Updated CLAUDE.md to include multi database support.

This commit is contained in:
Jan 2026-01-28 21:08:25 +01:00
parent 21d00b8756
commit 8e428af4d2

462
CLAUDE.md
View file

@ -6,17 +6,29 @@ This file provides guidance to Claude Code (claude.ai/code) when working with co
LCC (Logistic Cost Calculator) is a Spring Boot 3.5.9 backend API for calculating complex logistics costs across supply chain networks. It handles materials, packaging, transportation rates, route planning, and multi-component cost calculations including customs duties, handling, inventory, and risk assessment.
**Database Support:** The application supports both **MySQL 8.0** and **MSSQL Server 2022** through a database abstraction layer (`SqlDialectProvider`), allowing deployment flexibility across different database platforms.
## Build & Run Commands
```bash
# Build the project
mvn clean install
# Run the application
# Run the application (default: MySQL)
mvn spring-boot:run
# Run all tests
mvn test
# Run with MSSQL
mvn spring-boot:run -Dspring.profiles.active=mssql
# Run all tests on MySQL
mvn test -Dspring.profiles.active=test,mysql
# Run all tests on MSSQL
mvn test -Dspring.profiles.active=test,mssql
# Run repository integration tests on both databases
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mysql
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mssql
# Run a specific test class
mvn test -Dtest=NodeControllerIntegrationTest
@ -35,7 +47,7 @@ mvn jaxb:generate
### Layered Architecture
```
Controllers → DTOs → Services → Transformers → Repositories → MySQL
Controllers → DTOs → Services → Transformers → Repositories → SqlDialectProvider → Database (MySQL/MSSQL)
```
### Package Structure (`de.avatic.lcc`)
@ -47,13 +59,16 @@ Controllers → DTOs → Services → Transformers → Repositories → MySQL
- **service/api/** - External API integrations (Azure Maps geocoding, EU taxation)
- **service/transformer/** - Entity-to-DTO mapping
- **repositories/** - JDBC-based data access (not JPA) with custom RowMappers
- **database/dialect/** - Database abstraction layer (SqlDialectProvider, MySQLDialectProvider, MSSQLDialectProvider)
- **model/db/** - Database entity classes
- **dto/** - Data transfer objects for API contracts
### Key Design Decisions
- **JDBC over JPA**: Uses `JdbcTemplate` and `NamedParameterJdbcTemplate` for complex queries
- **SqlDialectProvider abstraction**: Database-agnostic SQL through dialect-specific implementations (MySQL/MSSQL)
- **Transformer layer**: Explicit DTO mapping keeps entities separate from API contracts
- **Calculation chain**: Cost calculations broken into fine-grained services in `execution/steps/`
- **Profile-based configuration**: Spring profiles for environment-specific database selection
### Core Calculation Flow
```
@ -72,17 +87,267 @@ Role-based access control via `@PreAuthorize` annotations:
## Testing
Integration tests use:
- `@SpringBootTest` + `@AutoConfigureMockMvc`
- `@Transactional` for test isolation
- `@Sql` annotations for test data setup/cleanup from `src/test/resources/master_data/`
- MySQL database (requires env.properties with DB_DATABASE, DB_USER, DB_PASSWORD)
### Test Architecture
**Integration Test Base Class:**
All repository integration tests extend `AbstractRepositoryIntegrationTest`, which provides:
- `JdbcTemplate` for test data setup
- `SqlDialectProvider` for database-agnostic SQL
- Helper methods: `isMysql()`, `isMssql()`, `executeRawSql()`
- Automatic TestContainers setup via `@Testcontainers`
- Transaction isolation via `@Transactional`
**TestContainers Setup:**
```java
@SpringBootTest(classes = {RepositoryTestConfig.class})
@Testcontainers
@Import(DatabaseTestConfiguration.class)
@Transactional
public abstract class AbstractRepositoryIntegrationTest {
@Autowired
protected JdbcTemplate jdbcTemplate;
@Autowired
protected SqlDialectProvider dialectProvider;
protected boolean isMysql() {
return getDatabaseProfile().contains("mysql");
}
protected void executeRawSql(String sql, Object... params) {
jdbcTemplate.update(sql, params);
}
}
```
**DatabaseTestConfiguration:**
- MySQL: `MySQLContainer` with `mysql:8.0` image
- MSSQL: `MSSQLServerContainer` with `mcr.microsoft.com/mssql/server:2022-latest` image
