Merge pull request #631 from CodeForPhilly/630-sanitize-phone-numbers

sanitize numbers

Author: c-simpson

src/server/models.py

@@ -20,6 +20,7 @@
 from sqlalchemy.dialects.postgresql import JSONB, insert
 from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.sql.functions import coalesce
+from utils import standardize_phone_number
 
 Base = declarative_base()
 
@@ -92,21 +93,6 @@ def dedup_consecutive(table, unique_id, id, order_by, dedup_on):
     return delete(table).where(unique_id == to_delete.c[0])
 
 
-def normalize_phone_number(number):
-    result = None
-
-    if number and str(number) != "nan":
-        number = re.sub("[() -.+]", "", str(number))
-
-        if number and number[0] == "1":
-            number = number[1:]
-
-        if number.isdigit() and len(number) == 10:
-            result = number
-
-    return result
-
-
 class PdpContacts(Base):
     __tablename__ = "pdp_contacts"
     __table_args__ = (
@@ -173,8 +159,10 @@ def insert_from_file_df(cls, df, conn):
         df = df[column_translation.keys()]
         df = df.rename(columns=column_translation)
 
-        df["phone"] = df["phone"].apply(normalize_phone_number)
-        df["mobile"] = df["mobile"].apply(normalize_phone_number)
+        phone_numbers = [standardize_phone_number(phone) for phone in df["phone"]]
+        mobile_numbers = [standardize_phone_number(phone) for phone in df["mobile"]]
+        df["phone"] = phone_numbers
+        df["mobile"] = mobile_numbers
 
         dedup_on = [col for col in cls.__table__.columns if col.name in df.columns]
         df["created_date"] = datetime.datetime.utcnow()
@@ -237,7 +225,8 @@ def insert_from_df(cls, df, conn):
         df = df[column_translation.keys()]
         df = df.rename(columns=column_translation)
 
-        df["phone"] = df["phone"].apply(normalize_phone_number)
+        phone_numbers = [standardize_phone_number(phone) for phone in df["phone"]]
+        df["phone"] = phone_numbers
 
         dedup_on = [col for col in cls.__table__.columns if col.name in df.columns]
         df["created_date"] = datetime.datetime.utcnow()

src/server/utils.py

@@ -0,0 +1,28 @@
+import re
+
+def standardize_phone_number(phone):
+    """Standardize phone number format.
+    
+    Args:
+        phone (str): The phone number to standardize.
+    
+    Returns:
+        str: The standardized phone number.
+    """
+    # Remove all non-numeric characters
+    phone = re.sub(r'\D', '', phone)
+
+    # if the phone number is less than 10 digits, it's invalid
+    if len(phone) < 10:
+        return None
+
+    # If the phone number is exactly 10 digits, return as is
+    if len(phone) == 10:
+        return phone
+
+    # if the phone number is greater than 10 digits, take the last 10 digits
+    if len(phone) > 10:
+        return f'{phone[-10:]}'
+
+    # anything else we ignore
+    return None

src/server/volgistics_importer.py

@@ -1,10 +1,10 @@
-import re
 from flask.globals import current_app
 from datetime import datetime, timedelta 
 from openpyxl import load_workbook
 from jellyfish import jaro_similarity
 
 from config import  engine
+from utils import standardize_phone_number
 
 import structlog
 
@@ -178,6 +178,10 @@ def volgistics_people_import(workbook):
     col_email = col['Email']
     time_stamp = datetime.utcnow()
 
+    home_phone = standardize_phone_number(r[col_home])
+    work_phone = standardize_phone_number(r[col_work])
+    cell_phone = standardize_phone_number(r[col_cell])
+
     try:
         for r in ws.iter_rows(min_row=2, max_col=42,values_only=True):
             insert_list.append(
@@ -194,9 +198,9 @@ def volgistics_people_import(workbook):
                     "state": r[col_state],
                     "zip": r[col_zip],
                     "all_phone_numbers": r[col_all_phones],
-                    "home": r[col_home],
-                    "work": r[col_work],
-                    "cell": r[col_cell],
+                    "home": home_phone,
+                    "work": work_phone,
+                    "cell": cell_phone,
                     "email": r[col_email],
                     "created_date" : time_stamp
                 }