Merge remote-tracking branch 'apache/main' into prepare_0.56.0

alamb · alamb · commit 6c49b829196f · 2025-04-29T14:59:39.000-04:00
diff --git a/dev/release/generate-changelog.py b/dev/release/generate-changelog.py
@@ -28,7 +28,8 @@ def print_pulls(repo_name, title, pulls):
         print()
         for (pull, commit) in pulls:
             url = "https://github.com/{}/pull/{}".format(repo_name, pull.number)
-            print("- {} [#{}]({}) ({})".format(pull.title, pull.number, url, commit.author.login))
+            author = f"({commit.author.login})" if commit.author else ''
+            print("- {} [#{}]({}) {}".format(pull.title, pull.number, url, author))
         print()
 
 
@@ -161,4 +162,4 @@ def cli(args=None):
     generate_changelog(repo, project, args.tag1, args.tag2, args.version)
 
 if __name__ == "__main__":
-    cli()
+    cli()
diff --git a/src/ast/mod.rs b/src/ast/mod.rs
@@ -930,12 +930,14 @@ pub enum Expr {
     Nested(Box<Expr>),
     /// A literal value, such as string, number, date or NULL
     Value(ValueWithSpan),
+    /// Prefixed expression, e.g. introducer strings, projection prefix
     /// <https://dev.mysql.com/doc/refman/8.0/en/charset-introducer.html>
-    IntroducedString {
-        introducer: String,
+    /// <https://docs.snowflake.com/en/sql-reference/constructs/connect-by>
+    Prefixed {
+        prefix: Ident,
         /// The value of the constant.
         /// Hint: you can unwrap the string value using `value.into_string()`.
-        value: Value,
+        value: Box<Expr>,
     },
     /// A constant of form `<data_type> 'value'`.
     /// This can represent ANSI SQL `DATE`, `TIME`, and `TIMESTAMP` literals (such as `DATE '2020-01-01'`),
@@ -1655,7 +1657,7 @@ impl fmt::Display for Expr {
             Expr::Collate { expr, collation } => write!(f, "{expr} COLLATE {collation}"),
             Expr::Nested(ast) => write!(f, "({ast})"),
             Expr::Value(v) => write!(f, "{v}"),
-            Expr::IntroducedString { introducer, value } => write!(f, "{introducer} {value}"),
+            Expr::Prefixed { prefix, value } => write!(f, "{prefix} {value}"),
             Expr::TypedString { data_type, value } => {
                 write!(f, "{data_type}")?;
                 write!(f, " {value}")
@@ -2472,10 +2474,11 @@ impl fmt::Display for DeclareAssignment {
 #[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
 #[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
 pub enum DeclareType {
-    /// Cursor variable type. e.g. [Snowflake] [PostgreSQL]
+    /// Cursor variable type. e.g. [Snowflake] [PostgreSQL] [MsSql]
     ///
     /// [Snowflake]: https://docs.snowflake.com/en/developer-guide/snowflake-scripting/cursors#declaring-a-cursor
     /// [PostgreSQL]: https://www.postgresql.org/docs/current/plpgsql-cursors.html
+    /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-cursor-transact-sql
     Cursor,
 
     /// Result set variable type. [Snowflake]
diff --git a/src/ast/spans.rs b/src/ast/spans.rs
@@ -1543,7 +1543,7 @@ impl Spanned for Expr {
                             .map(|items| union_spans(items.iter().map(|i| i.span()))),
                     ),
             ),
-            Expr::IntroducedString { value, .. } => value.span(),
+            Expr::Prefixed { value, .. } => value.span(),
             Expr::Case {
                 operand,
                 conditions,
diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs
@@ -888,6 +888,12 @@ pub trait Dialect: Debug + Any {
         keywords::RESERVED_FOR_TABLE_FACTOR
     }
 
+    /// Returns reserved keywords that may prefix a select item expression
+    /// e.g. `SELECT CONNECT_BY_ROOT name FROM Tbl2` (Snowflake)
+    fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] {
+        &[]
+    }
+
     /// Returns true if this dialect supports the `TABLESAMPLE` option
     /// before the table alias option. For example:
     ///
diff --git a/src/dialect/snowflake.rs b/src/dialect/snowflake.rs
@@ -44,6 +44,7 @@ use alloc::{format, vec};
 use super::keywords::RESERVED_FOR_IDENTIFIER;
 use sqlparser::ast::StorageSerializationPolicy;
 
