Hive: Support MAP<K, V> column types (#2346)

wugeer · web-flow · commit 5d61fdded630 · 2026-05-27T10:39:08.000Z
diff --git a/src/ast/data_type.rs b/src/ast/data_type.rs
@@ -439,10 +439,11 @@ pub enum DataType {
     Custom(ObjectName, Vec<String>),
     /// Arrays.
     Array(ArrayElemTypeDef),
-    /// Map, see [ClickHouse].
+    /// Map, see [ClickHouse], [Hive].
     ///
     /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/map
-    Map(Box<DataType>, Box<DataType>),
+    /// [Hive]: https://hive.apache.org/docs/latest/language/languagemanual-types/
+    Map(Box<DataType>, Box<DataType>, MapBracketKind),
     /// Tuple, see [ClickHouse].
     ///
     /// [ClickHouse]: https://clickhouse.com/docs/en/sql-reference/data-types/tuple
@@ -785,9 +786,14 @@ impl fmt::Display for DataType {
             DataType::LowCardinality(data_type) => {
                 write!(f, "LowCardinality({data_type})")
             }
-            DataType::Map(key_data_type, value_data_type) => {
-                write!(f, "Map({key_data_type}, {value_data_type})")
-            }
+            DataType::Map(key_data_type, value_data_type, bracket) => match bracket {
+                MapBracketKind::Parentheses => {
+                    write!(f, "Map({key_data_type}, {value_data_type})")
+                }
+                MapBracketKind::AngleBrackets => {
+                    write!(f, "MAP<{key_data_type}, {value_data_type}>")
+                }
+            },
             DataType::Tuple(fields) => {
                 write!(f, "Tuple({})", display_comma_separated(fields))
             }
@@ -904,6 +910,17 @@ pub enum StructBracketKind {
     AngleBrackets,
 }
 
+/// Type of brackets used for `MAP` types.
+#[derive(Debug, Clone, Copy, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub enum MapBracketKind {
+    /// Example: `Map(String, UInt16)`
+    Parentheses,
+    /// Example: `MAP<STRING, INT>`
+    AngleBrackets,
+}
+
 /// Timestamp and Time data types information about TimeZone formatting.
 ///
 /// This is more related to a display information than real differences between each variant. To
diff --git a/src/ast/mod.rs b/src/ast/mod.rs
@@ -53,7 +53,7 @@ use crate::{
 
 pub use self::data_type::{
     ArrayElemTypeDef, BinaryLength, CharLengthUnits, CharacterLength, DataType, EnumMember,
-    ExactNumberInfo, IntervalFields, StructBracketKind, TimezoneInfo,
+    ExactNumberInfo, IntervalFields, MapBracketKind, StructBracketKind, TimezoneInfo,
 };
 pub use self::dcl::{
     AlterRoleOperation, CreateRole, Grant, ResetConfig, Revoke, RoleOption, SecondaryRoles,
diff --git a/src/dialect/hive.rs b/src/dialect/hive.rs
@@ -79,4 +79,9 @@ impl Dialect for HiveDialect {
     fn supports_from_first_insert(&self) -> bool {
         true
     }
+
+    /// See <https://hive.apache.org/docs/latest/language/languagemanual-types/>
+    fn supports_map_literal_with_angle_brackets(&self) -> bool {
+        true
+    }
 }
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
@@ -12748,6 +12748,7 @@ impl<'a> Parser<'a> {
                     Ok(DataType::Map(
                         Box::new(key_data_type),
                         Box::new(value_data_type),
+                        MapBracketKind::AngleBrackets,
                     ))
                 }
                 Keyword::MAP if dialect_is!(dialect is ClickHouseDialect | GenericDialect) => {
@@ -12756,6 +12757,7 @@ impl<'a> Parser<'a> {
                     Ok(DataType::Map(
                         Box::new(key_data_type),
                         Box::new(value_data_type),
+                        MapBracketKind::Parentheses,
                     ))
                 }
                 Keyword::NESTED if dialect_is!(dialect is ClickHouseDialect | GenericDialect) => {
diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs
@@ -767,7 +767,8 @@ fn parse_create_table_with_nested_data_types() {
                         name: Ident::new("m"),
                         data_type: DataType::Map(
                             Box::new(DataType::String(None)),
-                            Box::new(DataType::UInt16)
+                            Box::new(DataType::UInt16),
+                            MapBracketKind::Parentheses
                         ),
                         options: vec![],
                     },
diff --git a/tests/sqlparser_hive.rs b/tests/sqlparser_hive.rs
@@ -562,6 +562,13 @@ fn test_tample_sample() {
     hive().verified_stmt("SELECT * FROM source TABLESAMPLE (10 ROWS)");
 }
 
+#[test]
+fn parse_create_table_with_map_column_comment() {
+    hive().verified_stmt(
+        "CREATE TABLE target (kv_map MAP<STRING, STRING> COMMENT 'kv col comment') COMMENT 'this is table comment'",
+    );
+}
+
 fn hive() -> TestedDialects {
     TestedDialects::new(vec![Box::new(HiveDialect {})])
 }

Original file line number	Diff line number	Diff line change
`@@ -79,4 +79,9 @@ impl Dialect for HiveDialect {`
`79`	`79`	`fn supports_from_first_insert(&self) -> bool {`
`80`	`80`	`true`
`81`	`81`	`}`
	`82`	`+`
	`83`	`+ /// See <https://hive.apache.org/docs/latest/language/languagemanual-types/>`
	`84`	`+ fn supports_map_literal_with_angle_brackets(&self) -> bool {`
	`85`	`+ true`
	`86`	`+ }`
`82`	`87`	`}`
Original file line number	Diff line number	Diff line change
`@@ -562,6 +562,13 @@ fn test_tample_sample() {`
`562`	`562`	`hive().verified_stmt("SELECT * FROM source TABLESAMPLE (10 ROWS)");`
`563`	`563`	`}`
`564`	`564`
	`565`	`+#[test]`
	`566`	`+fn parse_create_table_with_map_column_comment() {`
	`567`	`+ hive().verified_stmt(`
	`568`	`+ "CREATE TABLE target (kv_map MAP<STRING, STRING> COMMENT 'kv col comment') COMMENT 'this is table comment'",`
	`569`	`+ );`
	`570`	`+}`
	`571`	`+`
`565`	`572`	`fn hive() -> TestedDialects {`
`566`	`573`	`TestedDialects::new(vec![Box::new(HiveDialect {})])`
`567`	`574`	`}`