Files
markitect-main/guides/LEGACY_AGENT_GUIDE.md
2025-10-03 03:39:43 +02:00

311 lines
8.6 KiB
Markdown

# Legacy Agent - Comprehensive Legacy Interface Management
## Overview
The Legacy Agent is a comprehensive system for managing legacy interface compatibility and lifecycle in the MarkiTect project. Built on top of the existing legacy compatibility system from Issue #39, it provides intelligent automation for deprecation management, migration assistance, and cleanup operations.
## Architecture
The legacy agent system consists of several interconnected components:
### Core Components
1. **LegacyRegistry** (`markitect/legacy/registry.py`)
- Central registry for all legacy interfaces and versions
- Tracks git commit bindings, deprecation status, and timelines
- Maintains SQLite database for persistence
- Records usage statistics for informed decision making
2. **LegacyAgent** (`markitect/legacy/agent.py`)
- Intelligent automation engine for legacy lifecycle management
- Handles deprecation progression, cleanup scheduling, and notifications
- Configurable automation policies
- Task queue system for scheduled operations
3. **LegacySwitch System** (`markitect/legacy/switches.py`)
- CLI switch management (`--legacy-v1`, `--legacy-v2`, etc.)
- Automatic switch generation based on registry
- Deprecation warning integration
- Legacy routing to appropriate implementations
4. **DeprecationManager** (`markitect/legacy/deprecation.py`)
- Graduated deprecation warning system
- Timeline-based status progression
- User notification management
5. **GitStateTracker** (`markitect/legacy/git_tracker.py`)
- Binds legacy versions to specific git commits
- Enables precise version restoration
- Validates compatibility snapshots
## CLI Interface
The legacy agent exposes comprehensive CLI commands under the `markitect legacy` namespace:
### Core Commands
#### `markitect legacy status`
Shows status of all legacy interfaces with comprehensive metadata:
```bash
# Table view (default)
markitect legacy status
# JSON output
markitect legacy status --format json
# Include removed interfaces
markitect legacy status --include-removed
```
#### `markitect legacy analyze`
Performs intelligent analysis of legacy interfaces:
```bash
# Analyze all interfaces
markitect legacy analyze
# Analyze specific command
markitect legacy analyze query
# Analyze specific version
markitect legacy analyze query v1.0
```
#### `markitect legacy migrate`
Provides migration guidance with breaking change documentation:
```bash
# Get migration guide
markitect legacy migrate query v1.0
# Migrate to specific version
markitect legacy migrate query v1.0 --to-version v2.0
```
#### `markitect legacy cleanup`
Safely removes legacy interfaces with backup options:
```bash
# Interactive cleanup
markitect legacy cleanup query v1.0
# Force cleanup without confirmation
markitect legacy cleanup query v1.0 --force
# Cleanup without backup
markitect legacy cleanup query v1.0 --no-backup
```
### Agent Management
#### `markitect legacy agent-run`
Executes automated maintenance cycles:
```bash
# Run maintenance
markitect legacy agent-run
# Preview mode (dry run)
markitect legacy agent-run --dry-run
```
#### `markitect legacy agent-status`
Shows agent configuration and task queue status:
```bash
# Table view
markitect legacy agent-status
# JSON output
markitect legacy agent-status --format json
```
### Analytics & Reporting
#### `markitect legacy usage-stats`
Displays usage patterns for informed decision making:
```bash
# All interfaces (30 days)
markitect legacy usage-stats
# Specific command
markitect legacy usage-stats --command query
# Extended period
markitect legacy usage-stats --days 90
```
#### `markitect legacy generate-guide`
Creates detailed migration documentation:
```bash
# Output to stdout
markitect legacy generate-guide query v1.0
# Save to file
markitect legacy generate-guide query v1.0 --output migration_guide.md
```
## Legacy Interface Lifecycle
The system manages interfaces through a comprehensive lifecycle:
### Status Progression
1. **CURRENT** - Active implementation
2. **DEPRECATED** - Marked for eventual removal, warnings shown
3. **LEGACY** - Requires explicit `--legacy-vX` flag
4. **SUNSET** - Final warning phase, scheduled for removal
5. **REMOVED** - No longer available
