fabrictools 0.7.2__tar.gz → 0.7.4__tar.gz

{fabrictools-0.7.2 → fabrictools-0.7.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fabrictools
-Version: 0.7.2
+Version: 0.7.4
 Summary: User-friendly PySpark helpers for Microsoft Fabric Lakehouses and Warehouses
 Author-email: Willy Kinfoussia <willy.kinfoussia@gmail.com>
 License: MIT

{fabrictools-0.7.2 → fabrictools-0.7.4}/fabrictools/_version.py

@@ -1,3 +1,3 @@
 """Single source of truth for package version."""
 
-__version__ = "0.7.2"
+__version__ = "0.7.4"

{fabrictools-0.7.2 → fabrictools-0.7.4}/fabrictools/io/lakehouse.py

@@ -186,6 +186,21 @@ def _dedupe_preserve_order(values: list[str]) -> list[str]:
     return deduped
 
 
+def _enable_delta_column_mapping_on_path(spark: SparkSession, full_path: str) -> None:
+    """Upgrade an existing Delta table path to column mapping mode ``name``."""
+    escaped_path = full_path.replace("`", "``")
+    spark.sql(
+        f"""
+        ALTER TABLE delta.`{escaped_path}`
+        SET TBLPROPERTIES (
+            'delta.columnMapping.mode' = 'name',
+            'delta.minReaderVersion' = '2',
+            'delta.minWriterVersion' = '5'
+        )
+        """
+    )
+
+
 def _detect_partition_columns(
     df: DataFrame, threshold_bytes: int = 1_073_741_824
 ) -> list[str]:
@@ -281,6 +296,7 @@ def write_lakehouse(
     spark: Optional[SparkSession] = None,
     *,
     normalize_column_names: bool = True,
+    enable_column_mapping: bool = False,
     auto_partition: bool = True,
     auto_partition_threshold_bytes: int = 1_073_741_824,
 ) -> None:
@@ -303,6 +319,9 @@ def write_lakehouse(
         :py:func:`fabrictools.rename_columns_normalized` before
         resolving ``partition_by`` and writing. If ``False``, keep physical column
         names unchanged.
+    :param enable_column_mapping: If ``True`` and ``format="delta"``, writes table
+        properties required for Delta column mapping (mode ``name``), allowing
+        column names with spaces or special characters.
     :param auto_partition: If ``True`` (default), automatically partition the data
         by detected date columns if they exist.
     :type df: ~pyspark.sql.DataFrame
@@ -313,6 +332,7 @@ def write_lakehouse(
     :type format: str
     :type spark: ~pyspark.sql.SparkSession | None
     :type normalize_column_names: bool
+    :type enable_column_mapping: bool
 
     .. rubric:: Example
 
@@ -367,6 +387,22 @@ def write_lakehouse(
     writer = df.write.format(format).option("overwriteSchema", "true").mode(mode)
     if format.lower() == "parquet":
         writer = writer.option("datetimeRebaseMode", "CORRECTED")
+    elif format.lower() == "delta" and enable_column_mapping:
+        # If target already exists as a Delta table, upgrade protocol first so
+        # overwrite with business-friendly names (spaces, capitals, etc.) works.
+        try:
+            from delta.tables import DeltaTable  # type: ignore[import-untyped]  # noqa: PLC0415
+
+            if DeltaTable.isDeltaTable(_, full_path):
+                _enable_delta_column_mapping_on_path(_, full_path)
+        except Exception:
+            # Non-blocking: the write options below still apply for new tables.
+            pass
+        writer = (
+            writer.option("delta.columnMapping.mode", "name")
+            .option("delta.minReaderVersion", "2")
+            .option("delta.minWriterVersion", "5")
+        )
         
     if effective_partition_by:
         writer = writer.partitionBy(*effective_partition_by)

{fabrictools-0.7.2 → fabrictools-0.7.4}/fabrictools/prepare/business.py

@@ -130,6 +130,7 @@ def make_business_ready(
                 mode=mode,
                 spark=_spark,
                 normalize_column_names=False,
+                enable_column_mapping=True,
             )
 
             processed_tables.append(

{fabrictools-0.7.2 → fabrictools-0.7.4}/fabrictools.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: fabrictools
-Version: 0.7.2
+Version: 0.7.4
 Summary: User-friendly PySpark helpers for Microsoft Fabric Lakehouses and Warehouses
 Author-email: Willy Kinfoussia <willy.kinfoussia@gmail.com>
 License: MIT