Substreams Testing Expert (SVM)

Expert assistant for testing Solana/SVM Substreams applications.

Testing Philosophy

1. Test early and often — Catch issues in development, not production
2. Test with real data — Solana edge cases are numerous (failed txs, inner instructions, CPI)
3. Test at multiple scales — Single blocks, small ranges, and large datasets
4. Performance testing — Solana's high throughput (400ms blocks, thousands of txs) demands it

Unit Tests

Testing IDL Decoders

#[cfg(test)]
mod tests {
    use super::*;
    use substreams_solana_idls::raydium;

    #[test]
    fn test_unpack_swap_base_in() {
        // Known instruction data for a SwapBaseIn
        let instruction_data: Vec<u8> = vec![/* raw bytes */];

        let result = raydium::amm::v4::instructions::unpack(&instruction_data);
        assert!(result.is_ok());

        match result.unwrap() {
            raydium::amm::v4::instructions::RaydiumV4Instruction::SwapBaseIn(swap) => {
                assert!(swap.amount_in > 0);
                assert!(swap.minimum_amount_out > 0);
            }
            _ => panic!("Expected SwapBaseIn instruction"),
        }
    }

    #[test]
    fn test_unpack_invalid_data() {
        let invalid_data: Vec<u8> = vec![0xFF, 0xFF];
        let result = raydium::amm::v4::instructions::unpack(&invalid_data);
        assert!(result.is_err());
    }
}

Testing Helper Functions

#[cfg(test)]
mod tests {
    use super::*;
    use substreams_solana::base58;

    #[test]
    fn test_base58_encoding() {
        let bytes = vec![0u8; 32];
        let encoded = base58::encode(&bytes);
        assert_eq!(encoded, "11111111111111111111111111111111");
    }

    #[test]
    fn test_to_global_sequence() {
        let clock = Clock {
            number: 100,
            ..Default::default()
        };
        let seq = to_global_sequence(&clock, 5);
        assert_eq!(seq, (100 << 32) + 5);
    }

    #[test]
    fn test_common_key_v2() {
        let clock = Clock {
            id: "abc123".to_string(),
            ..Default::default()
        };
        let keys = common_key_v2(&clock, 0, 1);
        assert_eq!(keys[0], ("block_hash", "abc123".to_string()));
        assert_eq!(keys[1], ("transaction_index", "0".to_string()));
        assert_eq!(keys[2], ("instruction_index", "1".to_string()));
    }
}

Testing DatabaseChanges Output

#[cfg(test)]
mod tests {
    use super::*;
    use substreams_database_change::tables::Tables;

    #[test]
    fn test_db_row_creation() {
        let mut tables = Tables::new();
        tables.create_row("raydium_amm_v4_swap", [
            ("block_hash", "test_hash".to_string()),
            ("transaction_index", "0".to_string()),
            ("instruction_index", "0".to_string()),
        ])
        .set("signature", "test_sig")
        .set("fee", 5000u64);

        let changes = tables.to_database_changes();
        assert_eq!(changes.table_changes.len(), 1);
        assert_eq!(changes.table_changes[0].table, "raydium_amm_v4_swap");
    }
}

Integration Testing

Running Against Live Data

# Test a single DEX module with a small block range
substreams run ./raydium/amm-v4/substreams.yaml \
  map_events \
  --start-block 250000000 \
  --stop-block +100

# Test the aggregate db_out module
substreams run ./db-svm-dex/substreams.yaml \
  db_out \
  --start-block 250000000 \
  --stop-block +100

Verifying Output

# Output as JSON for inspection
substreams run ./raydium/amm-v4/substreams.yaml \
  map_events \
  --start-block 250000000 \
  --stop-block +10 \
  --output json

Build Verification

Full Build Test

# Build all workspace crates
cargo build --target wasm32-unknown-unknown --release

# Run all unit tests (native target)
cargo test

# Build specific crate
cargo build --target wasm32-unknown-unknown --release -p raydium-amm-v4

Schema Validation

For ClickHouse sinks with split schema files, verify schemas concatenate correctly:

cd db-svm-dex-clickhouse
make schema  # Concatenates split files into schema.sql

Performance Testing

Block Range Recommendations

Solana-Specific Considerations

• Vote transactions: Use blockswithoutvotes to exclude ~80% of transactions
• High instruction count: Some Solana blocks have thousands of instructions
• CPI depth: Inner instructions (CPI) can be deeply nested
• Failed transactions: Always verify failed transaction handling

Common Issues

Build Failures

# Missing wasm target
rustup target add wasm32-unknown-unknown

# Outdated dependencies
cargo update

# Protobuf generation issues
cd proto && make protogen

Runtime Issues

• Empty output: Check program ID filter matches expected program
• Missing instructions: Verify walk_instructions() covers inner instructions (CPI)
• Base58 encoding: Ensure addresses are base58-encoded, not hex
• Failed txs in output: Add is_failed() check if you want to skip them