utils.py

import fitz
import json
import os
import re
import logging
import hid
from model import CLIENT,simple_llm
from jsonschema import validate, ValidationError

def init_logger(name,path,log_level=logging.DEBUG,console_level=logging.DEBUG,file_level=logging.DEBUG,silence=False):
    # 创建logger对象
    logger = logging.getLogger(name)
    logger.setLevel(log_level)  # 设置最低级别

    formatter = logging.Formatter('%(asctime)s - %(levelname)s - %(message)s')

    # 创建控制台处理器并设置级别
    if not silence:
        console_handler = logging.StreamHandler()
        console_handler.setLevel(console_level)
        console_handler.setFormatter(formatter)
        logger.addHandler(console_handler)

    # 创建文件处理器并设置级别
    file_handler = logging.FileHandler(path, mode='a', encoding='utf-8')
    file_handler.setLevel(file_level)  # 只记录错误及以上的日志
    file_handler.setFormatter(formatter)
    logger.addHandler(file_handler)

    return logger

#deep merge,using for patch
def deep_merge(dict1, dict2):
    """
    递归地合并两个字典。如果两个字典有相同的键且对应的值都是字典，
    则递归合并这些子字典；否则，dict2 中的值会覆盖 dict1 中的值。
    """
    for key in dict2:
        if key in dict1:
            if isinstance(dict1[key], dict) and isinstance(dict2[key], dict):
                deep_merge(dict1[key], dict2[key])
            else:
                dict1[key] = dict2[key]
        else:
            dict1[key] = dict2[key]
    return dict1

#merge a list of json dict
def merge_json_dicts(file_paths):
    merged_dict = {}
    for file_path in file_paths:
        # print(file_path)
        if not os.path.exists(file_path):
            print(f"Warning: File {file_path} does not exist.")
            continue
        with open(file_path, 'r', encoding='utf-8') as file:
            data = json.load(file)
            if isinstance(data, dict):
                merged_dict = deep_merge(merged_dict, data)
            else:
                print(f"Warning: {file_path} does not contain a dictionary.")
    return merged_dict

def get_pdf_links(pdf_path):
    # 打开PDF文件
    document = fitz.open(pdf_path)
    links = []

    for page_num in range(len(document)):
        page = document[page_num]
        # 获取当前页面的链接
        for link in page.get_links():
            links.append((page_num, link))

    document.close()
    return links

def get_toc(pdf_file_path):
    pdf_file = pdf_file_path
    pdf = fitz.open(pdf_file)
    toc = pdf.get_toc()
    dic = {}
    for item in toc:
        title = item[1]
        page = item[2]
        level = item[0]
        dic[title] = page 
        #print(f"title: {title}, page: {page}, level: {level}")
        if level == 1:
            print(f"{title} - {page}")
        elif level == 2:
            print(f"    {title} - {page}")
        elif level == 3:
            print(f"        {title} - {page}")
        elif level == 4:
            print(f"            {title} - {page}")
    print(len(dic))
    pdf.close()
    # print(dic["2.1 - Unguided Bomb (MK-82) - CCIP / Computer Pilot"])

def clean_filename(filename):
    # 替换或删除特殊字符
    invalid_chars = '<>:"/\\|?*'
    for char in invalid_chars:
        filename = filename.replace(char, '')
    return filename

def split_pdf(pdf_file_path,save_path,page_list=None):
    pdf_file = pdf_file_path
    pdf = fitz.open(pdf_file)
    if page_list is not None:
        new_pdf = fitz.open()
        new_pdf.insert_pdf(pdf, from_page=page_list[0], to_page=page_list[1])
        filename = clean_filename(pdf_file)
        new_pdf.save(f"{save_path}\\{filename}.pdf")
        
    else:
        toc = pdf.get_toc()
        dic = []
        for item in toc:
            title = item[1]
            page = item[2]
            level = item[0]
            dic.append((title, page, level))

