Refactor Telegram bot to fetch hot Solana pairs - #37

main.py (3)

119-119: Wrap polling in __main__ guard for testability.

Calling bot.infinity_polling() at module level means the bot starts immediately on import, preventing unit testing or reuse of the module's functions.

♻️ Proposed fix

-bot.infinity_polling()
+if __name__ == "__main__":
+    bot.infinity_polling()

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@main.py` at line 119, The module currently calls bot.infinity_polling() at
import time which prevents importing for tests; move that call into a guarded
entrypoint: create a main() function that calls bot.infinity_polling() and wrap
its invocation in if __name__ == "__main__": main(). Update any startup logic
tied to bot.infinity_polling() accordingly so tests can import functions/classes
without starting the bot automatically.

---

82-91: Consider batching messages to avoid potential rate limits.

Sending up to 5 individual messages in rapid succession could trigger Telegram's rate limiting. Consider combining results into a single formatted message.

♻️ Proposed batched message approach

-    for idx, token in enumerate(results, start=1):
-        msg = (
-            f"#{idx} {token['name']} ({token['symbol']})\n"
-            f"Price: ${token['price']}\n"
-            f"Liquidity: ${token['liquidity']:,.0f}\n"
-            f"5m Volume: ${token['volume_5m']:,.0f}\n"
-            f"5m Change: {token['price_change_5m']:.2f}%\n"
-            f"Chart: {token['url']}"
-        )
-        bot.send_message(message.chat.id, msg)
+    lines = []
+    for idx, token in enumerate(results, start=1):
+        lines.append(
+            f"#{idx} {token['name']} ({token['symbol']})\n"
+            f"💰 Price: ${token['price']}\n"
+            f"💧 Liquidity: ${token['liquidity']:,.0f}\n"
+            f"📊 5m Volume: ${token['volume_5m']:,.0f}\n"
+            f"📈 5m Change: {token['price_change_5m']:.2f}%\n"
+            f"🔗 {token['url']}"
+        )
+    bot.send_message(message.chat.id, "\n\n".join(lines))

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@main.py` around lines 82 - 91, The loop sending individual messages (for idx,
token in enumerate(results, start=1): ... bot.send_message(message.chat.id,
msg)) can hit Telegram rate limits; instead build a single formatted string by
iterating over results and appending each token's formatted block (include idx,
token['name'], token['symbol'], price, liquidity, volume_5m, price_change_5m,
url) into one message body, then call bot.send_message once to message.chat.id;
ensure you respect Telegram max message length (chunk into multiple batched
messages if the combined string exceeds the API limit) and reuse the same send
operation rather than multiple sends in the loop.

---

42-43: Silent exception swallowing hinders debugging.

Catching all exceptions and continuing without any logging makes it difficult to diagnose issues with malformed pair data. Consider logging the exception or at least the pair identifier that failed parsing.

♻️ Proposed fix to add logging

+import logging
+
+logger = logging.getLogger(__name__)

Then in the exception handler:

             except Exception:
-                continue
+                logger.debug("Skipping malformed pair data: %s", p.get("pairAddress", "unknown"))
+                continue

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@main.py` around lines 42 - 43, Replace the silent catch-all "except
Exception:" in main.py with "except Exception as e:" and log the error and the
related pair identifier before continuing; e.g., call an existing logger
(logger.exception or logger.error) with a message that includes the pair
variable name used in the parsing loop (e.g., pair or pair_id) and the exception
(or stack trace); if no logger exists, write the same information to stderr so
malformed pair data failures are visible while still continuing.