sqlframe 3.42.0__tar.gz → 3.43.1__tar.gz

{sqlframe-3.42.0 → sqlframe-3.43.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sqlframe
-Version: 3.42.0
+Version: 3.43.1
 Summary: Turning PySpark Into a Universal DataFrame API
 Home-page: https://github.com/eakmanrq/sqlframe
 Author: Ryan Eakman

{sqlframe-3.42.0 → sqlframe-3.43.1}/docs/bigquery.md

@@ -526,7 +526,6 @@ See something that you would like to see supported? [Open an issue](https://gith
 * [to_date](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_date.html)
 * [to_timestamp](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_timestamp.html)
 * [to_timestamp_ntz](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_timestamp_ntz.html)
-* [transform](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.transform.html)
 * [translate](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.translate.html)
 * [trim](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.trim.html)
 * [trunc](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.trunc.html)

{sqlframe-3.42.0 → sqlframe-3.43.1}/docs/snowflake.md

@@ -522,7 +522,6 @@ See something that you would like to see supported? [Open an issue](https://gith
 * [to_number](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_number.html)
 * [to_timestamp](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_timestamp.html)
 * [to_timestamp_ntz](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.to_timestamp_ntz.html)
-* [transform](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.transform.html)
 * [translate](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.translate.html)
 * [trim](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.trim.html)
 * [trunc](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.functions.trunc.html)

sqlframe-3.43.1/mypy.ini

@@ -0,0 +1,13 @@
+[mypy]
+
+[mypy-duckdb.*]
+ignore_missing_imports = True
+
+[mypy-pyarrow.*]
+ignore_missing_imports = True
+
+[mypy-findspark.*]
+ignore_missing_imports = True
+
+[mypy-redshift_connector.*]
+ignore_missing_imports = True

{sqlframe-3.42.0 → sqlframe-3.43.1}/setup.py

@@ -20,7 +20,7 @@ setup(
     python_requires=">=3.9",
     install_requires=[
         "prettytable<4",
-        "sqlglot>=24.0.0,<27.15",
+        "sqlglot>=24.0.0,<27.17",
         "typing_extensions",
         "more-itertools",
     ],
@@ -30,7 +30,7 @@ setup(
             "google-cloud-bigquery-storage>=2,<3",
         ],
         "dev": [
-            "duckdb>=1.2,<1.4",
+            "duckdb>=1.2,<1.5",
             "findspark>=2,<3",
             "mypy>=1.10.0,<1.19",
             "openai>=1.30,<2",
@@ -56,7 +56,7 @@ setup(
             "pymdown-extensions",
         ],
         "duckdb": [
-            "duckdb>=1.2,<1.4",
+            "duckdb>=1.2,<1.5",
             "pandas>=2,<3",
         ],
         "openai": [

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe/_version.py

@@ -28,7 +28,7 @@ version_tuple: VERSION_TUPLE
 commit_id: COMMIT_ID
 __commit_id__: COMMIT_ID
 
-__version__ = version = '3.42.0'
-__version_tuple__ = version_tuple = (3, 42, 0)
+__version__ = version = '3.43.1'
+__version_tuple__ = version_tuple = (3, 43, 1)
 
-__commit_id__ = commit_id = 'gadde0dd66'
+__commit_id__ = commit_id = 'g07ceeb0ec'

