chore: cleanup tests since interface changes

tests/core/test_configuration.py

@@ -17,20 +17,4 @@ def test_load_config():
     project = project.unwrap()
 
     # Project config can be indexed directly, this gets the project name
-    assert project["name"] == "cdf-test"
-
-    # Namespaced access by indexing with a tuple (workspace, key)
-    assert project[("workspace1", "name")] == "workspace1"
-
-    # Namespaced access falls back to the project settings
-    # in this way, workspace settings can override project settings if needed
-    # This is the expected access pattern
-    assert project[("workspace1", "filesystem.provider")] == "local"
-
-    # Access by attribute works too, but does not support the
-    # behavior of falling back to the project settings since we are directly accessing
-    # the workspace settings
-    assert (
-        project[("workspace1", "pipelines.us_cities")]
-        == project.workspace1.pipelines.us_cities
-    )
+    assert project["project"]["name"] == "cdf-test"

tests/core/test_project.py

@@ -1,83 +1 @@
-import pathlib
-import typing as t
 
-import dlt
-import dlt.common.configuration
-
-import cdf.core.context as context
-from cdf.core.configuration import load_config
-from cdf.core.feature_flag import load_feature_flag_provider
-from cdf.core.filesystem import load_filesystem_provider
-
-
-# Q: Is a Project a subclass of a Workspace? or vice versa?
-#    No, instead we rely on mixins which are shared between Project & Workspace
-def test_project_ideal_interface():
-    r = (
-        # A project class is created by get_project
-        # This class defines the top-level configuration shared by all workspaces
-        # it also injects itself into the dlt config providers container until __del__
-        get_project("examples/sandbox")
-        # A workspace class is created by get_workspace
-        # This class has config defined as a ChainMap of the workspace and project settings
-        # It also stores its root path
-        >> (lambda project: project.get_workspace("workspace1"))
-        # We can then access the pipeline by name
-        # This returns a PipelineSpecification object
-        # and leverages the root path and workspace settings
-        >> (lambda workspace: workspace.get_pipeline("us_cities"))
-        # The run method sets all the necessary context variables such that
-        # the cdf.pipeline function will function in the script as expected
-        >> (lambda pipeline: pipeline.run())
-    )
-
-
-def test_project_interface():
-    project = load_config("examples/sandbox").unwrap()
-
-    context.active_workspace.set("workspace1")
-    context.inject_cdf_config_provider(project)
-
-    source_config = project["pipelines"]["us_cities"]
-
-    @dlt.common.configuration.with_config(auto_pipeline_section=True)
-    def foo(
-        pipeline_name: str, x: int = dlt.config.value, y: int = dlt.config.value
-    ) -> t.Tuple[int, int]:
-        return (x, y)
-
-    assert foo("us_cities") == (100, 2)
-
-    dlt.config.config_providers[-1].set_value("test123", 123, "us_cities")
-    assert dlt.config["pipelines.us_cities.options.test123"] == 123
-
-    dlt.config.config_providers[-1].set_value("test123", 123, "")
-    assert dlt.config["test123"] == project["test123"]
-
-    # set in pipeline options, which is very interesting
-    pipeline = dlt.pipeline("us_cities")
-    assert pipeline.runtime_config["dlthub_telemetry"] is False
-    assert pipeline.destination.destination_type.endswith("duckdb")
-
-    ff = load_feature_flag_provider("file", options={"path": "feature_flags.json"})
-    fs = load_filesystem_provider("file", options={"compress": True})
-
-    p = pipeline("us_cities", source_config)  # TODO: we need the root path?
-
-
-def test_project():
-    from cdf.core.project import get_project
-
-    pipeline_ = (
-        get_project("examples/sandbox")
-        .bind(lambda project: project.get_workspace("workspace1"))
-        .bind(lambda workspace: workspace.get_pipeline("us_cities"))
-        .unwrap()
-    )
-    # TODO: add destination? and dataset?
-    # oh well this is a pipeline specification
-    # we need to investigate the cdf.pipeline entrypoint
-    # but thats in user-code, so we need to communicate with contextvars
-    # So maybe spec.main() *args and **kwargs are stored in contextvars
-    # and we pull them out in the cdf.pipeline's .run() method and equivalents
-    pipeline_.main(destination="...")