Issue #82 - Done more work making errors more type-specific to aid fu…

…ture APIs providing tailored error messages to users.

csvqb/csvqb/models/cube/__init__.py

@@ -2,3 +2,4 @@
 from .cube import Cube
 from .catalog import CatalogMetadataBase
 from .csvqb import *
+from .validationerrors import *

csvqb/csvqb/models/cube/csvqb/__init__.py

@@ -2,5 +2,6 @@
 from .components import *
 from .catalog import CatalogMetadata
 from ..cube import Cube
+from .validationerrors import *
 
 QbCube = Cube[CatalogMetadata]

csvqb/csvqb/models/cube/csvqb/validationerrors.py

@@ -1,32 +1,23 @@
-from abc import ABC
-from dataclasses import dataclass, field
-from typing import Type, Optional, Union
+from dataclasses import dataclass
+from typing import Optional, Type, Union
 
-from csvqb.models.cube import (
-    QbDataStructureDefinition,
+from csvqb.models.cube.csvqb.components import (
     QbObservationValue,
     QbMultiUnits,
+    QbDataStructureDefinition,
 )
-from csvqb.models.validationerror import ValidationError
-
+from csvqb.models.validationerror import SpecificValidationError
 
 ComponentTypeDescription = Union[str, Type[QbDataStructureDefinition]]
 
 
-def _get_component_type_description(t: ComponentTypeDescription) -> str:
+def get_description_for_component(t: ComponentTypeDescription) -> str:
     if isinstance(t, str):
         return t
 
     return t.__name__
 
 
-@dataclass
-class SpecificValidationError(ValidationError, ABC):
-    """Abstract base class to represent ValidationErrors which are more specific and so can be interpreted by code."""
-
-    message: str = field(init=False)
-
-
 @dataclass
 class OutputUriTemplateMissingError(SpecificValidationError):
     """
@@ -39,13 +30,13 @@ class OutputUriTemplateMissingError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"'{self.csv_column_name}' - an {_get_component_type_description(self.component_type)} must have an "
+            f"'{self.csv_column_name}' - a {get_description_for_component(self.component_type)} must have an "
             + "output_uri_template defined."
         )
 
 
 @dataclass
-class MaximumNumberOfComponentsError(SpecificValidationError):
+class MaxNumComponentsExceededError(SpecificValidationError):
     """
     Represents an error where the user can define a maximum number of components of a given type, but has defined
     too many.
@@ -58,15 +49,15 @@ class MaximumNumberOfComponentsError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"Found {self.actual_number} of {_get_component_type_description(self.component_type)}s. "
-            + f"Expected maximum {self.maximum_number}."
+            f"Found {self.actual_number} of {get_description_for_component(self.component_type)}s. "
+            + f"Expected a maximum of {self.maximum_number}."
         )
         if self.additional_explanation is not None:
             self.message += " " + self.additional_explanation
 
 
 @dataclass
-class MinimumNumberOfComponentsError(SpecificValidationError):
+class MinNumComponentsNotSatisfiedError(SpecificValidationError):
     """
     Represents an error where the user must define a minimum number of components of a given type, but has not done so.
     """
@@ -78,7 +69,7 @@ class MinimumNumberOfComponentsError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"At least {self.minimum_number} {_get_component_type_description(self.component_type)}s must be defined."
+            f"At least {self.minimum_number} {get_description_for_component(self.component_type)}s must be defined."
             + f" Found {self.actual_number}."
         )
         if self.additional_explanation is not None:
@@ -98,8 +89,8 @@ class WrongNumberComponentsError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"Found {self.actual_number} {_get_component_type_description(self.component_type)}s."
-            + f" Expected {self.expected_number}."
+            f"Found {self.actual_number} {get_description_for_component(self.component_type)}s."
+            + f" Expected exactly {self.expected_number}."
         )
         if self.additional_explanation is not None:
             self.message += " " + self.additional_explanation
@@ -117,8 +108,8 @@ class NeitherDefinedError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"Found neither {_get_component_type_description(self.component_one)} "
-            + f"nor {_get_component_type_description(self.component_two)} defined. "
+            f"Found neither {get_description_for_component(self.component_one)} "
+            + f"nor {get_description_for_component(self.component_two)} defined. "
             + "One of these must be defined."
         )
 
