renaming

Removes the Write generics argument in PostingsSerializer.
This removes useless generic.
Prepares the path for codecs.
Removes one useless CountingWrite layer.
etc.

.github/workflows/test.yml

@@ -39,11 +39,11 @@ jobs:
 
     - name: Check Formatting
       run: cargo +nightly fmt --all -- --check
-    
+
     - name: Check Stable Compilation
       run: cargo build --all-features
 
-    
+
     - name: Check Bench Compilation
       run: cargo +nightly bench --no-run --profile=dev --all-features
 
@@ -59,10 +59,10 @@ jobs:
 
     strategy:
       matrix:
-        features: [
-            { label: "all", flags: "mmap,stopwords,lz4-compression,zstd-compression,failpoints" },
-            { label: "quickwit", flags: "mmap,quickwit,failpoints" }
-        ]
+        features:
+          - { label: "all", flags: "mmap,stopwords,lz4-compression,zstd-compression,failpoints,stemmer" }
+          - { label: "quickwit", flags: "mmap,quickwit,failpoints" }
+          - { label: "none", flags: "" }
 
     name: test-${{ matrix.features.label}}
 
@@ -80,7 +80,21 @@ jobs:
     - uses: Swatinem/rust-cache@v2
 
     - name: Run tests
-      run: cargo +stable nextest run --features ${{ matrix.features.flags }} --verbose --workspace
+      run: |
+        # if matrix.feature.flags is empty then run on --lib to avoid compiling examples
+        # (as most of them rely on mmap) otherwise run all
+        if [ -z "${{ matrix.features.flags }}" ]; then
+          cargo +stable nextest run --lib --no-default-features --verbose --workspace
+        else
+          cargo +stable nextest run --features ${{ matrix.features.flags }} --no-default-features --verbose --workspace
+        fi
 
     - name: Run doctests
-      run: cargo +stable test --doc --features ${{ matrix.features.flags }} --verbose --workspace
+      run: |
+        # if matrix.feature.flags is empty then run on --lib to avoid compiling examples
+        # (as most of them rely on mmap) otherwise run all
+        if [ -z "${{ matrix.features.flags }}" ]; then
+          echo "no doctest for no feature flag"
+        else
+          cargo +stable test --doc --features ${{ matrix.features.flags }} --verbose --workspace
+        fi

Cargo.toml

@@ -37,9 +37,9 @@ fs4 = { version = "0.13.1", optional = true }
 levenshtein_automata = "0.2.1"
 uuid = { version = "1.0.0", features = ["v4", "serde"] }
 crossbeam-channel = "0.5.4"
-rust-stemmers = "1.2.0"
+rust-stemmers = { version = "1.2.0", optional = true }
 downcast-rs = "2.0.1"
-bitpacking = { version = "0.9.2", default-features = false, features = [
+bitpacking = { version = "0.9.3", default-features = false, features = [
     "bitpacker4x",
 ] }
 census = "0.4.2"
@@ -113,7 +113,8 @@ debug-assertions = true
 overflow-checks = true
 
 [features]
-default = ["mmap", "stopwords", "lz4-compression", "columnar-zstd-compression"]
+default = ["mmap", "stopwords", "lz4-compression", "columnar-zstd-compression", "stemmer"]
+stemmer = ["rust-stemmers"]
 mmap = ["fs4", "tempfile", "memmap2"]
 stopwords = []
 

benches/agg_bench.rs

@@ -54,33 +54,33 @@ fn bench_agg(mut group: InputGroup<Index>) {
     register!(group, stats_f64);
     register!(group, extendedstats_f64);
     register!(group, percentiles_f64);
-    register!(group, terms_few);
+    register!(group, terms_7);
     register!(group, terms_all_unique);
-    register!(group, terms_many);
+    register!(group, terms_150_000);
     register!(group, terms_many_top_1000);
     register!(group, terms_many_order_by_term);
     register!(group, terms_many_with_top_hits);
     register!(group, terms_all_unique_with_avg_sub_agg);
     register!(group, terms_many_with_avg_sub_agg);
-    register!(group, terms_few_with_avg_sub_agg);
     register!(group, terms_status_with_avg_sub_agg);
-    register!(group, terms_status);
-    register!(group, terms_few_with_histogram);
     register!(group, terms_status_with_histogram);
+    register!(group, terms_zipf_1000);
+    register!(group, terms_zipf_1000_with_histogram);
+    register!(group, terms_zipf_1000_with_avg_sub_agg);
 
     register!(group, terms_many_json_mixed_type_with_avg_sub_agg);
 
     register!(group, cardinality_agg);
-    register!(group, terms_few_with_cardinality_agg);
+    register!(group, terms_status_with_cardinality_agg);
 
     register!(group, range_agg);
     register!(group, range_agg_with_avg_sub_agg);
-    register!(group, range_agg_with_term_agg_few);
+    register!(group, range_agg_with_term_agg_status);
     register!(group, range_agg_with_term_agg_many);
     register!(group, histogram);
     register!(group, histogram_hard_bounds);
     register!(group, histogram_with_avg_sub_agg);
-    register!(group, histogram_with_term_agg_few);
+    register!(group, histogram_with_term_agg_status);
     register!(group, avg_and_range_with_avg_sub_agg);
 
     // Filter aggregation benchmarks
@@ -159,10 +159,10 @@ fn cardinality_agg(index: &Index) {
     });
     execute_agg(index, agg_req);
 }
-fn terms_few_with_cardinality_agg(index: &Index) {
+fn terms_status_with_cardinality_agg(index: &Index) {
     let agg_req = json!({
         "my_texts": {
-            "terms": { "field": "text_few_terms" },
+            "terms": { "field": "text_few_terms_status" },
             "aggs": {
                 "cardinality": {
                     "cardinality": {
@@ -175,13 +175,7 @@ fn terms_few_with_cardinality_agg(index: &Index) {
     execute_agg(index, agg_req);
 }
 
-fn terms_few(index: &Index) {
-    let agg_req = json!({
-        "my_texts": { "terms": { "field": "text_few_terms" } },
-    });
-    execute_agg(index, agg_req);
-}
-fn terms_status(index: &Index) {
+fn terms_7(index: &Index) {
     let agg_req = json!({
         "my_texts": { "terms": { "field": "text_few_terms_status" } },
     });
@@ -194,7 +188,7 @@ fn terms_all_unique(index: &Index) {
     execute_agg(index, agg_req);
 }
 
-fn terms_many(index: &Index) {
+fn terms_150_000(index: &Index) {
     let agg_req = json!({
         "my_texts": { "terms": { "field": "text_many_terms" } },
     });
@@ -253,17 +247,6 @@ fn terms_all_unique_with_avg_sub_agg(index: &Index) {
     });
     execute_agg(index, agg_req);
 }
-fn terms_few_with_histogram(index: &Index) {
-    let agg_req = json!({
-        "my_texts": {
-            "terms": { "field": "text_few_terms" },
-            "aggs": {
-                "histo": {"histogram": { "field": "score_f64", "interval": 10 }}
-            }
-        }
-    });
-    execute_agg(index, agg_req);
-}
 fn terms_status_with_histogram(index: &Index) {
     let agg_req = json!({
         "my_texts": {
@@ -276,17 +259,18 @@ fn terms_status_with_histogram(index: &Index) {
     execute_agg(index, agg_req);
 }
 
-fn terms_few_with_avg_sub_agg(index: &Index) {
+fn terms_zipf_1000_with_histogram(index: &Index) {
     let agg_req = json!({
         "my_texts": {
-            "terms": { "field": "text_few_terms" },
+            "terms": { "field": "text_1000_terms_zipf" },
             "aggs": {
-                "average_f64": { "avg": { "field": "score_f64" } }
+                "histo": {"histogram": { "field": "score_f64", "interval": 10 }}
             }
-        },
+        }
     });
     execute_agg(index, agg_req);
 }
+
 fn terms_status_with_avg_sub_agg(index: &Index) {
     let agg_req = json!({
         "my_texts": {
@@ -299,6 +283,25 @@ fn terms_status_with_avg_sub_agg(index: &Index) {
     execute_agg(index, agg_req);
 }
 
+fn terms_zipf_1000_with_avg_sub_agg(index: &Index) {
+    let agg_req = json!({
+        "my_texts": {
+            "terms": { "field": "text_1000_terms_zipf" },
+            "aggs": {
+                "average_f64": { "avg": { "field": "score_f64" } }
+            }
+        },
+    });
+    execute_agg(index, agg_req);
+}
+
+fn terms_zipf_1000(index: &Index) {
+    let agg_req = json!({
+        "my_texts": { "terms": { "field": "text_1000_terms_zipf" } },
+    });
+    execute_agg(index, agg_req);
+}
+
 fn terms_many_json_mixed_type_with_avg_sub_agg(index: &Index) {
     let agg_req = json!({
         "my_texts": {
@@ -354,7 +357,7 @@ fn range_agg_with_avg_sub_agg(index: &Index) {
     execute_agg(index, agg_req);
 }
 
-fn range_agg_with_term_agg_few(index: &Index) {
+fn range_agg_with_term_agg_status(index: &Index) {
     let agg_req = json!({
         "rangef64": {
             "range": {
@@ -369,7 +372,7 @@ fn range_agg_with_term_agg_few(index: &Index) {
                 ]
             },
             "aggs": {
-                "my_texts": { "terms": { "field": "text_few_terms" } },
+                "my_texts": { "terms": { "field": "text_few_terms_status" } },
             }
         },
     });
@@ -425,12 +428,12 @@ fn histogram_with_avg_sub_agg(index: &Index) {
     });
     execute_agg(index, agg_req);
 }
-fn histogram_with_term_agg_few(index: &Index) {
+fn histogram_with_term_agg_status(index: &Index) {
     let agg_req = json!({
         "rangef64": {
             "histogram": { "field": "score_f64", "interval": 10 },
             "aggs": {
-                "my_texts": { "terms": { "field": "text_few_terms" } }
+                "my_texts": { "terms": { "field": "text_few_terms_status" } }
             }
         }
     });
@@ -475,6 +478,13 @@ fn get_collector(agg_req: Aggregations) -> AggregationCollector {
 }
 
 fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
+    // Flag to use existing index
+    let reuse_index = std::env::var("REUSE_AGG_BENCH_INDEX").is_ok();
+    if reuse_index && std::path::Path::new("agg_bench").exists() {
+        return Index::open_in_dir("agg_bench");
+    }
+    // crreate dir
+    std::fs::create_dir_all("agg_bench")?;
     let mut schema_builder = Schema::builder();
     let text_fieldtype = tantivy::schema::TextOptions::default()
         .set_indexing_options(
@@ -486,24 +496,44 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
     let text_field_all_unique_terms =
         schema_builder.add_text_field("text_all_unique_terms", STRING | FAST);
     let text_field_many_terms = schema_builder.add_text_field("text_many_terms", STRING | FAST);
-    let text_field_many_terms = schema_builder.add_text_field("text_many_terms", STRING | FAST);
-    let text_field_few_terms = schema_builder.add_text_field("text_few_terms", STRING | FAST);
     let text_field_few_terms_status =
         schema_builder.add_text_field("text_few_terms_status", STRING | FAST);
+    let text_field_1000_terms_zipf =
+        schema_builder.add_text_field("text_1000_terms_zipf", STRING | FAST);
     let score_fieldtype = tantivy::schema::NumericOptions::default().set_fast();
     let score_field = schema_builder.add_u64_field("score", score_fieldtype.clone());
     let score_field_f64 = schema_builder.add_f64_field("score_f64", score_fieldtype.clone());
     let score_field_i64 = schema_builder.add_i64_field("score_i64", score_fieldtype);
-    let index = Index::create_from_tempdir(schema_builder.build())?;
-    let few_terms_data = ["INFO", "ERROR", "WARN", "DEBUG"];
-    // Approximate production log proportions: INFO dominant, WARN and DEBUG occasional, ERROR rare.
-    let log_level_distribution = WeightedIndex::new([80u32, 3, 12, 5]).unwrap();
+    // use tmp dir
+    let index = if reuse_index {
+        Index::create_in_dir("agg_bench", schema_builder.build())?
+    } else {
+        Index::create_from_tempdir(schema_builder.build())?
+    };
+    // Approximate log proportions
+    let status_field_data = [
+        ("INFO", 8000),
+        ("ERROR", 300),
+        ("WARN", 1200),
+        ("DEBUG", 500),
+        ("OK", 500),
+        ("CRITICAL", 20),
+        ("EMERGENCY", 1),
+    ];
+    let log_level_distribution =
+        WeightedIndex::new(status_field_data.iter().map(|item| item.1)).unwrap();
 
     let lg_norm = rand_distr::LogNormal::new(2.996f64, 0.979f64).unwrap();
 
     let many_terms_data = (0..150_000)
         .map(|num| format!("author{num}"))
         .collect::<Vec<_>>();
+
+    // Prepare 1000 unique terms sampled using a Zipf distribution.
+    // Exponent ~1.1 approximates top-20 terms covering around ~20%.
+    let terms_1000: Vec<String> = (1..=1000).map(|i| format!("term_{i}")).collect();
+    let zipf_1000 = rand_distr::Zipf::new(1000, 1.1f64).unwrap();
+
     {
         let mut rng = StdRng::from_seed([1u8; 32]);
         let mut index_writer = index.writer_with_num_threads(1, 200_000_000)?;
@@ -513,8 +543,12 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
             index_writer.add_document(doc!())?;
         }
         if cardinality == Cardinality::Multivalued {
-            let log_level_sample_a = few_terms_data[log_level_distribution.sample(&mut rng)];
-            let log_level_sample_b = few_terms_data[log_level_distribution.sample(&mut rng)];
+            let log_level_sample_a = status_field_data[log_level_distribution.sample(&mut rng)].0;
+            let log_level_sample_b = status_field_data[log_level_distribution.sample(&mut rng)].0;
+            let idx_a = zipf_1000.sample(&mut rng) as usize - 1;
+            let idx_b = zipf_1000.sample(&mut rng) as usize - 1;
+            let term_1000_a = &terms_1000[idx_a];
+            let term_1000_b = &terms_1000[idx_b];
             index_writer.add_document(doc!(
                 json_field => json!({"mixed_type": 10.0}),
                 json_field => json!({"mixed_type": 10.0}),
@@ -524,10 +558,10 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
                 text_field_all_unique_terms => "coolo",
                 text_field_many_terms => "cool",
                 text_field_many_terms => "cool",
-                text_field_few_terms => "cool",
-                text_field_few_terms => "cool",
                 text_field_few_terms_status => log_level_sample_a,
                 text_field_few_terms_status => log_level_sample_b,
+                text_field_1000_terms_zipf => term_1000_a.as_str(),
+                text_field_1000_terms_zipf => term_1000_b.as_str(),
                 score_field => 1u64,
                 score_field => 1u64,
                 score_field_f64 => lg_norm.sample(&mut rng),
@@ -554,8 +588,8 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
                 json_field => json,
                 text_field_all_unique_terms => format!("unique_term_{}", rng.gen::<u64>()),
                 text_field_many_terms => many_terms_data.choose(&mut rng).unwrap().to_string(),
-                text_field_few_terms => few_terms_data.choose(&mut rng).unwrap().to_string(),
-                text_field_few_terms_status => few_terms_data[log_level_distribution.sample(&mut rng)],
+                text_field_few_terms_status => status_field_data[log_level_distribution.sample(&mut rng)].0,
+                text_field_1000_terms_zipf => terms_1000[zipf_1000.sample(&mut rng) as usize - 1].as_str(),
                 score_field => val as u64,
                 score_field_f64 => lg_norm.sample(&mut rng),
                 score_field_i64 => val as i64,
@@ -607,7 +641,7 @@ fn filter_agg_all_query_with_sub_aggs(index: &Index) {
                 "avg_score": { "avg": { "field": "score" } },
                 "stats_score": { "stats": { "field": "score_f64" } },
                 "terms_text": {
-                    "terms": { "field": "text_few_terms" }
+                    "terms": { "field": "text_few_terms_status" }
                 }
             }
         }
@@ -623,7 +657,7 @@ fn filter_agg_term_query_with_sub_aggs(index: &Index) {
                 "avg_score": { "avg": { "field": "score" } },
                 "stats_score": { "stats": { "field": "score_f64" } },
                 "terms_text": {
-                    "terms": { "field": "text_few_terms" }
+                    "terms": { "field": "text_few_terms_status" }
                 }
             }
         }

columnar/src/block_accessor.rs

@@ -29,12 +29,20 @@ impl<T: PartialOrd + Copy + std::fmt::Debug + Send + Sync + 'static + Default>
         }
     }
     #[inline]
-    pub fn fetch_block_with_missing(&mut self, docs: &[u32], accessor: &Column<T>, missing: T) {
+    pub fn fetch_block_with_missing(
+        &mut self,
+        docs: &[u32],
+        accessor: &Column<T>,
+        missing: Option<T>,
+    ) {
         self.fetch_block(docs, accessor);
         // no missing values
         if accessor.index.get_cardinality().is_full() {
             return;
         }
+        let Some(missing) = missing else {
+            return;
+        };
 
         // We can compare docid_cache length with docs to find missing docs
         // For multi value columns we can't rely on the length and always need to scan

columnar/src/column/mod.rs

@@ -85,8 +85,8 @@ impl<T: PartialOrd + Copy + Debug + Send + Sync + 'static> Column<T> {
     }
 
     #[inline]
-    pub fn first(&self, row_id: RowId) -> Option<T> {
-        self.values_for_doc(row_id).next()
+    pub fn first(&self, doc_id: DocId) -> Option<T> {
+        self.values_for_doc(doc_id).next()
     }
 
     /// Load the first value for each docid in the provided slice.

columnar/src/tests.rs

@@ -60,7 +60,7 @@ fn test_dataframe_writer_bool() {
     let DynamicColumn::Bool(bool_col) = dyn_bool_col else {
         panic!();
     };
-    let vals: Vec<Option<bool>> = (0..5).map(|row_id| bool_col.first(row_id)).collect();
+    let vals: Vec<Option<bool>> = (0..5).map(|doc_id| bool_col.first(doc_id)).collect();
     assert_eq!(&vals, &[None, Some(false), None, Some(true), None,]);
 }
 
@@ -108,7 +108,7 @@ fn test_dataframe_writer_ip_addr() {
     let DynamicColumn::IpAddr(ip_col) = dyn_bool_col else {
         panic!();
     };
-    let vals: Vec<Option<Ipv6Addr>> = (0..5).map(|row_id| ip_col.first(row_id)).collect();
+    let vals: Vec<Option<Ipv6Addr>> = (0..5).map(|doc_id| ip_col.first(doc_id)).collect();
     assert_eq!(
         &vals,
         &[
@@ -169,7 +169,7 @@ fn test_dictionary_encoded_str() {
     let DynamicColumn::Str(str_col) = col_handles[0].open().unwrap() else {
         panic!();
     };
-    let index: Vec<Option<u64>> = (0..5).map(|row_id| str_col.ords().first(row_id)).collect();
+    let index: Vec<Option<u64>> = (0..5).map(|doc_id| str_col.ords().first(doc_id)).collect();
     assert_eq!(index, &[None, Some(0), None, Some(2), Some(1)]);
     assert_eq!(str_col.num_rows(), 5);
     let mut term_buffer = String::new();
@@ -204,7 +204,7 @@ fn test_dictionary_encoded_bytes() {
         panic!();
     };
     let index: Vec<Option<u64>> = (0..5)
-        .map(|row_id| bytes_col.ords().first(row_id))
+        .map(|doc_id| bytes_col.ords().first(doc_id))
         .collect();
     assert_eq!(index, &[None, Some(0), None, Some(2), Some(1)]);
     assert_eq!(bytes_col.num_rows(), 5);

common/src/bitset.rs

@@ -181,6 +181,14 @@ pub struct BitSet {
     len: u64,
     max_value: u32,
 }
+impl std::fmt::Debug for BitSet {
+    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+        f.debug_struct("BitSet")
+            .field("len", &self.len)
+            .field("max_value", &self.max_value)
+            .finish()
+    }
+}
 
 fn num_buckets(max_val: u32) -> u32 {
     max_val.div_ceil(64u32)

examples/iterating_docs_and_positions.rs

@@ -91,46 +91,10 @@ fn main() -> tantivy::Result<()> {
         }
     }
 
-    // A `Term` is a text token associated with a field.
-    // Let's go through all docs containing the term `title:the` and access their position
-    let term_the = Term::from_field_text(title, "the");
-
-    // Some other powerful operations (especially `.skip_to`) may be useful to consume these
+    // Some other powerful operations (especially `.seek`) may be useful to consume these
     // posting lists rapidly.
     // You can check for them in the [`DocSet`](https://docs.rs/tantivy/~0/tantivy/trait.DocSet.html) trait
     // and the [`Postings`](https://docs.rs/tantivy/~0/tantivy/trait.Postings.html) trait
 
-    // Also, for some VERY specific high performance use case like an OLAP analysis of logs,
-    // you can get better performance by accessing directly the blocks of doc ids.
-    for segment_reader in searcher.segment_readers() {
-        // A segment contains different data structure.
-        // Inverted index stands for the combination of
-        // - the term dictionary
-        // - the inverted lists associated with each terms and their positions
-        let inverted_index = segment_reader.inverted_index(title)?;
-
-        // This segment posting object is like a cursor over the documents matching the term.
-        // The `IndexRecordOption` arguments tells tantivy we will be interested in both term
-        // frequencies and positions.
-        //
-        // If you don't need all this information, you may get better performance by decompressing
-        // less information.
-        if let Some(mut block_segment_postings) =
-            inverted_index.read_block_postings(&term_the, IndexRecordOption::Basic)?
-        {
-            loop {
-                let docs = block_segment_postings.docs();
-                if docs.is_empty() {
-                    break;
-                }
-                // Once again these docs MAY contains deleted documents as well.
-                let docs = block_segment_postings.docs();
-                // Prints `Docs [0, 2].`
-                println!("Docs {docs:?}");
-                block_segment_postings.advance();
-            }
-        }
-    }
-
     Ok(())
 }

src/aggregation/agg_data.rs

@@ -1,4 +1,4 @@
-use columnar::{Column, ColumnType, StrColumn};
+use columnar::{Column, ColumnBlockAccessor, ColumnType, StrColumn};
 use common::BitSet;
 use rustc_hash::FxHashSet;
 use serde::Serialize;
@@ -10,16 +10,16 @@ use crate::aggregation::accessor_helpers::{
 };
 use crate::aggregation::agg_req::{Aggregation, AggregationVariants, Aggregations};
 use crate::aggregation::bucket::{
-    FilterAggReqData, HistogramAggReqData, HistogramBounds, IncludeExcludeParam,
-    MissingTermAggReqData, RangeAggReqData, SegmentFilterCollector, SegmentHistogramCollector,
-    SegmentRangeCollector, TermMissingAgg, TermsAggReqData, TermsAggregation,
+    build_segment_filter_collector, build_segment_range_collector, FilterAggReqData,
+    HistogramAggReqData, HistogramBounds, IncludeExcludeParam, MissingTermAggReqData,
+    RangeAggReqData, SegmentHistogramCollector, TermMissingAgg, TermsAggReqData, TermsAggregation,
     TermsAggregationInternal,
 };
 use crate::aggregation::metric::{
-    AverageAggregation, CardinalityAggReqData, CardinalityAggregationReq, CountAggregation,
-    ExtendedStatsAggregation, MaxAggregation, MetricAggReqData, MinAggregation,
-    SegmentCardinalityCollector, SegmentExtendedStatsCollector, SegmentPercentilesCollector,
-    SegmentStatsCollector, StatsAggregation, StatsType, SumAggregation, TopHitsAggReqData,
+    build_segment_stats_collector, AverageAggregation, CardinalityAggReqData,
+    CardinalityAggregationReq, CountAggregation, ExtendedStatsAggregation, MaxAggregation,
+    MetricAggReqData, MinAggregation, SegmentCardinalityCollector, SegmentExtendedStatsCollector,
+    SegmentPercentilesCollector, StatsAggregation, StatsType, SumAggregation, TopHitsAggReqData,
     TopHitsSegmentCollector,
 };
 use crate::aggregation::segment_agg_result::{
@@ -35,6 +35,7 @@ pub struct AggregationsSegmentCtx {
     /// Request data for each aggregation type.
     pub per_request: PerRequestAggSegCtx,
     pub context: AggContextParams,
+    pub column_block_accessor: ColumnBlockAccessor<u64>,
 }
 
 impl AggregationsSegmentCtx {
@@ -107,21 +108,14 @@ impl AggregationsSegmentCtx {
             .as_deref()
             .expect("range_req_data slot is empty (taken)")
     }
-    #[inline]
-    pub(crate) fn get_filter_req_data(&self, idx: usize) -> &FilterAggReqData {
-        self.per_request.filter_req_data[idx]
-            .as_deref()
-            .expect("filter_req_data slot is empty (taken)")
-    }
 
     // ---------- mutable getters ----------
 
     #[inline]
-    pub(crate) fn get_term_req_data_mut(&mut self, idx: usize) -> &mut TermsAggReqData {
-        self.per_request.term_req_data[idx]
-            .as_deref_mut()
-            .expect("term_req_data slot is empty (taken)")
+    pub(crate) fn get_metric_req_data_mut(&mut self, idx: usize) -> &mut MetricAggReqData {
+        &mut self.per_request.stats_metric_req_data[idx]
     }
+
     #[inline]
     pub(crate) fn get_cardinality_req_data_mut(
         &mut self,
@@ -129,10 +123,7 @@ impl AggregationsSegmentCtx {
     ) -> &mut CardinalityAggReqData {
         &mut self.per_request.cardinality_req_data[idx]
     }
-    #[inline]
-    pub(crate) fn get_metric_req_data_mut(&mut self, idx: usize) -> &mut MetricAggReqData {
-        &mut self.per_request.stats_metric_req_data[idx]
-    }
+
     #[inline]
     pub(crate) fn get_histogram_req_data_mut(&mut self, idx: usize) -> &mut HistogramAggReqData {
         self.per_request.histogram_req_data[idx]
@@ -142,21 +133,6 @@ impl AggregationsSegmentCtx {
 
     // ---------- take / put (terms, histogram, range) ----------
 
-    /// Move out the boxed Terms request at `idx`, leaving `None`.
-    #[inline]
-    pub(crate) fn take_term_req_data(&mut self, idx: usize) -> Box<TermsAggReqData> {
-        self.per_request.term_req_data[idx]
-            .take()
-            .expect("term_req_data slot is empty (taken)")
-    }
-
-    /// Put back a Terms request into an empty slot at `idx`.
-    #[inline]
-    pub(crate) fn put_back_term_req_data(&mut self, idx: usize, value: Box<TermsAggReqData>) {
-        debug_assert!(self.per_request.term_req_data[idx].is_none());
-        self.per_request.term_req_data[idx] = Some(value);
-    }
-
     /// Move out the boxed Histogram request at `idx`, leaving `None`.
     #[inline]
     pub(crate) fn take_histogram_req_data(&mut self, idx: usize) -> Box<HistogramAggReqData> {
@@ -320,6 +296,7 @@ impl PerRequestAggSegCtx {
 
     /// Convert the aggregation tree into a serializable struct representation.
     /// Each node contains: { name, kind, children }.
+    #[allow(dead_code)]
     pub fn get_view_tree(&self) -> Vec<AggTreeViewNode> {
         fn node_to_view(node: &AggRefNode, pr: &PerRequestAggSegCtx) -> AggTreeViewNode {
             let mut children: Vec<AggTreeViewNode> =
@@ -345,12 +322,19 @@ impl PerRequestAggSegCtx {
 pub(crate) fn build_segment_agg_collectors_root(
     req: &mut AggregationsSegmentCtx,
 ) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
-    build_segment_agg_collectors(req, &req.per_request.agg_tree.clone())
+    build_segment_agg_collectors_generic(req, &req.per_request.agg_tree.clone())
 }
 
 pub(crate) fn build_segment_agg_collectors(
     req: &mut AggregationsSegmentCtx,
     nodes: &[AggRefNode],
+) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
+    build_segment_agg_collectors_generic(req, nodes)
+}
+
+fn build_segment_agg_collectors_generic(
+    req: &mut AggregationsSegmentCtx,
+    nodes: &[AggRefNode],
 ) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
     let mut collectors = Vec::new();
     for node in nodes.iter() {
@@ -388,6 +372,8 @@ pub(crate) fn build_segment_agg_collector(
             Ok(Box::new(SegmentCardinalityCollector::from_req(
                 req_data.column_type,
                 node.idx_in_req_data,
+                req_data.accessor.clone(),
+                req_data.missing_value_for_accessor,
             )))
         }
         AggKind::StatsKind(stats_type) => {
@@ -398,20 +384,21 @@ pub(crate) fn build_segment_agg_collector(
                 | StatsType::Count
                 | StatsType::Max
                 | StatsType::Min
-                | StatsType::Stats => Ok(Box::new(SegmentStatsCollector::from_req(
-                    node.idx_in_req_data,
-                ))),
-                StatsType::ExtendedStats(sigma) => {
-                    Ok(Box::new(SegmentExtendedStatsCollector::from_req(
-                        req_data.field_type,
-                        sigma,
-                        node.idx_in_req_data,
-                        req_data.missing,
-                    )))
-                }
-                StatsType::Percentiles => Ok(Box::new(
-                    SegmentPercentilesCollector::from_req_and_validate(node.idx_in_req_data)?,
+                | StatsType::Stats => build_segment_stats_collector(req_data),
+                StatsType::ExtendedStats(sigma) => Ok(Box::new(
+                    SegmentExtendedStatsCollector::from_req(req_data, sigma),
                 )),
+                StatsType::Percentiles => {
+                    let req_data = req.get_metric_req_data_mut(node.idx_in_req_data);
+                    Ok(Box::new(
+                        SegmentPercentilesCollector::from_req_and_validate(
+                            req_data.field_type,
+                            req_data.missing_u64,
+                            req_data.accessor.clone(),
+                            node.idx_in_req_data,
+                        ),
+                    ))
+                }
             }
         }
         AggKind::TopHits => {
@@ -428,12 +415,8 @@ pub(crate) fn build_segment_agg_collector(
         AggKind::DateHistogram => Ok(Box::new(SegmentHistogramCollector::from_req_and_validate(
             req, node,
         )?)),
-        AggKind::Range => Ok(Box::new(SegmentRangeCollector::from_req_and_validate(
-            req, node,
-        )?)),
-        AggKind::Filter => Ok(Box::new(SegmentFilterCollector::from_req_and_validate(
-            req, node,
-        )?)),
+        AggKind::Range => Ok(build_segment_range_collector(req, node)?),
+        AggKind::Filter => build_segment_filter_collector(req, node),
     }
 }
 
@@ -493,6 +476,7 @@ pub(crate) fn build_aggregations_data_from_req(
     let mut data = AggregationsSegmentCtx {
         per_request: Default::default(),
         context,
+        column_block_accessor: ColumnBlockAccessor::default(),
     };
 
     for (name, agg) in aggs.iter() {
@@ -521,9 +505,9 @@ fn build_nodes(
             let idx_in_req_data = data.push_range_req_data(RangeAggReqData {
                 accessor,
                 field_type,
-                column_block_accessor: Default::default(),
                 name: agg_name.to_string(),
                 req: range_req.clone(),
+                is_top_level,
             });
             let children = build_children(&req.sub_aggregation, reader, segment_ordinal, data)?;
             Ok(vec![AggRefNode {
@@ -541,9 +525,7 @@ fn build_nodes(
             let idx_in_req_data = data.push_histogram_req_data(HistogramAggReqData {
                 accessor,
                 field_type,
-                column_block_accessor: Default::default(),
                 name: agg_name.to_string(),
-                sub_aggregation_blueprint: None,
                 req: histo_req.clone(),
                 is_date_histogram: false,
                 bounds: HistogramBounds {
@@ -568,9 +550,7 @@ fn build_nodes(
             let idx_in_req_data = data.push_histogram_req_data(HistogramAggReqData {
                 accessor,
                 field_type,
-                column_block_accessor: Default::default(),
                 name: agg_name.to_string(),
-                sub_aggregation_blueprint: None,
                 req: histo_req,
                 is_date_histogram: true,
                 bounds: HistogramBounds {
@@ -650,7 +630,6 @@ fn build_nodes(
             let idx_in_req_data = data.push_metric_req_data(MetricAggReqData {
                 accessor,
                 field_type,
-                column_block_accessor: Default::default(),
                 name: agg_name.to_string(),
                 collecting_for,
                 missing: *missing,
@@ -678,7 +657,6 @@ fn build_nodes(
             let idx_in_req_data = data.push_metric_req_data(MetricAggReqData {
                 accessor,
                 field_type,
-                column_block_accessor: Default::default(),
                 name: agg_name.to_string(),
                 collecting_for: StatsType::Percentiles,
                 missing: percentiles_req.missing,
@@ -753,6 +731,7 @@ fn build_nodes(
                 segment_reader: reader.clone(),
                 evaluator,
                 matching_docs_buffer,
+                is_top_level,
             });
             let children = build_children(&req.sub_aggregation, reader, segment_ordinal, data)?;
             Ok(vec![AggRefNode {
@@ -895,7 +874,7 @@ fn build_terms_or_cardinality_nodes(
         });
     }
 
-    // Add one node per accessor to mirror previous behavior and allow per-type missing handling.
+    // Add one node per accessor
     for (accessor, column_type) in column_and_types {
         let missing_value_for_accessor = if use_special_missing_agg {
             None
@@ -926,11 +905,8 @@ fn build_terms_or_cardinality_nodes(
                     column_type,
                     str_dict_column: str_dict_column.clone(),
                     missing_value_for_accessor,
-                    column_block_accessor: Default::default(),
                     name: agg_name.to_string(),
                     req: TermsAggregationInternal::from_req(req),
-                    // Will be filled later when building collectors
-                    sub_aggregation_blueprint: None,
                     sug_aggregations: sub_aggs.clone(),
                     allowed_term_ids,
                     is_top_level,
@@ -943,7 +919,6 @@ fn build_terms_or_cardinality_nodes(
                     column_type,
                     str_dict_column: str_dict_column.clone(),
                     missing_value_for_accessor,
-                    column_block_accessor: Default::default(),
                     name: agg_name.to_string(),
                     req: req.clone(),
                 });

src/aggregation/agg_tests.rs

@@ -2,15 +2,441 @@ use serde_json::Value;
 
 use crate::aggregation::agg_req::{Aggregation, Aggregations};
 use crate::aggregation::agg_result::AggregationResults;
-use crate::aggregation::buf_collector::DOC_BLOCK_SIZE;
 use crate::aggregation::collector::AggregationCollector;
 use crate::aggregation::intermediate_agg_result::IntermediateAggregationResults;
 use crate::aggregation::tests::{get_test_index_2_segments, get_test_index_from_values_and_terms};
 use crate::aggregation::DistributedAggregationCollector;
+use crate::docset::COLLECT_BLOCK_BUFFER_LEN;
 use crate::query::{AllQuery, TermQuery};
 use crate::schema::{IndexRecordOption, Schema, FAST};
 use crate::{Index, IndexWriter, Term};
 
