refactor: replace regex datetime parsing with fromisoformat, closes #191

CHANGELOG.md

@@ -7,6 +7,7 @@ and this project adheres to [Semantic Versioning](http://semver.org/).
 ## [Unreleased]
 
 - service: let FunctionRequests return a list of EntityProxies instead of the raw json, when the `ReturnType` is a Collection. - Emil B.
+- model: replace regexp-based ISO datetime parsing with `datetime.fromisoformat` for `Edm.DateTime` and `Edm.DateTimeOffset` - Petr Hanak
 
 ## [1.11.2]
 

pyodata/v2/model.py

@@ -396,15 +396,14 @@ def ms_since_epoch_to_datetime(value, tzinfo):
 
 def parse_datetime_literal(value):
     try:
-        return datetime.datetime.strptime(value, '%Y-%m-%dT%H:%M:%S.%f')
+        # fromisoformat on Python < 3.11 requires fractional seconds to be exactly
+        # 3 or 6 digits; pad to 6 if needed
+        if '.' in value:
+            dt_part, frac = value.rsplit('.', 1)
+            value = f'{dt_part}.{frac.ljust(6, "0")}'
+        return datetime.datetime.fromisoformat(value)
     except ValueError:
-        try:
-            return datetime.datetime.strptime(value, '%Y-%m-%dT%H:%M:%S')
-        except ValueError:
-            try:
-                return datetime.datetime.strptime(value, '%Y-%m-%dT%H:%M')
-            except ValueError:
-                raise PyODataModelError(f'Cannot decode datetime from value {value}.')
+        raise PyODataModelError(f'Cannot decode datetime from value {value}.')
 
 
 class EdmDateTimeTypTraits(EdmPrefixedTypTraits):
@@ -487,7 +486,6 @@ def from_literal(self, value):
 
         value = super(EdmDateTimeTypTraits, self).from_literal(value)
 
-        # Note: parse_datetime_literal raises a PyODataModelError exception on invalid formats
         return parse_datetime_literal(value).replace(tzinfo=datetime.timezone.utc)
 
 
@@ -557,21 +555,13 @@ def from_literal(self, value):
         value = super(EdmDateTimeOffsetTypTraits, self).from_literal(value)
 
         try:
-            # Note: parse_datetime_literal raises a PyODataModelError exception on invalid formats
-            if re.match(r'\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}Z', value, flags=re.ASCII | re.IGNORECASE):
-                datetime_part = value[:-1]
-                tz_info = datetime.timezone.utc
-            else:
-                match = re.match(r'(?P<datetime>.+)(?P<sign>[\\+-])(?P<hours>\d{2}):(?P<minutes>\d{2})',
-                                 value,
-                                 flags=re.ASCII)
-                datetime_part = match.group('datetime')
-                tz_offset = datetime.timedelta(hours=int(match.group('hours')),
-                                               minutes=int(match.group('minutes')))
-                tz_sign = -1 if match.group('sign') == '-' else 1
-                tz_info = datetime.timezone(tz_sign * tz_offset)
-            return parse_datetime_literal(datetime_part).replace(tzinfo=tz_info)
-        except (ValueError, AttributeError):
+            normalized = value.upper().replace('Z', '+00:00')
+            if '.' in normalized:
+                dt_part, rest = normalized.split('.', 1)
+                frac, tz_suffix = rest[:6].ljust(6, '0'), rest[6:]
+                normalized = f'{dt_part}.{frac}{tz_suffix}'
+            return datetime.datetime.fromisoformat(normalized)
+        except ValueError:
             raise PyODataModelError(f'Cannot decode datetimeoffset from value {value}.')
 
