Add HealthBridge iOS app for intelligent health data synchronization

Complete implementation of a SwiftUI iOS app that serves as a "Single Source
of Truth" for health data. The app reads from all Apple Health sources,
detects conflicts between devices, merges data using configurable strategies,
and writes cleaned data back.

Features:
- Phase 1: HealthKit integration with automatic source discovery
- Phase 2: DataReader with conflict detection (time-window based)
- Phase 3: RuleEngine with 8 merge strategies (exclusive, priority, higher wins, etc.)
- Phase 4: MergeEngine for conflict resolution + DataWriter for HealthKit writes
- Phase 5: SwiftUI UI for dashboard, conflicts, rules, and sources management
- Phase 6: Background sync with configurable intervals and push notifications
- Phase 7: Complete rule editor and polished UI components

Supported data types:
- Steps, Heart Rate, Blood Pressure, SpO2, Sleep
- Distance, Floors Climbed, Active Energy, HRV, Respiratory Rate

Architecture: SourceManager -> DataReader -> RuleEngine -> MergeEngine -> DataWriter

HealthBridge/Services/DataReader.swift

@@ -0,0 +1,488 @@
+import Foundation
+import HealthKit
+import Combine
+
+// MARK: - Data Reader
+@MainActor
+class DataReader: ObservableObject {
+    static let shared = DataReader()
+
+    private let healthKitManager = HealthKitManager.shared
+    private let sourceManager = SourceManager.shared
+
+    @Published var isReading = false
+    @Published var lastReadDate: Date?
+    @Published var detectedConflicts: [Conflict] = []
+    @Published var readingProgress: Double = 0
+
+    private init() {}
+
+    // MARK: - Fetch Data by Type and Date Range
+
+    func fetchData(
+        for dataType: HealthDataType,
+        from startDate: Date,
+        to endDate: Date,
+        groupByWindow intervalMinutes: Int = 15
+    ) async throws -> [TimeWindowData] {
+        isReading = true
+        defer { isReading = false }
+
+        let samples = try await healthKitManager.fetchSamples(
+            for: dataType,
+            from: startDate,
+            to: endDate
+        )
+
+        // Group samples by source
+        let groupedBySource = groupBySource(samples: samples, dataType: dataType)
+
+        // Create time windows
+        let windows = generateTimeWindows(from: startDate, to: endDate, intervalMinutes: intervalMinutes)
+
+        // Assign samples to windows
+        var windowDataList: [TimeWindowData] = []
+
+        for window in windows {
+            let windowData = createWindowData(
+                window: window,
+                dataType: dataType,
+                groupedBySource: groupedBySource
+            )
+            windowDataList.append(windowData)
+        }
+
+        lastReadDate = Date()
+        return windowDataList
+    }
+
+    private func groupBySource(samples: [HKSample], dataType: HealthDataType) -> [String: [HKSample]] {
+        var grouped: [String: [HKSample]] = [:]
+
+        for sample in samples {
+            let sourceId = sample.sourceRevision.source.bundleIdentifier
+            if grouped[sourceId] == nil {
+                grouped[sourceId] = []
+            }
+            grouped[sourceId]?.append(sample)
+        }
+
+        return grouped
+    }
+
+    private func generateTimeWindows(from start: Date, to end: Date, intervalMinutes: Int) -> [TimeWindow] {
+        var windows: [TimeWindow] = []
+        var current = start
+
+        while current < end {
+            let windowEnd = min(
+                Calendar.current.date(byAdding: .minute, value: intervalMinutes, to: current)!,
+                end
+            )
+            windows.append(TimeWindow(start: current, end: windowEnd))
+            current = windowEnd
+        }
+
+        return windows
+    }
+
+    private func createWindowData(
+        window: TimeWindow,
+        dataType: HealthDataType,
+        groupedBySource: [String: [HKSample]]
+    ) -> TimeWindowData {
+        var readings: [SourceReading] = []
+
+        for (sourceId, samples) in groupedBySource {
+            let windowSamples = samples.filter { sample in
+                sample.startDate < window.end && sample.endDate > window.start
+            }
+
+            if !windowSamples.isEmpty {
+                let reading = createReading(
+                    from: windowSamples,
+                    sourceId: sourceId,
+                    dataType: dataType,
+                    window: window
+                )
+                readings.append(reading)
+            }
+        }
+
+        let hasConflict = detectConflict(in: readings, dataType: dataType)
+
+        return TimeWindowData(
+            timeWindow: window,
+            dataType: dataType,
+            readings: readings,
+            hasConflict: hasConflict
+        )
+    }
+
+    private func createReading(
+        from samples: [HKSample],
+        sourceId: String,
+        dataType: HealthDataType,
+        window: TimeWindow
+    ) -> SourceReading {
