zbparse/flask_app/main/table_content_extraction.py

from docx import Document
import json


def read_tables_from_docx(file_path):
    """读取DOCX文件中的表格数据，并以嵌套字典的形式返回."""
    doc = Document(file_path)
    table_list = {}
    cur_title = []
    header = None

    for table in doc.tables:
        for i, row in enumerate(table.rows):
            cell_texts = [cell.text.strip() for cell in row.cells]

            # 检查是否是表头
            if header is None:
                header = cell_texts
                continue  # 跳过第一个表头行

            # 如果遇到与第一个表头相同的行，跳过
            if cell_texts == header:
                continue

            cur_level = table_list
            temp_title = []
            for j, cell in enumerate(row.cells):
                text_str = cell.text.strip().replace(' ', '').replace('\n', '')
                if j < len(row.cells) - 1:
                    if text_str == "":
                        text_str = cur_title[j] if j < len(cur_title) else "<未识别到上级标题>"
                    if text_str not in cur_level:
                        cur_level[text_str] = {}
                    cur_level = cur_level[text_str]
                    temp_title.append(text_str)
                else:
                    cell_text = cell.text.strip().replace(' ', '')
                    if len(temp_title) > 0:
                        last_key = temp_title[-1]
                        if last_key in cur_level:
                            if isinstance(cur_level[last_key], dict):
                                cur_level[last_key] = f"\n{cell_text}"
                            else:
                                cur_level[last_key] += f"\n{cell_text}"
                        else:
                            cur_level[last_key] = cell_text
                    else:
                        last_key = f"第{i}行内容"
                        if last_key in cur_level:
                            if isinstance(cur_level[last_key], dict):
                                cur_level[last_key] = f"\n{cell_text}"
                            else:
                                cur_level[last_key] += f"\n{cell_text}"
                        else:
                            cur_level[last_key] = cell_text
            cur_title = temp_title[:]

    return table_list

def flatten_nested_dicts(d):
    """平坦化嵌套字典，以便更简洁地保存为JSON."""
    keys_to_remove = []
    items_to_add = {}

    for key, value in list(d.items()):
        if isinstance(value, dict):
            value = flatten_nested_dicts(value)
            if len(value) == 1 and key in value:
                keys_to_remove.append(key)
                items_to_add[key] = value[key]
            elif len(value) == 1 and list(value.keys())[0] == list(value.values())[0]:
                items_to_add[key] = list(value.values())[0]

    for key in keys_to_remove:
        del d[key]
    d.update(items_to_add)

    return d

def save_data_to_json(data, filename):
    """将数据保存到JSON文件中."""
    with open(filename, 'w', encoding='utf-8') as file:
        json.dump(data, file, ensure_ascii=False, indent=4)

def extract_tables_main(path, output_filename):
    # 读取文档表格数据
    table_data = read_tables_from_docx(path)

    # 平坦化嵌套字典
    flattened_data = flatten_nested_dicts(table_data)

    # 保存平坦化后的数据到JSON文件
    save_data_to_json(flattened_data, output_filename)

    print(f"The data has been processed and saved to '{output_filename}'.")

if __name__ == "__main__":
    path = 'C:\\Users\\Administrator\\Desktop\\招标文件\\招标test文件夹\\zbtest20\\zbtest20_17-22.docx'
    output_filename = "C:\\Users\\Administrator\\Desktop\\招标文件\\招标test文件夹\\zbtest20\\truncate_output.json"  # 前附表json文件
    extract_tables_main(path, output_filename)