Class: DotStrings::Parser

Inherits:
Object
  • Object
show all
Defined in:
lib/dotstrings/parser.rb

Overview

Parser for .strings files.

You can use this class directly, but it is recommended to use File.parse and File.parse_file wrappers instead.

Constant Summary collapse

TOK_SLASH =

Special tokens

'/'
TOK_ASTERISK = 
'*'
TOK_QUOTE = 
'"'
TOK_SINGLE_QUOTE = 
"'"
TOK_BACKSLASH = 
'\\'
TOK_EQUALS = 
'='
TOK_SEMICOLON = 
';'
TOK_NEW_LINE = 
"\n"
TOK_N = 
'n'
TOK_R = 
'r'
TOK_T = 
't'
TOK_CAP_U = 
'U'
TOK_ZERO = 
'0'
TOK_HEX_DIGIT = 
/[0-9a-fA-F]/.freeze
STATE_START =

States

0
STATE_COMMENT_START = 
1
STATE_COMMENT = 
2
STATE_MULTILINE_COMMENT = 
3
STATE_COMMENT_END = 
4
STATE_KEY = 
5
STATE_KEY_END = 
6
STATE_VALUE_SEPARATOR = 
7
STATE_VALUE = 
8
STATE_VALUE_END = 
9
STATE_UNICODE = 
10
STATE_UNICODE_SURROGATE = 
11
STATE_UNICODE_SURROGATE_U = 
12

Instance Method Summary collapse

Constructor Details

#initialize(strict: true) ⇒ Parser

Returns a new Parser instance.

Parameters:

  • strict (Boolean) (defaults to: true) —

    Whether to parse in strict mode.



49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
 
# File 'lib/dotstrings/parser.rb', line 49

def initialize(strict: true)
  @strict = strict

  @state = STATE_START
  @temp_state = nil

  @buffer = []
  @unicode_buffer = []
  @high_surrogate = nil

  @escaping = false

  @current_comment = nil
  @current_key = nil
  @current_value = nil

  @item_block = nil

  @offset = 0
  @line = 1
  @column = 1
end

Instance Method Details

#<<(data) ⇒ Object

Feeds data to the parser.



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
 
# File 'lib/dotstrings/parser.rb', line 82

def <<(data)
  data.each_char do |ch|
    case @state
    when STATE_START
      start_value(ch)
    when STATE_COMMENT_START
      case ch
      when TOK_SLASH
        @state = STATE_COMMENT
      when TOK_ASTERISK
        @state = STATE_MULTILINE_COMMENT
      else
        raise_error("Unexpected character '#{ch}'")
      end
    when STATE_COMMENT
      if ch == TOK_NEW_LINE
        @state = STATE_COMMENT_END
        @current_comment = @buffer.join.strip
        @buffer.clear
      else
        @buffer << ch
      end
    when STATE_MULTILINE_COMMENT
      if ch == TOK_SLASH && @buffer.last == TOK_ASTERISK
        @state = STATE_COMMENT_END
        @current_comment = @buffer.slice(0, @buffer.length - 1).join.strip
        @buffer.clear
      else
        @buffer << ch
      end
    when STATE_COMMENT_END
      comment_end(ch)
    when STATE_KEY
      parse_string(ch) do |key|
        @current_key = key
        @state = STATE_KEY_END
      end
    when STATE_KEY_END
      if ch == TOK_EQUALS
        @state = STATE_VALUE_SEPARATOR
      else
        raise_error("Unexpected character '#{ch}', expecting '#{TOK_EQUALS}'") unless whitespace?(ch)
      end
    when STATE_VALUE_SEPARATOR
      if ch == TOK_QUOTE
        @state = STATE_VALUE
      else
        raise_error("Unexpected character '#{ch}'") unless whitespace?(ch)
      end
    when STATE_VALUE
      parse_string(ch) do |value|
        @current_value = value
        @state = STATE_VALUE_END

        @item_block&.call(Item.new(
          comment: @current_comment,
          key: @current_key,
          value: @current_value
        ))
      end
    when STATE_VALUE_END
      if ch == TOK_SEMICOLON
        @state = STATE_START
      else
        raise_error("Unexpected character '#{ch}', expecting '#{TOK_SEMICOLON}'") unless whitespace?(ch)
      end
    when STATE_UNICODE
      parse_unicode(ch) do |unicode_ch|
        @buffer << unicode_ch
        # Restore state
        @state = @temp_state
      end
    when STATE_UNICODE_SURROGATE
      if ch == TOK_BACKSLASH
        @state = STATE_UNICODE_SURROGATE_U
      else
        raise_error("Unexpected character '#{ch}', expecting another unicode codepoint")
      end
    when STATE_UNICODE_SURROGATE_U
      if ch == TOK_CAP_U
        @state = STATE_UNICODE
      else
        raise_error("Unexpected character '#{ch}', expecting '#{TOK_CAP_U}'")
      end
    end

    update_position(ch)
  end
end

#on_item(&block) ⇒ Object

Specifies a block to be called when a new item is parsed.



74
75
76
 
# File 'lib/dotstrings/parser.rb', line 74

def on_item(&block)
  @item_block = block
end