From 1280c6edd8e8f7e0c0dcf78f967c6e91064052ee Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 27 Jul 2021 09:31:08 -0400
Subject: [PATCH 1/3] Add a test grouping on null

---
 datafusion/tests/sql.rs | 45 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 45 insertions(+)

diff --git a/datafusion/tests/sql.rs b/datafusion/tests/sql.rs
index d9f7c6ea41211..0e373c91ffdfc 100644
--- a/datafusion/tests/sql.rs
+++ b/datafusion/tests/sql.rs
@@ -2966,6 +2966,45 @@ async fn query_count_distinct() -> Result<()> {
     Ok(())
 }
 
+#[tokio::test]
+async fn query_group_on_null() -> Result<()> {
+    let schema = Arc::new(Schema::new(vec![Field::new("c1", DataType::Int32, true)]));
+
+    let data = RecordBatch::try_new(
+        schema.clone(),
+        vec![Arc::new(Int32Array::from(vec![
+            Some(0),
+            Some(3),
+            None,
+            Some(1),
+            Some(3),
+        ]))],
+    )?;
+
+    let table = MemTable::try_new(schema, vec![vec![data]])?;
+
+    let mut ctx = ExecutionContext::new();
+    ctx.register_table("test", Arc::new(table))?;
+    let sql = "SELECT COUNT(*), c1 FROM test GROUP BY c1";
+
+    let actual = execute_to_batches(&mut ctx, sql).await;
+
+    // this is incorrect: the results should also
+    // include a row for NULL (c1=NULL, count = 1)
+    // https://github.com/apache/arrow-datafusion/issues/782
+    let expected = vec![
+        "+-----------------+----+",
+        "| COUNT(UInt8(1)) | c1 |",
+        "+-----------------+----+",
+        "| 2               | 3  |",
+        "| 2               | 0  |",
+        "| 1               | 1  |",
+        "+-----------------+----+",
+    ];
+    assert_batches_eq!(expected, &actual);
+    Ok(())
+}
+
 #[tokio::test]
 async fn query_on_string_dictionary() -> Result<()> {
     // Test to ensure DataFusion can operate on dictionary types
@@ -3019,6 +3058,12 @@ async fn query_on_string_dictionary() -> Result<()> {
     let expected = vec![vec!["2"]];
     assert_eq!(expected, actual);
 
+    // grouping
+    let sql = "SELECT d1, COUNT(*) FROM test group by d1";
+    let actual = execute(&mut ctx, sql).await;
+    let expected = vec![vec!["one", "1"], vec!["three", "1"]];
+    assert_eq!(expected, actual);
+
     Ok(())
 }
 

From 5ed43ac9af6a4255856a7e83f280eeafe1f3afa6 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 27 Jul 2021 09:33:12 -0400
Subject: [PATCH 2/3] Remove mistakenly added test

---
 datafusion/tests/sql.rs | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/datafusion/tests/sql.rs b/datafusion/tests/sql.rs
index 0e373c91ffdfc..ea7f7f36c679f 100644
--- a/datafusion/tests/sql.rs
+++ b/datafusion/tests/sql.rs
@@ -3058,12 +3058,6 @@ async fn query_on_string_dictionary() -> Result<()> {
     let expected = vec![vec!["2"]];
     assert_eq!(expected, actual);
 
-    // grouping
-    let sql = "SELECT d1, COUNT(*) FROM test group by d1";
-    let actual = execute(&mut ctx, sql).await;
-    let expected = vec![vec!["one", "1"], vec!["three", "1"]];
-    assert_eq!(expected, actual);
-
     Ok(())
 }
 

From 9f2067afca1c0d5c9fd4f2ea8806dd4dd4c67257 Mon Sep 17 00:00:00 2001
From: Andrew Lamb <andrew@nerdnetworks.org>
Date: Tue, 27 Jul 2021 11:00:26 -0400
Subject: [PATCH 3/3] use sort blind comparison

---
 datafusion/tests/sql.rs | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/datafusion/tests/sql.rs b/datafusion/tests/sql.rs
index ea7f7f36c679f..ffa25115a4aa0 100644
--- a/datafusion/tests/sql.rs
+++ b/datafusion/tests/sql.rs
@@ -3001,7 +3001,7 @@ async fn query_group_on_null() -> Result<()> {
         "| 1               | 1  |",
         "+-----------------+----+",
     ];
-    assert_batches_eq!(expected, &actual);
+    assert_batches_sorted_eq!(expected, &actual);
     Ok(())
 }