Class: Predicator::Lexer

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

Constant Summary collapse

SPACE = 
/[ \t\r\n]/
DOT = 
/\./
LPAREN = 
/\(/
RPAREN = 
/\)/
NEQ = 
/!=/
GEQ = 
/>=/
LEQ = 
/<=/
GT = 
/>/
LT = 
/</
EQ = 
/=/
BANG = 
/!/
DATE = 
/(\d{4})-(\d{2})-(\d{2})/i
FLOAT = 
/[+-]?(?:[0-9_]+\.[0-9_]*|\.[0-9_]+|\d+(?=[eE]))(?:[eE][+-]?[0-9_]+)?\b/
INTEGER = 
/[+-]?\d(_?\d)*\b/
TRUE = 
/true\b/
FALSE = 
/false\b/
AND = 
/and/i
OR = 
/or/i
BETWEEN = 
/between/i
BLANK = 
/blank\?/i
PRESENT = 
/present\?/i
STRING = 
/(["])(?:\\?.)*?\1/
IDENTIFIER = 
/[a-z][A-Za-z0-9_]*/

Instance Method Summary collapse

Constructor Details

#initialize(string_or_io) ⇒ Lexer

Returns a new instance of Lexer.



30
31
32
33
34
35
36
37
38
 
# File 'lib/predicator/lexer.rb', line 30

def initialize string_or_io
  io = string_or_io.is_a?(String) ?
    StringIO.new(string_or_io) :
    string_or_io

  @ss = StringScanner.new io.read
  @tokens = []
  tokenize
end

Instance Method Details

#next_token ⇒ Object 



40
41
42
 
# File 'lib/predicator/lexer.rb', line 40

def next_token
  @tokens.shift
end

#tokenize ⇒ Object 



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
 
# File 'lib/predicator/lexer.rb', line 44

def tokenize
  until @ss.eos?
    case
    when @ss.scan(SPACE)
      # ignore space

    when text = @ss.scan(DOT)
      @tokens.push [:tDOT, text]

    when text = @ss.scan(LPAREN)
      @tokens.push [:tLPAREN, text]

    when text = @ss.scan(RPAREN)
      @tokens.push [:tRPAREN, text]

    when text = @ss.scan(NEQ)
      @tokens.push [:tNEQ, text]

    when text = @ss.scan(GEQ)
      @tokens.push [:tGEQ, text]

    when text = @ss.scan(LEQ)
      @tokens.push [:tLEQ, text]

    when text = @ss.scan(GT)
      @tokens.push [:tGT, text]

    when text = @ss.scan(LT)
      @tokens.push [:tLT, text]

    when text = @ss.scan(EQ)
      @tokens.push [:tEQ, text]

    when text = @ss.scan(BANG)
      @tokens.push [:tBANG, text]

    when text = @ss.scan(DATE)
      args = [ @ss[1], @ss[2], @ss[3] ].map(&:to_i)
      @tokens.push [:tDATE, args]

    when text = @ss.scan(FLOAT)
      @tokens.push [:tFLOAT, text]

    when text = @ss.scan(INTEGER)
      @tokens.push [:tINTEGER, text]

    when text = @ss.scan(TRUE)
      @tokens.push [:tTRUE, text]

    when text = @ss.scan(FALSE)
      @tokens.push [:tFALSE, text]

    when text = @ss.scan(AND)
      @tokens.push [:tAND, text]

    when text = @ss.scan(OR)
      @tokens.push [:tOR, text]

    when text = @ss.scan(BETWEEN)
      @tokens.push [:tBETWEEN, text]

    when text = @ss.scan(BLANK)
      @tokens.push [:tBLANK, text]

    when text = @ss.scan(PRESENT)
      @tokens.push [:tPRESENT, text]

    when text = @ss.scan(STRING)
      @tokens.push [:tSTRING, text[1..-2]]

    when text = @ss.scan(IDENTIFIER)
      @tokens.push [:tIDENTIFIER, text]

    else
      raise "Unexpected characters: #{@ss.peek(5).inspect}"
    end
  end

  @tokens.push [false, false]
end