[SPARK-51850] Fix DataFrame.execute to reset previously received Arrow batch data

### What changes were proposed in this pull request?

This PR aims to fix `DataFrame.execute` to reset previously received Arrow batch data.

### Why are the changes needed?

To prevent wrong data duplication when repeating execution on the same `DataFrame`.

### Does this PR introduce _any_ user-facing change?

This, this is a bug fix.

### How was this patch tested?

Pass the CIs.

### Was this patch authored or co-authored using generative AI tooling?

No.

Closes #72 from dongjoon-hyun/SPARK-51850.

Authored-by: Dongjoon Hyun <[email protected]>
Signed-off-by: Dongjoon Hyun <[email protected]>

Author: dongjoon-hyun

Sources/SparkConnect/DataFrame.swift

@@ -151,6 +151,9 @@ public actor DataFrame: Sendable {
 
   /// Execute the plan and try to fill `schema` and `batches`.
   private func execute() async throws {
+    // Clear all existing batches.
+    self.batches.removeAll()
+
     try await withGRPCClient(
       transport: .http2NIOPosix(
         target: .dns(host: spark.client.host, port: spark.client.port),

Tests/SparkConnectTests/DataFrameTests.swift

@@ -311,6 +311,15 @@ struct DataFrameTests {
     await spark.stop()
   }
 
+  @Test
+  func collectMultiple() async throws {
+    let spark = try await SparkSession.builder.getOrCreate()
+    let df = try await spark.range(1)
+    #expect(try await df.collect().count == 1)
+    #expect(try await df.collect().count == 1)
+    await spark.stop()
+  }
+
   @Test
   func head() async throws {
     let spark = try await SparkSession.builder.getOrCreate()