[CHORE] Remove usage of credentials in workflows to allow PRs from fo…

…rks to be green

.github/workflows/python-package.yml

@@ -258,11 +258,6 @@ jobs:
         python-version: ['3.8'] # can't use 3.7 due to requiring anon mode for adlfs
         daft-runner: [py, ray]
         micropartitions: [1, 0]
-    # These permissions are needed to interact with GitHub's OIDC Token endpoint.
-    # This is used in the step "Assume GitHub Actions AWS Credentials"
-    permissions:
-      id-token: write
-      contents: read
     steps:
     - uses: actions/checkout@v4
       with:
@@ -295,20 +290,6 @@ jobs:
       run: |
         mkdir -p /tmp/daft-integration-testing/nginx
         chmod +rw /tmp/daft-integration-testing/nginx
-    - name: Assume GitHub Actions AWS Credentials
-      uses: aws-actions/configure-aws-credentials@v4
-      with:
-        aws-region: us-west-2
-        role-to-assume: ${{ secrets.ACTIONS_AWS_ROLE_ARN }}
-        role-session-name: DaftPythonPackageGitHubWorkflow
-    - name: Assume GitHub Actions GCloud Credentials
-      uses: google-github-actions/auth@v1
-      with:
-        credentials_json: ${{ secrets.ACTIONS_GCP_SERVICE_ACCOUNT_JSON }}
-        # NOTE: Workload Identity seems to be having problems with our Rust crate, so we use JSON instead
-        # See issue: https://github.com/yoshidan/google-cloud-rust/issues/171#issuecomment-1730511655
-        # workload_identity_provider: ${{ secrets.ACTIONS_GCP_WORKLOAD_IDENTITY_PROVIDER }}
-        # service_account: ${{ secrets.ACTIONS_GCP_SERVICE_ACCOUNT }}
     - name: Spin up IO services
       uses: isbang/[email protected]
       with:

tests/integration/io/conftest.py

@@ -41,6 +41,8 @@ def aws_public_s3_config() -> daft.io.IOConfig:
         s3=daft.io.S3Config(
             # NOTE: no keys or endpoints specified for an AWS public s3 bucket
             region_name="us-west-2",
+            # Use anonymous mode to avoid having to search for credentials in the Github Runner
+            anonymous=True,
         )
     )
 

tests/integration/io/test_list_files_gcs.py

@@ -6,7 +6,6 @@
 from daft.daft import GCSConfig, IOConfig, io_glob
 
 BUCKET = "daft-public-data-gs"
-DEFAULT_GCS_CONFIG = GCSConfig(project_id=None, anonymous=None)
 ANON_GCS_CONFIG = GCSConfig(project_id=None, anonymous=True)
 
 
@@ -52,29 +51,26 @@ def compare_gcs_result(daft_ls_result: list, fsspec_result: list):
 )
 @pytest.mark.parametrize("recursive", [False, True])
 @pytest.mark.parametrize("fanout_limit", [None, 1])
-@pytest.mark.parametrize("gcs_config", [DEFAULT_GCS_CONFIG, ANON_GCS_CONFIG])
-def test_gs_flat_directory_listing(path, recursive, gcs_config, fanout_limit):
+def test_gs_flat_directory_listing(path, recursive, fanout_limit):
     fs = gcsfs.GCSFileSystem()
     glob_path = path.rstrip("/") + "/**" if recursive else path
-    daft_ls_result = io_glob(glob_path, io_config=IOConfig(gcs=gcs_config), fanout_limit=fanout_limit)
+    daft_ls_result = io_glob(glob_path, io_config=IOConfig(gcs=ANON_GCS_CONFIG), fanout_limit=fanout_limit)
     fsspec_result = gcsfs_recursive_list(fs, path) if recursive else fs.ls(path, detail=True)
     compare_gcs_result(daft_ls_result, fsspec_result)
 
 
 @pytest.mark.integration()
 @pytest.mark.parametrize("recursive", [False, True])
-@pytest.mark.parametrize("gcs_config", [DEFAULT_GCS_CONFIG, ANON_GCS_CONFIG])
-def test_gs_single_file_listing(recursive, gcs_config):
+def test_gs_single_file_listing(recursive):
     path = f"gs://{BUCKET}/test_ls/file.txt"
     fs = gcsfs.GCSFileSystem()
-    daft_ls_result = io_glob(path, io_config=IOConfig(gcs=gcs_config))
+    daft_ls_result = io_glob(path, io_config=IOConfig(gcs=ANON_GCS_CONFIG))
     fsspec_result = gcsfs_recursive_list(fs, path) if recursive else fs.ls(path, detail=True)
     compare_gcs_result(daft_ls_result, fsspec_result)
 
 
 @pytest.mark.integration()
-@pytest.mark.parametrize("gcs_config", [DEFAULT_GCS_CONFIG, ANON_GCS_CONFIG])
-def test_gs_notfound(gcs_config):
+def test_gs_notfound():
     path = f"gs://{BUCKET}/test_"
     with pytest.raises(FileNotFoundError, match=path):
-        io_glob(path, io_config=IOConfig(gcs=gcs_config))
+        io_glob(path, io_config=IOConfig(gcs=ANON_GCS_CONFIG))