import { useState, useEffect, useRef } from 'react';
import { aiCardOCR, ollamaCardOCR, puterCardOCR, geminiCardOCR } from '../lib/ai-ocr';

export default function CameraScanner({ onCardScanned, onError }) {
  const [isStreaming, setIsStreaming] = useState(false);
  const [isProcessing, setIsProcessing] = useState(false);
  const [scanResult, setScanResult] = useState(null);
  const [capturedImage, setCapturedImage] = useState(null);
  const [isAutoScanning, setIsAutoScanning] = useState(true);
  const [detectedCard, setDetectedCard] = useState(null);
  const [scanningAnimation, setScanningAnimation] = useState(false);
  const [toast, setToast] = useState(null);
  const [isDetecting, setIsDetecting] = useState(false);
  const [ocrSettings, setOcrSettings] = useState({
    service: 'gemini', // Default to Gemini
    openaiApiKey: '',
    geminiApiKey: '',
    ollamaUrl: 'http://localhost:11434'
  });

  const videoRef = useRef(null);
  const canvasRef = useRef(null);
  const detectionCanvasRef = useRef(null); // Separate canvas for computer vision
  const streamRef = useRef(null);
  const autoScanIntervalRef = useRef(null);
  const detectionIntervalRef = useRef(null);
  const lastScanTimeRef = useRef(0);
  const lastDetectionTimeRef = useRef(0);
  const detectionHistoryRef = useRef([]); // Track detection stability
  const stableDetectionRef = useRef(null); // Current stable detection

  // Load OCR settings from localStorage
  useEffect(() => {
    const loadOcrSettings = async () => {
      let settings = {
        service: 'gemini', // Default to Gemini
        openaiApiKey: '',
        geminiApiKey: '',
        ollamaUrl: 'http://localhost:11434'
      };

      // Load saved settings
      const savedSettings = localStorage.getItem('ocrSettings');
      if (savedSettings) {
        try {
          const parsed = JSON.parse(savedSettings);
          settings = { ...settings, ...parsed };
        } catch (error) {
          console.error('Failed to load OCR settings:', error);
        }
      }

      // Try to auto-load Gemini API key from environment if not already set
      if (!settings.geminiApiKey) {
        try {
          const response = await fetch('/api/config/gemini');
          if (response.ok) {
            const data = await response.json();
            if (data.hasKey && data.apiKey) {
              settings.geminiApiKey = data.apiKey;
              settings.service = 'gemini'; // Ensure Gemini is selected
              console.log('✅ Auto-configured Gemini API key from environment');
            }
          }
        } catch (error) {
          console.log('Could not auto-load Gemini API key:', error);
        }
      }

      setOcrSettings(settings);
      
      // Configure AI services
      if (settings.openaiApiKey) {
        aiCardOCR.setApiKey(settings.openaiApiKey);
      }
      if (settings.geminiApiKey) {
        geminiCardOCR.setApiKey(settings.geminiApiKey);
      }
      if (settings.ollamaUrl) {
        ollamaCardOCR.setBaseUrl(settings.ollamaUrl);
      }
    };

    loadOcrSettings();
  }, []);

  // Configure canvas contexts for optimal performance
  useEffect(() => {
    if (canvasRef.current) {
      const ctx = canvasRef.current.getContext('2d', { willReadFrequently: true });
    }
    if (detectionCanvasRef.current) {
      const ctx = detectionCanvasRef.current.getContext('2d', { willReadFrequently: true });
    }
  }, []);

  // Computer vision-based card detection (fast, no API calls)
  const detectCardWithComputerVision = () => {
    if (!videoRef.current || !detectionCanvasRef.current) return null;

    const video = videoRef.current;
    const canvas = detectionCanvasRef.current;
    const ctx = canvas.getContext('2d');

    // Set canvas size for detection (smaller for performance)
    canvas.width = 320;
    canvas.height = 240;

    // Draw current video frame
    ctx.drawImage(video, 0, 0, canvas.width, canvas.height);

    // Get image data for analysis
    const imageData = ctx.getImageData(0, 0, canvas.width, canvas.height);
    const data = imageData.data;

    // Card detection algorithm using computer vision
    const cardDetection = analyzeImageForCard(data, canvas.width, canvas.height);
    
    return cardDetection;
  };

  // Computer vision algorithm to detect rectangular card-like objects
  const analyzeImageForCard = (imageData, width, height) => {
    // Convert to grayscale and detect edges
    const grayscale = [];
    const edges = [];
    
