3 ÝPfö ã @ s0 d dl Z d dlZddlmZ G dd„ deƒZdS )é Né )ÚProbingStatec @ sn e Zd ZdZddd„Zdd„ Zedd„ ƒZd d „ Zedd„ ƒZ d d„ Z edd„ ƒZedd„ ƒZ edd„ ƒZdS )Ú CharSetProbergffffffî?Nc C s d | _ || _tjtƒ| _d S )N)Ú_stateÚlang_filterÚloggingZ getLoggerÚ__name__Zlogger)Úselfr © r ú#/usr/lib/python3.6/charsetprober.pyÚ__init__' s zCharSetProber.__init__c C s t j| _d S )N)r Z DETECTINGr )r r r r Úreset, s zCharSetProber.resetc C s d S )Nr )r r r r Úcharset_name/ s zCharSetProber.charset_namec C s d S )Nr )r Úbufr r r Úfeed3 s zCharSetProber.feedc C s | j S )N)r )r r r r Ústate6 s zCharSetProber.statec C s dS )Ng r )r r r r Úget_confidence: s zCharSetProber.get_confidencec C s t jdd| ƒ} | S )Ns ([ -])+ó )ÚreÚsub)r r r r Úfilter_high_byte_only= s z#CharSetProber.filter_high_byte_onlyc C sb t ƒ }tjd| ƒ}xJ|D ]B}|j|dd… ƒ |dd… }|jƒ rP|dk rPd}|j|ƒ qW |S )u9 We define three types of bytes: alphabet: english alphabets [a-zA-Z] international: international characters [€-ÿ] marker: everything else [^a-zA-Z€-ÿ] The input buffer can be thought to contain a series of words delimited by markers. This function works to filter all words that contain at least one international character. All contiguous sequences of markers are replaced by a single space ascii character. This filter applies to all scripts which do not use English characters. s% [a-zA-Z]*[€-ÿ]+[a-zA-Z]*[^a-zA-Z€-ÿ]?Nr ó €r éÿÿÿÿr )Ú bytearrayr ÚfindallÚextendÚisalpha)r ÚfilteredZwordsZwordZ last_charr r r Úfilter_international_wordsB s z(CharSetProber.filter_international_wordsc C s¬ t ƒ }d}d}x‚tt| ƒƒD ]r}| ||d … }|dkr>d}n|dkrJd}|dk r|jƒ r||kr†| r†|j| ||… ƒ |jdƒ |d }qW |s¨|j| |d … ƒ |S ) aÈ Returns a copy of ``buf`` that retains only the sequences of English alphabet and high byte characters that are not between <> characters. Also retains English alphabet and high byte characters immediately before occurrences of >. This filter can be applied to all scripts which contain both English characters and extended ASCII characters, but is currently only used by ``Latin1Prober``. Fr r ó >ó