framespec 0.0.1__tar.gz → 0.0.2__tar.gz

{framespec-0.0.1 → framespec-0.0.2}/PKG-INFO

@@ -1,8 +1,12 @@
-Metadata-Version: 2.3
+Metadata-Version: 2.4
 Name: framespec
-Version: 0.0.1
+Version: 0.0.2
 Summary: Declarative DataFrame specifications for PySpark.
-Author: Ryan
+Author: RyPy
+License-Expression: MIT
+Requires-Dist: pyspark>=4.0.0
+Requires-Dist: packaging>=24.0
+Requires-Dist: typing-extensions>=4.10.0
 Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 

framespec-0.0.2/pyproject.toml

@@ -0,0 +1,27 @@
+[project]
+name = "framespec"
+version = "0.0.2"
+description = "Declarative DataFrame specifications for PySpark."
+readme = "README.md"
+requires-python = ">=3.12"
+license = "MIT"
+dependencies = [
+    "pyspark >= 4.0.0",
+    "packaging >= 24.0",
+    "typing-extensions >= 4.10.0",
+]
+
+[[project.authors]]
+name = "RyPy"
+
+[dependency-groups]
+dev = [
+    "pytest >= 8.0",
+    "pytest-cov >= 5.0",
+    "mypy >= 1.10",
+    "ruff >= 0.5.0",
+]
+
+[build-system]
+requires = ["uv-build"]
+build-backend = "uv_build"

framespec-0.0.2/pyproject.toml.orig

@@ -0,0 +1,26 @@
+[project]
+name = "framespec"
+version = "0.0.2"
+description = "Declarative DataFrame specifications for PySpark."
+readme = "README.md"
+authors = [{ name = "RyPy" }]
+requires-python = ">=3.12"
+license = "MIT"
+
+dependencies = [
+    "pyspark >= 4.0.0",
+    "packaging >= 24.0",
+    "typing-extensions >= 4.10.0",
+]
+
+[dependency-groups]
+dev = [
+    "pytest >= 8.0",
+    "pytest-cov >= 5.0",
+    "mypy >= 1.10",
+    "ruff >= 0.5.0",
+]
+
+[build-system]
+requires = ["uv-build"]
+build-backend = "uv_build"