        last_page = 0
        last_title = ""
        cnt = 1
        for item in dic:
            title = item[0]
            page = item[1] - 1
            if last_title and last_page != page:
                new_pdf = fitz.open()
                new_pdf.insert_pdf(pdf, from_page=last_page, to_page=page - 1)
                filename = clean_filename(last_title)
                new_pdf.save(f"{save_path}\{cnt:03d}_{filename}.pdf")
                cnt = cnt + 1
                print(f"Saved {last_title}.pdf")
                new_pdf.close()    
            last_page = page
            last_title = title
    pdf.close()

def crop_pdf(pdf_file_path,save_path):
    pdf_path = pdf_file_path  # 这里替换成你的PDF文件路径
    pdf_doc = fitz.open(pdf_path)
    for page_number in range(len(pdf_doc)):
        page = pdf_doc[page_number]  # 获取单页
        new_rect = fitz.Rect(64,54,548,740)
        # 设置裁剪区域
        page.set_cropbox(new_rect)
    pdf_doc.save(save_path)
    pdf_doc.close()

def get_text_from_pdf(filename):
    pdf = fitz.open(filename)
    text = ''
    for page_number in range(len(pdf)):
        page = pdf[page_number]  # 获取单页
        text = text + page.get_text()  # 提取页面文本

        # print(f"Page {page_number + 1}:")
        # print(text)
    # 关闭PDF文件
    # print(text)
    pdf.close()
    return text

#syntax correction
def syntax_correction(data):
    #data cleaning
    data.replace('–','-')
    cleaned_data = re.sub(r'^```json(.*?)```$', r'\1', data, flags=re.DOTALL).strip()

    return cleaned_data

#always return none unless occur format wrong
def format_correction(json_data):
    output = None
    ve = None
    #check schema
    schema = {
        "type": "object",
        "properties": {
            "name": {"type": "string"},
            "type": {"type": "string"},
            "content":{
                "type": "array",
                "items": {
                    "type": "object",
                    "properties": {
                        "step": {"type": "integer","minimum": 1},
                        "agent": {"type": "string",
                                  "enum": ["PILOT","RIO","JESTER"]},
                        "instruction": {"type": "string"},
                        "substeps": {
                            "type": "array",
                            "items": {
                                "type": "string"
                                     }
                                    },
                        "note": {"type": "string"}
                                },
                    "required": ["step","instruction","agent"]
                        }
                },
            "description": {"type":"string"}
        },
        "required": ["name","type", "content","description"]
    }
    try:
        validate(instance=json_data, schema=schema)
    except ValidationError as e:
        ve = e
        # print(ve)
        tmp = '''
        将给你一段jsonschema的报错信息，请根据报错信息修改json文件，并且最后请输出json格式的数据,除了修改后的json文件外不要输出任何信息
        **error**:
        {error}
        **json**:
        {data}
        '''
        tmp = tmp.format(error=ve,data=json.dumps(json_data,indent=4))
        client,model = CLIENT("deepseek")
        chat_completion = client.chat.completions.create(
        messages=[
            {
                "role":"user",
                "content": tmp
            }
        ],
        model=model,
        response_format={"type": "json_object"},)
        output = syntax_correction(chat_completion.choices[0].message.content)
    return output,ve

#get single subtask from single pdf, always return null unless get error
def get_subtask_from_pdf(pdf_path,save_path,logger):
    tmp1 = '''
        You're a fighter pilot assistant and you'll need to learn some specialized knowledge, and you'll be given some documentation from which you'll need to extract the relevant knowledge of the sub-missions.
    Please extract the complete operation steps contained in the possible subtasks in each document, give these collections of steps a suitable name and description, but do not keep the serial number for the title.
    Pay attention to the format of the output, e.g. when exporting ", please replace it with ', replace – with -, etc. Do not output any other information other than that.
    The output is in json format, here's an example:
    **example1**
    *input*:
    {input1}
    *output*:
    {output1}
    **end of example1**
    Note the distinction between operations with [J], which usually refers to another form of RIO, Jester, here is an example:
    **example2**
    *input*:
    {input2}
    *output*:
    {output2}
    **end of example2**
    Sometimes you need to combine the respective functions of the two drivers and our mission scenario to deduce who the executor of the command is, in the following example it can be inferred that the radar is operated by the backseat RIO, so it can be inferred that the executor is RIO:
    **example3**
    *input*:
    {input3}
    *output*:
    {output3}
    **end of example3**
'''
    