+        let value: Double
+        var secondaryValue: Double? = nil
+
+        switch dataType {
+        case .steps, .floorsClimbed, .activeEnergy, .distance:
+            // Sum up values for cumulative types
+            value = samples.compactMap { sample -> Double? in
+                guard let quantitySample = sample as? HKQuantitySample else { return nil }
+                return quantitySample.quantity.doubleValue(for: dataType.hkUnit)
+            }.reduce(0, +)
+
+        case .heartRate, .restingHeartRate, .respiratoryRate, .heartRateVariability, .bloodOxygen:
+            // Average for rate-based types
+            let values = samples.compactMap { sample -> Double? in
+                guard let quantitySample = sample as? HKQuantitySample else { return nil }
+                return quantitySample.quantity.doubleValue(for: dataType.hkUnit)
+            }
+            value = values.isEmpty ? 0 : values.reduce(0, +) / Double(values.count)
+
+        case .bloodPressureSystolic, .bloodPressureDiastolic:
+            // For blood pressure, we need to handle correlations
+            let values = samples.compactMap { sample -> Double? in
+                guard let quantitySample = sample as? HKQuantitySample else { return nil }
+                return quantitySample.quantity.doubleValue(for: dataType.hkUnit)
+            }
+            value = values.isEmpty ? 0 : values.reduce(0, +) / Double(values.count)
+
+        case .sleep:
+            // Sum up sleep duration
+            value = samples.reduce(0) { acc, sample in
+                acc + sample.endDate.timeIntervalSince(sample.startDate) / 3600
+            }
+        }
+
+        let source = sourceManager.sources.first { $0.bundleIdentifier == sourceId }
+        let category = source?.category ?? sourceManager.classifySource(sourceId)
+
+        return SourceReading(
+            sourceId: sourceId,
+            sourceName: source?.name ?? sourceId,
+            sourceCategory: category,
+            value: value,
+            secondaryValue: secondaryValue,
+            timestamp: window.start,
+            originalRecordId: samples.first?.uuid.uuidString,
+            quality: samples.isEmpty ? .missing : .complete
+        )
+    }
+
+    // MARK: - Conflict Detection
+
+    private func detectConflict(in readings: [SourceReading], dataType: HealthDataType) -> Bool {
+        // No conflict if less than 2 readings
+        guard readings.count >= 2 else { return false }
+
+        // Filter out zero values (device wasn't tracking)
+        let nonZeroReadings = readings.filter { $0.value > 0 }
+        guard nonZeroReadings.count >= 2 else { return false }
+
+        // Check if values differ significantly
+        let values = nonZeroReadings.map { $0.value }
+        guard let minVal = values.min(), let maxVal = values.max() else { return false }
+
+        // Threshold varies by data type
+        let threshold = conflictThreshold(for: dataType)
+
+        if minVal == 0 {
+            return maxVal > threshold.absoluteThreshold
+        }
+
+        let percentDiff = (maxVal - minVal) / minVal * 100
+        return percentDiff > threshold.percentageThreshold
+    }
+
+    private func conflictThreshold(for dataType: HealthDataType) -> ConflictThreshold {
+        switch dataType {
+        case .steps:
+            return ConflictThreshold(percentageThreshold: 10, absoluteThreshold: 100)
+        case .distance:
+            return ConflictThreshold(percentageThreshold: 10, absoluteThreshold: 0.1) // 100m
+        case .heartRate:
+            return ConflictThreshold(percentageThreshold: 15, absoluteThreshold: 10)
+        case .bloodPressureSystolic, .bloodPressureDiastolic:
+            return ConflictThreshold(percentageThreshold: 5, absoluteThreshold: 5)
+        case .bloodOxygen:
+            return ConflictThreshold(percentageThreshold: 2, absoluteThreshold: 2)
+        case .floorsClimbed:
+            return ConflictThreshold(percentageThreshold: 20, absoluteThreshold: 2)
+        case .activeEnergy:
+            return ConflictThreshold(percentageThreshold: 15, absoluteThreshold: 50)
+        case .sleep:
+            return ConflictThreshold(percentageThreshold: 10, absoluteThreshold: 0.5) // 30 min
+        case .restingHeartRate:
+            return ConflictThreshold(percentageThreshold: 10, absoluteThreshold: 5)
+        case .heartRateVariability:
+            return ConflictThreshold(percentageThreshold: 20, absoluteThreshold: 10)
+        case .respiratoryRate:
+            return ConflictThreshold(percentageThreshold: 15, absoluteThreshold: 2)
+        }
+    }
+
+    // MARK: - Detect All Conflicts
+
+    func detectConflicts(
+        for date: Date,
+        dataTypes: [HealthDataType] = HealthDataType.allCases
+    ) async throws -> [Conflict] {
+        let calendar = Calendar.current
+        let startOfDay = calendar.startOfDay(for: date)
+        let endOfDay = calendar.date(byAdding: .day, value: 1, to: startOfDay)!