+// The following tests ensure that each bucket aggregation type correctly functions as a
+// sub-aggregation of another bucket aggregation in two scenarios:
+// 1) The parent has more buckets than the child sub-aggregation
+// 2) The child sub-aggregation has more buckets than the parent
+//
+// These scenarios exercise the bucket id mapping and sub-aggregation routing logic.
+
+#[test]
+fn test_terms_as_subagg_parent_more_vs_child_more() -> crate::Result<()> {
+    let index = get_test_index_2_segments(false)?;
+
+    // Case A: parent has more buckets than child
+    // Parent: range with 4 buckets
+    // Child: terms on text -> 2 buckets
+    let agg_parent_more: Aggregations = serde_json::from_value(json!({
+        "parent_range": {
+            "range": {
+                "field": "score",
+                "ranges": [
+                    {"to": 3.0},
+                    {"from": 3.0, "to": 7.0},
+                    {"from": 7.0, "to": 20.0},
+                    {"from": 20.0}
+                ]
+            },
+            "aggs": {
+                "child_terms": {"terms": {"field": "text", "order": {"_key": "asc"}}}
+            }
+        }
+    }))
+    .unwrap();
+
+    let res = crate::aggregation::tests::exec_request(agg_parent_more, &index)?;
+    // Exact expected structure and counts
+    assert_eq!(
+        res["parent_range"]["buckets"],
+        json!([
+            {
+                "key": "*-3",
+                "doc_count": 1,
+                "to": 3.0,
+                "child_terms": {
+                    "buckets": [
+                        {"doc_count": 1, "key": "cool"}
+                    ],
+                    "sum_other_doc_count": 0
+                }
+            },
+            {
+                "key": "3-7",
+                "doc_count": 3,
+                "from": 3.0,
+                "to": 7.0,
+                "child_terms": {
+                    "buckets": [
+                        {"doc_count": 2, "key": "cool"},
+                        {"doc_count": 1, "key": "nohit"}
+                    ],
+                    "sum_other_doc_count": 0
+                }
+            },
+            {
+                "key": "7-20",
+                "doc_count": 3,
+                "from": 7.0,
+                "to": 20.0,
+                "child_terms": {
+                    "buckets": [
+                        {"doc_count": 3, "key": "cool"}
+                    ],
+                    "sum_other_doc_count": 0
+                }
+            },
+            {
+                "key": "20-*",
+                "doc_count": 2,
+                "from": 20.0,
+                "child_terms": {
+                    "buckets": [
+                        {"doc_count": 1, "key": "cool"},
+                        {"doc_count": 1, "key": "nohit"}
+                    ],
+                    "sum_other_doc_count": 0
+                }
+            }
+        ])
+    );
+
+    // Case B: child has more buckets than parent
+    // Parent: histogram on score with large interval -> 1 bucket
+    // Child: terms on text -> 2 buckets (cool/nohit)
+    let agg_child_more: Aggregations = serde_json::from_value(json!({
+        "parent_hist": {
+            "histogram": {"field": "score", "interval": 100.0},
+            "aggs": {
+                "child_terms": {"terms": {"field": "text", "order": {"_key": "asc"}}}
+            }
+        }
+    }))
+    .unwrap();
+
+    let res = crate::aggregation::tests::exec_request(agg_child_more, &index)?;
+    assert_eq!(
+        res["parent_hist"],
+        json!({
+            "buckets": [
+                {
+                    "key": 0.0,
+                    "doc_count": 9,
+                    "child_terms": {
+                        "buckets": [
+                            {"doc_count": 7, "key": "cool"},
+                            {"doc_count": 2, "key": "nohit"}
+                        ],
+                        "sum_other_doc_count": 0
+                    }
+                }
+            ]
+        })
+    );
+
+    Ok(())
+}
+
+#[test]
+fn test_range_as_subagg_parent_more_vs_child_more() -> crate::Result<()> {
+    let index = get_test_index_2_segments(false)?;
+
+    // Case A: parent has more buckets than child
+    // Parent: range with 5 buckets
+    // Child: coarse range with 3 buckets
+    let agg_parent_more: Aggregations = serde_json::from_value(json!({
+        "parent_range": {
+            "range": {
+                "field": "score",
+                "ranges": [
+                    {"to": 3.0},
+                    {"from": 3.0, "to": 7.0},
+                    {"from": 7.0, "to": 11.0},
+                    {"from": 11.0, "to": 20.0},
+                    {"from": 20.0}
+                ]
+            },
+            "aggs": {
+                "child_range": {
+                    "range": {
+                        "field": "score",
+                        "ranges": [
+                            {"to": 3.0},
+                            {"from": 3.0, "to": 20.0}
+                        ]
+                    }
+                }
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_parent_more, &index)?;
+    assert_eq!(
+        res["parent_range"]["buckets"],
+        json!([
+            {"key": "*-3", "doc_count": 1, "to": 3.0,
+                "child_range": {"buckets": [
+                    {"key": "*-3", "doc_count": 1, "to": 3.0},
+                    {"key": "3-20", "doc_count": 0, "from": 3.0, "to": 20.0},
+                    {"key": "20-*", "doc_count": 0, "from": 20.0}
+                ]}
+            },
+            {"key": "3-7", "doc_count": 3, "from": 3.0, "to": 7.0,
+                "child_range": {"buckets": [
+                    {"key": "*-3", "doc_count": 0, "to": 3.0},
+                    {"key": "3-20", "doc_count": 3, "from": 3.0, "to": 20.0},
+                    {"key": "20-*", "doc_count": 0, "from": 20.0}
+                ]}
+            },
+            {"key": "7-11", "doc_count": 1, "from": 7.0, "to": 11.0,
+                "child_range": {"buckets": [
+                    {"key": "*-3", "doc_count": 0, "to": 3.0},
+                    {"key": "3-20", "doc_count": 1, "from": 3.0, "to": 20.0},
+                    {"key": "20-*", "doc_count": 0, "from": 20.0}
+                ]}
+            },
+            {"key": "11-20", "doc_count": 2, "from": 11.0, "to": 20.0,
+                "child_range": {"buckets": [
+                    {"key": "*-3", "doc_count": 0, "to": 3.0},
+                    {"key": "3-20", "doc_count": 2, "from": 3.0, "to": 20.0},
+                    {"key": "20-*", "doc_count": 0, "from": 20.0}
+                ]}
+            },
+            {"key": "20-*", "doc_count": 2, "from": 20.0,
+                "child_range": {"buckets": [
+                    {"key": "*-3", "doc_count": 0, "to": 3.0},
+                    {"key": "3-20", "doc_count": 0, "from": 3.0, "to": 20.0},
+                    {"key": "20-*", "doc_count": 2, "from": 20.0}
+                ]}
+            }
+        ])
+    );
+
+    // Case B: child has more buckets than parent
+    // Parent: terms on text (2 buckets)
+    // Child: range with 4 buckets
+    let agg_child_more: Aggregations = serde_json::from_value(json!({
+        "parent_terms": {
+            "terms": {"field": "text"},
+            "aggs": {
+                "child_range": {
+                    "range": {
+                        "field": "score",
+                        "ranges": [
+                            {"to": 3.0},
+                            {"from": 3.0, "to": 7.0},
+                            {"from": 7.0, "to": 20.0}
+                        ]
+                    }
+                }
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_child_more, &index)?;
+
+    assert_eq!(
+        res["parent_terms"],
+        json!({
+            "buckets": [
+                {
+                    "key": "cool",
+                    "doc_count": 7,
+                    "child_range": {
+                        "buckets": [
+                            {"key": "*-3", "doc_count": 1, "to": 3.0},
+                            {"key": "3-7", "doc_count": 2, "from": 3.0, "to": 7.0},
+                            {"key": "7-20", "doc_count": 3, "from": 7.0, "to": 20.0},
+                            {"key": "20-*", "doc_count": 1, "from": 20.0}
+                        ]
+                    }
+                },
+                {
+                    "key": "nohit",
+                    "doc_count": 2,
+                    "child_range": {
+                        "buckets": [
+                            {"key": "*-3", "doc_count": 0, "to": 3.0},
+                            {"key": "3-7", "doc_count": 1, "from": 3.0, "to": 7.0},
+                            {"key": "7-20", "doc_count": 0, "from": 7.0, "to": 20.0},
+                            {"key": "20-*", "doc_count": 1, "from": 20.0}
+                        ]
+                    }
+                }
+            ],
+            "doc_count_error_upper_bound": 0,
+            "sum_other_doc_count": 0
+        })
+    );
+
+    Ok(())
+}
+
+#[test]
+fn test_histogram_as_subagg_parent_more_vs_child_more() -> crate::Result<()> {
+    let index = get_test_index_2_segments(false)?;
+
+    // Case A: parent has more buckets than child
+    // Parent: range with several ranges
+    // Child: histogram with large interval (single bucket per parent)
+    let agg_parent_more: Aggregations = serde_json::from_value(json!({
+        "parent_range": {
+            "range": {
+                "field": "score",
+                "ranges": [
+                    {"to": 3.0},
+                    {"from": 3.0, "to": 7.0},
+                    {"from": 7.0, "to": 11.0},
+                    {"from": 11.0, "to": 20.0},
+                    {"from": 20.0}
+                ]
+            },
+            "aggs": {
+                "child_hist": {"histogram": {"field": "score", "interval": 100.0}}
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_parent_more, &index)?;
+    assert_eq!(
+        res["parent_range"]["buckets"],
+        json!([
+            {"key": "*-3", "doc_count": 1, "to": 3.0,
+                "child_hist": {"buckets": [ {"key": 0.0, "doc_count": 1} ]}
+            },
+            {"key": "3-7", "doc_count": 3, "from": 3.0, "to": 7.0,
+                "child_hist": {"buckets": [ {"key": 0.0, "doc_count": 3} ]}
+            },
+            {"key": "7-11", "doc_count": 1, "from": 7.0, "to": 11.0,
+                "child_hist": {"buckets": [ {"key": 0.0, "doc_count": 1} ]}
+            },
+            {"key": "11-20", "doc_count": 2, "from": 11.0, "to": 20.0,
+                "child_hist": {"buckets": [ {"key": 0.0, "doc_count": 2} ]}
+            },
+            {"key": "20-*", "doc_count": 2, "from": 20.0,
+                "child_hist": {"buckets": [ {"key": 0.0, "doc_count": 2} ]}
+            }
+        ])
+    );
+
+    // Case B: child has more buckets than parent
+    // Parent: terms on text -> 2 buckets
+    // Child: histogram with small interval -> multiple buckets including empties
+    let agg_child_more: Aggregations = serde_json::from_value(json!({
+        "parent_terms": {
+            "terms": {"field": "text"},
+            "aggs": {
+                "child_hist": {"histogram": {"field": "score", "interval": 10.0}}
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_child_more, &index)?;
+    assert_eq!(
+        res["parent_terms"],
+        json!({
+            "buckets": [
+                {
+                    "key": "cool",
+                    "doc_count": 7,
+                    "child_hist": {
+                        "buckets": [
+                            {"key": 0.0, "doc_count": 4},
+                            {"key": 10.0, "doc_count": 2},
+                            {"key": 20.0, "doc_count": 0},
+                            {"key": 30.0, "doc_count": 0},
+                            {"key": 40.0, "doc_count": 1}
+                        ]
+                    }
+                },
+                {
+                    "key": "nohit",
+                    "doc_count": 2,
+                    "child_hist": {
+                        "buckets": [
+                            {"key": 0.0, "doc_count": 1},
+                            {"key": 10.0, "doc_count": 0},
+                            {"key": 20.0, "doc_count": 0},
+                            {"key": 30.0, "doc_count": 0},
+                            {"key": 40.0, "doc_count": 1}
+                        ]
+                    }
+                }
+            ],
+            "doc_count_error_upper_bound": 0,
+            "sum_other_doc_count": 0
+        })
+    );
+
+    Ok(())
+}
+
+#[test]
+fn test_date_histogram_as_subagg_parent_more_vs_child_more() -> crate::Result<()> {
+    let index = get_test_index_2_segments(false)?;
+
+    // Case A: parent has more buckets than child
+    // Parent: range with several buckets
+    // Child: date_histogram with 30d -> single bucket per parent
+    let agg_parent_more: Aggregations = serde_json::from_value(json!({
+        "parent_range": {
+            "range": {
+                "field": "score",
+                "ranges": [
+                    {"to": 3.0},
+                    {"from": 3.0, "to": 7.0},
+                    {"from": 7.0, "to": 11.0},
+                    {"from": 11.0, "to": 20.0},
+                    {"from": 20.0}
+                ]
+            },
+            "aggs": {
+                "child_date_hist": {"date_histogram": {"field": "date", "fixed_interval": "30d"}}
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_parent_more, &index)?;
+    let buckets = res["parent_range"]["buckets"].as_array().unwrap();
+    // Verify each parent bucket has exactly one child date bucket with matching doc_count
+    for bucket in buckets {
+        let parent_count = bucket["doc_count"].as_u64().unwrap();
+        let child_buckets = bucket["child_date_hist"]["buckets"].as_array().unwrap();
+        assert_eq!(child_buckets.len(), 1);
+        assert_eq!(child_buckets[0]["doc_count"], parent_count);
+    }
+
+    // Case B: child has more buckets than parent
+    // Parent: terms on text (2 buckets)
+    // Child: date_histogram with 1d -> multiple buckets
+    let agg_child_more: Aggregations = serde_json::from_value(json!({
+        "parent_terms": {
+            "terms": {"field": "text"},
+            "aggs": {
+                "child_date_hist": {"date_histogram": {"field": "date", "fixed_interval": "1d"}}
+            }
+        }
+    }))
+    .unwrap();
+    let res = crate::aggregation::tests::exec_request(agg_child_more, &index)?;
+    let buckets = res["parent_terms"]["buckets"].as_array().unwrap();
+
+    // cool bucket
+    assert_eq!(buckets[0]["key"], "cool");
+    let cool_buckets = buckets[0]["child_date_hist"]["buckets"].as_array().unwrap();
+    assert_eq!(cool_buckets.len(), 3);
+    assert_eq!(cool_buckets[0]["doc_count"], 1); // day 0
+    assert_eq!(cool_buckets[1]["doc_count"], 4); // day 1
+    assert_eq!(cool_buckets[2]["doc_count"], 2); // day 2
+
+    // nohit bucket
+    assert_eq!(buckets[1]["key"], "nohit");
+    let nohit_buckets = buckets[1]["child_date_hist"]["buckets"].as_array().unwrap();
+    assert_eq!(nohit_buckets.len(), 2);
+    assert_eq!(nohit_buckets[0]["doc_count"], 1); // day 1
+    assert_eq!(nohit_buckets[1]["doc_count"], 1); // day 2
+
+    Ok(())
+}
+
 fn get_avg_req(field_name: &str) -> Aggregation {
     serde_json::from_value(json!({
         "avg": {
@@ -25,6 +451,10 @@ fn get_collector(agg_req: Aggregations) -> AggregationCollector {
 }
 
 // *** EVERY BUCKET-TYPE SHOULD BE TESTED HERE ***
+// Note: The flushng part of these  tests are outdated, since the buffering change after converting
+// the collection into one collector per request instead of per bucket.
+//
+// However they are useful as they test a complex aggregation requests.
 fn test_aggregation_flushing(
     merge_segments: bool,
     use_distributed_collector: bool,
@@ -37,8 +467,9 @@ fn test_aggregation_flushing(
 
     let reader = index.reader()?;
 
-    assert_eq!(DOC_BLOCK_SIZE, 64);
-    // In the tree we cache Documents of DOC_BLOCK_SIZE, before passing them down as one block.
+    assert_eq!(COLLECT_BLOCK_BUFFER_LEN, 64);
+    // In the tree we cache documents of COLLECT_BLOCK_BUFFER_LEN before passing them down as one
+    // block.
     //
     // Build a request so that on the first level we have one full cache, which is then flushed.
     // The same cache should have some residue docs at the end, which are flushed (Range 0-70)

src/aggregation/bucket/filter.rs

@@ -6,10 +6,14 @@ use serde::{Deserialize, Deserializer, Serialize, Serializer};
 use crate::aggregation::agg_data::{
     build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
 };
+use crate::aggregation::cached_sub_aggs::{
+    CachedSubAggs, HighCardSubAggCache, LowCardSubAggCache, SubAggCache,
+};
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
 };
-use crate::aggregation::segment_agg_result::{CollectorClone, SegmentAggregationCollector};
+use crate::aggregation::segment_agg_result::{BucketIdProvider, SegmentAggregationCollector};
+use crate::aggregation::BucketId;
 use crate::docset::DocSet;
 use crate::query::{AllQuery, EnableScoring, Query, QueryParser};
 use crate::schema::Schema;
@@ -404,15 +408,18 @@ pub struct FilterAggReqData {
     pub evaluator: DocumentQueryEvaluator,
     /// Reusable buffer for matching documents to minimize allocations during collection
     pub matching_docs_buffer: Vec<DocId>,
+    /// True if this filter aggregation is at the top level of the aggregation tree (not nested).
+    pub is_top_level: bool,
 }
 
 impl FilterAggReqData {
     pub(crate) fn get_memory_consumption(&self) -> usize {
         // Estimate: name + segment reader reference + bitset + buffer capacity
         self.name.len()
-            + std::mem::size_of::<SegmentReader>()
-            + self.evaluator.bitset.len() / 8 // BitSet memory (bits to bytes)
-            + self.matching_docs_buffer.capacity() * std::mem::size_of::<DocId>()
+        + std::mem::size_of::<SegmentReader>()
+        + self.evaluator.bitset.len() / 8 // BitSet memory (bits to bytes)
+        + self.matching_docs_buffer.capacity() * std::mem::size_of::<DocId>()
+        + std::mem::size_of::<bool>()
     }
 }
 
@@ -446,7 +453,7 @@ impl DocumentQueryEvaluator {
         let weight = query.weight(EnableScoring::disabled_from_schema(&schema))?;
 
         // Get a scorer that iterates over matching documents
-        let mut scorer = weight.scorer(segment_reader, 1.0, 0)?;
+        let mut scorer = weight.scorer(segment_reader, 1.0)?;
 
         // Create a BitSet to hold all matching documents
         let mut bitset = BitSet::with_max_value(max_doc);
@@ -489,17 +496,24 @@ impl Debug for DocumentQueryEvaluator {
     }
 }
 
-/// Segment collector for filter aggregation
-pub struct SegmentFilterCollector {
-    /// Document count in this bucket
+#[derive(Debug, Clone, PartialEq, Copy)]
+struct DocCount {
     doc_count: u64,
+    bucket_id: BucketId,
+}
+
+/// Segment collector for filter aggregation
+pub struct SegmentFilterCollector<C: SubAggCache> {
+    /// Document counts per parent bucket
+    parent_buckets: Vec<DocCount>,
     /// Sub-aggregation collectors
-    sub_aggregations: Option<Box<dyn SegmentAggregationCollector>>,
+    sub_aggregations: Option<CachedSubAggs<C>>,
+    bucket_id_provider: BucketIdProvider,
     /// Accessor index for this filter aggregation (to access FilterAggReqData)
     accessor_idx: usize,
 }
 
-impl SegmentFilterCollector {
+impl<C: SubAggCache> SegmentFilterCollector<C> {
     /// Create a new filter segment collector following the new agg_data pattern
     pub(crate) fn from_req_and_validate(
         req: &mut AggregationsSegmentCtx,
@@ -511,47 +525,75 @@ impl SegmentFilterCollector {
         } else {
             None
         };
+        let sub_agg_collector = sub_agg_collector.map(CachedSubAggs::new);
 
         Ok(SegmentFilterCollector {
-            doc_count: 0,
+            parent_buckets: Vec::new(),
             sub_aggregations: sub_agg_collector,
             accessor_idx: node.idx_in_req_data,
+            bucket_id_provider: BucketIdProvider::default(),
         })
     }
 }
 
-impl Debug for SegmentFilterCollector {
+pub(crate) fn build_segment_filter_collector(
+    req: &mut AggregationsSegmentCtx,
+    node: &AggRefNode,
+) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
+    let is_top_level = req.per_request.filter_req_data[node.idx_in_req_data]
+        .as_ref()
+        .expect("filter_req_data slot is empty")
+        .is_top_level;
+
+    if is_top_level {
+        Ok(Box::new(
+            SegmentFilterCollector::<LowCardSubAggCache>::from_req_and_validate(req, node)?,
+        ))
+    } else {
+        Ok(Box::new(
+            SegmentFilterCollector::<HighCardSubAggCache>::from_req_and_validate(req, node)?,
+        ))
+    }
+}
+
+impl<C: SubAggCache> Debug for SegmentFilterCollector<C> {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         f.debug_struct("SegmentFilterCollector")
-            .field("doc_count", &self.doc_count)
+            .field("buckets", &self.parent_buckets)
             .field("has_sub_aggs", &self.sub_aggregations.is_some())
             .field("accessor_idx", &self.accessor_idx)
             .finish()
     }
 }
 
-impl CollectorClone for SegmentFilterCollector {
-    fn clone_box(&self) -> Box<dyn SegmentAggregationCollector> {
-        // For now, panic - this needs proper implementation with weight recreation
-        panic!("SegmentFilterCollector cloning not yet implemented - requires weight recreation")
-    }
-}
-
-impl SegmentAggregationCollector for SegmentFilterCollector {
+impl<C: SubAggCache> SegmentAggregationCollector for SegmentFilterCollector<C> {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
         let mut sub_results = IntermediateAggregationResults::default();
+        let bucket_opt = self.parent_buckets.get(parent_bucket_id as usize);
 
-        if let Some(sub_aggs) = self.sub_aggregations {
-            sub_aggs.add_intermediate_aggregation_result(agg_data, &mut sub_results)?;
+        if let Some(sub_aggs) = &mut self.sub_aggregations {
+            sub_aggs
+                .get_sub_agg_collector()
+                .add_intermediate_aggregation_result(
+                    agg_data,
+                    &mut sub_results,
+                    // Here we create a new bucket ID for sub-aggregations if the bucket doesn't
+                    // exist, so that sub-aggregations can still produce results (e.g., zero doc
+                    // count)
+                    bucket_opt
+                        .map(|bucket| bucket.bucket_id)
+                        .unwrap_or(self.bucket_id_provider.next_bucket_id()),
+                )?;
         }
 
         // Create the filter bucket result
         let filter_bucket_result = IntermediateBucketResult::Filter {
-            doc_count: self.doc_count,
+            doc_count: bucket_opt.map(|b| b.doc_count).unwrap_or(0),
             sub_aggregations: sub_results,
         };
 
@@ -570,32 +612,17 @@ impl SegmentAggregationCollector for SegmentFilterCollector {
         Ok(())
     }
 
-    fn collect(&mut self, doc: DocId, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
-        // Access the evaluator from FilterAggReqData
-        let req_data = agg_data.get_filter_req_data(self.accessor_idx);
-
-        // O(1) BitSet lookup to check if document matches filter
-        if req_data.evaluator.matches_document(doc) {
-            self.doc_count += 1;
-
-            // If we have sub-aggregations, collect on them for this filtered document
-            if let Some(sub_aggs) = &mut self.sub_aggregations {
-                sub_aggs.collect(doc, agg_data)?;
-            }
-        }
-        Ok(())
-    }
-
-    #[inline]
-    fn collect_block(
+    fn collect(
         &mut self,
-        docs: &[DocId],
+        parent_bucket_id: BucketId,
+        docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
         if docs.is_empty() {
             return Ok(());
         }
 
+        let mut bucket = self.parent_buckets[parent_bucket_id as usize];
         // Take the request data to avoid borrow checker issues with sub-aggregations
         let mut req = agg_data.take_filter_req_data(self.accessor_idx);
 
@@ -604,18 +631,24 @@ impl SegmentAggregationCollector for SegmentFilterCollector {
         req.evaluator
             .filter_batch(docs, &mut req.matching_docs_buffer);
 
-        self.doc_count += req.matching_docs_buffer.len() as u64;
+        bucket.doc_count += req.matching_docs_buffer.len() as u64;
 
         // Batch process sub-aggregations if we have matches
         if !req.matching_docs_buffer.is_empty() {
             if let Some(sub_aggs) = &mut self.sub_aggregations {
-                // Use collect_block for better sub-aggregation performance
-                sub_aggs.collect_block(&req.matching_docs_buffer, agg_data)?;
+                for &doc_id in &req.matching_docs_buffer {
+                    sub_aggs.push(bucket.bucket_id, doc_id);
+                }
             }
         }
 
         // Put the request data back
         agg_data.put_back_filter_req_data(self.accessor_idx, req);
+        if let Some(sub_aggs) = &mut self.sub_aggregations {
+            sub_aggs.check_flush_local(agg_data)?;
+        }
+        // put back bucket
+        self.parent_buckets[parent_bucket_id as usize] = bucket;
 
         Ok(())
     }
@@ -626,6 +659,21 @@ impl SegmentAggregationCollector for SegmentFilterCollector {
         }
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        while self.parent_buckets.len() <= max_bucket as usize {
+            let bucket_id = self.bucket_id_provider.next_bucket_id();
+            self.parent_buckets.push(DocCount {
+                doc_count: 0,
+                bucket_id,
+            });
+        }
+        Ok(())
+    }
 }
 
 /// Intermediate result for filter aggregation
@@ -1519,9 +1567,9 @@ mod tests {
         let searcher = reader.searcher();
 
         let agg = json!({
-            "test": {
-                "filter": deserialized,
-                "aggs": { "count": { "value_count": { "field": "brand" } } }
+                "test": {
+                    "filter": deserialized,
+                    "aggs": { "count": { "value_count": { "field": "brand" } } }
             }
         });
 

src/aggregation/bucket/histogram/histogram.rs

@@ -1,6 +1,6 @@
 use std::cmp::Ordering;
 
-use columnar::{Column, ColumnBlockAccessor, ColumnType};
+use columnar::{Column, ColumnType};
 use rustc_hash::FxHashMap;
 use serde::{Deserialize, Serialize};
 use tantivy_bitpacker::minmax;
@@ -8,14 +8,14 @@ use tantivy_bitpacker::minmax;
 use crate::aggregation::agg_data::{
     build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
 };
-use crate::aggregation::agg_limits::MemoryConsumption;
 use crate::aggregation::agg_req::Aggregations;
 use crate::aggregation::agg_result::BucketEntry;
+use crate::aggregation::cached_sub_aggs::{CachedSubAggs, HighCardCachedSubAggs};
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
     IntermediateHistogramBucketEntry,
 };
-use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
+use crate::aggregation::segment_agg_result::{BucketIdProvider, SegmentAggregationCollector};
 use crate::aggregation::*;
 use crate::TantivyError;
 
@@ -26,13 +26,8 @@ pub struct HistogramAggReqData {
     pub accessor: Column<u64>,
     /// The field type of the fast field.
     pub field_type: ColumnType,
-    /// The column block accessor to access the fast field values.
-    pub column_block_accessor: ColumnBlockAccessor<u64>,
     /// The name of the aggregation.
     pub name: String,
-    /// The sub aggregation blueprint, used to create sub aggregations for each bucket.
-    /// Will be filled during initialization of the collector.
-    pub sub_aggregation_blueprint: Option<Box<dyn SegmentAggregationCollector>>,
     /// The histogram aggregation request.
     pub req: HistogramAggregation,
     /// True if this is a date_histogram aggregation.
@@ -257,18 +252,24 @@ impl HistogramBounds {
 pub(crate) struct SegmentHistogramBucketEntry {
     pub key: f64,
     pub doc_count: u64,
+    pub bucket_id: BucketId,
 }
 
 impl SegmentHistogramBucketEntry {
     pub(crate) fn into_intermediate_bucket_entry(
         self,
-        sub_aggregation: Option<Box<dyn SegmentAggregationCollector>>,
+        sub_aggregation: &mut Option<HighCardCachedSubAggs>,
         agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<IntermediateHistogramBucketEntry> {
         let mut sub_aggregation_res = IntermediateAggregationResults::default();
         if let Some(sub_aggregation) = sub_aggregation {
             sub_aggregation
-                .add_intermediate_aggregation_result(agg_data, &mut sub_aggregation_res)?;
+                .get_sub_agg_collector()
+                .add_intermediate_aggregation_result(
+                    agg_data,
+                    &mut sub_aggregation_res,
+                    self.bucket_id,
+                )?;
         }
         Ok(IntermediateHistogramBucketEntry {
             key: self.key,
@@ -278,27 +279,38 @@ impl SegmentHistogramBucketEntry {
     }
 }
 
+#[derive(Clone, Debug, Default)]
+struct HistogramBuckets {
+    pub buckets: FxHashMap<i64, SegmentHistogramBucketEntry>,
+}
+
 /// The collector puts values from the fast field into the correct buckets and does a conversion to
 /// the correct datatype.
-#[derive(Clone, Debug)]
+#[derive(Debug)]
 pub struct SegmentHistogramCollector {
     /// The buckets containing the aggregation data.
-    buckets: FxHashMap<i64, SegmentHistogramBucketEntry>,
-    sub_aggregations: FxHashMap<i64, Box<dyn SegmentAggregationCollector>>,
+    /// One Histogram bucket per parent bucket id.
+    parent_buckets: Vec<HistogramBuckets>,
+    sub_agg: Option<HighCardCachedSubAggs>,
     accessor_idx: usize,
+    bucket_id_provider: BucketIdProvider,
 }
 
 impl SegmentAggregationCollector for SegmentHistogramCollector {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
         let name = agg_data
             .get_histogram_req_data(self.accessor_idx)
             .name
             .clone();
-        let bucket = self.into_intermediate_bucket_result(agg_data)?;
+        // TODO: avoid prepare_max_bucket here and handle empty buckets.
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
+        let histogram = std::mem::take(&mut self.parent_buckets[parent_bucket_id as usize]);
+        let bucket = self.add_intermediate_bucket_result(agg_data, histogram)?;
         results.push(name, IntermediateAggregationResult::Bucket(bucket))?;
 
         Ok(())
@@ -307,44 +319,40 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
     #[inline]
     fn collect(
         &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.collect_block(&[doc], agg_data)
-    }
-
-    #[inline]
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let mut req = agg_data.take_histogram_req_data(self.accessor_idx);
+        let req = agg_data.take_histogram_req_data(self.accessor_idx);
         let mem_pre = self.get_memory_consumption();
+        let buckets = &mut self.parent_buckets[parent_bucket_id as usize].buckets;
 
         let bounds = req.bounds;
         let interval = req.req.interval;
         let offset = req.offset;
         let get_bucket_pos = |val| get_bucket_pos_f64(val, interval, offset) as i64;
 
-        req.column_block_accessor.fetch_block(docs, &req.accessor);
-        for (doc, val) in req
+        agg_data
+            .column_block_accessor
+            .fetch_block(docs, &req.accessor);
+        for (doc, val) in agg_data
             .column_block_accessor
             .iter_docid_vals(docs, &req.accessor)
         {
-            let val = f64_from_fastfield_u64(val, &req.field_type);
+            let val = f64_from_fastfield_u64(val, req.field_type);
             let bucket_pos = get_bucket_pos(val);
             if bounds.contains(val) {
-                let bucket = self.buckets.entry(bucket_pos).or_insert_with(|| {
+                let bucket = buckets.entry(bucket_pos).or_insert_with(|| {
                     let key = get_bucket_key_from_pos(bucket_pos as f64, interval, offset);
-                    SegmentHistogramBucketEntry { key, doc_count: 0 }
+                    SegmentHistogramBucketEntry {
+                        key,
+                        doc_count: 0,
+                        bucket_id: self.bucket_id_provider.next_bucket_id(),
+                    }
                 });
                 bucket.doc_count += 1;
-                if let Some(sub_aggregation_blueprint) = req.sub_aggregation_blueprint.as_ref() {
-                    self.sub_aggregations
-                        .entry(bucket_pos)
-                        .or_insert_with(|| sub_aggregation_blueprint.clone())
-                        .collect(doc, agg_data)?;
+                if let Some(sub_agg) = &mut self.sub_agg {
+                    sub_agg.push(bucket.bucket_id, doc);
                 }
             }
         }
@@ -358,14 +366,30 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
                 .add_memory_consumed(mem_delta as u64)?;
         }
 
+        if let Some(sub_agg) = &mut self.sub_agg {
+            sub_agg.check_flush_local(agg_data)?;
+        }
+
         Ok(())
     }
 
     fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
-        for sub_aggregation in self.sub_aggregations.values_mut() {
+        if let Some(sub_aggregation) = &mut self.sub_agg {
             sub_aggregation.flush(agg_data)?;
         }
+        Ok(())
+    }
 
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        while self.parent_buckets.len() <= max_bucket as usize {
+            self.parent_buckets.push(HistogramBuckets {
+                buckets: FxHashMap::default(),
+            });
+        }
         Ok(())
     }
 }
@@ -373,22 +397,19 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
 impl SegmentHistogramCollector {
     fn get_memory_consumption(&self) -> usize {
         let self_mem = std::mem::size_of::<Self>();
-        let sub_aggs_mem = self.sub_aggregations.memory_consumption();
-        let buckets_mem = self.buckets.memory_consumption();
-        self_mem + sub_aggs_mem + buckets_mem
+        let buckets_mem = self.parent_buckets.len() * std::mem::size_of::<HistogramBuckets>();
+        self_mem + buckets_mem
     }
     /// Converts the collector result into a intermediate bucket result.
-    pub fn into_intermediate_bucket_result(
-        self,
+    fn add_intermediate_bucket_result(
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
+        histogram: HistogramBuckets,
     ) -> crate::Result<IntermediateBucketResult> {
-        let mut buckets = Vec::with_capacity(self.buckets.len());
+        let mut buckets = Vec::with_capacity(histogram.buckets.len());
 
-        for (bucket_pos, bucket) in self.buckets {
-            let bucket_res = bucket.into_intermediate_bucket_entry(
-                self.sub_aggregations.get(&bucket_pos).cloned(),
-                agg_data,
-            );
+        for bucket in histogram.buckets.into_values() {
+            let bucket_res = bucket.into_intermediate_bucket_entry(&mut self.sub_agg, agg_data);
 
             buckets.push(bucket_res?);
         }
@@ -408,7 +429,7 @@ impl SegmentHistogramCollector {
         agg_data: &mut AggregationsSegmentCtx,
         node: &AggRefNode,
     ) -> crate::Result<Self> {
-        let blueprint = if !node.children.is_empty() {
+        let sub_agg = if !node.children.is_empty() {
             Some(build_segment_agg_collectors(agg_data, &node.children)?)
         } else {
             None
@@ -423,13 +444,13 @@ impl SegmentHistogramCollector {
             max: f64::MAX,
         });
         req_data.offset = req_data.req.offset.unwrap_or(0.0);
-
-        req_data.sub_aggregation_blueprint = blueprint;
+        let sub_agg = sub_agg.map(CachedSubAggs::new);
 
         Ok(Self {
-            buckets: Default::default(),
-            sub_aggregations: Default::default(),
+            parent_buckets: Default::default(),
+            sub_agg,
             accessor_idx: node.idx_in_req_data,
+            bucket_id_provider: BucketIdProvider::default(),
         })
     }
 }

src/aggregation/bucket/range.rs

@@ -1,18 +1,22 @@
 use std::fmt::Debug;
 use std::ops::Range;
 
-use columnar::{Column, ColumnBlockAccessor, ColumnType};
+use columnar::{Column, ColumnType};
 use rustc_hash::FxHashMap;
 use serde::{Deserialize, Serialize};
 
 use crate::aggregation::agg_data::{
     build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
 };
+use crate::aggregation::agg_limits::AggregationLimitsGuard;
+use crate::aggregation::cached_sub_aggs::{
+    CachedSubAggs, HighCardSubAggCache, LowCardCachedSubAggs, LowCardSubAggCache, SubAggCache,
+};
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
     IntermediateRangeBucketEntry, IntermediateRangeBucketResult,
 };
-use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
+use crate::aggregation::segment_agg_result::{BucketIdProvider, SegmentAggregationCollector};
 use crate::aggregation::*;
 use crate::TantivyError;
 
@@ -23,12 +27,12 @@ pub struct RangeAggReqData {
     pub accessor: Column<u64>,
     /// The type of the fast field.
     pub field_type: ColumnType,
-    /// The column block accessor to access the fast field values.
-    pub column_block_accessor: ColumnBlockAccessor<u64>,
     /// The range aggregation request.
     pub req: RangeAggregation,
     /// The name of the aggregation.
     pub name: String,
+    /// Whether this is a top-level aggregation.
+    pub is_top_level: bool,
 }
 
 impl RangeAggReqData {
@@ -151,19 +155,47 @@ pub(crate) struct SegmentRangeAndBucketEntry {
 
 /// The collector puts values from the fast field into the correct buckets and does a conversion to
 /// the correct datatype.
-#[derive(Clone, Debug)]
-pub struct SegmentRangeCollector {
+pub struct SegmentRangeCollector<C: SubAggCache> {
     /// The buckets containing the aggregation data.
-    buckets: Vec<SegmentRangeAndBucketEntry>,
+    /// One for each ParentBucketId
+    parent_buckets: Vec<Vec<SegmentRangeAndBucketEntry>>,
     column_type: ColumnType,
     pub(crate) accessor_idx: usize,
+    sub_agg: Option<CachedSubAggs<C>>,
+    /// Here things get a bit weird. We need to assign unique bucket ids across all
+    /// parent buckets. So we keep track of the next available bucket id here.
+    /// This allows a kind of flattening of the bucket ids across all parent buckets.
+    /// E.g. in nested aggregations:
+    /// Term Agg -> Range aggregation -> Stats aggregation
+    /// E.g. the Term Agg creates 3 buckets ["INFO", "ERROR", "WARN"], each of these has a Range
+    /// aggregation with 4 buckets. The Range aggregation will create buckets with ids:
+    /// - INFO: 0,1,2,3
+    /// - ERROR: 4,5,6,7
+    /// - WARN: 8,9,10,11
+    ///
+    /// This allows the Stats aggregation to have unique bucket ids to refer to.
+    bucket_id_provider: BucketIdProvider,
+    limits: AggregationLimitsGuard,
 }
 
+impl<C: SubAggCache> Debug for SegmentRangeCollector<C> {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        f.debug_struct("SegmentRangeCollector")
+            .field("parent_buckets_len", &self.parent_buckets.len())
+            .field("column_type", &self.column_type)
+            .field("accessor_idx", &self.accessor_idx)
+            .field("has_sub_agg", &self.sub_agg.is_some())
+            .finish()
+    }
+}
+
+/// TODO: Bad naming, there's also SegmentRangeAndBucketEntry
 #[derive(Clone)]
 pub(crate) struct SegmentRangeBucketEntry {
     pub key: Key,
     pub doc_count: u64,
-    pub sub_aggregation: Option<Box<dyn SegmentAggregationCollector>>,
+    // pub sub_aggregation: Option<Box<dyn SegmentAggregationCollector>>,
+    pub bucket_id: BucketId,
     /// The from range of the bucket. Equals `f64::MIN` when `None`.
     pub from: Option<f64>,
     /// The to range of the bucket. Equals `f64::MAX` when `None`. Open interval, `to` is not
@@ -184,48 +216,50 @@ impl Debug for SegmentRangeBucketEntry {
 impl SegmentRangeBucketEntry {
     pub(crate) fn into_intermediate_bucket_entry(
         self,
-        agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<IntermediateRangeBucketEntry> {
-        let mut sub_aggregation_res = IntermediateAggregationResults::default();
-        if let Some(sub_aggregation) = self.sub_aggregation {
-            sub_aggregation
-                .add_intermediate_aggregation_result(agg_data, &mut sub_aggregation_res)?
-        } else {
-            Default::default()
-        };
+        let sub_aggregation = IntermediateAggregationResults::default();
 
         Ok(IntermediateRangeBucketEntry {
             key: self.key.into(),
             doc_count: self.doc_count,
-            sub_aggregation: sub_aggregation_res,
+            sub_aggregation_res: sub_aggregation,
             from: self.from,
             to: self.to,
         })
     }
 }
 
-impl SegmentAggregationCollector for SegmentRangeCollector {
+impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
         let field_type = self.column_type;
         let name = agg_data
             .get_range_req_data(self.accessor_idx)
             .name
             .to_string();
 
-        let buckets: FxHashMap<SerializedKey, IntermediateRangeBucketEntry> = self
-            .buckets
+        let buckets = std::mem::take(&mut self.parent_buckets[parent_bucket_id as usize]);
+
+        let buckets: FxHashMap<SerializedKey, IntermediateRangeBucketEntry> = buckets
             .into_iter()
-            .map(move |range_bucket| {
-                Ok((
-                    range_to_string(&range_bucket.range, &field_type)?,
-                    range_bucket
-                        .bucket
-                        .into_intermediate_bucket_entry(agg_data)?,
-                ))
+            .map(|range_bucket| {
+                let bucket_id = range_bucket.bucket.bucket_id;
+                let mut agg = range_bucket.bucket.into_intermediate_bucket_entry()?;
+                if let Some(sub_aggregation) = &mut self.sub_agg {
+                    sub_aggregation
+                        .get_sub_agg_collector()
+                        .add_intermediate_aggregation_result(
+                            agg_data,
+                            &mut agg.sub_aggregation_res,
+                            bucket_id,
+                        )?;
+                }
+                Ok((range_to_string(&range_bucket.range, &field_type)?, agg))
             })
             .collect::<crate::Result<_>>()?;
 
@@ -242,73 +276,114 @@ impl SegmentAggregationCollector for SegmentRangeCollector {
     #[inline]
     fn collect(
         &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.collect_block(&[doc], agg_data)
-    }
-
-    #[inline]
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        // Take request data to avoid borrow conflicts during sub-aggregation
-        let mut req = agg_data.take_range_req_data(self.accessor_idx);
+        let req = agg_data.take_range_req_data(self.accessor_idx);
 
-        req.column_block_accessor.fetch_block(docs, &req.accessor);
+        agg_data
+            .column_block_accessor
+            .fetch_block(docs, &req.accessor);
 
-        for (doc, val) in req
+        let buckets = &mut self.parent_buckets[parent_bucket_id as usize];
+
+        for (doc, val) in agg_data
             .column_block_accessor
             .iter_docid_vals(docs, &req.accessor)
         {
-            let bucket_pos = self.get_bucket_pos(val);
-            let bucket = &mut self.buckets[bucket_pos];
+            let bucket_pos = get_bucket_pos(val, buckets);
+            let bucket = &mut buckets[bucket_pos];
             bucket.bucket.doc_count += 1;
-            if let Some(sub_agg) = bucket.bucket.sub_aggregation.as_mut() {
-                sub_agg.collect(doc, agg_data)?;
+            if let Some(sub_agg) = self.sub_agg.as_mut() {
+                sub_agg.push(bucket.bucket.bucket_id, doc);
             }
         }
 
         agg_data.put_back_range_req_data(self.accessor_idx, req);
+        if let Some(sub_agg) = self.sub_agg.as_mut() {
+            sub_agg.check_flush_local(agg_data)?;
+        }
 
         Ok(())
     }
 
     fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
-        for bucket in self.buckets.iter_mut() {
-            if let Some(sub_agg) = bucket.bucket.sub_aggregation.as_mut() {
-                sub_agg.flush(agg_data)?;
-            }
+        if let Some(sub_agg) = self.sub_agg.as_mut() {
+            sub_agg.flush(agg_data)?;
         }
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        while self.parent_buckets.len() <= max_bucket as usize {
+            let new_buckets = self.create_new_buckets(agg_data)?;
+            self.parent_buckets.push(new_buckets);
+        }
+
+        Ok(())
+    }
+}
+/// Build a concrete `SegmentRangeCollector` with either a Vec- or HashMap-backed
+/// bucket storage, depending on the column type and aggregation level.
+pub(crate) fn build_segment_range_collector(
+    agg_data: &mut AggregationsSegmentCtx,
+    node: &AggRefNode,
+) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
+    let accessor_idx = node.idx_in_req_data;
+    let req_data = agg_data.get_range_req_data(node.idx_in_req_data);
+    let field_type = req_data.field_type;
+
+    // TODO: A better metric instead of is_top_level would be the number of buckets expected.
+    // E.g. If range agg is not top level, but the parent is a bucket agg with less than 10 buckets,
+    // we can are still in low cardinality territory.
+    let is_low_card = req_data.is_top_level && req_data.req.ranges.len() <= 64;
+
+    let sub_agg = if !node.children.is_empty() {
+        Some(build_segment_agg_collectors(agg_data, &node.children)?)
+    } else {
+        None
+    };
+
+    if is_low_card {
+        Ok(Box::new(SegmentRangeCollector::<LowCardSubAggCache> {
+            sub_agg: sub_agg.map(LowCardCachedSubAggs::new),
+            column_type: field_type,
+            accessor_idx,
+            parent_buckets: Vec::new(),
+            bucket_id_provider: BucketIdProvider::default(),
+            limits: agg_data.context.limits.clone(),
+        }))
+    } else {
+        Ok(Box::new(SegmentRangeCollector::<HighCardSubAggCache> {
+            sub_agg: sub_agg.map(CachedSubAggs::new),
+            column_type: field_type,
+            accessor_idx,
+            parent_buckets: Vec::new(),
+            bucket_id_provider: BucketIdProvider::default(),
+            limits: agg_data.context.limits.clone(),
+        }))
+    }
 }
 
-impl SegmentRangeCollector {
-    pub(crate) fn from_req_and_validate(
-        req_data: &mut AggregationsSegmentCtx,
-        node: &AggRefNode,
-    ) -> crate::Result<Self> {
-        let accessor_idx = node.idx_in_req_data;
-        let (field_type, ranges) = {
-            let req_view = req_data.get_range_req_data(node.idx_in_req_data);
-            (req_view.field_type, req_view.req.ranges.clone())
-        };
-
+impl<C: SubAggCache> SegmentRangeCollector<C> {
+    pub(crate) fn create_new_buckets(
+        &mut self,
+        agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<Vec<SegmentRangeAndBucketEntry>> {
+        let field_type = self.column_type;
+        let req_data = agg_data.get_range_req_data(self.accessor_idx);
         // The range input on the request is f64.
         // We need to convert to u64 ranges, because we read the values as u64.
         // The mapping from the conversion is monotonic so ordering is preserved.
-        let sub_agg_prototype = if !node.children.is_empty() {
-            Some(build_segment_agg_collectors(req_data, &node.children)?)
-        } else {
-            None
-        };
-
-        let buckets: Vec<_> = extend_validate_ranges(&ranges, &field_type)?
+        let buckets: Vec<_> = extend_validate_ranges(&req_data.req.ranges, &field_type)?
             .iter()
             .map(|range| {
+                let bucket_id = self.bucket_id_provider.next_bucket_id();
                 let key = range
                     .key
                     .clone()
@@ -317,20 +392,20 @@ impl SegmentRangeCollector {
                 let to = if range.range.end == u64::MAX {
                     None
                 } else {
-                    Some(f64_from_fastfield_u64(range.range.end, &field_type))
+                    Some(f64_from_fastfield_u64(range.range.end, field_type))
                 };
                 let from = if range.range.start == u64::MIN {
                     None
                 } else {
-                    Some(f64_from_fastfield_u64(range.range.start, &field_type))
+                    Some(f64_from_fastfield_u64(range.range.start, field_type))
                 };
-                let sub_aggregation = sub_agg_prototype.clone();
+                // let sub_aggregation = sub_agg_prototype.clone();
 
                 Ok(SegmentRangeAndBucketEntry {
                     range: range.range.clone(),
                     bucket: SegmentRangeBucketEntry {
                         doc_count: 0,
-                        sub_aggregation,
+                        bucket_id,
                         key,
                         from,
                         to,
@@ -339,27 +414,20 @@ impl SegmentRangeCollector {
             })
             .collect::<crate::Result<_>>()?;
 
-        req_data.context.limits.add_memory_consumed(
+        self.limits.add_memory_consumed(
             buckets.len() as u64 * std::mem::size_of::<SegmentRangeAndBucketEntry>() as u64,
         )?;
-
-        Ok(SegmentRangeCollector {
-            buckets,
-            column_type: field_type,
-            accessor_idx,
-        })
-    }
-
-    #[inline]
-    fn get_bucket_pos(&self, val: u64) -> usize {
-        let pos = self
-            .buckets
-            .binary_search_by_key(&val, |probe| probe.range.start)
-            .unwrap_or_else(|pos| pos - 1);
-        debug_assert!(self.buckets[pos].range.contains(&val));
-        pos
+        Ok(buckets)
     }
 }
+#[inline]
+fn get_bucket_pos(val: u64, buckets: &[SegmentRangeAndBucketEntry]) -> usize {
+    let pos = buckets
+        .binary_search_by_key(&val, |probe| probe.range.start)
+        .unwrap_or_else(|pos| pos - 1);
+    debug_assert!(buckets[pos].range.contains(&val));
+    pos
+}
 