 

tests/test_model_v2.py

@@ -491,6 +491,7 @@ def test_traits():
     ('2001-02-03T04:05:06.000007', datetime(2001, 2, 3, 4, 5, 6, microsecond=7)),
     ('2001-02-03T04:05:06', datetime(2001, 2, 3, 4, 5, 6, 0)),
     ('2001-02-03T04:05', datetime(2001, 2, 3, 4, 5, 0, 0)),
+    ('2001-02-03T04:05:06.1', datetime(2001, 2, 3, 4, 5, 6, microsecond=100000)),
 ])
 def test_parse_datetime_literal(datetime_literal, expected):
     assert parse_datetime_literal(datetime_literal) == expected
@@ -561,6 +562,15 @@ def test_traits_datetime(type_date_time):
     assert testdate.microsecond == 0
     assert testdate.tzinfo == timezone.utc
 
+    # parsing sub-6-digit fractional seconds
+    testdate = type_date_time.traits.from_literal("datetime'1976-11-23T03:33:06.1'")
+    assert testdate.second == 6
+    assert testdate.microsecond == 100000
+    assert testdate.tzinfo == timezone.utc
+
+    # None returns None
+    assert type_date_time.traits.from_literal(None) is None
+
     # parsing invalid value
     with pytest.raises(PyODataModelError) as e_info:
         type_date_time.traits.from_literal('xyz')
@@ -732,6 +742,8 @@ def test_traits_datetimeoffset_to_json(type_date_time_offset, python_datetime, e
     ("datetimeoffset'1976-11-23t03:33:06+12:00'", datetime(1976, 11, 23, 3, 33, 6, tzinfo=timezone(timedelta(hours=12))), 'On dateline'),
     ("datetimeoffset'1976-11-23t03:33:06-12:00'", datetime(1976, 11, 23, 3, 33, 6, tzinfo=timezone(timedelta(hours=-12))), 'Minimum offset'),
     ("datetimeoffset'1976-11-23t03:33:06+14:00'", datetime(1976, 11, 23, 3, 33, 6, tzinfo=timezone(timedelta(hours=14))), 'Maximum offset'),
+    ("datetimeoffset'1976-11-23T03:33+01:00'", datetime(1976, 11, 23, 3, 33, 0, tzinfo=timezone(timedelta(hours=1))), 'No seconds with offset'),
+    ("datetimeoffset'1976-11-23T03:33:06.654321Z'", datetime(1976, 11, 23, 3, 33, 6, microsecond=654321, tzinfo=timezone.utc), 'Microseconds with Z'),
 ])
 def test_traits_datetimeoffset_from_literal(type_date_time_offset, literal, expected, comment):
     """Test Edm.DateTimeOffset trait: literal -> Python"""
@@ -749,6 +761,10 @@ def test_traits_datetimeoffset_from_invalid_literal(type_date_time_offset):
     assert str(e_info.value).startswith('Cannot decode datetimeoffset from value xyz')
 
 
+def test_traits_datetimeoffset_from_literal_none(type_date_time_offset):
+    assert type_date_time_offset.traits.from_literal(None) is None
+
+
 def test_traits_datetimeoffset_from_json(type_date_time_offset):
     """Test Edm.DateTimeOffset trait: OData -> Python"""
 

tests/test_model_v2_EdmStructTypeSerializer.py

@@ -12,7 +12,7 @@ def complex_type_property_declarations():
         'TestString': (Types.parse_type_name('Edm.String'), "'FooBar'", "'FooBar'", 'FooBar'),
         'TestBoolean': (Types.parse_type_name('Edm.Boolean'), False, 'false', False),
         'TestInt64': (Types.parse_type_name('Edm.Int64'), '123L', '123L', 123),
-        'TestDateTime': (Types.parse_type_name('Edm.DateTime'), "/Date(2147483647000)/", "datetime'2038-01-19T3:14:7'",
+        'TestDateTime': (Types.parse_type_name('Edm.DateTime'), "/Date(2147483647000)/", "datetime'2038-01-19T03:14:07'",
                          datetime.datetime(2038, 1, 19, hour=3, minute=14, second=7, tzinfo=datetime.timezone.utc))
     }