Develop v0.3.0

CHANGELOG.md

@@ -1,3 +1,13 @@
+## v0.3.0 (2025-11-19)
+
+### Feat
+
+- new domain type formattedtime for time only data
+
+### Refactor
+
+- small tweak to allow use of dynamic fields in select rules
+
 ## v0.2.0 (2025-11-12)
 
 ### Refactor

docs/detailed_guidance/domain_types.md

@@ -4,24 +4,24 @@ Domain types are custom defined pydantic types that solve common problems with u
 This might include Postcodes, NHS Numbers, dates with specific formats etc.
 
 Below is a list of defined types, their output type and any contraints. Nested beneath them are any constraints that area allowed and their default values if there are any.
-| Defined Type | Output Type | Contraints & Defaults |
-| ------------ | ----------- | --------------------- |
-| NHSNumber | str |
-| permissive_nhs_number | str | <li> warn_on_test_numbers = False </li> |
-| Postcode | str |
-| OrgId | str |
-| conformatteddate | date | <li>date_format: str</li><li>ge: date</li><li>le: date</li><li>gt: date</li><li>lt: date</li> |
-| formatteddatetime | datetime | <li>date_format: str </li><li>timezone_treatment: one_of ["forbid", "permit", "require] = "permit"</li> |
-| reportingperiod | date | <li>reporting_period_type: one_of ["start", "end"]</li><li>date_format: str = "%Y-%m-%d"</li> |
-| alphanumeric | str | <li>min_digits : NonNegativeInt = 1</li><li>max_digits: PositiveInt = 1</li> |
-| identifier | str | <li>min_digits : NonNegativeInt = 1</li><li>max_digits: PositiveInt = 1</li>
+| Defined Type | Output Type | Contraints & Defaults | Supported Implementations |
+| ------------ | ----------- | --------------------- | ------------------------- |
+| NHSNumber | str | | Spark, DuckDB |
+| permissive_nhs_number | str | <li> warn_on_test_numbers = False </li> | Spark, DuckDB |
+| Postcode | str | | Spark, DuckDB |
+| OrgId | str | | Spark, DuckDB |
+| conformatteddate | date | <li>date_format: str</li><li>ge: date</li><li>le: date</li><li>gt: date</li><li>lt: date</li> | Spark, DuckDB |
+| formatteddatetime | datetime | <li>date_format: str </li><li>timezone_treatment: one_of ["forbid", "permit", "require] = "permit"</li> | Spark, DuckDB |
+| formattedtime | time | <li>time_format: str </li><li>timezone_treatment: one_of ["forbid", "permit", "require"] = "permit" | DuckDB |
+| reportingperiod | date | <li>reporting_period_type: one_of ["start", "end"]</li><li>date_format: str = "%Y-%m-%d"</li> | Spark, DuckDB |
+| alphanumeric | str | <li>min_digits : NonNegativeInt = 1</li><li>max_digits: PositiveInt = 1</li> | Spark, DuckDB |
+| identifier | str | <li>min_digits : NonNegativeInt = 1</li><li>max_digits: PositiveInt = 1</li> | Spark, DuckDB |
 
-Other types that are allowed include:
+**Other types that are allowed include:**
 - str
 - int
 - date
 - datetime
 - Decimal
 - float
-
-And any types that are included in [pydantic version 1.10](https://docs.pydantic.dev/1.10/usage/types/#pydantic-types)
+- Any types that are included in [pydantic version 1.10](https://docs.pydantic.dev/1.10/usage/types/#pydantic-types)

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "nhs_dve"
-version = "0.2.0"
+version = "0.3.0"
 description = "`nhs data validation engine` is a framework used to validate data"
 authors = ["NHS England <england.contactus@nhs.net>"]
 readme = "README.md"

src/dve/core_engine/backends/implementations/duckdb/__init__.py

@@ -1,4 +1,5 @@
 """Implementation of duckdb backend"""
+
 from dve.core_engine.backends.implementations.duckdb.readers.json import DuckDBJSONReader
 from dve.core_engine.backends.readers import register_reader
 

src/dve/core_engine/backends/implementations/duckdb/duckdb_helpers.py

@@ -4,7 +4,7 @@
 """Helper objects for duckdb data contract implementation"""
 from collections.abc import Generator, Iterator
 from dataclasses import is_dataclass
-from datetime import date, datetime
+from datetime import date, datetime, time
 from decimal import Decimal
 from pathlib import Path
 from typing import Any, ClassVar, Union
@@ -87,6 +87,7 @@ def __call__(self):
     date: ddbtyp.DATE,
     datetime: ddbtyp.TIMESTAMP,
     Decimal: DDBDecimal()(),
+    time: ddbtyp.TIME,
 }
 """A mapping of Python types to the equivalent DuckDB types."""
 