    // Convert to grayscale
    for (let i = 0; i < imageData.length; i += 4) {
      const gray = Math.round(0.299 * imageData[i] + 0.587 * imageData[i + 1] + 0.114 * imageData[i + 2]);
      grayscale.push(gray);
    }

    // Enhanced edge detection with stronger thresholds
    for (let y = 1; y < height - 1; y++) {
      for (let x = 1; x < width - 1; x++) {
        const idx = y * width + x;
        
        // Horizontal gradient
        const gx = -grayscale[idx - width - 1] - 2 * grayscale[idx - 1] - grayscale[idx + width - 1] +
                   grayscale[idx - width + 1] + 2 * grayscale[idx + 1] + grayscale[idx + width + 1];
        
        // Vertical gradient  
        const gy = -grayscale[idx - width - 1] - 2 * grayscale[idx - width] - grayscale[idx - width + 1] +
                   grayscale[idx + width - 1] + 2 * grayscale[idx + width] + grayscale[idx + width + 1];
        
        // Edge magnitude
        const magnitude = Math.sqrt(gx * gx + gy * gy);
        edges[idx] = magnitude > 80 ? 255 : 0; // Increased threshold from 50 to 80
      }
    }

    // Look for rectangular regions with stricter criteria
    const cardCandidates = findRectangularRegions(edges, width, height);
    
    // Score candidates based on card-like properties with higher standards
    const bestCandidate = scoreCardCandidates(cardCandidates, width, height);
    
    return bestCandidate;
  };

  // Find rectangular regions that could be cards with stricter criteria
  const findRectangularRegions = (edges, width, height) => {
    const candidates = [];
    const minCardWidth = Math.floor(width * 0.2); // Increased from 15% to 20%
    const maxCardWidth = Math.floor(width * 0.7);  // Decreased from 80% to 70%
    const minCardHeight = Math.floor(height * 0.25); // Increased from 20% to 25%
    const maxCardHeight = Math.floor(height * 0.8); // Decreased from 90% to 80%

    // Scan for edge-dense rectangular regions with larger steps for performance
    for (let y = 0; y < height - minCardHeight; y += 15) { // Increased step from 10 to 15
      for (let x = 0; x < width - minCardWidth; x += 15) { // Increased step from 10 to 15
        for (let w = minCardWidth; w <= maxCardWidth && x + w < width; w += 25) { // Increased step from 20 to 25
          for (let h = minCardHeight; h <= maxCardHeight && y + h < height; h += 25) { // Increased step from 20 to 25
            
            // Stricter card-like aspect ratio check
            const aspectRatio = w / h;
            if (aspectRatio < 0.65 || aspectRatio > 0.77) continue; // Narrowed from 0.6-0.8 to 0.65-0.77
            
            // Count edges in this region with more selective sampling
            let edgeCount = 0;
            let totalPixels = 0;
            let cornerEdges = 0; // Count edges near corners (cards have defined corners)
            
            // Sample the region (not every pixel for performance)
            for (let sy = y; sy < y + h; sy += 4) { // Increased step from 3 to 4
              for (let sx = x; sx < x + w; sx += 4) { // Increased step from 3 to 4
                const idx = sy * width + sx;
                if (edges[idx] === 255) {
                  edgeCount++;
                  
                  // Check if this edge is near a corner (cards have distinct corners)
                  const isNearCorner = (
                    (sx < x + w * 0.2 && sy < y + h * 0.2) || // Top-left
                    (sx > x + w * 0.8 && sy < y + h * 0.2) || // Top-right
                    (sx < x + w * 0.2 && sy > y + h * 0.8) || // Bottom-left
                    (sx > x + w * 0.8 && sy > y + h * 0.8)    // Bottom-right
                  );
                  if (isNearCorner) cornerEdges++;
                }
                totalPixels++;
              }
            }
            
            const edgeDensity = edgeCount / totalPixels;
            const cornerEdgeDensity = cornerEdges / (edgeCount || 1);
            
            // Stricter criteria: cards should have moderate edge density AND corner definition
            if (edgeDensity > 0.08 && edgeDensity < 0.25 && cornerEdgeDensity > 0.1) { // Increased min from 0.05 to 0.08, added corner requirement
              
              // Additional check: look for rectangular perimeter (cards have clear borders)
              const perimeterStrength = checkRectangularPerimeter(edges, x, y, w, h, width);
              
              if (perimeterStrength > 0.3) { // Only accept if perimeter is well-defined
                candidates.push({
                  x: x,
                  y: y,
                  width: w,
                  height: h,
                  edgeDensity: edgeDensity,
                  aspectRatio: aspectRatio,
                  cornerEdgeDensity: cornerEdgeDensity,
                  perimeterStrength: perimeterStrength
                });
              }
            }
          }
        }
      }
    }

    return candidates;
  };