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe/base/dataframe.py

@@ -1665,10 +1665,34 @@ class BaseDataFrame(t.Generic[SESSION, WRITER, NA, STAT, GROUP_DATA]):
         def should_drop_expression(expr: exp.Expression) -> bool:
             # Check against fully qualified Column objects and
             # Check against unqualified string column names (drop ALL columns with this name)
-            if expr.sql() in drop_sql or (
-                isinstance(expr, exp.Column) and expr.alias_or_name in column_names
-            ):
+            if expr.sql() in drop_sql:
                 return True
+
+            if isinstance(expr, exp.Column) and (alias_or_name := expr.alias_or_name):
+                # Check direct match first
+                if alias_or_name in column_names:
+                    return True
+
+                # Handle string column references that contain aliases
+                for col_name in column_names:
+                    if ("." in col_name) and alias_or_name == (col_name.split(".", maxsplit=1)[-1]):
+                        # Extract the column name part after the last dot
+                        return True
+
+                # Handle case where normalized columns have table qualifiers but actual expressions
+                # are unqualified. This happens when using aliased column references like
+                # f.col('df.foo')
+
+                # Check if any drop column matches by column name AND table qualifier
+                for drop_col in drop_cols:
+                    if ((drop_expression := drop_col.expression).alias_or_name) == alias_or_name:
+                        if expr_table := expr.table:
+                            drop_table = drop_expression.args.get("table")
+                            if (not drop_table) or (expr_table == drop_table):
+                                return True
+                        else:
+                            return True
+
             return False
 
         new_expressions = [expr for expr in current_expressions if not should_drop_expression(expr)]

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe/base/functions.py

@@ -2729,7 +2729,7 @@ def aggregate(
     return Column.invoke_expression_over_column(col, expression.Reduce, **kwargs)
 
 
-@meta(unsupported_engines="postgres")
+@meta(unsupported_engines=["bigquery", "postgres", "snowflake"])
 def transform(
     col: ColumnOrName,
     f: t.Union[t.Callable[[Column], Column], t.Callable[[Column, Column], Column]],
@@ -7169,12 +7169,32 @@ def _lambda_quoted(value: str) -> t.Optional[bool]:
 
 @meta()
 def _get_lambda_from_func(lambda_expression: t.Callable):
-    variables = [
-        expression.to_identifier(x, quoted=_lambda_quoted(x))
-        for x in lambda_expression.__code__.co_varnames
-    ]
+    import inspect
+
+    # Get the function signature
+    sig = inspect.signature(lambda_expression)
+    param_names = list(sig.parameters.keys())
+
+    # Check if this looks like a column function (single 'col' parameter)
+    if len(param_names) == 1 and param_names[0] in ["col", "column"]:
+        # Wrap column functions to work with transform
+        variables = [expression.to_identifier("x", quoted=_lambda_quoted("x"))]
+        result = lambda_expression(Column("x"))
+        return expression.Lambda(
+            this=result.column_expression,
+            expressions=variables,
+        )
+
+    # Handle regular functions and lambdas
+    var_names = lambda_expression.__code__.co_varnames[: lambda_expression.__code__.co_argcount]
+
+    variables = [expression.to_identifier(x, quoted=_lambda_quoted(x)) for x in var_names]
+
+    # Call with Column objects for each parameter
+    result = lambda_expression(*[Column(x) for x in var_names])
+
     return expression.Lambda(
-        this=lambda_expression(*[Column(x) for x in variables]).column_expression,
+        this=result.column_expression,
         expressions=variables,
     )
 

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe/base/session.py

@@ -437,7 +437,7 @@ class _BaseSession(t.Generic[CATALOG, READER, WRITER, DF, TABLE, CONN, UDF_REGIS
                 for cte in df.expression.ctes:
                     if cte.alias_or_name not in expression_ctes:
                         ctes_to_add.append(cte)
-                expression.set("with", exp.With(expressions=expression.ctes + ctes_to_add))  # type: ignore
+                expression.set("with", exp.With(expressions=ctes_to_add + expression.ctes))  # type: ignore
 
             def replace_temp_view_name_with_cte(node: exp.Expression) -> exp.Expression:
                 if isinstance(node, exp.Table):

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe/duckdb/dataframe.py

@@ -61,5 +61,5 @@ class DuckDBDataFrame(
     def toArrow(self, batch_size: t.Optional[int] = None) -> t.Union[ArrowTable, RecordBatchReader]:
         self._collect(skip_rows=True)
         if not batch_size:
-            return self.session._last_result.arrow()
+            return self.session._last_result.fetch_arrow_table()
         return self.session._last_result.fetch_record_batch(batch_size)