src/dve/core_engine/backends/implementations/duckdb/utilities.py

@@ -1,5 +1,7 @@
 """Utility objects for use with duckdb backend"""
 
+import itertools
+
 from dve.core_engine.backends.base.utilities import _split_multiexpr_string
 
 
@@ -24,7 +26,11 @@ def expr_mapping_to_columns(expressions: dict) -> list[str]:
 
 def expr_array_to_columns(expressions: list[str]) -> list[str]:
     """Create list of duckdb expressions from list of expressions"""
-    return [f"{expression}" for expression in expressions]
+    return list(
+        itertools.chain.from_iterable(
+            _split_multiexpr_string(expression) for expression in expressions
+        )
+    )
 
 
 def multiexpr_string_to_columns(expressions: str) -> list[str]:

src/dve/core_engine/backends/implementations/spark/spark_helpers.py

@@ -12,14 +12,7 @@
 from dataclasses import dataclass, is_dataclass
 from decimal import Decimal
 from functools import wraps
-from typing import (
-    Any,
-    ClassVar,
-    Optional,
-    TypeVar,
-    Union,
-    overload,
-)
+from typing import Any, ClassVar, Optional, TypeVar, Union, overload
 
 from delta.exceptions import ConcurrentAppendException, DeltaConcurrentModificationException
 from pydantic import BaseModel

src/dve/core_engine/backends/implementations/spark/utilities.py

@@ -1,6 +1,7 @@
 """Some utilities which are useful for implementing Spark transformations."""
 
 import datetime as dt
+import itertools
 from collections.abc import Callable
 from json import JSONEncoder
 from operator import and_, or_
@@ -70,7 +71,13 @@ def expr_mapping_to_columns(expressions: ExpressionMapping) -> list[Column]:
 
 def expr_array_to_columns(expressions: ExpressionArray) -> list[Column]:
     """Convert an array of expressions to a list of columns."""
-    return list(map(sf.expr, expressions))
+
+    _expr_list = list(
+        itertools.chain.from_iterable(
+            _split_multiexpr_string(expression) for expression in expressions
+        )
+    )
+    return list(map(sf.expr, _expr_list))
 
 
 def multiexpr_string_to_columns(expressions: MultiExpression) -> list[Column]:

src/dve/core_engine/backends/readers/xml.py

@@ -3,14 +3,7 @@
 
 import re
 from collections.abc import Collection, Iterator
-from typing import (
-    IO,
-    Any,
-    GenericAlias,  # type: ignore
-    Optional,
-    Union,
-    overload
-)
+from typing import IO, Any, GenericAlias, Optional, Union, overload  # type: ignore
 
 import polars as pl
 from lxml import etree  # type: ignore

src/dve/core_engine/backends/utilities.py

@@ -2,10 +2,10 @@
 
 import sys
 from dataclasses import is_dataclass
-from datetime import date, datetime
+from datetime import date, datetime, time
 from decimal import Decimal
-from typing import Any, ClassVar, Union
 from typing import GenericAlias  # type: ignore
+from typing import Any, ClassVar, Union
 
 import polars as pl  # type: ignore
 from polars.datatypes.classes import DataTypeClass as PolarsType
@@ -33,13 +33,16 @@
     date: pl.Date,  # type: ignore
     datetime: pl.Datetime,  # type: ignore
     Decimal: pl.Utf8,  # type: ignore
+    time: pl.Time,  # type: ignore
 }
 """A mapping of Python types to the equivalent Polars types."""
 