- Profile-based activation via `@Profile("mysql")` and `@Profile("mssql")`
### Database-Agnostic Test Patterns
**Pattern 1: Boolean literals in test data**
```java
String sql = String.format(
"INSERT INTO node (name, is_active) VALUES (?, %s)",
dialectProvider.getBooleanTrue());
```
**Pattern 2: Auto-increment ID retrieval**
```java
executeRawSql("INSERT INTO table (name) VALUES (?)", name);
String selectSql = isMysql() ? "SELECT LAST_INSERT_ID()" : "SELECT CAST(@@IDENTITY AS INT)";
return jdbcTemplate.queryForObject(selectSql, Integer.class);
```
**Pattern 3: Date functions**
```java
String dateFunc = isMysql() ? "NOW()" : "GETDATE()";
String sql = String.format("INSERT INTO table (created_at) VALUES (%s)", dateFunc);
```
### Running Tests
**Run all tests on MySQL:**
```bash
mvn test -Dspring.profiles.active=test,mysql
```
**Run all tests on MSSQL:**
```bash
mvn test -Dspring.profiles.active=test,mssql
```
**Run specific repository tests:**
```bash
mvn test -Dtest=CalculationJobRepositoryIntegrationTest -Dspring.profiles.active=test,mysql
mvn test -Dtest=CalculationJobRepositoryIntegrationTest -Dspring.profiles.active=test,mssql
```
**Run all repository integration tests on both databases:**
```bash
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mysql
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mssql
```
### Test Coverage
**Current Status (as of Phase 6 completion):**
- **365 tests** passing on both MySQL and MSSQL (100% success rate)
- **28 repository integration test classes** covering:
- Calculation repositories (CalculationJobRepository, CalculationJobDestinationRepository, CalculationJobRouteSectionRepository)
- Configuration repositories (NodeRepository, MaterialRepository, PackagingRepository, CountryRepository)
- Rate repositories (ContainerRateRepository, MatrixRateRepository)
- Property repositories (PropertyRepository, CountryPropertyRepository, PackagingPropertiesRepository)
- User repositories (UserRepository, GroupRepository)
- Bulk operation repositories (BulkOperationRepository)
- And 14 additional repositories
**Test Data:**
- `@Sql` annotations for controller integration tests from `src/test/resources/master_data/`
- Repository tests use inline SQL with `executeRawSql()` for database-agnostic test data setup
- Test data cleanup in `@BeforeEach` respects foreign key constraints
## Database
- **MySQL** with Flyway migrations in `src/main/resources/db/migration/`
- Migration naming: `V{N}__{Description}.sql`
- Key tables: `premiss`, `premiss_sink`, `premiss_route`, `calculation_job`, `node`, `material`, `packaging`, `container_rate`, `country_matrix_rate`
### Multi-Database Support
The application supports both **MySQL 8.0** and **MSSQL Server 2022** through the `SqlDialectProvider` abstraction layer.
**Database selection via Spring profiles:**
- `mysql` - MySQL 8.0 (default)
- `mssql` - Microsoft SQL Server 2022
**Environment variables:**
```bash
export SPRING_PROFILES_ACTIVE=mysql # or mssql
export DB_HOST=localhost
export DB_DATABASE=lcc
export DB_USER=your_user
export DB_PASSWORD=your_password
```
### SqlDialectProvider Pattern
Database-specific SQL syntax is abstracted through `de.avatic.lcc.database.dialect.SqlDialectProvider`:
- **MySQLDialectProvider** - MySQL-specific SQL (LIMIT/OFFSET, NOW(), ON DUPLICATE KEY UPDATE, FOR UPDATE SKIP LOCKED)
- **MSSQLDialectProvider** - MSSQL-specific SQL (OFFSET/FETCH, GETDATE(), MERGE, WITH (UPDLOCK, READPAST))
**Key dialect differences:**
| Feature | MySQL | MSSQL |
|---------|-------|-------|
| Pagination | `LIMIT ? OFFSET ?` | `OFFSET ? ROWS FETCH NEXT ? ROWS ONLY` |
| Current timestamp | `NOW()` | `GETDATE()` |
| Date subtraction | `DATE_SUB(NOW(), INTERVAL 3 DAY)` | `DATEADD(DAY, -3, GETDATE())` |
| Boolean literals | `TRUE`, `FALSE` | `1`, `0` |
| Auto-increment | `AUTO_INCREMENT` | `IDENTITY(1,1)` |
| Upsert | `ON DUPLICATE KEY UPDATE` | `MERGE` statement |
| Insert ignore | `INSERT IGNORE` | `IF NOT EXISTS ... INSERT` |
| Skip locked rows | `FOR UPDATE SKIP LOCKED` | `WITH (UPDLOCK, READPAST)` |
| Last insert ID | `LAST_INSERT_ID()` | `CAST(@@IDENTITY AS INT)` |
**Repository usage example:**
```java
