Got it, let's walk through building this inline keyboard flow step by step. I'll use the python-telegram-bot library (version 13.x or later—callback handlers work smoothly here) to make this work as you described.

First, Setup the Basic Bot Structure #

First, install the library if you haven't already:

pip install python-telegram-bot==13.7

Then, start with the core bot setup, including your main menu (I'll include a quick example of the custom keyboard you mentioned):

from telegram import Update, InlineKeyboardButton, InlineKeyboardMarkup, ReplyKeyboardMarkup
from telegram.ext import Updater, CommandHandler, CallbackQueryHandler, MessageHandler, Filters, CallbackContext

# Your main menu custom keyboard
def get_main_menu():
    keyboard = [['Feedback']]
    return ReplyKeyboardMarkup(keyboard, resize_keyboard=True)

# Start command to load the main menu
def start(update: Update, context: CallbackContext):
    update.message.reply_text("Welcome to the main menu!", reply_markup=get_main_menu())

Handle the Feedback Button Click #

Next, we need a handler that triggers when the user taps the "Feedback" button in the main menu. This will send the first inline keyboard with your question:

def handle_feedback_trigger(update: Update, context: CallbackContext):
    # Build the first inline keyboard
    inline_keyboard = [
        [InlineKeyboardButton("No", callback_data='feedback_no')],
        [InlineKeyboardButton("Yes", callback_data='feedback_yes')],
        [InlineKeyboardButton("Leave Comments", callback_data='feedback_comments')]
    ]
    reply_markup = InlineKeyboardMarkup(inline_keyboard)
    # Send the question with the inline keyboard attached
    update.message.reply_text("Message: Do you find this app useful?", reply_markup=reply_markup)

Handle Inline Button Callback Queries #

Now, we need to catch the user's selections from the inline keyboard. We'll use a CallbackQueryHandler to process each button press:

def handle_feedback_actions(update: Update, context: CallbackContext):
    query = update.callback_query
    query.answer()  # Always acknowledge the callback to avoid loading spinners for the user

    if query.data == 'feedback_no':
        # Send the thank-you message and return to main menu
        query.edit_message_text("Thank you and we hope you can leave some comments how to improve the app")
        query.message.reply_text("Back to main menu:", reply_markup=get_main_menu())
    
    elif query.data == 'feedback_yes':
        # Show the second inline keyboard (I filled in your partial prompt)
        vote_keyboard = [
            [InlineKeyboardButton("⭐ 1 Star", callback_data='vote_1')],
            [InlineKeyboardButton("⭐⭐ 2 Stars", callback_data='vote_2')],
            [InlineKeyboardButton("⭐⭐⭐ 3 Stars", callback_data='vote_3')],
            [InlineKeyboardButton("⭐⭐⭐⭐ 4 Stars", callback_data='vote_4')],
            [InlineKeyboardButton("⭐⭐⭐⭐⭐ 5 Stars", callback_data='vote_5')]
        ]
        reply_markup = InlineKeyboardMarkup(vote_keyboard)
        query.edit_message_text("Please vote for our app:", reply_markup=reply_markup)
    
    elif query.data == 'feedback_comments':
        # Prompt user to enter their comments
        query.edit_message_text("Please type your comments below:")
        # Track that we're waiting for a comment using user data
        context.user_data['awaiting_comment'] = True

# Optional: Capture and handle the user's comment
def handle_comment_input(update: Update, context: CallbackContext):
    if context.user_data.get('awaiting_comment'):
        comment = update.message.text
        # Here you can save the comment to a database, log it, etc.
        update.message.reply_text("Thank you for your valuable feedback!", reply_markup=get_main_menu())
        context.user_data['awaiting_comment'] = False

Wire Up All Handlers and Launch the Bot #

Finally, register all handlers and start the bot:

def main():
    # Replace with your bot token from @BotFather
    updater = Updater("YOUR_BOT_TOKEN_HERE")
    dp = updater.dispatcher

    # Register all handlers
    dp.add_handler(CommandHandler("start", start))
    dp.add_handler(MessageHandler(Filters.regex('^Feedback$'), handle_feedback_trigger))
    dp.add_handler(CallbackQueryHandler(handle_feedback_actions))
    dp.add_handler(MessageHandler(Filters.text & ~Filters.command, handle_comment_input))

    # Start the bot
    updater.start_polling()
    updater.idle()

if __name__ == '__main__':
    main()

Quick Tips #

• Always call query.answer() when handling callbacks—it’s required to let Telegram know the button press was processed.
• Using edit_message_text replaces the existing message instead of sending a new one, keeping the chat thread clean.
• The context.user_data dict is a simple way to track conversation state (like waiting for a comment) between interactions.

You can extend the vote handling by adding more callback cases for the star buttons, following the same pattern as the initial feedback options.

内容的提问来源于stack exchange，提问作者David Tang