+
+        var allConflicts: [Conflict] = []
+
+        for (index, dataType) in dataTypes.enumerated() {
+            readingProgress = Double(index) / Double(dataTypes.count)
+
+            do {
+                let windowData = try await fetchData(
+                    for: dataType,
+                    from: startOfDay,
+                    to: endOfDay
+                )
+
+                let conflicts = windowData
+                    .filter { $0.hasConflict }
+                    .map { data in
+                        Conflict(
+                            dataType: dataType,
+                            timeWindow: data.timeWindow,
+                            readings: data.readings,
+                            status: .pending
+                        )
+                    }
+
+                allConflicts.append(contentsOf: conflicts)
+            } catch {
+                print("Failed to detect conflicts for \(dataType): \(error)")
+            }
+        }
+
+        readingProgress = 1.0
+        detectedConflicts = allConflicts
+        return allConflicts
+    }
+
+    // MARK: - Data Gaps Detection
+
+    func detectGaps(
+        for dataType: HealthDataType,
+        from startDate: Date,
+        to endDate: Date,
+        expectedIntervalMinutes: Int = 15
+    ) async throws -> [DataGap] {
+        let samples = try await healthKitManager.fetchSamples(
+            for: dataType,
+            from: startDate,
+            to: endDate
+        )
+
+        guard !samples.isEmpty else {
+            return [DataGap(
+                dataType: dataType,
+                timeWindow: TimeWindow(start: startDate, end: endDate),
+                expectedRecordCount: 0,
+                actualRecordCount: 0
+            )]
+        }
+
+        let sortedSamples = samples.sorted { $0.startDate < $1.startDate }
+        var gaps: [DataGap] = []
+        let expectedInterval = TimeInterval(expectedIntervalMinutes * 60)
+
+        // Check gap at start
+        if let firstSample = sortedSamples.first,
+           firstSample.startDate.timeIntervalSince(startDate) > expectedInterval * 2 {
+            gaps.append(DataGap(
+                dataType: dataType,
+                timeWindow: TimeWindow(start: startDate, end: firstSample.startDate),
+                expectedRecordCount: Int(firstSample.startDate.timeIntervalSince(startDate) / expectedInterval),
+                actualRecordCount: 0
+            ))
+        }
+
+        // Check gaps between samples
+        for i in 0..<(sortedSamples.count - 1) {
+            let current = sortedSamples[i]
+            let next = sortedSamples[i + 1]
+            let gap = next.startDate.timeIntervalSince(current.endDate)
+
+            if gap > expectedInterval * 2 {
+                gaps.append(DataGap(
+                    dataType: dataType,
+                    timeWindow: TimeWindow(start: current.endDate, end: next.startDate),
+                    expectedRecordCount: Int(gap / expectedInterval),
+                    actualRecordCount: 0
+                ))
+            }
+        }
+
+        // Check gap at end
+        if let lastSample = sortedSamples.last,
+           endDate.timeIntervalSince(lastSample.endDate) > expectedInterval * 2 {
+            gaps.append(DataGap(
+                dataType: dataType,
+                timeWindow: TimeWindow(start: lastSample.endDate, end: endDate),
+                expectedRecordCount: Int(endDate.timeIntervalSince(lastSample.endDate) / expectedInterval),
+                actualRecordCount: 0
+            ))
+        }
+
+        return gaps
+    }
+
+    // MARK: - Aggregated Data
+
+    func fetchDailySummary(for date: Date) async throws -> DailySummary {
+        let calendar = Calendar.current
+        let startOfDay = calendar.startOfDay(for: date)
+        let endOfDay = calendar.date(byAdding: .day, value: 1, to: startOfDay)!
