Improve MySQL option parsing in index definitions (apache#1997)

Author: mvzink

src/ast/ddl.rs

@@ -24,9 +24,8 @@ use serde::{Deserialize, Serialize};
 #[cfg(feature = "visitor")]
 use sqlparser_derive::{Visit, VisitMut};
 
-use super::super::dml::CreateTable;
 use crate::ast::{
-    ClusteredBy, ColumnDef, CommentDef, CreateTableOptions, Expr, FileFormat,
+    ClusteredBy, ColumnDef, CommentDef, CreateTable, CreateTableOptions, Expr, FileFormat,
     HiveDistributionStyle, HiveFormat, Ident, ObjectName, OnCommit, OneOrManyWithParens, Query,
     RowAccessPolicy, Statement, StorageSerializationPolicy, TableConstraint, Tag,
     WrappedCollection,

src/ast/dml.rs

@@ -63,14 +63,14 @@ pub use self::ddl::{
     AlterTypeAddValuePosition, AlterTypeOperation, AlterTypeRename, AlterTypeRenameValue,
     ClusteredBy, ColumnDef, ColumnOption, ColumnOptionDef, ColumnOptions, ColumnPolicy,
     ColumnPolicyProperty, ConstraintCharacteristics, CreateConnector, CreateDomain, CreateFunction,
-    Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs, GeneratedExpressionMode,
-    IdentityParameters, IdentityProperty, IdentityPropertyFormatKind, IdentityPropertyKind,
-    IdentityPropertyOrder, IndexOption, IndexType, KeyOrIndexDisplay, NullsDistinctOption, Owner,
-    Partition, ProcedureParam, ReferentialAction, RenameTableNameKind, ReplicaIdentity,
-    TableConstraint, TagsColumnOption, UserDefinedTypeCompositeAttributeDef,
-    UserDefinedTypeRepresentation, ViewColumnDef,
+    CreateIndex, CreateTable, Deduplicate, DeferrableInitial, DropBehavior, GeneratedAs,
+    GeneratedExpressionMode, IdentityParameters, IdentityProperty, IdentityPropertyFormatKind,
+    IdentityPropertyKind, IdentityPropertyOrder, IndexColumn, IndexOption, IndexType,
+    KeyOrIndexDisplay, NullsDistinctOption, Owner, Partition, ProcedureParam, ReferentialAction,
+    RenameTableNameKind, ReplicaIdentity, TableConstraint, TagsColumnOption,
+    UserDefinedTypeCompositeAttributeDef, UserDefinedTypeRepresentation, ViewColumnDef,
 };
-pub use self::dml::{CreateIndex, CreateTable, Delete, IndexColumn, Insert};
+pub use self::dml::{Delete, Insert};
 pub use self::operator::{BinaryOperator, UnaryOperator};
 pub use self::query::{
     AfterMatchSkip, ConnectBy, Cte, CteAsMaterialized, Distinct, EmptyMatchesMode,

src/ast/helpers/stmt_create_table.rs

@@ -713,6 +713,7 @@ impl Spanned for TableConstraint {
                 name,
                 index_type: _,
                 columns,
+                index_options: _,
             } => union_spans(
                 name.iter()
                     .map(|i| i.span)
@@ -747,6 +748,8 @@ impl Spanned for CreateIndex {
             nulls_distinct: _, // bool
             with,
             predicate,
+            index_options: _,
+            alter_options,
         } = self;
 
         union_spans(
@@ -756,7 +759,8 @@ impl Spanned for CreateIndex {
                 .chain(columns.iter().map(|i| i.column.span()))
                 .chain(include.iter().map(|i| i.span))
                 .chain(with.iter().map(|i| i.span()))
-                .chain(predicate.iter().map(|i| i.span())),
+                .chain(predicate.iter().map(|i| i.span()))
+                .chain(alter_options.iter().map(|i| i.span())),
         )
     }
 }

src/ast/mod.rs

@@ -7114,6 +7114,22 @@ impl<'a> Parser<'a> {
             None
         };
 
+        // MySQL options (including the modern style of `USING` after the column list instead of
+        // before, which is deprecated) shouldn't conflict with other preceding options (e.g. `WITH
+        // PARSER` won't be caught by the above `WITH` clause parsing because MySQL doesn't set that
+        // support flag). This is probably invalid syntax for other dialects, but it is simpler to
+        // parse it anyway (as we do inside `ALTER TABLE` and `CREATE TABLE` parsing).
+        let index_options = self.parse_index_options()?;
+
+        // MySQL allows `ALGORITHM` and `LOCK` options. Unlike in `ALTER TABLE`, they need not be comma separated.
+        let mut alter_options = Vec::new();
+        while self
+            .peek_one_of_keywords(&[Keyword::ALGORITHM, Keyword::LOCK])
+            .is_some()
+        {
+            alter_options.push(self.parse_alter_table_operation()?)
+        }
+
         Ok(Statement::CreateIndex(CreateIndex {
             name: index_name,
             table_name,
@@ -7126,6 +7142,8 @@ impl<'a> Parser<'a> {
             nulls_distinct,
             with,
             predicate,
+            index_options,
+            alter_options,
         }))
     }
 
