prop_extraction.py 文件源码

def main():

    args = docopt(__doc__)

    # Parse args
    inp = args['--in']
    out = args['--out']
    include_id = args["--id"] is not None

    logging.info("Loading spaCy...")
    pe = prop_extraction(include_id)

    # Differentiate between single input file and directories
    if os.path.isdir(inp):
        logging.debug('Running on directories:')
        num_of_lines = num_of_extractions = 0

        for input_fn in glob(os.path.join(inp, '*.txt')):
            output_fn = os.path.join(out, path_leaf(input_fn).replace('.txt', '.prop'))
            logging.debug('input file: {}\noutput file:{}'.format(input_fn, output_fn))

            cur_line_counter, cur_extractions_counter = run_single_file(input_fn, output_fn, pe)
            num_of_lines += cur_line_counter
            num_of_extractions += cur_extractions_counter

    else:
        logging.debug('Running on single files:')
        num_of_lines, num_of_extractions = run_single_file(inp, out, pe)

    logging.info('# Sentences: {} \t #Extractions: {} \t Extractions/sentence Ratio: {}'.
                 format(num_of_lines, num_of_extractions, float(num_of_extractions) / num_of_lines))