 
 def stringify_type(type_: Union[type, GenericAlias]) -> type:
     """Stringify an individual type."""
-    if isinstance(type_, type) and not isinstance(type_, GenericAlias):  # A model, return the contents.  # pylint: disable=C0301
+    if isinstance(type_, type) and not isinstance(
+        type_, GenericAlias
+    ):  # A model, return the contents.  # pylint: disable=C0301
         if issubclass(type_, BaseModel):
             return stringify_model(type_)
 

src/dve/core_engine/message.py

@@ -2,8 +2,8 @@
 
 import copy
 import datetime as dt
-import operator
 import json
+import operator
 from collections.abc import Callable
 from decimal import Decimal
 from functools import reduce

src/dve/core_engine/type_hints.py

@@ -6,12 +6,12 @@
 from pathlib import Path
 from queue import Queue as ThreadQueue
 from typing import TYPE_CHECKING, Any, List, Optional, TypeVar, Union  # pylint: disable=W1901
-# TODO - cannot remove List from Typing. See L60 for details.
 
 from pyspark.sql import DataFrame
 from pyspark.sql.types import StructType
 from typing_extensions import Literal, ParamSpec, get_args
 
+# TODO - cannot remove List from Typing. See L60 for details.
 
 if TYPE_CHECKING:  # pragma: no cover
     from dve.core_engine.message import FeedbackMessage

src/dve/metadata_parser/domain_types.py

@@ -392,6 +392,97 @@ def __get_validators__(cls) -> Iterator[classmethod]:
         yield cls.validate  # type: ignore
 
 
+class FormattedTime(dt.time):
+    """A time, provided as a datetime or a string in a specific format."""
+
+    TIME_FORMAT: ClassVar[Optional[str]] = None
+    """The specific format of the time."""
+    TIMEZONE_TREATMENT: ClassVar[Literal["forbid", "permit", "require"]] = "permit"
+    """How to treat the presence of timezone-related information."""
+    DEFAULT_PATTERNS: Sequence[str] = list(
+        # 24 hour time pattern combinations
+        map(
+            "".join,
+            itertools.product(
+                ("%H:%M:%S", "%H%M%S"),
+                ("", ".%f"),
+                ("%p", "%P", ""),
+                ("%z", ""),
+            ),
+        )
+    ) + list(
+        # 12 hour time pattern combinations
+        map(
+            "".join,
+            itertools.product(
+                ("%I:%M:%S", "%I%M%S"),
+                ("", ".%f"),
+                ("%z", ""),
+                (" %p", "%p", "%P", " %P", ""),
+            ),
+        )
+    )
+    """A sequence of time format patterns to try if `TIME_FORMAT` is unset."""
+
+    @classmethod
+    def convert_to_time(cls, value: dt.datetime) -> dt.time:
+        """
+        Convert `datetime.datetime` to `datetime.time`. If datetime contains timezone info, that
+        will be retained.
+        """
+        if value.tzinfo:
+            return value.timetz()
+
+        return value.time()
+
+    @classmethod
+    def parse_time(cls, string: str) -> dt.time:
+        """Attempt to parse a datetime using various formats in sequence."""
+        string = string.strip()
+        if string.endswith("Z"):  # Convert 'zulu' time to UTC.
+            string = string[:-1] + "+00:00"
+
+        for pattern in cls.DEFAULT_PATTERNS:
+            try:
+                datetime = dt.datetime.strptime(string, pattern)
+            except ValueError:
+                continue
+
+            time = cls.convert_to_time(datetime)
+
+            return time  # pragma: no cover
+        raise ValueError("Unable to parse provided time")
+
+    @classmethod
+    def validate(cls, value: Union[dt.time, dt.datetime, str]) -> dt.time | None:
+        """Validate a passed time, datetime or string."""
+        if value is None:
+            return value
+
+        if isinstance(value, dt.time):
+            new_time = value
+        elif isinstance(value, dt.datetime):
+            new_time = cls.convert_to_time(value)
+        else:
+            if cls.TIME_FORMAT is not None:
+                try:
+                    new_time = dt.datetime.strptime(value, cls.TIME_FORMAT)  # type: ignore
+                    new_time = cls.convert_to_time(new_time)  # type: ignore
+                except ValueError as err:
+                    raise ValueError(
+                        f"Unable to parse provided time in format {cls.TIME_FORMAT}"
+                    ) from err
+            else:
+                new_time = cls.parse_time(value)
+
+        if cls.TIMEZONE_TREATMENT == "forbid" and new_time.tzinfo:
+            raise ValueError("Provided time has timezone, but this is forbidden for this field")
+        if cls.TIMEZONE_TREATMENT == "require" and not new_time.tzinfo:
+            raise ValueError("Provided time missing timezone, but this is required for this field")
+
+        return new_time
+
+
 @lru_cache()
 @validate_arguments
 def formatteddatetime(
@@ -412,6 +503,23 @@ def formatteddatetime(
     return type("FormattedDatetime", (FormattedDatetime, *FormattedDatetime.__bases__), dict_)
 
