Path of the image * * @return instance New Tesseract * ---------------------------------------------------------*/ public function __construct($fname=null){ /* [1] Check argument =========================================================*/ { /* (1) Check type */ if( !is_string($fname) ) throw new \Exception("Tesseract.__construct() expected but Tesseract.__construct(<".gettype($fname).">) received"); /* (2) Check file validity */ if( !file_exists($fname) ) throw new \Exception("Tesseract.__construct() but is not valid"); } /* [2] Store as attribute =========================================================*/ $this->fname = $fname; } /* (2) Read the image file * * @return read The read content * ---------------------------------------------------------*/ public function read(){ /* [1] Record the text from the image =========================================================*/ /* (1) Process tesseract */ $read = shell_exec("tesseract ".$this->fname." stdout -l fra --user-words ".__ROOT__."/config/edt.user-words -c language_model_penalty_non_freq_dict_word=0.1 -c language_model_penalty_non_dict_word=.15 -c tessedit_char_whitelist=abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789 2>/dev/null"); /* (2) If empty */ if( is_null($read) || !preg_match('@\n@', $read) ) throw new \Exception("Nothing read"); /* (3) Split by lines */ $by_line = explode("\n", $read); /* (4) Get first line (title) */ $title = $by_line[0]; /* (5) Get last non-empty line */ for( $i = count($by_line)-1 ; $i > 0 ; $i-- ){ // {1} Check not empty // if( empty($by_line[$i]) ) continue; // {2} Matches // if( preg_match('@^amphi@i', $by_line[$i]) || // 'amphi A', 'amphi 600 droit' preg_match('@^S\d+@i', $by_line[$i]) // 'S10', 'S22' ) return [ $title, $by_line[$i] ]; } return [ $title, 'unknown' ]; } }