From 1b76fc9127efe7de3c13e386a930b20234002f6f Mon Sep 17 00:00:00 2001
From: "google-labs-jules[bot]"
 <161369871+google-labs-jules[bot]@users.noreply.github.com>
Date: Sun, 28 Jun 2026 15:03:00 +0000
Subject: [PATCH] Add test for CDCSodaExtractor.save_to_csv

Co-authored-by: Data-Science-Link <61164085+Data-Science-Link@users.noreply.github.com>
---
 .../cdc_api/test_soda_extractor.py            | 37 +++++++++++++++++++
 1 file changed, 37 insertions(+)
 create mode 100644 data_engineering/data_sources/cdc_api/test_soda_extractor.py

diff --git a/data_engineering/data_sources/cdc_api/test_soda_extractor.py b/data_engineering/data_sources/cdc_api/test_soda_extractor.py
new file mode 100644
index 0000000..55961a8
--- /dev/null
+++ b/data_engineering/data_sources/cdc_api/test_soda_extractor.py
@@ -0,0 +1,37 @@
+import pandas as pd
+from pathlib import Path
+from data_engineering.data_sources.cdc_api.soda_extractor import CDCSodaExtractor
+
+def test_save_to_csv(tmp_path: Path):
+    """Test saving a DataFrame to a CSV file."""
+    extractor = CDCSodaExtractor()
+
+    # Create a dummy DataFrame
+    df = pd.DataFrame({
+        "indicator": ["Synthetic opioids, excl. methadone (T40.4)", "Synthetic opioids, excl. methadone (T40.4)"],
+        "year": [2021, 2022],
+        "deaths": [1000, 1200]
+    })
+
+    # Define a test output path inside a non-existent subdirectory
+    # to test parent directory creation
+    output_path = tmp_path / "test_dir" / "test_output.csv"
+
+    # Save to CSV
+    extractor.save_to_csv(df, output_path)
+
+    # Assert the file was created
+    assert output_path.exists()
+    assert output_path.is_file()
+
+    # Read back the saved CSV and verify its contents
+    saved_df = pd.read_csv(output_path)
+
+    # Verify shape and columns
+    assert len(saved_df) == 2
+    assert list(saved_df.columns) == ["indicator", "year", "deaths"]
+
+    # Verify data types and values
+    assert saved_df["year"].tolist() == [2021, 2022]
+    assert saved_df["deaths"].tolist() == [1000, 1200]
+    assert saved_df["indicator"].tolist() == ["Synthetic opioids, excl. methadone (T40.4)", "Synthetic opioids, excl. methadone (T40.4)"]