Add support for MSSQL's JSON_ARRAY/JSON_OBJECT expr (apache#1507)

gaoqiangz · iffyio · web-flow · commit 92be237cfc08 · 2024-11-18T09:22:18.000-05:00
Co-authored-by: Ifeanyi Ubah &lt;ify1992@yahoo.com&gt;
diff --git a/src/ast/mod.rs b/src/ast/mod.rs
@@ -5449,6 +5449,8 @@ pub enum FunctionArgOperator {
     RightArrow,
     /// function(arg1 := value1)
     Assignment,
+    /// function(arg1 : value1)
+    Colon,
 }
 
 impl fmt::Display for FunctionArgOperator {
@@ -5457,6 +5459,7 @@ impl fmt::Display for FunctionArgOperator {
             FunctionArgOperator::Equals => f.write_str("="),
             FunctionArgOperator::RightArrow => f.write_str("=>"),
             FunctionArgOperator::Assignment => f.write_str(":="),
+            FunctionArgOperator::Colon => f.write_str(":"),
         }
     }
 }
@@ -5465,11 +5468,22 @@ impl fmt::Display for FunctionArgOperator {
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub enum FunctionArg {
+    /// `name` is identifier
+    ///
+    /// Enabled when `Dialect::supports_named_fn_args_with_expr_name` returns 'false'
     Named {
         name: Ident,
         arg: FunctionArgExpr,
         operator: FunctionArgOperator,
     },
+    /// `name` is arbitrary expression
+    ///
+    /// Enabled when `Dialect::supports_named_fn_args_with_expr_name` returns 'true'
+    ExprNamed {
+        name: Expr,
+        arg: FunctionArgExpr,
+        operator: FunctionArgOperator,
+    },
     Unnamed(FunctionArgExpr),
 }
 
@@ -5481,6 +5495,11 @@ impl fmt::Display for FunctionArg {
                 arg,
                 operator,
             } => write!(f, "{name} {operator} {arg}"),
+            FunctionArg::ExprNamed {
+                name,
+                arg,
+                operator,
+            } => write!(f, "{name} {operator} {arg}"),
             FunctionArg::Unnamed(unnamed_arg) => write!(f, "{unnamed_arg}"),
         }
     }
@@ -5619,7 +5638,10 @@ impl fmt::Display for FunctionArgumentList {
         }
         write!(f, "{}", display_comma_separated(&self.args))?;
         if !self.clauses.is_empty() {
-            write!(f, " {}", display_separated(&self.clauses, " "))?;
+            if !self.args.is_empty() {
+                write!(f, " ")?;
+            }
+            write!(f, "{}", display_separated(&self.clauses, " "))?;
         }
         Ok(())
     }
@@ -5661,6 +5683,11 @@ pub enum FunctionArgumentClause {
     ///
     /// [`GROUP_CONCAT`]: https://dev.mysql.com/doc/refman/8.0/en/aggregate-functions.html#function_group-concat
     Separator(Value),
+    /// The json-null-clause to the [`JSON_ARRAY`]/[`JSON_OBJECT`] function in MSSQL.
+    ///
+    /// [`JSON_ARRAY`]: <https://learn.microsoft.com/en-us/sql/t-sql/functions/json-array-transact-sql?view=sql-server-ver16>
+    /// [`JSON_OBJECT`]: <https://learn.microsoft.com/en-us/sql/t-sql/functions/json-object-transact-sql?view=sql-server-ver16>
+    JsonNullClause(JsonNullClause),
 }
 
 impl fmt::Display for FunctionArgumentClause {
@@ -5676,6 +5703,7 @@ impl fmt::Display for FunctionArgumentClause {
             FunctionArgumentClause::OnOverflow(on_overflow) => write!(f, "{on_overflow}"),
             FunctionArgumentClause::Having(bound) => write!(f, "{bound}"),
             FunctionArgumentClause::Separator(sep) => write!(f, "SEPARATOR {sep}"),
+            FunctionArgumentClause::JsonNullClause(null_clause) => write!(f, "{null_clause}"),
         }
     }
 }
@@ -7564,6 +7592,32 @@ impl fmt::Display for ShowStatementIn {
     }
 }
 
+/// MSSQL's json null clause
+///
+/// ```plaintext
+/// <json_null_clause> ::=
+///       NULL ON NULL
+///     | ABSENT ON NULL
+/// ```
+///
+/// <https://learn.microsoft.com/en-us/sql/t-sql/functions/json-object-transact-sql?view=sql-server-ver16#json_null_clause>
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub enum JsonNullClause {
+    NullOnNull,
+    AbsentOnNull,
+}
+
+impl Display for JsonNullClause {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        match self {
+            JsonNullClause::NullOnNull => write!(f, "NULL ON NULL"),
+            JsonNullClause::AbsentOnNull => write!(f, "ABSENT ON NULL"),
+        }
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
diff --git a/src/dialect/duckdb.rs b/src/dialect/duckdb.rs
@@ -47,6 +47,10 @@ impl Dialect for DuckDbDialect {
         true
     }
 