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sqlframe
-Version: 3.42.0
+Version: 3.43.1
 Summary: Turning PySpark Into a Universal DataFrame API
 Home-page: https://github.com/eakmanrq/sqlframe
 Author: Ryan Eakman

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe.egg-info/SOURCES.txt

@@ -5,6 +5,7 @@ LICENSE
 Makefile
 README.md
 mkdocs.yml
+mypy.ini
 pytest.ini
 renovate.json
 setup.cfg

{sqlframe-3.42.0 → sqlframe-3.43.1}/sqlframe.egg-info/requires.txt

@@ -1,6 +1,6 @@
 more-itertools
 prettytable<4
-sqlglot<27.15,>=24.0.0
+sqlglot<27.17,>=24.0.0
 typing_extensions
 
 [bigquery]
@@ -11,7 +11,7 @@ google-cloud-bigquery[pandas]<4,>=3
 databricks-sql-connector[pyarrow]<5,>=3.6
 
 [dev]
-duckdb<1.4,>=1.2
+duckdb<1.5,>=1.2
 findspark<3,>=2
 mypy<1.19,>=1.10.0
 openai<2,>=1.30
@@ -37,7 +37,7 @@ mkdocs==1.4.2
 pymdown-extensions
 
 [duckdb]
-duckdb<1.4,>=1.2
+duckdb<1.5,>=1.2
 pandas<3,>=2
 
 [openai]

{sqlframe-3.42.0 → sqlframe-3.43.1}/tests/integration/engines/duck/test_duckdb_dataframe.py

@@ -262,7 +262,7 @@ def test_explain(duckdb_employee: DuckDBDataFrame, capsys):
 ┌───────────────────────────┐
 │      COLUMN_DATA_SCAN     │
 │    ────────────────────   │
-│          ~5 Rows          │
+│          ~5 rows          │
 └───────────────────────────┘
 """.strip()
     )

{sqlframe-3.42.0 → sqlframe-3.43.1}/tests/integration/engines/duck/test_duckdb_session.py

@@ -1,5 +1,5 @@
 import pytest
-from duckdb.duckdb import ConnectionException
+from duckdb import ConnectionException
 from sqlglot import exp
 
 from sqlframe.base.types import Row

{sqlframe-3.42.0 → sqlframe-3.43.1}/tests/integration/engines/test_int_functions.py

@@ -2941,19 +2941,22 @@ def test_aggregate(get_session_and_func, get_func):
 def test_transform(get_session_and_func, get_func):
     session, transform = get_session_and_func("transform")
     when = get_func("when", session)
+    initcap = get_func("initcap", session)
+    df = session.createDataFrame([(["ab cd"],)], ["a"])
+    assert df.select(transform("a", f=lambda w: initcap(w)).alias("v")).first()[0] == ["Ab Cd"]
+    assert df.select(transform("a", f=initcap).alias("v")).first()[0] == ["Ab Cd"]
     df = session.createDataFrame([(1, [1, 2, 3, 4])], ("key", "some_values"))
     assert df.select(transform("some_values", lambda x: x * 2).alias("doubled")).collect() == [
         Row(doubled=[2, 4, 6, 8])
     ]
 
-    if isinstance(session, PySparkSession):
-
-        def alternate(x, i):
-            return when(i % 2 == 0, x).otherwise(-x)
+    def alternate(x, i):
+        return when(i % 2 == 0, x).otherwise(-x)
 
-        assert df.select(transform("some_values", alternate).alias("alternated")).collect() == [
-            Row(alternated=[1, -2, 3, -4])
-        ]
+    assert df.select(transform("some_values", alternate).alias("alternated")).collect() in [
+        [Row(alternated=[1, -2, 3, -4])],
+        [Row(alternated=[-1, 2, -3, 4])],
+    ]
 
 
 def test_exists(get_session_and_func):