framespec-0.0.2/src/framespec/core/types.py

@@ -0,0 +1,225 @@
+from dataclasses import dataclass
+from pyspark.sql import types as T
+
+
+# -- Base Type --------------------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class ColumnType:
+    """
+    Base class for all framespec column types.
+
+    Declarative only:
+    - immutable
+    - engine-agnostic except for .dtype (Spark DataType)
+    """
+    nullable: bool = True
+
+    @property
+    def dtype(self) -> T.DataType:
+        raise NotImplementedError
+
+
+# -- Integer Types ----------------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Byte(ColumnType):
+    min_value: int | None = None
+    max_value: int | None = None
+
+    @property
+    def dtype(self) -> T.ByteType:
+        return T.ByteType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Short(ColumnType):
+    min_value: int | None = None
+    max_value: int | None = None
+
+    @property
+    def dtype(self) -> T.ShortType:
+        return T.ShortType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Int(ColumnType):
+    min_value: int | None = None
+    max_value: int | None = None
+
+    @property
+    def dtype(self) -> T.IntegerType:
+        return T.IntegerType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Long(ColumnType):
+    min_value: int | None = None
+    max_value: int | None = None
+
+    @property
+    def dtype(self) -> T.LongType:
+        return T.LongType()
+
+
+# -- Floating Types ---------------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Float(ColumnType):
+    min_value: float | None = None
+    max_value: float | None = None
+
+    @property
+    def dtype(self) -> T.FloatType:
+        return T.FloatType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Double(ColumnType):
+    min_value: float | None = None
+    max_value: float | None = None
+
+    @property
+    def dtype(self) -> T.DoubleType:
+        return T.DoubleType()
+
+
+# -- Decimal Type -----------------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Decimal(ColumnType):
+    precision: int
+    scale: int = 0
+    min_value: float | None = None
+    max_value: float | None = None
+
+    @property
+    def dtype(self) -> T.DecimalType:
+        return T.DecimalType(self.precision, self.scale)
+
+
+# -- String & Binary Types --------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class String(ColumnType):
+    min_length: int | None = None
+    max_length: int | None = None
+    pattern: str | None = None
+    allowed_values: list[str] | None = None
+
+    @property
+    def dtype(self) -> T.StringType:
+        return T.StringType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Binary(ColumnType):
+    min_length: int | None = None
+    max_length: int | None = None
+
+    @property
+    def dtype(self) -> T.BinaryType:
+        return T.BinaryType()
+
+
+# -- Boolean Type -----------------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Boolean(ColumnType):
+    @property
+    def dtype(self) -> T.BooleanType:
+        return T.BooleanType()
+
+
+# -- Date & Timestamp Types -------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Date(ColumnType):
+    @property
+    def dtype(self) -> T.DateType:
+        return T.DateType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class Timestamp(ColumnType):
+    @property
+    def dtype(self) -> T.TimestampType:
+        return T.TimestampType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class TimestampNTZ(ColumnType):
+    @property
+    def dtype(self) -> T.TimestampNTZType:
+        return T.TimestampNTZType()
+
+
+# -- Complex Types: Array, Map, Struct -------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Array(ColumnType):
+    element_type: ColumnType
+    min_items: int | None = None
+    max_items: int | None = None
+    unique_items: bool = False
+
+    @property
+    def dtype(self) -> T.ArrayType:
+        return T.ArrayType(
+            elementType=self.element_type.dtype,
+            containsNull=self.element_type.nullable,
+        )
+
+
+@dataclass(frozen=True, kw_only=True)
+class Map(ColumnType):
+    key_type: ColumnType
+    value_type: ColumnType
+    min_properties: int | None = None
+    max_properties: int | None = None
+
+    @property
+    def dtype(self) -> T.MapType:
+        return T.MapType(
+            keyType=self.key_type.dtype,
+            valueType=self.value_type.dtype,
+            valueContainsNull=self.value_type.nullable,
+        )
+
+
+@dataclass(frozen=True, kw_only=True)
+class Struct(ColumnType):
+    fields: dict[str, ColumnType]
+    required: list[str] | None = None
+    min_properties: int | None = None
+    max_properties: int | None = None
+
+    @property
+    def dtype(self) -> T.StructType:
+        return T.StructType([
+            T.StructField(name, type_.dtype, type_.nullable)
+            for name, type_ in self.fields.items()
+        ])
+
+
+# -- Special Spark Types ----------------------------------------------------
+
+@dataclass(frozen=True, kw_only=True)
+class Null(ColumnType):
+    @property
+    def dtype(self) -> T.NullType:
+        return T.NullType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class CalendarInterval(ColumnType):
+    @property
+    def dtype(self) -> T.CalendarIntervalType:
+        return T.CalendarIntervalType()
+
+
+@dataclass(frozen=True, kw_only=True)
+class DayTimeInterval(ColumnType):
+    @property
+    def dtype(self) -> T.DayTimeIntervalType:
+        return T.DayTimeIntervalType()

framespec-0.0.1/pyproject.toml

@@ -1,14 +0,0 @@
-[project]
-name = "framespec"
-version = "0.0.1"
-description = "Declarative DataFrame specifications for PySpark."
-readme = "README.md"
-requires-python = ">=3.12"
-dependencies = []
-
-[[project.authors]]
-name = "Ryan"
-
-[build-system]
-requires = ["uv_build>=0.12.5,<0.13.0"]
-build-backend = "uv_build"

framespec-0.0.1/pyproject.toml.orig

@@ -1,12 +0,0 @@
-[project]
-name = "framespec"
-version = "0.0.1"
-description = "Declarative DataFrame specifications for PySpark."
-readme = "README.md"
-authors = [{ name = "Ryan" }]
-requires-python = ">=3.12"
-dependencies = []
-
-[build-system]
-requires = ["uv_build>=0.12.5,<0.13.0"]
-build-backend = "uv_build"