knowledge/tests/trunck_files.py

import os
import re

def rename_doc_files(directory):
    for root, dirs, files in os.walk(directory):
        for file in files:
            if file.endswith('.doc'):
                old_path = os.path.join(root, file)
                #文件名前面的数字及.去掉
                new_name = re.sub(r'^\d+\.', '', file).strip()
                if new_name != file:
                    new_path = os.path.join(root, new_name)
                    os.rename(old_path, new_path)
                    print(f'Renamed: {old_path} -> {new_path}')

def move_doc_to_ocr(directory):
    for root, dirs, files in os.walk(directory):
        for file in files:
            if file.endswith('.doc'):
                file_name = os.path.splitext(file)[0]
                folder_path = os.path.join(root, file_name)
                ocr_path = os.path.join(folder_path, 'ocr')
                if not os.path.exists(folder_path):
                    os.makedirs(ocr_path)
                old_path = os.path.join(root, file)
                new_path = os.path.join(ocr_path, file)
                os.rename(old_path, new_path)
                print(f'Moved: {old_path} -> {new_path}')
    for root, dirs, files in os.walk(directory):
        for file in files:
            if file.endswith('.doc'):
                old_path = os.path.join(root, file)
                new_name = re.sub(r'^\d+\.', '', file)
                if new_name != file:
                    new_path = os.path.join(root, new_name)
                    os.rename(old_path, new_path)
                    print(f'Renamed: {old_path} -> {new_path}')

if __name__ == '__main__':
    directory = 'E:\急诊科资料\中华医学期刊数据库'
    #rename_doc_files(directory)
    move_doc_to_ocr(directory)