 /// Converts the user provided f64 range value to fast field value space.
 ///
@@ -456,7 +524,7 @@ pub(crate) fn range_to_string(
             let val = i64::from_u64(val);
             format_date(val)
         } else {
-            Ok(f64_from_fastfield_u64(val, field_type).to_string())
+            Ok(f64_from_fastfield_u64(val, *field_type).to_string())
         }
     };
 
@@ -486,7 +554,7 @@ mod tests {
     pub fn get_collector_from_ranges(
         ranges: Vec<RangeAggregationRange>,
         field_type: ColumnType,
-    ) -> SegmentRangeCollector {
+    ) -> SegmentRangeCollector<HighCardSubAggCache> {
         let req = RangeAggregation {
             field: "dummy".to_string(),
             ranges,
@@ -506,30 +574,33 @@ mod tests {
                 let to = if range.range.end == u64::MAX {
                     None
                 } else {
-                    Some(f64_from_fastfield_u64(range.range.end, &field_type))
+                    Some(f64_from_fastfield_u64(range.range.end, field_type))
                 };
                 let from = if range.range.start == u64::MIN {
                     None
                 } else {
-                    Some(f64_from_fastfield_u64(range.range.start, &field_type))
+                    Some(f64_from_fastfield_u64(range.range.start, field_type))
                 };
                 SegmentRangeAndBucketEntry {
                     range: range.range.clone(),
                     bucket: SegmentRangeBucketEntry {
                         doc_count: 0,
-                        sub_aggregation: None,
                         key,
                         from,
                         to,
+                        bucket_id: 0,
                     },
                 }
             })
             .collect();
 
         SegmentRangeCollector {
-            buckets,
+            parent_buckets: vec![buckets],
             column_type: field_type,
             accessor_idx: 0,
+            sub_agg: None,
+            bucket_id_provider: Default::default(),
+            limits: AggregationLimitsGuard::default(),
         }
     }
 
@@ -776,7 +847,7 @@ mod tests {
         let buckets = vec![(10f64..20f64).into(), (30f64..40f64).into()];
         let collector = get_collector_from_ranges(buckets, ColumnType::F64);
 
-        let buckets = collector.buckets;
+        let buckets = collector.parent_buckets[0].clone();
         assert_eq!(buckets[0].range.start, u64::MIN);
         assert_eq!(buckets[0].range.end, 10f64.to_u64());
         assert_eq!(buckets[1].range.start, 10f64.to_u64());
@@ -799,7 +870,7 @@ mod tests {
         ];
         let collector = get_collector_from_ranges(buckets, ColumnType::F64);
 
-        let buckets = collector.buckets;
+        let buckets = collector.parent_buckets[0].clone();
         assert_eq!(buckets[0].range.start, u64::MIN);
         assert_eq!(buckets[0].range.end, 10f64.to_u64());
         assert_eq!(buckets[1].range.start, 10f64.to_u64());
@@ -814,7 +885,7 @@ mod tests {
         let buckets = vec![(-10f64..-1f64).into()];
         let collector = get_collector_from_ranges(buckets, ColumnType::F64);
 
-        let buckets = collector.buckets;
+        let buckets = collector.parent_buckets[0].clone();
         assert_eq!(&buckets[0].bucket.key.to_string(), "*--10");
         assert_eq!(&buckets[buckets.len() - 1].bucket.key.to_string(), "-1-*");
     }
@@ -823,7 +894,7 @@ mod tests {
         let buckets = vec![(0f64..10f64).into()];
         let collector = get_collector_from_ranges(buckets, ColumnType::F64);
 
-        let buckets = collector.buckets;
+        let buckets = collector.parent_buckets[0].clone();
         assert_eq!(&buckets[0].bucket.key.to_string(), "*-0");
         assert_eq!(&buckets[buckets.len() - 1].bucket.key.to_string(), "10-*");
     }
@@ -832,7 +903,7 @@ mod tests {
     fn range_binary_search_test_u64() {
         let check_ranges = |ranges: Vec<RangeAggregationRange>| {
             let collector = get_collector_from_ranges(ranges, ColumnType::U64);
-            let search = |val: u64| collector.get_bucket_pos(val);
+            let search = |val: u64| get_bucket_pos(val, &collector.parent_buckets[0]);
 
             assert_eq!(search(u64::MIN), 0);
             assert_eq!(search(9), 0);
@@ -878,7 +949,7 @@ mod tests {
         let ranges = vec![(10.0..100.0).into()];
 
         let collector = get_collector_from_ranges(ranges, ColumnType::F64);
-        let search = |val: u64| collector.get_bucket_pos(val);
+        let search = |val: u64| get_bucket_pos(val, &collector.parent_buckets[0]);
 
         assert_eq!(search(u64::MIN), 0);
         assert_eq!(search(9f64.to_u64()), 0);
@@ -890,63 +961,3 @@ mod tests {
                                              // the max value
     }
 }
-
-#[cfg(all(test, feature = "unstable"))]
-mod bench {
-
-    use itertools::Itertools;
-    use rand::seq::SliceRandom;
-    use rand::thread_rng;
-
-    use super::*;
-    use crate::aggregation::bucket::range::tests::get_collector_from_ranges;
-
-    const TOTAL_DOCS: u64 = 1_000_000u64;
-    const NUM_DOCS: u64 = 50_000u64;
-
-    fn get_collector_with_buckets(num_buckets: u64, num_docs: u64) -> SegmentRangeCollector {
-        let bucket_size = num_docs / num_buckets;
-        let mut buckets: Vec<RangeAggregationRange> = vec![];
-        for i in 0..num_buckets {
-            let bucket_start = (i * bucket_size) as f64;
-            buckets.push((bucket_start..bucket_start + bucket_size as f64).into())
-        }
-
-        get_collector_from_ranges(buckets, ColumnType::U64)
-    }
-
-    fn get_rand_docs(total_docs: u64, num_docs_returned: u64) -> Vec<u64> {
-        let mut rng = thread_rng();
-
-        let all_docs = (0..total_docs - 1).collect_vec();
-        let mut vals = all_docs
-            .as_slice()
-            .choose_multiple(&mut rng, num_docs_returned as usize)
-            .cloned()
-            .collect_vec();
-        vals.sort();
-        vals
-    }
-
-    fn bench_range_binary_search(b: &mut test::Bencher, num_buckets: u64) {
-        let collector = get_collector_with_buckets(num_buckets, TOTAL_DOCS);
-        let vals = get_rand_docs(TOTAL_DOCS, NUM_DOCS);
-        b.iter(|| {
-            let mut bucket_pos = 0;
-            for val in &vals {
-                bucket_pos = collector.get_bucket_pos(*val);
-            }
-            bucket_pos
-        })
-    }
-
-    #[bench]
-    fn bench_range_100_buckets(b: &mut test::Bencher) {
-        bench_range_binary_search(b, 100)
-    }
-
-    #[bench]
-    fn bench_range_10_buckets(b: &mut test::Bencher) {
-        bench_range_binary_search(b, 10)
-    }
-}

src/aggregation/bucket/term_missing_agg.rs

@@ -5,11 +5,13 @@ use crate::aggregation::agg_data::{
     build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
 };
 use crate::aggregation::bucket::term_agg::TermsAggregation;
+use crate::aggregation::cached_sub_aggs::{CachedSubAggs, HighCardCachedSubAggs};
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
     IntermediateKey, IntermediateTermBucketEntry, IntermediateTermBucketResult,
 };
-use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
+use crate::aggregation::segment_agg_result::{BucketIdProvider, SegmentAggregationCollector};
+use crate::aggregation::BucketId;
 
 /// Special aggregation to handle missing values for term aggregations.
 /// This missing aggregation will check multiple columns for existence.
@@ -35,41 +37,55 @@ impl MissingTermAggReqData {
     }
 }
 
-/// The specialized missing term aggregation.
 #[derive(Default, Debug, Clone)]
-pub struct TermMissingAgg {
+struct MissingCount {
     missing_count: u32,
+    bucket_id: BucketId,
+}
+
+/// The specialized missing term aggregation.
+#[derive(Default, Debug)]
+pub struct TermMissingAgg {
     accessor_idx: usize,
-    sub_agg: Option<Box<dyn SegmentAggregationCollector>>,
+    sub_agg: Option<HighCardCachedSubAggs>,
+    /// Idx = parent bucket id, Value = missing count for that bucket
+    missing_count_per_bucket: Vec<MissingCount>,
+    bucket_id_provider: BucketIdProvider,
 }
 impl TermMissingAgg {
     pub(crate) fn new(
-        req_data: &mut AggregationsSegmentCtx,
+        agg_data: &mut AggregationsSegmentCtx,
         node: &AggRefNode,
     ) -> crate::Result<Self> {
         let has_sub_aggregations = !node.children.is_empty();
         let accessor_idx = node.idx_in_req_data;
         let sub_agg = if has_sub_aggregations {
-            let sub_aggregation = build_segment_agg_collectors(req_data, &node.children)?;
+            let sub_aggregation = build_segment_agg_collectors(agg_data, &node.children)?;
             Some(sub_aggregation)
         } else {
             None
         };
 
+        let sub_agg = sub_agg.map(CachedSubAggs::new);
+        let bucket_id_provider = BucketIdProvider::default();
+
         Ok(Self {
             accessor_idx,
             sub_agg,
-            ..Default::default()
+            missing_count_per_bucket: Vec::new(),
+            bucket_id_provider,
         })
     }
 }
 
 impl SegmentAggregationCollector for TermMissingAgg {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
         let req_data = agg_data.get_missing_term_req_data(self.accessor_idx);
         let term_agg = &req_data.req;
         let missing = term_agg
@@ -80,13 +96,16 @@ impl SegmentAggregationCollector for TermMissingAgg {
         let mut entries: FxHashMap<IntermediateKey, IntermediateTermBucketEntry> =
             Default::default();
 
+        let missing_count = &self.missing_count_per_bucket[parent_bucket_id as usize];
         let mut missing_entry = IntermediateTermBucketEntry {
-            doc_count: self.missing_count,
+            doc_count: missing_count.missing_count,
             sub_aggregation: Default::default(),
         };
-        if let Some(sub_agg) = self.sub_agg {
+        if let Some(sub_agg) = &mut self.sub_agg {
             let mut res = IntermediateAggregationResults::default();
-            sub_agg.add_intermediate_aggregation_result(agg_data, &mut res)?;
+            sub_agg
+                .get_sub_agg_collector()
+                .add_intermediate_aggregation_result(agg_data, &mut res, missing_count.bucket_id)?;
             missing_entry.sub_aggregation = res;
         }
         entries.insert(missing.into(), missing_entry);
@@ -109,30 +128,52 @@ impl SegmentAggregationCollector for TermMissingAgg {
 
     fn collect(
         &mut self,
-        doc: crate::DocId,
+        parent_bucket_id: BucketId,
+        docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
+        let bucket = &mut self.missing_count_per_bucket[parent_bucket_id as usize];
         let req_data = agg_data.get_missing_term_req_data(self.accessor_idx);
-        let has_value = req_data
-            .accessors
-            .iter()
-            .any(|(acc, _)| acc.index.has_value(doc));
-        if !has_value {
-            self.missing_count += 1;
-            if let Some(sub_agg) = self.sub_agg.as_mut() {
-                sub_agg.collect(doc, agg_data)?;
+
+        for doc in docs {
+            let doc = *doc;
+            let has_value = req_data
+                .accessors
+                .iter()
+                .any(|(acc, _)| acc.index.has_value(doc));
+            if !has_value {
+                bucket.missing_count += 1;
+
+                if let Some(sub_agg) = self.sub_agg.as_mut() {
+                    sub_agg.push(bucket.bucket_id, doc);
+                }
             }
         }
+
+        if let Some(sub_agg) = self.sub_agg.as_mut() {
+            sub_agg.check_flush_local(agg_data)?;
+        }
         Ok(())
     }
 
-    fn collect_block(
+    fn prepare_max_bucket(
         &mut self,
-        docs: &[crate::DocId],
-        agg_data: &mut AggregationsSegmentCtx,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        for doc in docs {
-            self.collect(*doc, agg_data)?;
+        while self.missing_count_per_bucket.len() <= max_bucket as usize {
+            let bucket_id = self.bucket_id_provider.next_bucket_id();
+            self.missing_count_per_bucket.push(MissingCount {
+                missing_count: 0,
+                bucket_id,
+            });
+        }
+        Ok(())
+    }
+
+    fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
+        if let Some(sub_agg) = self.sub_agg.as_mut() {
+            sub_agg.flush(agg_data)?;
         }
         Ok(())
     }

src/aggregation/buf_collector.rs

@@ -1,87 +0,0 @@
-use super::intermediate_agg_result::IntermediateAggregationResults;
-use super::segment_agg_result::SegmentAggregationCollector;
-use crate::aggregation::agg_data::AggregationsSegmentCtx;
-use crate::DocId;
-
-#[cfg(test)]
-pub(crate) const DOC_BLOCK_SIZE: usize = 64;
-
-#[cfg(not(test))]
-pub(crate) const DOC_BLOCK_SIZE: usize = 256;
-
-pub(crate) type DocBlock = [DocId; DOC_BLOCK_SIZE];
-
-/// BufAggregationCollector buffers documents before calling collect_block().
-#[derive(Clone)]
-pub(crate) struct BufAggregationCollector {
-    pub(crate) collector: Box<dyn SegmentAggregationCollector>,
-    staged_docs: DocBlock,
-    num_staged_docs: usize,
-}
-
-impl std::fmt::Debug for BufAggregationCollector {
-    fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
-        f.debug_struct("SegmentAggregationResultsCollector")
-            .field("staged_docs", &&self.staged_docs[..self.num_staged_docs])
-            .field("num_staged_docs", &self.num_staged_docs)
-            .finish()
-    }
-}
-
-impl BufAggregationCollector {
-    pub fn new(collector: Box<dyn SegmentAggregationCollector>) -> Self {
-        Self {
-            collector,
-            num_staged_docs: 0,
-            staged_docs: [0; DOC_BLOCK_SIZE],
-        }
-    }
-}
-
-impl SegmentAggregationCollector for BufAggregationCollector {
-    #[inline]
-    fn add_intermediate_aggregation_result(
-        self: Box<Self>,
-        agg_data: &AggregationsSegmentCtx,
-        results: &mut IntermediateAggregationResults,
-    ) -> crate::Result<()> {
-        Box::new(self.collector).add_intermediate_aggregation_result(agg_data, results)
-    }
-
-    #[inline]
-    fn collect(
-        &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.staged_docs[self.num_staged_docs] = doc;
-        self.num_staged_docs += 1;
-        if self.num_staged_docs == self.staged_docs.len() {
-            self.collector
-                .collect_block(&self.staged_docs[..self.num_staged_docs], agg_data)?;
-            self.num_staged_docs = 0;
-        }
-        Ok(())
-    }
-
-    #[inline]
-    fn collect_block(
-        &mut self,
-        docs: &[crate::DocId],
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.collector.collect_block(docs, agg_data)?;
-        Ok(())
-    }
-
-    #[inline]
-    fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
-        self.collector
-            .collect_block(&self.staged_docs[..self.num_staged_docs], agg_data)?;
-        self.num_staged_docs = 0;
-
-        self.collector.flush(agg_data)?;
-
-        Ok(())
-    }
-}

src/aggregation/cached_sub_aggs.rs

@@ -0,0 +1,245 @@
+use std::fmt::Debug;
+
+use super::segment_agg_result::SegmentAggregationCollector;
+use crate::aggregation::agg_data::AggregationsSegmentCtx;
+use crate::aggregation::bucket::MAX_NUM_TERMS_FOR_VEC;
+use crate::aggregation::BucketId;
+use crate::DocId;
+
+/// A cache for sub-aggregations, storing doc ids per bucket id.
+/// Depending on the cardinality of the parent aggregation, we use different
+/// storage strategies.
+///
+/// ## Low Cardinality
+/// Cardinality here refers to the number of unique flattened buckets that can be created
+/// by the parent aggregation.
+/// Flattened buckets are the result of combining all buckets per collector
+/// into a single list of buckets, where each bucket is identified by its BucketId.
+///
+/// ## Usage
+/// Since this is caching for sub-aggregations, it is only used by bucket
+/// aggregations.
+///
+/// TODO: consider using a more advanced data structure for high cardinality
+/// aggregations.
+/// What this datastructure does in general is to group docs by bucket id.
+#[derive(Debug)]
+pub(crate) struct CachedSubAggs<C: SubAggCache> {
+    cache: C,
+    sub_agg_collector: Box<dyn SegmentAggregationCollector>,
+    num_docs: usize,
+}
+
+pub type LowCardCachedSubAggs = CachedSubAggs<LowCardSubAggCache>;
+pub type HighCardCachedSubAggs = CachedSubAggs<HighCardSubAggCache>;
+
+const FLUSH_THRESHOLD: usize = 2048;
+
+/// A trait for caching sub-aggregation doc ids per bucket id.
+/// Different implementations can be used depending on the cardinality
+/// of the parent aggregation.
+pub trait SubAggCache: Debug {
+    fn new() -> Self;
+    fn push(&mut self, bucket_id: BucketId, doc_id: DocId);
+    fn flush_local(
+        &mut self,
+        sub_agg: &mut Box<dyn SegmentAggregationCollector>,
+        agg_data: &mut AggregationsSegmentCtx,
+        force: bool,
+    ) -> crate::Result<()>;
+}
+
+impl<Backend: SubAggCache + Debug> CachedSubAggs<Backend> {
+    pub fn new(sub_agg: Box<dyn SegmentAggregationCollector>) -> Self {
+        Self {
+            cache: Backend::new(),
+            sub_agg_collector: sub_agg,
+            num_docs: 0,
+        }
+    }
+
+    pub fn get_sub_agg_collector(&mut self) -> &mut Box<dyn SegmentAggregationCollector> {
+        &mut self.sub_agg_collector
+    }
+
+    #[inline]
+    pub fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
+        self.cache.push(bucket_id, doc_id);
+        self.num_docs += 1;
+    }
+
+    /// Check if we need to flush based on the number of documents cached.
+    /// If so, flushes the cache to the provided aggregation collector.
+    pub fn check_flush_local(
+        &mut self,
+        agg_data: &mut AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        if self.num_docs >= FLUSH_THRESHOLD {
+            self.cache
+                .flush_local(&mut self.sub_agg_collector, agg_data, false)?;
+            self.num_docs = 0;
+        }
+        Ok(())
+    }
+
+    /// Note: this _does_ flush the sub aggregations.
+    pub fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
+        if self.num_docs != 0 {
+            self.cache
+                .flush_local(&mut self.sub_agg_collector, agg_data, true)?;
+            self.num_docs = 0;
+        }
+        self.sub_agg_collector.flush(agg_data)?;
+        Ok(())
+    }
+}
+
+/// Number of partitions for high cardinality sub-aggregation cache.
+const NUM_PARTITIONS: usize = 16;
+
+#[derive(Debug)]
+pub(crate) struct HighCardSubAggCache {
+    /// This weird partitioning is used to do some cheap grouping on the bucket ids.
+    /// bucket ids are dense, e.g. when we don't detect the cardinality as low cardinality,
+    /// but there are just 16 bucket ids, each bucket id will go to its own partition.
+    ///
+    /// We want to keep this cheap, because high cardinality aggregations can have a lot of
+    /// buckets, and there may be nothing to group.
+    partitions: Box<[PartitionEntry; NUM_PARTITIONS]>,
+}
+
+impl HighCardSubAggCache {
+    #[inline]
+    fn clear(&mut self) {
+        for partition in self.partitions.iter_mut() {
+            partition.clear();
+        }
+    }
+}
+
+#[derive(Debug, Clone, Default)]
+struct PartitionEntry {
+    bucket_ids: Vec<BucketId>,
+    docs: Vec<DocId>,
+}
+
+impl PartitionEntry {
+    #[inline]
+    fn clear(&mut self) {
+        self.bucket_ids.clear();
+        self.docs.clear();
+    }
+}
+
+impl SubAggCache for HighCardSubAggCache {
+    fn new() -> Self {
+        Self {
+            partitions: Box::new(core::array::from_fn(|_| PartitionEntry::default())),
+        }
+    }
+
+    fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
+        let idx = bucket_id % NUM_PARTITIONS as u32;
+        let slot = &mut self.partitions[idx as usize];
+        slot.bucket_ids.push(bucket_id);
+        slot.docs.push(doc_id);
+    }
+
+    fn flush_local(
+        &mut self,
+        sub_agg: &mut Box<dyn SegmentAggregationCollector>,
+        agg_data: &mut AggregationsSegmentCtx,
+        _force: bool,
+    ) -> crate::Result<()> {
+        let mut max_bucket = 0u32;
+        for partition in self.partitions.iter() {
+            if let Some(&local_max) = partition.bucket_ids.iter().max() {
+                max_bucket = max_bucket.max(local_max);
+            }
+        }
+
+        sub_agg.prepare_max_bucket(max_bucket, agg_data)?;
+
+        for slot in self.partitions.iter() {
+            if !slot.bucket_ids.is_empty() {
+                // Reduce dynamic dispatch overhead by collecting a full partition in one call.
+                sub_agg.collect_multiple(&slot.bucket_ids, &slot.docs, agg_data)?;
+            }
+        }
+
+        self.clear();
+        Ok(())
+    }
+}
+
+#[derive(Debug)]
+pub(crate) struct LowCardSubAggCache {
+    /// Cache doc ids per bucket for sub-aggregations.
+    ///
+    /// The outer Vec is indexed by BucketId.
+    per_bucket_docs: Vec<Vec<DocId>>,
+}
+
+impl LowCardSubAggCache {
+    #[inline]
+    fn clear(&mut self) {
+        for v in &mut self.per_bucket_docs {
+            v.clear();
+        }
+    }
+}
+
+impl SubAggCache for LowCardSubAggCache {
+    fn new() -> Self {
+        Self {
+            per_bucket_docs: Vec::new(),
+        }
+    }
+
+    fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
+        let idx = bucket_id as usize;
+        if self.per_bucket_docs.len() <= idx {
+            self.per_bucket_docs.resize_with(idx + 1, Vec::new);
+        }
+        self.per_bucket_docs[idx].push(doc_id);
+    }
+
+    fn flush_local(
+        &mut self,
+        sub_agg: &mut Box<dyn SegmentAggregationCollector>,
+        agg_data: &mut AggregationsSegmentCtx,
+        force: bool,
+    ) -> crate::Result<()> {
+        // Pre-aggregated: call collect per bucket.
+        let max_bucket = (self.per_bucket_docs.len() as BucketId).saturating_sub(1);
+        sub_agg.prepare_max_bucket(max_bucket, agg_data)?;
+        // The threshold above which we flush buckets individually.
+        // Note: We need to make sure that we don't lock ourselves into a situation where we hit
+        // the FLUSH_THRESHOLD, but never flush any buckets. (except the final flush)
+        let mut bucket_treshold = FLUSH_THRESHOLD / (self.per_bucket_docs.len().max(1) * 2);
+        const _: () = {
+            // MAX_NUM_TERMS_FOR_VEC threshold is used for term aggregations
+            // Note: There may be other flexible values, for other aggregations, but we can use the
+            // const value here as a upper bound. (better than nothing)
+            let bucket_treshold_limit = FLUSH_THRESHOLD / (MAX_NUM_TERMS_FOR_VEC as usize * 2);
+            assert!(
+                bucket_treshold_limit > 0,
+                "Bucket threshold must be greater than 0"
+            );
+        };
+        if force {
+            bucket_treshold = 0;
+        }
+        for (bucket_id, docs) in self
+            .per_bucket_docs
+            .iter()
+            .enumerate()
+            .filter(|(_, docs)| docs.len() > bucket_treshold)
+        {
+            sub_agg.collect(bucket_id as BucketId, docs, agg_data)?;
+        }
+
+        self.clear();
+        Ok(())
+    }
+}

src/aggregation/collector.rs

@@ -1,9 +1,9 @@
 use super::agg_req::Aggregations;
 use super::agg_result::AggregationResults;
-use super::buf_collector::BufAggregationCollector;
+use super::cached_sub_aggs::LowCardCachedSubAggs;
 use super::intermediate_agg_result::IntermediateAggregationResults;
-use super::segment_agg_result::SegmentAggregationCollector;
 use super::AggContextParams;
+// group buffering strategy is chosen explicitly by callers; no need to hash-group on the fly.
 use crate::aggregation::agg_data::{
     build_aggregations_data_from_req, build_segment_agg_collectors_root, AggregationsSegmentCtx,
 };
@@ -136,7 +136,7 @@ fn merge_fruits(
 /// `AggregationSegmentCollector` does the aggregation collection on a segment.
 pub struct AggregationSegmentCollector {
     aggs_with_accessor: AggregationsSegmentCtx,
-    agg_collector: BufAggregationCollector,
+    agg_collector: LowCardCachedSubAggs,
     error: Option<TantivyError>,
 }
 
@@ -151,8 +151,11 @@ impl AggregationSegmentCollector {
     ) -> crate::Result<Self> {
         let mut agg_data =
             build_aggregations_data_from_req(agg, reader, segment_ordinal, context.clone())?;
-        let result =
-            BufAggregationCollector::new(build_segment_agg_collectors_root(&mut agg_data)?);
+        let mut result =
+            LowCardCachedSubAggs::new(build_segment_agg_collectors_root(&mut agg_data)?);
+        result
+            .get_sub_agg_collector()
+            .prepare_max_bucket(0, &agg_data)?; // prepare for bucket zero
 
         Ok(AggregationSegmentCollector {
             aggs_with_accessor: agg_data,
@@ -170,26 +173,31 @@ impl SegmentCollector for AggregationSegmentCollector {
         if self.error.is_some() {
             return;
         }
-        if let Err(err) = self
+        self.agg_collector.push(0, doc);
+        match self
             .agg_collector
-            .collect(doc, &mut self.aggs_with_accessor)
+            .check_flush_local(&mut self.aggs_with_accessor)
         {
-            self.error = Some(err);
+            Ok(_) => {}
+            Err(e) => {
+                self.error = Some(e);
+            }
         }
     }
-
-    /// The query pushes the documents to the collector via this method.
-    ///
-    /// Only valid for Collectors that ignore docs
     fn collect_block(&mut self, docs: &[DocId]) {
         if self.error.is_some() {
             return;
         }
-        if let Err(err) = self
-            .agg_collector
-            .collect_block(docs, &mut self.aggs_with_accessor)
-        {
-            self.error = Some(err);
+
+        match self.agg_collector.get_sub_agg_collector().collect(
+            0,
+            docs,
+            &mut self.aggs_with_accessor,
+        ) {
+            Ok(_) => {}
+            Err(e) => {
+                self.error = Some(e);
+            }
         }
     }
 
@@ -200,10 +208,13 @@ impl SegmentCollector for AggregationSegmentCollector {
         self.agg_collector.flush(&mut self.aggs_with_accessor)?;
 
         let mut sub_aggregation_res = IntermediateAggregationResults::default();
-        Box::new(self.agg_collector).add_intermediate_aggregation_result(
-            &self.aggs_with_accessor,
-            &mut sub_aggregation_res,
-        )?;
+        self.agg_collector
+            .get_sub_agg_collector()
+            .add_intermediate_aggregation_result(
+                &self.aggs_with_accessor,
+                &mut sub_aggregation_res,
+                0,
+            )?;
 
         Ok(sub_aggregation_res)
     }

src/aggregation/intermediate_agg_result.rs

@@ -792,7 +792,7 @@ pub struct IntermediateRangeBucketEntry {
     /// The number of documents in the bucket.
     pub doc_count: u64,
     /// The sub_aggregation in this bucket.
-    pub sub_aggregation: IntermediateAggregationResults,
+    pub sub_aggregation_res: IntermediateAggregationResults,
     /// The from range of the bucket. Equals `f64::MIN` when `None`.
     pub from: Option<f64>,
     /// The to range of the bucket. Equals `f64::MAX` when `None`.
@@ -811,7 +811,7 @@ impl IntermediateRangeBucketEntry {
             key: self.key.into(),
             doc_count: self.doc_count,
             sub_aggregation: self
-                .sub_aggregation
+                .sub_aggregation_res
                 .into_final_result_internal(req, limits)?,
             to: self.to,
             from: self.from,
@@ -857,7 +857,8 @@ impl MergeFruits for IntermediateTermBucketEntry {
 impl MergeFruits for IntermediateRangeBucketEntry {
     fn merge_fruits(&mut self, other: IntermediateRangeBucketEntry) -> crate::Result<()> {
         self.doc_count += other.doc_count;
-        self.sub_aggregation.merge_fruits(other.sub_aggregation)?;
+        self.sub_aggregation_res
+            .merge_fruits(other.sub_aggregation_res)?;
         Ok(())
     }
 }
@@ -887,7 +888,7 @@ mod tests {
                 IntermediateRangeBucketEntry {
                     key: IntermediateKey::Str(key.to_string()),
                     doc_count: *doc_count,
-                    sub_aggregation: Default::default(),
+                    sub_aggregation_res: Default::default(),
                     from: None,
                     to: None,
                 },
@@ -920,7 +921,7 @@ mod tests {
                     doc_count: *doc_count,
                     from: None,
                     to: None,
-                    sub_aggregation: get_sub_test_tree(&[(
+                    sub_aggregation_res: get_sub_test_tree(&[(
                         sub_aggregation_key.to_string(),
                         *sub_aggregation_count,
                     )]),

src/aggregation/metric/average.rs

@@ -52,10 +52,8 @@ pub struct IntermediateAverage {
 
 impl IntermediateAverage {
     /// Creates a new [`IntermediateAverage`] instance from a [`SegmentStatsCollector`].
-    pub(crate) fn from_collector(collector: SegmentStatsCollector) -> Self {
-        Self {
-            stats: collector.stats,
-        }
+    pub(crate) fn from_stats(stats: IntermediateStats) -> Self {
+        Self { stats }
     }
     /// Merges the other intermediate result into self.
     pub fn merge_fruits(&mut self, other: IntermediateAverage) {

src/aggregation/metric/cardinality.rs

@@ -2,7 +2,7 @@ use std::collections::hash_map::DefaultHasher;
 use std::hash::{BuildHasher, Hasher};
 
 use columnar::column_values::CompactSpaceU64Accessor;
-use columnar::{Column, ColumnBlockAccessor, ColumnType, Dictionary, StrColumn};
+use columnar::{Column, ColumnType, Dictionary, StrColumn};
 use common::f64_to_u64;
 use hyperloglogplus::{HyperLogLog, HyperLogLogPlus};
 use rustc_hash::FxHashSet;
@@ -106,8 +106,6 @@ pub struct CardinalityAggReqData {
     pub str_dict_column: Option<StrColumn>,
     /// The missing value normalized to the internal u64 representation of the field type.
     pub missing_value_for_accessor: Option<u64>,
-    /// The column block accessor to access the fast field values.
-    pub(crate) column_block_accessor: ColumnBlockAccessor<u64>,
     /// The name of the aggregation.
     pub name: String,
     /// The aggregation request.
@@ -135,45 +133,34 @@ impl CardinalityAggregationReq {
     }
 }
 
-#[derive(Clone, Debug, PartialEq)]
+#[derive(Clone, Debug)]
 pub(crate) struct SegmentCardinalityCollector {
-    cardinality: CardinalityCollector,
-    entries: FxHashSet<u64>,
+    buckets: Vec<SegmentCardinalityCollectorBucket>,
     accessor_idx: usize,
+    /// The column accessor to access the fast field values.
+    accessor: Column<u64>,
+    /// The column_type of the field.
+    column_type: ColumnType,
+    /// The missing value normalized to the internal u64 representation of the field type.
+    missing_value_for_accessor: Option<u64>,
 }
 
-impl SegmentCardinalityCollector {
-    pub fn from_req(column_type: ColumnType, accessor_idx: usize) -> Self {
+#[derive(Clone, Debug, PartialEq, Default)]
+pub(crate) struct SegmentCardinalityCollectorBucket {
+    cardinality: CardinalityCollector,
+    entries: FxHashSet<u64>,
+}
+impl SegmentCardinalityCollectorBucket {
+    pub fn new(column_type: ColumnType) -> Self {
         Self {
             cardinality: CardinalityCollector::new(column_type as u8),
-            entries: Default::default(),
-            accessor_idx,
+            entries: FxHashSet::default(),
         }
     }
-
-    fn fetch_block_with_field(
-        &mut self,
-        docs: &[crate::DocId],
-        agg_data: &mut CardinalityAggReqData,
-    ) {
-        if let Some(missing) = agg_data.missing_value_for_accessor {
-            agg_data.column_block_accessor.fetch_block_with_missing(
-                docs,
-                &agg_data.accessor,
-                missing,
-            );
-        } else {
-            agg_data
-                .column_block_accessor
-                .fetch_block(docs, &agg_data.accessor);
-        }
-    }
-
     fn into_intermediate_metric_result(
         mut self,
-        agg_data: &AggregationsSegmentCtx,
+        req_data: &CardinalityAggReqData,
     ) -> crate::Result<IntermediateMetricResult> {
-        let req_data = &agg_data.get_cardinality_req_data(self.accessor_idx);
         if req_data.column_type == ColumnType::Str {
             let fallback_dict = Dictionary::empty();
             let dict = req_data
@@ -194,6 +181,7 @@ impl SegmentCardinalityCollector {
                     term_ids.push(term_ord as u32);
                 }
             }
+
             term_ids.sort_unstable();
             dict.sorted_ords_to_term_cb(term_ids.iter().map(|term| *term as u64), |term| {
                 self.cardinality.sketch.insert_any(&term);
@@ -227,16 +215,49 @@ impl SegmentCardinalityCollector {
     }
 }
 
+impl SegmentCardinalityCollector {
+    pub fn from_req(
+        column_type: ColumnType,
+        accessor_idx: usize,
+        accessor: Column<u64>,
+        missing_value_for_accessor: Option<u64>,
+    ) -> Self {
+        Self {
+            buckets: vec![SegmentCardinalityCollectorBucket::new(column_type); 1],
+            column_type,
+            accessor_idx,
+            accessor,
+            missing_value_for_accessor,
+        }
+    }
+
+    fn fetch_block_with_field(
+        &mut self,
+        docs: &[crate::DocId],
+        agg_data: &mut AggregationsSegmentCtx,
+    ) {
+        agg_data.column_block_accessor.fetch_block_with_missing(
+            docs,
+            &self.accessor,
+            self.missing_value_for_accessor,
+        );
+    }
+}
+
 impl SegmentAggregationCollector for SegmentCardinalityCollector {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
         let req_data = &agg_data.get_cardinality_req_data(self.accessor_idx);
         let name = req_data.name.to_string();
+        // take the bucket in buckets and replace it with a new empty one
+        let bucket = std::mem::take(&mut self.buckets[parent_bucket_id as usize]);
 
-        let intermediate_result = self.into_intermediate_metric_result(agg_data)?;
+        let intermediate_result = bucket.into_intermediate_metric_result(req_data)?;
         results.push(
             name,
             IntermediateAggregationResult::Metric(intermediate_result),
@@ -247,27 +268,20 @@ impl SegmentAggregationCollector for SegmentCardinalityCollector {
 
     fn collect(
         &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.collect_block(&[doc], agg_data)
-    }
-
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_cardinality_req_data_mut(self.accessor_idx);
-        self.fetch_block_with_field(docs, req_data);
+        self.fetch_block_with_field(docs, agg_data);
+        let bucket = &mut self.buckets[parent_bucket_id as usize];
 
-        let col_block_accessor = &req_data.column_block_accessor;
-        if req_data.column_type == ColumnType::Str {
+        let col_block_accessor = &agg_data.column_block_accessor;
+        if self.column_type == ColumnType::Str {
             for term_ord in col_block_accessor.iter_vals() {
-                self.entries.insert(term_ord);
+                bucket.entries.insert(term_ord);
             }
-        } else if req_data.column_type == ColumnType::IpAddr {
-            let compact_space_accessor = req_data
+        } else if self.column_type == ColumnType::IpAddr {
+            let compact_space_accessor = self
                 .accessor
                 .values
                 .clone()
@@ -282,16 +296,29 @@ impl SegmentAggregationCollector for SegmentCardinalityCollector {
                 })?;
             for val in col_block_accessor.iter_vals() {
                 let val: u128 = compact_space_accessor.compact_to_u128(val as u32);
-                self.cardinality.sketch.insert_any(&val);
+                bucket.cardinality.sketch.insert_any(&val);
             }
         } else {
             for val in col_block_accessor.iter_vals() {
-                self.cardinality.sketch.insert_any(&val);
+                bucket.cardinality.sketch.insert_any(&val);
             }
         }
 
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        if max_bucket as usize >= self.buckets.len() {
+            self.buckets.resize_with(max_bucket as usize + 1, || {
+                SegmentCardinalityCollectorBucket::new(self.column_type)
+            });
+        }
+        Ok(())
+    }
 }
 
 #[derive(Clone, Debug, Serialize, Deserialize)]

src/aggregation/metric/count.rs

@@ -52,10 +52,8 @@ pub struct IntermediateCount {
 
 impl IntermediateCount {
     /// Creates a new [`IntermediateCount`] instance from a [`SegmentStatsCollector`].
-    pub(crate) fn from_collector(collector: SegmentStatsCollector) -> Self {
-        Self {
-            stats: collector.stats,
-        }
+    pub(crate) fn from_stats(stats: IntermediateStats) -> Self {
+        Self { stats }
     }
     /// Merges the other intermediate result into self.
     pub fn merge_fruits(&mut self, other: IntermediateCount) {

src/aggregation/metric/extended_stats.rs

@@ -8,10 +8,9 @@ use crate::aggregation::agg_data::AggregationsSegmentCtx;
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateMetricResult,
 };
-use crate::aggregation::metric::MetricAggReqData;
 use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
 use crate::aggregation::*;
-use crate::{DocId, TantivyError};
+use crate::TantivyError;
 
 /// A multi-value metric aggregation that computes a collection of extended statistics
 /// on numeric values that are extracted
@@ -318,51 +317,28 @@ impl IntermediateExtendedStats {
     }
 }
 
-#[derive(Clone, Debug, PartialEq)]
+#[derive(Clone, Debug)]
 pub(crate) struct SegmentExtendedStatsCollector {
+    name: String,
     missing: Option<u64>,
     field_type: ColumnType,
-    pub(crate) extended_stats: IntermediateExtendedStats,
-    pub(crate) accessor_idx: usize,
-    val_cache: Vec<u64>,
+    accessor: columnar::Column<u64>,
+    buckets: Vec<IntermediateExtendedStats>,
+    sigma: Option<f64>,
 }
 
 impl SegmentExtendedStatsCollector {
-    pub fn from_req(
-        field_type: ColumnType,
-        sigma: Option<f64>,
-        accessor_idx: usize,
-        missing: Option<f64>,
-    ) -> Self {
-        let missing = missing.and_then(|val| f64_to_fastfield_u64(val, &field_type));
+    pub fn from_req(req: &MetricAggReqData, sigma: Option<f64>) -> Self {
+        let missing = req
+            .missing
+            .and_then(|val| f64_to_fastfield_u64(val, &req.field_type));
         Self {
-            field_type,
-            extended_stats: IntermediateExtendedStats::with_sigma(sigma),
-            accessor_idx,
+            name: req.name.clone(),
+            field_type: req.field_type,
+            accessor: req.accessor.clone(),
             missing,
-            val_cache: Default::default(),
-        }
-    }
-    #[inline]
-    pub(crate) fn collect_block_with_field(
-        &mut self,
-        docs: &[DocId],
-        req_data: &mut MetricAggReqData,
-    ) {
-        if let Some(missing) = self.missing.as_ref() {
-            req_data.column_block_accessor.fetch_block_with_missing(
-                docs,
-                &req_data.accessor,
-                *missing,
-            );
-        } else {
-            req_data
-                .column_block_accessor
-                .fetch_block(docs, &req_data.accessor);
-        }
-        for val in req_data.column_block_accessor.iter_vals() {
-            let val1 = f64_from_fastfield_u64(val, &self.field_type);
-            self.extended_stats.collect(val1);
+            buckets: vec![IntermediateExtendedStats::with_sigma(sigma); 16],
+            sigma,
         }
     }
 }
@@ -370,15 +346,18 @@ impl SegmentExtendedStatsCollector {
 impl SegmentAggregationCollector for SegmentExtendedStatsCollector {
     #[inline]
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
-        let name = agg_data.get_metric_req_data(self.accessor_idx).name.clone();
+        let name = self.name.clone();
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
+        let extended_stats = std::mem::take(&mut self.buckets[parent_bucket_id as usize]);
         results.push(
             name,
             IntermediateAggregationResult::Metric(IntermediateMetricResult::ExtendedStats(
-                self.extended_stats,
+                extended_stats,
             )),
         )?;
 
@@ -388,39 +367,36 @@ impl SegmentAggregationCollector for SegmentExtendedStatsCollector {
     #[inline]
     fn collect(
         &mut self,
-        doc: crate::DocId,
+        parent_bucket_id: BucketId,
+        docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data(self.accessor_idx);
-        if let Some(missing) = self.missing {
-            let mut has_val = false;
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &self.field_type);
-                self.extended_stats.collect(val1);
-                has_val = true;
-            }
-            if !has_val {
-                self.extended_stats
-                    .collect(f64_from_fastfield_u64(missing, &self.field_type));
-            }
-        } else {
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &self.field_type);
-                self.extended_stats.collect(val1);
-            }
+        let mut extended_stats = self.buckets[parent_bucket_id as usize].clone();
+
+        agg_data
+            .column_block_accessor
+            .fetch_block_with_missing(docs, &self.accessor, self.missing);
+        for val in agg_data.column_block_accessor.iter_vals() {
+            let val1 = f64_from_fastfield_u64(val, self.field_type);
+            extended_stats.collect(val1);
         }
 
+        // store back
+        self.buckets[parent_bucket_id as usize] = extended_stats;
+
         Ok(())
     }
 
