Class: GraphQL::Language::Lexer

Inherits:
Object
  • Object
show all
Defined in:
lib/graphql/language/lexer.rb

Defined Under Namespace

Modules: ByteFor, Punctuation

Constant Summary collapse

ESCAPES = 
/\\["\\\/bfnrt]/
ESCAPES_REPLACE = 
{
  '\\"' => '"',
  "\\\\" => "\\",
  "\\/" => '/',
  "\\b" => "\b",
  "\\f" => "\f",
  "\\n" => "\n",
  "\\r" => "\r",
  "\\t" => "\t",
}
UTF_8 = 
/\\u(?:([\dAa-f]{4})|\{([\da-f]{4,})\})(?:\\u([\dAa-f]{4}))?/i
VALID_STRING = 
/\A(?:[^\\]|#{ESCAPES}|#{UTF_8})*\z/o
ESCAPED = 
/(?:#{ESCAPES}|#{UTF_8})/o
IGNORE_REGEXP = 
%r{
  (?:
    [, \c\r\n\t]+ |
    \#.*$
  )*
}x
IDENTIFIER_REGEXP = 
/[_A-Za-z][_0-9A-Za-z]*/
INT_REGEXP = 
/-?(?:[0]|[1-9][0-9]*)/
FLOAT_DECIMAL_REGEXP = 
/[.][0-9]+/
FLOAT_EXP_REGEXP = 
/[eE][+-]?[0-9]+/
NUMERIC_REGEXP = 
/#{INT_REGEXP}(#{FLOAT_DECIMAL_REGEXP}#{FLOAT_EXP_REGEXP}|#{FLOAT_DECIMAL_REGEXP}|#{FLOAT_EXP_REGEXP})?/
KEYWORDS = 
[
  "on",
  "fragment",
  "true",
  "false",
  "null",
  "query",
  "mutation",
  "subscription",
  "schema",
  "scalar",
  "type",
  "extend",
  "implements",
  "interface",
  "union",
  "enum",
  "input",
  "directive",
  "repeatable"
].freeze
KEYWORD_REGEXP = 
/#{Regexp.union(KEYWORDS.sort)}\b/
KEYWORD_BY_TWO_BYTES = 
[
  :INTERFACE,
  :MUTATION,
  :EXTEND,
  :FALSE,
  :ENUM,
  :TRUE,
  :NULL,
  nil,
  nil,
  nil,
  nil,
  nil,
  nil,
  nil,
  :QUERY,
  nil,
  nil,
  :REPEATABLE,
  :IMPLEMENTS,
  :INPUT,
  :TYPE,
  :SCHEMA,
  nil,
  nil,
  nil,
  :DIRECTIVE,
  :UNION,
  nil,
  nil,
  :SCALAR,
  nil,
  :FRAGMENT
]
PUNCTUATION_NAME_FOR_BYTE =

A sparse array mapping the bytes for each punctuation to a symbol name for that punctuation

Punctuation.constants.each_with_object([]) { |name, arr|
  punct = Punctuation.const_get(name)
  arr[punct.ord] = name
}
QUOTE = 
'"'
UNICODE_DIGIT = 
/[0-9A-Za-z]/
FOUR_DIGIT_UNICODE = 
/#{UNICODE_DIGIT}{4}/
N_DIGIT_UNICODE = 
%r{#{Punctuation::LCURLY}#{UNICODE_DIGIT}{4,}#{Punctuation::RCURLY}}x
UNICODE_ESCAPE = 
%r{\\u(?:#{FOUR_DIGIT_UNICODE}|#{N_DIGIT_UNICODE})}
STRING_ESCAPE =

https://graphql.github.io/graphql-spec/June2018/#sec-String-Value

%r{[\\][\\/bfnrt]}
BLOCK_QUOTE = 
'"""'
ESCAPED_QUOTE = 
/\\"/
STRING_CHAR = 
/#{ESCAPED_QUOTE}|[^"\\\n\r]|#{UNICODE_ESCAPE}|#{STRING_ESCAPE}/
QUOTED_STRING_REGEXP = 
%r{#{QUOTE} (?:#{STRING_CHAR})* #{QUOTE}}x
BLOCK_STRING_REGEXP = 
%r{
  #{BLOCK_QUOTE}
  (?: [^"\\]               |  # Any characters that aren't a quote or slash
     (?<!") ["]{1,2} (?!") |  # Any quotes that don't have quotes next to them
     \\"{0,3}(?!")         |  # A slash followed by <= 3 quotes that aren't followed by a quote
     \\                    |  # A slash
     "{1,2}(?!")              # 1 or 2 " followed by something that isn't a quote
  )*
  (?:"")?
  #{BLOCK_QUOTE}
}xm
FIRST_BYTES =