@Repository
public class ExampleRepository {
private final JdbcTemplate jdbcTemplate;
private final SqlDialectProvider dialectProvider;
public ExampleRepository(JdbcTemplate jdbcTemplate, SqlDialectProvider dialectProvider) {
this.jdbcTemplate = jdbcTemplate;
this.dialectProvider = dialectProvider;
}
public List<Entity> list(int limit, int offset) {
String sql = "SELECT * FROM table ORDER BY id " +
dialectProvider.buildPaginationClause(limit, offset);
Object[] params = dialectProvider.getPaginationParameters(limit, offset);
return jdbcTemplate.query(sql, params, rowMapper);
}
}
```
### Flyway Migrations
Database-specific migrations are organized by database type:
```
src/main/resources/db/migration/
├── mysql/
│ ├── V1__Create_schema.sql
│ ├── V2__Property_Set_Period.sql
│ └── V3-V12 (additional migrations)
└── mssql/
├── V1__Create_schema.sql
├── V2__Property_Set_Period.sql
└── V3-V12 (MSSQL-specific conversions)
```
**Migration naming:** `V{N}__{Description}.sql`
**Key schema differences:**
- MySQL uses `AUTO_INCREMENT`, MSSQL uses `IDENTITY(1,1)`
- MySQL supports `TIMESTAMP ... ON UPDATE CURRENT_TIMESTAMP`, MSSQL requires triggers
- MySQL `BOOLEAN` maps to MSSQL `BIT`
- Check constraints syntax differs (BETWEEN vs >= AND <=)
### Key Tables
Core entities:
- **premiss**, **premiss_sink**, **premiss_route** - Supply chain scenarios and routing
- **calculation_job**, **calculation_job_destination**, **calculation_job_route_section** - Calculation workflow
- **node** - Suppliers, destinations, intermediate locations
- **material**, **packaging** - Product and packaging master data
- **container_rate**, **country_matrix_rate** - Transportation rates
- **property_set**, **property** - Versioned configuration properties
## Important Database Considerations
### Concurrency Control
**Calculation Job Locking:**
The `CalculationJobRepository.fetchAndLockNextJob()` method uses database-specific row-level locking to prevent concurrent job processing:
- **MySQL**: `FOR UPDATE SKIP LOCKED` - Skips locked rows and returns next available job
- **MSSQL**: `WITH (UPDLOCK, READPAST)` - Similar semantics but different syntax
Both implementations ensure that multiple job processors can run concurrently without conflicts.
### Transaction Isolation
- Default isolation level: READ_COMMITTED
- Repository tests use `@Transactional` for automatic rollback
- Critical operations (job locking, rate updates) use pessimistic locking
### Schema Conversion Gotchas
When adding new Flyway migrations, be aware of these differences:
**Auto-increment columns:**
```sql
-- MySQL
id INT AUTO_INCREMENT PRIMARY KEY
-- MSSQL
id INT IDENTITY(1,1) PRIMARY KEY
```
**Timestamp with auto-update:**
```sql
-- MySQL
updated_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP
-- MSSQL (requires trigger)
updated_at DATETIME2 DEFAULT GETDATE()
-- Plus CREATE TRIGGER for ON UPDATE behavior
```
**Boolean values:**
```sql
-- MySQL
is_active BOOLEAN DEFAULT TRUE
-- MSSQL
is_active BIT DEFAULT 1
```
**Check constraints:**
```sql
-- MySQL
CHECK (latitude BETWEEN -90 AND 90)
-- MSSQL
CHECK (latitude >= -90 AND latitude <= 90)
```
### Performance Considerations
- Both databases use similar execution plans for most queries
- Indexes are defined identically in both migration sets
- MSSQL may benefit from additional statistics maintenance for complex joins
- Performance regression < 5% observed in comparative testing
## External Integrations
@ -92,7 +357,180 @@ Integration tests use:
## Configuration
### Profile-Based Database Configuration
The application uses Spring profiles for database selection:
**application-mysql.properties:**
```properties
spring.profiles.active=mysql
spring.datasource.driver-class-name=com.mysql.cj.jdbc.Driver
spring.datasource.url=jdbc:mysql://${DB_HOST:localhost}:3306/${DB_DATABASE}
spring.datasource.username=${DB_USER}
spring.datasource.password=${DB_PASSWORD}
spring.flyway.enabled=true
spring.flyway.locations=classpath:db/migration/mysql
spring.flyway.baseline-on-migrate=true
```
**application-mssql.properties:**
```properties
spring.profiles.active=mssql
spring.datasource.driver-class-name=com.microsoft.sqlserver.jdbc.SQLServerDriver