{sqlframe-3.42.0 → sqlframe-3.43.1}/tests/integration/test_int_dataframe.py

@@ -1865,6 +1865,18 @@ def test_drop_column_single(
     compare_frames(df, dfs)
 
 
+def test_drop_alias_column(
+    pyspark_employee: PySparkDataFrame,
+    get_df: t.Callable[[str], BaseDataFrame],
+    compare_frames: t.Callable,
+):
+    employee = get_df("employee")
+    df = pyspark_employee.select(F.col("fname"), F.col("age")).drop("age")
+
+    dfs = employee.alias("df").select(SF.col("fname"), SF.col("age")).drop("df.age")
+    compare_frames(df, dfs)
+
+
 def test_drop_column_reference_join(
     pyspark_employee: PySparkDataFrame,
     pyspark_store: PySparkDataFrame,
@@ -1892,6 +1904,61 @@ def test_drop_column_reference_join(
     compare_frames(df, dfs, sort=True)
 
 
+def test_drop_column_join_column_df_reference(
+    pyspark_employee: PySparkDataFrame,
+    get_df: t.Callable[[str], BaseDataFrame],
+    compare_frames: t.Callable,
+):
+    df1 = pyspark_employee.sparkSession.createDataFrame(  # type: ignore
+        [{"foo": 0, "bar": "a"}, {"foo": 1, "bar": "b"}]
+    ).alias("df1")
+    df2 = pyspark_employee.sparkSession.createDataFrame([{"foo": 0, "baz": 1.5}]).alias("df2")  # type: ignore
+    df_joined = (
+        df1.join(df2, on=F.col("df1.foo") == F.col("df2.foo"), how="left")
+        .drop(df1.foo)
+        .select(df1.bar, df2.baz, df2.foo)
+    )
+
+    employee = get_df("employee")
+    dfs1 = employee.sparkSession.createDataFrame(
+        [{"foo": 0, "bar": "a"}, {"foo": 1, "bar": "b"}]
+    ).alias("dfs1")
+    dfs2 = employee.sparkSession.createDataFrame([{"foo": 0, "baz": 1.5}]).alias("dfs2")
+    dfs_joined = (
+        dfs1.join(dfs2, on=SF.col("dfs1.foo") == SF.col("dfs2.foo"), how="left")
+        .drop(dfs1.foo)
+        .select(dfs1.bar, dfs2.baz, dfs2.foo)
+    )
+
+    compare_frames(df_joined, dfs_joined, sort=True, compare_schema=False)
+
+
+def test_drop_join_column_unqualified(
+    pyspark_employee: PySparkDataFrame,
+    get_df: t.Callable[[str], BaseDataFrame],
+    compare_frames: t.Callable,
+):
+    df1 = pyspark_employee.sparkSession.createDataFrame(  # type: ignore
+        [{"foo": 0, "bar": "a"}, {"foo": 1, "bar": "b"}]
+    ).alias("df1")
+    df2 = pyspark_employee.sparkSession.createDataFrame([{"foo": 0, "baz": 1.5}]).alias("df2")  # type: ignore
+    df_joined = (
+        df1.join(df2, on=F.col("df1.foo") == F.col("df2.foo"), how="left").drop("foo")
+        # select the columns to work around column order bug
+    )
+
+    employee = get_df("employee")
+    dfs1 = employee.sparkSession.createDataFrame(
+        [{"foo": 0, "bar": "a"}, {"foo": 1, "bar": "b"}]
+    ).alias("dfs1")
+    dfs2 = employee.sparkSession.createDataFrame([{"foo": 0, "baz": 1.5}]).alias("dfs2")
+    dfs_joined = dfs1.join(dfs2, on=SF.col("dfs1.foo") == SF.col("dfs2.foo"), how="left").drop(
+        "foo"
+    )
+
+    compare_frames(df_joined, dfs_joined, sort=True, compare_schema=False)
+
+
 def test_limit(
     pyspark_employee: PySparkDataFrame,
     get_df: t.Callable[[str], BaseDataFrame],