Module: JSON::Repair::StringUtils

Included in:
JSON::Repairer
Defined in:
lib/json/repair/string_utils.rb

Constant Summary collapse

BACKSLASH =

Constants for character chars

'\\'
SLASH =

0x5c

'/'
ASTERISK =

0x2f

'*'
OPENING_BRACE =

0x2a

'{'
CLOSING_BRACE =

0x7b

'}'
OPENING_BRACKET =

0x7d

'['
CLOSING_BRACKET =

0x5b

']'
OPEN_PARENTHESIS =

0x5d

'('
CLOSE_PARENTHESIS =

0x28

')'
SPACE =

0x29

' '
NEWLINE =

0x20

"\n"
TAB =

0xa

"\t"
RETURN =

0x9

"\r"
BACKSPACE =

0xd

"\b"
FORM_FEED =

0x08

"\f"
DOUBLE_QUOTE =

0x0c

'"'
PLUS =

0x0022

'+'
MINUS =

0x2b

'-'
QUOTE =

0x2d

"'"
ZERO =

0x27

'0'
NINE =

0x30

'9'
COMMA =

0x39

','
DOT =

0x2c

'.'
COLON =

0x2e

':'
SEMICOLON =

0x3a

';'
UPPERCASE_A =

0x3b

'A'
LOWERCASE_A =

0x41

'a'
UPPERCASE_E =

0x61

'E'
LOWERCASE_E =

0x45

'e'
UPPERCASE_F =

0x65

'F'
LOWERCASE_F =

0x46

'f'
NON_BREAKING_SPACE =

0x66

' '
MONGOLIAN_VOWEL_SEPARATOR =

0xa0

'᠎'
EN_QUAD =

0x180e

' '
ZERO_WIDTH_SPACE =

0x2000

'​'
NARROW_NO_BREAK_SPACE =

0x200b

' '
MEDIUM_MATHEMATICAL_SPACE =

0x202f

' '
IDEOGRAPHIC_SPACE =

0x205f

'　'
ZERO_WIDTH_NO_BREAK_SPACE =

0x3000

''
DOUBLE_QUOTE_LEFT =

0xfeff

'“'
DOUBLE_QUOTE_RIGHT =

0x201c

'”'
QUOTE_LEFT =

0x201d

'‘'
QUOTE_RIGHT =

0x2018

'’'
GRAVE_ACCENT =

0x2019

'`'
ACUTE_ACCENT =

0x0060

'´'
REGEX_DELIMITER =

0x00b4

%r{^[,:\[\]/{}()\n+]+$}
REGEX_UNQUOTED_STRING_DELIMITER = 
%r{^[,\[\]/{}\n+]+$}
REGEX_START_OF_VALUE = 
/^[\[{\w-]$/
REGEX_URL_START =

matches “https://” and other schemas

%r{^(http|https|ftp|mailto|file|data|irc)://$}
REGEX_URL_CHAR =

matches all valid URL characters EXCEPT “[”, “]”, and “,” (important JSON delimiters)

%r{^[A-Za-z0-9\-._~:/?#@!$&'()*+;=]$}
REGEX_FUNCTION_NAME_CHAR_START = 
/\A[a-zA-Z_$]\z/
REGEX_FUNCTION_NAME_CHAR = 
/\A[a-zA-Z0-9_$]\z/

Instance Method Summary collapse

Instance Method Details

#control_character?(char) ⇒ Boolean

Returns:

  • (Boolean) 


99
100
101
 
# File 'lib/json/repair/string_utils.rb', line 99

def control_character?(char)
  [NEWLINE, RETURN, TAB, BACKSPACE, FORM_FEED].include?(char)
end

#delimiter?(char) ⇒ Boolean

Returns:

  • (Boolean) 


76
77
78
 
# File 'lib/json/repair/string_utils.rb', line 76

def delimiter?(char)
  REGEX_DELIMITER.match?(char)
end

#digit?(char) ⇒ Boolean

Returns:

  • (Boolean) 


68
69
70
 
# File 'lib/json/repair/string_utils.rb', line 68

def digit?(char)
  char && char >= ZERO && char <= NINE
end

#double_quote?(char) ⇒ Boolean

Returns:

  • (Boolean) 


129
130
131
 
# File 'lib/json/repair/string_utils.rb', line 129

def double_quote?(char)
  char == DOUBLE_QUOTE
end

#double_quote_like?(char) ⇒ Boolean

Returns:

  • (Boolean) 


137
138
139
 
# File 'lib/json/repair/string_utils.rb', line 137

def double_quote_like?(char)
  [DOUBLE_QUOTE, DOUBLE_QUOTE_LEFT, DOUBLE_QUOTE_RIGHT].include?(char)
end

#ends_with_comma_or_newline?(text) ⇒ Boolean

Returns:

  • (Boolean) 


193
194
195
 
# File 'lib/json/repair/string_utils.rb', line 193

def ends_with_comma_or_newline?(text)
  /[,\n][ \t\r]*$/.match?(text)
end

#function_name_char?(char) ⇒ Boolean

Returns:

  • (Boolean) 


91
92
93
 
# File 'lib/json/repair/string_utils.rb', line 91

def function_name_char?(char)
  !char.nil? && REGEX_FUNCTION_NAME_CHAR.match?(char)
end

#function_name_char_start?(char) ⇒ Boolean

Returns:

  • (Boolean) 


87
88
89
 
# File 'lib/json/repair/string_utils.rb', line 87

def function_name_char_start?(char)
  !char.nil? && REGEX_FUNCTION_NAME_CHAR_START.match?(char)
end