    tmp2 ='''
    This is the content of the document you need to extract:
    {content}
    '''
    #提取PILOT/RIO操作示例
    example_input1 = get_text_from_pdf('data\\pdf\\cropped_TomcatGuide_split\\099_2.7 - TALD Decoys.pdf')
    with open('data\\json\\PR_example.json', 'r',encoding='utf-8') as file:
        data = json.load(file)
    example_output1 = json.dumps(data, indent=4,ensure_ascii=False)
    #提取PILOT/JESTER操作示例
    example_input2 = get_text_from_pdf('data\\pdf\\cropped_TomcatGuide_split\\015_1 - Pilot Pre-Start.pdf')
    with open('data\\json\\PJ_example.json', 'r',encoding='utf-8') as file:
        data = json.load(file)
    example_output2 = json.dumps(data, indent=4,ensure_ascii=False)
    #需要结合上下文找出单人操作示例
    example_input3 = get_text_from_pdf("data\\pdf\\cropped_TomcatGuide_split\\055_2.4.4 - Pulse Doppler STT (Single Target Track) Lock.pdf")
    with open('data\\json\\single_example.json', 'r',encoding='utf-8') as file:
        data = json.load(file)
    example_output3 = json.dumps(data, indent=4,ensure_ascii=False)

    tmp1 = tmp1.format(input1 = example_input1, output1 = example_output1,
                       input2 = example_input2, output2 = example_output2,
                       input3 = example_input3, output3 = example_output3)

    text = get_text_from_pdf(pdf_path)
    tmp2 = tmp2.format(content = text)

    client,model = CLIENT("deepseek")
    chat_completion = client.chat.completions.create(
        messages=[
            {
                "role": "system",
                "content": tmp1,
            },
            {
                "role":"user",
                "content": tmp2
            }
        ],
        model=model,
        response_format={"type": "json_object"},
    )

    llm_output = syntax_correction(chat_completion.choices[0].message.content)
    if llm_output == '{}':#if output is empty
        logger.info(f'[E]No subtasks found in {pdf_path}')
        return
    
    #show the output
    # print(llm_output)

    #load the output into json and check the format
    loop_cnt = 0
    while llm_output is not None:
        if loop_cnt > 5:
            logger.error(f"ERROR in {pdf_path}, Can't get correct format in limit times. Please check the format of the output manually:\n{llm_output}")
            return
        if loop_cnt > 0:
            logger.warning(f"WARN in {pdf_path}, Can't get correct format in {loop_cnt} times from:\n{llm_output}\nBUG INFO:\n{ve}\n")
            print(f"trying {loop_cnt} times to get correct format")
        try:
            new_object = json.loads(llm_output)
        except json.JSONDecodeError:
            logger.error(f"ERROR in {pdf_path}, Can't load into json from:\n{llm_output}")
            return
        llm_output,ve = format_correction(new_object)
        loop_cnt += 1

    #save the output to subtasks.json
    with open(save_path, 'r', encoding='utf-8') as file:
        data = json.load(file)