### Automated Progression
The agent can automatically progress interfaces based on:
- Time-based rules (e.g., deprecated → legacy after 90 days)
- Usage patterns (e.g., unused interfaces move to sunset)
- Manual scheduling
- Policy configuration
### Agent Configuration
The agent behavior is configurable via `AgentConfig`:
```python
config = AgentConfig(
auto_progression=True, # Auto-progress deprecations
cleanup_unused_days=180, # Cleanup after 6 months unused
migration_guide_auto_generation=True, # Auto-generate guides
notification_threshold_days=30, # Notify 30 days before removal
max_concurrent_migrations=3, # Limit concurrent operations
backup_before_cleanup=True # Always backup before removal
)
```
## Integration with Existing Legacy System
The agent builds on the foundation from Issue #39:
### Existing Components (Issue #39)
- `markitect/legacy_compat.py` - Basic legacy switches and warnings
- `LegacyVersions` registry with git commit binding
- `LegacyMode` state management
- Environment-based test detection
### New Agent Components
- Advanced registry with database persistence
- Intelligent lifecycle automation
- Comprehensive CLI interface
- Usage analytics and reporting
- Migration assistance tools
### Compatibility
The new system is fully backward compatible with existing legacy switches:
- `--legacy-v39-pre` continues to work as before
- Existing deprecation warnings are preserved
- Test environment detection still functions
## Usage Examples
### Setting Up Legacy Interface
```python
from markitect.legacy import LegacyRegistry, LegacyStatus
registry = LegacyRegistry()
# Register a legacy version
registry.register_legacy_interface(
command='query',
version='v1.0',
git_commit='a1b2c3d4',
status=LegacyStatus.DEPRECATED,
deprecated_date='2025-09-30',
removal_date='2025-12-30',
breaking_changes=['Parameter renamed', 'Output format changed'],
description='Legacy query with old parameter names'
)
```
### Adding Legacy Support to CLI Command
```python
from markitect.legacy import legacy_option, with_legacy_support
@click.command()
@click.argument('sql', type=str)
@legacy_option('v1.0', 'Use v1.0 legacy behavior')
@with_legacy_support('query')
def query_command(sql, legacy_v1_0=False):
# Modern implementation - legacy routing is automatic
return execute_modern_query(sql)
```
### Running Maintenance
```bash
# Check what would be done
markitect legacy agent-run --dry-run
# Execute maintenance
markitect legacy agent-run
# Check results
markitect legacy agent-status
```
## Benefits
1. **Automated Management** - Reduces manual overhead of legacy maintenance
2. **Data-Driven Decisions** - Usage analytics inform deprecation timelines
3. **User Experience** - Clear migration paths and gradual warnings
4. **Safety** - Backup and rollback capabilities
5. **Comprehensive Tracking** - Complete audit trail of all operations
6. **Policy Enforcement** - Consistent application of deprecation policies
## Technical Implementation
### Database Schema
The registry uses SQLite with tables for:
- `legacy_interfaces` - Interface definitions and metadata
- `legacy_usage` - Usage tracking for analytics
### Task System
The agent uses a persistent task queue for:
- Scheduled deprecation progressions
- Cleanup operations
- Notification delivery
- Migration guide generation
### Git Integration
Version bindings enable:
- Precise restoration of legacy behavior
- Validation of compatibility snapshots
- Audit trail of changes
## Future Enhancements
1. **Integration with CI/CD** - Automated testing of legacy interfaces
2. **User Notification System** - Email/webhook notifications
3. **Migration Assistance** - Interactive migration wizards
4. **Advanced Analytics** - Usage heat maps and trend analysis
5. **Policy Templates** - Pre-configured deprecation policies
6. **Cross-Project Support** - Legacy management across multiple projects
## Getting Started
1. **Check Current Status**:
```bash
markitect legacy status
```
2. **Run Analysis**:
```bash
markitect legacy analyze
```
3. **Configure Agent**:
```bash
markitect legacy agent-status
```
4. **Run Maintenance**:
```bash
markitect legacy agent-run --dry-run
markitect legacy agent-run
```
The legacy agent provides a complete solution for managing the entire lifecycle of deprecated interfaces, ensuring smooth transitions while maintaining backward compatibility.