Reject window captures that composite extra windows

ScreenCaptureKit's desktopIndependentWindow filter can include owned
child windows (e.g. an app dialog plus its parent), producing a capture
much larger than the picked window. Detect the overflow by comparing the
returned image size against the window's frame and fall back to the
screen crop with rounded mask. Hover changes are logged with the picked
window id and tracked count for diagnosis.

💘 Generated with Crush

Assisted-by: Crush:deepseek-v4-flash-0731

Sources/Kapture/CaptureController.swift

@@ -14,7 +14,8 @@ final class CaptureController: NSObject {
     // Selection state (driven by the event tap; AppKit global coordinates)
     private var marqueeStart: NSPoint?
     private(set) var marqueeRect: NSRect?
-    private(set) var hoverFrame: NSRect?
+    private(set) var hoverWindow: (id: CGWindowID, frame: NSRect)?
+    var hoverFrame: NSRect? { hoverWindow?.frame }
     private var trackedWindows: [(id: CGWindowID, frame: NSRect)] = []
     private var lastWindowScan = Date.distantPast
     private var eventTap: CFMachPort?
@@ -102,14 +103,15 @@ final class CaptureController: NSObject {
         }
     }
 
-    /// A window was picked from hover — capture it with rounded corners.
+    /// A window was picked from hover — capture it directly (real rounded
-    private func finishWindowPick(_ rect: NSRect) {
+    /// corners come from the window's own alpha).
+    private func finishWindowPick(_ windowID: CGWindowID, _ rect: NSRect) {
         let ids = overlayIDs
         hideOverlays()
         isSelecting = false
         restorePreviousApp()
         DispatchQueue.main.async { [self] in
-            capture(rect, excludingWindowIDs: ids, roundedCorners: true)
+            capture(rect, excludingWindowIDs: ids, windowID: windowID)
         }
     }
 
@@ -136,7 +138,7 @@ final class CaptureController: NSObject {
         overlays.forEach { $0.orderOut(nil) }
         overlays = []
         marqueeRect = nil
-        hoverFrame = nil
+        hoverWindow = nil
         marqueeStart = nil
     }
 
@@ -246,10 +248,10 @@ final class CaptureController: NSObject {
         return NSRect(x: rect.origin.x, y: y, width: rect.width, height: rect.height)
     }
 
-    private func topmostWindowContaining(_ appKitPoint: NSPoint) -> NSRect? {
+    private func topmostWindowContaining(_ appKitPoint: NSPoint) -> (id: CGWindowID, frame: NSRect)? {
-        for (_, frame) in trackedWindows {
+        for entry in trackedWindows {
-            if frame.contains(appKitPoint) {
+            if entry.frame.contains(appKitPoint) {
-                return frame
+                return entry
             }
         }
         return nil
@@ -282,9 +284,14 @@ final class CaptureController: NSObject {
             }
             let appKitPoint = cgPointToAppKit(NSPoint(x: event.location.x, y: event.location.y))
             if type == .mouseMoved {
-                let frame = topmostWindowContaining(appKitPoint)
+                let window = topmostWindowContaining(appKitPoint)
-                if frame != hoverFrame {
+                if window?.frame != hoverFrame {
-                    hoverFrame = frame
+                    hoverWindow = window
+                    if let w = window {
+                        CaptureController.log("hover: picked id=\(w.id) frame=\(w.frame) (tracked=\(trackedWindows.count))")
+                    } else {
+                        CaptureController.log("hover: none (tracked=\(trackedWindows.count))")
+                    }
                     renderOverlays()
                 }
                 return event
@@ -298,7 +305,7 @@ final class CaptureController: NSObject {
                 height: abs(appKitPoint.y - start.y)
             )
             marqueeRect = rect
-            hoverFrame = nil
+            hoverWindow = nil
             renderOverlays()
             return nil
         case .leftMouseDown:
@@ -313,9 +320,9 @@ final class CaptureController: NSObject {
             if let rect = marqueeRect, dragDistance >= 4, rect.width >= 2, rect.height >= 2 {
                 CaptureController.log("tap: marquee complete \(rect)")
                 finishSelection(rect)
-            } else if let hover = hoverFrame {
+            } else if let hover = hoverWindow {
-                CaptureController.log("tap: window pick \(hover)")
+                CaptureController.log("tap: window pick id=\(hover.id) \(hover.frame)")
-                finishWindowPick(hover)
+                finishWindowPick(hover.id, hover.frame)
             } else {
                 CaptureController.log("tap: click with no hover/marquee — cancel")
                 cancelSelection()
@@ -328,8 +335,8 @@ final class CaptureController: NSObject {
 
