S3 bucket - Failed to get directory contents - No such file or directory: test-subfolder

[gregorycottone]

After properly set the S3 bucket in the Advanced Settings Editor:

{
  "resources": [
    {
      "name": "s3 bucket",
      "url": "s3://test-bucket"
    }
  ]
}

The jupyter-fs icon appears on the left bar of the jupyterlab UI, then, if I click it, I can properly see the files and subfolders inside the bucket.

The issue appears when I try to open one of the subfolders inside the S3 bucket.
Indeed what I get is a "Failed to get directory contents - No such file or directory: test-subfolder" error message.

So, let's suppose I have an S3 bucket called "test-bucket" that contains one file called "test-file" and one subfolder called "test-subfolder", I'm able to open the "test-bucket" via jupyter-fs icon, I'm also able to open the "test-file", but I'm not able to open the "test-subfolder", indeed when I click in the + icon next to the "test-subfolder", I get the message:

"Failed to get directory contents - No such file or directory: test-subfolder"

By checking the networking tab of the inspect tools of the browser I see a 404 Not Found error:

Request URL: https://my-domain-name/user/user-name/api/contents/4eba04d6%3Atest-subfolder?content=1&hash=0&1737376389969
Request Method: GET
Status Code: 404 Not Found
Response body: {"message": "No such file or directory: test-subfolder", "reason": null}

Now, I'm sure that the "test-subfolder" folder from the "test-bucket" S3 bucket contains files, either because I have access to the S3 bucket and I can see what's inside from the AWS S3 console, but also because if I run the following code snippet from a notebook running on jupyterlab:

import boto3

bucket_name = 'test-bucket'
s3 = boto3.client('s3')
response = s3.list_objects_v2(Bucket=bucket_name)
print(response)

I can print the list of all the files inside the "test-bucket" bucket.

Also, if I run this command from the terminal of jupyterlab: aws s3 ls s3://test-bucket/ --recursive I can see the list of all the files stored inside the "test-subfolder" folder.

A last important information is that if I set the following configuration in the Advance Settings Editor:

{
  "resources": [
    {
      "name": "s3 bucket",
      "url": "s3://test-bucket/test-subfolder/"
    }
  ]
}

Then I can see the list of files contained in the S3 bucket from the jupyter-fs icon.

Here a screenshot of the error I get (note the name of the subfolder is not "test-subfolder", but "textract_output".

Desktop (please complete the following information):

• OS of the machine running jupyter: [Amazon Linux 2023]
• Browser [Microsoft Edge]
• Version [jupyter-fs==1.0.0.post1]