  // Check if the region has a well-defined rectangular perimeter (like a card border)
  const checkRectangularPerimeter = (edges, x, y, w, h, frameWidth) => {
    let perimeterEdges = 0;
    let perimeterPixels = 0;
    
    // Check top and bottom edges
    for (let sx = x; sx < x + w; sx += 2) {
      // Top edge
      const topIdx = y * frameWidth + sx;
      if (topIdx >= 0 && topIdx < edges.length) {
        if (edges[topIdx] === 255) perimeterEdges++;
        perimeterPixels++;
      }
      
      // Bottom edge
      const bottomIdx = (y + h - 1) * frameWidth + sx;
      if (bottomIdx >= 0 && bottomIdx < edges.length) {
        if (edges[bottomIdx] === 255) perimeterEdges++;
        perimeterPixels++;
      }
    }
    
    // Check left and right edges
    for (let sy = y; sy < y + h; sy += 2) {
      // Left edge
      const leftIdx = sy * frameWidth + x;
      if (leftIdx >= 0 && leftIdx < edges.length) {
        if (edges[leftIdx] === 255) perimeterEdges++;
        perimeterPixels++;
      }
      
      // Right edge
      const rightIdx = sy * frameWidth + (x + w - 1);
      if (rightIdx >= 0 && rightIdx < edges.length) {
        if (edges[rightIdx] === 255) perimeterEdges++;
        perimeterPixels++;
      }
    }
    
    return perimeterPixels > 0 ? perimeterEdges / perimeterPixels : 0;
  };

  // Score and select the best card candidate with higher standards
  const scoreCardCandidates = (candidates, frameWidth, frameHeight) => {
    if (candidates.length === 0) return null;

    let bestCandidate = null;
    let bestScore = 0;

    for (const candidate of candidates) {
      let score = 0;
      
      // Prefer candidates with good aspect ratio (closer to 0.71 - typical card ratio)
      const aspectRatioScore = 1 - Math.abs(candidate.aspectRatio - 0.71);
      score += aspectRatioScore * 35; // Increased weight from 30 to 35
      
      // Prefer moderate edge density (not too sparse, not too dense)
      const edgeDensityScore = Math.min(candidate.edgeDensity * 100, 25);
      score += edgeDensityScore;
      
      // Reward corner definition (cards have clear corners)
      const cornerScore = candidate.cornerEdgeDensity * 25;
      score += cornerScore;
      
      // Reward strong perimeter (cards have borders)
      const perimeterScore = candidate.perimeterStrength * 20;
      score += perimeterScore;
      
      // Prefer cards that are reasonably sized
      const cardArea = (candidate.width / frameWidth) * (candidate.height / frameHeight);
      const sizeScore = cardArea > 0.05 && cardArea < 0.4 ? 15 : 0; // Reward reasonable size
      score += sizeScore;
      
      // Prefer cards closer to center (people usually center cards when scanning)
      const centerX = frameWidth / 2;
      const centerY = frameHeight / 2;
      const cardCenterX = candidate.x + candidate.width / 2;
      const cardCenterY = candidate.y + candidate.height / 2;
      const distanceFromCenter = Math.sqrt(
        Math.pow(cardCenterX - centerX, 2) + Math.pow(cardCenterY - centerY, 2)
      );
      const maxDistance = Math.sqrt(Math.pow(centerX, 2) + Math.pow(centerY, 2));
      const centerScore = (1 - distanceFromCenter / maxDistance) * 10; // Reduced weight from 15 to 10
      score += centerScore;

      if (score > bestScore) {
        bestScore = score;
        bestCandidate = candidate;
      }
    }

    // Much higher threshold - only return candidate if it's very likely to be a card
    return bestScore > 70 ? bestCandidate : null; // Increased from 40 to 70
  };

  // Convert canvas coordinates to video coordinates (percentage)
  const convertToVideoCoordinates = (candidate) => {
    if (!candidate || !detectionCanvasRef.current || !videoRef.current) return null;

    const canvas = detectionCanvasRef.current;
    const video = videoRef.current;

    // Convert from detection canvas coordinates to video percentage
    return {
      x: (candidate.x / canvas.width) * 100,
      y: (candidate.y / canvas.height) * 100,
      width: (candidate.width / canvas.width) * 100,
      height: (candidate.height / canvas.height) * 100
    };
  };