@@ -146,8 +137,8 @@ class IncompatibleComponentsError(SpecificValidationError):
 
     def __post_init__(self):
         self.message = (
-            f"Both {_get_component_type_description(self.component_one)} "
-            + f"and {_get_component_type_description(self.component_two)} have been defined. "
+            f"Both {get_description_for_component(self.component_one)} "
+            + f"and {get_description_for_component(self.component_two)} have been defined. "
             + f"These components cannot be used together."
         )
 

csvqb/csvqb/models/cube/cube.py

@@ -7,9 +7,13 @@
 import pandas as pd
 
 from csvqb.models.validationerror import ValidationError
+from .validationerrors import (
+    DuplicateColumnTitleError,
+    ColumnNotFoundInDataError,
+    MissingColumnDefinitionError,
+)
 from .columns import CsvColumn
-from csvqb.models.cube.catalog import CatalogMetadataBase
-from csvqb.inputs import pandas_input_to_columnar
+from .catalog import CatalogMetadataBase
 from ..pydanticmodel import PydanticModel
 
 TMetadata = TypeVar("TMetadata", bound=CatalogMetadataBase, covariant=True)
@@ -35,9 +39,7 @@ def _validate_columns(self) -> List[ValidationError]:
         existing_col_titles: Set[str] = set()
         for col in self.columns:
             if col.csv_column_title in existing_col_titles:
-                errors.append(
-                    ValidationError(f"Duplicate column title '{col.csv_column_title}'")
-                )
+                errors.append(DuplicateColumnTitleError(col.csv_column_title))
             else:
                 existing_col_titles.add(col.csv_column_title)
 
@@ -46,22 +48,14 @@ def _validate_columns(self) -> List[ValidationError]:
                 if col.csv_column_title in self.data.columns:
                     maybe_column_data = self.data[col.csv_column_title]
                 else:
-                    errors.append(
-                        ValidationError(
-                            f"Column '{col.csv_column_title}' not found in data provided."
-                        )
-                    )
+                    errors.append(ColumnNotFoundInDataError(col.csv_column_title))
 
             errors += col.validate_data(maybe_column_data)
 
         if self.data is not None:
             defined_column_titles = [c.csv_column_title for c in self.columns]
             for column in list(self.data.columns):
                 if column not in defined_column_titles:
-                    errors.append(
-                        ValidationError(
-                            f"Column '{column}' does not have a mapping defined."
-                        )
-                    )
+                    errors.append(MissingColumnDefinitionError(column))
 
         return errors

csvqb/csvqb/models/cube/validationerrors.py

@@ -0,0 +1,41 @@
+from dataclasses import dataclass
+
+from csvqb.models.validationerror import SpecificValidationError
+
+
+@dataclass
+class DuplicateColumnTitleError(SpecificValidationError):
+    """
+    An error to inform the user that they have defined two instances of the same column.
+    """
+
+    csv_column_title: str
+
+    def __post_init__(self):
+        self.message = f"Duplicate column title '{self.csv_column_title}'"
+
+
+@dataclass
+class ColumnNotFoundInDataError(SpecificValidationError):
+    """
+    An error to inform the user that they have defined a column which cannot be found in the provided data.
+    """
+
+    csv_column_title: str
+
+    def __post_init__(self):
+        self.message = f"Column '{self.csv_column_title}' not found in data provided."
+
+
+@dataclass
+class MissingColumnDefinitionError(SpecificValidationError):
+    """
+    An error to inform the user that there is a column in their data that does not have a mapping specified.
+    """
+
+    csv_column_title: str
+
+    def __post_init__(self):
+        self.message = (
+            f"Column '{self.csv_column_title}' does not have a mapping defined."
+        )

csvqb/csvqb/models/validationerror.py

@@ -2,9 +2,19 @@
 ValidationError
 ---------------
 """
-from dataclasses import dataclass
+from dataclasses import dataclass, field
+from abc import ABC
 
 
 @dataclass
 class ValidationError:
+    """Class representing an error validating a model."""
+
     message: str
+
+
+@dataclass
+class SpecificValidationError(ValidationError, ABC):
+    """Abstract base class to represent ValidationErrors which are more specific and so can be interpreted by code."""
+
+    message: str = field(init=False)

csvqb/csvqb/tests/unit/cube/qb/test_cubeqb_errorvalidation.py

@@ -7,10 +7,10 @@
 from csvqb.models.cube import *
 from csvqb.models.cube.csvqb.validationerrors import (
     OutputUriTemplateMissingError,
-    MinimumNumberOfComponentsError,
+    MinNumComponentsNotSatisfiedError,
     UnitsNotDefinedError,
     BothUnitTypesDefinedError,
-    MaximumNumberOfComponentsError,
+    MaxNumComponentsExceededError,
     WrongNumberComponentsError,
     IncompatibleComponentsError,
 )
@@ -148,7 +148,7 @@ def test_no_dimensions_validation_error():
 
