handle bom in phone number field

app/aws/s3.py

@@ -105,24 +105,20 @@ def extract_phones(job):
     current_app.logger.info(f"HEADERS {first_row}")
     phone_index = 0
     for item in first_row:
-        if item.lower() == "phone number":
+        # Note: may contain a BOM and look like \ufeffphone number
+        if "phone number" in item.lower():
             break
         phone_index = phone_index + 1
+
     phones = {}
     job_row = 0
     for row in job:
         row = row.split(",")
-        # TODO WHY ARE WE CALCULATING PHONE INDEX IN THE LOOP?
-        phone_index = 0
-        for item in first_row:
-            if item.lower() == "phone number":
-                break
-            phone_index = phone_index + 1
         current_app.logger.info(f"PHONE INDEX IS NOW {phone_index}")
         current_app.logger.info(f"LENGTH OF ROW IS {len(row)}")
         if phone_index >= len(row):
             phones[job_row] = "Error: can't retrieve phone number"
-            current_app.logger.error("Corrupt csv file, missing columns job_id {job_id} service_id {service_id}")
+            current_app.logger.error("Corrupt csv file")
         else:
             my_phone = row[phone_index]
             my_phone = re.sub(r"[\+\s\(\)\-\.]*", "", my_phone)

tests/app/aws/test_s3.py

@@ -67,6 +67,13 @@ def test_get_s3_file_makes_correct_call(notify_api, mocker):
             0,
             "15553333333",
         ),
+        (
+            # simulate file saved with utf8withbom
+            "\\ufeffPHONE NUMBER,Name\r\n5555555550,T 1\r\n5555555551,T 5,3/31/2024\r\n5555555552,T 2",
+            "eee",
+            2,
+            "5555555552",
+        ),
     ],
 )
 def test_get_phone_number_from_s3(