anoopj commented on code in PR #3015:
URL: https://github.com/apache/iceberg-rust/pull/3015#discussion_r3800087733
##########
crates/iceberg/src/arrow/caching_delete_file_loader.rs:
##########
@@ -374,16 +383,43 @@ impl CachingDeleteFileLoader {
));
}
- result
- .entry(file_path.to_string())
- .or_default()
- .insert(pos as u64);
+ if run_path != Some(file_path) {
+ if let Some(run_path) = run_path {
+ Self::merge_delete_positions(&mut result, run_path,
&run_positions);
+ run_positions.clear();
+ }
+
+ run_path = Some(file_path);
+ }
+
+ run_positions.push(pos as u64);
+ }
+
+ if let Some(run_path) = run_path {
+ Self::merge_delete_positions(&mut result, run_path,
&run_positions);
}
}
Ok(result)
}
+ /// Marks every position in `positions` as deleted for `file_path`, merging
+ /// into any delete vector already recorded for that file.
+ fn merge_delete_positions(
+ result: &mut HashMap<String, DeleteVector>,
+ file_path: &str,
+ positions: &[u64],
+ ) {
+ if positions.is_empty() {
+ return;
+ }
+
+ let delete_vector = result.entry(file_path.to_string()).or_default();
+ for &pos in positions {
+ delete_vector.insert(pos);
+ }
Review Comment:
Thanks for the pointer. We now call `insert_positions` first and only falls
back to the per-element loop on `Err`.
##########
crates/iceberg/src/arrow/caching_delete_file_loader.rs:
##########
@@ -360,6 +359,16 @@ impl CachingDeleteFileLoader {
));
};
+ // Within a batch, positional deletes are sorted by (file_path,
pos),
+ // so the rows for one data file form a contiguous run. Buffer each
+ // run and merge it with a single map lookup, allocating and
hashing
+ // the key once per run instead of once per row. Grouping is per
+ // batch, not across the whole stream: a run never spans batch
+ // boundaries, so a path that also appears in another batch merges
+ // into its existing delete vector (order does not affect the
result).
+ let mut run_path: Option<&str> = None;
+ let mut run_positions: Vec<u64> = Vec::new();
Review Comment:
Done.
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
