Make arrow from relation return record batch reader

Author: evertlammerts

duckdb/__init__.pyi

@@ -453,7 +453,7 @@ class DuckDBPyRelation:
     def set_alias(self, alias: str) -> DuckDBPyRelation: ...
     def show(self, max_width: Optional[int] = None, max_rows: Optional[int] = None, max_col_width: Optional[int] = None, null_value: Optional[str] = None, render_mode: Optional[RenderMode] = None) -> None: ...
     def sql_query(self) -> str: ...
-    def to_arrow_table(self, batch_size: int = ...) -> pyarrow.lib.Table: ...
+    def to_arrow_table(self, batch_size: int = ...) -> pyarrow.lib.RecordBatchReader: ...
     def to_csv(
             self,
             file_name: str,

duckdb/experimental/spark/sql/dataframe.py

@@ -75,7 +75,7 @@ def toArrow(self) -> "pa.Table":
         age: [[2,5]]
         name: [["Alice","Bob"]]
         """
-        return self.relation.arrow()
+        return self.relation.to_arrow_table()
 
     def createOrReplaceTempView(self, name: str) -> None:
         """Creates or replaces a local temporary view with this :class:`DataFrame`.

src/duckdb_py/pyrelation/initialize.cpp

@@ -61,7 +61,7 @@ static void InitializeConsumers(py::class_<DuckDBPyRelation> &m) {
 	         py::arg("date_as_object") = false)
 	    .def("fetch_df_chunk", &DuckDBPyRelation::FetchDFChunk, "Execute and fetch a chunk of the rows",
 	         py::arg("vectors_per_chunk") = 1, py::kw_only(), py::arg("date_as_object") = false)
-	    .def("arrow", &DuckDBPyRelation::ToArrowTable, "Execute and fetch all rows as an Arrow Table",
+	    .def("arrow", &DuckDBPyRelation::ToRecordBatch, "Execute and return an Arrow Record Batch Reader that yields all rows",
 	         py::arg("batch_size") = 1000000)
 	    .def("fetch_arrow_table", &DuckDBPyRelation::ToArrowTable, "Execute and fetch all rows as an Arrow Table",
 	         py::arg("batch_size") = 1000000)
@@ -78,7 +78,7 @@ static void InitializeConsumers(py::class_<DuckDBPyRelation> &m) {
 		)";
 	m.def("__arrow_c_stream__", &DuckDBPyRelation::ToArrowCapsule, capsule_docs,
 	      py::arg("requested_schema") = py::none());
-	m.def("record_batch", &DuckDBPyRelation::ToRecordBatch,
+	m.def("fetch_record_batch", &DuckDBPyRelation::ToRecordBatch,
 	      "Execute and return an Arrow Record Batch Reader that yields all rows", py::arg("batch_size") = 1000000)
 	    .def("fetch_arrow_reader", &DuckDBPyRelation::ToRecordBatch,
 	         "Execute and return an Arrow Record Batch Reader that yields all rows", py::arg("batch_size") = 1000000);