Use this array to check, for a given byte that will start a token, what kind of token might it start?

Array.new(255)

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(graphql_str, filename: nil) ⇒ Lexer

Returns a new instance of Lexer.



6
7
8
9
10
11
12
13
14
 
# File 'lib/graphql/language/lexer.rb', line 6

def initialize(graphql_str, filename: nil)
  if !(graphql_str.encoding == Encoding::UTF_8 || graphql_str.ascii_only?)
    graphql_str = graphql_str.dup.force_encoding(Encoding::UTF_8)
  end
  @string = graphql_str
  @filename = filename
  @scanner = StringScanner.new(graphql_str)
  @pos = nil
end

Instance Attribute Details

#pos ⇒ Object (readonly)

Returns the value of attribute pos.



20
21
22
 
# File 'lib/graphql/language/lexer.rb', line 20

def pos
  @pos
end

Class Method Details

.replace_escaped_characters_in_place(raw_string) ⇒ Object

Replace any escaped unicode or whitespace with the actual characters To avoid allocating more strings, this modifies the string passed into it



301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
 
# File 'lib/graphql/language/lexer.rb', line 301

def self.replace_escaped_characters_in_place(raw_string)
  raw_string.gsub!(ESCAPED) do |matched_str|
    if (point_str_1 = $1 || $2)
      codepoint_1 = point_str_1.to_i(16)
      if (codepoint_2 = $3)
        codepoint_2 = codepoint_2.to_i(16)
        if (codepoint_1 >= 0xD800 && codepoint_1 <= 0xDBFF) && # leading surrogate
            (codepoint_2 >= 0xDC00 && codepoint_2 <= 0xDFFF) # trailing surrogate
          # A surrogate pair
          combined = ((codepoint_1 - 0xD800) * 0x400) + (codepoint_2 - 0xDC00) + 0x10000
          [combined].pack('U'.freeze)
        else
          # Two separate code points
          [codepoint_1].pack('U'.freeze) + [codepoint_2].pack('U'.freeze)
        end
      else
        [codepoint_1].pack('U'.freeze)
      end
    else
      ESCAPES_REPLACE[matched_str]
    end
  end
  nil
end

.tokenize(string) ⇒ Object

This is not used during parsing because the parser doesn’t actually need tokens.



328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
 
# File 'lib/graphql/language/lexer.rb', line 328

def self.tokenize(string)
  lexer = GraphQL::Language::Lexer.new(string)
  tokens = []
  prev_token = nil
  while (token_name = lexer.advance)
    new_token = [
      token_name,
      lexer.line_number,
      lexer.column_number,
      lexer.debug_token_value(token_name),
      prev_token,
    ]
    tokens << new_token
    prev_token = new_token
  end
  tokens
end

Instance Method Details

#_hash(key) ⇒ Object

This produces a unique integer for bytes 2 and 3 of each keyword string See https://tenderlovemaking.com/2023/09/02/fast-tokenizers-with-stringscanner.html



221
222
223
 
# File 'lib/graphql/language/lexer.rb', line 221

def _hash key
  (key * 18592990) >> 27 & 0x1f
end

#advance ⇒ Object 



22
23
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
 
# File 'lib/graphql/language/lexer.rb', line 22

def advance
  @scanner.skip(IGNORE_REGEXP)
  return false if @scanner.eos?
  @pos = @scanner.pos
  next_byte = @string.getbyte(@pos)
  next_byte_is_for = FIRST_BYTES[next_byte]
  case next_byte_is_for
  when ByteFor::PUNCTUATION
    @scanner.pos += 1
    PUNCTUATION_NAME_FOR_BYTE[next_byte]
  when ByteFor::NAME
    if len = @scanner.skip(KEYWORD_REGEXP)
      case len
      when 2
        :ON
      when 12
        :SUBSCRIPTION
      else
        pos = @pos

