feat: support export data for bigquery

Author: chenkovsky

src/ast/mod.rs

@@ -4355,6 +4355,15 @@ pub enum Statement {
     ///
     /// See [ReturnStatement]
     Return(ReturnStatement),
+
+    /// Export data statement
+    ///
+    /// Example:
+    /// ```sql
+    /// EXPORT DATA OPTIONS(uri='gs://bucket/folder/*', format='PARQUET', overwrite=true) AS
+    /// SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10
+    /// ```
+    ExportData(ExportData),
 }
 
 /// ```sql
@@ -6193,6 +6202,7 @@ impl fmt::Display for Statement {
             Statement::Return(r) => write!(f, "{r}"),
             Statement::List(command) => write!(f, "LIST {command}"),
             Statement::Remove(command) => write!(f, "REMOVE {command}"),
+            Statement::ExportData(e) => write!(f, "{e}"),
         }
     }
 }
@@ -10125,6 +10135,20 @@ impl fmt::Display for MemberOf {
     }
 }
 
+#[derive(Debug, Clone, PartialEq, PartialOrd, Eq, Ord, Hash)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[cfg_attr(feature = "visitor", derive(Visit, VisitMut))]
+pub struct ExportData {
+    pub options: Vec<SqlOption>,
+    pub query: Box<Query>,
+}
+
+impl fmt::Display for ExportData {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        write!(f, "EXPORT DATA OPTIONS({}) AS {}", display_comma_separated(&self.options), self.query)
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use crate::tokenizer::Location;

src/ast/spans.rs

@@ -15,7 +15,7 @@
 // specific language governing permissions and limitations
 // under the License.
 
-use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions};
+use crate::ast::{query::SelectItemQualifiedWildcardKind, ColumnOptions, ExportData};
 use core::iter;
 
 use crate::tokenizer::Span;
@@ -531,6 +531,9 @@ impl Spanned for Statement {
             Statement::Print { .. } => Span::empty(),
             Statement::Return { .. } => Span::empty(),
             Statement::List(..) | Statement::Remove(..) => Span::empty(),
+            Statement::ExportData(ExportData { options, query }) => union_spans(
+                options.iter().map(|i| i.span()).chain(core::iter::once(query.span()))
+            ),
         }
     }
 }

src/parser/mod.rs

@@ -640,6 +640,7 @@ impl<'a> Parser<'a> {
                 Keyword::COMMENT if self.dialect.supports_comment_on() => self.parse_comment(),
                 Keyword::PRINT => self.parse_print(),
                 Keyword::RETURN => self.parse_return(),
+                Keyword::EXPORT => self.parse_export(),
                 _ => self.expected("an SQL statement", next_token),
             },
             Token::LParen => {
@@ -16472,6 +16473,17 @@ impl<'a> Parser<'a> {
         }
     }
 
+    fn parse_export(&mut self) -> Result<Statement, ParserError> {
+        self.expect_keyword(Keyword::DATA)?;
+        self.expect_keyword(Keyword::OPTIONS)?;
+        self.expect_token(&Token::LParen)?;
+        let options = self.parse_comma_separated(|p| p.parse_sql_option())?;
+        self.expect_token(&Token::RParen)?;
+        self.expect_keyword(Keyword::AS)?;
+        let query = self.parse_query()?;
+        Ok(Statement::ExportData(ExportData { options, query }))
+    }
+
     /// Consume the parser and return its underlying token buffer
     pub fn into_tokens(self) -> Vec<TokenWithSpan> {
         self.tokens

tests/sqlparser_bigquery.rs

@@ -2566,3 +2566,15 @@ fn test_struct_trailing_and_nested_bracket() {
         )
     );
 }
+
+#[test]
+fn test_export() {
+    bigquery().verified_stmt(concat!(
+        "EXPORT DATA OPTIONS(",
+        "uri = 'gs://bucket/folder/*', ",
+        "format = 'PARQUET', ",
+        "overwrite = true",
+        ") AS ",
+        "SELECT field1, field2 FROM mydataset.table1 ORDER BY field1 LIMIT 10",
+    ));
+}