Optimize time ranged leaf search queries

When the search request contains a time range, we aborted the
optimization of converting unneeded split searches into count queries.

Author: tontinton

quickwit/quickwit-search/src/leaf.rs

@@ -942,11 +942,6 @@ fn is_simple_all_query(search_request: &SearchRequest) -> bool {
         return false;
     }
 
-    // TODO: Update the logic to handle start_timestamp end_timestamp ranges
-    if search_request.start_timestamp.is_some() || search_request.end_timestamp.is_some() {
-        return false;
-    }
-
     let Ok(query_ast) = serde_json::from_str(&search_request.query_ast) else {
         return false;
     };
@@ -1000,6 +995,29 @@ impl CanSplitDoBetter {
         }
     }
 
+    fn is_split_contained_in_search_time_range(
+        split: &SplitIdAndFooterOffsets,
+        search_request: &SearchRequest,
+    ) -> bool {
+        if let Some(start) = search_request.start_timestamp {
+            let Some(split_start) = split.timestamp_start else {
+                return false;
+            };
+            if split_start < start {
+                return false;
+            }
+        }
+        if let Some(end) = search_request.end_timestamp {
+            let Some(split_end) = split.timestamp_end else {
+                return false;
+            };
+            if split_end >= end {
+                return false;
+            }
+        }
+        true
+    }
+
     fn to_splits_and_request(
         splits: Vec<SplitIdAndFooterOffsets>,
         request: Arc<SearchRequest>,
@@ -1011,23 +1029,33 @@ impl CanSplitDoBetter {
     }
 
     /// Calculate the number of splits which are guaranteed to deliver enough documents.
+    ///
+    /// If there's a time range and not enough splits contain at least the number of requested
+    /// documents, return None.
     fn get_min_required_splits(
         splits: &[SplitIdAndFooterOffsets],
         request: &SearchRequest,
-    ) -> usize {
+    ) -> Option<usize> {
         let num_requested_docs = request.start_offset + request.max_hits;
 
-        splits
-            .into_iter()
-            .map(|split| split.num_docs)
-            // computing the partial sum
-            .scan(0u64, |partial_sum: &mut u64, num_docs_in_split: u64| {
-                *partial_sum += num_docs_in_split;
-                Some(*partial_sum)
-            })
-            .take_while(|partial_sum| *partial_sum < num_requested_docs)
-            .count()
-            + 1
+        let mut min_required_splits = 0;
+        let mut partial_sum = 0u64;
+
+        for split in splits.iter() {
+            if !Self::is_split_contained_in_search_time_range(split, &request) {
+                continue;
+            }
+
+            partial_sum += split.num_docs;
+
+            if partial_sum >= num_requested_docs {
+                return Some(min_required_splits + 1);
+            }
+
+            min_required_splits += 1;
+        }
+
+        None
     }
 
     fn optimize_split_id_higher(
@@ -1042,7 +1070,10 @@ impl CanSplitDoBetter {
             return Ok(Self::to_splits_and_request(splits, request));
         }
 
-        let min_required_splits = Self::get_min_required_splits(&splits, &request);
+        let Some(min_required_splits) = Self::get_min_required_splits(&splits, &request) else {
+            // not enough splits contained in time range.
+            return Ok(Self::to_splits_and_request(splits, request));
+        };
         let mut split_with_req = Self::to_splits_and_request(splits, request);
 
         // In this case there is no sort order, we order by split id.
@@ -1060,14 +1091,20 @@ impl CanSplitDoBetter {
         request: Arc<SearchRequest>,
         mut splits: Vec<SplitIdAndFooterOffsets>,
     ) -> Result<Vec<(SplitIdAndFooterOffsets, SearchRequest)>, SearchError> {
-        splits.sort_unstable_by_key(|split| std::cmp::Reverse(split.timestamp_end()));
+        splits.sort_unstable_by_key(|split| {
+            let contained = Self::is_split_contained_in_search_time_range(split, &request);
+            (!contained, std::cmp::Reverse(split.timestamp_end()))
+        });
 
         if !is_simple_all_query(&request) {
             // no optimization opportunity here.
             return Ok(Self::to_splits_and_request(splits, request));
         }
 
-        let min_required_splits = Self::get_min_required_splits(&splits, &request);
+        let Some(min_required_splits) = Self::get_min_required_splits(&splits, &request) else {
+            // not enough splits contained in time range.
+            return Ok(Self::to_splits_and_request(splits, request));
+        };
         let mut split_with_req = Self::to_splits_and_request(splits, request);
 
         // We order by timestamp desc. split_with_req is sorted by timestamp_end desc.
@@ -1097,14 +1134,20 @@ impl CanSplitDoBetter {
         request: Arc<SearchRequest>,
         mut splits: Vec<SplitIdAndFooterOffsets>,
     ) -> Result<Vec<(SplitIdAndFooterOffsets, SearchRequest)>, SearchError> {
-        splits.sort_unstable_by_key(|split| split.timestamp_start());
+        splits.sort_unstable_by_key(|split| {
+            let contained = Self::is_split_contained_in_search_time_range(split, &request);
+            (!contained, split.timestamp_start())
+        });
 
         if !is_simple_all_query(&request) {
             // no optimization opportunity here.
             return Ok(Self::to_splits_and_request(splits, request));
         }
 
-        let min_required_splits = Self::get_min_required_splits(&splits, &request);
+        let Some(min_required_splits) = Self::get_min_required_splits(&splits, &request) else {
+            // not enough splits contained in time range.
+            return Ok(Self::to_splits_and_request(splits, request));
+        };
         let mut split_with_req = Self::to_splits_and_request(splits, request);
 
         // We order by timestamp asc. split_with_req is sorted by timestamp_start.
@@ -1141,7 +1184,10 @@ impl CanSplitDoBetter {
         request: Arc<SearchRequest>,
         mut splits: Vec<SplitIdAndFooterOffsets>,
     ) -> Result<Vec<(SplitIdAndFooterOffsets, SearchRequest)>, SearchError> {
-        splits.sort_unstable_by_key(|split| std::cmp::Reverse(split.timestamp_end()));
+        splits.sort_unstable_by_key(|split| {
+            let contained = Self::is_split_contained_in_search_time_range(split, &request);
+            (!contained, std::cmp::Reverse(split.timestamp_end()))
+        });
 
         if !is_simple_all_query(&request) {
             // no optimization opportunity here.
@@ -1154,6 +1200,9 @@ impl CanSplitDoBetter {
     /// This function tries to detect upfront which splits contain the top n hits and convert other
     /// split searches to count only searches. It also optimizes split order.
     ///
+    /// To skip splits in time ranged queries, we sort the splits first by whether they are
+    /// contained in the search request time range.
+    ///
     /// Returns the search_requests with their split.
     fn optimize(
         &self,