introduce optional collect_block in segmentcollector

add collect_block in segment_collector to handle groups of documents as performance optimization

Author: PSeitz

src/collector/count_collector.rs

@@ -70,6 +70,11 @@ impl SegmentCollector for SegmentCountCollector {
         Ok(())
     }
 
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        self.count += docs.len();
+        Ok(())
+    }
+
     fn harvest(self) -> usize {
         self.count
     }

src/collector/mod.rs

@@ -172,19 +172,33 @@ pub trait Collector: Sync + Send {
     ) -> crate::Result<<Self::Child as SegmentCollector>::Fruit> {
         let mut segment_collector = self.for_segment(segment_ord as u32, reader)?;
 
+        let mut cache_pos = 0;
+        let mut cache = [(0, 0.0); 64];
+
         if let Some(alive_bitset) = reader.alive_bitset() {
             weight.for_each(reader, &mut |doc, score| {
                 if alive_bitset.is_alive(doc) {
-                    segment_collector.collect(doc, score)?;
+                    cache[cache_pos] = (doc, score);
+                    cache_pos += 1;
+                    if cache_pos == 64 {
+                        segment_collector.collect_block(&cache)?;
+                        cache_pos = 0;
+                    }
                 }
                 Ok(())
             })?;
         } else {
             weight.for_each(reader, &mut |doc, score| {
-                segment_collector.collect(doc, score)?;
+                cache[cache_pos] = (doc, score);
+                cache_pos += 1;
+                if cache_pos == 64 {
+                    segment_collector.collect_block(&cache)?;
+                    cache_pos = 0;
+                }
                 Ok(())
             })?;
         }
+        segment_collector.collect_block(&cache[..cache_pos])?;
         Ok(segment_collector.harvest())
     }
 }
@@ -258,6 +272,14 @@ pub trait SegmentCollector: 'static {
     /// The query pushes the scored document to the collector via this method.
     fn collect(&mut self, doc: DocId, score: Score) -> crate::Result<()>;
 
+    /// The query pushes the scored document to the collector via this method.
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        for (doc, score) in docs {
+            self.collect(*doc, *score)?;
+        }
+        Ok(())
+    }
+
     /// Extract the fruit of the collection from the `SegmentCollector`.
     fn harvest(self) -> Self::Fruit;
 }
@@ -317,6 +339,12 @@ where
         Ok(())
     }
 
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        self.0.collect_block(docs)?;
+        self.1.collect_block(docs)?;
+        Ok(())
+    }
+
     fn harvest(self) -> <Self as SegmentCollector>::Fruit {
         (self.0.harvest(), self.1.harvest())
     }
@@ -383,6 +411,13 @@ where
         Ok(())
     }
 
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        self.0.collect_block(docs)?;
+        self.1.collect_block(docs)?;
+        self.2.collect_block(docs)?;
+        Ok(())
+    }
+
     fn harvest(self) -> <Self as SegmentCollector>::Fruit {
         (self.0.harvest(), self.1.harvest(), self.2.harvest())
     }
@@ -459,6 +494,14 @@ where
         Ok(())
     }
 
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        self.0.collect_block(docs)?;
+        self.1.collect_block(docs)?;
+        self.2.collect_block(docs)?;
+        self.3.collect_block(docs)?;
+        Ok(())
+    }
+
     fn harvest(self) -> <Self as SegmentCollector>::Fruit {
         (
             self.0.harvest(),

src/collector/top_collector.rs

@@ -60,7 +60,8 @@ pub(crate) struct TopCollector<T> {
 }
 
 impl<T> TopCollector<T>
-where T: PartialOrd + Clone
+where
+    T: PartialOrd + Clone,
 {
     /// Creates a top collector, with a number of documents equal to "limit".
     ///

src/collector/top_score_collector.rs

@@ -704,6 +704,14 @@ impl SegmentCollector for TopScoreSegmentCollector {
         Ok(())
     }
 
+    #[inline]
+    fn collect_block(&mut self, docs: &[(DocId, Score)]) -> crate::Result<()> {
+        for (doc, score) in docs {
+            self.0.collect(*doc, *score);
+        }
+        Ok(())
+    }
+
     fn harvest(self) -> Vec<(Score, DocAddress)> {
         self.0.harvest()
     }