-    #[inline]
-    fn collect_block(
+    fn prepare_max_bucket(
         &mut self,
-        docs: &[crate::DocId],
-        agg_data: &mut AggregationsSegmentCtx,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data_mut(self.accessor_idx);
-        self.collect_block_with_field(docs, req_data);
+        if self.buckets.len() <= max_bucket as usize {
+            self.buckets.resize_with(max_bucket as usize + 1, || {
+                IntermediateExtendedStats::with_sigma(self.sigma)
+            });
+        }
         Ok(())
     }
 }

src/aggregation/metric/max.rs

@@ -52,10 +52,8 @@ pub struct IntermediateMax {
 
 impl IntermediateMax {
     /// Creates a new [`IntermediateMax`] instance from a [`SegmentStatsCollector`].
-    pub(crate) fn from_collector(collector: SegmentStatsCollector) -> Self {
-        Self {
-            stats: collector.stats,
-        }
+    pub(crate) fn from_stats(stats: IntermediateStats) -> Self {
+        Self { stats }
     }
     /// Merges the other intermediate result into self.
     pub fn merge_fruits(&mut self, other: IntermediateMax) {

src/aggregation/metric/min.rs

@@ -52,10 +52,8 @@ pub struct IntermediateMin {
 
 impl IntermediateMin {
     /// Creates a new [`IntermediateMin`] instance from a [`SegmentStatsCollector`].
-    pub(crate) fn from_collector(collector: SegmentStatsCollector) -> Self {
-        Self {
-            stats: collector.stats,
-        }
+    pub(crate) fn from_stats(stats: IntermediateStats) -> Self {
+        Self { stats }
     }
     /// Merges the other intermediate result into self.
     pub fn merge_fruits(&mut self, other: IntermediateMin) {

src/aggregation/metric/mod.rs

@@ -31,7 +31,7 @@ use std::collections::HashMap;
 
 pub use average::*;
 pub use cardinality::*;
-use columnar::{Column, ColumnBlockAccessor, ColumnType};
+use columnar::{Column, ColumnType};
 pub use count::*;
 pub use extended_stats::*;
 pub use max::*;
@@ -55,8 +55,6 @@ pub struct MetricAggReqData {
     pub field_type: ColumnType,
     /// The missing value normalized to the internal u64 representation of the field type.
     pub missing_u64: Option<u64>,
-    /// The column block accessor to access the fast field values.
-    pub column_block_accessor: ColumnBlockAccessor<u64>,
     /// The column accessor to access the fast field values.
     pub accessor: Column<u64>,
     /// Used when converting to intermediate result

src/aggregation/metric/percentiles.rs

@@ -7,10 +7,9 @@ use crate::aggregation::agg_data::AggregationsSegmentCtx;
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateMetricResult,
 };
-use crate::aggregation::metric::MetricAggReqData;
 use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
 use crate::aggregation::*;
-use crate::{DocId, TantivyError};
+use crate::TantivyError;
 
 /// # Percentiles
 ///
@@ -131,10 +130,16 @@ impl PercentilesAggregationReq {
     }
 }
 
-#[derive(Clone, Debug, PartialEq)]
+#[derive(Clone, Debug)]
 pub(crate) struct SegmentPercentilesCollector {
-    pub(crate) percentiles: PercentilesCollector,
+    pub(crate) buckets: Vec<PercentilesCollector>,
     pub(crate) accessor_idx: usize,
+    /// The type of the field.
+    pub field_type: ColumnType,
+    /// The missing value normalized to the internal u64 representation of the field type.
+    pub missing_u64: Option<u64>,
+    /// The column accessor to access the fast field values.
+    pub accessor: Column<u64>,
 }
 
 #[derive(Clone, Serialize, Deserialize)]
@@ -229,33 +234,18 @@ impl PercentilesCollector {
 }
 
 impl SegmentPercentilesCollector {
-    pub fn from_req_and_validate(accessor_idx: usize) -> crate::Result<Self> {
-        Ok(Self {
-            percentiles: PercentilesCollector::new(),
+    pub fn from_req_and_validate(
+        field_type: ColumnType,
+        missing_u64: Option<u64>,
+        accessor: Column<u64>,
+        accessor_idx: usize,
+    ) -> Self {
+        Self {
+            buckets: Vec::with_capacity(64),
+            field_type,
+            missing_u64,
+            accessor,
             accessor_idx,
-        })
-    }
-    #[inline]
-    pub(crate) fn collect_block_with_field(
-        &mut self,
-        docs: &[DocId],
-        req_data: &mut MetricAggReqData,
-    ) {
-        if let Some(missing) = req_data.missing_u64.as_ref() {
-            req_data.column_block_accessor.fetch_block_with_missing(
-                docs,
-                &req_data.accessor,
-                *missing,
-            );
-        } else {
-            req_data
-                .column_block_accessor
-                .fetch_block(docs, &req_data.accessor);
-        }
-
-        for val in req_data.column_block_accessor.iter_vals() {
-            let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-            self.percentiles.collect(val1);
         }
     }
 }
@@ -263,12 +253,18 @@ impl SegmentPercentilesCollector {
 impl SegmentAggregationCollector for SegmentPercentilesCollector {
     #[inline]
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
         let name = agg_data.get_metric_req_data(self.accessor_idx).name.clone();
-        let intermediate_metric_result = IntermediateMetricResult::Percentiles(self.percentiles);
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
+        // Swap collector with an empty one to avoid cloning
+        let percentiles_collector = std::mem::take(&mut self.buckets[parent_bucket_id as usize]);
+
+        let intermediate_metric_result =
+            IntermediateMetricResult::Percentiles(percentiles_collector);
 
         results.push(
             name,
@@ -281,40 +277,33 @@ impl SegmentAggregationCollector for SegmentPercentilesCollector {
     #[inline]
     fn collect(
         &mut self,
-        doc: crate::DocId,
+        parent_bucket_id: BucketId,
+        docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data(self.accessor_idx);
+        let percentiles = &mut self.buckets[parent_bucket_id as usize];
+        agg_data.column_block_accessor.fetch_block_with_missing(
+            docs,
+            &self.accessor,
+            self.missing_u64,
+        );
 
-        if let Some(missing) = req_data.missing_u64 {
-            let mut has_val = false;
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-                self.percentiles.collect(val1);
-                has_val = true;
-            }
-            if !has_val {
-                self.percentiles
-                    .collect(f64_from_fastfield_u64(missing, &req_data.field_type));
-            }
-        } else {
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-                self.percentiles.collect(val1);
-            }
+        for val in agg_data.column_block_accessor.iter_vals() {
+            let val1 = f64_from_fastfield_u64(val, self.field_type);
+            percentiles.collect(val1);
         }
 
         Ok(())
     }
 
-    #[inline]
-    fn collect_block(
+    fn prepare_max_bucket(
         &mut self,
-        docs: &[crate::DocId],
-        agg_data: &mut AggregationsSegmentCtx,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data_mut(self.accessor_idx);
-        self.collect_block_with_field(docs, req_data);
+        while self.buckets.len() <= max_bucket as usize {
+            self.buckets.push(PercentilesCollector::new());
+        }
         Ok(())
     }
 }

src/aggregation/metric/stats.rs

@@ -1,5 +1,6 @@
 use std::fmt::Debug;
 
+use columnar::{Column, ColumnType};
 use serde::{Deserialize, Serialize};
 
 use super::*;
@@ -7,10 +8,9 @@ use crate::aggregation::agg_data::AggregationsSegmentCtx;
 use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateAggregationResults, IntermediateMetricResult,
 };
-use crate::aggregation::metric::MetricAggReqData;
 use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
 use crate::aggregation::*;
-use crate::{DocId, TantivyError};
+use crate::TantivyError;
 
 /// A multi-value metric aggregation that computes a collection of statistics on numeric values that
 /// are extracted from the aggregated documents.
@@ -83,7 +83,7 @@ impl Stats {
 
 /// Intermediate result of the stats aggregation that can be combined with other intermediate
 /// results.
-#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
+#[derive(Clone, Copy, Debug, PartialEq, Serialize, Deserialize)]
 pub struct IntermediateStats {
     /// The number of extracted values.
     pub(crate) count: u64,
@@ -187,75 +187,75 @@ pub enum StatsType {
     Percentiles,
 }
 
+fn create_collector<const TYPE_ID: u8>(
+    req: &MetricAggReqData,
+) -> Box<dyn SegmentAggregationCollector> {
+    Box::new(SegmentStatsCollector::<TYPE_ID> {
+        name: req.name.clone(),
+        collecting_for: req.collecting_for,
+        is_number_or_date_type: req.is_number_or_date_type,
+        missing_u64: req.missing_u64,
+        accessor: req.accessor.clone(),
+        buckets: vec![IntermediateStats::default()],
+    })
+}
+
+/// Build a concrete `SegmentStatsCollector` depending on the column type.
+pub(crate) fn build_segment_stats_collector(
+    req: &MetricAggReqData,
+) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
+    match req.field_type {
+        ColumnType::I64 => Ok(create_collector::<{ ColumnType::I64 as u8 }>(req)),
+        ColumnType::U64 => Ok(create_collector::<{ ColumnType::U64 as u8 }>(req)),
+        ColumnType::F64 => Ok(create_collector::<{ ColumnType::F64 as u8 }>(req)),
+        ColumnType::Bool => Ok(create_collector::<{ ColumnType::Bool as u8 }>(req)),
+        ColumnType::DateTime => Ok(create_collector::<{ ColumnType::DateTime as u8 }>(req)),
+        ColumnType::Bytes => Ok(create_collector::<{ ColumnType::Bytes as u8 }>(req)),
+        ColumnType::Str => Ok(create_collector::<{ ColumnType::Str as u8 }>(req)),
+        ColumnType::IpAddr => Ok(create_collector::<{ ColumnType::IpAddr as u8 }>(req)),
+    }
+}
+
+#[repr(C)]
 #[derive(Clone, Debug)]
-pub(crate) struct SegmentStatsCollector {
-    pub(crate) stats: IntermediateStats,
-    pub(crate) accessor_idx: usize,
+pub(crate) struct SegmentStatsCollector<const COLUMN_TYPE_ID: u8> {
+    pub(crate) missing_u64: Option<u64>,
+    pub(crate) accessor: Column<u64>,
+    pub(crate) is_number_or_date_type: bool,
+    pub(crate) buckets: Vec<IntermediateStats>,
+    pub(crate) name: String,
+    pub(crate) collecting_for: StatsType,
 }
 
-impl SegmentStatsCollector {
-    pub fn from_req(accessor_idx: usize) -> Self {
-        Self {
-            stats: IntermediateStats::default(),
-            accessor_idx,
-        }
-    }
-    #[inline]
-    pub(crate) fn collect_block_with_field(
-        &mut self,
-        docs: &[DocId],
-        req_data: &mut MetricAggReqData,
-    ) {
-        if let Some(missing) = req_data.missing_u64.as_ref() {
-            req_data.column_block_accessor.fetch_block_with_missing(
-                docs,
-                &req_data.accessor,
-                *missing,
-            );
-        } else {
-            req_data
-                .column_block_accessor
-                .fetch_block(docs, &req_data.accessor);
-        }
-        if req_data.is_number_or_date_type {
-            for val in req_data.column_block_accessor.iter_vals() {
-                let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-                self.stats.collect(val1);
-            }
-        } else {
-            for _val in req_data.column_block_accessor.iter_vals() {
-                // we ignore the value and simply record that we got something
-                self.stats.collect(0.0);
-            }
-        }
-    }
-}
-
-impl SegmentAggregationCollector for SegmentStatsCollector {
+impl<const COLUMN_TYPE_ID: u8> SegmentAggregationCollector
+    for SegmentStatsCollector<COLUMN_TYPE_ID>
+{
     #[inline]
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
-        let req = agg_data.get_metric_req_data(self.accessor_idx);
-        let name = req.name.clone();
+        let name = self.name.clone();
 
-        let intermediate_metric_result = match req.collecting_for {
+        self.prepare_max_bucket(parent_bucket_id, agg_data)?;
+        let stats = self.buckets[parent_bucket_id as usize];
+        let intermediate_metric_result = match self.collecting_for {
             StatsType::Average => {
-                IntermediateMetricResult::Average(IntermediateAverage::from_collector(*self))
+                IntermediateMetricResult::Average(IntermediateAverage::from_stats(stats))
             }
             StatsType::Count => {
-                IntermediateMetricResult::Count(IntermediateCount::from_collector(*self))
+                IntermediateMetricResult::Count(IntermediateCount::from_stats(stats))
             }
-            StatsType::Max => IntermediateMetricResult::Max(IntermediateMax::from_collector(*self)),
-            StatsType::Min => IntermediateMetricResult::Min(IntermediateMin::from_collector(*self)),
-            StatsType::Stats => IntermediateMetricResult::Stats(self.stats),
-            StatsType::Sum => IntermediateMetricResult::Sum(IntermediateSum::from_collector(*self)),
+            StatsType::Max => IntermediateMetricResult::Max(IntermediateMax::from_stats(stats)),
+            StatsType::Min => IntermediateMetricResult::Min(IntermediateMin::from_stats(stats)),
+            StatsType::Stats => IntermediateMetricResult::Stats(stats),
+            StatsType::Sum => IntermediateMetricResult::Sum(IntermediateSum::from_stats(stats)),
             _ => {
                 return Err(TantivyError::InvalidArgument(format!(
                     "Unsupported stats type for stats aggregation: {:?}",
-                    req.collecting_for
+                    self.collecting_for
                 )))
             }
         };
@@ -271,41 +271,67 @@ impl SegmentAggregationCollector for SegmentStatsCollector {
     #[inline]
     fn collect(
         &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data(self.accessor_idx);
-        if let Some(missing) = req_data.missing_u64 {
-            let mut has_val = false;
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-                self.stats.collect(val1);
-                has_val = true;
-            }
-            if !has_val {
-                self.stats
-                    .collect(f64_from_fastfield_u64(missing, &req_data.field_type));
-            }
-        } else {
-            for val in req_data.accessor.values_for_doc(doc) {
-                let val1 = f64_from_fastfield_u64(val, &req_data.field_type);
-                self.stats.collect(val1);
-            }
-        }
-
-        Ok(())
-    }
-
-    #[inline]
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
-        let req_data = agg_data.get_metric_req_data_mut(self.accessor_idx);
-        self.collect_block_with_field(docs, req_data);
+        // TODO: remove once we fetch all values for all bucket ids in one go
+        if docs.len() == 1 && self.missing_u64.is_none() {
+            collect_stats::<COLUMN_TYPE_ID>(
+                &mut self.buckets[parent_bucket_id as usize],
+                self.accessor.values_for_doc(docs[0]),
+                self.is_number_or_date_type,
+            )?;
+
+            return Ok(());
+        }
+        agg_data.column_block_accessor.fetch_block_with_missing(
+            docs,
+            &self.accessor,
+            self.missing_u64,
+        );
+        collect_stats::<COLUMN_TYPE_ID>(
+            &mut self.buckets[parent_bucket_id as usize],
+            agg_data.column_block_accessor.iter_vals(),
+            self.is_number_or_date_type,
+        )?;
+
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        let required_buckets = (max_bucket as usize) + 1;
+        if self.buckets.len() < required_buckets {
+            self.buckets
+                .resize_with(required_buckets, IntermediateStats::default);
+        }
+        Ok(())
+    }
+}
+
+#[inline]
+fn collect_stats<const COLUMN_TYPE_ID: u8>(
+    stats: &mut IntermediateStats,
+    vals: impl Iterator<Item = u64>,
+    is_number_or_date_type: bool,
+) -> crate::Result<()> {
+    if is_number_or_date_type {
+        for val in vals {
+            let val1 = convert_to_f64::<COLUMN_TYPE_ID>(val);
+            stats.collect(val1);
+        }
+    } else {
+        for _val in vals {
+            // we ignore the value and simply record that we got something
+            stats.collect(0.0);
+        }
+    }
+
+    Ok(())
 }
 
 #[cfg(test)]

src/aggregation/metric/sum.rs

@@ -52,10 +52,8 @@ pub struct IntermediateSum {
 
 impl IntermediateSum {
     /// Creates a new [`IntermediateSum`] instance from a [`SegmentStatsCollector`].
-    pub(crate) fn from_collector(collector: SegmentStatsCollector) -> Self {
-        Self {
-            stats: collector.stats,
-        }
+    pub(crate) fn from_stats(stats: IntermediateStats) -> Self {
+        Self { stats }
     }
     /// Merges the other intermediate result into self.
     pub fn merge_fruits(&mut self, other: IntermediateSum) {

src/aggregation/metric/top_hits.rs

@@ -15,12 +15,11 @@ use crate::aggregation::intermediate_agg_result::{
     IntermediateAggregationResult, IntermediateMetricResult,
 };
 use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
-use crate::aggregation::AggregationError;
+use crate::aggregation::{AggregationError, BucketId};
 use crate::collector::sort_key::ReverseComparator;
 use crate::collector::TopNComputer;
 use crate::schema::OwnedValue;
 use crate::{DocAddress, DocId, SegmentOrdinal};
-// duplicate import removed; already imported above
 
 /// Contains all information required by the TopHitsSegmentCollector to perform the
 /// top_hits aggregation on a segment.
@@ -472,7 +471,10 @@ impl TopHitsTopNComputer {
     /// Create a new TopHitsCollector
     pub fn new(req: &TopHitsAggregationReq) -> Self {
         Self {
-            top_n: TopNComputer::new(req.size + req.from.unwrap_or(0)),
+            top_n: TopNComputer::new_with_comparator(
+                req.size + req.from.unwrap_or(0),
+                ReverseComparator,
+            ),
             req: req.clone(),
         }
     }
@@ -518,7 +520,8 @@ impl TopHitsTopNComputer {
 pub(crate) struct TopHitsSegmentCollector {
     segment_ordinal: SegmentOrdinal,
     accessor_idx: usize,
-    top_n: TopNComputer<Vec<DocValueAndOrder>, DocAddress, ReverseComparator>,
+    buckets: Vec<TopNComputer<Vec<DocValueAndOrder>, DocAddress, ReverseComparator>>,
+    num_hits: usize,
 }
 
 impl TopHitsSegmentCollector {
@@ -527,19 +530,29 @@ impl TopHitsSegmentCollector {
         accessor_idx: usize,
         segment_ordinal: SegmentOrdinal,
     ) -> Self {
+        let num_hits = req.size + req.from.unwrap_or(0);
         Self {
-            top_n: TopNComputer::new(req.size + req.from.unwrap_or(0)),
+            num_hits,
             segment_ordinal,
             accessor_idx,
+            buckets: vec![TopNComputer::new_with_comparator(num_hits, ReverseComparator); 1],
         }
     }
-    fn into_top_hits_collector(
-        self,
+    fn get_top_hits_computer(
+        &mut self,
+        parent_bucket_id: BucketId,
         value_accessors: &HashMap<String, Vec<DynamicColumn>>,
         req: &TopHitsAggregationReq,
     ) -> TopHitsTopNComputer {
+        if parent_bucket_id as usize >= self.buckets.len() {
+            return TopHitsTopNComputer::new(req);
+        }
+        let top_n = std::mem::replace(
+            &mut self.buckets[parent_bucket_id as usize],
+            TopNComputer::new(0),
+        );
         let mut top_hits_computer = TopHitsTopNComputer::new(req);
-        let top_results = self.top_n.into_vec();
+        let top_results = top_n.into_vec();
 
         for res in top_results {
             let doc_value_fields = req.get_document_field_data(value_accessors, res.doc.doc_id);
@@ -554,54 +567,24 @@ impl TopHitsSegmentCollector {
 
         top_hits_computer
     }
-
-    /// TODO add a specialized variant for a single sort field
-    fn collect_with(
-        &mut self,
-        doc_id: crate::DocId,
-        req: &TopHitsAggregationReq,
-        accessors: &[(Column<u64>, ColumnType)],
-    ) -> crate::Result<()> {
-        let sorts: Vec<DocValueAndOrder> = req
-            .sort
-            .iter()
-            .enumerate()
-            .map(|(idx, KeyOrder { order, .. })| {
-                let order = *order;
-                let value = accessors
-                    .get(idx)
-                    .expect("could not find field in accessors")
-                    .0
-                    .values_for_doc(doc_id)
-                    .next();
-                DocValueAndOrder { value, order }
-            })
-            .collect();
-
-        self.top_n.push(
-            sorts,
-            DocAddress {
-                segment_ord: self.segment_ordinal,
-                doc_id,
-            },
-        );
-        Ok(())
-    }
 }
 
 impl SegmentAggregationCollector for TopHitsSegmentCollector {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut crate::aggregation::intermediate_agg_result::IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
         let req_data = agg_data.get_top_hits_req_data(self.accessor_idx);
 
         let value_accessors = &req_data.value_accessors;
 
-        let intermediate_result = IntermediateMetricResult::TopHits(
-            self.into_top_hits_collector(value_accessors, &req_data.req),
-        );
+        let intermediate_result = IntermediateMetricResult::TopHits(self.get_top_hits_computer(
+            parent_bucket_id,
+            value_accessors,
+            &req_data.req,
+        ));
         results.push(
             req_data.name.to_string(),
             IntermediateAggregationResult::Metric(intermediate_result),
@@ -611,26 +594,56 @@ impl SegmentAggregationCollector for TopHitsSegmentCollector {
     /// TODO: Consider a caching layer to reduce the call overhead
     fn collect(
         &mut self,
-        doc_id: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        let req_data = agg_data.get_top_hits_req_data(self.accessor_idx);
-        self.collect_with(doc_id, &req_data.req, &req_data.accessors)?;
-        Ok(())
-    }
-
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
+        let top_n = &mut self.buckets[parent_bucket_id as usize];
         let req_data = agg_data.get_top_hits_req_data(self.accessor_idx);
-        // TODO: Consider getting fields with the column block accessor.
-        for doc in docs {
-            self.collect_with(*doc, &req_data.req, &req_data.accessors)?;
+        let req = &req_data.req;
+        let accessors = &req_data.accessors;
+        for &doc_id in docs {
+            // TODO: this is terrible, a new vec is allocated for every doc
+            // We can fetch blocks instead
+            // We don't need to store the order for every value
+            let sorts: Vec<DocValueAndOrder> = req
+                .sort
+                .iter()
+                .enumerate()
+                .map(|(idx, KeyOrder { order, .. })| {
+                    let order = *order;
+                    let value = accessors
+                        .get(idx)
+                        .expect("could not find field in accessors")
+                        .0
+                        .values_for_doc(doc_id)
+                        .next();
+                    DocValueAndOrder { value, order }
+                })
+                .collect();
+
+            top_n.push(
+                sorts,
+                DocAddress {
+                    segment_ord: self.segment_ordinal,
+                    doc_id,
+                },
+            );
         }
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        _agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        self.buckets.resize(
+            (max_bucket as usize) + 1,
+            TopNComputer::new_with_comparator(self.num_hits, ReverseComparator),
+        );
+        Ok(())
+    }
 }
 
 #[cfg(test)]
@@ -746,7 +759,7 @@ mod tests {
                     ],
                     "from": 0,
                 }
-        }
+            }
         }))
         .unwrap();
 
@@ -875,7 +888,7 @@ mod tests {
                         "mixed.*",
                     ],
                 }
-        }
+            }
         }))?;
 
         let collector = AggregationCollector::from_aggs(d, Default::default());

src/aggregation/mod.rs

@@ -133,7 +133,7 @@ mod agg_limits;
 pub mod agg_req;
 pub mod agg_result;
 pub mod bucket;
-mod buf_collector;
+pub(crate) mod cached_sub_aggs;
 mod collector;
 mod date;
 mod error;
@@ -162,6 +162,19 @@ use serde::{Deserialize, Deserializer, Serialize};
 
 use crate::tokenizer::TokenizerManager;
 
+/// A bucket id is a dense identifier for a bucket within an aggregation.
+/// It is used to index into a Vec that hold per-bucket data.
+///
+/// For example, in a terms aggregation, each unique term will be assigned a incremental BucketId.
+/// This BucketId will be forwarded to sub-aggregations to identify the parent bucket.
+///
+/// This allows to have a single AggregationCollector instance per aggregation,
+/// that can handle multiple buckets efficiently.
+///
+/// The API to call sub-aggregations is therefore a &[(BucketId, &[DocId])].
+/// For that we'll need a buffer. One Vec per bucket aggregation is needed.
+pub type BucketId = u32;
+
 /// Context parameters for aggregation execution
 ///
 /// This struct holds shared resources needed during aggregation execution:
@@ -335,19 +348,37 @@ impl Display for Key {
     }
 }
 
+pub(crate) fn convert_to_f64<const COLUMN_TYPE_ID: u8>(val: u64) -> f64 {
+    if COLUMN_TYPE_ID == ColumnType::U64 as u8 {
+        val as f64
+    } else if COLUMN_TYPE_ID == ColumnType::I64 as u8
+        || COLUMN_TYPE_ID == ColumnType::DateTime as u8
+    {
+        i64::from_u64(val) as f64
+    } else if COLUMN_TYPE_ID == ColumnType::F64 as u8 {
+        f64::from_u64(val)
+    } else if COLUMN_TYPE_ID == ColumnType::Bool as u8 {
+        val as f64
+    } else {
+        panic!(
+            "ColumnType ID {} cannot be converted to f64 metric",
+            COLUMN_TYPE_ID
+        )
+    }
+}
+
 /// Inverse of `to_fastfield_u64`. Used to convert to `f64` for metrics.
 ///
 /// # Panics
 /// Only `u64`, `f64`, `date`, and `i64` are supported.
-pub(crate) fn f64_from_fastfield_u64(val: u64, field_type: &ColumnType) -> f64 {
+pub(crate) fn f64_from_fastfield_u64(val: u64, field_type: ColumnType) -> f64 {
     match field_type {
-        ColumnType::U64 => val as f64,
-        ColumnType::I64 | ColumnType::DateTime => i64::from_u64(val) as f64,
-        ColumnType::F64 => f64::from_u64(val),
-        ColumnType::Bool => val as f64,
-        _ => {
-            panic!("unexpected type {field_type:?}. This should not happen")
-        }
+        ColumnType::U64 => convert_to_f64::<{ ColumnType::U64 as u8 }>(val),
+        ColumnType::I64 => convert_to_f64::<{ ColumnType::I64 as u8 }>(val),
+        ColumnType::F64 => convert_to_f64::<{ ColumnType::F64 as u8 }>(val),
+        ColumnType::Bool => convert_to_f64::<{ ColumnType::Bool as u8 }>(val),
+        ColumnType::DateTime => convert_to_f64::<{ ColumnType::DateTime as u8 }>(val),
+        _ => panic!("unexpected type {field_type:?}. This should not happen"),
     }
 }
 

src/aggregation/segment_agg_result.rs

@@ -8,25 +8,67 @@ use std::fmt::Debug;
 pub(crate) use super::agg_limits::AggregationLimitsGuard;
 use super::intermediate_agg_result::IntermediateAggregationResults;
 use crate::aggregation::agg_data::AggregationsSegmentCtx;
+use crate::aggregation::BucketId;
+
+/// Monotonically increasing provider of BucketIds.
+#[derive(Debug, Clone, Default)]
+pub struct BucketIdProvider(u32);
+impl BucketIdProvider {
+    /// Get the next BucketId.
+    pub fn next_bucket_id(&mut self) -> BucketId {
+        let bucket_id = self.0;
+        self.0 += 1;
+        bucket_id
+    }
+}
 
 /// A SegmentAggregationCollector is used to collect aggregation results.
-pub trait SegmentAggregationCollector: CollectorClone + Debug {
+pub trait SegmentAggregationCollector: Debug {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()>;
 
+    /// Note: The caller needs to call `prepare_max_bucket` before calling `collect`.
     fn collect(
         &mut self,
-        doc: crate::DocId,
+        parent_bucket_id: BucketId,
+        docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()>;
 
-    fn collect_block(
+    /// Collect docs for multiple buckets in one call.
+    /// Minimizes dynamic dispatch overhead when collecting many buckets.
+    ///
+    /// Note: The caller needs to call `prepare_max_bucket` before calling `collect`.
+    fn collect_multiple(
         &mut self,
+        bucket_ids: &[BucketId],
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        debug_assert_eq!(bucket_ids.len(), docs.len());
+        let mut start = 0;
+        while start < bucket_ids.len() {
+            let bucket_id = bucket_ids[start];
+            let mut end = start + 1;
+            while end < bucket_ids.len() && bucket_ids[end] == bucket_id {
+                end += 1;
+            }
+            self.collect(bucket_id, &docs[start..end], agg_data)?;
+            start = end;
+        }
+        Ok(())
+    }
+
+    /// Prepare the collector for collecting up to BucketId `max_bucket`.
+    /// This is useful so we can split allocation ahead of time of collecting.
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        agg_data: &AggregationsSegmentCtx,
     ) -> crate::Result<()>;
 
     /// Finalize method. Some Aggregator collect blocks of docs before calling `collect_block`.
@@ -36,26 +78,7 @@ pub trait SegmentAggregationCollector: CollectorClone + Debug {
     }
 }
 
-/// A helper trait to enable cloning of Box<dyn SegmentAggregationCollector>
-pub trait CollectorClone {
-    fn clone_box(&self) -> Box<dyn SegmentAggregationCollector>;
-}
-
-impl<T> CollectorClone for T
-where T: 'static + SegmentAggregationCollector + Clone
-{
-    fn clone_box(&self) -> Box<dyn SegmentAggregationCollector> {
-        Box::new(self.clone())
-    }
-}
-
-impl Clone for Box<dyn SegmentAggregationCollector> {
-    fn clone(&self) -> Box<dyn SegmentAggregationCollector> {
-        self.clone_box()
-    }
-}
-
-#[derive(Clone, Default)]
+#[derive(Default)]
 /// The GenericSegmentAggregationResultsCollector is the generic version of the collector, which
 /// can handle arbitrary complexity of  sub-aggregations. Ideally we never have to pick this one
 /// and can provide specialized versions instead, that remove some of its overhead.
@@ -73,12 +96,13 @@ impl Debug for GenericSegmentAggregationResultsCollector {
 
 impl SegmentAggregationCollector for GenericSegmentAggregationResultsCollector {
     fn add_intermediate_aggregation_result(
-        self: Box<Self>,
+        &mut self,
         agg_data: &AggregationsSegmentCtx,
         results: &mut IntermediateAggregationResults,
+        parent_bucket_id: BucketId,
     ) -> crate::Result<()> {
-        for agg in self.aggs {
-            agg.add_intermediate_aggregation_result(agg_data, results)?;
+        for agg in &mut self.aggs {
+            agg.add_intermediate_aggregation_result(agg_data, results, parent_bucket_id)?;
         }
 
         Ok(())
@@ -86,23 +110,13 @@ impl SegmentAggregationCollector for GenericSegmentAggregationResultsCollector {
 
     fn collect(
         &mut self,
-        doc: crate::DocId,
-        agg_data: &mut AggregationsSegmentCtx,
-    ) -> crate::Result<()> {
-        self.collect_block(&[doc], agg_data)?;
-
-        Ok(())
-    }
-
-    fn collect_block(
-        &mut self,
+        parent_bucket_id: BucketId,
         docs: &[crate::DocId],
         agg_data: &mut AggregationsSegmentCtx,
     ) -> crate::Result<()> {
         for collector in &mut self.aggs {
-            collector.collect_block(docs, agg_data)?;
+            collector.collect(parent_bucket_id, docs, agg_data)?;
         }
-
         Ok(())
     }
 
@@ -112,4 +126,15 @@ impl SegmentAggregationCollector for GenericSegmentAggregationResultsCollector {
         }
         Ok(())
     }
+
+    fn prepare_max_bucket(
+        &mut self,
+        max_bucket: BucketId,
+        agg_data: &AggregationsSegmentCtx,
+    ) -> crate::Result<()> {
+        for collector in &mut self.aggs {
+            collector.prepare_max_bucket(max_bucket, agg_data)?;
+        }
+        Ok(())
+    }
 }

src/codec/mod.rs

@@ -0,0 +1,90 @@
+pub mod postings;
+pub mod standard;
+
+use std::borrow::Cow;
+use std::io;
+
+use common::OwnedBytes;
+use serde::{Deserialize, Serialize};
+pub use standard::StandardCodec;
+
+use crate::codec::postings::PostingsCodec;
+use crate::postings::Postings;
+use crate::schema::IndexRecordOption;
+
+pub trait Codec: Clone + std::fmt::Debug + Send + Sync + 'static {
+    type PostingsCodec: PostingsCodec;
+
+    const NAME: &'static str;
+
+    fn from_json_props(json_value: &serde_json::Value) -> crate::Result<Self>;
+    fn to_json_props(&self) -> serde_json::Value;
+
+    fn postings_codec(&self) -> &Self::PostingsCodec;
+}
+
+#[derive(Serialize, Deserialize, Clone, Debug)]
+pub struct CodecConfiguration {
+    name: Cow<'static, str>,
+    #[serde(default, skip_serializing_if = "serde_json::Value::is_null")]
+    props: serde_json::Value,
+}
+
+impl CodecConfiguration {
+    pub fn from_codec<C: Codec>(codec: &C) -> Self {
+        CodecConfiguration {
+            name: Cow::Borrowed(C::NAME),
+            props: codec.to_json_props(),
+        }
+    }
+
+    pub fn to_codec<C: Codec>(&self) -> crate::Result<C> {
+        if self.name != C::NAME {
+            return Err(crate::TantivyError::InvalidArgument(format!(
+                "Codec name mismatch: expected {}, got {}",
+                C::NAME,
+                self.name
+            )));
+        }
+        C::from_json_props(&self.props)
+    }
+}
+
+impl Default for CodecConfiguration {
+    fn default() -> Self {
+        CodecConfiguration::from_codec(&StandardCodec)
+    }
+}
+
+pub trait ObjectSafeCodec: 'static + Send + Sync {
+    fn load_postings_type_erased(
+        &self,
+        doc_freq: u32,
+        postings_data: OwnedBytes,
+        record_option: IndexRecordOption,
+        requested_option: IndexRecordOption,
+        positions_data: Option<OwnedBytes>,
+    ) -> io::Result<Box<dyn Postings>>;
+}
+
+impl<TCodec: Codec> ObjectSafeCodec for TCodec {
+    fn load_postings_type_erased(
+        &self,
+        doc_freq: u32,
+        postings_data: OwnedBytes,
+        record_option: IndexRecordOption,
+        requested_option: IndexRecordOption,
+        positions_data: Option<OwnedBytes>,
+    ) -> io::Result<Box<dyn Postings>> {
+        let postings: <<Self as Codec>::PostingsCodec as PostingsCodec>::Postings =
+            self.postings_codec().load_postings(
+                doc_freq,
+                postings_data,
+                record_option,
+                requested_option,
+                positions_data,
+            )?;
+        let boxed_postings: Box<dyn Postings> = Box::new(postings);
+        Ok(boxed_postings)
+    }
+}

src/codec/postings/mod.rs

@@ -0,0 +1,40 @@
+use std::io;
+
+use common::OwnedBytes;
+
+use crate::fieldnorm::FieldNormReader;
+use crate::postings::{FreqReadingOption, Postings};
+use crate::query::Bm25Weight;
+use crate::schema::IndexRecordOption;
+use crate::{DocId, Score};
+
+pub trait PostingsCodec: Send + Sync + 'static {
+    type PostingsSerializer: PostingsSerializer;
+    type Postings: Postings;
+
+    fn new_serializer(
+        &self,
+        avg_fieldnorm: Score,
+        mode: IndexRecordOption,
+        fieldnorm_reader: Option<FieldNormReader>,
+    ) -> Self::PostingsSerializer;
+
+    fn load_postings(
+        &self,
+        doc_freq: u32,
+        postings_data: OwnedBytes,
+        record_option: IndexRecordOption,
+        requested_option: IndexRecordOption,
+        positions_data: Option<OwnedBytes>,
+    ) -> io::Result<Self::Postings>;
+}
+
+pub trait PostingsSerializer {
+    fn new_term(&mut self, term_doc_freq: u32, record_term_freq: bool);
+
+    fn write_doc(&mut self, doc_id: DocId, term_freq: u32);
+
+    fn close_term(&mut self, doc_freq: u32, wrt: &mut impl io::Write) -> io::Result<()>;
+
+    fn clear(&mut self);
+}

src/codec/standard/mod.rs

@@ -0,0 +1,33 @@
+use serde::{Deserialize, Serialize};
+
+use crate::codec::standard::postings::StandardPostingsCodec;
+use crate::codec::Codec;
+
+pub mod postings;
+
+#[derive(Debug, Default, Clone, Serialize, Deserialize)]
+pub struct StandardCodec;
+
+impl Codec for StandardCodec {
+    type PostingsCodec = StandardPostingsCodec;
+
+    const NAME: &'static str = "standard";
+
+    fn from_json_props(json_value: &serde_json::Value) -> crate::Result<Self> {
+        if !json_value.is_null() {
+            return Err(crate::TantivyError::InvalidArgument(format!(
+                "Codec property for the StandardCodec are unexpected. expected null, got {}",
+                json_value.as_str().unwrap_or("null")
+            )));
+        }
+        Ok(StandardCodec)
+    }
+
+    fn to_json_props(&self) -> serde_json::Value {
+        serde_json::Value::Null
+    }
+
+    fn postings_codec(&self) -> &Self::PostingsCodec {
+        &StandardPostingsCodec
+    }
+}

src/codec/standard/postings/block.rs

@@ -0,0 +1,50 @@
+use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
+use crate::DocId;
+
+pub struct Block {
+    doc_ids: [DocId; COMPRESSION_BLOCK_SIZE],
+    term_freqs: [u32; COMPRESSION_BLOCK_SIZE],
+    len: usize,
+}
+
+impl Block {
+    pub fn new() -> Self {
+        Block {
+            doc_ids: [0u32; COMPRESSION_BLOCK_SIZE],
+            term_freqs: [0u32; COMPRESSION_BLOCK_SIZE],
+            len: 0,
+        }
+    }
+
+    pub fn doc_ids(&self) -> &[DocId] {
+        &self.doc_ids[..self.len]
+    }
+
+    pub fn term_freqs(&self) -> &[u32] {
+        &self.term_freqs[..self.len]
+    }
+
+    pub fn clear(&mut self) {
+        self.len = 0;
+    }
+
+    pub fn append_doc(&mut self, doc: DocId, term_freq: u32) {
+        let len = self.len;
+        self.doc_ids[len] = doc;
+        self.term_freqs[len] = term_freq;
+        self.len = len + 1;
+    }
+
+    pub fn is_full(&self) -> bool {
+        self.len == COMPRESSION_BLOCK_SIZE
+    }
+
+    pub fn is_empty(&self) -> bool {
+        self.len == 0
+    }
+
+    pub fn last_doc(&self) -> DocId {
+        assert_eq!(self.len, COMPRESSION_BLOCK_SIZE);
+        self.doc_ids[COMPRESSION_BLOCK_SIZE - 1]
+    }
+}

src/codec/standard/postings/block_segment_postings.rs

@@ -1,11 +1,11 @@
 use std::io;
 
-use common::VInt;
+use common::{OwnedBytes, VInt};
 
-use crate::directory::{FileSlice, OwnedBytes};
+use crate::codec::standard::postings::skip::{BlockInfo, SkipReader};
 use crate::fieldnorm::FieldNormReader;
-use crate::postings::compression::{BlockDecoder, VIntDecoder, COMPRESSION_BLOCK_SIZE};
-use crate::postings::{BlockInfo, FreqReadingOption, SkipReader};
+use crate::postings::compression::{BlockDecoder, VIntDecoder as _, COMPRESSION_BLOCK_SIZE};
+use crate::postings::FreqReadingOption;
 use crate::query::Bm25Weight;
 use crate::schema::IndexRecordOption;
 use crate::{DocId, Score, TERMINATED};
@@ -16,13 +16,8 @@ fn max_score<I: Iterator<Item = Score>>(mut it: I) -> Option<Score> {
 
 /// `BlockSegmentPostings` is a cursor iterating over blocks
 /// of documents.
-///
-/// # Warning
-///
-/// While it is useful for some very specific high-performance
-/// use cases, you should prefer using `SegmentPostings` for most usage.
 #[derive(Clone)]
-pub struct BlockSegmentPostings {
+pub(crate) struct BlockSegmentPostings {
     pub(crate) doc_decoder: BlockDecoder,
     block_loaded: bool,
     freq_decoder: BlockDecoder,
@@ -88,7 +83,7 @@ fn split_into_skips_and_postings(
 }
 
 impl BlockSegmentPostings {
-    /// Opens a `BlockSegmentPostings`.
+    /// Opens a `StandardPostingsReader`.
     /// `doc_freq` is the number of documents in the posting list.
     /// `record_option` represents the amount of data available according to the schema.
     /// `requested_option` is the amount of data requested by the user.
@@ -96,12 +91,10 @@ impl BlockSegmentPostings {
     /// term frequency blocks.
     pub(crate) fn open(
         doc_freq: u32,
-        data: FileSlice,
+        bytes: OwnedBytes,
         mut record_option: IndexRecordOption,
         requested_option: IndexRecordOption,
-        seek_doc: DocId,
-    ) -> io::Result<(BlockSegmentPostings, usize)> {
-        let bytes = data.read_bytes()?;
+    ) -> io::Result<BlockSegmentPostings> {
         let (skip_data_opt, postings_data) = split_into_skips_and_postings(doc_freq, bytes)?;
         let skip_reader = match skip_data_opt {
             Some(skip_data) => {
@@ -126,7 +119,7 @@ impl BlockSegmentPostings {
             (_, _) => FreqReadingOption::ReadFreq,
         };
 
