Class: UniversalDetector::EUCTWDistributionAnalysis

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

Constant Summary

Constants inherited from CharDistributionAnalysis

CharDistributionAnalysis::ENOUGH_DATA_THRESHOLD, CharDistributionAnalysis::SURE_NO, CharDistributionAnalysis::SURE_YES

Instance Method Summary collapse

Methods inherited from CharDistributionAnalysis

#feed, #get_confidence, #got_enough_data, #reset

Constructor Details

#initializeEUCTWDistributionAnalysis

Returns a new instance of EUCTWDistributionAnalysis.



109
110
111
112
113
114
# File 'lib/CharDistributionAnalysis.rb', line 109

def initialize
    super
    @_mCharToFreqOrder = EUCTWCharToFreqOrder
    @_mTableSize = EUCTW_TABLE_SIZE
    @_mTypicalDistributionRatio = EUCTW_TYPICAL_DISTRIBUTION_RATIO
end

Instance Method Details

#get_order(aStr) ⇒ Object



116
117
118
119
120
121
122
123
124
125
126
# File 'lib/CharDistributionAnalysis.rb', line 116

def get_order(aStr)
    # for euc-TW encoding, we are interested 
    #   first  byte range: 0xc4 -- 0xfe
    #   second byte range: 0xa1 -- 0xfe
    # no validation needed here. State machine has done that
    if aStr[0] >= 0xC4
        return 94 * (aStr[0] - 0xC4) + aStr[1] - 0xA1
    else
        return -1
    end
end