+    fn supports_named_fn_args_with_assignment_operator(&self) -> bool {
+        true
+    }
+
     // DuckDB uses this syntax for `STRUCT`s.
     //
     // https://duckdb.org/docs/sql/data_types/struct.html#creating-structs
diff --git a/src/dialect/generic.rs b/src/dialect/generic.rs
@@ -119,4 +119,8 @@ impl Dialect for GenericDialect {
     fn supports_load_extension(&self) -> bool {
         true
     }
+
+    fn supports_named_fn_args_with_assignment_operator(&self) -> bool {
+        true
+    }
 }
diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs
@@ -231,11 +231,34 @@ pub trait Dialect: Debug + Any {
         false
     }
 
-    /// Returns true if the dialect supports named arguments of the form FUN(a = '1', b = '2').
+    /// Returns true if the dialect supports named arguments of the form `FUN(a = '1', b = '2')`.
     fn supports_named_fn_args_with_eq_operator(&self) -> bool {
         false
     }
 
+    /// Returns true if the dialect supports named arguments of the form `FUN(a : '1', b : '2')`.
+    fn supports_named_fn_args_with_colon_operator(&self) -> bool {
+        false
+    }
+
+    /// Returns true if the dialect supports named arguments of the form `FUN(a := '1', b := '2')`.
+    fn supports_named_fn_args_with_assignment_operator(&self) -> bool {
+        false
+    }
+
+    /// Returns true if the dialect supports named arguments of the form `FUN(a => '1', b => '2')`.
+    fn supports_named_fn_args_with_rarrow_operator(&self) -> bool {
+        true
+    }
+
+    /// Returns true if dialect supports argument name as arbitrary expression.
+    /// e.g. `FUN(LOWER('a'):'1',  b:'2')`
+    /// Such function arguments are represented in the AST by the `FunctionArg::ExprNamed` variant,
+    /// otherwise use the `FunctionArg::Named` variant (compatible reason).
+    fn supports_named_fn_args_with_expr_name(&self) -> bool {
+        false
+    }
+
     /// Returns true if the dialect supports identifiers starting with a numeric
     /// prefix such as tables named `59901_user_login`
     fn supports_numeric_prefix(&self) -> bool {
diff --git a/src/dialect/mssql.rs b/src/dialect/mssql.rs
@@ -66,4 +66,16 @@ impl Dialect for MsSqlDialect {
     fn supports_methods(&self) -> bool {
         true
     }
+
+    fn supports_named_fn_args_with_colon_operator(&self) -> bool {
+        true
+    }
+
+    fn supports_named_fn_args_with_expr_name(&self) -> bool {
+        true
+    }
+
+    fn supports_named_fn_args_with_rarrow_operator(&self) -> bool {
+        false
+    }
 }
diff --git a/src/keywords.rs b/src/keywords.rs
@@ -74,6 +74,7 @@ macro_rules! define_keywords {
 define_keywords!(
     ABORT,
     ABS,
+    ABSENT,
     ABSOLUTE,
     ACCESS,
     ACCOUNT,
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
@@ -11321,45 +11321,58 @@ impl<'a> Parser<'a> {
     }
 
     pub fn parse_function_args(&mut self) -> Result<FunctionArg, ParserError> {
-        if self.peek_nth_token(1) == Token::RArrow {
-            let name = self.parse_identifier(false)?;
-
-            self.expect_token(&Token::RArrow)?;
-            let arg = self.parse_wildcard_expr()?.into();
-
-            Ok(FunctionArg::Named {
-                name,
-                arg,
-                operator: FunctionArgOperator::RightArrow,
-            })
-        } else if self.dialect.supports_named_fn_args_with_eq_operator()
-            && self.peek_nth_token(1) == Token::Eq
-        {
-            let name = self.parse_identifier(false)?;
-
-            self.expect_token(&Token::Eq)?;
-            let arg = self.parse_wildcard_expr()?.into();
-
-            Ok(FunctionArg::Named {
-                name,
-                arg,
-                operator: FunctionArgOperator::Equals,
-            })
-        } else if dialect_of!(self is DuckDbDialect | GenericDialect)
-            && self.peek_nth_token(1) == Token::Assignment
-        {
-            let name = self.parse_identifier(false)?;
-
-            self.expect_token(&Token::Assignment)?;
-            let arg = self.parse_expr()?.into();
-
-            Ok(FunctionArg::Named {
-                name,
-                arg,
-                operator: FunctionArgOperator::Assignment,
-            })
+        let arg = if self.dialect.supports_named_fn_args_with_expr_name() {
+            self.maybe_parse(|p| {
+                let name = p.parse_expr()?;
+                let operator = p.parse_function_named_arg_operator()?;
+                let arg = p.parse_wildcard_expr()?.into();
+                Ok(FunctionArg::ExprNamed {
+                    name,
+                    arg,
+                    operator,
+                })
+            })?
         } else {
-            Ok(FunctionArg::Unnamed(self.parse_wildcard_expr()?.into()))
+            self.maybe_parse(|p| {
+                let name = p.parse_identifier(false)?;
+                let operator = p.parse_function_named_arg_operator()?;
+                let arg = p.parse_wildcard_expr()?.into();
+                Ok(FunctionArg::Named {
+                    name,
+                    arg,
+                    operator,
+                })
+            })?
+        };
+        if let Some(arg) = arg {
+            return Ok(arg);
+        }
+        Ok(FunctionArg::Unnamed(self.parse_wildcard_expr()?.into()))
+    }
+
+    fn parse_function_named_arg_operator(&mut self) -> Result<FunctionArgOperator, ParserError> {
+        let tok = self.next_token();
+        match tok.token {
+            Token::RArrow if self.dialect.supports_named_fn_args_with_rarrow_operator() => {
+                Ok(FunctionArgOperator::RightArrow)
+            }
+            Token::Eq if self.dialect.supports_named_fn_args_with_eq_operator() => {
+                Ok(FunctionArgOperator::Equals)
+            }
+            Token::Assignment
+                if self
+                    .dialect
+                    .supports_named_fn_args_with_assignment_operator() =>
+            {
+                Ok(FunctionArgOperator::Assignment)
+            }
+            Token::Colon if self.dialect.supports_named_fn_args_with_colon_operator() => {
+                Ok(FunctionArgOperator::Colon)
+            }
+            _ => {
+                self.prev_token();
+                self.expected("argument operator", tok)
+            }
         }
     }
 
