param  [Integer] position
  # @return [[Integer, Integer]] `[line, column]`
  #
  def decompose_position(position)
    line_no, line_begin = line_for(position)

    [ @first_line + line_no, position - line_begin ]
  end

  ##
  # Extract line `lineno` from source, taking `first_line` into account.
  #
  # @param  [Integer] lineno
  # @return [String]
  # @raise  [IndexError] if `lineno` is out of bounds
  #
  def source_line(lineno)
    unless @lines
      @lines = @source.lines.to_a
      @lines.each { |line| line.chomp!("\n") }

      # If a file ends with a newline, the EOF token will appear
      # to be one line further than the end of file.
      @lines << ""
    end

    @lines.fetch(lineno - @first_line).dup
  end

  private

  def line_begins
    unless @line_begins
      @line_begins, index = [ [ 0, 0 ] ], 1

      @source.each_char do |char|
        if char == "\n"
          @line_begins.unshift [ @line_begins.length, index ]
        end

        index += 1
      end
    end

    @line_begins
  end

  def line_for(position)
    if line_begins.respond_to? :bsearch
      # Fast O(log n) variant for Ruby >=2.0.
      line_begins.bsearch do |line, line_begin|
        line_begin <= position
      end
    else
      # Slower O(n) variant for Ruby <2.0.
      line_begins.find do |line, line_begin|
        line_begin <= position
      end
    end
  end
end