1576fffbcSThomas Huth# ... 2576fffbcSThomas Huth# 3576fffbcSThomas Huth# Copyright (c) 2019 Philippe Mathieu-Daudé <f4bug@amsat.org> 4576fffbcSThomas Huth# 5576fffbcSThomas Huth# This work is licensed under the terms of the GNU GPL, version 2 or 6576fffbcSThomas Huth# later. See the COPYING file in the top-level directory. 7576fffbcSThomas Huth 8576fffbcSThomas Huthimport logging 9*37e9b19cSDaniel P. Berrangéfrom subprocess import run 10576fffbcSThomas Huth 11576fffbcSThomas Huth 12576fffbcSThomas Huthdef tesseract_ocr(image_path, tesseract_args=''): 13576fffbcSThomas Huth console_logger = logging.getLogger('console') 14576fffbcSThomas Huth console_logger.debug(image_path) 15*37e9b19cSDaniel P. Berrangé proc = run(['tesseract', image_path, 'stdout'], 16*37e9b19cSDaniel P. Berrangé capture_output=True, encoding='utf8') 17*37e9b19cSDaniel P. Berrangé if proc.returncode: 18576fffbcSThomas Huth return None 19576fffbcSThomas Huth lines = [] 20*37e9b19cSDaniel P. Berrangé for line in proc.stdout.split('\n'): 21576fffbcSThomas Huth sline = line.strip() 22576fffbcSThomas Huth if len(sline): 23576fffbcSThomas Huth console_logger.debug(sline) 24576fffbcSThomas Huth lines += [sline] 25576fffbcSThomas Huth return lines 26