        # Use bytes 2 and 3 as a unique identifier for this keyword
        bytes = (@string.getbyte(pos + 2) << 8) | @string.getbyte(pos + 1)
        KEYWORD_BY_TWO_BYTES[_hash(bytes)]
      end
    else
      @scanner.skip(IDENTIFIER_REGEXP)
      :IDENTIFIER
    end
  when ByteFor::IDENTIFIER
    @scanner.skip(IDENTIFIER_REGEXP)
    :IDENTIFIER
  when ByteFor::NUMBER
    @scanner.skip(NUMERIC_REGEXP)
    # Check for a matched decimal:
    @scanner[1] ? :FLOAT : :INT
  when ByteFor::ELLIPSIS
    if @string.getbyte(@pos + 1) != 46 || @string.getbyte(@pos + 2) != 46
      raise_parse_error("Expected `...`, actual: #{@string[@pos..@pos + 2].inspect}")
    end
    @scanner.pos += 3
    :ELLIPSIS
  when ByteFor::STRING
    if @scanner.skip(BLOCK_STRING_REGEXP) || @scanner.skip(QUOTED_STRING_REGEXP)
      :STRING
    else
      raise_parse_error("Expected string or block string, but it was malformed")
    end
  else
    @scanner.pos += 1
    :UNKNOWN_CHAR
  end
rescue ArgumentError => err
  if err.message == "invalid byte sequence in UTF-8"
    raise_parse_error("Parse error on bad Unicode escape sequence", nil, nil)
  end
end

#column_number ⇒ Object 



141
142
143
 
# File 'lib/graphql/language/lexer.rb', line 141

def column_number
  @scanner.string[0..@pos].split("\n").last.length
end

#debug_token_value(token_name) ⇒ Object 



85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/graphql/language/lexer.rb', line 85

def debug_token_value(token_name)
  if token_name && Lexer::Punctuation.const_defined?(token_name)
    Lexer::Punctuation.const_get(token_name)
  elsif token_name == :ELLIPSIS
    "..."
  elsif token_name == :STRING
    string_value
  elsif @scanner.matched_size.nil?
    @scanner.peek(1)
  else
    token_value
  end
end

#eos? ⇒ Boolean

Returns:

  • (Boolean) 


16
17
18
 
# File 'lib/graphql/language/lexer.rb', line 16

def eos?
  @scanner.eos?
end

#line_number ⇒ Object 



137
138
139
 
# File 'lib/graphql/language/lexer.rb', line 137

def line_number
  @scanner.string[0..@pos].count("\n") + 1
end

#raise_parse_error(message, line = line_number, col = column_number) ⇒ Object

Raises:



145
146
147
 
# File 'lib/graphql/language/lexer.rb', line 145

def raise_parse_error(message, line = line_number, col = column_number)
  raise GraphQL::ParseError.new(message, line, col, @string, filename: @filename)
end

#string_value ⇒ Object 



114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
 
# File 'lib/graphql/language/lexer.rb', line 114

def string_value
  str = token_value
  is_block = str.start_with?('"""')
  if is_block
    str.gsub!(/\A"""|"""\z/, '')
    return Language::BlockString.trim_whitespace(str)
  else
    str.gsub!(/\A"|"\z/, '')

    if !str.valid_encoding? || !str.match?(VALID_STRING)
      raise_parse_error("Bad unicode escape in #{str.inspect}")
    else
      Lexer.replace_escaped_characters_in_place(str)

      if !str.valid_encoding?
        raise_parse_error("Bad unicode escape in #{str.inspect}")
      else
        str
      end
    end
  end
end

#token_value ⇒ Object 



79
80
81
82
83
 
# File 'lib/graphql/language/lexer.rb', line 79

def token_value
  @string.byteslice(@scanner.pos - @scanner.matched_size, @scanner.matched_size)
rescue StandardError => err
  raise GraphQL::Error, "(token_value failed: #{err.class}: #{err.message})"
end