Skip to main content

Python SCALE Codec Library

Project description

Python SCALE Codec

Build Status Latest Version Supported Python versions License

Python SCALE Codec Library

Description

Most of the data that the Substrate RPCs output is encoded with the SCALE Codec. This codec is used by the Substrate nodes' internal runtime. In order to get to meaningful data this data will need to be decoded. The Python SCALE Codec Library will specialize in this task.

Documentation

https://polkascan.github.io/py-scale-codec/

Installation

pip install scalecodec

Examples (MetadataV14 runtimes and higher)

Encode a Call

runtime_config = RuntimeConfigurationObject()
# This types are all hardcoded types needed to decode metadata types
runtime_config.update_type_registry(load_type_registry_preset(name="metadata_types"))

# Decode retrieved metadata from the RPC
metadata = runtime_config.create_scale_object(
    'MetadataVersioned', data=ScaleBytes(response.get('result'))
)
metadata.decode()

# Add the embedded type registry to the runtime config
runtime_config.add_portable_registry(metadata)

call = runtime_config.create_scale_object(
    "Call", metadata=metadata
)
call.encode({
    "call_module": "Balances",
    "call_function": "transfer",
    "call_args": {"dest": "5GNJqTPyNqANBkUVMN1LPPrxXnFouWXoe2wNSmmEoLctxiZY", "value": 3},
})

Decode the result of a state_getStorageAt RPC call

event_data = "0x2000000000000000b0338609000000000200000001000000000080b2e60e0000000002000000020000000003be1957935299d0be2f35b8856751feab95fc7089239366b52b72ca98249b94300000020000000500be1957935299d0be2f35b8856751feab95fc7089239366b52b72ca98249b943000264d2823000000000000000000000000000200000005027a9650a6bd43f1e0b4546affb88f8c14213e1fb60512692c2b39fbfcfc56b703be1957935299d0be2f35b8856751feab95fc7089239366b52b72ca98249b943000264d2823000000000000000000000000000200000013060c4c700700000000000000000000000000000200000005047b8441d5110c178c29be709793a41d73ae8b3119a971b18fbd20945ea5d622f00313dc01000000000000000000000000000002000000000010016b0b00000000000000"

system_pallet = [p for p in metadata.pallets if p['name'] == 'System'][0]
event_storage_function = [s for s in system_pallet['storage']['entries'] if s['name'] == "Events"][0]


event = runtime_config.create_scale_object(
    event_storage_function.get_value_type_string(), data=ScaleBytes(event_data), metadata=metadata
)
print(event.decode())

Examples (prior to MetadataV14)

Decode a SCALE-encoded Compact<Balance>

RuntimeConfiguration().update_type_registry(load_type_registry_preset("default"))
RuntimeConfiguration().update_type_registry(load_type_registry_preset("kusama"))
obj = RuntimeConfiguration().create_scale_object('Compact<Balance>', data=ScaleBytes("0x130080cd103d71bc22"))
obj.decode()
print(obj.value)

Encode to Compact<Balance>

RuntimeConfiguration().update_type_registry(load_type_registry_preset("default"))
obj = RuntimeConfiguration().create_scale_object('Compact<Balance>')
scale_data = obj.encode(2503000000000000000)
print(scale_data)

Encode to Vec<Bytes>

RuntimeConfiguration().update_type_registry(load_type_registry_preset("default"))
value = ['test', 'vec']
obj = RuntimeConfiguration().create_scale_object('Vec<Bytes>')
scale_data = obj.encode(value)
print(scale_data)

Add custom types to type registry

RuntimeConfiguration().update_type_registry(load_type_registry_preset("default"))

custom_types = {
    "types": {
        "MyCustomType": "u32",
        "CustomNextAuthority": {
          "type": "struct",
          "type_mapping": [
             ["AuthorityId", "AuthorityId"],
             ["weight", "AuthorityWeight"]
          ]
        }
    }   
}

RuntimeConfiguration().update_type_registry(custom_types)

Or from a custom JSON file

RuntimeConfiguration().update_type_registry(load_type_registry_preset("default"))
RuntimeConfiguration().update_type_registry(load_type_registry_file("/path/to/type_registry.json"))

Multiple runtime configurations

By default a singleton is used to maintain the configuration, for multiple instances:

# Kusama runtime config
runtime_config_kusama = RuntimeConfigurationObject()
runtime_config_kusama.update_type_registry(load_type_registry_preset("default"))
runtime_config_kusama.update_type_registry(load_type_registry_preset("kusama"))


# Polkadot runtime config
runtime_config_polkadot = RuntimeConfigurationObject()
runtime_config_polkadot.update_type_registry(load_type_registry_preset("default"))
runtime_config_polkadot.update_type_registry(load_type_registry_preset("polkadot"))

# Decode extrinsic using Kusama runtime configuration
extrinsic = runtime_config_kusama.create_scale_object(
    type_string='Extrinsic', 
    metadata=metadata_decoder
)
extrinsic.decode(ScaleBytes(extrinsic_data))

License

https://github.com/polkascan/py-scale-codec/blob/master/LICENSE

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

scalecodec-1.0.30.tar.gz (125.0 kB view details)

Uploaded Source

Built Distribution

scalecodec-1.0.30-py3-none-any.whl (74.8 kB view details)

Uploaded Python 3

File details

Details for the file scalecodec-1.0.30.tar.gz.

File metadata

  • Download URL: scalecodec-1.0.30.tar.gz
  • Upload date:
  • Size: 125.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.7.1 importlib_metadata/4.10.1 pkginfo/1.8.2 requests/2.27.1 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.10.2

File hashes

Hashes for scalecodec-1.0.30.tar.gz
Algorithm Hash digest
SHA256 0b3c7e9ab8825f5247835f304e5dedf67a9b4122df4025d625ebdfc9487b84ad
MD5 d8a48d0404d0da1ae90ad6b9b9ed5afa
BLAKE2b-256 0b38650d7dc31f3a68c920cbca2580b08392b21f7fec21f229ed1e25715401fe

See more details on using hashes here.

File details

Details for the file scalecodec-1.0.30-py3-none-any.whl.

File metadata

  • Download URL: scalecodec-1.0.30-py3-none-any.whl
  • Upload date:
  • Size: 74.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.7.1 importlib_metadata/4.10.1 pkginfo/1.8.2 requests/2.27.1 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.10.2

File hashes

Hashes for scalecodec-1.0.30-py3-none-any.whl
Algorithm Hash digest
SHA256 c35c72dbdc75d6fb4a91db9f3506492adee7a86c22bb12e46ec413003e11c697
MD5 7958d3eae34b2ee7e7d031f113f0b47b
BLAKE2b-256 5a15deb13e2ab8ed0e3300e1eb1de70946ab82f9abd601933f9b34ff6f39a36a

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page