Class: UniversalDetector::EUCJPProber

Inherits:
MultiByteCharSetProber show all
Defined in:
lib/EUCJPProber.rb

Instance Method Summary collapse

Methods inherited from CharSetProber

#filter_high_bit_only, #filter_with_english_letters, #filter_without_english_letters, #get_state

Constructor Details

#initialize ⇒ EUCJPProber

Returns a new instance of EUCJPProber.



38
39
40
41
42
43
44
 
# File 'lib/EUCJPProber.rb', line 38

def initialize
    super
    @_mCodingSM = CodingStateMachine.new(EUCJPSMModel)
    @_mDistributionAnalyzer = EUCJPDistributionAnalysis.new
    @_mContextAnalyzer = EUCJPContextAnalysis.new
    reset()
end

Instance Method Details

#feed(aBuf) ⇒ Object 



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
 
# File 'lib/EUCJPProber.rb', line 55

def feed(aBuf)
    aLen = aBuf.length
    for i in 0...aLen
        codingState = @_mCodingSM.next_state(aBuf[i])
        if codingState == :Error
            if DEBUG
                p(get_charset_name() + ' prober hit error at byte ' + i.to_s + '\n')
            end
            @_mState = :NotMe
            break
        elsif codingState == :ItsMe
            @_mState = :FoundIt
            break
        elsif codingState == :Start
            charLen = @_mCodingSM.get_current_charlen()
            if i == 0
                @_mLastChar[1] = aBuf[0]
                @_mContextAnalyzer.feed(@_mLastChar, charLen)
                @_mDistributionAnalyzer.feed(@_mLastChar, charLen)                        
            else
                @_mContextAnalyzer.feed(aBuf[i-1..i+1], charLen)
                @_mDistributionAnalyzer.feed(aBuf[i-1..i+1], charLen)
            end
        end
    end

    @_mLastChar[0] = aBuf[aLen - 1]

    if get_state() == :Detecting
        if @_mContextAnalyzer.got_enough_data() && (get_confidence() > SHORTCUT_THRESHOLD)
            @_mState = :FoundIt
        end
    end
    return get_state()
end

#get_charset_name ⇒ Object 



51
52
53
 
# File 'lib/EUCJPProber.rb', line 51

def get_charset_name
    return "EUC-JP"
end

#get_confidence ⇒ Object 



91
92
93
94
95
 
# File 'lib/EUCJPProber.rb', line 91

def get_confidence
    contxtCf = @_mContextAnalyzer.get_confidence()
    distribCf = @_mDistributionAnalyzer.get_confidence()
    return [contxtCf, distribCf].max
end

#reset ⇒ Object 



46
47
48
49
 
# File 'lib/EUCJPProber.rb', line 46

def reset
    super
    @_mContextAnalyzer.reset()
end