feat: support INSERT INTO [TABLE] FUNCTION of Clickhouse (apache#1633)

byte-sourcerer · iffyio · ayman-sigma · commit bbece9033b1c · 2025-04-10T14:47:53.000-07:00
Co-authored-by: Kermit &lt;chenjiawei1@xiaohongshu.com&gt;
Co-authored-by: Ifeanyi Ubah &lt;ify1992@yahoo.com&gt;
diff --git a/src/ast/dml.rs b/src/ast/dml.rs
@@ -36,7 +36,7 @@ use super::{
     FileFormat, FromTable, HiveDistributionStyle, HiveFormat, HiveIOFormat, HiveRowFormat, Ident,
     InsertAliases, MysqlInsertPriority, ObjectName, OnCommit, OnInsert, OneOrManyWithParens,
     OrderByExpr, Query, RowAccessPolicy, SelectItem, SqlOption, SqliteOnConflict, TableEngine,
-    TableWithJoins, Tag, WrappedCollection,
+    TableObject, TableWithJoins, Tag, WrappedCollection,
 };
 
 /// CREATE INDEX statement.
@@ -470,8 +470,7 @@ pub struct Insert {
     /// INTO - optional keyword
     pub into: bool,
     /// TABLE
-    #[cfg_attr(feature = "visitor", visit(with = "visit_relation"))]
-    pub table_name: ObjectName,
+    pub table: TableObject,
     /// table_name as foo (for PostgreSQL)
     pub table_alias: Option<Ident>,
     /// COLUMNS
@@ -488,7 +487,7 @@ pub struct Insert {
     /// Columns defined after PARTITION
     pub after_columns: Vec<Ident>,
     /// whether the insert has the table keyword (Hive)
-    pub table: bool,
+    pub has_table_keyword: bool,
     pub on: Option<OnInsert>,
     /// RETURNING
     pub returning: Option<Vec<SelectItem>>,
@@ -503,9 +502,9 @@ pub struct Insert {
 impl Display for Insert {
     fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         let table_name = if let Some(alias) = &self.table_alias {
-            format!("{0} AS {alias}", self.table_name)
+            format!("{0} AS {alias}", self.table)
         } else {
-            self.table_name.to_string()
+            self.table.to_string()
         };
 
         if let Some(on_conflict) = self.or {
@@ -531,7 +530,7 @@ impl Display for Insert {
                 ignore = if self.ignore { " IGNORE" } else { "" },
                 over = if self.overwrite { " OVERWRITE" } else { "" },
                 int = if self.into { " INTO" } else { "" },
-                tbl = if self.table { " TABLE" } else { "" },
+                tbl = if self.has_table_keyword { " TABLE" } else { "" },
             )?;
         }
         if !self.columns.is_empty() {
diff --git a/src/ast/mod.rs b/src/ast/mod.rs
@@ -7925,6 +7925,36 @@ impl fmt::Display for RenameTable {
     }
 }
 
+/// Represents the referenced table in an `INSERT INTO` statement
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub enum TableObject {
+    /// Table specified by name.
+    /// Example:
+    /// ```sql
+    /// INSERT INTO my_table
+    /// ```
+    TableName(#[cfg_attr(feature = "visitor", visit(with = "visit_relation"))] ObjectName),
+
+    /// Table specified as a function.
+    /// Example:
+    /// ```sql
+    /// INSERT INTO TABLE FUNCTION remote('localhost', default.simple_table)
+    /// ```
+    /// [Clickhouse](https://clickhouse.com/docs/en/sql-reference/table-functions)
+    TableFunction(Function),
+}
+
+impl fmt::Display for TableObject {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        match self {
+            Self::TableName(table_name) => write!(f, "{table_name}"),
+            Self::TableFunction(func) => write!(f, "FUNCTION {}", func),
+        }
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
diff --git a/src/ast/spans.rs b/src/ast/spans.rs
@@ -32,8 +32,9 @@ use super::{
     OrderBy, OrderByExpr, Partition, PivotValueSource, ProjectionSelect, Query, ReferentialAction,
     RenameSelectItem, ReplaceSelectElement, ReplaceSelectItem, Select, SelectInto, SelectItem,
     SetExpr, SqlOption, Statement, Subscript, SymbolDefinition, TableAlias, TableAliasColumnDef,
-    TableConstraint, TableFactor, TableOptionsClustered, TableWithJoins, UpdateTableFromKind, Use,
-    Value, Values, ViewColumnDef, WildcardAdditionalOptions, With, WithFill,
+    TableConstraint, TableFactor, TableObject, TableOptionsClustered, TableWithJoins,
+    UpdateTableFromKind, Use, Value, Values, ViewColumnDef, WildcardAdditionalOptions, With,
+    WithFill,
 };
 
