MohamedAbdeen21 commented on code in PR #9912:
URL: https://github.com/apache/arrow-datafusion/pull/9912#discussion_r1553540658
##########
datafusion/core/src/datasource/listing/table.rs:
##########
@@ -438,6 +439,112 @@ impl ListingOptions {
self.format.infer_schema(state, &store, &files).await
}
+
+ /// Infers the partition columns stored in `LOCATION` and compares
+ /// them with the columns provided in `PARTITIONED BY` to help prevent
+ /// accidental corrupts of partitioned tables.
+ ///
+ /// Allows specifying partial partitions.
+ pub async fn validate_partitions(
+ &self,
+ state: &SessionState,
+ table_path: &ListingTableUrl,
+ ) -> Result<()> {
+ if self.table_partition_cols.is_empty() {
+ return Ok(());
+ }
+
+ if !table_path.is_collection() {
+ return plan_err!(
+ "Can't create a partitioned table backed by a single file, \
+ perhaps the URL is missing a trailing slash?"
+ );
+ }
+
+ let inferred = self.infer_partitions(state, table_path).await?;
+
+ // no partitioned files found on disk
+ if inferred.is_empty() {
+ return Ok(());
+ }
+
+ let table_partition_names = self
+ .table_partition_cols
+ .iter()
+ .map(|(col_name, _)| col_name.clone())
+ .collect_vec();
Review Comment:
Makes sense, but then we'll have to iterate on a clone because it's used in
the error message. I prefer having the collect over the clone.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