  // Fast computer vision detection effect
  useEffect(() => {
    if (isStreaming && isAutoScanning && !isProcessing) {
      // Run computer vision detection every 500ms (fast, no API calls)
      detectionIntervalRef.current = setInterval(() => {
        const candidate = detectCardWithComputerVision();
        
        // Add to detection history for stability tracking
        const now = Date.now();
        detectionHistoryRef.current.push({
          timestamp: now,
          detected: !!candidate,
          candidate: candidate
        });
        
        // Keep only last 6 detections (3 seconds of history at 500ms intervals)
        detectionHistoryRef.current = detectionHistoryRef.current.filter(
          detection => now - detection.timestamp < 3000
        );
        
        // Check for stable detection (at least 4 out of last 6 detections must be positive)
        const recentDetections = detectionHistoryRef.current.slice(-6);
        const positiveDetections = recentDetections.filter(d => d.detected).length;
        const isStableDetection = recentDetections.length >= 4 && positiveDetections >= 4;
        
        if (isStableDetection && candidate) {
          // Only update if this is a new stable detection or significantly different position
          const shouldUpdate = !stableDetectionRef.current || 
            Math.abs(candidate.x - stableDetectionRef.current.x) > 10 ||
            Math.abs(candidate.y - stableDetectionRef.current.y) > 10;
            
          if (shouldUpdate) {
            const videoCoords = convertToVideoCoordinates(candidate);
            if (videoCoords) {
              setDetectedCard(videoCoords);
              stableDetectionRef.current = candidate;
            }
          }
        } else if (!isStableDetection) {
          // Clear detection if not stable
          setDetectedCard(null);
          stableDetectionRef.current = null;
        }
      }, 500);

      // Run AI analysis only when card is detected and stable for 3 seconds
      autoScanIntervalRef.current = setInterval(() => {
        const now = Date.now();
        if (detectedCard && now - lastScanTimeRef.current > 3000) {
          processDetectedCard();
        }
      }, 1000);
    } else {
      if (detectionIntervalRef.current) {
        clearInterval(detectionIntervalRef.current);
        detectionIntervalRef.current = null;
      }
      if (autoScanIntervalRef.current) {
        clearInterval(autoScanIntervalRef.current);
        autoScanIntervalRef.current = null;
      }
      // Clear detection history when not scanning
      detectionHistoryRef.current = [];
      stableDetectionRef.current = null;
    }

    return () => {
      if (detectionIntervalRef.current) {
        clearInterval(detectionIntervalRef.current);
      }
      if (autoScanIntervalRef.current) {
        clearInterval(autoScanIntervalRef.current);
      }
    };
  }, [isStreaming, isAutoScanning, isProcessing, detectedCard]);

  // Show toast notification
  const showToast = (message, type = 'success') => {
    setToast({ message, type });
    setTimeout(() => setToast(null), 3000);
  };

  // Start camera stream with card aspect ratio
  const startCamera = async () => {
    try {
      const stream = await navigator.mediaDevices.getUserMedia({
        video: {
          facingMode: 'environment', // Use back camera on mobile
          width: { ideal: 1280 },
          height: { ideal: 720 }, // 16:9 aspect ratio, good for cards
          aspectRatio: { ideal: 16/9 }
        }
      });

      if (videoRef.current) {
        videoRef.current.srcObject = stream;
        streamRef.current = stream;
        
        // Wait for video to be ready
        videoRef.current.onloadedmetadata = () => {
          videoRef.current?.play().then(() => {
            setIsStreaming(true);
          }).catch((err) => {
            onError(`Video playback failed: ${err.message}`);
          });
        };
        
        videoRef.current.onerror = (err) => {
          onError('Video element error occurred');
        };
      } else {
        onError('Video element not available');
      }
    } catch (err) {
      console.error('Camera access error:', err);
      onError(`Unable to access camera: ${err.message}`);
    }
  };

  // Stop camera stream
  const stopCamera = () => {
    if (streamRef.current) {
      streamRef.current.getTracks().forEach(track => track.stop());
      streamRef.current = null;
    }
    if (autoScanIntervalRef.current) {
      clearInterval(autoScanIntervalRef.current);
      autoScanIntervalRef.current = null;
    }
    if (detectionIntervalRef.current) {
      clearInterval(detectionIntervalRef.current);
      detectionIntervalRef.current = null;
    }
    setIsStreaming(false);
    setCapturedImage(null);
    setScanResult(null);
    setDetectedCard(null);
    setIsDetecting(false);
    // Clear detection state
    detectionHistoryRef.current = [];
    stableDetectionRef.current = null;
  };

