fix(image_layers) image layers crash on Android

[ekuleshov]

Description

Related Issue: Closes #118

Type of Change

• ✨ New feature (non-breaking change which adds functionality)
• 🛠️ Bug fix (non-breaking change which fixes an issue)
• ❌ Breaking change (fix or feature that would cause existing functionality to change)
• 🧹 Code refactor
• ✅ Build configuration change
• 📝 Documentation
• 🗑️ Chore

[hm21]

Thanks for the PR! I looked into the Bitmap.scale() extension from androidx.core.graphics it delegates to Bitmap.createScaledBitmap(), which returns the same object when width and height already match. So calling recycle() on that would indeed crash. The defensive !== check in change 1 (line ~92) is a valid fix for that edge case.

However, I want to point out that change 2 (line ~110) is logically redundant, the scale() call is already inside an if block that checks sizedBitmap.width != videoWidth || sizedBitmap.height != videoHeight, so at least one dimension always differs, meaning createScaledBitmap will always return a new object there. The added guard can never trigger. It's harmless but i think it's unnecessary. Could you try remove it and check if it's still crash then?

[ekuleshov]

However, I want to point out that change 2 (line ~110) is logically redundant, the scale() call is already inside an if block that checks sizedBitmap.width != videoWidth || sizedBitmap.height != videoHeight...

You are right. Though apparently the check for "same image" in the scale() code is a bit more elaborate than the outer condition here. So, I figured that removing outer condition and let scale() deal with the same size is cleaner approach, if you are okay with that.

Also noticed that generate thumbnails code doesn't do recycle() on interim bitmaps either. I've pushed a chenge to this PR as well.

Let me know if you don't want these changes. I'll revert.

[hm21]

LGTM and thanks for the update