     // MARK: - Capture
 
-    private func capture(_ rect: NSRect, excludingWindowIDs ids: [Int], roundedCorners: Bool = false) {
+    private func capture(_ rect: NSRect, excludingWindowIDs ids: [Int], roundedCorners: Bool = false, windowID: CGWindowID? = nil) {
-        ScreenCapture.shared.capture(rect, excludingWindowIDs: ids, roundedCorners: roundedCorners) { [weak self] image in
+        ScreenCapture.shared.capture(rect, excludingWindowIDs: ids, roundedCorners: roundedCorners, windowID: windowID) { [weak self] image in
             DispatchQueue.main.async {
                 self?.isSelecting = false
                 self?.restorePreviousApp()
@@ -414,7 +421,7 @@ final class ScreenCapture {
         return content.displays.first(where: { $0.displayID == cgID }) ?? content.displays.first
     }
 
-    func capture(_ rect: NSRect, excludingWindowIDs excludedIDs: [Int], roundedCorners: Bool = false, completion: @escaping (CGImage?) -> Void) {
+    func capture(_ rect: NSRect, excludingWindowIDs excludedIDs: [Int], roundedCorners: Bool = false, windowID: CGWindowID? = nil, completion: @escaping (CGImage?) -> Void) {
         if #available(macOS 14.0, *) {
             Task {
                 guard let content = try? await SCShareableContent.current,
@@ -449,6 +456,37 @@ final class ScreenCapture {
                     height: rect.height
                 )
 
+                // Window pick: capture the window DIRECTLY — the image comes
+                // with the app's real rounded-corner alpha (no mask guessing).
+                if let windowID,
+                   let scWindow = content.windows.first(where: { $0.windowID == windowID }) {
+                    var windowConfig = SCStreamConfiguration()
+                    windowConfig.width = Int(scWindow.frame.width * backingScale)
+                    windowConfig.height = Int(scWindow.frame.height * backingScale)
+                            windowConfig.showsCursor = false
+                    if let windowImage = try? await SCScreenshotManager.captureImage(
+                        contentFilter: SCContentFilter(desktopIndependentWindow: scWindow),
+                        configuration: windowConfig
+                    ) {
+                        let expectedW = CGFloat(scWindow.frame.width) * backingScale
+                        let expectedH = CGFloat(scWindow.frame.height) * backingScale
+                        let overflow = max(
+                            CGFloat(windowImage.width) / max(expectedW, 1),
+                            CGFloat(windowImage.height) / max(expectedH, 1)
+                        )
+                        if overflow > 1.35 {
+                            // SCKit composited extra windows (owned popups etc.)
+                            // — reject and use the screen crop instead.
+                            CaptureController.log("capture: WINDOW id=\(windowID) overflow=\(overflow) (\(windowImage.width)x\(windowImage.height) vs \(Int(expectedW))x\(Int(expectedH))) — rejecting, using crop+mask")
+                        } else {
+                            CaptureController.log("capture: WINDOW id=\(windowID) image=\(windowImage.width)x\(windowImage.height) (native alpha corners)")
+                            completion(windowImage)
+                            return
+                        }
+                    }
+                    CaptureController.log("capture: window filter failed for id=\(windowID) — falling back to crop+mask")
+                }
+
                 guard let image = try? await SCScreenshotManager.captureImage(contentFilter: filter, configuration: config) else {
                     completion(nil)
                     return

build-app.sh

@@ -108,9 +108,9 @@ cat > "$APP/Contents/Info.plist" <<'PLIST'
     <key>CFBundleIconFile</key>
     <string>Icon</string>
     <key>CFBundleShortVersionString</key>
-    <string>1.7.4</string>
+    <string>1.8.1</string>
     <key>CFBundleVersion</key>
-    <string>74</string>
+    <string>81</string>
     <key>LSUIElement</key>
     <true/>
     <key>NSHighResolutionCapable</key>