Add retry logic with exponential backoff for Telegram network exceptions

[Copilot]

The bot crashes on Telegram network exceptions without recovery. Added automatic retry with exponential backoff for all network operations.

Changes

Retry Decorator (services/utils.py)

• Created async_retry() decorator: 3 attempts, 1s→2s→4s exponential backoff
• Configurable exception types and timing parameters
• Logs retry attempts via logging module

HTTP API Calls (services/api.py)

• Applied retry to all 28 API functions
• Catches NetworkError, TimeoutException, HTTPError
• Added explicit 30s timeout to all AsyncClient instances
• Created retry-wrapped bot helpers: _retry_bot_send_message(), _retry_bot_edit_message_text()

Bot Polling (main.py)

• Automatic restart on TelegramNetworkError
• Retry initialization (webhook deletion, command setup)
• Infinite recovery loop with 5s delay between restart attempts

Notification Handlers (handlers/participant/notification.py)

• Replaced direct bot calls with retry-wrapped versions

Example

@async_retry(max_attempts=3, delay=1.0, backoff=2.0, 
             exceptions=(NetworkError, TimeoutException, HTTPError))
async def get_all_events(count, chat_id):
    async with AsyncClient(timeout=30.0) as client:
        response = await client.get(url, headers=headers)
        return response.json() if response.status_code == 200 else None

Network errors now trigger automatic retry (1s, 2s, 4s delays) instead of crashing. Bot polling restarts automatically on Telegram API failures.

Original prompt

i usually get telegram network exception in bot. fix it. i suppose you should add retries

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.