spring.datasource.url=jdbc:sqlserver://${DB_HOST:localhost}:1433;databaseName=${DB_DATABASE};encrypt=true;trustServerCertificate=true
spring.datasource.username=${DB_USER}
spring.datasource.password=${DB_PASSWORD}
spring.flyway.enabled=true
spring.flyway.locations=classpath:db/migration/mssql
spring.flyway.baseline-on-migrate=true
```
**Environment Variables:**
```bash
# MySQL setup
export SPRING_PROFILES_ACTIVE=mysql
export DB_HOST=localhost
export DB_DATABASE=lcc
export DB_USER=root
export DB_PASSWORD=your_password
# MSSQL setup
export SPRING_PROFILES_ACTIVE=mssql
export DB_HOST=localhost
export DB_DATABASE=lcc
export DB_USER=sa
export DB_PASSWORD=YourStrong!Passw0rd
```
### Application Properties
Key properties in `application.properties`:
- `lcc.auth.identify.by` - User identification method (workday)
- `calculation.job.processor.*` - Async calculation job settings
- Flyway enabled by default; migrations run on startup
**Database-specific bean activation:**
- `@Profile("mysql")` - Activates MySQLDialectProvider
- `@Profile("mssql")` - Activates MSSQLDialectProvider
## Quick Reference
### Switching Databases
**Switch from MySQL to MSSQL:**
```bash
# Update environment
export SPRING_PROFILES_ACTIVE=mssql
export DB_HOST=localhost
export DB_DATABASE=lcc
export DB_USER=sa
export DB_PASSWORD=YourStrong!Passw0rd
# Run application
mvn spring-boot:run
```
**Switch back to MySQL:**
```bash
export SPRING_PROFILES_ACTIVE=mysql
export DB_HOST=localhost
export DB_DATABASE=lcc
export DB_USER=root
export DB_PASSWORD=your_password
mvn spring-boot:run
```
### Running Migrations
Migrations run automatically on application startup when Flyway is enabled.
**Manual migration with Flyway CLI:**
```bash
# MySQL
flyway -url=jdbc:mysql://localhost:3306/lcc -user=root -password=pass -locations=filesystem:src/main/resources/db/migration/mysql migrate
# MSSQL
flyway -url=jdbc:sqlserver://localhost:1433;databaseName=lcc -user=sa -password=pass -locations=filesystem:src/main/resources/db/migration/mssql migrate
```
### Testing Checklist
When modifying repositories or adding new database-dependent code:
1. **Run unit tests** (if applicable)
```bash
mvn test -Dtest=MySQLDialectProviderTest
mvn test -Dtest=MSSQLDialectProviderTest
```
2. **Run repository integration tests on MySQL**
```bash
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mysql
```
3. **Run repository integration tests on MSSQL**
```bash
mvn test -Dtest="*RepositoryIntegrationTest" -Dspring.profiles.active=test,mssql
```
4. **Run full test suite on both databases**
```bash
mvn test -Dspring.profiles.active=test,mysql
mvn test -Dspring.profiles.active=test,mssql
```
### Common Repository Patterns
**Pattern 1: Constructor injection with SqlDialectProvider**
```java
@Repository
public class ExampleRepository {
private final JdbcTemplate jdbcTemplate;
private final SqlDialectProvider dialectProvider;
public ExampleRepository(JdbcTemplate jdbcTemplate, SqlDialectProvider dialectProvider) {
this.jdbcTemplate = jdbcTemplate;
this.dialectProvider = dialectProvider;
}
}
```
**Pattern 2: Pagination queries**
```java
public List<Entity> list(int limit, int offset) {
String sql = "SELECT * FROM table WHERE condition ORDER BY id " +
dialectProvider.buildPaginationClause(limit, offset);
Object[] params = ArrayUtils.addAll(
new Object[]{conditionValue},
dialectProvider.getPaginationParameters(limit, offset)
);
return jdbcTemplate.query(sql, params, rowMapper);
}
```
**Pattern 3: Insert with ID retrieval**
```java
public Integer create(Entity entity) {
String sql = "INSERT INTO table (name, is_active) VALUES (?, ?)";
jdbcTemplate.update(sql, entity.getName(), entity.isActive());
String idSql = dialectProvider.getLastInsertIdQuery();
return jdbcTemplate.queryForObject(idSql, Integer.class);
}
```
**Pattern 4: Upsert operations**
```java
public void upsert(Entity entity) {
String sql = dialectProvider.buildUpsertStatement(
"table_name",
List.of("unique_col1", "unique_col2"), // unique columns
List.of("unique_col1", "unique_col2", "value"), // insert columns
List.of("value") // update columns
);
jdbcTemplate.update(sql, entity.getCol1(), entity.getCol2(), entity.getValue());
}
```