-        let mut block_segment_postings: BlockSegmentPostings = BlockSegmentPostings {
+        let mut block_segment_postings = BlockSegmentPostings {
             doc_decoder: BlockDecoder::with_val(TERMINATED),
             block_loaded: false,
             freq_decoder: BlockDecoder::with_val(1),
@@ -136,13 +129,89 @@ impl BlockSegmentPostings {
             data: postings_data,
             skip_reader,
         };
-        let inner_pos = if seek_doc == 0 {
-            block_segment_postings.load_block();
-            0
-        } else {
-            block_segment_postings.seek(seek_doc)
-        };
-        Ok((block_segment_postings, inner_pos))
+        block_segment_postings.load_block();
+        Ok(block_segment_postings)
+    }
+}
+
+impl BlockSegmentPostings {
+    pub fn freq_reading_option(&self) -> FreqReadingOption {
+        self.freq_reading_option
+    }
+
+    /// Returns the overall number of documents in the block postings.
+    /// It does not take in account whether documents are deleted or not.
+    ///
+    /// This `doc_freq` is simply the sum of the length of all of the blocks
+    /// length, and it does not take in account deleted documents.
+    pub fn doc_freq(&self) -> u32 {
+        self.doc_freq
+    }
+
+    /// Returns the array of docs in the current block.
+    ///
+    /// Before the first call to `.advance()`, the block
+    /// returned by `.docs()` is empty.
+    #[inline]
+    pub fn docs(&self) -> &[DocId] {
+        debug_assert!(self.block_loaded);
+        self.doc_decoder.output_array()
+    }
+
+    /// Return the document at index `idx` of the block.
+    #[inline]
+    pub fn doc(&self, idx: usize) -> u32 {
+        debug_assert!(self.block_loaded);
+        self.doc_decoder.output(idx)
+    }
+
+    /// Return the array of `term freq` in the block.
+    #[inline]
+    pub fn freqs(&self) -> &[u32] {
+        debug_assert!(self.block_loaded);
+        self.freq_decoder.output_array()
+    }
+
+    /// Return the frequency at index `idx` of the block.
+    #[inline]
+    pub fn freq(&self, idx: usize) -> u32 {
+        debug_assert!(self.block_loaded);
+        self.freq_decoder.output(idx)
+    }
+
+    /// Position on a block that may contains `target_doc`.
+    ///
+    /// If all docs are smaller than target, the block loaded may be empty,
+    /// or be the last an incomplete VInt block.
+    pub fn seek(&mut self, target_doc: DocId) -> usize {
+        // Move to the block that might contain our document.
+        self.seek_block(target_doc);
+        self.load_block();
+
+        // At this point we are on the block that might contain our document.
+        let doc = self.doc_decoder.seek_within_block(target_doc);
+
+        // The last block is not full and padded with TERMINATED,
+        // so we are guaranteed to have at least one value (real or padding)
+        // that is >= target_doc.
+        debug_assert!(doc < COMPRESSION_BLOCK_SIZE);
+
+        // `doc` is now the first element >= `target_doc`.
+        // If all docs are smaller than target, the current block is incomplete and padded
+        // with TERMINATED. After the search, the cursor points to the first TERMINATED.
+        doc
+    }
+
+    pub fn position_offset(&self) -> u64 {
+        self.skip_reader.position_offset()
+    }
+
+    /// Advance to the next block.
+    pub fn advance(&mut self) {
+        self.skip_reader.advance();
+        self.block_loaded = false;
+        self.block_max_score_cache = None;
+        self.load_block();
     }
 
     /// Returns the block_max_score for the current block.
@@ -166,7 +235,7 @@ impl BlockSegmentPostings {
         }
         // this is the last block of the segment posting list.
         // If it is actually loaded, we can compute block max manually.
-        if self.block_is_loaded() {
+        if self.block_loaded {
             let docs = self.doc_decoder.output_array().iter().cloned();
             let freqs = self.freq_decoder.output_array().iter().cloned();
             let bm25_scores = docs.zip(freqs).map(|(doc, term_freq)| {
@@ -183,114 +252,25 @@ impl BlockSegmentPostings {
         // We do not cache it however, so that it gets computed when once block is loaded.
         bm25_weight.max_score()
     }
+}
 
-    pub(crate) fn freq_reading_option(&self) -> FreqReadingOption {
-        self.freq_reading_option
-    }
-
-    // Resets the block segment postings on another position
-    // in the postings file.
-    //
-    // This is useful for enumerating through a list of terms,
-    // and consuming the associated posting lists while avoiding
-    // reallocating a `BlockSegmentPostings`.
-    //
-    // # Warning
-    //
-    // This does not reset the positions list.
-    pub(crate) fn reset(&mut self, doc_freq: u32, postings_data: OwnedBytes) -> io::Result<()> {
-        let (skip_data_opt, postings_data) =
-            split_into_skips_and_postings(doc_freq, postings_data)?;
-        self.data = postings_data;
-        self.block_max_score_cache = None;
-        self.block_loaded = false;
-        if let Some(skip_data) = skip_data_opt {
-            self.skip_reader.reset(skip_data, doc_freq);
-        } else {
-            self.skip_reader.reset(OwnedBytes::empty(), doc_freq);
+impl BlockSegmentPostings {
+    /// Returns an empty segment postings object
+    pub fn empty() -> BlockSegmentPostings {
+        BlockSegmentPostings {
+            doc_decoder: BlockDecoder::with_val(TERMINATED),
+            block_loaded: true,
+            freq_decoder: BlockDecoder::with_val(1),
+            freq_reading_option: FreqReadingOption::NoFreq,
+            block_max_score_cache: None,
+            doc_freq: 0,
+            data: OwnedBytes::empty(),
+            skip_reader: SkipReader::new(OwnedBytes::empty(), 0, IndexRecordOption::Basic),
         }
-        self.doc_freq = doc_freq;
-        self.load_block();
-        Ok(())
     }
 
-    /// Returns the overall number of documents in the block postings.
-    /// It does not take in account whether documents are deleted or not.
-    ///
-    /// This `doc_freq` is simply the sum of the length of all of the blocks
-    /// length, and it does not take in account deleted documents.
-    pub fn doc_freq(&self) -> u32 {
-        self.doc_freq
-    }
-
-    /// Returns the array of docs in the current block.
-    ///
-    /// Before the first call to `.advance()`, the block
-    /// returned by `.docs()` is empty.
-    #[inline]
-    pub fn docs(&self) -> &[DocId] {
-        debug_assert!(self.block_is_loaded());
-        self.doc_decoder.output_array()
-    }
-
-    /// Return the document at index `idx` of the block.
-    #[inline]
-    pub fn doc(&self, idx: usize) -> u32 {
-        self.doc_decoder.output(idx)
-    }
-
-    /// Return the array of `term freq` in the block.
-    #[inline]
-    pub fn freqs(&self) -> &[u32] {
-        debug_assert!(self.block_is_loaded());
-        self.freq_decoder.output_array()
-    }
-
-    /// Return the frequency at index `idx` of the block.
-    #[inline]
-    pub fn freq(&self, idx: usize) -> u32 {
-        debug_assert!(self.block_is_loaded());
-        self.freq_decoder.output(idx)
-    }
-
-    /// Returns the length of the current block.
-    ///
-    /// All blocks have a length of `NUM_DOCS_PER_BLOCK`,
-    /// except the last block that may have a length
-    /// of any number between 1 and `NUM_DOCS_PER_BLOCK - 1`
-    #[inline]
-    pub fn block_len(&self) -> usize {
-        debug_assert!(self.block_is_loaded());
-        self.doc_decoder.output_len
-    }
-
-    /// Position on a block that may contains `target_doc`, and returns the
-    /// position of the first document greater than or equal to `target_doc`
-    /// within that block.
-    ///
-    /// If all docs are smaller than target, the block loaded may be empty,
-    /// or be the last an incomplete VInt block.
-    pub fn seek(&mut self, target_doc: DocId) -> usize {
-        // Move to the block that might contain our document.
-        self.seek_block(target_doc);
-        self.load_block();
-
-        // At this point we are on the block that might contain our document.
-        let doc = self.doc_decoder.seek_within_block(target_doc);
-
-        // The last block is not full and padded with TERMINATED,
-        // so we are guaranteed to have at least one value (real or padding)
-        // that is >= target_doc.
-        debug_assert!(doc < COMPRESSION_BLOCK_SIZE);
-
-        // `doc` is now the first element >= `target_doc`.
-        // If all docs are smaller than target, the current block is incomplete and padded
-        // with TERMINATED. After the search, the cursor points to the first TERMINATED.
-        doc
-    }
-
-    pub(crate) fn position_offset(&self) -> u64 {
-        self.skip_reader.position_offset()
+    pub(crate) fn skip_reader(&self) -> &SkipReader {
+        &self.skip_reader
     }
 
     /// Dangerous API! This calls seeks the next block on the skip list,
@@ -306,15 +286,11 @@ impl BlockSegmentPostings {
         }
     }
 
-    pub(crate) fn block_is_loaded(&self) -> bool {
-        self.block_loaded
-    }
-
     pub(crate) fn load_block(&mut self) {
-        let offset = self.skip_reader.byte_offset();
-        if self.block_is_loaded() {
+        if self.block_loaded {
             return;
         }
+        let offset = self.skip_reader.byte_offset();
         match self.skip_reader.block_info() {
             BlockInfo::BitPacked {
                 doc_num_bits,
@@ -359,32 +335,6 @@ impl BlockSegmentPostings {
         }
         self.block_loaded = true;
     }
-
-    /// Advance to the next block.
-    pub fn advance(&mut self) {
-        self.skip_reader.advance();
-        self.block_loaded = false;
-        self.block_max_score_cache = None;
-        self.load_block();
-    }
-
-    /// Returns an empty segment postings object
-    pub fn empty() -> BlockSegmentPostings {
-        BlockSegmentPostings {
-            doc_decoder: BlockDecoder::with_val(TERMINATED),
-            block_loaded: true,
-            freq_decoder: BlockDecoder::with_val(1),
-            freq_reading_option: FreqReadingOption::NoFreq,
-            block_max_score_cache: None,
-            doc_freq: 0,
-            data: OwnedBytes::empty(),
-            skip_reader: SkipReader::new(OwnedBytes::empty(), 0, IndexRecordOption::Basic),
-        }
-    }
-
-    pub(crate) fn skip_reader(&self) -> &SkipReader {
-        &self.skip_reader
-    }
 }
 
 #[cfg(test)]
@@ -395,10 +345,8 @@ mod tests {
     use crate::docset::{DocSet, TERMINATED};
     use crate::index::Index;
     use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
-    use crate::postings::postings::Postings;
-    use crate::postings::SegmentPostings;
+    use crate::postings::{Postings as _, SegmentPostings};
     use crate::schema::{IndexRecordOption, Schema, Term, INDEXED};
-    use crate::DocId;
 
     #[test]
     fn test_empty_segment_postings() {
@@ -461,7 +409,7 @@ mod tests {
         doc_ids.push(130);
         {
             let block_segments = build_block_postings(&doc_ids)?;
-            let mut docset = SegmentPostings::from_block_postings(block_segments, None, 0);
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
             assert_eq!(docset.seek(128), 129);
             assert_eq!(docset.doc(), 129);
             assert_eq!(docset.advance(), 130);
@@ -469,8 +417,8 @@ mod tests {
             assert_eq!(docset.advance(), TERMINATED);
         }
         {
-            let block_segments = build_block_postings(&doc_ids)?;
-            let mut docset = SegmentPostings::from_block_postings(block_segments, None, 0);
+            let block_segments = build_block_postings(&doc_ids).unwrap();
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
             assert_eq!(docset.seek(129), 129);
             assert_eq!(docset.doc(), 129);
             assert_eq!(docset.advance(), 130);
@@ -479,7 +427,7 @@ mod tests {
         }
         {
             let block_segments = build_block_postings(&doc_ids)?;
-            let mut docset = SegmentPostings::from_block_postings(block_segments, None, 0);
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
             assert_eq!(docset.doc(), 0);
             assert_eq!(docset.seek(131), TERMINATED);
             assert_eq!(docset.doc(), TERMINATED);
@@ -487,31 +435,6 @@ mod tests {
         Ok(())
     }
 
-    fn build_block_postings(docs: &[DocId]) -> crate::Result<BlockSegmentPostings> {
-        let mut schema_builder = Schema::builder();
-        let int_field = schema_builder.add_u64_field("id", INDEXED);
-        let schema = schema_builder.build();
-        let index = Index::create_in_ram(schema);
-        let mut index_writer = index.writer_for_tests()?;
-        let mut last_doc = 0u32;
-        for &doc in docs {
-            for _ in last_doc..doc {
-                index_writer.add_document(doc!(int_field=>1u64))?;
-            }
-            index_writer.add_document(doc!(int_field=>0u64))?;
-            last_doc = doc + 1;
-        }
-        index_writer.commit()?;
-        let searcher = index.reader()?.searcher();
-        let segment_reader = searcher.segment_reader(0);
-        let inverted_index = segment_reader.inverted_index(int_field).unwrap();
-        let term = Term::from_field_u64(int_field, 0u64);
-        let term_info = inverted_index.get_term_info(&term)?.unwrap();
-        let block_postings = inverted_index
-            .read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)?;
-        Ok(block_postings)
-    }
-
     #[test]
     fn test_block_segment_postings_seek() -> crate::Result<()> {
         let mut docs = vec![0];
@@ -547,7 +470,7 @@ mod tests {
         let searcher = index.reader()?.searcher();
         let segment_reader = searcher.segment_reader(0);
 
-        let mut block_segments;
+        let block_segments;
         {
             let term = Term::from_field_u64(int_field, 0u64);
             let inverted_index = segment_reader.inverted_index(int_field)?;
@@ -556,13 +479,6 @@ mod tests {
                 .read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)?;
         }
         assert_eq!(block_segments.docs(), &[0, 2, 4]);
-        {
-            let term = Term::from_field_u64(int_field, 1u64);
-            let inverted_index = segment_reader.inverted_index(int_field)?;
-            let term_info = inverted_index.get_term_info(&term)?.unwrap();
-            inverted_index.reset_block_postings_from_terminfo(&term_info, &mut block_segments)?;
-        }
-        assert_eq!(block_segments.docs(), &[1, 3, 5]);
         Ok(())
     }
 }

src/codec/standard/postings/mod.rs

@@ -0,0 +1,52 @@
+use std::io;
+
+use crate::codec::postings::PostingsCodec;
+use crate::codec::standard::postings::block_segment_postings::BlockSegmentPostings;
+use crate::codec::standard::postings::segment_postings::SegmentPostings;
+use crate::fieldnorm::FieldNormReader;
+use crate::positions::PositionReader;
+use crate::schema::IndexRecordOption;
+use crate::Score;
+
+mod block;
+mod block_segment_postings;
+mod segment_postings;
+mod skip;
+mod standard_postings_serializer;
+
+pub use standard_postings_serializer::StandardPostingsSerializer;
+
+pub struct StandardPostingsCodec;
+
+impl PostingsCodec for StandardPostingsCodec {
+    type PostingsSerializer = StandardPostingsSerializer;
+    type Postings = SegmentPostings;
+
+    fn new_serializer(
+        &self,
+        avg_fieldnorm: Score,
+        mode: IndexRecordOption,
+        fieldnorm_reader: Option<FieldNormReader>,
+    ) -> Self::PostingsSerializer {
+        StandardPostingsSerializer::new(avg_fieldnorm, mode, fieldnorm_reader)
+    }
+
+    fn load_postings(
+        &self,
+        doc_freq: u32,
+        postings_data: common::OwnedBytes,
+        record_option: IndexRecordOption,
+        requested_option: IndexRecordOption,
+        positions_data_opt: Option<common::OwnedBytes>,
+    ) -> io::Result<Self::Postings> {
+        // Rationalize record_option/requested_option.
+        let record_option = requested_option.downgrade(record_option);
+        let block_segment_postings =
+            BlockSegmentPostings::open(doc_freq, postings_data, record_option, requested_option)?;
+        let position_reader = positions_data_opt.map(PositionReader::open).transpose()?;
+        Ok(SegmentPostings::from_block_postings(
+            block_segment_postings,
+            position_reader,
+        ))
+    }
+}

src/codec/standard/postings/segment_postings.rs

@@ -1,11 +1,13 @@
-use common::HasLen;
+use common::{BitSet, HasLen};
 
+use super::BlockSegmentPostings;
 use crate::docset::DocSet;
-use crate::fastfield::AliveBitSet;
+use crate::fieldnorm::FieldNormReader;
 use crate::positions::PositionReader;
 use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
-use crate::postings::{BlockSegmentPostings, Postings};
-use crate::{DocId, TERMINATED};
+use crate::postings::{FreqReadingOption, Postings};
+use crate::query::Bm25Weight;
+use crate::{DocId, Score, TERMINATED};
 
 /// `SegmentPostings` represents the inverted list or postings associated with
 /// a term in a `Segment`.
@@ -29,31 +31,6 @@ impl SegmentPostings {
         }
     }
 
-    /// Compute the number of non-deleted documents.
-    ///
-    /// This method will clone and scan through the posting lists.
-    /// (this is a rather expensive operation).
-    pub fn doc_freq_given_deletes(&self, alive_bitset: &AliveBitSet) -> u32 {
-        let mut docset = self.clone();
-        let mut doc_freq = 0;
-        loop {
-            let doc = docset.doc();
-            if doc == TERMINATED {
-                return doc_freq;
-            }
-            if alive_bitset.is_alive(doc) {
-                doc_freq += 1u32;
-            }
-            docset.advance();
-        }
-    }
-
-    /// Returns the overall number of documents in the block postings.
-    /// It does not take in account whether documents are deleted or not.
-    pub fn doc_freq(&self) -> u32 {
-        self.block_cursor.doc_freq()
-    }
-
     /// Creates a segment postings object with the given documents
     /// and no frequency encoded.
     ///
@@ -64,30 +41,35 @@ impl SegmentPostings {
     /// buffer with the serialized data.
     #[cfg(test)]
     pub fn create_from_docs(docs: &[u32]) -> SegmentPostings {
-        use crate::directory::FileSlice;
-        use crate::postings::serializer::PostingsSerializer;
+        use common::OwnedBytes;
+
         use crate::schema::IndexRecordOption;
         let mut buffer = Vec::new();
         {
+            use crate::codec::postings::PostingsSerializer;
+
             let mut postings_serializer =
-                PostingsSerializer::new(&mut buffer, 0.0, IndexRecordOption::Basic, None);
+                crate::codec::standard::postings::StandardPostingsSerializer::new(
+                    0.0,
+                    IndexRecordOption::Basic,
+                    None,
+                );
             postings_serializer.new_term(docs.len() as u32, false);
             for &doc in docs {
                 postings_serializer.write_doc(doc, 1u32);
             }
             postings_serializer
-                .close_term(docs.len() as u32)
+                .close_term(docs.len() as u32, &mut buffer)
                 .expect("In memory Serialization should never fail.");
         }
-        let (block_segment_postings, position_within_block) = BlockSegmentPostings::open(
+        let block_segment_postings = BlockSegmentPostings::open(
             docs.len() as u32,
-            FileSlice::from(buffer),
+            OwnedBytes::new(buffer),
             IndexRecordOption::Basic,
             IndexRecordOption::Basic,
-            0u32,
         )
         .unwrap();
-        SegmentPostings::from_block_postings(block_segment_postings, None, position_within_block)
+        SegmentPostings::from_block_postings(block_segment_postings, None)
     }
 
     /// Helper functions to create `SegmentPostings` for tests.
@@ -96,9 +78,11 @@ impl SegmentPostings {
         doc_and_tfs: &[(u32, u32)],
         fieldnorms: Option<&[u32]>,
     ) -> SegmentPostings {
-        use crate::directory::FileSlice;
+        use common::OwnedBytes;
+
+        use crate::codec::postings::PostingsSerializer as _;
+        use crate::codec::standard::postings::StandardPostingsSerializer;
         use crate::fieldnorm::FieldNormReader;
-        use crate::postings::serializer::PostingsSerializer;
         use crate::schema::IndexRecordOption;
         use crate::Score;
         let mut buffer: Vec<u8> = Vec::new();
@@ -115,8 +99,7 @@ impl SegmentPostings {
                 total_num_tokens as Score / fieldnorms.len() as Score
             })
             .unwrap_or(0.0);
-        let mut postings_serializer = PostingsSerializer::new(
-            &mut buffer,
+        let mut postings_serializer = StandardPostingsSerializer::new(
             average_field_norm,
             IndexRecordOption::WithFreqs,
             fieldnorm_reader,
@@ -126,31 +109,30 @@ impl SegmentPostings {
             postings_serializer.write_doc(doc, tf);
         }
         postings_serializer
-            .close_term(doc_and_tfs.len() as u32)
+            .close_term(doc_and_tfs.len() as u32, &mut buffer)
             .unwrap();
-        let (block_segment_postings, position_within_block) = BlockSegmentPostings::open(
+        let block_segment_postings = BlockSegmentPostings::open(
             doc_and_tfs.len() as u32,
-            FileSlice::from(buffer),
+            OwnedBytes::new(buffer),
             IndexRecordOption::WithFreqs,
             IndexRecordOption::WithFreqs,
-            0u32,
         )
         .unwrap();
-        SegmentPostings::from_block_postings(block_segment_postings, None, position_within_block)
+        SegmentPostings::from_block_postings(block_segment_postings, None)
     }
 
-    /// Creates a Segment Postings from a
-    /// - `BlockSegmentPostings`,
-    /// - a position reader
-    /// - a target document to seek to
+    /// Reads a Segment postings from an &[u8]
+    ///
+    /// * `len` - number of document in the posting lists.
+    /// * `data` - data array. The complete data is not necessarily used.
+    /// * `freq_handler` - the freq handler is in charge of decoding frequencies and/or positions
     pub(crate) fn from_block_postings(
         segment_block_postings: BlockSegmentPostings,
         position_reader: Option<PositionReader>,
-        position_within_block: usize,
     ) -> SegmentPostings {
         SegmentPostings {
             block_cursor: segment_block_postings,
-            cur: position_within_block,
+            cur: 0, // cursor within the block
             position_reader,
         }
     }
@@ -161,7 +143,6 @@ impl DocSet for SegmentPostings {
     // next needs to be called a first time to point to the correct element.
     #[inline]
     fn advance(&mut self) -> DocId {
-        debug_assert!(self.block_cursor.block_is_loaded());
         if self.cur == COMPRESSION_BLOCK_SIZE - 1 {
             self.cur = 0;
             self.block_cursor.advance();
@@ -194,6 +175,19 @@ impl DocSet for SegmentPostings {
     fn size_hint(&self) -> u32 {
         self.len() as u32
     }
+
+    fn fill_bitset(&mut self, bitset: &mut BitSet) {
+        loop {
+            let docs = self.block_cursor.docs();
+            if docs.is_empty() {
+                break;
+            }
+            for &doc in docs {
+                bitset.insert(doc);
+            }
+            self.block_cursor.advance();
+        }
+    }
 }
 
 impl HasLen for SegmentPostings {
@@ -203,6 +197,15 @@ impl HasLen for SegmentPostings {
 }
 
 impl Postings for SegmentPostings {
+    fn new_term_scorer(
+        self: Box<Self>,
+        fieldnorm_reader: FieldNormReader,
+        similarity_weight: Bm25Weight,
+    ) -> Box<dyn crate::query::Scorer> {
+        use crate::query::term_query::TermScorer;
+        Box::new(TermScorer::new(*self, fieldnorm_reader, similarity_weight))
+    }
+
     /// Returns the frequency associated with the current document.
     /// If the schema is set up so that no frequency have been encoded,
     /// this method should always return 1.
@@ -224,6 +227,12 @@ impl Postings for SegmentPostings {
         self.block_cursor.freq(self.cur)
     }
 
+    /// Returns the overall number of documents in the block postings.
+    /// It does not take in account whether documents are deleted or not.
+    fn doc_freq(&self) -> u32 {
+        self.block_cursor.doc_freq()
+    }
+
     fn append_positions_with_offset(&mut self, offset: u32, output: &mut Vec<u32>) {
         let term_freq = self.term_freq();
         let prev_len = output.len();
@@ -247,6 +256,29 @@ impl Postings for SegmentPostings {
             }
         }
     }
+
+    fn supports_block_max(&self) -> bool {
+        true
+    }
+
+    fn seek_block(
+        &mut self,
+        target_doc: crate::DocId,
+        fieldnorm_reader: &FieldNormReader,
+        similarity_weight: &Bm25Weight,
+    ) -> Score {
+        self.block_cursor.seek_block(target_doc);
+        self.block_cursor
+            .block_max_score(&fieldnorm_reader, &similarity_weight)
+    }
+
+    fn last_doc_in_block(&self) -> crate::DocId {
+        self.block_cursor.skip_reader().last_doc_in_block()
+    }
+
+    fn freq_reading_option(&self) -> FreqReadingOption {
+        self.block_cursor.freq_reading_option()
+    }
 }
 
 #[cfg(test)]

src/codec/standard/postings/skip.rs

@@ -6,17 +6,21 @@ use crate::{DocId, Score, TERMINATED};
 
 // doc num bits uses the following encoding:
 // given 0b a b cdefgh
-//         |1|2|   3  |
+//         |1|2|3|  4  |
 // - 1: unused
 // - 2: is delta-1 encoded. 0 if not, 1, if yes
-// - 3: a 6 bit number in 0..=32, the actual bitwidth
+// - 3: unused
+// - 4: a 5 bit number in 0..32, the actual bitwidth. Bitpacking could in theory say this is 32
+//   (requiring a 6th bit), but the biggest doc_id we can want to encode is TERMINATED-1, which can
+//   be represented on 31b without delta encoding.
 fn encode_bitwidth(bitwidth: u8, delta_1: bool) -> u8 {
+    assert!(bitwidth < 32);
     bitwidth | ((delta_1 as u8) << 6)
 }
 
 fn decode_bitwidth(raw_bitwidth: u8) -> (u8, bool) {
     let delta_1 = ((raw_bitwidth >> 6) & 1) != 0;
-    let bitwidth = raw_bitwidth & 0x3f;
+    let bitwidth = raw_bitwidth & 0x1f;
     (bitwidth, delta_1)
 }
 
@@ -430,7 +434,7 @@ mod tests {
 
     #[test]
     fn test_encode_decode_bitwidth() {
-        for bitwidth in 0..=32 {
+        for bitwidth in 0..32 {
             for delta_1 in [false, true] {
                 assert_eq!(
                     (bitwidth, delta_1),

src/codec/standard/postings/standard_postings_serializer.rs

@@ -0,0 +1,186 @@
+use std::cmp::Ordering;
+use std::io::{self, Write as _};
+
+use common::{BinarySerializable as _, VInt};
+
+use crate::codec::postings::PostingsSerializer;
+use crate::codec::standard::postings::block::Block;
+use crate::codec::standard::postings::skip::SkipSerializer;
+use crate::fieldnorm::FieldNormReader;
+use crate::postings::compression::{BlockEncoder, VIntEncoder as _, COMPRESSION_BLOCK_SIZE};
+use crate::query::Bm25Weight;
+use crate::schema::IndexRecordOption;
+use crate::{DocId, Score};
+
+pub struct StandardPostingsSerializer {
+    last_doc_id_encoded: u32,
+
+    block_encoder: BlockEncoder,
+    block: Box<Block>,
+
+    postings_write: Vec<u8>,
+    skip_write: SkipSerializer,
+
+    mode: IndexRecordOption,
+    fieldnorm_reader: Option<FieldNormReader>,
+
+    bm25_weight: Option<Bm25Weight>,
+    avg_fieldnorm: Score, /* Average number of term in the field for that segment.
+                           * this value is used to compute the block wand information. */
+    term_has_freq: bool,
+}
+
+impl StandardPostingsSerializer {
+    pub fn new(
+        avg_fieldnorm: Score,
+        mode: IndexRecordOption,
+        fieldnorm_reader: Option<FieldNormReader>,
+    ) -> StandardPostingsSerializer {
+        Self {
+            last_doc_id_encoded: 0,
+            block_encoder: BlockEncoder::new(),
+            block: Box::new(Block::new()),
+            postings_write: Vec::new(),
+            skip_write: SkipSerializer::new(),
+            mode,
+            fieldnorm_reader,
+            bm25_weight: None,
+            avg_fieldnorm,
+            term_has_freq: false,
+        }
+    }
+}
+
+impl PostingsSerializer for StandardPostingsSerializer {
+    fn new_term(&mut self, term_doc_freq: u32, record_term_freq: bool) {
+        self.bm25_weight = None;
+
+        self.term_has_freq = self.mode.has_freq() && record_term_freq;
+        if !self.term_has_freq {
+            return;
+        }
+
+        let num_docs_in_segment: u64 =
+            if let Some(fieldnorm_reader) = self.fieldnorm_reader.as_ref() {
+                fieldnorm_reader.num_docs() as u64
+            } else {
+                return;
+            };
+
+        if num_docs_in_segment == 0 {
+            return;
+        }
+
+        self.bm25_weight = Some(Bm25Weight::for_one_term_without_explain(
+            term_doc_freq as u64,
+            num_docs_in_segment,
+            self.avg_fieldnorm,
+        ));
+    }
+
+    fn write_doc(&mut self, doc_id: DocId, term_freq: u32) {
+        self.block.append_doc(doc_id, term_freq);
+        if self.block.is_full() {
+            self.write_block();
+        }
+    }
+
+    fn close_term(
+        &mut self,
+        doc_freq: u32,
+        output_write: &mut impl std::io::Write,
+    ) -> io::Result<()> {
+        if !self.block.is_empty() {
+            // we have doc ids waiting to be written
+            // this happens when the number of doc ids is
+            // not a perfect multiple of our block size.
+            //
+            // In that case, the remaining part is encoded
+            // using variable int encoding.
+            {
+                let block_encoded = self
+                    .block_encoder
+                    .compress_vint_sorted(self.block.doc_ids(), self.last_doc_id_encoded);
+                self.postings_write.write_all(block_encoded)?;
+            }
+            // ... Idem for term frequencies
+            if self.term_has_freq {
+                let block_encoded = self
+                    .block_encoder
+                    .compress_vint_unsorted(self.block.term_freqs());
+                self.postings_write.write_all(block_encoded)?;
+            }
+            self.block.clear();
+        }
+        if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
+            let skip_data = self.skip_write.data();
+            VInt(skip_data.len() as u64).serialize(output_write)?;
+            output_write.write_all(skip_data)?;
+        }
+        output_write.write_all(&self.postings_write[..])?;
+        self.skip_write.clear();
+        self.postings_write.clear();
+        self.bm25_weight = None;
+        Ok(())
+    }
+
+    fn clear(&mut self) {
+        self.block.clear();
+        self.last_doc_id_encoded = 0;
+    }
+}
+
+impl StandardPostingsSerializer {
+    fn write_block(&mut self) {
+        {
+            // encode the doc ids
+            let (num_bits, block_encoded): (u8, &[u8]) = self
+                .block_encoder
+                .compress_block_sorted(self.block.doc_ids(), self.last_doc_id_encoded);
+            self.last_doc_id_encoded = self.block.last_doc();
+            self.skip_write
+                .write_doc(self.last_doc_id_encoded, num_bits);
+            // last el block 0, offset block 1,
+            self.postings_write.extend(block_encoded);
+        }
+        if self.term_has_freq {
+            let (num_bits, block_encoded): (u8, &[u8]) = self
+                .block_encoder
+                .compress_block_unsorted(self.block.term_freqs(), true);
+            self.postings_write.extend(block_encoded);
+            self.skip_write.write_term_freq(num_bits);
+            if self.mode.has_positions() {
+                // We serialize the sum of term freqs within the skip information
+                // in order to navigate through positions.
+                let sum_freq = self.block.term_freqs().iter().cloned().sum();
+                self.skip_write.write_total_term_freq(sum_freq);
+            }
+            let mut blockwand_params = (0u8, 0u32);
+            if let Some(bm25_weight) = self.bm25_weight.as_ref() {
+                if let Some(fieldnorm_reader) = self.fieldnorm_reader.as_ref() {
+                    let docs = self.block.doc_ids().iter().cloned();
+                    let term_freqs = self.block.term_freqs().iter().cloned();
+                    let fieldnorms = docs.map(|doc| fieldnorm_reader.fieldnorm_id(doc));
+                    blockwand_params = fieldnorms
+                        .zip(term_freqs)
+                        .max_by(
+                            |(left_fieldnorm_id, left_term_freq),
+                             (right_fieldnorm_id, right_term_freq)| {
+                                let left_score =
+                                    bm25_weight.tf_factor(*left_fieldnorm_id, *left_term_freq);
+                                let right_score =
+                                    bm25_weight.tf_factor(*right_fieldnorm_id, *right_term_freq);
+                                left_score
+                                    .partial_cmp(&right_score)
+                                    .unwrap_or(Ordering::Equal)
+                            },
+                        )
+                        .unwrap();
+                }
+            }
+            let (fieldnorm_id, term_freq) = blockwand_params;
+            self.skip_write.write_blockwand_max(fieldnorm_id, term_freq);
+        }
+        self.block.clear();
+    }
+}

src/collector/sort_key/mod.rs

@@ -1,10 +1,12 @@
 mod order;
+mod sort_by_erased_type;
 mod sort_by_score;
 mod sort_by_static_fast_value;
 mod sort_by_string;
 mod sort_key_computer;
 
 pub use order::*;
+pub use sort_by_erased_type::SortByErasedType;
 pub use sort_by_score::SortBySimilarityScore;
 pub use sort_by_static_fast_value::SortByStaticFastValue;
 pub use sort_by_string::SortByString;
@@ -34,11 +36,13 @@ pub(crate) mod tests {
     use std::collections::HashMap;
     use std::ops::Range;
 
-    use crate::collector::sort_key::{SortBySimilarityScore, SortByStaticFastValue, SortByString};
+    use crate::collector::sort_key::{
+        SortByErasedType, SortBySimilarityScore, SortByStaticFastValue, SortByString,
+    };
     use crate::collector::{ComparableDoc, DocSetCollector, TopDocs};
     use crate::indexer::NoMergePolicy;
     use crate::query::{AllQuery, QueryParser};
-    use crate::schema::{Schema, FAST, TEXT};
+    use crate::schema::{OwnedValue, Schema, FAST, TEXT};
     use crate::{DocAddress, Document, Index, Order, Score, Searcher};
 
     fn make_index() -> crate::Result<Index> {
@@ -313,11 +317,9 @@ pub(crate) mod tests {
                 (SortBySimilarityScore, score_order),
                 (SortByString::for_field("city"), city_order),
             ));
-            Ok(searcher
-                .search(&AllQuery, &top_collector)?
-                .into_iter()
-                .map(|(f, doc)| (f, ids[&doc]))
-                .collect())
+            let results: Vec<((Score, Option<String>), DocAddress)> =
+                searcher.search(&AllQuery, &top_collector)?;
+            Ok(results.into_iter().map(|(f, doc)| (f, ids[&doc])).collect())
         }
 
         assert_eq!(
@@ -342,6 +344,51 @@ pub(crate) mod tests {
         Ok(())
     }
 
+    #[test]
+    fn test_order_by_score_then_owned_value() -> crate::Result<()> {
+        let index = make_index()?;
+
+        type SortKey = (Score, OwnedValue);
+
+        fn query(
+            index: &Index,
+            score_order: Order,
+            city_order: Order,
+        ) -> crate::Result<Vec<(SortKey, u64)>> {
+            let searcher = index.reader()?.searcher();
+            let ids = id_mapping(&searcher);
+
+            let top_collector = TopDocs::with_limit(4).order_by::<(Score, OwnedValue)>((
+                (SortBySimilarityScore, score_order),
+                (SortByErasedType::for_field("city"), city_order),
+            ));
+            let results: Vec<((Score, OwnedValue), DocAddress)> =
+                searcher.search(&AllQuery, &top_collector)?;
+            Ok(results.into_iter().map(|(f, doc)| (f, ids[&doc])).collect())
+        }
+
+        assert_eq!(
+            &query(&index, Order::Asc, Order::Asc)?,
+            &[
+                ((1.0, OwnedValue::Str("austin".to_owned())), 0),
+                ((1.0, OwnedValue::Str("greenville".to_owned())), 1),
+                ((1.0, OwnedValue::Str("tokyo".to_owned())), 2),
+                ((1.0, OwnedValue::Null), 3),
+            ]
+        );
+
+        assert_eq!(
+            &query(&index, Order::Asc, Order::Desc)?,
+            &[
+                ((1.0, OwnedValue::Str("tokyo".to_owned())), 2),
+                ((1.0, OwnedValue::Str("greenville".to_owned())), 1),
+                ((1.0, OwnedValue::Str("austin".to_owned())), 0),
+                ((1.0, OwnedValue::Null), 3),
+            ]
+        );
+        Ok(())
+    }
+
     use proptest::prelude::*;
 
     proptest! {

src/collector/sort_key/order.rs

@@ -1,11 +1,70 @@
 use std::cmp::Ordering;
 
+use columnar::MonotonicallyMappableToU64;
 use serde::{Deserialize, Serialize};
 
 use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
-use crate::schema::Schema;
+use crate::schema::{OwnedValue, Schema};
 use crate::{DocId, Order, Score};
 
+fn compare_owned_value<const NULLS_FIRST: bool>(lhs: &OwnedValue, rhs: &OwnedValue) -> Ordering {
+    match (lhs, rhs) {
+        (OwnedValue::Null, OwnedValue::Null) => Ordering::Equal,
+        (OwnedValue::Null, _) => {
+            if NULLS_FIRST {
+                Ordering::Less
+            } else {
+                Ordering::Greater
+            }
+        }
+        (_, OwnedValue::Null) => {
+            if NULLS_FIRST {
+                Ordering::Greater
+            } else {
+                Ordering::Less
+            }
+        }
+        (OwnedValue::Str(a), OwnedValue::Str(b)) => a.cmp(b),
+        (OwnedValue::PreTokStr(a), OwnedValue::PreTokStr(b)) => a.cmp(b),
+        (OwnedValue::U64(a), OwnedValue::U64(b)) => a.cmp(b),
+        (OwnedValue::I64(a), OwnedValue::I64(b)) => a.cmp(b),
+        (OwnedValue::F64(a), OwnedValue::F64(b)) => a.to_u64().cmp(&b.to_u64()),
+        (OwnedValue::Bool(a), OwnedValue::Bool(b)) => a.cmp(b),
+        (OwnedValue::Date(a), OwnedValue::Date(b)) => a.cmp(b),
+        (OwnedValue::Facet(a), OwnedValue::Facet(b)) => a.cmp(b),
+        (OwnedValue::Bytes(a), OwnedValue::Bytes(b)) => a.cmp(b),
+        (OwnedValue::IpAddr(a), OwnedValue::IpAddr(b)) => a.cmp(b),
+        (OwnedValue::U64(a), OwnedValue::I64(b)) => {
+            if *b < 0 {
+                Ordering::Greater
+            } else {
+                a.cmp(&(*b as u64))
+            }
+        }
+        (OwnedValue::I64(a), OwnedValue::U64(b)) => {
+            if *a < 0 {
+                Ordering::Less
+            } else {
+                (*a as u64).cmp(b)
+            }
+        }
+        (OwnedValue::U64(a), OwnedValue::F64(b)) => (*a as f64).to_u64().cmp(&b.to_u64()),
+        (OwnedValue::F64(a), OwnedValue::U64(b)) => a.to_u64().cmp(&(*b as f64).to_u64()),
+        (OwnedValue::I64(a), OwnedValue::F64(b)) => (*a as f64).to_u64().cmp(&b.to_u64()),
+        (OwnedValue::F64(a), OwnedValue::I64(b)) => a.to_u64().cmp(&(*b as f64).to_u64()),
+        (a, b) => {
+            let ord = a.discriminant_value().cmp(&b.discriminant_value());
+            // If the discriminant is equal, it's because a new type was added, but hasn't been
+            // included in this `match` statement.
+            assert!(
+                ord != Ordering::Equal,
+                "Unimplemented comparison for type of {a:?}, {b:?}"
+            );
+            ord
+        }
+    }
+}
+
 /// Comparator trait defining the order in which documents should be ordered.
 pub trait Comparator<T>: Send + Sync + std::fmt::Debug + Default {
     /// Return the order between two values.
@@ -25,7 +84,18 @@ pub struct NaturalComparator;
 impl<T: PartialOrd> Comparator<T> for NaturalComparator {
     #[inline(always)]
     fn compare(&self, lhs: &T, rhs: &T) -> Ordering {
-        lhs.partial_cmp(rhs).unwrap()
+        lhs.partial_cmp(rhs).unwrap_or(Ordering::Equal)
+    }
+}
+
+/// A (partial) implementation of comparison for OwnedValue.
+///
+/// Intended for use within columns of homogenous types, and so will panic for OwnedValues with
+/// mismatched types. The one exception is Null, for which we do define all comparisons.
+impl Comparator<OwnedValue> for NaturalComparator {
+    #[inline(always)]
+    fn compare(&self, lhs: &OwnedValue, rhs: &OwnedValue) -> Ordering {
+        compare_owned_value::</* NULLS_FIRST= */ true>(lhs, rhs)
     }
 }
 
@@ -121,6 +191,13 @@ impl Comparator<String> for ReverseNoneIsLowerComparator {
     }
 }
 
+impl Comparator<OwnedValue> for ReverseNoneIsLowerComparator {
+    #[inline(always)]
+    fn compare(&self, lhs: &OwnedValue, rhs: &OwnedValue) -> Ordering {
+        compare_owned_value::</* NULLS_FIRST= */ false>(rhs, lhs)
+    }
+}
+
 /// Compare values naturally, but treating `None` as higher than `Some`.
 ///
 /// When used with `TopDocs`, which reverses the order, this results in a
@@ -185,6 +262,13 @@ impl Comparator<String> for NaturalNoneIsHigherComparator {
     }
 }
 