    #repeat check
    for obj in data:
        if obj['name'] == new_object['name'] and obj['type'] == new_object['type']:
            logger.info(f'[R]Subtask from {pdf_path} already exists in the database')
            return
        
    data.append(new_object)
    with open(save_path, 'w', encoding='utf-8') as file:
        json.dump(data, file, indent=4, ensure_ascii=False)
    logger.info(f'[S]Subtasks from {pdf_path} have been saved')
    return

def get_subtasks_from_pdf(folder_path,save_path,log_path,target_list = None):
    pdf_cnt = 0
    #init logger
    logger = init_logger("debug for subtasks generate",console_level=logging.INFO)

    for filename in os.listdir(folder_path):
        if filename.endswith('.pdf'):
            pdf_cnt += 1
            if target_list is not None and pdf_cnt not in target_list:#allow user to choose target pdf manually
                print(f"skip {pdf_cnt}th pdf")
                continue
            pdf_path = folder_path + filename
            get_subtask_from_pdf(pdf_path,save_path,logger)
            # print(pdf_cnt)

def get_state_list_from_pdf():
    tmp ="""
    You are a veteran fighter pilot, you need to classify the instruments and switches in the cockpit according to their functions or characteristics,
    so that you can quickly query the status of all the instruments and switches required for a certain type of operation,
    Please note that the JSON content you output should be brief, without any explanatory content, and you don't need to explain the function of each meter switch.
    this is the list of instrument switches in the back cockpit(RIO) of the F-14, please convert the output to json format:
    <content>
    {content}
    </content>
    <example>
    "category": "TACAN",
    "description": "TACAN(Tactical Air Navigation) related information",
    "items":[
        "ALR-67 knob",
        "SW knob",
        "V/UHF 2 knob",
        "TACAN CMD switch",
        "Dual rotary switch",
        "GO & NO-GO lights",
        "BIT button",
        "MODE switches",
        "VOL knob",
        "Mode knob"]
    </example>
    """
    # print(output)
    output = get_text_from_pdf('data\\pdf\\TomcatManual_split\\cropped_pilot.pdf')
    tmp = tmp.format(content = output)

    with open('data\\txt\\test.txt', 'w', encoding='utf-8') as file:
        file.write(tmp)

def show_subtasks_names(file_name,silent=True,with_tag=None):
    with open(file_name, 'r',encoding='utf-8') as file:
            data = json.load(file)
    
    # 如果with_tag不为None且是文件路径，读取tag文件
    tagged_tasks = []
    if with_tag is not None:
        try:
            with open(with_tag, 'r', encoding='utf-8') as tag_file:
                tagged_tasks = json.load(tag_file)
        except Exception as e:
            print(f"Error loading tag file: {e}")
    
    RIO_TASKS = []
    PILOT_TASKS = []
    PR_TASKS = []
    PJ_TASKS = []
    
    for obj in data:
        task_name = obj['name']
        # 如果任务名在tag文件中，则跳过
        if task_name in tagged_tasks:
            continue
        param = obj.get('param', '')
        if param:
            task_name += f'({param})'
                        
        if obj['agents'] == 'RIO':
            RIO_TASKS.append(task_name)
        elif obj['agents'] == 'PILOT':
            PILOT_TASKS.append(task_name)
        elif obj['agents'] == 'PILOT and RIO':
            PR_TASKS.append(task_name)
        elif obj['agents'] == 'PILOT and JESTER':
            PJ_TASKS.append(task_name)
            
    if not silent:
        print("**********************RIO_TASKS:**********************")
        for i in RIO_TASKS:
            print(i)
        print("**********************PILOT_TASKS:**********************")
        for i in PILOT_TASKS:
            print(i)
        print("**********************PR_TASKS:**********************")
        for i in PR_TASKS:  
            print(i)
        print("**********************PJ_TASKS:**********************")
        for i in PJ_TASKS:
            print(i)
            
    return RIO_TASKS,PILOT_TASKS,PR_TASKS

def rename_subtasks(file_path,save_path):
    tmp = '''
    You are a fighter assist system, and you need to assist two pilots, PILOT and RIO.
    There are currently some records of the operation of two pilots, which are used as sub-plans for the subsequent development of the battle plan, but the titles in these records are not clear, or are not suitable for the development of the battle plan, please help me rewrite this title to make this set of operations sound like an operation that can complete a certain mission. Of course, if you think the title is already clear, you can also leave it alone.
    You only need to output the modified name, you don't need to output any additional information, if you think you don't need to change it, please output the original name, be careful not to lose the information
    Here's the json file you need to try to modify, with the title "name" indicated:
    {content}
    '''

