Support Map literal syntax for DuckDB and Generic (#1344)

src/ast/mod.rs

@@ -329,6 +329,37 @@ impl fmt::Display for DictionaryField {
     }
 }
 
+/// Represents a Map expression.
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub struct Map {
+    pub entries: Vec<MapEntry>,
+}
+
+impl Display for Map {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        write!(f, "MAP {{{}}}", display_comma_separated(&self.entries))
+    }
+}
+
+/// A map field within a map.
+///
+/// [duckdb]: https://duckdb.org/docs/sql/data_types/map.html#creating-maps
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub struct MapEntry {
+    pub key: Box<Expr>,
+    pub value: Box<Expr>,
+}
+
+impl fmt::Display for MapEntry {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        write!(f, "{}: {}", self.key, self.value)
+    }
+}
+
 /// Options for `CAST` / `TRY_CAST`
 /// BigQuery: <https://cloud.google.com/bigquery/docs/reference/standard-sql/format-elements#formatting_syntax>
 #[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
@@ -764,6 +795,14 @@ pub enum Expr {
     /// ```
     /// [1]: https://duckdb.org/docs/sql/data_types/struct#creating-structs
     Dictionary(Vec<DictionaryField>),
+    /// `DuckDB` specific `Map` literal expression [1]
+    ///
+    /// Syntax:
+    /// ```sql
+    /// syntax: Map {key1: value1[, ... ]}
+    /// ```
+    /// [1]: https://duckdb.org/docs/sql/data_types/map#creating-maps
+    Map(Map),
     /// An access of nested data using subscript syntax, for example `array[2]`.
     Subscript {
         expr: Box<Expr>,
@@ -1331,6 +1370,9 @@ impl fmt::Display for Expr {
             Expr::Dictionary(fields) => {
                 write!(f, "{{{}}}", display_comma_separated(fields))
             }
+            Expr::Map(map) => {
+                write!(f, "{map}")
+            }
             Expr::Subscript {
                 expr,
                 subscript: key,

src/dialect/duckdb.rs

@@ -48,4 +48,11 @@ impl Dialect for DuckDbDialect {
     fn supports_dictionary_syntax(&self) -> bool {
         true
     }
+
+    // DuckDB uses this syntax for `MAP`s.
+    //
+    // https://duckdb.org/docs/sql/data_types/map.html#creating-maps
+    fn support_map_literal_syntax(&self) -> bool {
+        true
+    }
 }

src/dialect/generic.rs

@@ -70,4 +70,8 @@ impl Dialect for GenericDialect {
     fn supports_select_wildcard_except(&self) -> bool {
         true
     }
+
+    fn support_map_literal_syntax(&self) -> bool {
+        true
+    }
 }

src/dialect/mod.rs

@@ -215,6 +215,11 @@ pub trait Dialect: Debug + Any {
     fn supports_dictionary_syntax(&self) -> bool {
         false
     }
+    /// Returns true if the dialect supports defining object using the
+    /// syntax like `Map {1: 10, 2: 20}`.
+    fn support_map_literal_syntax(&self) -> bool {
+        false
+    }
     /// Returns true if the dialect supports lambda functions, for example:
     ///
     /// ```sql

src/parser/mod.rs

@@ -1078,6 +1078,9 @@ impl<'a> Parser<'a> {
                     let expr = self.parse_subexpr(Self::PLUS_MINUS_PREC)?;
                     Ok(Expr::Prior(Box::new(expr)))
                 }
+                Keyword::MAP if self.peek_token() == Token::LBrace && self.dialect.support_map_literal_syntax() => {
+                    self.parse_duckdb_map_literal()
+                }
                 // Here `w` is a word, check if it's a part of a multipart
                 // identifier, a function call, or a simple identifier:
                 _ => match self.peek_token().token {
@@ -2322,6 +2325,47 @@ impl<'a> Parser<'a> {
         })
     }
 
+    /// DuckDB specific: Parse a duckdb [map]
+    ///
+    /// Syntax:
+    ///
+    /// ```sql
+    /// Map {key1: value1[, ... ]}
+    /// ```
+    ///
+    /// [map]: https://duckdb.org/docs/sql/data_types/map.html#creating-maps
+    fn parse_duckdb_map_literal(&mut self) -> Result<Expr, ParserError> {
+        self.expect_token(&Token::LBrace)?;
+
+        let fields = self.parse_comma_separated(Self::parse_duckdb_map_field)?;
+
+        self.expect_token(&Token::RBrace)?;
+
+        Ok(Expr::Map(Map { entries: fields }))
+    }
+
+    /// Parse a field for a duckdb [map]
+    ///
+    /// Syntax
+    ///
+    /// ```sql
+    /// key: value
+    /// ```
+    ///
+    /// [map]: https://duckdb.org/docs/sql/data_types/map.html#creating-maps
+    fn parse_duckdb_map_field(&mut self) -> Result<MapEntry, ParserError> {
+        let key = self.parse_expr()?;
+
+        self.expect_token(&Token::Colon)?;
+
+        let value = self.parse_expr()?;
+
+        Ok(MapEntry {
+            key: Box::new(key),
+            value: Box::new(value),
+        })
+    }
+
     /// Parse clickhouse [map]
     ///
     /// Syntax