+impl Comparator<OwnedValue> for NaturalNoneIsHigherComparator {
+    #[inline(always)]
+    fn compare(&self, lhs: &OwnedValue, rhs: &OwnedValue) -> Ordering {
+        compare_owned_value::</* NULLS_FIRST= */ false>(lhs, rhs)
+    }
+}
+
 /// An enum representing the different sort orders.
 #[derive(Debug, Clone, Copy, Eq, PartialEq, Default)]
 pub enum ComparatorEnum {
@@ -404,11 +488,12 @@ impl<TSegmentSortKeyComputer, TSegmentSortKey, TComparator> SegmentSortKeyComput
     for SegmentSortKeyComputerWithComparator<TSegmentSortKeyComputer, TComparator>
 where
     TSegmentSortKeyComputer: SegmentSortKeyComputer<SegmentSortKey = TSegmentSortKey>,
-    TSegmentSortKey: PartialOrd + Clone + 'static + Sync + Send,
+    TSegmentSortKey: Clone + 'static + Sync + Send,
     TComparator: Comparator<TSegmentSortKey> + 'static + Sync + Send,
 {
     type SortKey = TSegmentSortKeyComputer::SortKey;
     type SegmentSortKey = TSegmentSortKey;
+    type SegmentComparator = TComparator;
 
     fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Self::SegmentSortKey {
         self.segment_sort_key_computer.segment_sort_key(doc, score)
@@ -432,6 +517,7 @@ where
 #[cfg(test)]
 mod tests {
     use super::*;
+    use crate::schema::OwnedValue;
 
     #[test]
     fn test_natural_none_is_higher() {
@@ -455,4 +541,27 @@ mod tests {
         // compare(None, None) should be Equal.
         assert_eq!(comp.compare(&null, &null), Ordering::Equal);
     }
+
+    #[test]
+    fn test_mixed_ownedvalue_compare() {
+        let u = OwnedValue::U64(10);
+        let i = OwnedValue::I64(10);
+        let f = OwnedValue::F64(10.0);
+
+        let nc = NaturalComparator;
+        assert_eq!(nc.compare(&u, &i), Ordering::Equal);
+        assert_eq!(nc.compare(&u, &f), Ordering::Equal);
+        assert_eq!(nc.compare(&i, &f), Ordering::Equal);
+
+        let u2 = OwnedValue::U64(11);
+        assert_eq!(nc.compare(&u2, &f), Ordering::Greater);
+
+        let s = OwnedValue::Str("a".to_string());
+        // Str < U64
+        assert_eq!(nc.compare(&s, &u), Ordering::Less);
+        // Str < I64
+        assert_eq!(nc.compare(&s, &i), Ordering::Less);
+        // Str < F64
+        assert_eq!(nc.compare(&s, &f), Ordering::Less);
+    }
 }

src/collector/sort_key/sort_by_erased_type.rs

@@ -0,0 +1,361 @@
+use columnar::{ColumnType, MonotonicallyMappableToU64};
+
+use crate::collector::sort_key::{
+    NaturalComparator, SortBySimilarityScore, SortByStaticFastValue, SortByString,
+};
+use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
+use crate::fastfield::FastFieldNotAvailableError;
+use crate::schema::OwnedValue;
+use crate::{DateTime, DocId, Score};
+
+/// Sort by the boxed / OwnedValue representation of either a fast field, or of the score.
+///
+/// Using the OwnedValue representation allows for type erasure, and can be useful when sort orders
+/// are not known until runtime. But it comes with a performance cost: wherever possible, prefer to
+/// use a SortKeyComputer implementation with a known-type at compile time.
+#[derive(Debug, Clone)]
+pub enum SortByErasedType {
+    /// Sort by a fast field
+    Field(String),
+    /// Sort by score
+    Score,
+}
+
+impl SortByErasedType {
+    /// Creates a new sort key computer which will sort by the given fast field column, with type
+    /// erasure.
+    pub fn for_field(column_name: impl ToString) -> Self {
+        Self::Field(column_name.to_string())
+    }
+
+    /// Creates a new sort key computer which will sort by score, with type erasure.
+    pub fn for_score() -> Self {
+        Self::Score
+    }
+}
+
+trait ErasedSegmentSortKeyComputer: Send + Sync {
+    fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Option<u64>;
+    fn convert_segment_sort_key(&self, sort_key: Option<u64>) -> OwnedValue;
+}
+
+struct ErasedSegmentSortKeyComputerWrapper<C, F> {
+    inner: C,
+    converter: F,
+}
+
+impl<C, F> ErasedSegmentSortKeyComputer for ErasedSegmentSortKeyComputerWrapper<C, F>
+where
+    C: SegmentSortKeyComputer<SegmentSortKey = Option<u64>> + Send + Sync,
+    F: Fn(C::SortKey) -> OwnedValue + Send + Sync + 'static,
+{
+    fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Option<u64> {
+        self.inner.segment_sort_key(doc, score)
+    }
+
+    fn convert_segment_sort_key(&self, sort_key: Option<u64>) -> OwnedValue {
+        let val = self.inner.convert_segment_sort_key(sort_key);
+        (self.converter)(val)
+    }
+}
+
+struct ScoreSegmentSortKeyComputer {
+    segment_computer: SortBySimilarityScore,
+}
+
+impl ErasedSegmentSortKeyComputer for ScoreSegmentSortKeyComputer {
+    fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Option<u64> {
+        let score_value: f64 = self.segment_computer.segment_sort_key(doc, score).into();
+        Some(score_value.to_u64())
+    }
+
+    fn convert_segment_sort_key(&self, sort_key: Option<u64>) -> OwnedValue {
+        let score_value: u64 = sort_key.expect("This implementation always produces a score.");
+        OwnedValue::F64(f64::from_u64(score_value))
+    }
+}
+
+impl SortKeyComputer for SortByErasedType {
+    type SortKey = OwnedValue;
+    type Child = ErasedColumnSegmentSortKeyComputer;
+    type Comparator = NaturalComparator;
+
+    fn requires_scoring(&self) -> bool {
+        matches!(self, Self::Score)
+    }
+
+    fn segment_sort_key_computer(
+        &self,
+        segment_reader: &crate::SegmentReader,
+    ) -> crate::Result<Self::Child> {
+        let inner: Box<dyn ErasedSegmentSortKeyComputer> = match self {
+            Self::Field(column_name) => {
+                let fast_fields = segment_reader.fast_fields();
+                // TODO: We currently double-open the column to avoid relying on the implementation
+                // details of `SortByString` or `SortByStaticFastValue`. Once
+                // https://github.com/quickwit-oss/tantivy/issues/2776 is resolved, we should
+                // consider directly constructing the appropriate `SegmentSortKeyComputer` type for
+                // the column that we open here.
+                let (_column, column_type) =
+                    fast_fields.u64_lenient(column_name)?.ok_or_else(|| {
+                        FastFieldNotAvailableError {
+                            field_name: column_name.to_owned(),
+                        }
+                    })?;
+
+                match column_type {
+                    ColumnType::Str => {
+                        let computer = SortByString::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<String>| {
+                                val.map(OwnedValue::Str).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    ColumnType::U64 => {
+                        let computer = SortByStaticFastValue::<u64>::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<u64>| {
+                                val.map(OwnedValue::U64).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    ColumnType::I64 => {
+                        let computer = SortByStaticFastValue::<i64>::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<i64>| {
+                                val.map(OwnedValue::I64).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    ColumnType::F64 => {
+                        let computer = SortByStaticFastValue::<f64>::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<f64>| {
+                                val.map(OwnedValue::F64).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    ColumnType::Bool => {
+                        let computer = SortByStaticFastValue::<bool>::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<bool>| {
+                                val.map(OwnedValue::Bool).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    ColumnType::DateTime => {
+                        let computer = SortByStaticFastValue::<DateTime>::for_field(column_name);
+                        let inner = computer.segment_sort_key_computer(segment_reader)?;
+                        Box::new(ErasedSegmentSortKeyComputerWrapper {
+                            inner,
+                            converter: |val: Option<DateTime>| {
+                                val.map(OwnedValue::Date).unwrap_or(OwnedValue::Null)
+                            },
+                        })
+                    }
+                    column_type => {
+                        return Err(crate::TantivyError::SchemaError(format!(
+                            "Field `{}` is of type {column_type:?}, which is not supported for \
+                             sorting by owned value yet.",
+                            column_name
+                        )))
+                    }
+                }
+            }
+            Self::Score => Box::new(ScoreSegmentSortKeyComputer {
+                segment_computer: SortBySimilarityScore,
+            }),
+        };
+        Ok(ErasedColumnSegmentSortKeyComputer { inner })
+    }
+}
+
+pub struct ErasedColumnSegmentSortKeyComputer {
+    inner: Box<dyn ErasedSegmentSortKeyComputer>,
+}
+
+impl SegmentSortKeyComputer for ErasedColumnSegmentSortKeyComputer {
+    type SortKey = OwnedValue;
+    type SegmentSortKey = Option<u64>;
+    type SegmentComparator = NaturalComparator;
+
+    #[inline(always)]
+    fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Option<u64> {
+        self.inner.segment_sort_key(doc, score)
+    }
+
+    fn convert_segment_sort_key(&self, segment_sort_key: Self::SegmentSortKey) -> OwnedValue {
+        self.inner.convert_segment_sort_key(segment_sort_key)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use crate::collector::sort_key::{ComparatorEnum, SortByErasedType};
+    use crate::collector::TopDocs;
+    use crate::query::AllQuery;
+    use crate::schema::{OwnedValue, Schema, FAST, TEXT};
+    use crate::Index;
+
+    #[test]
+    fn test_sort_by_owned_u64() {
+        let mut schema_builder = Schema::builder();
+        let id_field = schema_builder.add_u64_field("id", FAST);
+        let schema = schema_builder.build();
+        let index = Index::create_in_ram(schema);
+        let mut writer = index.writer_for_tests().unwrap();
+        writer.add_document(doc!(id_field => 10u64)).unwrap();
+        writer.add_document(doc!(id_field => 2u64)).unwrap();
+        writer.add_document(doc!()).unwrap();
+        writer.commit().unwrap();
+
+        let reader = index.reader().unwrap();
+        let searcher = reader.searcher();
+
+        let collector = TopDocs::with_limit(10)
+            .order_by((SortByErasedType::for_field("id"), ComparatorEnum::Natural));
+        let top_docs = searcher.search(&AllQuery, &collector).unwrap();
+
+        let values: Vec<OwnedValue> = top_docs.into_iter().map(|(key, _)| key).collect();
+
+        assert_eq!(
+            values,
+            vec![OwnedValue::U64(10), OwnedValue::U64(2), OwnedValue::Null]
+        );
+
+        let collector = TopDocs::with_limit(10).order_by((
+            SortByErasedType::for_field("id"),
+            ComparatorEnum::ReverseNoneLower,
+        ));
+        let top_docs = searcher.search(&AllQuery, &collector).unwrap();
+
+        let values: Vec<OwnedValue> = top_docs.into_iter().map(|(key, _)| key).collect();
+
+        assert_eq!(
+            values,
+            vec![OwnedValue::U64(2), OwnedValue::U64(10), OwnedValue::Null]
+        );
+    }
+
+    #[test]
+    fn test_sort_by_owned_string() {
+        let mut schema_builder = Schema::builder();
+        let city_field = schema_builder.add_text_field("city", FAST | TEXT);
+        let schema = schema_builder.build();
+        let index = Index::create_in_ram(schema);
+        let mut writer = index.writer_for_tests().unwrap();
+        writer.add_document(doc!(city_field => "tokyo")).unwrap();
+        writer.add_document(doc!(city_field => "austin")).unwrap();
+        writer.add_document(doc!()).unwrap();
+        writer.commit().unwrap();
+
+        let reader = index.reader().unwrap();
+        let searcher = reader.searcher();
+
+        let collector = TopDocs::with_limit(10).order_by((
+            SortByErasedType::for_field("city"),
+            ComparatorEnum::ReverseNoneLower,
+        ));
+        let top_docs = searcher.search(&AllQuery, &collector).unwrap();
+
+        let values: Vec<OwnedValue> = top_docs.into_iter().map(|(key, _)| key).collect();
+
+        assert_eq!(
+            values,
+            vec![
+                OwnedValue::Str("austin".to_string()),
+                OwnedValue::Str("tokyo".to_string()),
+                OwnedValue::Null
+            ]
+        );
+    }
+
+    #[test]
+    fn test_sort_by_owned_reverse() {
+        let mut schema_builder = Schema::builder();
+        let id_field = schema_builder.add_u64_field("id", FAST);
+        let schema = schema_builder.build();
+        let index = Index::create_in_ram(schema);
+        let mut writer = index.writer_for_tests().unwrap();
+        writer.add_document(doc!(id_field => 10u64)).unwrap();
+        writer.add_document(doc!(id_field => 2u64)).unwrap();
+        writer.add_document(doc!()).unwrap();
+        writer.commit().unwrap();
+
+        let reader = index.reader().unwrap();
+        let searcher = reader.searcher();
+
+        let collector = TopDocs::with_limit(10)
+            .order_by((SortByErasedType::for_field("id"), ComparatorEnum::Reverse));
+        let top_docs = searcher.search(&AllQuery, &collector).unwrap();
+
+        let values: Vec<OwnedValue> = top_docs.into_iter().map(|(key, _)| key).collect();
+
+        assert_eq!(
+            values,
+            vec![OwnedValue::Null, OwnedValue::U64(2), OwnedValue::U64(10)]
+        );
+    }
+
+    #[test]
+    fn test_sort_by_owned_score() {
+        let mut schema_builder = Schema::builder();
+        let body_field = schema_builder.add_text_field("body", TEXT);
+        let schema = schema_builder.build();
+        let index = Index::create_in_ram(schema);
+        let mut writer = index.writer_for_tests().unwrap();
+        writer.add_document(doc!(body_field => "a a")).unwrap();
+        writer.add_document(doc!(body_field => "a")).unwrap();
+        writer.commit().unwrap();
+
+        let reader = index.reader().unwrap();
+        let searcher = reader.searcher();
+        let query_parser = crate::query::QueryParser::for_index(&index, vec![body_field]);
+        let query = query_parser.parse_query("a").unwrap();
+
+        // Sort by score descending (Natural)
+        let collector = TopDocs::with_limit(10)
+            .order_by((SortByErasedType::for_score(), ComparatorEnum::Natural));
+        let top_docs = searcher.search(&query, &collector).unwrap();
+
+        let values: Vec<f64> = top_docs
+            .into_iter()
+            .map(|(key, _)| match key {
+                OwnedValue::F64(val) => val,
+                _ => panic!("Wrong type {key:?}"),
+            })
+            .collect();
+
+        assert_eq!(values.len(), 2);
+        assert!(values[0] > values[1]);
+
+        // Sort by score ascending (ReverseNoneLower)
+        let collector = TopDocs::with_limit(10).order_by((
+            SortByErasedType::for_score(),
+            ComparatorEnum::ReverseNoneLower,
+        ));
+        let top_docs = searcher.search(&query, &collector).unwrap();
+
+        let values: Vec<f64> = top_docs
+            .into_iter()
+            .map(|(key, _)| match key {
+                OwnedValue::F64(val) => val,
+                _ => panic!("Wrong type {key:?}"),
+            })
+            .collect();
+
+        assert_eq!(values.len(), 2);
+        assert!(values[0] < values[1]);
+    }
+}

src/collector/sort_key/sort_by_score.rs

@@ -63,8 +63,8 @@ impl SortKeyComputer for SortBySimilarityScore {
 
 impl SegmentSortKeyComputer for SortBySimilarityScore {
     type SortKey = Score;
-
     type SegmentSortKey = Score;
+    type SegmentComparator = NaturalComparator;
 
     #[inline(always)]
     fn segment_sort_key(&mut self, _doc: DocId, score: Score) -> Score {

src/collector/sort_key/sort_by_static_fast_value.rs

@@ -34,9 +34,7 @@ impl<T: FastValue> SortByStaticFastValue<T> {
 
 impl<T: FastValue> SortKeyComputer for SortByStaticFastValue<T> {
     type Child = SortByFastValueSegmentSortKeyComputer<T>;
-
     type SortKey = Option<T>;
-
     type Comparator = NaturalComparator;
 
     fn check_schema(&self, schema: &crate::schema::Schema) -> crate::Result<()> {
@@ -84,8 +82,8 @@ pub struct SortByFastValueSegmentSortKeyComputer<T> {
 
 impl<T: FastValue> SegmentSortKeyComputer for SortByFastValueSegmentSortKeyComputer<T> {
     type SortKey = Option<T>;
-
     type SegmentSortKey = Option<u64>;
+    type SegmentComparator = NaturalComparator;
 
     #[inline(always)]
     fn segment_sort_key(&mut self, doc: DocId, _score: Score) -> Self::SegmentSortKey {

src/collector/sort_key/sort_by_string.rs

@@ -30,9 +30,7 @@ impl SortByString {
 
 impl SortKeyComputer for SortByString {
     type SortKey = Option<String>;
-
     type Child = ByStringColumnSegmentSortKeyComputer;
-
     type Comparator = NaturalComparator;
 
     fn segment_sort_key_computer(
@@ -50,8 +48,8 @@ pub struct ByStringColumnSegmentSortKeyComputer {
 
 impl SegmentSortKeyComputer for ByStringColumnSegmentSortKeyComputer {
     type SortKey = Option<String>;
-
     type SegmentSortKey = Option<TermOrdinal>;
+    type SegmentComparator = NaturalComparator;
 
     #[inline(always)]
     fn segment_sort_key(&mut self, doc: DocId, _score: Score) -> Option<TermOrdinal> {
@@ -60,6 +58,8 @@ impl SegmentSortKeyComputer for ByStringColumnSegmentSortKeyComputer {
     }
 
     fn convert_segment_sort_key(&self, term_ord_opt: Option<TermOrdinal>) -> Option<String> {
+        // TODO: Individual lookups to the dictionary like this are very likely to repeatedly
+        // decompress the same blocks. See https://github.com/quickwit-oss/tantivy/issues/2776
         let term_ord = term_ord_opt?;
         let str_column = self.str_column_opt.as_ref()?;
         let mut bytes = Vec::new();

src/collector/sort_key/sort_key_computer.rs

@@ -12,13 +12,21 @@ use crate::{DocAddress, DocId, Result, Score, SegmentReader};
 /// It is the segment local version of the [`SortKeyComputer`].
 pub trait SegmentSortKeyComputer: 'static {
     /// The final score being emitted.
-    type SortKey: 'static + PartialOrd + Send + Sync + Clone;
+    type SortKey: 'static + Send + Sync + Clone;
 
     /// Sort key used by at the segment level by the `SegmentSortKeyComputer`.
     ///
     /// It is typically small like a `u64`, and is meant to be converted
     /// to the final score at the end of the collection of the segment.
-    type SegmentSortKey: 'static + PartialOrd + Clone + Send + Sync + Clone;
+    type SegmentSortKey: 'static + Clone + Send + Sync + Clone;
+
+    /// Comparator type.
+    type SegmentComparator: Comparator<Self::SegmentSortKey> + 'static;
+
+    /// Returns the segment sort key comparator.
+    fn segment_comparator(&self) -> Self::SegmentComparator {
+        Self::SegmentComparator::default()
+    }
 
     /// Computes the sort key for the given document and score.
     fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Self::SegmentSortKey;
@@ -47,7 +55,7 @@ pub trait SegmentSortKeyComputer: 'static {
         left: &Self::SegmentSortKey,
         right: &Self::SegmentSortKey,
     ) -> Ordering {
-        NaturalComparator.compare(left, right)
+        self.segment_comparator().compare(left, right)
     }
 
     /// Implementing this method makes it possible to avoid computing
@@ -81,7 +89,7 @@ pub trait SegmentSortKeyComputer: 'static {
 /// the sort key at a segment scale.
 pub trait SortKeyComputer: Sync {
     /// The sort key type.
-    type SortKey: 'static + Send + Sync + PartialOrd + Clone + std::fmt::Debug;
+    type SortKey: 'static + Send + Sync + Clone + std::fmt::Debug;
     /// Type of the associated [`SegmentSortKeyComputer`].
     type Child: SegmentSortKeyComputer<SortKey = Self::SortKey>;
     /// Comparator type.
@@ -136,10 +144,7 @@ where
     HeadSortKeyComputer: SortKeyComputer,
     TailSortKeyComputer: SortKeyComputer,
 {
-    type SortKey = (
-        <HeadSortKeyComputer::Child as SegmentSortKeyComputer>::SortKey,
-        <TailSortKeyComputer::Child as SegmentSortKeyComputer>::SortKey,
-    );
+    type SortKey = (HeadSortKeyComputer::SortKey, TailSortKeyComputer::SortKey);
     type Child = (HeadSortKeyComputer::Child, TailSortKeyComputer::Child);
 
     type Comparator = (
@@ -188,6 +193,11 @@ where
         TailSegmentSortKeyComputer::SegmentSortKey,
     );
 
+    type SegmentComparator = (
+        HeadSegmentSortKeyComputer::SegmentComparator,
+        TailSegmentSortKeyComputer::SegmentComparator,
+    );
+
     /// A SegmentSortKeyComputer maps to a SegmentSortKey, but it can also decide on
     /// its ordering.
     ///
@@ -269,11 +279,12 @@ impl<T, PreviousScore, NewScore> SegmentSortKeyComputer
     for MappedSegmentSortKeyComputer<T, PreviousScore, NewScore>
 where
     T: SegmentSortKeyComputer<SortKey = PreviousScore>,
-    PreviousScore: 'static + Clone + Send + Sync + PartialOrd,
-    NewScore: 'static + Clone + Send + Sync + PartialOrd,
+    PreviousScore: 'static + Clone + Send + Sync,
+    NewScore: 'static + Clone + Send + Sync,
 {
     type SortKey = NewScore;
     type SegmentSortKey = T::SegmentSortKey;
+    type SegmentComparator = T::SegmentComparator;
 
     fn segment_sort_key(&mut self, doc: DocId, score: Score) -> Self::SegmentSortKey {
         self.sort_key_computer.segment_sort_key(doc, score)
@@ -463,6 +474,7 @@ where
 {
     type SortKey = TSortKey;
     type SegmentSortKey = TSortKey;
+    type SegmentComparator = NaturalComparator;
 
     fn segment_sort_key(&mut self, doc: DocId, _score: Score) -> TSortKey {
         (self)(doc)

src/collector/top_score_collector.rs

@@ -324,7 +324,7 @@ impl TopDocs {
         sort_key_computer: impl SortKeyComputer<SortKey = TSortKey> + Send + 'static,
     ) -> impl Collector<Fruit = Vec<(TSortKey, DocAddress)>>
     where
-        TSortKey: 'static + Clone + Send + Sync + PartialOrd + std::fmt::Debug,
+        TSortKey: 'static + Clone + Send + Sync + std::fmt::Debug,
     {
         TopBySortKeyCollector::new(sort_key_computer, self.doc_range())
     }
@@ -445,7 +445,7 @@ where
     F: 'static + Send + Sync + Fn(&SegmentReader) -> TTweakScoreSortKeyFn,
     TTweakScoreSortKeyFn: 'static + Fn(DocId, Score) -> TSortKey,
     TweakScoreSegmentSortKeyComputer<TTweakScoreSortKeyFn>:
-        SegmentSortKeyComputer<SortKey = TSortKey>,
+        SegmentSortKeyComputer<SortKey = TSortKey, SegmentSortKey = TSortKey>,
     TSortKey: 'static + PartialOrd + Clone + Send + Sync + std::fmt::Debug,
 {
     type SortKey = TSortKey;
@@ -480,6 +480,7 @@ where
 {
     type SortKey = TSortKey;
     type SegmentSortKey = TSortKey;
+    type SegmentComparator = NaturalComparator;
 
     fn segment_sort_key(&mut self, doc: DocId, score: Score) -> TSortKey {
         (self.sort_key_fn)(doc, score)

src/core/executor.rs

@@ -48,7 +48,15 @@ impl Executor {
         F: Sized + Sync + Fn(A) -> crate::Result<R>,
     {
         match self {
-            Executor::SingleThread => args.map(f).collect::<crate::Result<_>>(),
+            Executor::SingleThread => {
+                // Avoid `collect`, since the stacktrace is blown up by it, which makes profiling
+                // harder.
+                let mut result = Vec::with_capacity(args.size_hint().0);
+                for arg in args {
+                    result.push(f(arg)?);
+                }
+                Ok(result)
+            }
             Executor::ThreadPool(pool) => {
                 let args: Vec<A> = args.collect();
                 let num_fruits = args.len();

src/core/json_utils.rs

@@ -4,7 +4,7 @@ use common::{replace_in_place, JsonPathWriter};
 use rustc_hash::FxHashMap;
 
 use crate::indexer::indexing_term::IndexingTerm;
-use crate::postings::{IndexingContext, IndexingPosition, PostingsWriter};
+use crate::postings::{IndexingContext, IndexingPosition, PostingsWriter as _, PostingsWriterEnum};
 use crate::schema::document::{ReferenceValue, ReferenceValueLeaf, Value};
 use crate::schema::{Type, DATE_TIME_PRECISION_INDEXED};
 use crate::time::format_description::well_known::Rfc3339;
@@ -80,7 +80,7 @@ fn index_json_object<'a, V: Value<'a>>(
     text_analyzer: &mut TextAnalyzer,
     term_buffer: &mut IndexingTerm,
     json_path_writer: &mut JsonPathWriter,
-    postings_writer: &mut dyn PostingsWriter,
+    postings_writer: &mut PostingsWriterEnum,
     ctx: &mut IndexingContext,
     positions_per_path: &mut IndexingPositionsPerPath,
 ) {
@@ -110,7 +110,7 @@ pub(crate) fn index_json_value<'a, V: Value<'a>>(
     text_analyzer: &mut TextAnalyzer,
     term_buffer: &mut IndexingTerm,
     json_path_writer: &mut JsonPathWriter,
-    postings_writer: &mut dyn PostingsWriter,
+    postings_writer: &mut PostingsWriterEnum,
     ctx: &mut IndexingContext,
     positions_per_path: &mut IndexingPositionsPerPath,
 ) {

src/directory/mmap_directory/mod.rs

@@ -1,3 +1,5 @@
+mod file_watcher;
+
 use std::collections::HashMap;
 use std::fmt;
 use std::fs::{self, File, OpenOptions};
@@ -7,6 +9,7 @@ use std::path::{Path, PathBuf};
 use std::sync::{Arc, RwLock, Weak};
 
 use common::StableDeref;
+use file_watcher::FileWatcher;
 use fs4::fs_std::FileExt;
 #[cfg(all(feature = "mmap", unix))]
 pub use memmap2::Advice;
@@ -18,7 +21,6 @@ use crate::core::META_FILEPATH;
 use crate::directory::error::{
     DeleteError, LockError, OpenDirectoryError, OpenReadError, OpenWriteError,
 };
-use crate::directory::file_watcher::FileWatcher;
 use crate::directory::{
     AntiCallToken, Directory, DirectoryLock, FileHandle, Lock, OwnedBytes, TerminatingWrite,
     WatchCallback, WatchHandle, WritePtr,

src/directory/mod.rs

@@ -5,7 +5,6 @@ mod mmap_directory;
 
 mod directory;
 mod directory_lock;
-mod file_watcher;
 pub mod footer;
 mod managed_directory;
 mod ram_directory;

src/docset.rs

@@ -1,5 +1,7 @@
 use std::borrow::{Borrow, BorrowMut};
 
+use common::BitSet;
+
 use crate::fastfield::AliveBitSet;
 use crate::DocId;
 
@@ -42,7 +44,6 @@ pub trait DocSet: Send {
     /// Calling `seek(TERMINATED)` is also legal and is the normal way to consume a `DocSet`.
     ///
     /// `target` has to be larger or equal to `.doc()` when calling `seek`.
-    /// If `target` is equal to `.doc()` then the DocSet should not advance.
     fn seek(&mut self, target: DocId) -> DocId {
         let mut doc = self.doc();
         debug_assert!(doc <= target);
@@ -107,6 +108,15 @@ pub trait DocSet: Send {
         buffer.len()
     }
 
+    /// TODO comment on the size of the bitset
+    fn fill_bitset(&mut self, bitset: &mut BitSet) {
+        let mut doc = self.doc();
+        while doc != TERMINATED {
+            bitset.insert(doc);
+            doc = self.advance();
+        }
+    }
+
     /// Returns the current document
     /// Right after creating a new `DocSet`, the docset points to the first document.
     ///
@@ -167,19 +177,6 @@ pub trait DocSet: Send {
     }
 }
 
-/// Consumes the `DocSet` and returns a Vec with all of the docs in the DocSet
-/// including the current doc.
-#[cfg(test)]
-pub fn docset_to_doc_vec(mut doc_set: Box<dyn DocSet>) -> Vec<DocId> {
-    let mut output = Vec::new();
-    let mut doc = doc_set.doc();
-    while doc != TERMINATED {
-        output.push(doc);
-        doc = doc_set.advance();
-    }
-    output
-}
-
 impl DocSet for &mut dyn DocSet {
     fn advance(&mut self) -> u32 {
         (**self).advance()

src/fieldnorm/mod.rs

@@ -113,7 +113,7 @@ mod tests {
             IndexRecordOption::WithFreqs,
         );
         let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-        let mut scorer = weight.scorer(searcher.segment_reader(0), 1.0f32, 0)?;
+        let mut scorer = weight.scorer(searcher.segment_reader(0), 1.0f32)?;
         assert_eq!(scorer.doc(), 0);
         assert!((scorer.score() - 0.22920431).abs() < 0.001f32);
         assert_eq!(scorer.advance(), 1);
@@ -142,7 +142,7 @@ mod tests {
             IndexRecordOption::WithFreqs,
         );
         let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-        let mut scorer = weight.scorer(searcher.segment_reader(0), 1.0f32, 0)?;
+        let mut scorer = weight.scorer(searcher.segment_reader(0), 1.0f32)?;
         assert_eq!(scorer.doc(), 0);
         assert!((scorer.score() - 0.22920431).abs() < 0.001f32);
         assert_eq!(scorer.advance(), 1);

src/index/index.rs

@@ -8,6 +8,7 @@ use std::thread::available_parallelism;
 use super::segment::Segment;
 use super::segment_reader::merge_field_meta_data;
 use super::{FieldMetadata, IndexSettings};
+use crate::codec::{CodecConfiguration, StandardCodec};
 use crate::core::{Executor, META_FILEPATH};
 use crate::directory::error::OpenReadError;
 #[cfg(feature = "mmap")]
@@ -59,6 +60,7 @@ fn save_new_metas(
     schema: Schema,
     index_settings: IndexSettings,
     directory: &dyn Directory,
+    codec: CodecConfiguration,
 ) -> crate::Result<()> {
     save_metas(
         &IndexMeta {
@@ -67,6 +69,7 @@ fn save_new_metas(
             schema,
             opstamp: 0u64,
             payload: None,
+            codec,
         },
         directory,
     )?;
@@ -101,18 +104,21 @@ fn save_new_metas(
 /// };
 /// let index = Index::builder().schema(schema).settings(settings).create_in_ram();
 /// ```
-pub struct IndexBuilder {
+pub struct IndexBuilder<Codec: crate::codec::Codec = StandardCodec> {
     schema: Option<Schema>,
     index_settings: IndexSettings,
     tokenizer_manager: TokenizerManager,
     fast_field_tokenizer_manager: TokenizerManager,
+    codec: Codec,
 }
-impl Default for IndexBuilder {
+
+impl Default for IndexBuilder<StandardCodec> {
     fn default() -> Self {
         IndexBuilder::new()
     }
 }
-impl IndexBuilder {
+
+impl IndexBuilder<StandardCodec> {
     /// Creates a new `IndexBuilder`
     pub fn new() -> Self {
         Self {
@@ -120,6 +126,21 @@ impl IndexBuilder {
             index_settings: IndexSettings::default(),
             tokenizer_manager: TokenizerManager::default(),
             fast_field_tokenizer_manager: TokenizerManager::default(),
+            codec: StandardCodec,
+        }
+    }
+}
+
+impl<Codec: crate::codec::Codec> IndexBuilder<Codec> {
+    /// Set the codec
+    #[must_use]
+    pub fn codec<NewCodec: crate::codec::Codec>(self, codec: NewCodec) -> IndexBuilder<NewCodec> {
+        IndexBuilder {
+            schema: self.schema,
+            index_settings: self.index_settings,
+            tokenizer_manager: self.tokenizer_manager,
+            fast_field_tokenizer_manager: self.fast_field_tokenizer_manager,
+            codec,
         }
     }
 
@@ -154,7 +175,7 @@ impl IndexBuilder {
     /// The index will be allocated in anonymous memory.
     /// This is useful for indexing small set of documents
     /// for instances like unit test or temporary in memory index.
-    pub fn create_in_ram(self) -> Result<Index, TantivyError> {
+    pub fn create_in_ram(self) -> Result<Index<Codec>, TantivyError> {
         let ram_directory = RamDirectory::create();
         self.create(ram_directory)
     }
@@ -165,7 +186,7 @@ impl IndexBuilder {
     /// If a previous index was in this directory, it returns an
     /// [`TantivyError::IndexAlreadyExists`] error.
     #[cfg(feature = "mmap")]
-    pub fn create_in_dir<P: AsRef<Path>>(self, directory_path: P) -> crate::Result<Index> {
+    pub fn create_in_dir<P: AsRef<Path>>(self, directory_path: P) -> crate::Result<Index<Codec>> {
         let mmap_directory: Box<dyn Directory> = Box::new(MmapDirectory::open(directory_path)?);
         if Index::exists(&*mmap_directory)? {
             return Err(TantivyError::IndexAlreadyExists);
@@ -186,7 +207,7 @@ impl IndexBuilder {
         self,
         dir: impl Into<Box<dyn Directory>>,
         mem_budget: usize,
-    ) -> crate::Result<SingleSegmentIndexWriter<D>> {
+    ) -> crate::Result<SingleSegmentIndexWriter<Codec, D>> {
         let index = self.create(dir)?;
         let index_simple_writer = SingleSegmentIndexWriter::new(index, mem_budget)?;
         Ok(index_simple_writer)
@@ -202,7 +223,7 @@ impl IndexBuilder {
     /// For other unit tests, prefer the [`RamDirectory`], see:
     /// [`IndexBuilder::create_in_ram()`].
     #[cfg(feature = "mmap")]
-    pub fn create_from_tempdir(self) -> crate::Result<Index> {
+    pub fn create_from_tempdir(self) -> crate::Result<Index<Codec>> {
         let mmap_directory: Box<dyn Directory> = Box::new(MmapDirectory::create_from_tempdir()?);
         self.create(mmap_directory)
     }
@@ -215,12 +236,15 @@ impl IndexBuilder {
     }
 
     /// Opens or creates a new index in the provided directory
-    pub fn open_or_create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
+    pub fn open_or_create<T: Into<Box<dyn Directory>>>(
+        self,
+        dir: T,
+    ) -> crate::Result<Index<Codec>> {
         let dir: Box<dyn Directory> = dir.into();
         if !Index::exists(&*dir)? {
             return self.create(dir);
         }
-        let mut index = Index::open(dir)?;
+        let mut index: Index<Codec> = Index::<Codec>::open_with_codec(dir)?;
         index.set_tokenizers(self.tokenizer_manager.clone());
         if index.schema() == self.get_expect_schema()? {
             Ok(index)
@@ -244,18 +268,26 @@ impl IndexBuilder {
     /// Creates a new index given an implementation of the trait `Directory`.
     ///
     /// If a directory previously existed, it will be erased.
-    fn create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
+    pub fn create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index<Codec>> {
+        self.create_avoid_monomorphization(dir.into())
+    }
+
+    fn create_avoid_monomorphization(self, dir: Box<dyn Directory>) -> crate::Result<Index<Codec>> {
         self.validate()?;
         let dir = dir.into();
         let directory = ManagedDirectory::wrap(dir)?;
+        let codec: CodecConfiguration = CodecConfiguration::from_codec(&self.codec);
         save_new_metas(
             self.get_expect_schema()?,
             self.index_settings.clone(),
             &directory,
+            codec,
         )?;
-        let mut metas = IndexMeta::with_schema(self.get_expect_schema()?);
+        let schema = self.get_expect_schema()?;
+        let mut metas = IndexMeta::with_schema_and_codec(schema, &self.codec);
         metas.index_settings = self.index_settings;
-        let mut index = Index::open_from_metas(directory, &metas, SegmentMetaInventory::default());
+        let mut index: Index<Codec> =
+            Index::<Codec>::open_from_metas(directory, &metas, SegmentMetaInventory::default())?;
         index.set_tokenizers(self.tokenizer_manager);
         index.set_fast_field_tokenizers(self.fast_field_tokenizer_manager);
         Ok(index)
@@ -264,7 +296,7 @@ impl IndexBuilder {
 
 /// Search Index
 #[derive(Clone)]
-pub struct Index {
+pub struct Index<Codec: crate::codec::Codec = crate::codec::StandardCodec> {
     directory: ManagedDirectory,
     schema: Schema,
     settings: IndexSettings,
@@ -272,6 +304,7 @@ pub struct Index {
     tokenizers: TokenizerManager,
     fast_field_tokenizers: TokenizerManager,
     inventory: SegmentMetaInventory,
+    codec: Codec,
 }
 
 impl Index {
@@ -279,41 +312,6 @@ impl Index {
     pub fn builder() -> IndexBuilder {
         IndexBuilder::new()
     }
-    /// Examines the directory to see if it contains an index.
-    ///
-    /// Effectively, it only checks for the presence of the `meta.json` file.
-    pub fn exists(dir: &dyn Directory) -> Result<bool, OpenReadError> {
-        dir.exists(&META_FILEPATH)
-    }
-
-    /// Accessor to the search executor.
-    ///
-    /// This pool is used by default when calling `searcher.search(...)`
-    /// to perform search on the individual segments.
-    ///
-    /// By default the executor is single thread, and simply runs in the calling thread.
-    pub fn search_executor(&self) -> &Executor {
-        &self.executor
-    }
-
-    /// Replace the default single thread search executor pool
-    /// by a thread pool with a given number of threads.
-    pub fn set_multithread_executor(&mut self, num_threads: usize) -> crate::Result<()> {
-        self.executor = Executor::multi_thread(num_threads, "tantivy-search-")?;
-        Ok(())
-    }
-
-    /// Custom thread pool by a outer thread pool.
-    pub fn set_executor(&mut self, executor: Executor) {
-        self.executor = executor;
-    }
-
-    /// Replace the default single thread search executor pool
-    /// by a thread pool with as many threads as there are CPUs on the system.
-    pub fn set_default_multithread_executor(&mut self) -> crate::Result<()> {
-        let default_num_threads = available_parallelism()?.get();
-        self.set_multithread_executor(default_num_threads)
-    }
 
     /// Creates a new index using the [`RamDirectory`].
     ///
@@ -324,6 +322,13 @@ impl Index {
         IndexBuilder::new().schema(schema).create_in_ram().unwrap()
     }
 
+    /// Examines the directory to see if it contains an index.
+    ///
+    /// Effectively, it only checks for the presence of the `meta.json` file.
+    pub fn exists(directory: &dyn Directory) -> Result<bool, OpenReadError> {
+        directory.exists(&META_FILEPATH)
+    }
+
     /// Creates a new index in a given filepath.
     /// The index will use the [`MmapDirectory`].
     ///
@@ -370,20 +375,107 @@ impl Index {
         schema: Schema,
         settings: IndexSettings,
     ) -> crate::Result<Index> {
-        let dir: Box<dyn Directory> = dir.into();
+        Self::create_to_avoid_monomorphization(dir.into(), schema, settings)
+    }
+
+    fn create_to_avoid_monomorphization(
+        dir: Box<dyn Directory>,
+        schema: Schema,
+        settings: IndexSettings,
+    ) -> crate::Result<Index> {
         let mut builder = IndexBuilder::new().schema(schema);
         builder = builder.settings(settings);
         builder.create(dir)
     }
 
+    /// Opens a new directory from an index path.
+    #[cfg(feature = "mmap")]
+    pub fn open_in_dir<P: AsRef<Path>>(directory_path: P) -> crate::Result<Index> {
+        Self::open_in_dir_to_avoid_monomorphization(directory_path.as_ref())
+    }
+
+    #[inline(never)]
+    fn open_in_dir_to_avoid_monomorphization(directory_path: &Path) -> crate::Result<Index> {
+        let mmap_directory = MmapDirectory::open(directory_path)?;
+        Index::open(mmap_directory)
+    }
+
+    /// Open the index using the provided directory
+    pub fn open<T: Into<Box<dyn Directory>>>(directory: T) -> crate::Result<Index> {
+        Index::<StandardCodec>::open_with_codec(directory.into())
+    }
+}
+
+impl<Codec: crate::codec::Codec> Index<Codec> {
+    /// Returns a version of this index with the standard codec.
+    /// This is useful when you need to pass the index to APIs that
+    /// don't care about the codec (e.g., for reading).
+    pub(crate) fn with_standard_codec(&self) -> Index<StandardCodec> {
+        Index {
+            directory: self.directory.clone(),
+            schema: self.schema.clone(),
+            settings: self.settings.clone(),
+            executor: self.executor.clone(),
+            tokenizers: self.tokenizers.clone(),
+            fast_field_tokenizers: self.fast_field_tokenizers.clone(),
+            inventory: self.inventory.clone(),
+            codec: StandardCodec::default(),
+        }
+    }
+
+    /// Open the index using the provided directory
+    #[inline(never)]
+    pub fn open_with_codec(directory: Box<dyn Directory>) -> crate::Result<Index<Codec>> {
+        let directory = ManagedDirectory::wrap(directory)?;
+        let inventory = SegmentMetaInventory::default();
+        let metas = load_metas(&directory, &inventory)?;
+        let index: Index<Codec> = Index::<Codec>::open_from_metas(directory, &metas, inventory)?;
+        Ok(index)
+    }
+
+    /// Accessor to the codec.
+    pub fn codec(&self) -> &Codec {
+        &self.codec
+    }
+
+    /// Accessor to the search executor.
+    ///
+    /// This pool is used by default when calling `searcher.search(...)`
+    /// to perform search on the individual segments.
+    ///
+    /// By default the executor is single thread, and simply runs in the calling thread.
+    pub fn search_executor(&self) -> &Executor {
+        &self.executor
+    }
+
+    /// Replace the default single thread search executor pool
+    /// by a thread pool with a given number of threads.
+    pub fn set_multithread_executor(&mut self, num_threads: usize) -> crate::Result<()> {
+        self.executor = Executor::multi_thread(num_threads, "tantivy-search-")?;
+        Ok(())
+    }
+
+    /// Custom thread pool by a outer thread pool.
+    pub fn set_executor(&mut self, executor: Executor) {
+        self.executor = executor;
+    }
+
+    /// Replace the default single thread search executor pool
+    /// by a thread pool with as many threads as there are CPUs on the system.
+    pub fn set_default_multithread_executor(&mut self) -> crate::Result<()> {
+        let default_num_threads = available_parallelism()?.get();
+        self.set_multithread_executor(default_num_threads)
+    }
+
     /// Creates a new index given a directory and an [`IndexMeta`].
-    fn open_from_metas(
+    fn open_from_metas<C: crate::codec::Codec>(
         directory: ManagedDirectory,
         metas: &IndexMeta,
         inventory: SegmentMetaInventory,
-    ) -> Index {
+    ) -> crate::Result<Index<C>> {
         let schema = metas.schema.clone();
-        Index {
+        let codec = metas.codec.to_codec::<C>()?;
+        Ok(Index {
             settings: metas.index_settings.clone(),
             directory,
             schema,
@@ -391,7 +483,8 @@ impl Index {
             fast_field_tokenizers: TokenizerManager::default(),
             executor: Executor::single_thread(),
             inventory,
-        }
+            codec,
+        })
     }
 