#hex?(char) ⇒ Boolean

Functions to check character chars

Returns:

  • (Boolean) 


62
63
64
65
66
 
# File 'lib/json/repair/string_utils.rb', line 62

def hex?(char)
  (char >= ZERO && char <= NINE) ||
    (char >= UPPERCASE_A && char <= UPPERCASE_F) ||
    (char >= LOWERCASE_A && char <= LOWERCASE_F)
end

#insert_before_last_whitespace(text, text_to_insert) ⇒ Object 



154
155
156
157
158
159
160
161
162
 
# File 'lib/json/repair/string_utils.rb', line 154

def insert_before_last_whitespace(text, text_to_insert)
  index = text.length

  return text + text_to_insert unless whitespace?(text[index - 1])

  index -= 1 while whitespace?(text[index - 1])

  text[0...index] + text_to_insert + text[index..]
end

#parse_keyword(name, value) ⇒ Object 



179
180
181
182
183
184
185
186
187
 
# File 'lib/json/repair/string_utils.rb', line 179

def parse_keyword(name, value)
  if @json[@index, name.length] == name
    @output << value
    @index += name.length
    true
  else
    false
  end
end

#parse_keywords ⇒ Object

Parse keywords true, false, null Repair Python keywords True, False, None Repair Ruby keyword nil



167
168
169
170
171
172
173
174
175
176
177
 
# File 'lib/json/repair/string_utils.rb', line 167

def parse_keywords
  parse_keyword('true', 'true') ||
    parse_keyword('false', 'false') ||
    parse_keyword('null', 'null') ||
    # Repair Python keywords True, False, None
    parse_keyword('True', 'true') ||
    parse_keyword('False', 'false') ||
    parse_keyword('None', 'null') ||
    # Repair Ruby keyword nil
    parse_keyword('nil', 'null')
end

#quote?(char) ⇒ Boolean

Returns:

  • (Boolean) 


125
126
127
 
# File 'lib/json/repair/string_utils.rb', line 125

def quote?(char)
  double_quote_like?(char) || single_quote_like?(char)
end

#remove_at_index(text, start, count) ⇒ Object 



189
190
191
 
# File 'lib/json/repair/string_utils.rb', line 189

def remove_at_index(text, start, count)
  text[0...start] + text[start + count..]
end

#single_quote?(char) ⇒ Boolean

Returns:

  • (Boolean) 


133
134
135
 
# File 'lib/json/repair/string_utils.rb', line 133

def single_quote?(char)
  char == QUOTE
end

#single_quote_like?(char) ⇒ Boolean

Returns:

  • (Boolean) 


141
142
143
 
# File 'lib/json/repair/string_utils.rb', line 141

def single_quote_like?(char)
  [QUOTE, QUOTE_LEFT, QUOTE_RIGHT, GRAVE_ACCENT, ACUTE_ACCENT].include?(char)
end

#special_whitespace?(char) ⇒ Boolean

Returns:

  • (Boolean) 


111
112
113
114
115
116
117
118
119
120
121
122
123
 
# File 'lib/json/repair/string_utils.rb', line 111

def special_whitespace?(char)
  return false unless char

  [
    NON_BREAKING_SPACE,
    MONGOLIAN_VOWEL_SEPARATOR,
    NARROW_NO_BREAK_SPACE,
    MEDIUM_MATHEMATICAL_SPACE,
    IDEOGRAPHIC_SPACE,
    ZERO_WIDTH_NO_BREAK_SPACE
  ].include?(char) ||
    (char >= EN_QUAD && char <= ZERO_WIDTH_SPACE)
end

#start_of_value?(char) ⇒ Boolean

Returns:

  • (Boolean) 


95
96
97
 
# File 'lib/json/repair/string_utils.rb', line 95

def start_of_value?(char)
  REGEX_START_OF_VALUE.match?(char) || (char && quote?(char))
end

#strip_last_occurrence(text, text_to_strip, strip_remaining_text: false) ⇒ Object

Strip last occurrence of text_to_strip from text



146
147
148
149
150
151
152
 
# File 'lib/json/repair/string_utils.rb', line 146

def strip_last_occurrence(text, text_to_strip, strip_remaining_text: false)
  index = text.rindex(text_to_strip)
  return text unless index

  remaining_text = strip_remaining_text ? '' : text[index + 1..]
  text[0...index] + remaining_text
end

#unquoted_string_delimiter?(char) ⇒ Boolean

Returns:

  • (Boolean) 


80
81
82
 
# File 'lib/json/repair/string_utils.rb', line 80

def unquoted_string_delimiter?(char)
  REGEX_UNQUOTED_STRING_DELIMITER.match?(char)
end

#valid_string_character?(char) ⇒ Boolean

Returns:

  • (Boolean) 


72
73
74
 
# File 'lib/json/repair/string_utils.rb', line 72

def valid_string_character?(char)
  char.ord >= 0x20 && char.ord <= 0x10ffff
end

#whitespace?(char) ⇒ Boolean

Returns:

  • (Boolean) 


103
104
105
 
# File 'lib/json/repair/string_utils.rb', line 103

def whitespace?(char)
  [SPACE, NEWLINE, TAB, RETURN].include?(char)
end

#whitespace_except_newline?(char) ⇒ Boolean

Returns:

  • (Boolean) 


107
108
109
 
# File 'lib/json/repair/string_utils.rb', line 107

def whitespace_except_newline?(char)
  [SPACE, TAB, RETURN].include?(char)
end