+const RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR: [Keyword; 1] = [Keyword::CONNECT_BY_ROOT];
 /// A [`Dialect`] for [Snowflake](https://www.snowflake.com/)
 #[derive(Debug, Default)]
 pub struct SnowflakeDialect;
@@ -346,6 +347,11 @@ impl Dialect for SnowflakeDialect {
     fn supports_group_by_expr(&self) -> bool {
         true
     }
+
+    /// See: <https://docs.snowflake.com/en/sql-reference/constructs/connect-by>
+    fn get_reserved_keywords_for_select_item_operator(&self) -> &[Keyword] {
+        &RESERVED_KEYWORDS_FOR_SELECT_ITEM_OPERATOR
+    }
 }
 
 fn parse_file_staging_command(kw: Keyword, parser: &mut Parser) -> Result<Statement, ParserError> {
diff --git a/src/keywords.rs b/src/keywords.rs
@@ -207,6 +207,7 @@ define_keywords!(
     CONNECT,
     CONNECTION,
     CONNECTOR,
+    CONNECT_BY_ROOT,
     CONSTRAINT,
     CONTAINS,
     CONTINUE,
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
@@ -1388,9 +1388,9 @@ impl<'a> Parser<'a> {
             | Token::HexStringLiteral(_)
                 if w.value.starts_with('_') =>
             {
-                Ok(Expr::IntroducedString {
-                    introducer: w.value.clone(),
-                    value: self.parse_introduced_string_value()?,
+                Ok(Expr::Prefixed {
+                    prefix: w.clone().into_ident(w_span),
+                    value: self.parse_introduced_string_expr()?.into(),
                 })
             }
             // string introducer https://dev.mysql.com/doc/refman/8.0/en/charset-introducer.html
@@ -1399,9 +1399,9 @@ impl<'a> Parser<'a> {
             | Token::HexStringLiteral(_)
                 if w.value.starts_with('_') =>
             {
-                Ok(Expr::IntroducedString {
-                    introducer: w.value.clone(),
-                    value: self.parse_introduced_string_value()?,
+                Ok(Expr::Prefixed {
+                    prefix: w.clone().into_ident(w_span),
+                    value: self.parse_introduced_string_expr()?.into(),
                 })
             }
             Token::Arrow if self.dialect.supports_lambda_functions() => {
@@ -6446,7 +6446,7 @@ impl<'a> Parser<'a> {
     /// DECLARE
     // {
     //   { @local_variable [AS] data_type [ = value ] }
-    //   | { @cursor_variable_name CURSOR }
+    //   | { @cursor_variable_name CURSOR [ FOR ] }
     // } [ ,...n ]
     /// ```
     /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-local-variable-transact-sql?view=sql-server-ver16
@@ -6462,14 +6462,19 @@ impl<'a> Parser<'a> {
     /// ```text
     // {
     //   { @local_variable [AS] data_type [ = value ] }
-    //   | { @cursor_variable_name CURSOR }
+    //   | { @cursor_variable_name CURSOR [ FOR ]}
     // } [ ,...n ]
     /// ```
     /// [MsSql]: https://learn.microsoft.com/en-us/sql/t-sql/language-elements/declare-local-variable-transact-sql?view=sql-server-ver16
     pub fn parse_mssql_declare_stmt(&mut self) -> Result<Declare, ParserError> {
         let name = {
             let ident = self.parse_identifier()?;
-            if !ident.value.starts_with('@') {
+            if !ident.value.starts_with('@')
+                && !matches!(
+                    self.peek_token().token,
+                    Token::Word(w) if w.keyword == Keyword::CURSOR
+                )
+            {
                 Err(ParserError::TokenizerError(
                     "Invalid MsSql variable declaration.".to_string(),
                 ))
@@ -6493,7 +6498,14 @@ impl<'a> Parser<'a> {
             _ => (None, Some(self.parse_data_type()?)),
         };
 
-        let assignment = self.parse_mssql_variable_declaration_expression()?;
+        let (for_query, assignment) = if self.peek_keyword(Keyword::FOR) {
+            self.next_token();
+            let query = Some(self.parse_query()?);
+            (query, None)
+        } else {
+            let assignment = self.parse_mssql_variable_declaration_expression()?;
+            (None, assignment)
+        };
 
         Ok(Declare {
             names: vec![name],
@@ -6504,7 +6516,7 @@ impl<'a> Parser<'a> {
             sensitive: None,
             scroll: None,
             hold: None,
-            for_query: None,
+            for_query,
         })
     }
 
@@ -9023,13 +9035,19 @@ impl<'a> Parser<'a> {
         }
     }
 
-    fn parse_introduced_string_value(&mut self) -> Result<Value, ParserError> {
+    fn parse_introduced_string_expr(&mut self) -> Result<Expr, ParserError> {
         let next_token = self.next_token();
         let span = next_token.span;
         match next_token.token {
-            Token::SingleQuotedString(ref s) => Ok(Value::SingleQuotedString(s.to_string())),
-            Token::DoubleQuotedString(ref s) => Ok(Value::DoubleQuotedString(s.to_string())),
-            Token::HexStringLiteral(ref s) => Ok(Value::HexStringLiteral(s.to_string())),
+            Token::SingleQuotedString(ref s) => Ok(Expr::Value(
+                Value::SingleQuotedString(s.to_string()).with_span(span),
+            )),
+            Token::DoubleQuotedString(ref s) => Ok(Expr::Value(
+                Value::DoubleQuotedString(s.to_string()).with_span(span),
+            )),
+            Token::HexStringLiteral(ref s) => Ok(Expr::Value(
+                Value::HexStringLiteral(s.to_string()).with_span(span),
+            )),
             unexpected => self.expected(
                 "a string value",
                 TokenWithSpan {
@@ -13956,6 +13974,13 @@ impl<'a> Parser<'a> {
 