     /// Setter for the tokenizer manager.
@@ -447,7 +540,7 @@ impl Index {
     /// Create a default [`IndexReader`] for the given index.
     ///
     /// See [`Index.reader_builder()`].
-    pub fn reader(&self) -> crate::Result<IndexReader> {
+    pub fn reader(&self) -> crate::Result<IndexReader<Codec>> {
         self.reader_builder().try_into()
     }
 
@@ -455,17 +548,10 @@ impl Index {
     ///
     /// Most project should create at most one reader for a given index.
     /// This method is typically called only once per `Index` instance.
-    pub fn reader_builder(&self) -> IndexReaderBuilder {
+    pub fn reader_builder(&self) -> IndexReaderBuilder<Codec> {
         IndexReaderBuilder::new(self.clone())
     }
 
-    /// Opens a new directory from an index path.
-    #[cfg(feature = "mmap")]
-    pub fn open_in_dir<P: AsRef<Path>>(directory_path: P) -> crate::Result<Index> {
-        let mmap_directory = MmapDirectory::open(directory_path)?;
-        Index::open(mmap_directory)
-    }
-
     /// Returns the list of the segment metas tracked by the index.
     ///
     /// Such segments can of course be part of the index,
@@ -506,16 +592,6 @@ impl Index {
         self.inventory.new_segment_meta(segment_id, max_doc)
     }
 
-    /// Open the index using the provided directory
-    pub fn open<T: Into<Box<dyn Directory>>>(directory: T) -> crate::Result<Index> {
-        let directory = directory.into();
-        let directory = ManagedDirectory::wrap(directory)?;
-        let inventory = SegmentMetaInventory::default();
-        let metas = load_metas(&directory, &inventory)?;
-        let index = Index::open_from_metas(directory, &metas, inventory);
-        Ok(index)
-    }
-
     /// Reads the index meta file from the directory.
     pub fn load_metas(&self) -> crate::Result<IndexMeta> {
         load_metas(self.directory(), &self.inventory)
@@ -539,7 +615,7 @@ impl Index {
     pub fn writer_with_options<D: Document>(
         &self,
         options: IndexWriterOptions,
-    ) -> crate::Result<IndexWriter<D>> {
+    ) -> crate::Result<IndexWriter<Codec, D>> {
         let directory_lock = self
             .directory
             .acquire_lock(&INDEX_WRITER_LOCK)
@@ -581,7 +657,7 @@ impl Index {
         &self,
         num_threads: usize,
         overall_memory_budget_in_bytes: usize,
-    ) -> crate::Result<IndexWriter<D>> {
+    ) -> crate::Result<IndexWriter<Codec, D>> {
         let memory_arena_in_bytes_per_thread = overall_memory_budget_in_bytes / num_threads;
         let options = IndexWriterOptions::builder()
             .num_worker_threads(num_threads)
@@ -595,7 +671,7 @@ impl Index {
     /// That index writer only simply has a single thread and a memory budget of 15 MB.
     /// Using a single thread gives us a deterministic allocation of DocId.
     #[cfg(test)]
-    pub fn writer_for_tests<D: Document>(&self) -> crate::Result<IndexWriter<D>> {
+    pub fn writer_for_tests<D: Document>(&self) -> crate::Result<IndexWriter<Codec, D>> {
         self.writer_with_num_threads(1, MEMORY_BUDGET_NUM_BYTES_MIN)
     }
 
@@ -613,7 +689,7 @@ impl Index {
     pub fn writer<D: Document>(
         &self,
         memory_budget_in_bytes: usize,
-    ) -> crate::Result<IndexWriter<D>> {
+    ) -> crate::Result<IndexWriter<Codec, D>> {
         let mut num_threads = std::cmp::min(available_parallelism()?.get(), MAX_NUM_THREAD);
         let memory_budget_num_bytes_per_thread = memory_budget_in_bytes / num_threads;
         if memory_budget_num_bytes_per_thread < MEMORY_BUDGET_NUM_BYTES_MIN {
@@ -640,7 +716,7 @@ impl Index {
     }
 
     /// Returns the list of segments that are searchable
-    pub fn searchable_segments(&self) -> crate::Result<Vec<Segment>> {
+    pub fn searchable_segments(&self) -> crate::Result<Vec<Segment<Codec>>> {
         Ok(self
             .searchable_segment_metas()?
             .into_iter()
@@ -649,12 +725,12 @@ impl Index {
     }
 
     #[doc(hidden)]
-    pub fn segment(&self, segment_meta: SegmentMeta) -> Segment {
+    pub fn segment(&self, segment_meta: SegmentMeta) -> Segment<Codec> {
         Segment::for_index(self.clone(), segment_meta)
     }
 
     /// Creates a new segment.
-    pub fn new_segment(&self) -> Segment {
+    pub fn new_segment(&self) -> Segment<Codec> {
         let segment_meta = self
             .inventory
             .new_segment_meta(SegmentId::generate_random(), 0);

src/index/index_meta.rs

@@ -7,6 +7,7 @@ use std::sync::Arc;
 use serde::{Deserialize, Serialize};
 
 use super::SegmentComponent;
+use crate::codec::{Codec, CodecConfiguration};
 use crate::index::SegmentId;
 use crate::schema::Schema;
 use crate::store::Compressor;
@@ -320,6 +321,7 @@ pub struct IndexMeta {
     /// This payload is entirely unused by tantivy.
     #[serde(skip_serializing_if = "Option::is_none")]
     pub payload: Option<String>,
+    pub codec: CodecConfiguration,
 }
 
 #[derive(Deserialize, Debug)]
@@ -331,6 +333,8 @@ struct UntrackedIndexMeta {
     pub opstamp: Opstamp,
     #[serde(skip_serializing_if = "Option::is_none")]
     pub payload: Option<String>,
+    #[serde(default)]
+    pub codec: CodecConfiguration,
 }
 
 impl UntrackedIndexMeta {
@@ -345,6 +349,7 @@ impl UntrackedIndexMeta {
             schema: self.schema,
             opstamp: self.opstamp,
             payload: self.payload,
+            codec: self.codec,
         }
     }
 }
@@ -355,13 +360,14 @@ impl IndexMeta {
     ///
     /// This new index does not contains any segments.
     /// Opstamp will the value `0u64`.
-    pub fn with_schema(schema: Schema) -> IndexMeta {
+    pub fn with_schema_and_codec<C: Codec>(schema: Schema, codec: &C) -> IndexMeta {
         IndexMeta {
             index_settings: IndexSettings::default(),
             segments: vec![],
             schema,
             opstamp: 0u64,
             payload: None,
+            codec: CodecConfiguration::from_codec(codec),
         }
     }
 
@@ -404,16 +410,20 @@ mod tests {
             schema_builder.build()
         };
         let index_metas = IndexMeta {
-            index_settings: IndexSettings::default(),
+            index_settings: IndexSettings {
+                docstore_compression: Compressor::None,
+                ..Default::default()
+            },
             segments: Vec::new(),
             schema,
             opstamp: 0u64,
             payload: None,
+            codec: Default::default(),
         };
         let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
         assert_eq!(
             json,
-            r#"{"index_settings":{"docstore_compression":"lz4","docstore_blocksize":16384},"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","fieldnorms":true,"tokenizer":"default"},"stored":false,"fast":false}}],"opstamp":0}"#
+            r#"{"index_settings":{"docstore_compression":"none","docstore_blocksize":16384},"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","fieldnorms":true,"tokenizer":"default"},"stored":false,"fast":false}}],"opstamp":0,"codec":{"name":"standard"}}"#
         );
 
         let deser_meta: UntrackedIndexMeta = serde_json::from_str(&json).unwrap();
@@ -494,6 +504,8 @@ mod tests {
     #[test]
     #[cfg(feature = "lz4-compression")]
     fn test_index_settings_default() {
+        use crate::store::Compressor;
+
         let mut index_settings = IndexSettings::default();
         assert_eq!(
             index_settings,

src/index/inverted_index_reader.rs

@@ -1,4 +1,5 @@
 use std::io;
+use std::sync::Arc;
 
 use common::json_path_writer::JSON_END_OF_PATH;
 use common::{BinarySerializable, ByteCount};
@@ -9,12 +10,11 @@ use itertools::Itertools;
 #[cfg(feature = "quickwit")]
 use tantivy_fst::automaton::{AlwaysMatch, Automaton};
 
+use crate::codec::{ObjectSafeCodec, StandardCodec};
 use crate::directory::FileSlice;
-use crate::positions::PositionReader;
-use crate::postings::{BlockSegmentPostings, SegmentPostings, TermInfo};
+use crate::postings::{Postings, TermInfo};
 use crate::schema::{IndexRecordOption, Term, Type};
 use crate::termdict::TermDictionary;
-use crate::DocId;
 
 /// The inverted index reader is in charge of accessing
 /// the inverted index associated with a specific field.
@@ -34,6 +34,7 @@ pub struct InvertedIndexReader {
     positions_file_slice: FileSlice,
     record_option: IndexRecordOption,
     total_num_tokens: u64,
+    codec: Arc<dyn ObjectSafeCodec>,
 }
 
 /// Object that records the amount of space used by a field in an inverted index.
@@ -69,6 +70,7 @@ impl InvertedIndexReader {
         postings_file_slice: FileSlice,
         positions_file_slice: FileSlice,
         record_option: IndexRecordOption,
+        codec: Arc<dyn ObjectSafeCodec>,
     ) -> io::Result<InvertedIndexReader> {
         let (total_num_tokens_slice, postings_body) = postings_file_slice.split(8);
         let total_num_tokens = u64::deserialize(&mut total_num_tokens_slice.read_bytes()?)?;
@@ -78,6 +80,7 @@ impl InvertedIndexReader {
             positions_file_slice,
             record_option,
             total_num_tokens,
+            codec,
         })
     }
 
@@ -90,6 +93,7 @@ impl InvertedIndexReader {
             positions_file_slice: FileSlice::empty(),
             record_option,
             total_num_tokens: 0u64,
+            codec: Arc::new(StandardCodec),
         }
     }
 
@@ -161,81 +165,26 @@ impl InvertedIndexReader {
         Ok(fields)
     }
 
-    /// Resets the block segment to another position of the postings
-    /// file.
-    ///
-    /// This is useful for enumerating through a list of terms,
-    /// and consuming the associated posting lists while avoiding
-    /// reallocating a [`BlockSegmentPostings`].
-    ///
-    /// # Warning
-    ///
-    /// This does not reset the positions list.
-    pub fn reset_block_postings_from_terminfo(
-        &self,
-        term_info: &TermInfo,
-        block_postings: &mut BlockSegmentPostings,
-    ) -> io::Result<()> {
-        let postings_slice = self
-            .postings_file_slice
-            .slice(term_info.postings_range.clone());
-        let postings_bytes = postings_slice.read_bytes()?;
-        block_postings.reset(term_info.doc_freq, postings_bytes)?;
-        Ok(())
-    }
-
-    /// Returns a block postings given a `Term`.
-    /// This method is for an advanced usage only.
-    ///
-    /// Most users should prefer using [`Self::read_postings()`] instead.
-    pub fn read_block_postings(
-        &self,
-        term: &Term,
-        option: IndexRecordOption,
-    ) -> io::Result<Option<BlockSegmentPostings>> {
-        let Some(term_info) = self.get_term_info(term)? else {
-            return Ok(None);
-        };
-        let block_postings_not_loaded =
-            self.read_block_postings_from_terminfo(&term_info, option)?;
-        Ok(Some(block_postings_not_loaded))
-    }
-
-    /// Returns a block postings given a `term_info`.
-    /// This method is for an advanced usage only.
-    ///
-    /// Most users should prefer using [`Self::read_postings()`] instead.
-    pub(crate) fn read_block_postings_from_terminfo_with_seek(
-        &self,
-        term_info: &TermInfo,
-        requested_option: IndexRecordOption,
-        seek_doc: DocId,
-    ) -> io::Result<(BlockSegmentPostings, usize)> {
-        let postings_data = self
-            .postings_file_slice
-            .slice(term_info.postings_range.clone());
-        BlockSegmentPostings::open(
-            term_info.doc_freq,
-            postings_data,
-            self.record_option,
-            requested_option,
-            seek_doc,
-        )
-    }
-
-    /// Returns a block postings given a `term_info`.
-    /// This method is for an advanced usage only.
-    ///
-    /// Most users should prefer using [`Self::read_postings()`] instead.
-    pub fn read_block_postings_from_terminfo(
-        &self,
-        term_info: &TermInfo,
-        requested_option: IndexRecordOption,
-    ) -> io::Result<BlockSegmentPostings> {
-        let (block_segment_postings, _) =
-            self.read_block_postings_from_terminfo_with_seek(term_info, requested_option, 0)?;
-        Ok(block_segment_postings)
-    }
+    // /// Returns a block postings given a `term_info`.
+    // /// This method is for an advanced usage only.
+    // ///
+    // /// Most users should prefer using [`Self::read_postings()`] instead.
+    // pub fn read_block_postings_from_terminfo(
+    //     &self,
+    //     term_info: &TermInfo,
+    //     requested_option: IndexRecordOption,
+    // ) -> io::Result<BlockSegmentPostings> {
+    //     let postings_data = self
+    //         .postings_file_slice
+    //         .slice(term_info.postings_range.clone())
+    //         .read_bytes()?;
+    //     BlockSegmentPostings::open(
+    //         term_info.doc_freq,
+    //         postings_data,
+    //         self.record_option,
+    //         requested_option,
+    //     )
+    // }
 
     /// Returns a posting object given a `term_info`.
     /// This method is for an advanced usage only.
@@ -244,27 +193,33 @@ impl InvertedIndexReader {
     pub fn read_postings_from_terminfo(
         &self,
         term_info: &TermInfo,
-        record_option: IndexRecordOption,
-        seek_doc: DocId,
-    ) -> io::Result<SegmentPostings> {
-        let (block_segment_postings, position_within_block) =
-            self.read_block_postings_from_terminfo_with_seek(term_info, record_option, seek_doc)?;
-        let position_reader = {
-            if record_option.has_positions() {
+        option: IndexRecordOption,
+    ) -> io::Result<Box<dyn Postings>> {
+        let option = option.downgrade(self.record_option);
+
+        let postings_data = self
+            .postings_file_slice
+            .slice(term_info.postings_range.clone())
+            .read_bytes()?;
+        let positions_data = {
+            if option.has_positions() {
                 let positions_data = self
                     .positions_file_slice
-                    .read_bytes_slice(term_info.positions_range.clone())?;
-                let position_reader = PositionReader::open(positions_data)?;
-                Some(position_reader)
+                    .slice(term_info.positions_range.clone())
+                    .read_bytes()?;
+                Some(positions_data)
             } else {
                 None
             }
         };
-        Ok(SegmentPostings::from_block_postings(
-            block_segment_postings,
-            position_reader,
-            position_within_block,
-        ))
+        let postings = self.codec.load_postings_type_erased(
+            term_info.doc_freq,
+            postings_data,
+            self.record_option,
+            option,
+            positions_data,
+        )?;
+        Ok(postings)
     }
 
     /// Returns the total number of tokens recorded for all documents
@@ -287,9 +242,9 @@ impl InvertedIndexReader {
         &self,
         term: &Term,
         option: IndexRecordOption,
-    ) -> io::Result<Option<SegmentPostings>> {
+    ) -> io::Result<Option<Box<dyn Postings>>> {
         self.get_term_info(term)?
-            .map(move |term_info| self.read_postings_from_terminfo(&term_info, option, 0u32))
+            .map(move |term_info| self.read_postings_from_terminfo(&term_info, option))
             .transpose()
     }
 

src/index/segment.rs

@@ -2,6 +2,7 @@ use std::fmt;
 use std::path::PathBuf;
 
 use super::SegmentComponent;
+use crate::codec::StandardCodec;
 use crate::directory::error::{OpenReadError, OpenWriteError};
 use crate::directory::{Directory, FileSlice, WritePtr};
 use crate::index::{Index, SegmentId, SegmentMeta};
@@ -10,25 +11,25 @@ use crate::Opstamp;
 
 /// A segment is a piece of the index.
 #[derive(Clone)]
-pub struct Segment {
-    index: Index,
+pub struct Segment<C: crate::codec::Codec = StandardCodec> {
+    index: Index<C>,
     meta: SegmentMeta,
 }
 
-impl fmt::Debug for Segment {
-    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
+impl<C: crate::codec::Codec> fmt::Debug for Segment<C> {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         write!(f, "Segment({:?})", self.id().uuid_string())
     }
 }
 
-impl Segment {
+impl<C: crate::codec::Codec> Segment<C> {
     /// Creates a new segment given an `Index` and a `SegmentId`
-    pub(crate) fn for_index(index: Index, meta: SegmentMeta) -> Segment {
+    pub(crate) fn for_index(index: Index<C>, meta: SegmentMeta) -> Segment<C> {
         Segment { index, meta }
     }
 
     /// Returns the index the segment belongs to.
-    pub fn index(&self) -> &Index {
+    pub fn index(&self) -> &Index<C> {
         &self.index
     }
 
@@ -46,7 +47,7 @@ impl Segment {
     ///
     /// This method is only used when updating `max_doc` from 0
     /// as we finalize a fresh new segment.
-    pub fn with_max_doc(self, max_doc: u32) -> Segment {
+    pub fn with_max_doc(self, max_doc: u32) -> Segment<C> {
         Segment {
             index: self.index,
             meta: self.meta.with_max_doc(max_doc),
@@ -55,7 +56,7 @@ impl Segment {
 
     #[doc(hidden)]
     #[must_use]
-    pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment {
+    pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment<C> {
         Segment {
             index: self.index,
             meta: self.meta.with_delete_meta(num_deleted_docs, opstamp),

src/index/segment_reader.rs

@@ -6,6 +6,7 @@ use common::{ByteCount, HasLen};
 use fnv::FnvHashMap;
 use itertools::Itertools;
 
+use crate::codec::ObjectSafeCodec;
 use crate::directory::{CompositeFile, FileSlice};
 use crate::error::DataCorruption;
 use crate::fastfield::{intersect_alive_bitsets, AliveBitSet, FacetReader, FastFieldReaders};
@@ -47,6 +48,8 @@ pub struct SegmentReader {
     store_file: FileSlice,
     alive_bitset_opt: Option<AliveBitSet>,
     schema: Schema,
+
+    codec: Arc<dyn ObjectSafeCodec>,
 }
 
 impl SegmentReader {
@@ -140,15 +143,16 @@ impl SegmentReader {
     }
 
     /// Open a new segment for reading.
-    pub fn open(segment: &Segment) -> crate::Result<SegmentReader> {
+    pub fn open<C: crate::codec::Codec>(segment: &Segment<C>) -> crate::Result<SegmentReader> {
         Self::open_with_custom_alive_set(segment, None)
     }
 
     /// Open a new segment for reading.
-    pub fn open_with_custom_alive_set(
-        segment: &Segment,
+    pub fn open_with_custom_alive_set<C: crate::codec::Codec>(
+        segment: &Segment<C>,
         custom_bitset: Option<AliveBitSet>,
     ) -> crate::Result<SegmentReader> {
+        let codec: Arc<dyn ObjectSafeCodec> = Arc::new(segment.index().codec().clone());
         let termdict_file = segment.open_read(SegmentComponent::Terms)?;
         let termdict_composite = CompositeFile::open(&termdict_file)?;
 
@@ -204,6 +208,7 @@ impl SegmentReader {
             alive_bitset_opt,
             positions_composite,
             schema,
+            codec,
         })
     }
 
@@ -273,6 +278,7 @@ impl SegmentReader {
             postings_file,
             positions_file,
             record_option,
+            self.codec.clone(),
         )?);
 
         // by releasing the lock in between, we may end up opening the inverting index

src/indexer/delete_queue.rs

@@ -4,19 +4,20 @@ use std::sync::{Arc, RwLock, Weak};
 use super::operation::DeleteOperation;
 use crate::Opstamp;
 
-// The DeleteQueue is similar in conceptually to a multiple
-// consumer single producer broadcast channel.
-//
-// All consumer will receive all messages.
-//
-// Consumer of the delete queue are holding a `DeleteCursor`,
-// which points to a specific place of the `DeleteQueue`.
-//
-// New consumer can be created in two ways
-// - calling `delete_queue.cursor()` returns a cursor, that will include all future delete operation
-//   (and some or none of the past operations... The client is in charge of checking the opstamps.).
-// - cloning an existing cursor returns a new cursor, that is at the exact same position, and can
-//   now advance independently from the original cursor.
+/// The DeleteQueue is similar in conceptually to a multiple
+/// consumer single producer broadcast channel.
+///
+/// All consumer will receive all messages.
+///
+/// Consumer of the delete queue are holding a `DeleteCursor`,
+/// which points to a specific place of the `DeleteQueue`.
+///
+/// New consumer can be created in two ways
+/// - calling `delete_queue.cursor()` returns a cursor, that will include all future delete
+///   operation (and some or none of the past operations... The client is in charge of checking the
+///   opstamps.).
+/// - cloning an existing cursor returns a new cursor, that is at the exact same position, and can
+///   now advance independently from the original cursor.
 #[derive(Default)]
 struct InnerDeleteQueue {
     writer: Vec<DeleteOperation>,
@@ -249,12 +250,7 @@ mod tests {
 
     struct DummyWeight;
     impl Weight for DummyWeight {
-        fn scorer(
-            &self,
-            _reader: &SegmentReader,
-            _boost: Score,
-            _seek_doc: DocId,
-        ) -> crate::Result<Box<dyn Scorer>> {
+        fn scorer(&self, _reader: &SegmentReader, _boost: Score) -> crate::Result<Box<dyn Scorer>> {
             Err(crate::TantivyError::InternalError("dummy impl".to_owned()))
         }
 

src/indexer/index_writer.rs

@@ -9,6 +9,7 @@ use smallvec::smallvec;
 use super::operation::{AddOperation, UserOperation};
 use super::segment_updater::SegmentUpdater;
 use super::{AddBatch, AddBatchReceiver, AddBatchSender, PreparedCommit};
+use crate::codec::{Codec, StandardCodec};
 use crate::directory::{DirectoryLock, GarbageCollectionResult, TerminatingWrite};
 use crate::error::TantivyError;
 use crate::fastfield::write_alive_bitset;
@@ -68,12 +69,12 @@ pub struct IndexWriterOptions {
 /// indexing queue.
 /// Each indexing thread builds its own independent [`Segment`], via
 /// a `SegmentWriter` object.
-pub struct IndexWriter<D: Document = TantivyDocument> {
+pub struct IndexWriter<C: Codec = StandardCodec, D: Document = TantivyDocument> {
     // the lock is just used to bind the
     // lifetime of the lock with that of the IndexWriter.
     _directory_lock: Option<DirectoryLock>,
 
-    index: Index,
+    index: Index<C>,
 
     options: IndexWriterOptions,
 
@@ -82,7 +83,7 @@ pub struct IndexWriter<D: Document = TantivyDocument> {
     index_writer_status: IndexWriterStatus<D>,
     operation_sender: AddBatchSender<D>,
 
-    segment_updater: SegmentUpdater,
+    segment_updater: SegmentUpdater<C>,
 
     worker_id: usize,
 
@@ -128,8 +129,8 @@ fn compute_deleted_bitset(
 /// is `==` target_opstamp.
 /// For instance, there was no delete operation between the state of the `segment_entry` and
 /// the `target_opstamp`, `segment_entry` is not updated.
-pub fn advance_deletes(
-    mut segment: Segment,
+pub fn advance_deletes<C: Codec>(
+    mut segment: Segment<C>,
     segment_entry: &mut SegmentEntry,
     target_opstamp: Opstamp,
 ) -> crate::Result<()> {
@@ -179,11 +180,11 @@ pub fn advance_deletes(
     Ok(())
 }
 
-fn index_documents<D: Document>(
+fn index_documents<C: crate::codec::Codec, D: Document>(
     memory_budget: usize,
-    segment: Segment,
+    segment: Segment<C>,
     grouped_document_iterator: &mut dyn Iterator<Item = AddBatch<D>>,
-    segment_updater: &SegmentUpdater,
+    segment_updater: &SegmentUpdater<C>,
     mut delete_cursor: DeleteCursor,
 ) -> crate::Result<()> {
     let mut segment_writer = SegmentWriter::for_segment(memory_budget, segment.clone())?;
@@ -226,8 +227,8 @@ fn index_documents<D: Document>(
 }
 
 /// `doc_opstamps` is required to be non-empty.
-fn apply_deletes(
-    segment: &Segment,
+fn apply_deletes<C: crate::codec::Codec>(
+    segment: &Segment<C>,
     delete_cursor: &mut DeleteCursor,
     doc_opstamps: &[Opstamp],
 ) -> crate::Result<Option<BitSet>> {
@@ -262,7 +263,7 @@ fn apply_deletes(
     })
 }
 
-impl<D: Document> IndexWriter<D> {
+impl<C: Codec, D: Document> IndexWriter<C, D> {
     /// Create a new index writer. Attempts to acquire a lockfile.
     ///
     /// The lockfile should be deleted on drop, but it is possible
@@ -278,7 +279,7 @@ impl<D: Document> IndexWriter<D> {
     /// If the memory arena per thread is too small or too big, returns
     /// `TantivyError::InvalidArgument`
     pub(crate) fn new(
-        index: &Index,
+        index: &Index<C>,
         options: IndexWriterOptions,
         directory_lock: DirectoryLock,
     ) -> crate::Result<Self> {
@@ -345,7 +346,7 @@ impl<D: Document> IndexWriter<D> {
     }
 
     /// Accessor to the index.
-    pub fn index(&self) -> &Index {
+    pub fn index(&self) -> &Index<C> {
         &self.index
     }
 
@@ -393,7 +394,7 @@ impl<D: Document> IndexWriter<D> {
     /// It is safe to start writing file associated with the new `Segment`.
     /// These will not be garbage collected as long as an instance object of
     /// `SegmentMeta` object associated with the new `Segment` is "alive".
-    pub fn new_segment(&self) -> Segment {
+    pub fn new_segment(&self) -> Segment<C> {
         self.index.new_segment()
     }
 
@@ -615,7 +616,7 @@ impl<D: Document> IndexWriter<D> {
     /// It is also possible to add a payload to the `commit`
     /// using this API.
     /// See [`PreparedCommit::set_payload()`].
-    pub fn prepare_commit(&mut self) -> crate::Result<PreparedCommit<'_, D>> {
+    pub fn prepare_commit(&mut self) -> crate::Result<PreparedCommit<'_, C, D>> {
         // Here, because we join all of the worker threads,
         // all of the segment update for this commit have been
         // sent.
@@ -665,7 +666,7 @@ impl<D: Document> IndexWriter<D> {
         self.prepare_commit()?.commit()
     }
 
-    pub(crate) fn segment_updater(&self) -> &SegmentUpdater {
+    pub(crate) fn segment_updater(&self) -> &SegmentUpdater<C> {
         &self.segment_updater
     }
 
@@ -804,7 +805,7 @@ impl<D: Document> IndexWriter<D> {
     }
 }
 
-impl<D: Document> Drop for IndexWriter<D> {
+impl<C: Codec, D: Document> Drop for IndexWriter<C, D> {
     fn drop(&mut self) {
         self.segment_updater.kill();
         self.drop_sender();

src/indexer/merger.rs

@@ -7,6 +7,7 @@ use common::ReadOnlyBitSet;
 use itertools::Itertools;
 use measure_time::debug_time;
 
+use crate::codec::{Codec, StandardCodec};
 use crate::directory::WritePtr;
 use crate::docset::{DocSet, TERMINATED};
 use crate::error::DataCorruption;
@@ -15,7 +16,7 @@ use crate::fieldnorm::{FieldNormReader, FieldNormReaders, FieldNormsSerializer,
 use crate::index::{Segment, SegmentComponent, SegmentReader};
 use crate::indexer::doc_id_mapping::{MappingType, SegmentDocIdMapping};
 use crate::indexer::SegmentSerializer;
-use crate::postings::{InvertedIndexSerializer, Postings, SegmentPostings};
+use crate::postings::{FreqReadingOption, InvertedIndexSerializer, Postings};
 use crate::schema::{value_type_to_column_type, Field, FieldType, Schema};
 use crate::store::StoreWriter;
 use crate::termdict::{TermMerger, TermOrdinal};
@@ -76,10 +77,11 @@ fn estimate_total_num_tokens(readers: &[SegmentReader], field: Field) -> crate::
     Ok(total_num_tokens)
 }
 
-pub struct IndexMerger {
+pub struct IndexMerger<C: Codec = StandardCodec> {
     schema: Schema,
     pub(crate) readers: Vec<SegmentReader>,
     max_doc: u32,
+    codec: C,
 }
 
 struct DeltaComputer {
@@ -144,8 +146,8 @@ fn extract_fast_field_required_columns(schema: &Schema) -> Vec<(String, ColumnTy
         .collect()
 }
 
-impl IndexMerger {
-    pub fn open(schema: Schema, segments: &[Segment]) -> crate::Result<IndexMerger> {
+impl<C: Codec> IndexMerger<C> {
+    pub fn open(schema: Schema, segments: &[Segment<C>]) -> crate::Result<IndexMerger<C>> {
         let alive_bitset = segments.iter().map(|_| None).collect_vec();
         Self::open_with_custom_alive_set(schema, segments, alive_bitset)
     }
@@ -162,11 +164,15 @@ impl IndexMerger {
     // This can be used to merge but also apply an additional filter.
     // One use case is demux, which is basically taking a list of
     // segments and partitions them e.g. by a value in a field.
+    //
+    // # Panics if segments is empty.
     pub fn open_with_custom_alive_set(
         schema: Schema,
-        segments: &[Segment],
+        segments: &[Segment<C>],
         alive_bitset_opt: Vec<Option<AliveBitSet>>,
-    ) -> crate::Result<IndexMerger> {
+    ) -> crate::Result<IndexMerger<C>> {
+        assert!(!segments.is_empty());
+        let codec = segments[0].index().codec().clone();
         let mut readers = vec![];
         for (segment, new_alive_bitset_opt) in segments.iter().zip(alive_bitset_opt) {
             if segment.meta().num_docs() > 0 {
@@ -189,6 +195,7 @@ impl IndexMerger {
             schema,
             readers,
             max_doc,
+            codec,
         })
     }
 
@@ -287,7 +294,7 @@ impl IndexMerger {
         &self,
         indexed_field: Field,
         _field_type: &FieldType,
-        serializer: &mut InvertedIndexSerializer,
+        serializer: &mut InvertedIndexSerializer<C>,
         fieldnorm_reader: Option<FieldNormReader>,
         doc_id_mapping: &SegmentDocIdMapping,
     ) -> crate::Result<()> {
@@ -355,7 +362,7 @@ impl IndexMerger {
                          indexed. Have you modified the schema?",
         );
 
-        let mut segment_postings_containing_the_term: Vec<(usize, SegmentPostings)> = vec![];
+        let mut segment_postings_containing_the_term: Vec<(usize, Box<dyn Postings>)> = vec![];
 
         while merged_terms.advance() {
             segment_postings_containing_the_term.clear();
@@ -367,11 +374,8 @@ impl IndexMerger {
             for (segment_ord, term_info) in merged_terms.current_segment_ords_and_term_infos() {
                 let segment_reader = &self.readers[segment_ord];
                 let inverted_index: &InvertedIndexReader = &field_readers[segment_ord];
-                let segment_postings = inverted_index.read_postings_from_terminfo(
-                    &term_info,
-                    segment_postings_option,
-                    0u32,
-                )?;
+                let segment_postings = inverted_index
+                    .read_postings_from_terminfo(&term_info, segment_postings_option)?;
                 let alive_bitset_opt = segment_reader.alive_bitset();
                 let doc_freq = if let Some(alive_bitset) = alive_bitset_opt {
                     segment_postings.doc_freq_given_deletes(alive_bitset)
@@ -398,11 +402,10 @@ impl IndexMerger {
             assert!(!segment_postings_containing_the_term.is_empty());
 
             let has_term_freq = {
-                let has_term_freq = !segment_postings_containing_the_term[0]
+                let has_term_freq = segment_postings_containing_the_term[0]
                     .1
-                    .block_cursor
-                    .freqs()
-                    .is_empty();
+                    .freq_reading_option()
+                    == FreqReadingOption::ReadFreq;
                 for (_, postings) in &segment_postings_containing_the_term[1..] {
                     // This may look at a strange way to test whether we have term freq or not.
                     // With JSON object, the schema is not sufficient to know whether a term
@@ -418,7 +421,7 @@ impl IndexMerger {
                     //
                     // Overall the reliable way to know if we have actual frequencies loaded or not
                     // is to check whether the actual decoded array is empty or not.
-                    if has_term_freq == postings.block_cursor.freqs().is_empty() {
+                    if postings.freq_reading_option() != FreqReadingOption::ReadFreq {
                         return Err(DataCorruption::comment_only(
                             "Term freqs are inconsistent across segments",
                         )
@@ -470,7 +473,7 @@ impl IndexMerger {
 
     fn write_postings(
         &self,
-        serializer: &mut InvertedIndexSerializer,
+        serializer: &mut InvertedIndexSerializer<C>,
         fieldnorm_readers: FieldNormReaders,
         doc_id_mapping: &SegmentDocIdMapping,
     ) -> crate::Result<()> {
@@ -528,7 +531,7 @@ impl IndexMerger {
     ///
     /// # Returns
     /// The number of documents in the resulting segment.
-    pub fn write(&self, mut serializer: SegmentSerializer) -> crate::Result<u32> {
+    pub fn write(&self, mut serializer: SegmentSerializer<C>) -> crate::Result<u32> {
         let doc_id_mapping = self.get_doc_id_from_concatenated_data()?;
         debug!("write-fieldnorms");
         if let Some(fieldnorms_serializer) = serializer.extract_fieldnorms_serializer() {
@@ -1524,7 +1527,7 @@ mod tests {
             .term_scorer_for_test(searcher.segment_reader(0u32), 1.0)?
             .unwrap();
         assert_eq!(term_scorer.doc(), 0);
-        assert_nearly_equals!(term_scorer.block_max_score(), 0.0079681855);
+        assert_nearly_equals!(term_scorer.seek_block(0), 0.0079681855);
         assert_nearly_equals!(term_scorer.score(), 0.0079681855);
         for _ in 0..81 {
             writer.add_document(doc!(text=>"hello happy tax payer"))?;
@@ -1543,7 +1546,7 @@ mod tests {
             // there.
             for doc in segment_reader.doc_ids_alive() {
                 assert_eq!(term_scorer.doc(), doc);
-                assert_nearly_equals!(term_scorer.block_max_score(), 0.003478312);
+                assert_nearly_equals!(term_scorer.seek_block(doc), 0.003478312);
                 assert_nearly_equals!(term_scorer.score(), 0.003478312);
                 term_scorer.advance();
             }
@@ -1568,7 +1571,7 @@ mod tests {
         // the difference compared to before is intrinsic to the bm25 formula. no worries there.
         for doc in segment_reader.doc_ids_alive() {
             assert_eq!(term_scorer.doc(), doc);
-            assert_nearly_equals!(term_scorer.block_max_score(), 0.003478312);
+            assert_nearly_equals!(term_scorer.seek_block(doc), 0.003478312);
             assert_nearly_equals!(term_scorer.score(), 0.003478312);
             term_scorer.advance();
         }

src/indexer/mod.rs

@@ -4,6 +4,7 @@
 //! `IndexWriter` is the main entry point for that, which created from
 //! [`Index::writer`](crate::Index::writer).
 
