Class: Pcut::LineParser

Inherits:
Object
  • Object
show all
Defined in:
lib/pcut/line_parser.rb

Constant Summary collapse

DOUBLE_QUOTE = 
"\""
SINGLE_QUOTE = 
"'"
QUOTES = 
{
  DOUBLE_QUOTE => [DOUBLE_QUOTE, DOUBLE_QUOTE],
  SINGLE_QUOTE => [SINGLE_QUOTE, SINGLE_QUOTE],
  "D" => [DOUBLE_QUOTE, DOUBLE_QUOTE], # shortcut
  "S" => [SINGLE_QUOTE, SINGLE_QUOTE], # shortcut
  "[" => ["[", "]"],
  "(" => ["(", ")"],
  "<" => ["<", ">"]
}
DELIMITER = 
"\t"

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ LineParser

Returns a new instance of LineParser.



25
26
27
28
29
30
31
32
 
# File 'lib/pcut/line_parser.rb', line 25

def initialize
  @quote_guard  = {}
  @delimiter    = DELIMITER
  # do not delete quoting chars
  @keep_quotes  = false
  # treat continuous delimiters as one delimiters
  @skip_continuous_delimiters = false
end

Instance Attribute Details

#keep_quotes ⇒ Object

Returns the value of attribute keep_quotes.



20
21
22
 
# File 'lib/pcut/line_parser.rb', line 20

def keep_quotes
  @keep_quotes
end

#quote_guard ⇒ Object

Returns the value of attribute quote_guard.



20
21
22
 
# File 'lib/pcut/line_parser.rb', line 20

def quote_guard
  @quote_guard
end

#skip_continuous_delimiters ⇒ Object

Returns the value of attribute skip_continuous_delimiters.



20
21
22
 
# File 'lib/pcut/line_parser.rb', line 20

def skip_continuous_delimiters
  @skip_continuous_delimiters
end

Instance Method Details

#parse(str) ⇒ Object 



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
 
# File 'lib/pcut/line_parser.rb', line 48

def parse(str)
  buffer = nil
  previous_char = nil
  current_quote = nil
  finalizer     = nil
  partials = []

  str.each_char do |c|

    # skip continuous delimiters
    if skip_continuous_delimiters &&
      @delimiter == c &&
      @delimiter == previous_char
      next
    end

    # check starting quote
    if @quote_guard.include?(c) &&
      previous_char != "\\" # not escaped
      starter = @quote_guard[c].first

      if !current_quote && c == starter
        current_quote = c
        finalizer = @quote_guard[c].last
        if @keep_quotes
          buffer ||= ""
          buffer += c
        end
        next
      end
    end

    # check end of quote
    if current_quote && c == finalizer
      current_quote = nil
      finalizer     = nil
      if @keep_quotes
        buffer ||= ""
        buffer += c
      end
      next
    end

    if !current_quote && c == @delimiter
      partials << buffer
      buffer = nil
    elsif current_quote || c != @delimiter
      buffer ||= ""
      buffer += c
    end
    previous_char = c
  end
  partials << buffer if buffer
  partials
end

#set_delimiter(char) ⇒ Object 



41
42
43
44
45
46
 
# File 'lib/pcut/line_parser.rb', line 41

def set_delimiter(char)
  if !char.is_a?(String) || char.size != 1
    raise ArgumentError, "invalid delimiter: #{char.to_s}"
  end
  @delimiter = char
end

#set_quote_guard(quote) ⇒ Object 



34
35
36
37
38
39
 
# File 'lib/pcut/line_parser.rb', line 34

def set_quote_guard(quote)
  unless QUOTES.has_key?(quote)
    raise ArgumentError, "invalid quote: #{quote}"
  end
  @quote_guard[QUOTES[quote].first] = QUOTES[quote]
end