Links
Comment on page

Python

Creating a Spice Function in Python

Function Handlers

In the function.yaml , code execution runtime and invocation handler is defined.
Example function.yaml for a Python function
# hello_world/function.yaml
output_datasets:
- {orgName}.{appName}.hello_world
# This will trigger the function to execute on every new Ethereum block.
triggers:
- path: eth
# This selects the runtime that will execute your code.
runtime: python3.11
# For Python, handler is [script_name].[method]
# The below will invoke a method `process` in a file `spice_function.py`.
handler: spice_function.process
Replace {orgName} and {appName} with the values for your specific org and app.
See Spice Functions YAML Specification for the full YAML schema.

Function Handlers

A Python invocation handler method has the following signature:
def process(context: dict,
duckdb: duckdb.DuckDBPyConnection,
spice_client: spicepy.Client):
  • context: A dictionary with context variables, E.g. block_number and block_hash
  • duckdb: A connection to the Spice Function's DuckDB instance.
  • spice_client: A spicepy client that is pre-initialized with the Spice app's API Key.
The handler method will be called on each triggered event.

Dependencies

Python dependencies can be specified by placing a requirements.txt adjacent to the code importing the dependencies.
The Python function runtime comes pre-loaded with these modules:
Currently, Python modules that depend on compiled C extensions are not supported to be defined in requirements.txt. A request for a particular missing dependency may be made on Discord.

Example Function

An example Python function that will produce the same outputs shown above:
# hello_world/spice_function.py
import duckdb
import spicepy
def process(context: dict,
duckdb: duckdb.DuckDBPyConnection,
spice_client: spicepy.Client):
duckdb.sql(f"INSERT INTO output.hello_world VALUES ({context['block_number']}, 'Hello!')")
print("Hello, World!")
Last modified 1mo ago