+
+        var summary = DailySummary(date: date)
+
+        for dataType in HealthDataType.allCases {
+            do {
+                let samples = try await healthKitManager.fetchSamples(
+                    for: dataType,
+                    from: startOfDay,
+                    to: endOfDay
+                )
+
+                let value = aggregateValue(samples: samples, dataType: dataType)
+                summary.values[dataType] = value
+
+                // Check for conflicts
+                let windowData = try await fetchData(
+                    for: dataType,
+                    from: startOfDay,
+                    to: endOfDay
+                )
+                let conflictCount = windowData.filter { $0.hasConflict }.count
+                summary.conflictCounts[dataType] = conflictCount
+            } catch {
+                print("Failed to fetch \(dataType) for summary: \(error)")
+            }
+        }
+
+        return summary
+    }
+
+    private func aggregateValue(samples: [HKSample], dataType: HealthDataType) -> Double {
+        switch dataType {
+        case .steps, .floorsClimbed, .activeEnergy, .distance:
+            return samples.compactMap { sample -> Double? in
+                guard let quantitySample = sample as? HKQuantitySample else { return nil }
+                return quantitySample.quantity.doubleValue(for: dataType.hkUnit)
+            }.reduce(0, +)
+
+        case .heartRate, .restingHeartRate, .respiratoryRate, .heartRateVariability,
+             .bloodOxygen, .bloodPressureSystolic, .bloodPressureDiastolic:
+            let values = samples.compactMap { sample -> Double? in
+                guard let quantitySample = sample as? HKQuantitySample else { return nil }
+                return quantitySample.quantity.doubleValue(for: dataType.hkUnit)
+            }
+            return values.isEmpty ? 0 : values.reduce(0, +) / Double(values.count)
+
+        case .sleep:
+            return samples.reduce(0) { acc, sample in
+                acc + sample.endDate.timeIntervalSince(sample.startDate) / 3600
+            }
+        }
+    }
+}
+
+// MARK: - Supporting Types
+
+struct TimeWindowData: Identifiable {
+    let id = UUID()
+    let timeWindow: TimeWindow
+    let dataType: HealthDataType
+    let readings: [SourceReading]
+    let hasConflict: Bool
+
+    var primaryReading: SourceReading? {
+        readings.max { $0.sourceCategory.priority < $1.sourceCategory.priority }
+    }
+
+    var conflictSeverity: ConflictSeverity? {
+        guard hasConflict, readings.count >= 2 else { return nil }
+
+        let values = readings.map { $0.value }.filter { $0 > 0 }
+        guard let min = values.min(), let max = values.max(), min > 0 else { return nil }
+
+        let percentDiff = (max - min) / min * 100
+
+        if percentDiff < 5 { return .minor }
+        if percentDiff < 20 { return .moderate }
+        if percentDiff < 50 { return .significant }
+        return .major
+    }
+}
+
+struct ConflictThreshold {
+    let percentageThreshold: Double
+    let absoluteThreshold: Double
+}
+
+struct DataGap: Identifiable {
+    let id = UUID()
+    let dataType: HealthDataType
+    let timeWindow: TimeWindow
+    let expectedRecordCount: Int
+    let actualRecordCount: Int
+
+    var severity: GapSeverity {
+        let duration = timeWindow.duration
+        if duration < 3600 { return .minor } // < 1 hour
+        if duration < 4 * 3600 { return .moderate } // < 4 hours
+        if duration < 12 * 3600 { return .significant } // < 12 hours
+        return .major
+    }
+
+    enum GapSeverity {
+        case minor, moderate, significant, major
+    }
+}
+
+struct DailySummary {
+    let date: Date
+    var values: [HealthDataType: Double] = [:]
+    var conflictCounts: [HealthDataType: Int] = [:]
+    var lastUpdated = Date()
+
+    var totalConflicts: Int {
+        conflictCounts.values.reduce(0, +)
+    }
+
+    func formattedValue(for dataType: HealthDataType) -> String {
+        guard let value = values[dataType] else { return "–" }
+
+        switch dataType {
+        case .steps, .floorsClimbed:
+            return String(format: "%.0f", value)
+        case .distance:
+            return String(format: "%.2f km", value)
+        case .heartRate, .restingHeartRate, .respiratoryRate:
+            return String(format: "%.0f %@", value, dataType.unit)
+        case .bloodPressureSystolic, .bloodPressureDiastolic:
+            return String(format: "%.0f mmHg", value)
+        case .bloodOxygen:
+            return String(format: "%.0f%%", value * 100)
+        case .activeEnergy:
+            return String(format: "%.0f kcal", value)
+        case .sleep:
+            let hours = Int(value)
+            let minutes = Int((value - Double(hours)) * 60)
+            return "\(hours)h \(minutes)min"
+        case .heartRateVariability:
+            return String(format: "%.0f ms", value)
+        }
+    }
+}