     /// Parse a comma-delimited list of projections after SELECT
     pub fn parse_select_item(&mut self) -> Result<SelectItem, ParserError> {
+        let prefix = self
+            .parse_one_of_keywords(
+                self.dialect
+                    .get_reserved_keywords_for_select_item_operator(),
+            )
+            .map(|keyword| Ident::new(format!("{:?}", keyword)));
+
         match self.parse_wildcard_expr()? {
             Expr::QualifiedWildcard(prefix, token) => Ok(SelectItem::QualifiedWildcard(
                 SelectItemQualifiedWildcardKind::ObjectName(prefix),
@@ -14000,8 +14025,11 @@ impl<'a> Parser<'a> {
             expr => self
                 .maybe_parse_select_item_alias()
                 .map(|alias| match alias {
-                    Some(alias) => SelectItem::ExprWithAlias { expr, alias },
-                    None => SelectItem::UnnamedExpr(expr),
+                    Some(alias) => SelectItem::ExprWithAlias {
+                        expr: maybe_prefixed_expr(expr, prefix),
+                        alias,
+                    },
+                    None => SelectItem::UnnamedExpr(maybe_prefixed_expr(expr, prefix)),
                 }),
         }
     }
@@ -15363,6 +15391,17 @@ impl<'a> Parser<'a> {
     }
 }
 
+fn maybe_prefixed_expr(expr: Expr, prefix: Option<Ident>) -> Expr {
+    if let Some(prefix) = prefix {
+        Expr::Prefixed {
+            prefix,
+            value: Box::new(expr),
+        }
+    } else {
+        expr
+    }
+}
+
 impl Word {
     #[deprecated(since = "0.54.0", note = "please use `into_ident` instead")]
     pub fn to_ident(&self, span: Span) -> Ident {
diff --git a/tests/sqlparser_mssql.rs b/tests/sqlparser_mssql.rs
@@ -1387,6 +1387,10 @@ fn parse_mssql_declare() {
         ],
         ast
     );
+
+    let declare_cursor_for_select =
+        "DECLARE vend_cursor CURSOR FOR SELECT * FROM Purchasing.Vendor";
+    let _ = ms().verified_stmt(declare_cursor_for_select);
 }
 
 #[test]
diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs
@@ -3020,9 +3020,12 @@ fn parse_hex_string_introducer() {
                 distinct: None,
                 top: None,
                 top_before_distinct: false,
-                projection: vec![SelectItem::UnnamedExpr(Expr::IntroducedString {
-                    introducer: "_latin1".to_string(),
-                    value: Value::HexStringLiteral("4D7953514C".to_string())
+                projection: vec![SelectItem::UnnamedExpr(Expr::Prefixed {
+                    prefix: Ident::from("_latin1"),
+                    value: Expr::Value(
+                        Value::HexStringLiteral("4D7953514C".to_string()).with_empty_span()
+                    )
+                    .into(),
                 })],
                 from: vec![],
                 lateral_views: vec![],
diff --git a/tests/sqlparser_snowflake.rs b/tests/sqlparser_snowflake.rs
@@ -3983,3 +3983,45 @@ fn test_nested_join_without_parentheses() {
         }],
     );
 }
+
+#[test]
+fn parse_connect_by_root_operator() {
+    let sql = "SELECT CONNECT_BY_ROOT name AS root_name FROM Tbl1";
+
+    match snowflake().verified_stmt(sql) {
+        Statement::Query(query) => {
+            assert_eq!(
+                query.body.as_select().unwrap().projection[0],
+                SelectItem::ExprWithAlias {
+                    expr: Expr::Prefixed {
+                        prefix: Ident::new("CONNECT_BY_ROOT"),
+                        value: Box::new(Expr::Identifier(Ident::new("name")))
+                    },
+                    alias: Ident::new("root_name"),
+                }
+            );
+        }
+        _ => unreachable!(),
+    }
+
+    let sql = "SELECT CONNECT_BY_ROOT name FROM Tbl2";
+    match snowflake().verified_stmt(sql) {
+        Statement::Query(query) => {
+            assert_eq!(
+                query.body.as_select().unwrap().projection[0],
+                SelectItem::UnnamedExpr(Expr::Prefixed {
+                    prefix: Ident::new("CONNECT_BY_ROOT"),
+                    value: Box::new(Expr::Identifier(Ident::new("name")))
+                })
+            );
+        }
+        _ => unreachable!(),
+    }
+
+    let sql = "SELECT CONNECT_BY_ROOT FROM Tbl2";
+    let res = snowflake().parse_sql_statements(sql);
+    assert_eq!(
+        res.unwrap_err().to_string(),
+        "sql parser error: Expected an expression, found: FROM"
+    );
+}
