houqp commented on a change in pull request #1665:
URL: https://github.com/apache/arrow-datafusion/pull/1665#discussion_r792300008
##########
File path: datafusion/src/datasource/object_store/mod.rs
##########
@@ -159,6 +159,16 @@ pub trait ObjectStore: Sync + Send + Debug {
/// Get object reader for one file
fn file_reader(&self, file: SizedFile) -> Result<Arc<dyn ObjectReader>>;
}
+/// Checks if we should filter out this path name
+///
https://github.com/apache/spark/blob/5b91381b9bcea27b7be6d9cef852efbd6c23d98a/core/src/main/scala/org/apache/spark/util/HadoopFSUtils.scala#L358
+fn do_path_filter_with_suffix(path: &str, suffix: &str) -> bool {
+ let exclude = path.starts_with('_') && !path.contains('=')
+ || path.starts_with('.')
+ || path.ends_with("._COPYING_");
+ let include = path.starts_with("_common_metadata") ||
path.starts_with("_metadata");
+ let is_suffix = path.ends_with(suffix);
+ (!exclude || include) && is_suffix
Review comment:
i recommend changing up the evaluation order so we can perform early
exit and avoid unnecessary string operations. For example:
```rust
if !path.ends_with(suffix) {
return false;
}
let exclude = ....;
if exlude {
let include = path.starts_with("_common_metadata") ||
path.starts_with("_metadata");
if !include {
return false;
}
}
return true;
```
--
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]
