Warrior_EA/DATABASE.md
AnimateDread 74c7395127 feat: add max-pooling and convolution OpenCL kernels, clean up barrier and signal code
- Define MAX_WEIGHT constant (1.0e6) for weight limits in clusters
- Remove redundant barrier from FeedForward kernel (prevents sync issues)
- Port FeedForwardProof and CalcInputGradientProof kernels for max-pooling (no weights, sliding max)
- Port FeedForwardConv kernel for convolution layers (shared weights, multiple output channels)
- Remove unused code and refactor signal condition logic (CSignalPAI)
2026-07-13 03:23:39 -04:00

3.1 KiB

Database Module Documentation

Overview

The Database module in Warrior_EA provides robust, fault-tolerant management of all persistent data required by the EA, including signal statistics, ranking, and versioning. It is designed for reliability, modularity, and ease of integration with the AI/ML and signal subsystems.


Key Components

1. DatabaseManager.mqh

  • CDatabaseManager: Orchestrates all database operations, including initialization, version management, file system operations, and transactional safety. Integrates:
    • CDatabaseConnectionManager: Handles opening/closing database connections and transaction state.
    • CDatabaseVersionManager: Manages versioning and upgrades of the database schema/files.
    • CDatabaseFileSystemManager: Handles directory and file operations with retry logic.
    • CDatabaseOperationsManager: Executes SQL queries and manages prepared statements.
  • Design Notes: Implements retry logic, version checks, and safe upgrades. Cleans or recreates directories as needed for version mismatches or training mode.

2. DatabaseConnectionManager.mqh

  • CDatabaseConnectionManager: Manages the lifecycle of a database connection, including retries, transaction begin/commit/rollback, and verbose logging. Ensures only one open handle at a time and provides robust error handling.

3. DatabaseFileSystemManager.mqh

  • CDatabaseFileSystemManager: Provides safe, retry-based wrappers for directory creation, cleaning, and file deletion. Ensures file system operations are robust against transient errors.

4. DatabaseOperationsManager.mqh

  • CDatabaseOperationsManager: Higher-level manager for executing and retrying SQL queries, creating tables, and managing prepared statements. Integrates with the connection manager for handle management.

5. DatabaseVersionManager.mqh

  • CDatabaseVersionManager: Handles reading and updating the database version file, with retry logic for file I/O. Ensures schema upgrades and version mismatches are handled safely.

Design Highlights

  • Fault Tolerance: All file and database operations use retry loops and error logging to maximize reliability.
  • Transactional Safety: Explicit transaction management with rollback on failure.
  • Versioning: Database version is tracked and enforced, with automatic upgrades and directory cleaning as needed.
  • Separation of Concerns: Each class has a clear, single responsibility, following SOLID principles.
  • Integration: Designed to work seamlessly with the AI/ML and signal modules for persistent storage of model weights, statistics, and rankings.

Usage Patterns

  • Initialize the database via CDatabaseManager.Init() with version and directory structure.
  • Use connection and operations managers for all SQL and file operations.
  • Always check return values for error handling and retries.

Next Steps

  • Add unit tests for all database operations and error paths.
  • Expand documentation for each SQL schema/table used.
  • Continue with documentation of Enumerations/ next.

This documentation will be updated as additional files are analyzed and integrated.