 
+@lru_cache()
+@validate_arguments
+def formattedtime(
+    time_format: Optional[str] = None,
+    timezone_treatment: Literal["forbid", "permit", "require"] = "permit",
+) -> type[FormattedTime]:
+    """Return a formatted time class with a set time format and timezone treatment."""
+    if time_format is None and timezone_treatment == "permit":
+        return FormattedTime
+
+    dict_ = FormattedTime.__dict__.copy()
+    dict_["TIME_FORMAT"] = time_format
+    dict_["TIMEZONE_TREATMENT"] = timezone_treatment
+
+    return type("FormattedTime", (FormattedTime, *FormattedTime.__bases__), dict_)
+
+
 class ReportingPeriod(dt.date):
     """A reporting period field, with the type of reporting period supplied"""
 

src/dve/metadata_parser/model_generator.py

@@ -72,6 +72,7 @@ def constr(
     "identifier": domain_types.identifier,
     "orgid": domain_types.OrgID,
     "formatteddatetime": domain_types.formatteddatetime,
+    "formattedtime": domain_types.formattedtime,
     "conformatteddate": domain_types.conformatteddate,
     "reportingperiodstart": domain_types.reportingperiod(reporting_period_type="start"),
     "reportingperiodend": domain_types.reportingperiod(reporting_period_type="end"),

tests/test_core_engine/test_backends/fixtures.py

@@ -2,7 +2,7 @@
 
 # pylint: disable=redefined-outer-name
 import json
-from datetime import date, datetime
+from datetime import date, datetime, time
 from pathlib import Path
 from tempfile import TemporaryDirectory
 from typing import Any, Dict, Iterator, List, Tuple
@@ -83,10 +83,10 @@ def temp_duckdb_dir():
 
 @pytest.fixture
 def temp_csv_file(temp_duckdb_dir: Path):
-    header: str = "ID,varchar_field,bigint_field,date_field,timestamp_field"
+    header: str = "ID,varchar_field,bigint_field,date_field,timestamp_field,time_field"
     typed_data = [
-        [1, "hi", 3, date(2023, 1, 3), datetime(2023, 1, 3, 12, 0, 3)],
-        [2, "bye", 4, date(2023, 3, 7), datetime(2023, 5, 9, 15, 21, 53)],
+        [1, "hi", 3, date(2023, 1, 3), datetime(2023, 1, 3, 12, 0, 3), time(12, 0, 0)],
+        [2, "bye", 4, date(2023, 3, 7), datetime(2023, 5, 9, 15, 21, 53), time(13, 0 ,0)],
     ]
 
     class SimpleModel(BaseModel):
@@ -95,6 +95,7 @@ class SimpleModel(BaseModel):
         bigint_field: int
         date_field: date
         timestamp_field: datetime
+        time_field: time
 
     with open(temp_duckdb_dir.joinpath("dummy.csv"), mode="w") as csv_file:
         csv_file.write(header + "\n")

tests/test_core_engine/test_backends/test_implementations/test_duckdb/test_data_contract.py

@@ -41,6 +41,14 @@ def test_duckdb_data_contract_csv(temp_csv_file):
                             "bigint_field": "NonNegativeInt",
                             "date_field": "date",
                             "timestamp_field": "datetime",
+                            "time_field": {
+                                "description": "test",
+                                "callable": "formattedtime",
+                                "constraints": {
+                                    "time_format": "%Y-%m-%d",
+                                    "timezone_treatment": "forbid"
+                                }
+                            }
                         },
                         "reader_config": {
                             ".csv": {