  // Process the detected card with full AI analysis
  const processDetectedCard = async () => {
    if (!detectedCard || isProcessing) return;

    lastScanTimeRef.current = Date.now();
    setScanningAnimation(true);

    // Capture image
    const imageDataUrl = captureImageData();
    if (imageDataUrl) {
      await processImage(imageDataUrl, true); // true for auto-scan
    }

    setTimeout(() => setScanningAnimation(false), 2000);
  };

  // Capture image from video stream
  const captureImageData = () => {
    if (!videoRef.current || !canvasRef.current) return null;

    const video = videoRef.current;
    const canvas = canvasRef.current;
    const ctx = canvas.getContext('2d');

    // Set canvas dimensions to match video
    canvas.width = video.videoWidth;
    canvas.height = video.videoHeight;

    // Draw current video frame to canvas
    ctx?.drawImage(video, 0, 0, canvas.width, canvas.height);

    // Get image data URL
    return canvas.toDataURL('image/jpeg', 0.8);
  };

  // Manual capture for testing
  const captureImage = () => {
    const imageDataUrl = captureImageData();
    if (imageDataUrl) {
      setCapturedImage(imageDataUrl);
      processImage(imageDataUrl, false);
    }
  };

  // Process image with AI OCR (only called when computer vision detects a card)
  const processImage = async (imageData, isAutoScan = false) => {
    setIsProcessing(true);
    setScanResult(null);

    try {
      let ocrResult;
      
      console.log('🔍 Current OCR settings:', ocrSettings);
      
      if (ocrSettings.service === 'puter') {
        console.log('🎯 Using Puter.js (Free AI Vision)...');
        ocrResult = await puterCardOCR.analyzeCard(imageData);
        console.log('✅ Puter.js Vision result:', ocrResult);
      } else if (ocrSettings.service === 'openai') {
        console.log('🤖 Using OpenAI Vision API...');
        ocrResult = await aiCardOCR.analyzeCard(imageData);
        console.log('✅ OpenAI Vision result:', ocrResult);
      } else if (ocrSettings.service === 'ollama') {
        console.log('🦙 Using Ollama Vision...');
        ocrResult = await ollamaCardOCR.analyzeCard(imageData);
        console.log('✅ Ollama Vision result:', ocrResult);
      } else if (ocrSettings.service === 'gemini') {
        console.log('🤖 Using Gemini Vision API...');
        ocrResult = await geminiCardOCR.analyzeCard(imageData);
        console.log('✅ Gemini Vision result:', ocrResult);
      } else {
        throw new Error('No OCR service configured');
      }

      // First check: Is this actually a trading card?
      if (!ocrResult.isCard) {
        if (isAutoScan) {
          console.log('❌ No trading card detected in auto-scan, continuing...');
          setDetectedCard(null); // Remove highlight
          return;
        } else {
          // Show error for manual scans
          onError(ocrResult.reason || 'No trading card detected in image');
          return;
        }
      }

      // Second check: Do we have a card name?
      if (!ocrResult.cardName || ocrResult.cardName.trim().length < 2) {
        if (isAutoScan) {
          console.log('❌ Card name not clear in auto-scan, continuing...');
          setDetectedCard(null);
          return;
        } else {
          onError('Could not read card name clearly. Please try again with better lighting.');
          return;
        }
      }

      // Third check: Cross-reference with database
      console.log('🔍 Cross-referencing with database...');
      const dbResponse = await fetch('/api/cards/find-or-create', {
        method: 'POST',
        headers: {
          'Content-Type': 'application/json',
          'Authorization': `Bearer ${localStorage.getItem('auth_token')}`
        },
        body: JSON.stringify({
          name: ocrResult.cardName.trim(),
          set: ocrResult.setName,
          setCode: ocrResult.setCode,
          cardNumber: ocrResult.cardNumber,
          game: ocrResult.game,
          cardType: ocrResult.cardType,
          rarity: ocrResult.rarity,
          hp: ocrResult.hp,
          manaCost: ocrResult.manaCost,
          ocrData: {
            confidence: ocrResult.confidence,
            rawText: ocrResult.rawText,
            abilities: ocrResult.abilities,
            flavorText: ocrResult.flavorText,
            artist: ocrResult.artist
          }
        })
      });

      if (!dbResponse.ok) {
        throw new Error(`Database lookup failed: ${dbResponse.status}`);
      }

      const dbResult = await dbResponse.json();

