bugfix/missing-comma-was-creating-a-new-file-format

[evamaxfield]

Description of Changes

Found while working on napari stuff which tests for png support. If you try to read a PNG directly with this package's Reader, it works! If you try to read a PNG with BioImage, it fails because it doesn't have a plugin listing for "png" but does have one for "pngppm" because if there isn't a comma, Python will combine strings together 🙃🙃🙃.

[evamaxfield]

I am also mirroring the change made here: AllenCellModeling/aicsimageio#543

In which imageio was bumped to a higher version to handle various ffmpeg differences

[toloudis]

I wonder if we can get the tests back to green

[evamaxfield]

I wonder if we can get the tests back to green

Was trying to before I got blocked on time :/ will return to this in ~2 weeks.

[TimMonko]

Any way I can help this get through? I'm lazy and weary of using other libraries just for png's.
The only CI failure I see is AssertionError [ERR_ASSERTION]: Unsupported architecture (only x64 is supported) for MacOS. But maybe the logs have expired?
Cheers-

[BrianWhitneyAI]

why'd we lose a scene here?

[toloudis]

S is not scene, it's "Samples", as in 3 means RGB and 4 means RGBA

[BrianWhitneyAI]

oh gotcha my mistake. I think my question persists though. What happened to our 4th sample?

[BrianWhitneyAI]

for example.gif we have an issue with the reader here.

            elif image_length > 1:
                # Read and stack all frames
                frames = []
                for frame in reader:
                    frames.append(frame)

                image_data = np.stack(frames)

We have a first frame that is uneven

[BrianWhitneyAI]

@evamaxfield @toloudis any thoughts here? np.stack expects same-size arrays, not sure why the first one is missing a dim.

[BrianWhitneyAI]

Any way I can help this get through? I'm lazy and weary of using other libraries just for png's. The only CI failure I see is AssertionError [ERR_ASSERTION]: Unsupported architecture (only x64 is supported) for MacOS. But maybe the logs have expired? Cheers-

I reran the tests and we seem to be having an issue with the first frame of image data here .

[TimMonko]

I tried to run tests locally from instructions in contributing.md (just build). My goal was to compare the present example.gif to the aicsimageio example.gif since they appear to be in different servers. However, I get this error:

[INFO: download_test_resources:  79 2024-12-12 11:34:01,094] Downloading test resources using top hash: 3fba62f10af9b4b85dfe07bffdc526a6f8efaea1b7ee7d06f9e9e5613e8339fc
[ERROR: download_test_resources:  95 2024-12-12 11:34:02,244] =============================================
[ERROR: download_test_resources:  99 2024-12-12 11:34:02,246]

S3 AccessDenied for S3Api.HEAD_OBJECT on bucket: bioio-dev-test-resources

[ERROR: download_test_resources: 100 2024-12-12 11:34:02,246] =============================================

Individually running download_test_resources.py also gives this error. Is access to the S3 bucket public? Perhaps there is some other setting I need to do. I have never worked with AWS before.

[BrianWhitneyAI]

I tried to run tests locally from instructions in contributing.md (just build). My goal was to compare the present example.gif to the aicsimageio example.gif since they appear to be in different servers. However, I get this error:

[INFO: download_test_resources:  79 2024-12-12 11:34:01,094] Downloading test resources using top hash: 3fba62f10af9b4b85dfe07bffdc526a6f8efaea1b7ee7d06f9e9e5613e8339fc
[ERROR: download_test_resources:  95 2024-12-12 11:34:02,244] =============================================
[ERROR: download_test_resources:  99 2024-12-12 11:34:02,246]

S3 AccessDenied for S3Api.HEAD_OBJECT on bucket: bioio-dev-test-resources

[ERROR: download_test_resources: 100 2024-12-12 11:34:02,246] =============================================

Individually running download_test_resources.py also gives this error. Is access to the S3 bucket public? Perhaps there is some other setting I need to do. I have never worked with AWS before.

We are working on this right now with our infrastructure team. They are not currently public, though if you open a branch you can run tests from there with the approval of a "Trusted Developer". Hopefully in the next few weeks, these files will all be public...