Polars EVM adds the evm namespace to polars on dataframes, lazyframes, series, and expressions.
This namespace has lots of functions for processing EVM data:
- binary ↔ hex conversions
- binary → float conversions (
u256,i256, etc) - event decoding
- transaction decoding
- keccak
pip install polars_evm
Just import polars_evm and then the evm namespace will be registered to polars.
import polars as pl
import polars_evm
addresses = [
b'\xda\xc1\x7f\x95\x8d.\xe5#\xa2 b\x06\x99E\x97\xc1=\x83\x1e\xc7',
b'\xa0\xb8i\x91\xc6!\x8b6\xc1\xd1\x9dJ.\x9e\xb0\xce6\x06\xebH',
b'_\x98\x80ZN\x8b\xe2U\xa3(\x80\xfd\xec\x7fg(\xc6V\x8b\xa0'
]
balances = [
b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00Jy\xb0\x9aq\x1e\xd1(",
b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xd1\xff\xf7\xfb\xa8O\x87",
b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01cEx]\x8a\x00\x00",
]
df = pl.DataFrame({'address': addresses, 'balance': balances})
print('perform conversions on dataframe:')
print(df.evm.binary_to_float({'balance': 'u256'}, replace=True).evm.binary_to_hex())
print()
print('perform conversions using expressions:')
print(df.select(pl.col.address.evm.binary_to_hex(), pl.col.balance.evm.binary_to_float('u256')))
print()
print('perform binary to hex conversion on series:')
print('hex series:', df['address'].evm.binary_to_hex())
print()
print('perform binary to float conversion on series:')
print('float series:', df['balance'].evm.binary_to_float('u256'))output:
perform conversions on dataframe:
shape: (3, 2)
┌────────────────────────────────────────────┬───────────┐
│ address ┆ balance │
│ --- ┆ --- │
│ str ┆ f64 │
╞════════════════════════════════════════════╪═══════════╡
│ 0xdac17f958d2ee523a2206206994597c13d831ec7 ┆ 5.3665e18 │
│ 0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48 ┆ 1.3117e17 │
│ 0x5f98805a4e8be255a32880fdec7f6728c6568ba0 ┆ 1.0000e17 │
└────────────────────────────────────────────┴───────────┘
perform conversions using expressions:
shape: (3, 2)
┌────────────────────────────────────────────┬───────────┐
│ address ┆ balance │
│ --- ┆ --- │
│ str ┆ f64 │
╞════════════════════════════════════════════╪═══════════╡
│ 0xdac17f958d2ee523a2206206994597c13d831ec7 ┆ 5.3665e18 │
│ 0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48 ┆ 1.3117e17 │
│ 0x5f98805a4e8be255a32880fdec7f6728c6568ba0 ┆ 1.0000e17 │
└────────────────────────────────────────────┴───────────┘
perform binary to hex conversion on series:
hex series: shape: (3,)
Series: 'address' [str]
[
"0xdac17f958d2ee523a2206206994597c13d831ec7"
"0xa0b86991c6218b36c1d19d4a2e9eb0ce3606eb48"
"0x5f98805a4e8be255a32880fdec7f6728c6568ba0"
]
perform binary to float conversion on series:
float series: shape: (3,)
Series: 'balance' [f64]
[
5.3665e18
1.3117e17
1.0000e17
]# DataFrame namespace
df.evm.binary_to_hex(prefix=True, columns=None)
df.evm.hex_to_binary(prefix=True, columns=None)
df.evm.binary_to_float({'column1': 'u256', 'column2': 'i256'}, replace=False, prefix=True)
df.evm.filter_binary(column1_name=hex_or_bytes, column2_name=list_of_values)
df.evm.decode_events(event_abi)
df.evm.decode_contract_events(event_abi)
df.evm.decode_transactions(function_abi_or_contract_abi)
# LazyFrame namespace
lf.evm.binary_to_hex(prefix=True, columns=None)
lf.evm.hex_to_binary(prefix=True, columns=None)
lf.evm.binary_to_float({'column1': 'u256', 'column2': 'i256'}, replace=False, prefix=True)
lf.evm.filter_binary(column1_name=hex_or_bytes, column2_name=list_of_values)
lf.evm.decode_events(event_abi)
# Series namespace
series.evm.binary_to_hex(prefix=True)
series.evm.hex_to_binary(prefix=True)
series.evm.binary_to_float('u256')
series.evm.keccak(output='hex', text=False)
# Expression namespace
pl.Expr.evm.binary_to_hex(prefix=True)
pl.Expr.evm.hex_to_binary(prefix=True)
pl.Expr.binary_to_float('u256')
pl.Expr.evm.keccak(output='hex', text=False)Beyond the evm namespace, polars_evm has the following utilities:
-
set_column_display_width(): set display width so that it fully displays tx hashes in jupyter notebooks and other printouts