      // Handle different database response scenarios
      if (dbResult.needsUserSelection) {
        // Multiple matches found - let user choose
        if (isAutoScan) {
          showToast(`⚠️ Multiple matches for "${ocrResult.cardName}" - use manual mode to select`, 'warning');
          setDetectedCard(null);
          return;
        } else {
          // Show selection modal for manual scans
          setScanResult({
            type: 'multipleMatches',
            cardName: ocrResult.cardName,
            matches: dbResult.matches,
            confidence: ocrResult.confidence
          });
          return;
        }
      }

      if (dbResult.needsUserInput) {
        // Low confidence or unclear card
        if (isAutoScan) {
          console.log('❌ Low confidence match, skipping auto-scan');
          setDetectedCard(null);
          return;
        } else {
          onError(dbResult.message || 'Card not found in database and confidence is too low');
          return;
        }
      }

      // Success! We have a confirmed card match
      const finalCard = dbResult.card;
      const cleanCardName = finalCard.name;

      setScanResult({
        type: 'success',
        cardName: cleanCardName,
        setName: finalCard.set_name,
        game: finalCard.game,
        rarity: finalCard.rarity,
        confidence: ocrResult.confidence,
        isExisting: dbResult.isExisting,
        message: dbResult.message
      });

      // Show success toast for auto-scan
      if (isAutoScan) {
        showToast(`📸 Captured: ${cleanCardName}`, 'success');
        setDetectedCard(null); // Remove highlight after successful scan
      }

      // Send the confirmed card data to parent
      onCardScanned({
        name: cleanCardName,
        set: finalCard.set_name,
        setCode: finalCard.set_code,
        cardNumber: finalCard.card_number,
        game: finalCard.game,
        cardType: finalCard.card_type,
        rarity: finalCard.rarity,
        hp: finalCard.hp || ocrResult.hp,
        manaCost: finalCard.mana_cost || ocrResult.manaCost,
        abilities: ocrResult.abilities,
        ocrText: ocrResult.rawText,
        confidence: ocrResult.confidence,
        capturedImage: imageData,
        image_url: finalCard.image_url, // Add the database card image
        databaseId: finalCard.id,
        isExisting: dbResult.isExisting
      });

    } catch (error) {
      console.error('Card processing error:', error);
      if (!isAutoScan) {
        onError(`Card processing failed: ${error.message}`);
      } else {
        // For auto-scans, just continue silently
        setDetectedCard(null);
      }
    } finally {
      setIsProcessing(false);
    }
  };

  // Cleanup on unmount
  useEffect(() => {
    return () => {
      stopCamera();
    };
  }, []);

  return (
    <div className="camera-scanner">
      {/* Camera Controls */}
      <div className="flex gap-4 mb-4 flex-wrap">
        {!isStreaming ? (
          <button
            onClick={startCamera}
            className="px-6 py-3 rounded-xl font-medium transition-all duration-200 flex items-center gap-2 hover:opacity-90"
            style={{ 
              backgroundColor: 'var(--accent-ember)', 
              color: 'white' 
            }}
          >
            <span>📹</span> Start Camera
          </button>
        ) : (
          <>
            <button
              onClick={() => setIsAutoScanning(!isAutoScanning)}
              className={`px-4 py-2 rounded-xl font-medium transition-all duration-200 flex items-center gap-2 hover:opacity-90`}
              style={{ 
                backgroundColor: isAutoScanning ? 'var(--accent-flame)' : 'var(--bg-tertiary)', 
                color: isAutoScanning ? 'white' : 'var(--text-primary)',
                border: isAutoScanning ? 'none' : '1px solid var(--border)'
              }}
            >
              <span>{isAutoScanning ? '🔄' : '⏸️'}</span>
              {isAutoScanning ? 'Auto Scanning' : 'Manual Mode'}
            </button>
            
            <button
              onClick={captureImage}
              disabled={isProcessing}
              className="px-4 py-2 rounded-xl font-medium transition-all duration-200 flex items-center gap-2 hover:opacity-90 disabled:opacity-50"
              style={{ 
                backgroundColor: 'var(--bg-tertiary)', 
                color: 'var(--text-primary)',
                border: '1px solid var(--border)'
              }}
            >
              <span>📸</span> 
              Manual Capture
            </button>
            
            <button
              onClick={stopCamera}
              className="px-4 py-2 rounded-xl font-medium transition-all duration-200 flex items-center gap-2 hover:opacity-90"
              style={{ 
                backgroundColor: 'var(--bg-tertiary)', 
                color: 'var(--text-primary)',
                border: '1px solid var(--border)'
              }}
            >
              <span>⏹️</span> Stop
            </button>
          </>
        )}
      </div>

