import re
import PyPDF2
import json

def extract_key_value_pairs(text):
    import re
    # 使用正则表达式来找到通常的键值对
    pattern = r'\d+\.\d+\s*([\w\s（）\u4e00-\u9fff]+)[\x01\x02☑]([\w\s\u4e00-\u9fff]+)'
    matches = re.findall(pattern, text)
    results = {}

    for key, value in matches:
        # 移除键中的数字和点序号
        key = re.sub(r'^\d+(\.\d+)*\s+', '', key)
        # 移除键中的多余空格和特殊字符
        cleaned_key = re.sub(r'\s+', '', key).replace('（', '').replace('）', '')
        # 清理值并停止提取到特定标点符号为止
        cleaned_value = re.split(r'[，。、,]', value)[0].strip()
        results[cleaned_key] = cleaned_value

    # 处理 '
' 或 '☑' 位于行首的特殊情况
    lines = text.split('\n')
    previous_lines = []
    last_serial_key = ""

    for line in lines:
        if re.match(r'^[\x01\x02☑√]', line):
            # 提取当前行的值，直到特定标点符号
            value = re.sub(r'^[\x01\x02☑√]\s*', '', line)
            cleaned_value = re.split(r'[，。、,]', value)[0].strip()

            # 使用最后一个包含序号的行作为键名的候选
            if last_serial_key:
                # 从最后一个有效序号行提取键名，并在特殊字符前停止
                key = re.sub(r'^\d+(\.\d+)*\s+', '', last_serial_key)
                key = re.split(r'[\x01\x02□]', key)[0].strip()  # 停止提取到特殊字符为止
                # 处理重复键名的情况
                original_key = key
                count = 1
                while key in results:
                    key = f"{original_key}{' ' * count}"
                    count += 1
                results[key] = cleaned_value
        else:
            # 保留最近的几行作为键的候选行
            if re.search(r'\d+\.\d+', line):
                last_serial_key = line  # 更新最后一个包含序号的行作为键名的候选
            previous_lines.append(line)
            if len(previous_lines) > 10:
                previous_lines.pop(0)

    return results



def read_pdf_and_judge_main(file_path, output_json_path):
    with open(file_path, 'rb') as file:
        reader = PyPDF2.PdfReader(file)
        num_pages = len(reader.pages)
        print(f"Total pages: {num_pages}")

        all_data = {}
        for page_num in range(num_pages):
            page = reader.pages[page_num]
            text = page.extract_text() if page.extract_text() else ""
            # 使用正则表达式删除每页开头的页码，格式通常为“数字+空格”
            cleaned_text = re.sub(r'^\d+\s+', '', text)

            key_value_pairs = extract_key_value_pairs( cleaned_text)
            all_data.update(key_value_pairs)
        # 和
有区别吗  
  
        # 将所有数据保存为 JSON 文件
        with open(output_json_path, "w", encoding="utf-8") as json_file:
            json.dump(all_data, json_file, ensure_ascii=False, indent=4)

        print(f"Data extraction complete and saved to '{output_json_path}'.")


if __name__ == "__main__":
    # 示例调用
    file_path = 'C:\\Users\\Administrator\\Desktop\\招标文件\\output1\\ztb_tobidders_notice_table.pdf'
    output_json_path = 'judge_exist.json'
    read_pdf_and_judge_main(file_path, output_json_path)