 /// Given an iterator of spans, return the [Span::union] of all spans.
@@ -1141,14 +1142,14 @@ impl Spanned for Insert {
             or: _,     // enum, sqlite specific
             ignore: _, // bool
             into: _,   // bool
-            table_name,
+            table,
             table_alias,
             columns,
             overwrite: _, // bool
             source,
             partitioned,
             after_columns,
-            table: _, // bool
+            has_table_keyword: _, // bool
             on,
             returning,
             replace_into: _, // bool
@@ -1158,7 +1159,7 @@ impl Spanned for Insert {
         } = self;
 
         union_spans(
-            core::iter::once(table_name.span())
+            core::iter::once(table.span())
                 .chain(table_alias.as_ref().map(|i| i.span))
                 .chain(columns.iter().map(|i| i.span))
                 .chain(source.as_ref().map(|q| q.span()))
@@ -2126,6 +2127,17 @@ impl Spanned for UpdateTableFromKind {
     }
 }
 
+impl Spanned for TableObject {
+    fn span(&self) -> Span {
+        match self {
+            TableObject::TableName(ObjectName(segments)) => {
+                union_spans(segments.iter().map(|i| i.span))
+            }
+            TableObject::TableFunction(func) => func.span(),
+        }
+    }
+}
+
 #[cfg(test)]
 pub mod tests {
     use crate::dialect::{Dialect, GenericDialect, SnowflakeDialect};
diff --git a/src/dialect/clickhouse.rs b/src/dialect/clickhouse.rs
@@ -50,4 +50,8 @@ impl Dialect for ClickHouseDialect {
     fn supports_limit_comma(&self) -> bool {
         true
     }
+
+    fn supports_insert_table_function(&self) -> bool {
+        true
+    }
 }
diff --git a/src/dialect/mod.rs b/src/dialect/mod.rs
@@ -792,6 +792,11 @@ pub trait Dialect: Debug + Any {
     fn supports_insert_set(&self) -> bool {
         false
     }
+
+    /// Does the dialect support table function in insertion?
+    fn supports_insert_table_function(&self) -> bool {
+        false
+    }
 }
 
 /// This represents the operators for which precedence must be defined
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
@@ -8947,6 +8947,18 @@ impl<'a> Parser<'a> {
         }
     }
 
+    /// Parse a table object for insetion
+    /// e.g. `some_database.some_table` or `FUNCTION some_table_func(...)`
+    pub fn parse_table_object(&mut self) -> Result<TableObject, ParserError> {
+        if self.dialect.supports_insert_table_function() && self.parse_keyword(Keyword::FUNCTION) {
+            let fn_name = self.parse_object_name(false)?;
+            self.parse_function_call(fn_name)
+                .map(TableObject::TableFunction)
+        } else {
+            self.parse_object_name(false).map(TableObject::TableName)
+        }
+    }
+
     /// Parse a possibly qualified, possibly quoted identifier, optionally allowing for wildcards,
     /// e.g. *, *.*, `foo`.*, or "foo"."bar"
     fn parse_object_name_with_wildcards(
@@ -12020,7 +12032,7 @@ impl<'a> Parser<'a> {
         } else {
             // Hive lets you put table here regardless
             let table = self.parse_keyword(Keyword::TABLE);
-            let table_name = self.parse_object_name(false)?;
+            let table_object = self.parse_table_object()?;
 
             let table_alias =
                 if dialect_of!(self is PostgreSqlDialect) && self.parse_keyword(Keyword::AS) {
@@ -12128,7 +12140,7 @@ impl<'a> Parser<'a> {
 
             Ok(Statement::Insert(Insert {
                 or,
-                table_name,
+                table: table_object,
                 table_alias,
                 ignore,
                 into,
@@ -12138,7 +12150,7 @@ impl<'a> Parser<'a> {
                 after_columns,
                 source,
                 assignments,
-                table,
+                has_table_keyword: table,
                 on,
                 returning,
                 replace_into,
diff --git a/src/test_utils.rs b/src/test_utils.rs
@@ -154,7 +154,6 @@ impl TestedDialects {
     pub fn one_statement_parses_to(&self, sql: &str, canonical: &str) -> Statement {
         let mut statements = self.parse_sql_statements(sql).expect(sql);
         assert_eq!(statements.len(), 1);
-
         if !canonical.is_empty() && sql != canonical {
             assert_eq!(self.parse_sql_statements(canonical).unwrap(), statements);
         }
diff --git a/tests/sqlparser_clickhouse.rs b/tests/sqlparser_clickhouse.rs
@@ -222,6 +222,12 @@ fn parse_create_table() {
     );
 }
 
