<_9Y � @ sF d d l m Z d d l m Z m Z m Z Gd d � d e � Z d S)� )�
CharSetProber)�CharacterCategory� ProbingState�SequenceLikelihoodc s� e Z d Z d Z d Z d Z d Z d d � f d d � Z � f d d
� Z e d d � � Z
e d
d � � Z d d � Z d d � Z
� S)�SingleByteCharSetProber�@ i gffffff�?g�������?FNc si t t | � j � | | _ | | _ | | _ d | _ d | _ d | _ d | _ d | _
| j � d S)N) �superr �__init__�_model� _reversed� _name_prober� _last_order�
_seq_counters� _total_seqs� _total_char�
_freq_char�reset)�self�model�reversedZ name_prober)� __class__� �/sbcharsetprober.pyr ' s z SingleByteCharSetProber.__init__c sQ t t | � j � d | _ d g t j � | _ d | _ d | _ d | _ d S)N� � )
r r r r
r Zget_num_categoriesr r r r )r )r r r r 5 s z SingleByteCharSetProber.resetc C s" | j r | j j S| j d Sd S)N� charset_name)r r r
)r r r r r ? s
z$SingleByteCharSetProber.charset_namec C s' | j r | j j S| j j d � Sd S)N�language)r r r
�get)r r r r r F s
z SingleByteCharSetProber.languagec C s� | j d s | j | � } | s) | j S| j d } x� t | � D]� \ } } | | } | t j k rw | j d 7_ | | j k r | j d 7_ | j | j k r | j
d 7_
| j s� | j | j | } | j d | } n% | | j | j } | j d | } | j | d 7<| | _ qC W| j d } | j t
j k r�| j
| j k r�| j � } | | j k r�| j j d | | � t
j | _ n7 | | j k r�| j j d | | | j � t
j | _ | j S)NZkeep_english_letter�char_to_order_mapr Zprecedence_matrixr z$%s confidence = %s, we have a winnerz9%s confidence = %s, below negative shortcut threshhold %s)r
Zfilter_international_words�state� enumerater ZCONTROLr � SAMPLE_SIZEr r
r r r r Z DETECTING�SB_ENOUGH_REL_THRESHOLD�get_confidence�POSITIVE_SHORTCUT_THRESHOLD�logger�debugZFOUND_IT�_state�NEGATIVE_SHORTCUT_THRESHOLDZNOT_ME) r Zbyte_strr �i�c�orderr r Z
confidencer r r �feedM sF
z SingleByteCharSetProber.feedc C se d } | j d k ra d | j t j | j | j d } | | j | j } | d k ra d } | S)Ng{�G�z�?r g �?Ztypical_positive_ratiog�G�z��?)r r r ZPOSITIVEr
r r )r �rr r r r# | s & z&SingleByteCharSetProber.get_confidence)�__name__�
__module__� __qualname__r! r" r$ r( r r �propertyr r r, r# r r )r r r ! s
/r N)Z
charsetproberr Zenumsr r r r r r r r �<module> s