import logging
import sys
from time import sleep

import pandas as pd
from web3 import Web3

log = logging.getLogger(__name__)


def process_transaction(tx, w3):
    if (
        # tx['to'] and
        # tx['to'].lower() == '0xA51afAFe0263b40EdaEf0Df8781eA9aa03E381a3'.lower() and
        tx['input'].startswith(bytes.fromhex('3593564c')) and
        (receipt := w3.eth.get_transaction_receipt(tx.hash))['status'] == 1
    ):
        return receipt['gasUsed']
    return None


if __name__ == '__main__':
    if len(sys.argv) != 2:
        print('Usage: python uniswap4_gas.py <rpc_url>')
    rpc_url = sys.argv[1]
    logging.basicConfig(level=logging.INFO)
    log.setLevel(logging.DEBUG)

    num_blocks = 10000
    w3 = Web3(Web3.HTTPProvider(rpc_url, request_kwargs={'timeout':15}))
    gas_data = pd.DataFrame(columns=['gas'])
    end_block = w3.eth.block_number
    start_block = end_block - num_blocks

    try:
        # Process events in pages of 2000 blocks
        PAGE_SIZE = 2000
        events = []

        for block in range(start_block, end_block + 1, PAGE_SIZE):
            current_end = min(block + PAGE_SIZE - 1, end_block)
            log.info(f'Querying events for blocks {block} to {current_end}...')

            event_filter = w3.eth.filter({
                'fromBlock': block,
                'toBlock': current_end,
                # 'address': '0x011f31D20C8778c8Beb1093b73E3A5690Ee6271b',
                'topics': ['0x40e9cecb9f5f1f1c5b9c97dec2917b7ee92e57ba5563708daca94dd84ad7112f'],
            })

            page_events = event_filter.get_all_entries()
            events.extend(page_events)
            log.info(f'Found {len(page_events)} events in this page')
            sleep(0.5)  # Add small delay between requests

        for event in events:
            try:
                log.info(f"Processing transaction from block {event['blockNumber']}")
                tx = w3.eth.get_transaction(event['transactionHash'])
                gas_used = process_transaction(tx, w3)

                if gas_used:
                    new_data = pd.DataFrame({'gas': [gas_used]})
                    gas_data = pd.concat([gas_data, new_data], ignore_index=True)
                    log.info(f"Transaction {event['transactionHash'].hex()}: Gas used {gas_used}")

            except Exception as e:
                log.error(f"Error processing transaction: {str(e)}")
            finally:
                sleep(0.2)

    finally:
        gas_data.to_csv('gas_data.csv', index=False)