substreams

README

Substreams - A streaming data engine for The Graph - by StreamingFast

DEVELOPER PREVIEW OF SUBSTREAMS

Think Fluvio for deterministic blockchain data.

The successor of https://github.com/streamingfast/sparkle, enabling greater composability, yet similar powers of parallelisation, and a much simpler model to work with.

Install client

This client will allow you to interact with Substreams endpoints, and stream data in real-time.x

Get a release.

From source:

git clone git@github.com:streamingfast/substreams
cd substreams
go install -v ./cmd/substreams

From source without checkout:

go install github.com/streamingfast/substreams/cmd/substreams@latest

Install dependencies to build Substreams

This will allow you to develop Substreams modules locally, and run them remotely.

Install rust

We're going to be using the Rust programming language, to develop some custom logic.

There are several ways to install Rust, but for the sake of brevity:

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh

Install wasm-pack

Finally, wasm-pack is a tool which allows us to work wi th Rust-generated WASM code.

Their quicktart docs can be found here, but again:

curl https://rustwasm.github.io/wasm-pack/installer/init.sh -sSf | sh

Run remotely

Using StreamingFast's infrastructure

Dump that somewhere like .bashrc:

export STREAMINGFAST_KEY=server_YOUR_KEY_HERE  # Ask us on Discord for a key
function sftoken {
    export FIREHOSE_API_TOKEN=$(curl https://auth.dfuse.io/v1/auth/issue -s --data-binary '{"api_key":"'$STREAMINGFAST_KEY'"}' | jq -r .token)
	export SUBSTREAMS_API_TOKEN=$FIREHOSE_API_TOKEN
    echo Token set on FIREHOSE_API_TOKEN and SUBSTREAMS_API_TOKEN
}

Then in your shell, load a key in an env var with:

sftoken

Then, try to run the PancakeSwap Substreams from our Substreams Playground project

The below commands will be run from substreams-playground

cd ./pcs-rust/ && ./build.sh
cd ../eth-token/ && ./build.sh
cd ..
substreams run -e bsc-dev.streamingfast.io:443 ./pcs-rust/substreams.yaml pairs,block_to_pairs,volumes,totals,db_out -s 6810706 -t 6810711

Run locally

You can run the substreams service locally this way:

Get a recent release of the Ethereum Firehose, and install sfeth.

Alternatively, you can use this Docker image: ghcr.io/streamingfast/sf-ethereum:6aa70ca, known to work with version v0.0.5-beta of the substreams release herein.

Get some data (merged blocks) to play with locally (here on BSC mainnet):

# Downloads 2.6GB of data
sfeth tools download-from-firehose bsc-dev.streamingfast.io:443 6810000 6820000 ./localblocks
sfeth tools generate-irreversible-index ./localblocks ./localirr 6810000 6819700

Then run the firehose service locally in a terminal, reading blocks from your disk:

sfeth start firehose  --config-file=  --log-to-file=false  --common-blockstream-addr=  --common-blocks-store-url=./localdata --firehose-grpc-listen-addr=:9000* --substreams-enabled --substreams-rpc-endpoint=https://URL.POINTING.TO.A.BSC.ARCHIVE.NODE/if-you/want-to-use/eth_call/within/substreams

And then run the substreams command against your local deployment (checkout substreams-playground in the Run remotely section above):

substreams run -k -e localhost:9000 wasm_substreams_manifest.yaml pairs,block_to_pairs,db_out,volumes,totals -s 6810706 -t 6810711

Documentation

Index

• type BlockHook
• type PostJobHook
• type ReturnFunc

Types

type BlockHook

type BlockHook func(ctx context.Context, clock *pbsubstreams.Clock) error

type PostJobHook

type PostJobHook func(ctx context.Context, clock *pbsubstreams.Clock) error

type ReturnFunc

type ReturnFunc func(any *pbsubstreams.BlockScopedData) error