    # 读取JSON文件
    with open(file_path, 'r', encoding='utf-8') as file:
        data = json.load(file)

    # 处理每个对象
    for obj in data:
        # 使用simple_llm函数生成新的name
        new_name = simple_llm("deepseek",tmp.format(content = json.dumps(obj, indent=4,ensure_ascii=False))).strip('"\'')
        print(f"change old name: {obj['name']} to new one: {new_name}")
        obj['name'] = new_name  # 更新对象的'name'键

    # 将更新后的数据写回文件
    with open(save_path, 'w', encoding='utf-8') as file:
        json.dump(data, file, indent=4, ensure_ascii=False)

def describe_subtasks(file_path,save_path):
    tmp="""
    You are a veteran fighter pilot, and I will show you some of the operations of the two pilots on the F14 fighter jet, and please give a general description of what they operated, so that I can directly know what these operations involve and what objectives are accomplished,
    Make sure your answers are concise and don't generate superfluous content
    Example session:
    <input>
    {example_json}
    </input>
    <output>The action involves the use of the LANTIRN (Low Altitude Navigation and Targeting Infrared for Night) pod, which provides advanced targeting and navigation capabilities, particularly in low-light or night-time conditions. The GBU-12 is a laser-guided bomb, and the LANTIRN pod is used to designate and track the target with a laser, guiding the bomb to its target.</output>

    Here is the operations you need to overview:
    <operations>
    {op}
    </operations>
    """
    with open(file_path, 'r',encoding='utf-8') as file:
        data = json.load(file)
        for obj in data:
                if obj["name"] == "GBU-12 LANTIRN-Guided Bombing Mission Execution Plan" and obj["type"] == "PILOT and RIO":
                    content = json.dumps(obj, indent=4,ensure_ascii=False).strip()
        for obj in data:
            if obj["name"] == "PRE-FLIGHT PREPARATION AND START-UP PROCEDURE" and obj["type"] == "PILOT and JESTER":
                tmp = tmp.format(example_json=content,op=json.dumps(obj, indent=4,ensure_ascii=False))
                # print(tmp)
                descri = simple_llm("deepseek",content=tmp)
                obj["description"] = descri
                print(obj["name"],descri)
    # with open(save_path, 'w', encoding='utf-8') as file:
    #     json.dump(data, file, indent=4, ensure_ascii=False)

def describe_data_category(file_path,output_file):
    tmp = '''This is a certain type of data for F14 fighters, and you need to summarize this data category very briefly:
{content}
And finally output should be like <category>: <describe>format,
in description, you need to describe what this data category is first, and then describe the data contained in this category,
for example:
**AHRS**: Attitude and Heading Reference System, providing aircraft orientation and heading data. 
**DDD**:Dual Digital Displays, providing radar and navigation data.
**Weapon Panel**:RIO Weapon Panel, providing weapon related data.
'''
    summaries = []
    cnt = 0
    with open(file_path, 'r',encoding='utf-8') as file:
        data = json.load(file)
        for obj in data:
            summ = simple_llm("volces-deepseek",tmp.format(content = json.dumps(obj, indent=4,ensure_ascii=False)))
            summaries.append(summ)
            cnt = cnt + 1
            print(f"{cnt}: {summ}")
    print(summaries)
    # with open(output_file, 'w', encoding='utf-8') as file:
    #     for summ in summaries:
    #         file.write(summ + '\n')

def describe_ops(file_path, chosen_category=[], output_file=None, silent=False,redescript_path=None):
    # 读取JSON文件
    cnt = 0
    if redescript_path is not None and os.path.exists(redescript_path):
        with open(redescript_path, 'r', encoding='utf-8') as file:
            redescript_data = json.load(file)
    else:
        print(f"no redescript file find in {redescript_path}")

    if not isinstance(file_path,list):
        file_path = [file_path]
    data = merge_json_dicts(file_path)

    if chosen_category == []:
        for cate,_ in data.items():
            chosen_category.append(cate)
        pass
    
