Update logic to update a Report object (scrapd#216)

Defines the logic to update a `Report` with the attributes from another
one.

Author: rgreinho

scrapd/core/apd.py

@@ -1,6 +1,5 @@
 """Define the module containing the function used to scrap data from the APD website."""
 import asyncio
-import datetime
 from pathlib import Path
 import re
 from urllib.parse import urljoin
@@ -136,7 +135,7 @@ def parse_page(page, url, dump=False):
     :return: a dictionary representing a fatality.
     :rtype: dict
     """
-    report = model.Report(case='19-123456', date=datetime.datetime.now().date())
+    report = model.Report(case='19-123456')
 
     # Parse the twitter fields.
     twitter_report, twitter_err = twitter.parse(page)

scrapd/core/model.py

@@ -68,7 +68,7 @@ class Report(BaseModel):
 
     case: str
     crash: int = 0
-    date: datetime.date
+    date: datetime.date = None
     fatalities: List[Fatality] = []
     link: str = ''
     latitude: float = 0.0
@@ -90,18 +90,41 @@ def compute_fatalities_age(self):
             # Compute the age.
             f.age = date_utils.compute_age(self.date, f.dob)
 
-    def update(self, other):
-        """Update a model with values from another one."""
+    def update(self, other, strict=False):
+        """
+        Update a model in place with values from another one.
+
+        Updates only the empty values of the `self` instance with the non-empty values of the `other` instance.
+
+        :param Report other: report to update with
+        :param bool strict: strict mode
+        """
+        # Do nothing if there is no other instance to update from.
         if not other:
             return
 
-        attrs = ['case', 'crash', 'date', 'fatalities', 'link', 'latitude', 'location', 'longitude', 'notes', 'time']
-        for attr in attrs:
-            # Only the required values can be overridden.
-            if attr in ('case', 'date'):
+        # Ensure other instance has the right type.
+        if not isinstance(other, Report):
+            raise TypeError(f'other instance is not of type "Report": {type(other)}')
+
+        # Define the list of attrs.
+        required_attrs = ['case']
+        attrs = ['crash', 'date', 'fatalities', 'link', 'latitude', 'location', 'longitude', 'notes', 'time']
+
+        # When strict...
+        if strict:
+            # The case number and date must be identical.
+            if not all([getattr(self, attr) == getattr(other, attr) for attr in required_attrs]):
+                raise ValueError(
+                    f'in strict mode the required attributes "({", ".join(required_attrs)})" must be identical')
+        else:
+            # Otherwise they are overridden.
+            for attr in required_attrs:
                 setattr(self, attr, getattr(other, attr))
-                continue
-            if not getattr(self, attr):
+
+        # Set the non-empty attributes of `other` into the empty attributes of the current instance.
+        for attr in attrs:
+            if not getattr(self, attr) and getattr(other, attr):
                 setattr(self, attr, getattr(other, attr))
 
     @validator('case')

tests/core/test_model.py

@@ -47,23 +47,34 @@
     },
 ]
 
+now = datetime.datetime.now()
 update_model_scenarios = [
     {
-        'input_': model.Report(case='19-123456', date=datetime.datetime.now().date()),
-        'other': model.Report(case='19-123456', date=datetime.datetime.now().date()),
-        'expected': model.Report(case='19-123456', date=datetime.datetime.now().date()),
-        'id': 'simple',
+        'input_': model.Report(case='19-123456', date=now.date()),
+        'other': model.Report(case='19-123456', date=now.date()),
+        'expected': model.Report(case='19-123456', date=now.date()),
+        'strict': False,
+        'id': 'identical-nonstrict',
     },
     {
-        'input_': model.Report(case='19-123456', date=datetime.datetime.now().date(), link='link'),
+        'input_': model.Report(case='19-123456', date=now.date()),
+        'other': model.Report(case='19-123456', date=now.date()),
+        'expected': model.Report(case='19-123456', date=now.date()),
+        'strict': True,
+        'id': 'identical-strict',
+    },
+    {
+        'input_': model.Report(case='19-123456', date=now.date(), link='link'),
         'other': None,
-        'expected': model.Report(case='19-123456', date=datetime.datetime.now().date(), link='link'),
+        'expected': model.Report(case='19-123456', date=now.date(), link='link'),
+        'strict': False,
         'id': 'None',
     },
     {
-        'input_': model.Report(case='19-123456', date=datetime.datetime.now().date(), link='link'),
-        'other': model.Report(case='19-123456', date=datetime.datetime.now().date(), link='other link', crash=1),
-        'expected': model.Report(case='19-123456', date=datetime.datetime.now().date(), link='link', crash=1),
+        'input_': model.Report(case='19-123456', date=now.date(), link='link'),
+        'other': model.Report(case='19-123456', date=now.date(), link='other link', crash=1),
+        'expected': model.Report(case='19-123456', date=now.date(), link='link', crash=1),
+        'strict': False,
         'id': 'complex',
     },
 ]
@@ -145,21 +156,28 @@ def test_compute_fatalities_age(self, input_, expected):
         assert [f.age for f in input_.fatalities] == expected
 
     @pytest.mark.parametrize(
-        'input_,other,expected',
-        [pytest.param(s['input_'], s['other'], s['expected'], id=s['id']) for s in update_model_scenarios],
+        'input_,other,strict,expected',
+        [pytest.param(s['input_'], s['other'], s['strict'], s['expected'], id=s['id']) for s in update_model_scenarios],
     )
-    def test_update_00(self, input_, other, expected):
+    def test_update_00(self, input_, other, strict, expected):
         """Ensure models can be updated."""
         actual = input_.copy(deep=True)
-        actual.update(other)
+        actual.update(other, strict)
         assert actual == expected
 
     def test_update_01(self):
-        """Ensure the case field gets updated."""
-        actual = model.Report(case='19-123456', date=datetime.datetime.now().date())
-        other = model.Report(case='19-123456', date=datetime.datetime.now().date())
-        actual.update(other)
-        assert actual == other
+        """Ensure required fields are identical in strict mode."""
+        actual = model.Report(case='19-123456', date=now.date())
+        other = model.Report(case='19-654321', date=now.date())
+        with pytest.raises(ValueError):
+            actual.update(other, True)
+
+    def test_update_02(self):
+        """Ensure both instances are of type Report."""
+        actual = model.Report(case='19-123456', date=now.date())
+        other = dict(case='19-654321', date=now.date())
+        with pytest.raises(TypeError):
+            actual.update(other)
 
     def test_invalid_case_number(self):
         """Ensure the case number has a valid format."""

tests/core/test_twitter.py

@@ -32,8 +32,8 @@
         'id': 'case-number-only',
         'page': 'traffic-fatality-2-3',
         'description': "Case:           19-0161105",
-        'expected': None,
-        'errors': 1,
+        'expected': model.Report(case="19-0161105", crash=2),
+        'errors': 0,
     },
     {
         'id': 'regular',
@@ -100,7 +100,7 @@
         See video of suspect vehicle here --> https://youtu.be/ezxaRW79PnI
         """,
         'expected': None,
-        'errors': 2,
+        'errors': 1,
     },
     {
         'id': 'multiple-fatalities',
@@ -142,7 +142,7 @@
         'page': None,
         'description': '',
         'expected': model.Report(case='19-123456', date=datetime.datetime.now().date()),
-        'errors': 2,
+        'errors': 1,
     },
     {
         'id': 'with-arrested',