@@ -11403,19 +11416,24 @@ impl<'a> Parser<'a> {
     /// FIRST_VALUE(x IGNORE NULL);
     /// ```
     fn parse_function_argument_list(&mut self) -> Result<FunctionArgumentList, ParserError> {
+        let mut clauses = vec![];
+
+        // For MSSQL empty argument list with json-null-clause case, e.g. `JSON_ARRAY(NULL ON NULL)`
+        if let Some(null_clause) = self.parse_json_null_clause() {
+            clauses.push(FunctionArgumentClause::JsonNullClause(null_clause));
+        }
+
         if self.consume_token(&Token::RParen) {
             return Ok(FunctionArgumentList {
                 duplicate_treatment: None,
                 args: vec![],
-                clauses: vec![],
+                clauses,
             });
         }
 
         let duplicate_treatment = self.parse_duplicate_treatment()?;
         let args = self.parse_comma_separated(Parser::parse_function_args)?;
 
-        let mut clauses = vec![];
-
         if self.dialect.supports_window_function_null_treatment_arg() {
             if let Some(null_treatment) = self.parse_null_treatment()? {
                 clauses.push(FunctionArgumentClause::IgnoreOrRespectNulls(null_treatment));
@@ -11456,6 +11474,10 @@ impl<'a> Parser<'a> {
             clauses.push(FunctionArgumentClause::OnOverflow(on_overflow));
         }
 
+        if let Some(null_clause) = self.parse_json_null_clause() {
+            clauses.push(FunctionArgumentClause::JsonNullClause(null_clause));
+        }
+
         self.expect_token(&Token::RParen)?;
         Ok(FunctionArgumentList {
             duplicate_treatment,
@@ -11464,6 +11486,17 @@ impl<'a> Parser<'a> {
         })
     }
 
+    /// Parses MSSQL's json-null-clause
+    fn parse_json_null_clause(&mut self) -> Option<JsonNullClause> {
+        if self.parse_keywords(&[Keyword::ABSENT, Keyword::ON, Keyword::NULL]) {
+            Some(JsonNullClause::AbsentOnNull)
+        } else if self.parse_keywords(&[Keyword::NULL, Keyword::ON, Keyword::NULL]) {
+            Some(JsonNullClause::NullOnNull)
+        } else {
+            None
+        }
+    }
+
     fn parse_duplicate_treatment(&mut self) -> Result<Option<DuplicateTreatment>, ParserError> {
         let loc = self.peek_token().location;
         match (
diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs
@@ -4402,8 +4402,9 @@ fn parse_explain_query_plan() {
 
 #[test]
 fn parse_named_argument_function() {
+    let dialects = all_dialects_where(|d| d.supports_named_fn_args_with_rarrow_operator());
     let sql = "SELECT FUN(a => '1', b => '2') FROM foo";
-    let select = verified_only_select(sql);
+    let select = dialects.verified_only_select(sql);
 
     assert_eq!(
         &Expr::Function(Function {
diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs

Original file line number	Diff line number	Diff line change
`@@ -119,4 +119,8 @@ impl Dialect for GenericDialect {`
`119`	`119`	`fn supports_load_extension(&self) -> bool {`
`120`	`120`	`true`
`121`	`121`	`}`
	`122`	`+`
	`123`	`+ fn supports_named_fn_args_with_assignment_operator(&self) -> bool {`
	`124`	`+ true`
	`125`	`+ }`
`122`	`126`	`}`