Class: CharDet::GB2312DistributionAnalysis

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

#initializeGB2312DistributionAnalysis

Returns a new instance of GB2312DistributionAnalysis.



143
144
145
146
147
148
# File 'lib/rchardet/chardistribution.rb', line 143

def initialize
  super()
  @_mCharToFreqOrder = GB2312CharToFreqOrder
  @_mTableSize = GB2312_TABLE_SIZE
  @_mTypicalDistributionRatio = GB2312_TYPICAL_DISTRIBUTION_RATIO
end

Instance Method Details

#get_order(aStr) ⇒ Object



150
151
152
153
154
155
156
157
158
159
160
# File 'lib/rchardet/chardistribution.rb', line 150

def get_order(aStr)
  # for GB2312 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..0] >= "\xB0") and (aStr[1..1] >= "\xA1")
	return 94 * (aStr[0] - 0xB0) + aStr[1] - 0xA1
  else
	return -1
  end
end