      {/* Camera Preview with Card-like Aspect Ratio */}
      <div 
        className="relative rounded-xl overflow-hidden mb-4" 
        style={{ 
          backgroundColor: 'var(--bg-secondary)',
          aspectRatio: '16/9', // Card-friendly aspect ratio
          maxHeight: '400px'
        }}
      >
        <video
          ref={videoRef}
          autoPlay
          playsInline
          muted
          className="w-full h-full object-cover"
          style={{ 
            minHeight: isStreaming ? 'auto' : '300px',
            aspectRatio: '16/9'
          }}
        />
        
        {/* Card Detection Highlight - Only show when card is actually detected */}
        {detectedCard && isStreaming && (
          <div 
            className="absolute border-4 rounded-xl transition-all duration-500 ease-in-out"
            style={{ 
              borderColor: 'var(--accent-ember)',
              left: `${detectedCard.x}%`,
              top: `${detectedCard.y}%`,
              width: `${detectedCard.width}%`,
              height: `${detectedCard.height}%`,
              backgroundColor: 'rgba(var(--accent-ember-rgb), 0.1)',
              boxShadow: '0 0 30px rgba(var(--accent-ember-rgb), 0.6)',
              borderRadius: '12px', // Card-like rounded corners
              animation: 'pulse 2s infinite'
            }}
          >
            {/* Card Detected Label */}
            <div 
              className="absolute -top-8 left-0 px-3 py-1 rounded-lg text-sm font-medium"
              style={{ 
                backgroundColor: 'var(--accent-ember)', 
                color: 'white',
                fontSize: '12px'
              }}
            >
              🃏 Card Detected
            </div>
            
            {/* Scanning Animation */}
            {scanningAnimation && (
              <div className="absolute inset-0 flex items-center justify-center">
                <div className="animate-spin rounded-full h-8 w-8 border-b-2 border-white"></div>
              </div>
            )}
          </div>
        )}
        
        {/* Detection Status Indicator */}
        {isStreaming && (
          <div className="absolute top-4 left-4 flex items-center gap-2">
            <div 
              className={`w-3 h-3 rounded-full transition-all duration-300 ${
                detectedCard ? 'animate-pulse' : ''
              }`}
              style={{ 
                backgroundColor: detectedCard ? '#10b981' : '#6b7280'
              }}
            />
            <span className="text-sm font-medium" style={{ color: 'white', textShadow: '0 1px 2px rgba(0,0,0,0.8)' }}>
              {detectedCard ? '🃏 Card Found' : '👁️ Watching'}
            </span>
          </div>
        )}
        
        {/* Show overlay when streaming but no card detected */}
        {isStreaming && !detectedCard && (
          <div className="absolute inset-0 pointer-events-none">
            <div className="absolute inset-8 border-2 border-dashed rounded-xl flex items-center justify-center" style={{ borderColor: 'rgba(var(--accent-ember-rgb), 0.5)' }}>
              <div className="px-6 py-4 rounded-xl text-center" style={{ backgroundColor: 'rgba(0,0,0,0.7)', color: 'white' }}>
                <div className="font-medium text-lg mb-2">🃏 Position Trading Card</div>
                <div className="text-sm opacity-75">
                  {isAutoScanning ? 'Computer vision will detect rectangular cards' : 'Position card and tap capture'}
                </div>
                <div className="text-xs opacity-60 mt-1">No API calls until card is detected</div>
              </div>
            </div>
          </div>
        )}
        
        {/* Show placeholder when not streaming */}
        {!isStreaming && (
          <div className="absolute inset-0 flex items-center justify-center">
            <div className="text-center" style={{ color: 'var(--text-secondary)' }}>
              <div className="text-4xl mb-2">📹</div>
              <div className="font-medium">Camera Preview</div>
              <div className="text-sm opacity-75">Click "Start Camera" to begin</div>
              <div className="text-xs opacity-60 mt-2">AI-optimized card detection</div>
            </div>
          </div>
        )}
      </div>

      {/* Toast Notification */}
      {toast && (
        <div 
          className={`fixed top-4 right-4 z-50 px-4 py-3 rounded-xl shadow-lg transition-all duration-300 ${
            toast.type === 'success' ? 'border-green-500' : toast.type === 'warning' ? 'border-yellow-500' : 'border-red-500'
          }`}
          style={{ 
            backgroundColor: 'var(--bg-secondary)', 
            borderColor: toast.type === 'success' ? '#10b981' : toast.type === 'warning' ? '#f59e0b' : '#ef4444',
            borderWidth: '2px',
            color: 'var(--text-primary)'
          }}
        >
          <div className="flex items-center gap-2">
            <span>{toast.type === 'success' ? '✅' : toast.type === 'warning' ? '⚠️' : '❌'}</span>
            <span className="font-medium">{toast.message}</span>
          </div>
        </div>
      )}

