chat_indicate/main.py

import re
import telebot
from config import Config
from ai_assistant import AI_assistant
import pdfplumber
import time
# 使用Token來初始化一個telebot.TeleBot物件，並將其儲存到bot這個變數中

BOT_TOKEN = '6701395239:AAFE30dqvNihDdni9vYoAbWssO-X5yAmwho'
# BOT_TOKEN = "6746720034:AAEMaoV2FwIZ8pz_PF18-bo2a6gFC1eVtVs"
#BOT_TOKEN = '6589162555:AAHGhrTQ0wYNtIUySMohnfpxQl1d6blr24Q'
bot = telebot.TeleBot(BOT_TOKEN)

user_url = None
user_answer = None




# 定義消息處理器，當收到網址時
# 考慮到medium也有提供短網址，所以暫不把qffers判斷寫進去
@bot.message_handler(func=lambda msg: re.search(r'http[s]?://(www\.)?(link\.)?medium\.com/', msg.text) if msg.text else False)
def handle_medium_url(message):
    # bot.reply_to(message, "想了解什麼訊息呢?問問我吧！")  # 這邊改到上面不確定會不會比較好

    global user_url, identify
    user_url = message.text
    #contents, lang, identify = get_contents(user_url)
    #save_to_storage(contents, identify)
    bot.reply_to(message, "想了解什麼訊息呢?問問我吧！")

@bot.message_handler(func=lambda msg: re.search(r'http[s]?://', msg.text) if msg.text else False)
def handle_other_url(message):
    bot.reply_to(message, "此網頁不支援唷😂😂\n請試試看輸入 https://link.medium.com/rxe98Z708Db ", disable_web_page_preview=True)



# 定義消息處理器，當用戶發送一個消息時，機器人將回覆相同的消息
@bot.message_handler(func=lambda msg: msg.text.lower() in ["hi", "hello", "嗨", "你好", "早上好", "晚上好", "早安", "晚安", "介紹", "誰"])
def reply_all(message):
    print(message) #你可以出來看看會有啥
    print(message.text) #單純擷取文字
    user_first_name = message.from_user.first_name
    intro =  f'嗨, {user_first_name}！👋👋\n我們是睿富者（QFFERS）\n歡迎到我們的Medium文章看看~ https://link.medium.com/rxe98Z708Db \n \n也歡迎試著貼上Medium文章網址問個問題吧！😂😂\n'
    result = start()
    print(result)
    bot.reply_to(message, result, disable_web_page_preview=True)

@bot.message_handler(func=lambda msg: True)
def handle_user_answer(message):
    global user_answer , startime
    user_answer = message.text
    # result = answering(user_answer)

    result = start()
    print("This is result", result)
    end =time.time()
    print("Time: ", end-startime)
    bot.reply_to(message, result)

# Using assistant API to answer question
def answering(query):
    files = ai_assistant.get_files()
    ai_assistant.create_assistant(
        name="QFFERS Bot",
        instructions="你是一個天問Bot機器人，你的任務是請基於用戶上傳的PDF上找尋用戶所要找尋的答案、數值。"
                     "任務說明:用戶提問時，請仔細分析問題並提供基於上傳PDF。如果答案來自PDF檔案請提供該篇PDF的段落，若沒有資料請回答:我不知道",
        tools=[{"type": "retrieval"}],
        files=files
    )
    ai_assistant.create_thread()
    ai_assistant.add_message_to_thread(
        role="user",
        content=query
    )
    ai_assistant.run_assistant(
        instructions="Please user's language to answer the question. You can only answer according to the uploaded files.")
    ai_assistant.check_run(thread_id=ai_assistant.thread.id, run_id=ai_assistant.run.id)
    total_price , content = ai_assistant.process_messages()
    return content

def start():
    global startime
    startime = time.time()
    url = "https://www.bea.gov/news/2024/personal-income-and-outlays-december-2023"
    text = find_indicate(url)
    return text
def find_indicate_bs4(url):
    from bs4 import BeautifulSoup
    import requests
    response = requests.get(url)
    soup = BeautifulSoup(response.text, 'html.parser')

    # 使用CSS選擇器找到元素
    element = soup.select("tr",class_ ='item-fact-row' )

    return element[1].text.split('+')[0]+" "+element[1].text.split('+')[1]
def find_indicate(url):
    from selenium import webdriver
    from selenium.webdriver.common.by import By
    from selenium.webdriver.chrome.options import Options
    import time
    options = Options()
    # options.add_argument("--headless")  # 啟用無頭模式
    driver =webdriver.Chrome(options=options)
    driver.get(url)
    # time.sleep(3)
    chat = driver.find_element(By.XPATH, '//*[@id="home"]/div[2]/div/div/div[1]/table/tbody/tr[13]/td[6]')
    return chat.text

# 是一個持續運行的迴圈，不斷從Telegram伺服器抓取新的消息
# 然後使用上面定義的消息處理器來處理這些消息。

def tg_bot(cfg: Config):
    """Run the Telegram bot."""
    global ai_assistant, storage ,bot
    print("Starting Telegram bot...")
    # ai_assistant = AI_assistant(cfg)
    # 啟動Telegram Bot
    bot.infinity_polling()
#非農就業人數
def read_pdf_nonfarm(month, year):
    pdf = pdfplumber.open(f"empsit/empsit_{month}_{year}.pdf")
    page = pdf.pages[0]
    text = page.extract_text().split('\n')
    text = (text[7]+text[8]).split(',')
    text = text[0]+text[1]+text[2]
    return text
def read_nonfarm():
    startimee = time.time()
    for i in range(7,13):
        print(f"2022年{i}月非農就業人數: ", end= "" )
        print(read_pdf_nonfarm(i, 23))

    endtimee = time.time()
    print("Time: ", endtimee-startimee)
    # print(text.split('\n')[7:9])
if __name__ == "__main__":
    #非農
    startimee = time.time()
    print(f"2023年7月非農就業人數: ", end= "" )
    print(read_pdf_nonfarm(7, 23))
    endtimee = time.time()
    print("Time_NonFarm: ", endtimee-startimee)
    startimee = time.time()
    print(find_indicate_bs4("https://www.bea.gov/data/personal-consumption-expenditures-price-index"))
    endtimee = time.time()
    print("Time_NonFarm: ", endtimee-startimee)
    # cfg = Config()
    # tg_bot(cfg)