Class: CharDet::EUCKRDistributionAnalysis

Inherits:
CharDistributionAnalysis show all
Defined in:
lib/rchardet/chardistribution.rb

Instance Method Summary collapse

Methods inherited from CharDistributionAnalysis

#feed, #get_confidence, #got_enough_data, #reset

Constructor Details

#initialize ⇒ EUCKRDistributionAnalysis

Returns a new instance of EUCKRDistributionAnalysis.



131
132
133
134
135
136
 
# File 'lib/rchardet/chardistribution.rb', line 131

def initialize
  super()
  @charToFreqOrder = EUCKRCharToFreqOrder
  @tableSize = EUCKR_TABLE_SIZE
  @typicalDistributionRatio = EUCKR_TYPICAL_DISTRIBUTION_RATIO
end

Instance Method Details

#get_order(aStr) ⇒ Object 



138
139
140
141
142
143
144
145
146
147
148
149
 
# File 'lib/rchardet/chardistribution.rb', line 138

def get_order(aStr)
  # for euc-KR encoding, we are interested 
  #   first  byte range: 0xb0 -- 0xfe
  #   second byte range: 0xa1 -- 0xfe
  # no validation needed here. State machine has done that
  if aStr[0, 1] >= "\xB0"
    bytes = aStr.bytes.to_a
    return 94 * (bytes[0] - 0xB0) + bytes[1] - 0xA1
  else
    return -1
  end
end