fix: replace empty '' by numpy.nan on column for google-sheet [TCTC-6…

…530] (#1228)

tests/google_sheets/test_google_sheets.py

@@ -5,6 +5,7 @@
 from datetime import datetime
 from os import path
 
+import numpy as np
 import pandas as pd
 from googleapiclient.http import HttpMock
 from pandas import DataFrame
@@ -15,6 +16,8 @@
 from toucan_connectors.google_sheets.google_sheets_connector import (
     GoogleSheetsConnector,
     GoogleSheetsDataSource,
+    parse_cell_value,
+    serial_number_to_date,
 )
 from toucan_connectors.json_wrapper import JsonWrapper
 
@@ -298,3 +301,35 @@ def test_get_form(mocker):
     )
     expected_results = ['sample data', 'animals']
     assert schema['definitions']['sheet']['enum'] == expected_results
+
+
+def test_numeric_dateformat_():
+    value = 44303  # Example serial number representing a date
+    format_ = {'numberFormat': {'type': 'DATE'}}
+    result = parse_cell_value(value, format_)
+    assert result == serial_number_to_date(value)
+
+
+def test_empty_string():
+    value = ''
+    format_ = {}
+    result = parse_cell_value(value, format_)
+    assert np.isnan(result)
+
+
+def test_other_types():
+    value = 'hello'
+    format_ = {'numberFormat': {'type': 'DATE'}}
+    result = parse_cell_value(value, format_)
+    assert result == value
+
+    value = 123
+    format_ = {}
+    result = parse_cell_value(value, format_)
+    assert result == value
+
+
+def test_default_format():
+    value = 44303  # Example serial number representing a date
+    result = parse_cell_value(value)
+    assert result == value

toucan_connectors/google_sheets/google_sheets_connector.py

@@ -2,6 +2,7 @@
 from datetime import datetime
 from typing import Any, Callable, Dict, List, Optional, Type
 
+import numpy as np
 import pandas as pd
 from dateutil.relativedelta import relativedelta
 from google.oauth2.credentials import Credentials
@@ -214,17 +215,19 @@ def serial_number_to_date(serial_number: float) -> datetime:
     return SERIAL_REFERENCE_DAY + relativedelta(days=int(serial_number))
 
 
-def parse_cell_value(value, format):
+def parse_cell_value(value: Any, format_: dict[str, Any] | None = None) -> Any:
     """
     Use the format (if provided) to parse the value in its intended type
     """
     if (
-        (type(value) is int or type(value) is float)
-        and format
-        and 'numberFormat' in format
-        and format['numberFormat']['type'] == 'DATE'
+        isinstance(value, (int, float))
+        and format_ is not None
+        and format_.get('numberFormat', {}).get('type') == 'DATE'
     ):
         return serial_number_to_date(value)
+    elif isinstance(value, str) and value == '':
+        return np.nan
+
     return value