+#[test]
+fn parse_insert_into_function() {
+    clickhouse().verified_stmt(r#"INSERT INTO TABLE FUNCTION remote('localhost', default.simple_table) VALUES (100, 'inserted via remote()')"#);
+    clickhouse().verified_stmt(r#"INSERT INTO FUNCTION remote('localhost', default.simple_table) VALUES (100, 'inserted via remote()')"#);
+}
+
 #[test]
 fn parse_alter_table_attach_and_detach_partition() {
     for operation in &["ATTACH", "DETACH"] {
diff --git a/tests/sqlparser_common.rs b/tests/sqlparser_common.rs
@@ -96,7 +96,7 @@ fn parse_insert_values() {
     ) {
         match verified_stmt(sql) {
             Statement::Insert(Insert {
-                table_name,
+                table: table_name,
                 columns,
                 source: Some(source),
                 ..
@@ -149,7 +149,7 @@ fn parse_insert_default_values() {
             partitioned,
             returning,
             source,
-            table_name,
+            table: table_name,
             ..
         }) => {
             assert_eq!(columns, vec![]);
@@ -158,7 +158,10 @@ fn parse_insert_default_values() {
             assert_eq!(partitioned, None);
             assert_eq!(returning, None);
             assert_eq!(source, None);
-            assert_eq!(table_name, ObjectName(vec!["test_table".into()]));
+            assert_eq!(
+                table_name,
+                TableObject::TableName(ObjectName(vec!["test_table".into()]))
+            );
         }
         _ => unreachable!(),
     }
@@ -174,7 +177,7 @@ fn parse_insert_default_values() {
             partitioned,
             returning,
             source,
-            table_name,
+            table: table_name,
             ..
         }) => {
             assert_eq!(after_columns, vec![]);
@@ -183,7 +186,10 @@ fn parse_insert_default_values() {
             assert_eq!(partitioned, None);
             assert!(returning.is_some());
             assert_eq!(source, None);
-            assert_eq!(table_name, ObjectName(vec!["test_table".into()]));
+            assert_eq!(
+                table_name,
+                TableObject::TableName(ObjectName(vec!["test_table".into()]))
+            );
         }
         _ => unreachable!(),
     }
@@ -199,7 +205,7 @@ fn parse_insert_default_values() {
             partitioned,
             returning,
             source,
-            table_name,
+            table: table_name,
             ..
         }) => {
             assert_eq!(after_columns, vec![]);
@@ -208,7 +214,10 @@ fn parse_insert_default_values() {
             assert_eq!(partitioned, None);
             assert_eq!(returning, None);
             assert_eq!(source, None);
-            assert_eq!(table_name, ObjectName(vec!["test_table".into()]));
+            assert_eq!(
+                table_name,
+                TableObject::TableName(ObjectName(vec!["test_table".into()]))
+            );
         }
         _ => unreachable!(),
     }
diff --git a/tests/sqlparser_mysql.rs b/tests/sqlparser_mysql.rs
diff --git a/tests/sqlparser_postgres.rs b/tests/sqlparser_postgres.rs

Original file line number	Diff line number	Diff line change
`@@ -50,4 +50,8 @@ impl Dialect for ClickHouseDialect {`
`50`	`50`	`fn supports_limit_comma(&self) -> bool {`
`51`	`51`	`true`
`52`	`52`	`}`
	`53`	`+`
	`54`	`+ fn supports_insert_table_function(&self) -> bool {`
	`55`	`+ true`
	`56`	`+ }`
`53`	`57`	`}`
Original file line number	Diff line number	Diff line change
`@@ -792,6 +792,11 @@ pub trait Dialect: Debug + Any {`
`792`	`792`	`fn supports_insert_set(&self) -> bool {`
`793`	`793`	`false`
`794`	`794`	`}`
	`795`	`+`
	`796`	`+ /// Does the dialect support table function in insertion?`
	`797`	`+ fn supports_insert_table_function(&self) -> bool {`
	`798`	`+ false`
	`799`	`+ }`
`795`	`800`	`}`
`796`	`801`
`797`	`802`	`/// This represents the operators for which precedence must be defined`
Original file line number	Diff line number	Diff line change
`@@ -154,7 +154,6 @@ impl TestedDialects {`
`154`	`154`	`pub fn one_statement_parses_to(&self, sql: &str, canonical: &str) -> Statement {`
`155`	`155`	`let mut statements = self.parse_sql_statements(sql).expect(sql);`
`156`	`156`	`assert_eq!(statements.len(), 1);`
`157`		`-`
`158`	`157`	`if !canonical.is_empty() && sql != canonical {`
`159`	`158`	`assert_eq!(self.parse_sql_statements(canonical).unwrap(), statements);`
`160`	`159`	`}`