class CharDet::SingleByteCharSetProber
NEGATIVE_CAT = 0
Public Class Methods
new(model, reversed=false, nameProber=nil)
click to toggle source
Calls superclass method
CharDet::CharSetProber::new
# File lib/rchardet/sbcharsetprober.rb, line 41 def initialize(model, reversed=false, nameProber=nil) super() @model = model @reversed = reversed # TRUE if we need to reverse every pair in the model lookup @nameProber = nameProber # Optional auxiliary prober for name decision reset() end
Public Instance Methods
feed(aBuf)
click to toggle source
# File lib/rchardet/sbcharsetprober.rb, line 66 def feed(aBuf) if !@model['keepEnglishLetter'] aBuf = filter_without_english_letters(aBuf) end aLen = aBuf.length if aLen == 0 return get_state() end aBuf.each_byte do |b| c = b.chr order = @model['charToOrderMap'][c.bytes.first] if order < SYMBOL_CAT_ORDER @totalChar += 1 end if order < SAMPLE_SIZE @freqChar += 1 if @lastOrder < SAMPLE_SIZE @totalSeqs += 1 if !@reversed @seqCounters[@model['precedenceMatrix'][(@lastOrder * SAMPLE_SIZE) + order]] += 1 else # reverse the order of the letters in the lookup @seqCounters[@model['precedenceMatrix'][(order * SAMPLE_SIZE) + @lastOrder]] += 1 end end end @lastOrder = order end if get_state() == EDetecting if @totalSeqs > SB_ENOUGH_REL_THRESHOLD cf = get_confidence() if cf > POSITIVE_SHORTCUT_THRESHOLD $stderr << "#{@model['charsetName']} confidence = #{cf}, we have a winner\n" if $debug @state = EFoundIt elsif cf < NEGATIVE_SHORTCUT_THRESHOLD $stderr << "#{@model['charsetName']} confidence = #{cf}, below negative shortcut threshold #{NEGATIVE_SHORTCUT_THRESHOLD}\n" if $debug @state = ENotMe end end end return get_state() end
get_charset_name()
click to toggle source
# File lib/rchardet/sbcharsetprober.rb, line 58 def get_charset_name if @nameProber return @nameProber.get_charset_name() else return @model['charsetName'] end end
get_confidence()
click to toggle source
# File lib/rchardet/sbcharsetprober.rb, line 110 def get_confidence r = 0.01 if @totalSeqs > 0 r = (1.0 * @seqCounters[POSITIVE_CAT]) / @totalSeqs / @model['mTypicalPositiveRatio'] r = r * @freqChar / @totalChar if r >= 1.0 r = 0.99 end end return r end
reset()
click to toggle source
Calls superclass method
CharDet::CharSetProber#reset
# File lib/rchardet/sbcharsetprober.rb, line 49 def reset super() @lastOrder = 255 # char order of last character @seqCounters = [0] * NUMBER_OF_SEQ_CAT @totalSeqs = 0 @totalChar = 0 @freqChar = 0 # characters that fall in our sampling range end