"""
Test Visual Geolocation with Bangkok Data
Simulate Erik's Bangkok images
"""

import sys
sys.path.insert(0, '/home/bernt/.openclaw/workspace/iom')

from visual_geolocation.pipeline import VisualGeolocationPipeline
from visual_geolocation.evidence_extractor import (
    EvidencePackage, ImageMetadata, VisualObject, 
    TextDetection, GeometricFeature, EnvironmentalSignal
)
from visual_geolocation.geolocation_engine import GeolocationEngine
from visual_geolocation.confidence_model import ConfidenceModel
from datetime import datetime


def create_bangkok_evidence():
    """Create evidence package for Bangkok image"""
    
    # Metadata from Bangkok image
    metadata = ImageMetadata(
        gps_lat=13.7563,
        gps_lng=100.5018,
        altitude=10.0,
        compass_heading=90.0,  # East-facing
        pitch=5.0,
        roll=2.0,
        timestamp="2026-06-26T20:00:00Z",  # Night time
        camera_model="iPhone15,2",
        focal_length=24.0,
        exposure=0.033,  # 1/30s
        aperture=1.78,
        iso=1600
    )
    
    # Visual objects detected
    visual_objects = [
        VisualObject(
            label="street_light",
            confidence=0.85,
            bbox=[100, 50, 150, 400],
            attributes={"type": "LED", "color": "white"}
        ),
        VisualObject(
            label="building",
            confidence=0.92,
            bbox=[200, 100, 500, 400],
            attributes={"type": "commercial", "floors": 5}
        ),
        VisualObject(
            label="car",
            confidence=0.78,
            bbox=[50, 300, 200, 450],
            attributes={"type": "sedan", "color": "white"}
        ),
        VisualObject(
            label="sign",
            confidence=0.88,
            bbox=[300, 50, 450, 150],
            attributes={"type": "commercial", "illuminated": True}
        )
    ]
    
    # Semantic objects
    semantic_objects = [
        VisualObject(
            label="street_view",
            confidence=0.95,
            bbox=[0, 0, 640, 480],
            attributes={"type": "scene", "lighting": "night"}
        ),
        VisualObject(
            label="commercial_area",
            confidence=0.82,
            bbox=[0, 0, 640, 480],
            attributes={"type": "zone", "density": "high"}
        )
    ]
    
    # OCR text detections
    text_detections = [
        TextDetection(
            text="Sukhumvit Road",
            confidence=0.92,
            bbox=[50, 50, 250, 100],
            language="en",
            text_type="street_name"
        ),
        TextDetection(
            text="Sainokuni",
            confidence=0.88,
            bbox=[300, 60, 450, 120],
            language="ja",
            text_type="business_name"
        ),
        TextDetection(
            text="MASSAGE",
            confidence=0.85,
            bbox=[500, 80, 600, 130],
            language="en",
            text_type="business_name"
        ),
        TextDetection(
            text="Bangkok 10110",
            confidence=0.90,
            bbox=[50, 400, 200, 450],
            language="en",
            text_type="postal_code"
        )
    ]
    
    # Geometric features
    geometric_features = [
        GeometricFeature(
            feature_type="horizon_line",
            coordinates=[0, 240, 640, 240],
            confidence=0.70
        ),
        GeometricFeature(
            feature_type="vanishing_point",
            coordinates=[320, 200],
            confidence=0.65
        ),
        GeometricFeature(
            feature_type="camera_height_estimate",
            coordinates=[1.6],  # meters
            confidence=0.60
        )
    ]
    
    # Environmental signals
    environmental_signals = [
        EnvironmentalSignal(
            signal_type="lighting",
            value={
                "type": "artificial",
                "sources": ["street_light", "commercial_signs"],
                "brightness": "medium",
                "is_night": True
            },
            confidence=0.95
        ),
        EnvironmentalSignal(
            signal_type="weather",
            value={
                "condition": "clear",
                "temperature_estimate": 28,  # celsius
                "humidity_estimate": 80
            },
            confidence=0.70
        ),
        EnvironmentalSignal(
            signal_type="urban_density",
            value={
                "level": "high",
                "building_height": "medium",
                "commercial_activity": "high"
            },
            confidence=0.85
        )
    ]
    
    # Temporal signals
    temporal_signals = {
        "timestamp": "2026-06-26T20:00:00Z",
        "time_of_day": "night",
        "season": "summer",
        "day_of_week": "Friday",
        "is_weekend": False,
        "local_time": "03:00",  # Bangkok time (UTC+7)
    }
    
    # Create evidence package
    evidence = EvidencePackage(
        image_id="bangkok_001",
        timestamp=datetime.now(),
        metadata=metadata,
        visual_objects=visual_objects,
        semantic_objects=semantic_objects,
        text_detections=text_detections,
        geometric_features=geometric_features,
        environmental_signals=environmental_signals,
        temporal_signals=temporal_signals
    )
    
    return evidence


def test_bangkok_geolocation():
    """Test geolocation with Bangkok evidence"""
    print("=" * 60)
    print("BANGKOK VISUAL GEOLOCATION TEST")
    print("=" * 60)
    
    # Create evidence
    evidence = create_bangkok_evidence()
    
    print("\n📸 EVIDENCE PACKAGE")
    print(f"   Image ID: {evidence.image_id}")
    print(f"   Timestamp: {evidence.timestamp}")
    print(f"   Visual objects: {len(evidence.visual_objects)}")
    print(f"   Semantic objects: {len(evidence.semantic_objects)}")
    print(f"   Text detections: {len(evidence.text_detections)}")
    print(f"   Geometric features: {len(evidence.geometric_features)}")
    print(f"   Environmental signals: {len(evidence.environmental_signals)}")
    
    # Geolocate
    print("\n🌍 GEOLOCATION")
    engine = GeolocationEngine()
    estimate = engine.geolocate(evidence)
    
    print(f"   Position: {estimate.lat:.6f}, {estimate.lng:.6f}")
    print(f"   Accuracy: ±{estimate.accuracy:.1f}m")
    print(f"   Confidence: {estimate.confidence:.1%}")
    print(f"   Method: {estimate.method}")
    
    # Confidence
    print("\n✅ CONFIDENCE")
    model = ConfidenceModel()
    report = model.calculate_confidence(estimate, evidence)
    
    print(f"   Overall: {report.overall_confidence:.1%}")
    print(f"   Uncertainty: ±{report.uncertainty_radius:.1f}m")
    print(f"   Supporting evidence: {len(report.supporting_evidence)}")
    for ev in report.supporting_evidence:
        print(f"     - {ev['type']}: {ev['description']}")
    print(f"   Contradicting evidence: {len(report.contradicting_evidence)}")
    print(f"   Unknown factors: {len(report.unknown_factors)}")
    for factor in report.unknown_factors:
        print(f"     - {factor}")
    
    # Summary
    print("\n" + "=" * 60)
    print("RESULT")
    print("=" * 60)
    print(f"📍 Bangkok, Sukhumvit Road")
    print(f"   Position: {estimate.lat:.6f}°N, {estimate.lng:.6f}°E")
    print(f"   Accuracy: ±{estimate.accuracy:.1f}m")
    print(f"   Confidence: {report.overall_confidence:.1%}")
    print(f"   Time: Night (Friday, Summer)")
    print(f"   Evidence: {len(evidence.visual_objects)} visual + {len(evidence.text_detections)} text")
    
    return estimate, report


if __name__ == '__main__':
    test_bangkok_geolocation()