    # 如果chosen_category不是列表，则将其转换为列表
    if not isinstance(chosen_category, list):
        chosen_category = [chosen_category]

    descriptions = []
    # 遍历所选类别
    for category in chosen_category:
        # 获取当前类别的数据
        category_data = data.get(category, {})
        
        # 存储当前类别下所有对象及其描述
        descriptions.append(f"***{category}***")
        if not silent:
            print(f"***{category}***")
        # 遍历当前类别下的所有对象
        for obj, item in category_data.items():
            if redescript_path is not None:
                description = redescript_data.get(obj, '')
            else:
                description = item.get('description', '')
            # print(item['inputs'])
            # max_value = item['outputs'][0].get('max_value',None)
            descriptions.append(f"{obj}: {description} ")#max_value:{max_value}")
            
            # 如果不静默模式，则打印描述
            if not silent:
                cnt += 1
                print(f"{cnt} {obj}: {description}")# max_value:{max_value}")
        
        # 将当前类别的描述添加到结果字典中
        results = descriptions
    
    # 如果指定了输出文件，则将结果写入文件
    if output_file is not None:
        with open(output_file, 'w', encoding='utf-8') as out_f:
            json.dump(results, out_f, ensure_ascii=False, indent=4)
    return results

def filter_panel_data(file_path, output_file=None):
    with open(file_path, 'r', encoding='utf-8') as file:
        data = json.load(file)
    # 遍历数据并收集需要删除的键
    to_delete = []
    for category, items in data.items():
        for name, subitem in items.items():
            outputs = subitem.get('outputs', [])
            # 如果outputs为空或者任何一个output的最大值为65535，则标记该项为待删除
            if any(output.get('max_value') == 65535 for output in outputs) or not outputs:
                to_delete.append((category, name))
    # 删除标记的项
    for category, name in to_delete:
        del data[category][name]
    # 如果没有提供输出文件路径，则覆盖原文件；否则写入新的文件
    output_file = output_file if output_file else file_path
    with open(output_file, 'w', encoding='utf-8') as file:
        json.dump(data, file, indent=2, ensure_ascii=False)

def find_empty_category(file_path):
    with open(file_path, 'r', encoding='utf-8') as file:
        data = json.load(file)
    for category, items in data.items():
        if items:
            print(category)
            pass
        else:
            print(category)
            pass

def show_api_category(file_path):
    with open(file_path, 'r', encoding='utf-8') as file:
        data = json.load(file)
    api_category = []
    for category, items in data.items():
        for name, item in items.items():
            if item["outputs"][0]["type"] == "string":
                api_category.append(name)
    print(api_category)

def list_all_hid_devices():
    """列出所有连接的 HID 设备详细信息"""
    print("所有连接的 HID 设备:")
    devices = hid.enumerate()
    
    if not devices:
        print("未找到 HID 设备")
        return
    
    for i, device in enumerate(devices):
        print(f"{i+1}. 厂商ID: 0x{device['vendor_id']:04x}, 产品ID: 0x{device['product_id']:04x}")
        print(f"   产品名称: {device.get('product_string', 'N/A')}")
        print(f"   制造商: {device.get('manufacturer_string', 'N/A')}")
        print(f"   接口编号: {device.get('interface_number', 'N/A')}")
        print(f"   使用类型: {device.get('usage_page', 'N/A')}, {device.get('usage', 'N/A')}")
        print(f"   路径: {device['path']}")
        print("-----------------------------------")

def load_prefix_task(task_name, file_path="data/json/evaluation/task/task.json"):
    try:
        with open(file_path, 'r', encoding='utf-8') as f:
            tasks = json.load(f)
        
