Thank you for your pull request and welcome to our community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. In order for us to review and merge your code, please sign up at https://code.facebook.com/cla - and if you have received this in error or have any questions, please drop us a line at [email protected]. Thanks!

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Facebook open source project. Thanks!

I'd like to write two tests for the same screen where there's both the 2x and 3x device scale cases. Is the pull request required for that?

@lyahdav I think this is what you're looking for.

The idea here is that the developer can indeed define views with different content scale factors or layers with different contents scales that are needed by the same instance of the app, and test them on any simulator/device regardless of what that device's main screen's scale is.

So inside a FBSnapshotTestCase you could write something like:

 - (void)testViewScale1
 {
      UIView* view = [self createMyViewWithScale:1];
      self.manualScale = 1;
      FBSnapshotVerifyView(view, nil);
 }

 - (void)testViewScale3
 {
      UIView* view = [self createMyViewWithScale:3];
      self.manualScale = 3;
      FBSnapshotVerifyView(view, nil);
 }

The two tests will go look for reference images called respectively testViewScale1.png and [email protected], where the contents of the latter are the same of the former rendered at 3x3 times the resolution as you'd expect.

Hello @djrtl
I really like your addition. I am using your fork right now but I run into a problem.

I have a very simple test:

func testProfile() {
    let customSizeFrame = ...
    let myView = ...
    view.frame = customSizeFrame
    manualScale = 2
    FBSnapshotVerifyView(view)
}

However, when I run it in a @2x scale device with recordMode = true to generate the image, and then switch to a @3x device (iPhone Plus basically) turning off recordMode, the test fails. It is very strange because both images are exactly the same (both 640x1136):

reference image

failed image

diff

It seems like it is applying a wrong scale to the reference image when comparing both. Any idea how to fix this?

Thank you!

Hello @victorBaro,

I made this quite some time ago and I thought I had tested that situation, but then again maybe not. I'm sorry for that.

Right now it will be hard for me to look into this issue, I'll try to find the time in a few days.

In the meantime if you want to look into it yourself, you might want to take a look at where the comparisons between the pixel resolutions of the reference and new snapshot are made to try and understand why they fail in that use case, as my guess is that that is what is happening.

Hello @victorBaro,

I think my colleague might have fixed the issue you reported. Could you please let us know if the version that includes his changes works correctly for you?