add backup test

Author: minhkhul

nchs_mortality/delphi_nchs_mortality/pull.py

@@ -73,8 +73,9 @@ def pull_nchs_mortality_data(
         results = client.get("r8kw-7aab", limit=10**10)
         df = pd.DataFrame.from_records(results)
 
-        create_backup_csv(df, backup_dir, custom_run=custom_run, logger=logger)
+    create_backup_csv(df, backup_dir, custom_run=custom_run, logger=logger)
 
+    if not test_file:
         # drop "By Total" rows
         df = df[df["group"].transform(str.lower) == "by week"]
 

nchs_mortality/tests/conftest.py

@@ -14,6 +14,7 @@
 
 PARAMS = {
   "common": {
+    "custom_run": True,
     "daily_export_dir": "./daily_receiving",
     "weekly_export_dir": "./receiving",
     "backup_dir": "./raw_data_backups"

nchs_mortality/tests/test_pull.py

@@ -1,3 +1,4 @@
+import os
 import pytest
 
 import pandas as pd
@@ -96,3 +97,14 @@ def test_bad_file_with_inconsistent_time_col(self):
     def test_bad_file_with_missing_cols(self):
         with pytest.raises(ValueError):
             pull_nchs_mortality_data(SOCRATA_TOKEN, backup_dir = "", custom_run = True, test_file = "bad_data_with_missing_cols.csv")
+
+    def test_backup_today_data(self):
+        today = pd.Timestamp.today().strftime("%Y%m%d")
+        backup_dir = "./raw_data_backups"
+        pull_nchs_mortality_data(SOCRATA_TOKEN, backup_dir = backup_dir, custom_run = False, test_file = "test_data.csv")
+        backup_file = f"{backup_dir}/{today}.csv.gz"
+        backup_df = pd.read_csv(backup_file)
+        source_df = pd.read_csv("test_data/test_data.csv")
+        pd.testing.assert_frame_equal(source_df, backup_df)
+        if os.path.exists(backup_file):
+            os.remove(backup_file)