      {/* Hidden canvases for image processing */}
      <canvas ref={canvasRef} className="hidden" />
      <canvas ref={detectionCanvasRef} className="hidden" />

      {/* Processing Status */}
      {isProcessing && !scanningAnimation && (
        <div className="rounded-xl p-4 mb-4" style={{ backgroundColor: 'var(--bg-tertiary)' }}>
          <div className="flex items-center gap-3">
            <div className="animate-spin rounded-full h-6 w-6 border-b-2" style={{ borderColor: 'var(--accent-ember)' }}></div>
            <div>
              <div className="font-medium" style={{ color: 'var(--text-primary)' }}>Processing with AI...</div>
              <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>Card detected by computer vision</div>
            </div>
          </div>
        </div>
      )}

      {/* Manual Capture Results (only show for manual captures) */}
      {scanResult && capturedImage && (
        <div className="rounded-xl border p-4 mb-4" style={{ backgroundColor: 'var(--bg-secondary)', borderColor: 'var(--border)' }}>
          <div className="font-medium mb-2" style={{ color: 'var(--text-primary)' }}>Manual Scan Results</div>
          
          {scanResult.type === 'multipleMatches' ? (
            <div className="space-y-2">
              <div className="flex items-center gap-2">
                <span className="text-yellow-600">⚠️</span>
                <div className="font-medium" style={{ color: 'var(--text-primary)' }}>Multiple matches found for "{scanResult.cardName}":</div>
              </div>
              {scanResult.matches.map((match, index) => (
                <div key={index} className="flex items-center gap-2 text-sm" style={{ color: 'var(--text-secondary)' }}>
                  <span>{match.name}</span>
                  <span>({match.set_name})</span>
                </div>
              ))}
              <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>
                Confidence: {Math.round(scanResult.confidence)}%
              </div>
            </div>
          ) : scanResult.type === 'success' ? (
            <div className="space-y-2">
              <div className="flex items-center gap-2">
                <span className="text-green-600">✅</span>
                <div>
                  <div className="font-medium" style={{ color: 'var(--text-primary)' }}>Card Found: {scanResult.cardName}</div>
                  {scanResult.setName && (
                    <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>Set: {scanResult.setName}</div>
                  )}
                  {scanResult.game && (
                    <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>Game: {scanResult.game}</div>
                  )}
                  {scanResult.rarity && (
                    <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>Rarity: {scanResult.rarity}</div>
                  )}
                  {scanResult.message && (
                    <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>{scanResult.message}</div>
                  )}
                </div>
              </div>
              <div className="text-sm" style={{ color: 'var(--text-secondary)' }}>
                Confidence: {Math.round(scanResult.confidence)}%
              </div>
            </div>
          ) : (
            <div className="space-y-2">
              <div className="flex items-center gap-2">
                <span className="text-yellow-600">⚠️</span>
                <div className="font-medium" style={{ color: 'var(--text-primary)' }}>Card not clearly recognized</div>
              </div>
              <details className="text-sm" style={{ color: 'var(--text-secondary)' }}>
                <summary className="cursor-pointer">View raw OCR response</summary>
                <pre className="mt-2 whitespace-pre-wrap p-2 rounded text-xs" style={{ backgroundColor: 'var(--bg-tertiary)' }}>
                  {scanResult.rawText || 'No raw text available'}
                </pre>
              </details>
            </div>
          )}
        </div>
      )}

      {/* Instructions */}
      <div className="rounded-xl p-4 mt-4" style={{ backgroundColor: 'var(--bg-tertiary)' }}>
        <div className="font-medium mb-2" style={{ color: 'var(--text-primary)' }}>💡 Efficient Card Detection</div>
        <ul className="text-sm space-y-1" style={{ color: 'var(--text-secondary)' }}>
          <li>• 👁️ Computer vision detects rectangular objects (no API calls)</li>
          <li>• 🎯 AI analysis only when card-like shape is found</li>
          <li>• ⚡ Fast detection every 500ms, AI scan every 3 seconds</li>
          <li>• 💰 Dramatically reduces API usage and costs</li>
          <li>• 🃏 Still maintains high accuracy for card recognition</li>
        </ul>
      </div>
    </div>
  );
}