tests/sqlparser_common.rs

@@ -10077,6 +10077,101 @@ fn test_dictionary_syntax() {
     )
 }
 
+#[test]
+fn test_map_syntax() {
+    fn check(sql: &str, expect: Expr) {
+        assert_eq!(
+            all_dialects_where(|d| d.support_map_literal_syntax()).verified_expr(sql),
+            expect
+        );
+    }
+
+    check(
+        "MAP {'Alberta': 'Edmonton', 'Manitoba': 'Winnipeg'}",
+        Expr::Map(Map {
+            entries: vec![
+                MapEntry {
+                    key: Box::new(Expr::Value(Value::SingleQuotedString("Alberta".to_owned()))),
+                    value: Box::new(Expr::Value(Value::SingleQuotedString(
+                        "Edmonton".to_owned(),
+                    ))),
+                },
+                MapEntry {
+                    key: Box::new(Expr::Value(Value::SingleQuotedString(
+                        "Manitoba".to_owned(),
+                    ))),
+                    value: Box::new(Expr::Value(Value::SingleQuotedString(
+                        "Winnipeg".to_owned(),
+                    ))),
+                },
+            ],
+        }),
+    );
+
+    fn number_expr(s: &str) -> Expr {
+        Expr::Value(number(s))
+    }
+
+    check(
+        "MAP {1: 10.0, 2: 20.0}",
+        Expr::Map(Map {
+            entries: vec![
+                MapEntry {
+                    key: Box::new(number_expr("1")),
+                    value: Box::new(number_expr("10.0")),
+                },
+                MapEntry {
+                    key: Box::new(number_expr("2")),
+                    value: Box::new(number_expr("20.0")),
+                },
+            ],
+        }),
+    );
+
+    check(
+        "MAP {[1, 2, 3]: 10.0, [4, 5, 6]: 20.0}",
+        Expr::Map(Map {
+            entries: vec![
+                MapEntry {
+                    key: Box::new(Expr::Array(Array {
+                        elem: vec![number_expr("1"), number_expr("2"), number_expr("3")],
+                        named: false,
+                    })),
+                    value: Box::new(Expr::Value(number("10.0"))),
+                },
+                MapEntry {
+                    key: Box::new(Expr::Array(Array {
+                        elem: vec![number_expr("4"), number_expr("5"), number_expr("6")],
+                        named: false,
+                    })),
+                    value: Box::new(Expr::Value(number("20.0"))),
+                },
+            ],
+        }),
+    );
+
+    check(
+        "MAP {'a': 10, 'b': 20}['a']",
+        Expr::Subscript {
+            expr: Box::new(Expr::Map(Map {
+                entries: vec![
+                    MapEntry {
+                        key: Box::new(Expr::Value(Value::SingleQuotedString("a".to_owned()))),
+                        value: Box::new(number_expr("10")),
+                    },
+                    MapEntry {
+                        key: Box::new(Expr::Value(Value::SingleQuotedString("b".to_owned()))),
+                        value: Box::new(number_expr("20")),
+                    },
+                ],
+            })),
+            subscript: Box::new(Subscript::Index {
+                index: Expr::Value(Value::SingleQuotedString("a".to_owned())),
+            }),
+        },
+    );
+}
+
 #[test]
 fn parse_within_group() {
     verified_expr("PERCENTILE_CONT(0.5) WITHIN GROUP (ORDER BY sales_amount)");

tests/sqlparser_custom_dialect.rs

@@ -125,6 +125,28 @@ fn custom_statement_parser() -> Result<(), ParserError> {
     Ok(())
 }
 
+#[test]
+fn test_map_syntax_not_support_default() -> Result<(), ParserError> {
+    #[derive(Debug)]
+    struct MyDialect {}
+
+    impl Dialect for MyDialect {
+        fn is_identifier_start(&self, ch: char) -> bool {
+            is_identifier_start(ch)
+        }
+
+        fn is_identifier_part(&self, ch: char) -> bool {
+            is_identifier_part(ch)
+        }
+    }
+
+    let dialect = MyDialect {};
+    let sql = "SELECT MAP {1: 2}";
+    let ast = Parser::parse_sql(&dialect, sql);
+    assert!(ast.is_err());
+    Ok(())
+}
+
 fn is_identifier_start(ch: char) -> bool {
     ch.is_ascii_lowercase() || ch.is_ascii_uppercase() || ch == '_'
 }