@@ -8438,12 +8456,14 @@ impl<'a> Parser<'a> {
 
                 let index_type = self.parse_optional_using_then_index_type()?;
                 let columns = self.parse_parenthesized_index_column_list()?;
+                let index_options = self.parse_index_options()?;
 
                 Ok(Some(TableConstraint::Index {
                     display_as_key,
                     name,
                     index_type,
                     columns,
+                    index_options,
                 }))
             }
             Token::Word(w)
@@ -17507,6 +17527,7 @@ mod tests {
                 name: None,
                 index_type: None,
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
 
@@ -17518,6 +17539,7 @@ mod tests {
                 name: None,
                 index_type: None,
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
 
@@ -17529,6 +17551,7 @@ mod tests {
                 name: Some(Ident::with_quote('\'', "index")),
                 index_type: None,
                 columns: vec![mk_expected_col("c1"), mk_expected_col("c2")],
+                index_options: vec![],
             }
         );
 
@@ -17540,6 +17563,7 @@ mod tests {
                 name: None,
                 index_type: Some(IndexType::BTree),
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
 
@@ -17551,6 +17575,7 @@ mod tests {
                 name: None,
                 index_type: Some(IndexType::Hash),
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
 
@@ -17562,6 +17587,7 @@ mod tests {
                 name: Some(Ident::new("idx_name")),
                 index_type: Some(IndexType::BTree),
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
 
@@ -17573,6 +17599,7 @@ mod tests {
                 name: Some(Ident::new("idx_name")),
                 index_type: Some(IndexType::Hash),
                 columns: vec![mk_expected_col("c1")],
+                index_options: vec![],
             }
         );
     }

src/ast/spans.rs

@@ -9190,7 +9190,7 @@ fn ensure_multiple_dialects_are_tested() {
 
 #[test]
 fn parse_create_index() {
-    let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test(name,age DESC)";
+    let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test(name, age DESC)";
     let indexed_columns: Vec<IndexColumn> = vec![
         IndexColumn {
             operator_class: None,
@@ -9236,7 +9236,7 @@ fn parse_create_index() {
 
 #[test]
 fn test_create_index_with_using_function() {
-    let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING BTREE (name,age DESC)";
+    let sql = "CREATE UNIQUE INDEX IF NOT EXISTS idx_name ON test USING BTREE (name, age DESC)";
     let indexed_columns: Vec<IndexColumn> = vec![
         IndexColumn {
             operator_class: None,
@@ -9274,6 +9274,8 @@ fn test_create_index_with_using_function() {
             nulls_distinct: None,
             with,
             predicate: None,
+            index_options,
+            alter_options,
         }) => {
             assert_eq!("idx_name", name.to_string());
             assert_eq!("test", table_name.to_string());
@@ -9284,6 +9286,8 @@ fn test_create_index_with_using_function() {
             assert!(if_not_exists);
             assert!(include.is_empty());
             assert!(with.is_empty());
+            assert!(index_options.is_empty());
+            assert!(alter_options.is_empty());
         }
         _ => unreachable!(),
     }
@@ -9325,6 +9329,8 @@ fn test_create_index_with_with_clause() {
             nulls_distinct: None,
             with,
             predicate: None,
+            index_options,
+            alter_options,
         }) => {
             pretty_assertions::assert_eq!("title_idx", name.to_string());
             pretty_assertions::assert_eq!("films", table_name.to_string());
@@ -9334,6 +9340,8 @@ fn test_create_index_with_with_clause() {
             assert!(!if_not_exists);
             assert!(include.is_empty());
             pretty_assertions::assert_eq!(with_parameters, with);
+            assert!(index_options.is_empty());
+            assert!(alter_options.is_empty());
         }
         _ => unreachable!(),
     }

src/parser/mod.rs

@@ -4215,3 +4215,31 @@ fn parse_show_charset() {
     mysql().verified_stmt("SHOW CHARSET WHERE charset = 'utf8mb4%'");
     mysql().verified_stmt("SHOW CHARSET LIKE 'utf8mb4%'");
 }
+
+#[test]
+fn test_ddl_with_index_using() {
+    let columns = "(name, age DESC)";
+    let using = "USING BTREE";
+
+    for sql in [
+        format!("CREATE INDEX idx_name ON test {using} {columns}"),
+        format!("CREATE TABLE foo (name VARCHAR(255), age INT, KEY idx_name {using} {columns})"),
+        format!("ALTER TABLE foo ADD KEY idx_name {using} {columns}"),
+        format!("CREATE INDEX idx_name ON test{columns} {using}"),
+        format!("CREATE TABLE foo (name VARCHAR(255), age INT, KEY idx_name {columns} {using})"),
+        format!("ALTER TABLE foo ADD KEY idx_name {columns} {using}"),
+    ] {
+        mysql_and_generic().verified_stmt(&sql);
+    }
+}
+
+#[test]
+fn test_create_index_options() {
+    mysql_and_generic()
+        .verified_stmt("CREATE INDEX idx_name ON t(c1, c2) USING HASH LOCK = SHARED");
+    mysql_and_generic()
+        .verified_stmt("CREATE INDEX idx_name ON t(c1, c2) USING BTREE ALGORITHM = INPLACE");
+    mysql_and_generic().verified_stmt(
+        "CREATE INDEX idx_name ON t(c1, c2) USING BTREE LOCK = EXCLUSIVE ALGORITHM = DEFAULT",
+    );
+}