Class: Tomlrb::Scanner

Inherits:
Object
  • Object
show all
Defined in:
lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb

Constant Summary collapse

COMMENT = 
/#[^\u0000-\u0008\u000A-\u001F\u007F]*/.freeze
IDENTIFIER = 
/[A-Za-z0-9_-]+/.freeze
SPACE = 
/[ \t]/.freeze
NEWLINE = 
/(?:[ \t]*(?:\r?\n)[ \t]*)+/.freeze
STRING_BASIC = 
/(")(?:\\?[^\u0000-\u0008\u000A-\u001F\u007F])*?\1/.freeze
STRING_MULTI = 
/"{3}([^\u0000-\u0008\u000B\u000C\u000E-\u001F\u007F]*?(?<!\\)"{3,5})/m.freeze
STRING_LITERAL = 
/(')(?:\\?[^\u0000-\u0008\u000A-\u001F\u007F])*?\1/.freeze
STRING_LITERAL_MULTI = 
/'{3}([^\u0000-\u0008\u000B\u000C\u000E-\u001F\u007F]*?'{3,5})/m.freeze
DATETIME = 
/(-?\d{4})-(\d{2})-(\d{2})(?:(?:t|\s)(\d{2}):(\d{2}):(\d{2}(?:\.\d+)?))?(z|[-+]\d{2}:\d{2})?/i.freeze
LOCAL_TIME = 
/(\d{2}):(\d{2}):(\d{2}(?:\.\d+)?)/.freeze
FLOAT = 
/[+-]?(?:(?:\d|[1-9](?:_?\d)*)\.\d(?:_?\d)*|\d+(?=[eE]))(?:[eE][+-]?[0-9]+(_[0-9])*[0-9]*)?(?!\w)/.freeze
FLOAT_KEYWORD = 
/[+-]?(?:inf|nan)\b/.freeze
INTEGER = 
/[+-]?([1-9](_?\d)*|0)(?![A-Za-z0-9_-]+)/.freeze
NON_DEC_INTEGER = 
/0(?:x[0-9A-Fa-f](_?[0-9A-Fa-f])*|o[0-7](_?[0-7])*|b[01](_?[01])*)/.freeze
BOOLEAN = 
/true|false/.freeze
SPACED_ARRAY_OF_TABLES_START = 
/^\[[ \t]+\[(#{IDENTIFIER}|#{STRING_BASIC}|#{STRING_LITERAL}|#{INTEGER}|#{NON_DEC_INTEGER}|#{FLOAT_KEYWORD}|#{BOOLEAN})\]\]$/.freeze
SPACED_ARRAY_OF_TABLES_END = 
/^\[\[(#{IDENTIFIER}|#{STRING_BASIC}|#{STRING_LITERAL}|#{INTEGER}|#{NON_DEC_INTEGER}|#{FLOAT_KEYWORD}|#{BOOLEAN})\][ \t]+\]$/.freeze
SPACED_ARRAY_OF_TABLES_BOTH = 
/^\[[ \t]+\[(#{IDENTIFIER}|#{STRING_BASIC}|#{STRING_LITERAL}|#{INTEGER}|#{NON_DEC_INTEGER}|#{FLOAT_KEYWORD}|#{BOOLEAN})\][ \t]+\]$/.freeze

Instance Method Summary collapse

Constructor Details

#initialize(io) ⇒ Scanner

Returns a new instance of Scanner.



44
45
46
47
 
# File 'lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb', line 44

def initialize(io)
  @ss = StringScanner.new(io.read)
  @eos = false
end

Instance Method Details

#next_token ⇒ Object 



49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
 
# File 'lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb', line 49

def next_token
  case
  when @ss.scan(NEWLINE) then [:NEWLINE, nil]
  when @ss.scan(SPACED_ARRAY_OF_TABLES_START) then raise ParseError.new('Array of tables has spaces in starting brackets')
  when @ss.scan(SPACED_ARRAY_OF_TABLES_END) then raise ParseError.new('Array of tables has spaces in ending brackets')
  when @ss.scan(SPACED_ARRAY_OF_TABLES_BOTH) then raise ParseError.new('Array of tables has spaces in starting and ending brackets')
  when @ss.scan(SPACE) then next_token
  when @ss.scan(COMMENT) then next_token
  when @ss.scan(DATETIME) then process_datetime
  when @ss.scan(LOCAL_TIME) then process_local_time
  when text = @ss.scan(STRING_MULTI) then [:STRING_MULTI, text[3..-4]]
  when text = @ss.scan(STRING_BASIC) then [:STRING_BASIC, text[1..-2]]
  when text = @ss.scan(STRING_LITERAL_MULTI) then [:STRING_LITERAL_MULTI, text[3..-4]]
  when text = @ss.scan(STRING_LITERAL) then [:STRING_LITERAL, text[1..-2]]
  when text = @ss.scan(FLOAT) then [:FLOAT, text]
  when text = @ss.scan(FLOAT_KEYWORD) then [:FLOAT_KEYWORD, text]
  when text = @ss.scan(INTEGER) then [:INTEGER, text]
  when text = @ss.scan(NON_DEC_INTEGER) then [:NON_DEC_INTEGER, text]
  when text = @ss.scan(BOOLEAN) then [:BOOLEAN, text]
  when text = @ss.scan(IDENTIFIER) then [:IDENTIFIER, text]
  when @ss.eos? then process_eos
  else x = @ss.getch; [x, x]
  end
end

#process_datetime ⇒ Object 



74
75
76
77
78
 
# File 'lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb', line 74

def process_datetime
  offset = @ss[7].gsub(/[zZ]/, '+00:00') if @ss[7]
  args = [@ss[1], @ss[2], @ss[3], @ss[4], @ss[5], @ss[6], offset]
  [:DATETIME, args]
end

#process_eos ⇒ Object 



85
86
87
88
89
90
 
# File 'lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb', line 85

def process_eos
  return if @eos

  @eos = true
  [:EOS, nil]
end

#process_local_time ⇒ Object 



80
81
82
83
 
# File 'lib/llm/shell/internal/tomlrb/lib/tomlrb/scanner.rb', line 80

def process_local_time
  args = [@ss[1], @ss[2], @ss[3].to_f]
  [:LOCAL_TIME, args]
end