# Testing polyfill-rs
This document describes how to run tests for polyfill-rs, with a focus on integration tests that verify our client can actually communicate with the real Polymarket API.
## Test Types
### Unit Tests
- **Location**: Scattered throughout source files (`src/*.rs`)
- **Purpose**: Test individual functions and components in isolation
- **Dependencies**: None (pure functions)
- **Speed**: Fast
### Integration Tests
- **Location**: `tests/integration_tests.rs`
- **Purpose**: Verify the client can communicate with the real Polymarket API
- **Dependencies**: Network connectivity + credentials (tests are `#[ignore]` by default)
- **Speed**: Slower (network calls)
## Running Tests
### Quick Start (Basic Tests)
```bash
# Run unit tests + doc tests (real-API tests are `#[ignore]` by default)
cargo test --all-features
# Run the "no-alloc hot paths" regression tests
cargo test --all-features --test no_alloc_hot_paths
# Compile-check all examples
cargo build --examples
```
### Full Integration Testing
#### 1. Set up Environment Variables
Create a `.env` file or export variables:
```bash
# Required for ignored real-API tests
export POLYMARKET_PRIVATE_KEY="your_private_key_here"
# Optional: API credentials (some tests/tools may use these)
export POLYMARKET_API_KEY="your_api_key"
export POLYMARKET_API_SECRET="your_api_secret"
export POLYMARKET_API_PASSPHRASE="your_passphrase"
# Backward-compatible aliases also accepted by test helpers
export POLYMARKET_SECRET="your_api_secret"
export POLYMARKET_PASSPHRASE="your_passphrase"
# Optional (defaults provided in test helpers)
export POLYMARKET_HOST="https://clob.polymarket.com"
export POLYMARKET_CHAIN_ID="137"
# Optional, but required when funds live in a Polymarket proxy/Safe wallet.
# 0 = EOA, 1 = Proxy, 2 = Gnosis Safe/browser wallet, 3 = Poly1271.
export POLYMARKET_SIGNATURE_TYPE="2"
# Optional: override the derived funder/proxy wallet address.
export POLYMARKET_FUNDER_ADDRESS="0x..."
# Optional: pin clob.polymarket.com to a known A record if local DNS is blocked.
# Keep the hostname in POLYMARKET_HOST so HTTPS/SNI still validates.
export POLYMARKET_RESOLVE_IP="104.18.34.205"
```
#### 2. Run Integration Tests
```bash
# Using the test runner script (runs ignored tests that hit the real API)
./scripts/run_integration_tests.sh
# Or directly with cargo
cargo test --all-features --test integration_tests -- --ignored --nocapture --test-threads=1
```
## Test Categories
### Always Run (No Auth Required)
- **API Connectivity**: Basic connection to Polymarket API
- **Market Data Endpoints**: Order book, prices, spreads, etc.
- **Error Handling**: Invalid requests and error responses
- **Rate Limiting**: Multiple rapid requests
- **API Compatibility**: Verify our API matches polymarket-rs-client
- **Performance**: Response time measurements
### Authentication Required
- **Authentication**: API key creation and validation
- **Advanced Client Features**: Full client configuration
- **WebSocket Connectivity**: Real-time data streaming
### API Credentials Required
- **Order Management**: Order creation and management (read-only tests)
## Test Results
### Success Indicators
```
API connectivity test passed
Market data endpoints test passed
Error handling test passed
Rate limiting test passed
API compatibility test passed
Performance test passed
Server time: 234ms
Markets request: 1.2s
Markets returned: 50
```
### Ignored Indicators (default)
```
test test_real_api_* ... ignored
```
### Failure Indicators
```
API connectivity test failed: Network error: connection refused
Market data endpoints test failed: API error (404): Token not found
```
## Performance Benchmarks
Our integration tests include performance measurements:
| Server Time | < 5s | 234ms |
| Markets Request | < 10s | 1.2s |
| Order Book | < 5s | 890ms |
| Price Quote | < 3s | 156ms |
## Troubleshooting
### Common Issues
#### Network Connectivity
```bash
# Test basic connectivity
curl -I https://clob.polymarket.com/
# Check DNS resolution
nslookup clob.polymarket.com
# If local DNS fails but outbound HTTPS works, get a current A record and pin it:
curl -sS -H 'accept:application/dns-json' \
'https://cloudflare-dns.com/dns-query?name=clob.polymarket.com&type=A'
export POLYMARKET_RESOLVE_IP="104.18.34.205"
```
#### Authentication Issues
```bash
# Verify private key format
# Run a small ignored auth smoke-test (requires real credentials)
cargo test --all-features --test simple_auth_test -- --ignored --nocapture --test-threads=1
```
#### Rate Limiting
```bash
# If tests fail due to rate limiting, consider adding delays between manual runs.
```
### Debug Mode
Run tests with detailed logging:
```bash
# Enable debug logging
RUST_LOG=debug cargo test --all-features --test integration_tests -- --ignored --nocapture --test-threads=1
# Enable trace logging for maximum detail
RUST_LOG=trace cargo test --all-features --test integration_tests -- --ignored --nocapture --test-threads=1
```
## Continuous Integration
### GitHub Actions
Our CI runs formatting, clippy, unit tests, docs, security audit, and a separate no-alloc job. Real-API integration tests are `#[ignore]` and are not run in CI.
```yaml
# .github/workflows/ci.yml
- name: Run tests (excluding no-alloc hot paths)
run: cargo test --all-features -- --skip no_alloc_
- name: Run no-alloc hot path tests
run: cargo test --all-features --test no_alloc_hot_paths
```
### Local CI
Run the same tests locally:
```bash
# Install cargo-nextest for faster test execution
cargo install cargo-nextest
# Run with nextest (ignored tests are not run by default)
cargo nextest run --all-features
```
## Test Coverage
Our integration tests cover:
- **API Endpoints**: All major REST endpoints
- **Authentication**: EIP-712 signing and API key management
- **Error Handling**: Network errors, API errors, validation errors
- **Performance**: Response time and throughput measurements
- **WebSocket**: Real-time data streaming (when available)
- **Compatibility**: API compatibility with polymarket-rs-client
## Adding New Tests
### Template for New Integration Test
```rust
#[tokio::test]
async fn test_new_feature() -> Result<()> {
let config = TestConfig::from_env();
// Skip if requirements not met
if !config.has_auth() {
TestReporter::skip("test_new_feature", "no private key");
return Ok(());
}
// Test implementation
let client = config.create_auth_client()?;
let result = client.some_new_method().await?;
// Assertions
assert!(result.is_valid());
TestReporter::success("test_new_feature");
Ok(())
}
```
### Best Practices
1. **Use TestConfig**: Always use the shared test configuration
2. **Handle Missing Credentials**: Skip tests gracefully when credentials aren't available
3. **Measure Performance**: Include timing measurements for performance-critical operations
4. **Provide Context**: Use descriptive test names and error messages
5. **Clean Up**: Don't leave test data in the system
## Security Notes
- **Never commit credentials**: All test credentials are loaded from environment variables
- **Use test accounts**: If testing with real credentials, use dedicated test accounts
- **Read-only tests**: Order management tests only create orders, they don't execute them
- **Rate limiting**: Tests include delays to respect API rate limits