declarative_sqlite 1.3.0

declarative_sqlite: ^1.3.0 copied to clipboard

Features #

• Constraint Violation Handling in bulkLoad: Added graceful constraint violation handling with ConstraintViolationStrategy
  • throwException (default): Maintains existing behavior - throws on constraint violations
  • skip: Silently skips problematic rows and continues processing valid ones
  • Comprehensive constraint detection for unique, primary key, check, foreign key, and NOT NULL violations
  • Detailed logging for monitoring and debugging constraint violation events
• Fixed Unique Constraint Generation: Resolved issue where unique keys weren't properly creating CREATE UNIQUE INDEX statements
  • Unique constraints now generate proper SQL: CREATE UNIQUE INDEX uniq_table_column ON table (column)
  • Improved migration script generation to handle both regular indexes and unique constraints

Developer Experience #

• Enhanced bulkLoad method for server synchronization scenarios
• Better error handling and logging for constraint violations
• Safer bulk loading operations with granular control over error handling

Bug Fixes #

• Fixed missing unique constraint generation in schema migration scripts
• Improved constraint violation detection and categorization

Documentation #

• Updated bulkLoad method documentation with constraint violation handling examples
• Enhanced migration guide with unique constraint best practices

Features #

• Unified Save Method: Enhanced DbRecord.save() to automatically handle both insert and update operations
  • save() now intelligently detects whether a record is new (needs INSERT) or existing (needs UPDATE)
  • Eliminates the need to manually choose between insert() and save() for updates
  • Added isNewRecord property to check whether a record needs insertion
  • Deprecated explicit insert() method in favor of unified save() approach
  • After successful insert via save(), record data is automatically refreshed with all system columns
  • Multiple consecutive save() calls work seamlessly on the same record

Developer Experience #

• Simplified CRUD workflow - just use save() for everything
• Reduced cognitive load - no need to track insert vs update state manually
• Better API consistency across create and update operations

Documentation #

• Updated data modeling guide with unified save examples
• Enhanced CRUD operations documentation with recommended patterns
• Added comprehensive example demonstrating unified save approach

Features #

• Non-LWW Column Protection: Added protection against updating non-LWW columns on server-origin rows
  • Prevents data corruption in distributed systems by restricting updates to LWW columns only on rows that originated from server
  • Local-origin rows can still update all columns freely
  • Throws StateError when attempting to update non-LWW columns on server-origin rows
  • Enhanced DbRecord.setValue() with origin validation
• Improved Dirty Row Tracking: Enhanced dirty row tracking to differentiate between full row updates (local origin) and partial updates (server origin)

Data Safety #

• Added isLocalOrigin property to DbRecord for checking row origin
• Enhanced synchronization safety with column-level update restrictions
• Maintained backward compatibility for existing local data operations

Breaking Changes #

• Removed foreign key functionality from KeyBuilder and DbKey classes
• Foreign key constraint handling removed from exception system
• Architecture now favors application-level relationship management

Documentation #

• Updated all documentation to remove foreign key references
• Improved examples focusing on application-domain relationships

Features #

• Complete rewrite with simplified, reactive database operations
• Declarative schema definition with automatic table creation
• Query builder with type-safe operations
• Streaming queries for reactive UIs
• File repository integration
• Comprehensive test coverage

API Changes #

• Removed transaction support for simplicity
• Removed sync manager and HLC support
• Streamlined DeclarativeDatabase.open() method
• Updated query methods: query(), queryMaps(), streamRecords()
• Simplified schema building with SchemaBuilder

Documentation #

• Updated all examples to use current API
• Added comprehensive test suite
• Updated README with correct usage patterns