        for task in tasks:
            if task.get('task') == task_name:
                return task
        
        print(f"警告: 在預設任務文件中找不到任務 '{task_name}'")
        return None
    except FileNotFoundError:
        print(f"警告: 找不到預設任務文件 {file_path}")
        return None
    except json.JSONDecodeError as e:
        print(f"警告: 預設任務文件格式錯誤: {e}")
        return None


if __name__ == "__main__":
    pass
    # filter_panel_data("data\\json\\dcs\\F-14.json","data\\json\\dcs\\F-14_filtered.json") 
    # show_api_category("data\json\dcs\oringin\F-14.json")
    # st= "RIO Indicator Lights"
    # st = "Screen"
    st = "Joystick"
    # describe_ops("data\json\dcs\wo_geague\F-14_wog.json",chosen_category=st.split(','))
    # describe_ops(file_path=["data\json\dcs\wo_geague\F-14_wog_re.json",
    #                         "data\json\dcs\patch\F-14_patch.json"],
    #              chosen_category=st.split(','),
    #              redescript_path="data\\json\\dcs\\data_redescript\\F-14_wog_redescript.json")
    
    # describe_ops(file_path=["data\\json\\dcs\\oringin\\CommonData.json"],
    #              chosen_category=[],
    #              redescript_path="data\\json\\dcs\\data_redescript\\F-14_wog_redescript.json")
    # find_empty_category("data\\json\\dcs\\wo_geague\\F-14_wog_re.json")

    # describe_data_category("data\json\dcs\\wo_geague\F-14_wog_re.json",output_file="data\\txt\F-14_category_summ_V3.txt")
    # describe_data_category("data\json\dcs\CommonData.json",output_file="data\\txt\CommonData_category_summ.txt")
    # describe_data_category("data\json\dcs\patch\F-14_patch.json",output_file="data\\txt\F-14_category_summ_V3.txt")

    # crop_pdf('.\data\pdf\F-14B Manual 1.0_compressed.pdf',
    #          save_path='.\data\pdf\cropped_F-14B Manual 1.0_compressed.pdf')
    # text = get_text_from_pdf('.\data\pdf\cropped_F-14B Manual 1.0_compressed.pdf')
    # with open('data/txt/cropped_TomcatManual.txt', 'w', encoding='utf-8') as file:
    #     file.write(text)

    # split_pdf(pdf_file_path='data\\pdf\\cropped_TomcatGuide_compressed.pdf',save_path='data\\pdf\\cropped_TomcatGuide_split')
    # get_toc('data\pdf\F-14B Manual 1.0_compressed.pdf')

    # get_subtask_from_pdf(pdf_path = 'data\\pdf\\cropped_TomcatGuide_split\\1_Disclaimer.pdf',
    #                      save_path = 'data\\json\\test.json',)
    # tg_list = [15,16,17,18,19,20,21,22,23,24,
    #            44,48,55,68,71,73,76,
    #            91,92,93,94,97,98,99,
    #            100,101,102,103,104,105,106,107,108,
    #            124,126,127,128,
    #            141,142,143,144,145,146,151,154,157,161,162,163,164]
    # get_subtasks_from_pdf(folder_path = 'data\\pdf\\cropped_TomcatGuide_split\\',
    #                       save_path = 'data\\json\\subtasksV2.json',
    #                       log_path='log\\test.log',
    #                       target_list=tg_list)
    # rename_subtasks(file_path='data\\json\\subtasks.json',
    #                 save_path='data\\json\\renamed_subtasks.json')
    show_subtasks_names('data\\json\\subtasksV3.json',silent=False,with_tag="data\\json\\Atag.json")
    #
    # describe_subtasks(file_path='data\\json\\renamed_subtasks.json',
    #                   save_path='data\\json\\renamed_describe_subtasks.json',)
    # print("if you look at this, you are a genius")

    # list_all_hid_devices() 
    pass