test: update unit tests to match refactoring

Author: BLMgithub

tests/test_assembly_stage.py

@@ -250,15 +250,14 @@ def test_merge_data_aggregates_duplicates(
 
 
 def test_derived_fields_correctness(valid_derived_df):
-    result = derive_fields(valid_derived_df, "20230101T120000")
+    result = derive_fields(valid_derived_df)
 
     if isinstance(result, pl.LazyFrame):
         result = result.collect()
 
     assert result["lead_time_days"].to_list() == [3, 5]
     assert result["approval_lag_days"].to_list() == [1, 1]
     assert result["delivery_delay_days"].to_list() == [1, 1]
-    assert result.select(pl.col("run_id").unique()).item() == "20230101T120000"
     assert "order_year_week" in result.columns
 
 
@@ -358,22 +357,14 @@ def test_assemble_data_fails_on_missing_column(
 def test_dimension_references_uniqueness():
     df = pl.DataFrame({"id": ["1", "1", "2"], "val": ["a", "a", "b"]})
 
-    result = dimension_references(df.lazy(), "test", ["id"], ["id", "val"])
+    result = dimension_references(df.lazy(), ["id"], ["id", "val"])
     if isinstance(result, pl.LazyFrame):
         result = result.collect()
     assert result.height == 2
 
     df_conflict = pl.DataFrame({"id": ["1", "1"], "val": ["a", "b"]})
 
-    result = dimension_references(df_conflict.lazy(), "test", ["id"], ["id", "val"])
+    result = dimension_references(df_conflict.lazy(), ["id"], ["id", "val"])
     if isinstance(result, pl.LazyFrame):
         result = result.collect()
     assert result.height == 1
-
-
-def test_dimension_references_fails_if_cols_missing():
-    df = pl.DataFrame({"id": ["1"]})
-    from polars.exceptions import ColumnNotFoundError
-
-    with pytest.raises((KeyError, ColumnNotFoundError)):
-        dimension_references(df.lazy(), "test", ["id"], ["id", "missing"])

tests/test_semantic_stage.py

@@ -166,21 +166,6 @@ def test_seller_semantic_model_grain_preserved_success(tmp_path, valid_assembled
     assert dim_df.height == expected_dim_len
 
 
-def test_seller_semantic_fails_on_multiple_run_ids(tmp_path, valid_assembled_df):
-    run_context = RunContext.create(base=tmp_path, run_id="20230101T120000")
-    # Clone and modify run_id
-    broken_df = valid_assembled_df.clone()
-    broken_df = broken_df.with_columns(
-        pl.when(pl.Series([False, True]))
-        .then(pl.lit("another_run").cast(pl.Categorical))
-        .otherwise(pl.col("run_id"))
-        .alias("run_id")
-    )
-
-    with pytest.raises(RuntimeError, match="Multiple run_ids detected"):
-        build_seller_semantic(broken_df.lazy(), run_context)
-
-
 # =============================================================================
 # BUILD BI SEMANTIC
 # =============================================================================
@@ -220,34 +205,6 @@ def test_build_semantic_layer_success(
             assert outputs_path.exists()
 
 
-def test_build_semantic_layer_fails_on_multiple_ids(tmp_path, valid_assembled_df):
-    run_id = "20230101T120000"
-    run_context = RunContext.create(base=tmp_path, run_id=run_id)
-    run_context.initialize_directories()
-
-    # Clone and modify run_id for Polars
-    broken_assembled = valid_assembled_df.clone()
-    broken_assembled = broken_assembled.with_columns(
-        pl.when(pl.Series([False, True]))
-        .then(pl.lit("another_run").cast(pl.Categorical))
-        .otherwise(pl.col("run_id"))
-        .alias("run_id")
-    )
-
-    broken_assembled.write_parquet(
-        run_context.assembled_path / "assembled_events_2023_01_01.parquet"
-    )
-
-    report = build_semantic_layer(run_context)
-
-    assert report["status"] == "failed"
-    assert (
-        report["modules"]["seller_semantic"]["seller_weekly_fact"]["build_stage"]
-        == False
-    )
-    assert any("Multiple run_ids detected" in error for error in report["errors"])
-
-
 def test_build_semantic_layer_fails_on_missing_columns(tmp_path, valid_assembled_df):
     run_id = "20230101T120000"
     run_context = RunContext.create(base=tmp_path, run_id=run_id)