     assert_num_validation_errors(errors, 1)
     error = errors[0]
-    assert isinstance(error, MinimumNumberOfComponentsError)
+    assert isinstance(error, MinNumComponentsNotSatisfiedError)
     assert error.component_type == QbDimension
     assert error.minimum_number == 1
     assert error.actual_number == 0
@@ -258,7 +258,7 @@ def test_multiple_units_columns():
 
     assert_num_validation_errors(errors, 1)
     error = errors[0]
-    assert isinstance(error, MaximumNumberOfComponentsError)
+    assert isinstance(error, MaxNumComponentsExceededError)
     assert error.component_type == QbMultiUnits
     assert error.maximum_number == 1
     assert error.actual_number == 2
@@ -393,7 +393,7 @@ def test_multi_measure_obs_val_with_multiple_measure_dimensions():
 
     assert_num_validation_errors(errors, 1)
     error = errors[0]
-    assert isinstance(error, MaximumNumberOfComponentsError)
+    assert isinstance(error, MaxNumComponentsExceededError)
     assert error.component_type == QbMultiMeasureDimension
     assert error.maximum_number == 1
     assert error.actual_number == 2
@@ -442,7 +442,7 @@ def test_measure_dimension_with_single_measure_obs_val():
     assert error.component_two == QbMultiMeasureDimension
     assert (
         error.additional_explanation
-        == "A single measure cube cannot have a measure dimension."
+        == "A single-measure cube cannot have a measure dimension."
     )
 
 

csvqb/csvqb/tests/unit/cube/test_cube_errorvalidation.py

@@ -20,6 +20,8 @@ def test_column_not_configured_error():
 
     assert len(validation_errors) == 1
     error = validation_errors[0]
+    assert isinstance(error, MissingColumnDefinitionError)
+    assert error.csv_column_title == "Some Dimension"
     assert "Column 'Some Dimension'" in error.message
 
 
@@ -37,6 +39,8 @@ def test_column_title_wrong_error():
 
     assert len(validation_errors) == 1
     error = validation_errors[0]
+    assert isinstance(error, ColumnNotFoundInDataError)
+    assert error.csv_column_title == "Some Column Title"
     assert "Column 'Some Column Title'" in error.message
 
 
@@ -59,6 +63,8 @@ def test_two_column_same_title():
 
     assert len(validation_errors) == 1
     error = validation_errors[0]
+    assert isinstance(error, DuplicateColumnTitleError)
+    assert error.csv_column_title == "Some Dimension"
     assert "Duplicate column title 'Some Dimension'" == error.message
 
 

csvqb/csvqb/utils/qb/cube.py

@@ -6,8 +6,8 @@
 
 from csvqb.models.cube.csvqb.validationerrors import (
     OutputUriTemplateMissingError,
-    MinimumNumberOfComponentsError,
-    MaximumNumberOfComponentsError,
+    MinNumComponentsNotSatisfiedError,
+    MaxNumComponentsExceededError,
     WrongNumberComponentsError,
     UnitsNotDefinedError,
     BothUnitTypesDefinedError,
@@ -74,7 +74,7 @@ def _validate_dimensions(cube: Cube) -> List[ValidationError]:
                 )
 
     if len(dimension_columns) == 0:
-        errors.append(MinimumNumberOfComponentsError(QbDimension, 1, 0))
+        errors.append(MinNumComponentsNotSatisfiedError(QbDimension, 1, 0))
     return errors
 
 
@@ -104,7 +104,7 @@ def _validate_observation_value_constraints(cube: Cube) -> List[ValidationError]
 
     if len(multi_units_columns) > 1:
         errors.append(
-            MaximumNumberOfComponentsError(QbMultiUnits, 1, len(multi_units_columns))
+            MaxNumComponentsExceededError(QbMultiUnits, 1, len(multi_units_columns))
         )
 
     if len(observed_value_columns) != 1:
@@ -149,7 +149,7 @@ def _validate_multi_measure_cube(
         )
     elif len(multi_measure_columns) > 1:
         errors.append(
-            MaximumNumberOfComponentsError(
+            MaxNumComponentsExceededError(
                 QbMultiMeasureDimension, 1, len(multi_measure_columns)
             )
         )
@@ -168,7 +168,7 @@ def _validate_single_measure_cube(
             IncompatibleComponentsError(
                 QbSingleMeasureObservationValue,
                 QbMultiMeasureDimension,
-                additional_explanation="A single measure cube cannot have a measure dimension.",
+                additional_explanation="A single-measure cube cannot have a measure dimension.",
             )
         )