+/// Delete queue implementation for broadcasting delete operations to consumers.
 pub(crate) mod delete_queue;
 pub(crate) mod path_to_unordered_id;
 

src/indexer/prepared_commit.rs

@@ -1,16 +1,17 @@
 use super::IndexWriter;
+use crate::codec::Codec;
 use crate::schema::document::Document;
 use crate::{FutureResult, Opstamp, TantivyDocument};
 
 /// A prepared commit
-pub struct PreparedCommit<'a, D: Document = TantivyDocument> {
-    index_writer: &'a mut IndexWriter<D>,
+pub struct PreparedCommit<'a, C: Codec, D: Document = TantivyDocument> {
+    index_writer: &'a mut IndexWriter<C, D>,
     payload: Option<String>,
     opstamp: Opstamp,
 }
 
-impl<'a, D: Document> PreparedCommit<'a, D> {
-    pub(crate) fn new(index_writer: &'a mut IndexWriter<D>, opstamp: Opstamp) -> Self {
+impl<'a, C: Codec, D: Document> PreparedCommit<'a, C, D> {
+    pub(crate) fn new(index_writer: &'a mut IndexWriter<C, D>, opstamp: Opstamp) -> Self {
         Self {
             index_writer,
             payload: None,

src/indexer/segment_serializer.rs

@@ -8,17 +8,17 @@ use crate::store::StoreWriter;
 
 /// Segment serializer is in charge of laying out on disk
 /// the data accumulated and sorted by the `SegmentWriter`.
-pub struct SegmentSerializer {
-    segment: Segment,
+pub struct SegmentSerializer<C: crate::codec::Codec> {
+    segment: Segment<C>,
     pub(crate) store_writer: StoreWriter,
     fast_field_write: WritePtr,
     fieldnorms_serializer: Option<FieldNormsSerializer>,
-    postings_serializer: InvertedIndexSerializer,
+    postings_serializer: InvertedIndexSerializer<C>,
 }
 
-impl SegmentSerializer {
+impl<C: crate::codec::Codec> SegmentSerializer<C> {
     /// Creates a new `SegmentSerializer`.
-    pub fn for_segment(mut segment: Segment) -> crate::Result<SegmentSerializer> {
+    pub fn for_segment(mut segment: Segment<C>) -> crate::Result<SegmentSerializer<C>> {
         let settings = segment.index().settings().clone();
         let store_writer = {
             let store_write = segment.open_write(SegmentComponent::Store)?;
@@ -50,12 +50,12 @@ impl SegmentSerializer {
         self.store_writer.mem_usage()
     }
 
-    pub fn segment(&self) -> &Segment {
+    pub fn segment(&self) -> &Segment<C> {
         &self.segment
     }
 
     /// Accessor to the `PostingsSerializer`.
-    pub fn get_postings_serializer(&mut self) -> &mut InvertedIndexSerializer {
+    pub fn get_postings_serializer(&mut self) -> &mut InvertedIndexSerializer<C> {
         &mut self.postings_serializer
     }
 

src/indexer/segment_updater.rs

@@ -10,6 +10,7 @@ use std::sync::{Arc, RwLock};
 use rayon::{ThreadPool, ThreadPoolBuilder};
 
 use super::segment_manager::SegmentManager;
+use crate::codec::{Codec, CodecConfiguration};
 use crate::core::META_FILEPATH;
 use crate::directory::{Directory, DirectoryClone, GarbageCollectionResult};
 use crate::fastfield::AliveBitSet;
@@ -61,10 +62,10 @@ pub(crate) fn save_metas(metas: &IndexMeta, directory: &dyn Directory) -> crate:
 // We voluntarily pass a merge_operation ref to guarantee that
 // the merge_operation is alive during the process
 #[derive(Clone)]
-pub(crate) struct SegmentUpdater(Arc<InnerSegmentUpdater>);
+pub(crate) struct SegmentUpdater<C: Codec>(Arc<InnerSegmentUpdater<C>>);
 
-impl Deref for SegmentUpdater {
-    type Target = InnerSegmentUpdater;
+impl<C: Codec> Deref for SegmentUpdater<C> {
+    type Target = InnerSegmentUpdater<C>;
 
     #[inline]
     fn deref(&self) -> &Self::Target {
@@ -72,8 +73,8 @@ impl Deref for SegmentUpdater {
     }
 }
 
-fn garbage_collect_files(
-    segment_updater: SegmentUpdater,
+fn garbage_collect_files<C: Codec>(
+    segment_updater: SegmentUpdater<C>,
 ) -> crate::Result<GarbageCollectionResult> {
     info!("Running garbage collection");
     let mut index = segment_updater.index.clone();
@@ -84,8 +85,8 @@ fn garbage_collect_files(
 
 /// Merges a list of segments the list of segment givens in the `segment_entries`.
 /// This function happens in the calling thread and is computationally expensive.
-fn merge(
-    index: &Index,
+fn merge<Codec: crate::codec::Codec>(
+    index: &Index<Codec>,
     mut segment_entries: Vec<SegmentEntry>,
     target_opstamp: Opstamp,
 ) -> crate::Result<Option<SegmentEntry>> {
@@ -108,13 +109,13 @@ fn merge(
 
     let delete_cursor = segment_entries[0].delete_cursor().clone();
 
-    let segments: Vec<Segment> = segment_entries
+    let segments: Vec<Segment<Codec>> = segment_entries
         .iter()
         .map(|segment_entry| index.segment(segment_entry.meta().clone()))
         .collect();
 
     // An IndexMerger is like a "view" of our merged segments.
-    let merger: IndexMerger = IndexMerger::open(index.schema(), &segments[..])?;
+    let merger: IndexMerger<Codec> = IndexMerger::open(index.schema(), &segments[..])?;
 
     // ... we just serialize this index merger in our new segment to merge the segments.
     let segment_serializer = SegmentSerializer::for_segment(merged_segment.clone())?;
@@ -139,10 +140,10 @@ fn merge(
 /// meant to work if you have an `IndexWriter` running for the origin indices, or
 /// the destination `Index`.
 #[doc(hidden)]
-pub fn merge_indices<T: Into<Box<dyn Directory>>>(
-    indices: &[Index],
-    output_directory: T,
-) -> crate::Result<Index> {
+pub fn merge_indices<Codec: crate::codec::Codec>(
+    indices: &[Index<Codec>],
+    output_directory: Box<dyn Directory>,
+) -> crate::Result<Index<Codec>> {
     if indices.is_empty() {
         // If there are no indices to merge, there is no need to do anything.
         return Err(crate::TantivyError::InvalidArgument(
@@ -163,7 +164,7 @@ pub fn merge_indices<T: Into<Box<dyn Directory>>>(
         ));
     }
 
-    let mut segments: Vec<Segment> = Vec::new();
+    let mut segments: Vec<Segment<Codec>> = Vec::new();
     for index in indices {
         segments.extend(index.searchable_segments()?);
     }
@@ -185,12 +186,12 @@ pub fn merge_indices<T: Into<Box<dyn Directory>>>(
 /// meant to work if you have an `IndexWriter` running for the origin indices, or
 /// the destination `Index`.
 #[doc(hidden)]
-pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
-    segments: &[Segment],
+pub fn merge_filtered_segments<C: crate::codec::Codec, T: Into<Box<dyn Directory>>>(
+    segments: &[Segment<C>],
     target_settings: IndexSettings,
     filter_doc_ids: Vec<Option<AliveBitSet>>,
     output_directory: T,
-) -> crate::Result<Index> {
+) -> crate::Result<Index<C>> {
     if segments.is_empty() {
         // If there are no indices to merge, there is no need to do anything.
         return Err(crate::TantivyError::InvalidArgument(
@@ -211,14 +212,15 @@ pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
         ));
     }
 
-    let mut merged_index = Index::create(
-        output_directory,
-        target_schema.clone(),
-        target_settings.clone(),
-    )?;
+    let mut merged_index: Index<C> = Index::builder()
+        .schema(target_schema.clone())
+        .codec(segments[0].index().codec().clone())
+        .settings(target_settings.clone())
+        .create(output_directory.into())?;
+
     let merged_segment = merged_index.new_segment();
     let merged_segment_id = merged_segment.id();
-    let merger: IndexMerger =
+    let merger: IndexMerger<C> =
         IndexMerger::open_with_custom_alive_set(merged_index.schema(), segments, filter_doc_ids)?;
     let segment_serializer = SegmentSerializer::for_segment(merged_segment)?;
     let num_docs = merger.write(segment_serializer)?;
@@ -235,6 +237,7 @@ pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
             ))
             .trim_end()
     );
+    let codec_configuration = CodecConfiguration::from_codec(segments[0].index().codec());
 
     let index_meta = IndexMeta {
         index_settings: target_settings, // index_settings of all segments should be the same
@@ -242,6 +245,7 @@ pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
         schema: target_schema,
         opstamp: 0u64,
         payload: Some(stats),
+        codec: codec_configuration,
     };
 
     // save the meta.json
@@ -250,7 +254,7 @@ pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
     Ok(merged_index)
 }
 
-pub(crate) struct InnerSegmentUpdater {
+pub(crate) struct InnerSegmentUpdater<C: Codec> {
     // we keep a copy of the current active IndexMeta to
     // avoid loading the file every time we need it in the
     // `SegmentUpdater`.
@@ -261,7 +265,7 @@ pub(crate) struct InnerSegmentUpdater {
     pool: ThreadPool,
     merge_thread_pool: ThreadPool,
 
-    index: Index,
+    index: Index<C>,
     segment_manager: SegmentManager,
     merge_policy: RwLock<Arc<dyn MergePolicy>>,
     killed: AtomicBool,
@@ -269,13 +273,13 @@ pub(crate) struct InnerSegmentUpdater {
     merge_operations: MergeOperationInventory,
 }
 
-impl SegmentUpdater {
+impl<Codec: crate::codec::Codec> SegmentUpdater<Codec> {
     pub fn create(
-        index: Index,
+        index: Index<Codec>,
         stamper: Stamper,
         delete_cursor: &DeleteCursor,
         num_merge_threads: usize,
-    ) -> crate::Result<SegmentUpdater> {
+    ) -> crate::Result<Self> {
         let segments = index.searchable_segment_metas()?;
         let segment_manager = SegmentManager::from_segments(segments, delete_cursor);
         let pool = ThreadPoolBuilder::new()
@@ -404,12 +408,14 @@ impl SegmentUpdater {
             //
             // Segment 1 from disk 1, Segment 1 from disk 2, etc.
             committed_segment_metas.sort_by_key(|segment_meta| -(segment_meta.max_doc() as i32));
+            let codec = CodecConfiguration::from_codec(index.codec());
             let index_meta = IndexMeta {
                 index_settings: index.settings().clone(),
                 segments: committed_segment_metas,
                 schema: index.schema(),
                 opstamp,
                 payload: commit_message,
+                codec,
             };
             // TODO add context to the error.
             save_metas(&index_meta, directory.box_clone().borrow_mut())?;
@@ -443,7 +449,7 @@ impl SegmentUpdater {
         opstamp: Opstamp,
         payload: Option<String>,
     ) -> FutureResult<Opstamp> {
-        let segment_updater: SegmentUpdater = self.clone();
+        let segment_updater: SegmentUpdater<Codec> = self.clone();
         self.schedule_task(move || {
             let segment_entries = segment_updater.purge_deletes(opstamp)?;
             segment_updater.segment_manager.commit(segment_entries);
@@ -702,6 +708,7 @@ impl SegmentUpdater {
 #[cfg(test)]
 mod tests {
     use super::merge_indices;
+    use crate::codec::StandardCodec;
     use crate::collector::TopDocs;
     use crate::directory::RamDirectory;
     use crate::fastfield::AliveBitSet;
@@ -915,7 +922,7 @@ mod tests {
 
     #[test]
     fn test_merge_empty_indices_array() {
-        let merge_result = merge_indices(&[], RamDirectory::default());
+        let merge_result = merge_indices::<StandardCodec>(&[], Box::new(RamDirectory::default()));
         assert!(merge_result.is_err());
     }
 
@@ -942,7 +949,10 @@ mod tests {
         };
 
         // mismatched schema index list
-        let result = merge_indices(&[first_index, second_index], RamDirectory::default());
+        let result = merge_indices(
+            &[first_index, second_index],
+            Box::new(RamDirectory::default()),
+        );
         assert!(result.is_err());
 
         Ok(())

src/indexer/segment_writer.rs

@@ -4,6 +4,7 @@ use itertools::Itertools;
 use tokenizer_api::BoxTokenStream;
 
 use super::operation::AddOperation;
+use crate::codec::Codec;
 use crate::fastfield::FastFieldsWriter;
 use crate::fieldnorm::{FieldNormReaders, FieldNormsWriter};
 use crate::index::{Segment, SegmentComponent};
@@ -12,7 +13,7 @@ use crate::indexer::segment_serializer::SegmentSerializer;
 use crate::json_utils::{index_json_value, IndexingPositionsPerPath};
 use crate::postings::{
     compute_table_memory_size, serialize_postings, IndexingContext, IndexingPosition,
-    PerFieldPostingsWriter, PostingsWriter,
+    PerFieldPostingsWriter, PostingsWriter, PostingsWriterEnum,
 };
 use crate::schema::document::{Document, Value};
 use crate::schema::{FieldEntry, FieldType, Schema, DATE_TIME_PRECISION_INDEXED};
@@ -45,11 +46,11 @@ fn compute_initial_table_size(per_thread_memory_budget: usize) -> crate::Result<
 ///
 /// They creates the postings list in anonymous memory.
 /// The segment is laid on disk when the segment gets `finalized`.
-pub struct SegmentWriter {
+pub struct SegmentWriter<Codec: crate::codec::Codec> {
     pub(crate) max_doc: DocId,
     pub(crate) ctx: IndexingContext,
     pub(crate) per_field_postings_writers: PerFieldPostingsWriter,
-    pub(crate) segment_serializer: SegmentSerializer,
+    pub(crate) segment_serializer: SegmentSerializer<Codec>,
     pub(crate) fast_field_writers: FastFieldsWriter,
     pub(crate) fieldnorms_writer: FieldNormsWriter,
     pub(crate) json_path_writer: JsonPathWriter,
@@ -60,7 +61,7 @@ pub struct SegmentWriter {
     schema: Schema,
 }
 
-impl SegmentWriter {
+impl<Codec: crate::codec::Codec> SegmentWriter<Codec> {
     /// Creates a new `SegmentWriter`
     ///
     /// The arguments are defined as follows
@@ -70,7 +71,10 @@ impl SegmentWriter {
     ///   behavior as a memory limit.
     /// - segment: The segment being written
     /// - schema
-    pub fn for_segment(memory_budget_in_bytes: usize, segment: Segment) -> crate::Result<Self> {
+    pub fn for_segment(
+        memory_budget_in_bytes: usize,
+        segment: Segment<Codec>,
+    ) -> crate::Result<Self> {
         let schema = segment.schema();
         let tokenizer_manager = segment.index().tokenizers().clone();
         let tokenizer_manager_fast_field = segment.index().fast_field_tokenizer().clone();
@@ -169,7 +173,7 @@ impl SegmentWriter {
             }
 
             let (term_buffer, ctx) = (&mut self.term_buffer, &mut self.ctx);
-            let postings_writer: &mut dyn PostingsWriter =
+            let postings_writer: &mut PostingsWriterEnum =
                 self.per_field_postings_writers.get_for_field_mut(field);
             term_buffer.clear_with_field(field);
 
@@ -386,13 +390,13 @@ impl SegmentWriter {
 /// to the `SegmentSerializer`.
 ///
 /// `doc_id_map` is used to map to the new doc_id order.
-fn remap_and_write(
+fn remap_and_write<C: Codec>(
     schema: Schema,
     per_field_postings_writers: &PerFieldPostingsWriter,
     ctx: IndexingContext,
     fast_field_writers: FastFieldsWriter,
     fieldnorms_writer: &FieldNormsWriter,
-    mut serializer: SegmentSerializer,
+    mut serializer: SegmentSerializer<C>,
 ) -> crate::Result<()> {
     debug!("remap-and-write");
     if let Some(fieldnorms_serializer) = serializer.extract_fieldnorms_serializer() {
@@ -421,10 +425,9 @@ fn remap_and_write(
 #[cfg(test)]
 mod tests {
     use std::collections::BTreeMap;
-    use std::path::{Path, PathBuf};
+    use std::path::Path;
 
     use columnar::ColumnType;
-    use tempfile::TempDir;
 
     use crate::collector::{Count, TopDocs};
     use crate::directory::RamDirectory;
@@ -1067,10 +1070,7 @@ mod tests {
         let mut schema_builder = Schema::builder();
         schema_builder.add_text_field("title", text_options);
         let schema = schema_builder.build();
-        let tempdir = TempDir::new().unwrap();
-        let tempdir_path = PathBuf::from(tempdir.path());
-        Index::create_in_dir(&tempdir_path, schema).unwrap();
-        let index = Index::open_in_dir(tempdir_path).unwrap();
+        let index = Index::create_in_ram(schema);
         let schema = index.schema();
         let mut index_writer = index.writer(50_000_000).unwrap();
         let title = schema.get_field("title").unwrap();

src/indexer/single_segment_index_writer.rs

@@ -1,5 +1,6 @@
 use std::marker::PhantomData;
 
+use crate::codec::CodecConfiguration;
 use crate::indexer::operation::AddOperation;
 use crate::indexer::segment_updater::save_metas;
 use crate::indexer::SegmentWriter;
@@ -7,22 +8,22 @@ use crate::schema::document::Document;
 use crate::{Directory, Index, IndexMeta, Opstamp, Segment, TantivyDocument};
 
 #[doc(hidden)]
-pub struct SingleSegmentIndexWriter<D: Document = TantivyDocument> {
-    segment_writer: SegmentWriter,
-    segment: Segment,
+pub struct SingleSegmentIndexWriter<Codec: crate::codec::Codec, D: Document = TantivyDocument> {
+    segment_writer: SegmentWriter<Codec>,
+    segment: Segment<Codec>,
     opstamp: Opstamp,
-    _phantom: PhantomData<D>,
+    _doc: PhantomData<D>,
 }
 
-impl<D: Document> SingleSegmentIndexWriter<D> {
-    pub fn new(index: Index, mem_budget: usize) -> crate::Result<Self> {
+impl<Codec: crate::codec::Codec, D: Document> SingleSegmentIndexWriter<Codec, D> {
+    pub fn new(index: Index<Codec>, mem_budget: usize) -> crate::Result<Self> {
         let segment = index.new_segment();
         let segment_writer = SegmentWriter::for_segment(mem_budget, segment.clone())?;
         Ok(Self {
             segment_writer,
             segment,
             opstamp: 0,
-            _phantom: PhantomData,
+            _doc: PhantomData,
         })
     }
 
@@ -37,10 +38,10 @@ impl<D: Document> SingleSegmentIndexWriter<D> {
             .add_document(AddOperation { opstamp, document })
     }
 
-    pub fn finalize(self) -> crate::Result<Index> {
+    pub fn finalize(self) -> crate::Result<Index<Codec>> {
         let max_doc = self.segment_writer.max_doc();
         self.segment_writer.finalize()?;
-        let segment: Segment = self.segment.with_max_doc(max_doc);
+        let segment: Segment<Codec> = self.segment.with_max_doc(max_doc);
         let index = segment.index();
         let index_meta = IndexMeta {
             index_settings: index.settings().clone(),
@@ -48,6 +49,7 @@ impl<D: Document> SingleSegmentIndexWriter<D> {
             schema: index.schema(),
             opstamp: 0,
             payload: None,
+            codec: CodecConfiguration::from_codec(index.codec()),
         };
         save_metas(&index_meta, index.directory())?;
         index.directory().sync_directory()?;

src/lib.rs

@@ -17,6 +17,7 @@
 //!
 //! ```rust
 //! # use std::path::Path;
+//! # use std::fs;
 //! # use tempfile::TempDir;
 //! # use tantivy::collector::TopDocs;
 //! # use tantivy::query::QueryParser;
@@ -27,8 +28,11 @@
 //! #     // Let's create a temporary directory for the
 //! #     // sake of this example
 //! #     if let Ok(dir) = TempDir::new() {
-//! #         run_example(dir.path()).unwrap();
-//! #         dir.close().unwrap();
+//! #         let index_path = dir.path().join("index");
+//! #         // In case the directory already exists, we remove it
+//! #         let _ = fs::remove_dir_all(&index_path);
+//! #         fs::create_dir_all(&index_path).unwrap();
+//! #         run_example(&index_path).unwrap();
 //! #     }
 //! # }
 //! #
@@ -162,6 +166,7 @@ mod functional_test;
 
 #[macro_use]
 mod macros;
+pub mod codec;
 mod future_result;
 
 // Re-exports
@@ -203,6 +208,7 @@ mod docset;
 mod reader;
 
 #[cfg(test)]
+#[cfg(feature = "mmap")]
 mod compat_tests;
 
 pub use self::reader::{IndexReader, IndexReaderBuilder, ReloadPolicy, Warmer};
@@ -1170,12 +1176,11 @@ pub mod tests {
 
     #[test]
     fn test_validate_checksum() -> crate::Result<()> {
-        let index_path = tempfile::tempdir().expect("dir");
         let mut builder = Schema::builder();
         let body = builder.add_text_field("body", TEXT | STORED);
         let schema = builder.build();
-        let index = Index::create_in_dir(&index_path, schema)?;
-        let mut writer: IndexWriter = index.writer(50_000_000)?;
+        let index = Index::create_in_ram(schema);
+        let mut writer: IndexWriter = index.writer_for_tests()?;
         writer.set_merge_policy(Box::new(NoMergePolicy));
         for _ in 0..5000 {
             writer.add_document(doc!(body => "foo"))?;

src/postings/json_postings_writer.rs

@@ -3,6 +3,7 @@ use std::io;
 use common::json_path_writer::JSON_END_OF_PATH;
 use stacker::Addr;
 
+use crate::codec::Codec;
 use crate::indexer::indexing_term::IndexingTerm;
 use crate::indexer::path_to_unordered_id::OrderedPathId;
 use crate::postings::postings_writer::SpecializedPostingsWriter;
@@ -22,12 +23,6 @@ pub(crate) struct JsonPostingsWriter<Rec: Recorder> {
     non_str_posting_writer: SpecializedPostingsWriter<DocIdRecorder>,
 }
 
-impl<Rec: Recorder> From<JsonPostingsWriter<Rec>> for Box<dyn PostingsWriter> {
-    fn from(json_postings_writer: JsonPostingsWriter<Rec>) -> Box<dyn PostingsWriter> {
-        Box::new(json_postings_writer)
-    }
-}
-
 impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
     #[inline]
     fn subscribe(
@@ -58,12 +53,12 @@ impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
     }
 
     /// The actual serialization format is handled by the `PostingsSerializer`.
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         ordered_term_addrs: &[(Field, OrderedPathId, &[u8], Addr)],
         ordered_id_to_path: &[&str],
         ctx: &IndexingContext,
-        serializer: &mut FieldSerializer,
+        serializer: &mut FieldSerializer<C>,
     ) -> io::Result<()> {
         let mut term_buffer = JsonTermSerializer(Vec::with_capacity(48));
         let mut buffer_lender = BufferLender::default();

src/postings/loaded_postings.rs

@@ -1,5 +1,8 @@
 use crate::docset::{DocSet, TERMINATED};
-use crate::postings::{Postings, SegmentPostings};
+use crate::fieldnorm::FieldNormReader;
+use crate::postings::Postings;
+use crate::query::term_query::TermScorer;
+use crate::query::{Bm25Weight, Scorer};
 use crate::DocId;
 
 /// `LoadedPostings` is a `DocSet` and `Postings` implementation.
@@ -22,19 +25,27 @@ pub struct LoadedPostings {
 }
 
 impl LoadedPostings {
+    fn new_term_scorer(
+        self: Box<Self>,
+        fieldnorm_reader: FieldNormReader,
+        similarity_weight: Bm25Weight,
+    ) -> Box<dyn Scorer> {
+        Box::new(TermScorer::new(*self, fieldnorm_reader, similarity_weight))
+    }
+
     /// Creates a new `LoadedPostings` from a `SegmentPostings`.
     ///
     /// It will also preload positions, if positions are available in the SegmentPostings.
-    pub fn load(segment_postings: &mut SegmentPostings) -> LoadedPostings {
-        let num_docs = segment_postings.doc_freq() as usize;
+    pub fn load(postings: &mut Box<dyn Postings>) -> LoadedPostings {
+        let num_docs = postings.doc_freq() as usize;
         let mut doc_ids = Vec::with_capacity(num_docs);
         let mut positions = Vec::with_capacity(num_docs);
         let mut position_offsets = Vec::with_capacity(num_docs);
-        while segment_postings.doc() != TERMINATED {
+        while postings.doc() != TERMINATED {
             position_offsets.push(positions.len() as u32);
-            doc_ids.push(segment_postings.doc());
-            segment_postings.append_positions_with_offset(0, &mut positions);
-            segment_postings.advance();
+            doc_ids.push(postings.doc());
+            postings.append_positions_with_offset(0, &mut positions);
+            postings.advance();
         }
         position_offsets.push(positions.len() as u32);
         LoadedPostings {
@@ -88,6 +99,14 @@ impl DocSet for LoadedPostings {
     }
 }
 impl Postings for LoadedPostings {
+    fn new_term_scorer(
+        self: Box<Self>,
+        fieldnorm_reader: FieldNormReader,
+        similarity_weight: Bm25Weight,
+    ) -> Box<dyn Scorer> {
+        Box::new(TermScorer::new(*self, fieldnorm_reader, similarity_weight))
+    }
+
     fn term_freq(&self) -> u32 {
         let start = self.position_offsets[self.cursor] as usize;
         let end = self.position_offsets[self.cursor + 1] as usize;
@@ -101,6 +120,10 @@ impl Postings for LoadedPostings {
             output.push(*pos + offset);
         }
     }
+
+    fn freq_reading_option(&self) -> super::FreqReadingOption {
+        super::FreqReadingOption::ReadFreq
+    }
 }
 
 #[cfg(test)]

src/postings/mod.rs

@@ -4,7 +4,7 @@ mod block_search;
 
 pub(crate) use self::block_search::branchless_binary_search;
 
-mod block_segment_postings;
+// mod block_segment_postings;
 pub(crate) mod compression;
 mod indexing_context;
 mod json_postings_writer;
@@ -13,27 +13,24 @@ mod per_field_postings_writer;
 mod postings;
 mod postings_writer;
 mod recorder;
-mod segment_postings;
 mod serializer;
-mod skip;
 mod term_info;
 
 pub(crate) use loaded_postings::LoadedPostings;
 pub(crate) use stacker::compute_table_memory_size;
 
-pub use self::block_segment_postings::BlockSegmentPostings;
 pub(crate) use self::indexing_context::IndexingContext;
 pub(crate) use self::per_field_postings_writer::PerFieldPostingsWriter;
 pub use self::postings::Postings;
-pub(crate) use self::postings_writer::{serialize_postings, IndexingPosition, PostingsWriter};
-pub use self::segment_postings::SegmentPostings;
+pub(crate) use self::postings_writer::{
+    serialize_postings, IndexingPosition, PostingsWriter, PostingsWriterEnum,
+};
 pub use self::serializer::{FieldSerializer, InvertedIndexSerializer};
-pub(crate) use self::skip::{BlockInfo, SkipReader};
 pub use self::term_info::TermInfo;
 
 #[expect(clippy::enum_variant_names)]
 #[derive(Debug, PartialEq, Clone, Copy, Eq)]
-pub(crate) enum FreqReadingOption {
+pub enum FreqReadingOption {
     NoFreq,
     SkipFreq,
     ReadFreq,
@@ -527,6 +524,7 @@ pub(crate) mod tests {
     }
 
     impl<TScorer: Scorer> Scorer for UnoptimizedDocSet<TScorer> {
+        #[inline]
         fn score(&mut self) -> Score {
             self.0.score()
         }

src/postings/per_field_postings_writer.rs

@@ -1,16 +1,15 @@
 use crate::postings::json_postings_writer::JsonPostingsWriter;
-use crate::postings::postings_writer::SpecializedPostingsWriter;
+use crate::postings::postings_writer::{PostingsWriterEnum, SpecializedPostingsWriter};
 use crate::postings::recorder::{DocIdRecorder, TermFrequencyRecorder, TfAndPositionRecorder};
-use crate::postings::PostingsWriter;
 use crate::schema::{Field, FieldEntry, FieldType, IndexRecordOption, Schema};
 
 pub(crate) struct PerFieldPostingsWriter {
-    per_field_postings_writers: Vec<Box<dyn PostingsWriter>>,
+    per_field_postings_writers: Vec<PostingsWriterEnum>,
 }
 
 impl PerFieldPostingsWriter {
     pub fn for_schema(schema: &Schema) -> Self {
-        let per_field_postings_writers = schema
+        let per_field_postings_writers: Vec<PostingsWriterEnum> = schema
             .fields()
             .map(|(_, field_entry)| posting_writer_from_field_entry(field_entry))
             .collect();
@@ -19,16 +18,16 @@ impl PerFieldPostingsWriter {
         }
     }
 
-    pub(crate) fn get_for_field(&self, field: Field) -> &dyn PostingsWriter {
-        self.per_field_postings_writers[field.field_id() as usize].as_ref()
+    pub(crate) fn get_for_field(&self, field: Field) -> &PostingsWriterEnum {
+        &self.per_field_postings_writers[field.field_id() as usize]
     }
 
-    pub(crate) fn get_for_field_mut(&mut self, field: Field) -> &mut dyn PostingsWriter {
-        self.per_field_postings_writers[field.field_id() as usize].as_mut()
+    pub(crate) fn get_for_field_mut(&mut self, field: Field) -> &mut PostingsWriterEnum {
+        &mut self.per_field_postings_writers[field.field_id() as usize]
     }
 }
 
-fn posting_writer_from_field_entry(field_entry: &FieldEntry) -> Box<dyn PostingsWriter> {
+fn posting_writer_from_field_entry(field_entry: &FieldEntry) -> PostingsWriterEnum {
     match *field_entry.field_type() {
         FieldType::Str(ref text_options) => text_options
             .get_indexing_options()
@@ -51,7 +50,7 @@ fn posting_writer_from_field_entry(field_entry: &FieldEntry) -> Box<dyn Postings
         | FieldType::Date(_)
         | FieldType::Bytes(_)
         | FieldType::IpAddr(_)
-        | FieldType::Facet(_) => Box::<SpecializedPostingsWriter<DocIdRecorder>>::default(),
+        | FieldType::Facet(_) => <SpecializedPostingsWriter<DocIdRecorder>>::default().into(),
         FieldType::JsonObject(ref json_object_options) => {
             if let Some(text_indexing_option) = json_object_options.get_text_indexing_options() {
                 match text_indexing_option.index_option() {

src/postings/postings.rs

@@ -1,4 +1,9 @@
 use crate::docset::DocSet;
+use crate::fastfield::AliveBitSet;
+use crate::fieldnorm::FieldNormReader;
+use crate::postings::FreqReadingOption;
+use crate::query::{Bm25Weight, Scorer};
+use crate::{Score, TERMINATED};
 
 /// Postings (also called inverted list)
 ///
@@ -11,9 +16,42 @@ use crate::docset::DocSet;
 /// but other implementations mocking `SegmentPostings` exist,
 /// for merging segments or for testing.
 pub trait Postings: DocSet + 'static {
+    fn new_term_scorer(
+        self: Box<Self>,
+        fieldnorm_reader: FieldNormReader,
+        similarity_weight: Bm25Weight,
+    ) -> Box<dyn Scorer> {
+        // let self_dyn: Box<dyn Postings>  = self;
+        todo!();
+    }
+
     /// The number of times the term appears in the document.
     fn term_freq(&self) -> u32;
 
+    fn doc_freq(&self) -> u32 {
+        todo!();
+    }
+
+    /// Compute the number of non-deleted documents.
+    ///
+    /// This method will clone and scan through the posting lists.
+    /// (this is a rather expensive operation).
+    fn doc_freq_given_deletes(&self, alive_bitset: &AliveBitSet) -> u32 {
+        todo!();
+        // let mut docset = self.clone();
+        // let mut doc_freq = 0;
+        // loop {
+        //     let doc = docset.doc();
+        //     if doc == TERMINATED {
+        //         return doc_freq;
+        //     }
+        //     if alive_bitset.is_alive(doc) {
+        //         doc_freq += 1u32;
+        //     }
+        //     docset.advance();
+        // }
+    }
+
     /// Returns the positions offsetted with a given value.
     /// It is not necessary to clear the `output` before calling this method.
     /// The output vector will be resized to the `term_freq`.
@@ -31,9 +69,41 @@ pub trait Postings: DocSet + 'static {
     fn positions(&mut self, output: &mut Vec<u32>) {
         self.positions_with_offset(0u32, output);
     }
+
+    fn freq_reading_option(&self) -> FreqReadingOption;
+
+    // TODO see if we can put that in a lift to PostingsWithBlockMax trait.
+    // supports Block-Wand
+    fn supports_block_max(&self) -> bool {
+        false
+    }
+    // TODO document
+    // Only allowed for block max.
+    fn seek_block(
+        &mut self,
+        target_doc: crate::DocId,
+        fieldnorm_reader: &FieldNormReader,
+        similarity_weight: &Bm25Weight,
+    ) -> Score {
+        unimplemented!()
+    }
+
+    // TODO
+    // Only allowed for block max.
+    fn last_doc_in_block(&self) -> crate::DocId {
+        unimplemented!()
+    }
 }
 
 impl Postings for Box<dyn Postings> {
+    fn new_term_scorer(
+        self: Box<Self>,
+        fieldnorm_reader: FieldNormReader,
+        similarity_weight: Bm25Weight,
+    ) -> Box<dyn Scorer> {
+        (*self).new_term_scorer(fieldnorm_reader, similarity_weight)
+    }
+
     fn term_freq(&self) -> u32 {
         (**self).term_freq()
     }
@@ -41,4 +111,25 @@ impl Postings for Box<dyn Postings> {
     fn append_positions_with_offset(&mut self, offset: u32, output: &mut Vec<u32>) {
         (**self).append_positions_with_offset(offset, output);
     }
+
+    fn supports_block_max(&self) -> bool {
+        (**self).supports_block_max()
+    }
+
+    fn seek_block(
+        &mut self,
+        target_doc: crate::DocId,
+        fieldnorm_reader: &FieldNormReader,
+        similarity_weight: &Bm25Weight,
+    ) -> Score {
+        (**self).seek_block(target_doc, fieldnorm_reader, similarity_weight)
+    }
+
+    fn last_doc_in_block(&self) -> crate::DocId {
+        (**self).last_doc_in_block()
+    }
+
+    fn freq_reading_option(&self) -> FreqReadingOption {
+        (**self).freq_reading_option()
+    }
 }

src/postings/postings_writer.rs

@@ -4,10 +4,14 @@ use std::ops::Range;
 
 use stacker::Addr;
 
+use crate::codec::Codec;
 use crate::fieldnorm::FieldNormReaders;
 use crate::indexer::indexing_term::IndexingTerm;
 use crate::indexer::path_to_unordered_id::OrderedPathId;
-use crate::postings::recorder::{BufferLender, Recorder};
+use crate::postings::json_postings_writer::JsonPostingsWriter;
+use crate::postings::recorder::{
+    BufferLender, DocIdRecorder, Recorder, TermFrequencyRecorder, TfAndPositionRecorder,
+};
 use crate::postings::{
     FieldSerializer, IndexingContext, InvertedIndexSerializer, PerFieldPostingsWriter,
 };
@@ -45,12 +49,12 @@ fn make_field_partition(
 /// Serialize the inverted index.
 /// It pushes all term, one field at a time, towards the
 /// postings serializer.
-pub(crate) fn serialize_postings(
+pub(crate) fn serialize_postings<C: Codec>(
     ctx: IndexingContext,
     schema: Schema,
     per_field_postings_writers: &PerFieldPostingsWriter,
     fieldnorm_readers: FieldNormReaders,
-    serializer: &mut InvertedIndexSerializer,
+    serializer: &mut InvertedIndexSerializer<C>,
 ) -> crate::Result<()> {
     // Replace unordered ids by ordered ids to be able to sort
     let unordered_id_to_ordered_id: Vec<OrderedPathId> =
@@ -100,6 +104,141 @@ pub(crate) struct IndexingPosition {
     pub end_position: u32,
 }
 
+pub enum PostingsWriterEnum {
+    DocId(SpecializedPostingsWriter<DocIdRecorder>),
+    DocIdTf(SpecializedPostingsWriter<TermFrequencyRecorder>),
+    DocTfAndPosition(SpecializedPostingsWriter<TfAndPositionRecorder>),
+    JsonDocId(JsonPostingsWriter<DocIdRecorder>),
+    JsonDocIdTf(JsonPostingsWriter<TermFrequencyRecorder>),
+    JsonDocTfAndPosition(JsonPostingsWriter<TfAndPositionRecorder>),
+}
+
+impl From<SpecializedPostingsWriter<DocIdRecorder>> for PostingsWriterEnum {
+    fn from(doc_id_recorder_writer: SpecializedPostingsWriter<DocIdRecorder>) -> Self {
+        PostingsWriterEnum::DocId(doc_id_recorder_writer)
+    }
+}
+
+impl From<SpecializedPostingsWriter<TermFrequencyRecorder>> for PostingsWriterEnum {
+    fn from(doc_id_tf_recorder_writer: SpecializedPostingsWriter<TermFrequencyRecorder>) -> Self {
+        PostingsWriterEnum::DocIdTf(doc_id_tf_recorder_writer)
+    }
+}
+
+impl From<SpecializedPostingsWriter<TfAndPositionRecorder>> for PostingsWriterEnum {
+    fn from(
+        doc_id_tf_and_positions_recorder_writer: SpecializedPostingsWriter<TfAndPositionRecorder>,
+    ) -> Self {
+        PostingsWriterEnum::DocTfAndPosition(doc_id_tf_and_positions_recorder_writer)
+    }
+}
+
+impl From<JsonPostingsWriter<DocIdRecorder>> for PostingsWriterEnum {
+    fn from(doc_id_recorder_writer: JsonPostingsWriter<DocIdRecorder>) -> Self {
+        PostingsWriterEnum::JsonDocId(doc_id_recorder_writer)
+    }
+}
+
+impl From<JsonPostingsWriter<TermFrequencyRecorder>> for PostingsWriterEnum {
+    fn from(doc_id_tf_recorder_writer: JsonPostingsWriter<TermFrequencyRecorder>) -> Self {
+        PostingsWriterEnum::JsonDocIdTf(doc_id_tf_recorder_writer)
+    }
+}
+
+impl From<JsonPostingsWriter<TfAndPositionRecorder>> for PostingsWriterEnum {
+    fn from(
+        doc_id_tf_and_positions_recorder_writer: JsonPostingsWriter<TfAndPositionRecorder>,
+    ) -> Self {
+        PostingsWriterEnum::JsonDocTfAndPosition(doc_id_tf_and_positions_recorder_writer)
+    }
+}
+
+impl PostingsWriter for PostingsWriterEnum {
+    fn subscribe(&mut self, doc: DocId, pos: u32, term: &IndexingTerm, ctx: &mut IndexingContext) {
+        match self {
+            PostingsWriterEnum::DocId(writer) => writer.subscribe(doc, pos, term, ctx),
+            PostingsWriterEnum::DocIdTf(writer) => writer.subscribe(doc, pos, term, ctx),
+            PostingsWriterEnum::DocTfAndPosition(writer) => writer.subscribe(doc, pos, term, ctx),
+            PostingsWriterEnum::JsonDocId(writer) => writer.subscribe(doc, pos, term, ctx),
+            PostingsWriterEnum::JsonDocIdTf(writer) => writer.subscribe(doc, pos, term, ctx),
+            PostingsWriterEnum::JsonDocTfAndPosition(writer) => {
+                writer.subscribe(doc, pos, term, ctx)
+            }
+        }
+    }
+
+    fn serialize<C: Codec>(
+        &self,
+        term_addrs: &[(Field, OrderedPathId, &[u8], Addr)],
+        ordered_id_to_path: &[&str],
+        ctx: &IndexingContext,
+        serializer: &mut FieldSerializer<C>,
+    ) -> io::Result<()> {
+        match self {
+            PostingsWriterEnum::DocId(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+            PostingsWriterEnum::DocIdTf(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+            PostingsWriterEnum::DocTfAndPosition(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+            PostingsWriterEnum::JsonDocId(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+            PostingsWriterEnum::JsonDocIdTf(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+            PostingsWriterEnum::JsonDocTfAndPosition(writer) => {
+                writer.serialize(term_addrs, ordered_id_to_path, ctx, serializer)
+            }
+        }
+    }
+
+    /// Tokenize a text and subscribe all of its token.
+    fn index_text(
+        &mut self,
+        doc_id: DocId,
+        token_stream: &mut dyn TokenStream,
+        term_buffer: &mut IndexingTerm,
+        ctx: &mut IndexingContext,
+        indexing_position: &mut IndexingPosition,
+    ) {
+        match self {
+            PostingsWriterEnum::DocId(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+            PostingsWriterEnum::DocIdTf(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+            PostingsWriterEnum::DocTfAndPosition(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+            PostingsWriterEnum::JsonDocId(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+            PostingsWriterEnum::JsonDocIdTf(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+            PostingsWriterEnum::JsonDocTfAndPosition(writer) => {
+                writer.index_text(doc_id, token_stream, term_buffer, ctx, indexing_position)
+            }
+        }
+    }
+
+    fn total_num_tokens(&self) -> u64 {
+        match self {
+            PostingsWriterEnum::DocId(writer) => writer.total_num_tokens(),
+            PostingsWriterEnum::DocIdTf(writer) => writer.total_num_tokens(),
+            PostingsWriterEnum::DocTfAndPosition(writer) => writer.total_num_tokens(),
+            PostingsWriterEnum::JsonDocId(writer) => writer.total_num_tokens(),
+            PostingsWriterEnum::JsonDocIdTf(writer) => writer.total_num_tokens(),
+            PostingsWriterEnum::JsonDocTfAndPosition(writer) => writer.total_num_tokens(),
+        }
+    }
+}
+
 /// The `PostingsWriter` is in charge of receiving documenting
 /// and building a `Segment` in anonymous memory.
 ///
@@ -116,12 +255,12 @@ pub(crate) trait PostingsWriter: Send + Sync {
 
     /// Serializes the postings on disk.
     /// The actual serialization format is handled by the `PostingsSerializer`.
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         term_addrs: &[(Field, OrderedPathId, &[u8], Addr)],
         ordered_id_to_path: &[&str],
         ctx: &IndexingContext,
-        serializer: &mut FieldSerializer,
+        serializer: &mut FieldSerializer<C>,
     ) -> io::Result<()>;
 
     /// Tokenize a text and subscribe all of its token.
@@ -171,22 +310,14 @@ pub(crate) struct SpecializedPostingsWriter<Rec: Recorder> {
     _recorder_type: PhantomData<Rec>,
 }
 
-impl<Rec: Recorder> From<SpecializedPostingsWriter<Rec>> for Box<dyn PostingsWriter> {
-    fn from(
-        specialized_postings_writer: SpecializedPostingsWriter<Rec>,
-    ) -> Box<dyn PostingsWriter> {
-        Box::new(specialized_postings_writer)
-    }
-}
-
 impl<Rec: Recorder> SpecializedPostingsWriter<Rec> {
     #[inline]
-    pub(crate) fn serialize_one_term(
+    pub(crate) fn serialize_one_term<C: Codec>(
         term: &[u8],
         addr: Addr,
         buffer_lender: &mut BufferLender,
         ctx: &IndexingContext,
-        serializer: &mut FieldSerializer,
+        serializer: &mut FieldSerializer<C>,
     ) -> io::Result<()> {
         let recorder: Rec = ctx.term_index.read(addr);
         let term_doc_freq = recorder.term_doc_freq().unwrap_or(0u32);
@@ -227,12 +358,12 @@ impl<Rec: Recorder> PostingsWriter for SpecializedPostingsWriter<Rec> {
         });
     }
 
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         term_addrs: &[(Field, OrderedPathId, &[u8], Addr)],
         _ordered_id_to_path: &[&str],
         ctx: &IndexingContext,
-        serializer: &mut FieldSerializer,
+        serializer: &mut FieldSerializer<C>,
     ) -> io::Result<()> {
         let mut buffer_lender = BufferLender::default();
         for (_field, _path_id, term, addr) in term_addrs {

src/postings/recorder.rs

@@ -1,6 +1,7 @@
 use common::read_u32_vint;
 use stacker::{ExpUnrolledLinkedList, MemoryArena};
 
+use crate::codec::Codec;
 use crate::postings::FieldSerializer;
 use crate::DocId;
 
@@ -67,10 +68,10 @@ pub(crate) trait Recorder: Copy + Default + Send + Sync + 'static {
     /// Close the document. It will help record the term frequency.
     fn close_doc(&mut self, arena: &mut MemoryArena);
     /// Pushes the postings information to the serializer.
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         arena: &MemoryArena,
-        serializer: &mut FieldSerializer<'_>,
+        serializer: &mut FieldSerializer<C>,
         buffer_lender: &mut BufferLender,
     );
     /// Returns the number of document containing this term.
@@ -110,10 +111,10 @@ impl Recorder for DocIdRecorder {
     #[inline]
     fn close_doc(&mut self, _arena: &mut MemoryArena) {}
 
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         arena: &MemoryArena,
-        serializer: &mut FieldSerializer<'_>,
+        serializer: &mut FieldSerializer<C>,
         buffer_lender: &mut BufferLender,
     ) {
         let buffer = buffer_lender.lend_u8();
@@ -178,10 +179,10 @@ impl Recorder for TermFrequencyRecorder {
         self.current_tf = 0;
     }
 
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         arena: &MemoryArena,
-        serializer: &mut FieldSerializer<'_>,
+        serializer: &mut FieldSerializer<C>,
         buffer_lender: &mut BufferLender,
     ) {
         let buffer = buffer_lender.lend_u8();
@@ -235,10 +236,10 @@ impl Recorder for TfAndPositionRecorder {
         self.stack.writer(arena).write_u32_vint(POSITION_END);
     }
 
-    fn serialize(
+    fn serialize<C: Codec>(
         &self,
         arena: &MemoryArena,
-        serializer: &mut FieldSerializer<'_>,
+        serializer: &mut FieldSerializer<C>,
         buffer_lender: &mut BufferLender,
     ) {
         let (buffer_u8, buffer_positions) = buffer_lender.lend_all();

src/postings/serializer.rs

@@ -1,16 +1,14 @@
-use std::cmp::Ordering;
 use std::io::{self, Write};
 
-use common::{BinarySerializable, CountingWriter, VInt};
+use common::{BinarySerializable, CountingWriter};
 
 use super::TermInfo;
+use crate::codec::postings::PostingsSerializer;
+use crate::codec::Codec;
 use crate::directory::{CompositeWrite, WritePtr};
 use crate::fieldnorm::FieldNormReader;
 use crate::index::Segment;
 use crate::positions::PositionSerializer;
-use crate::postings::compression::{BlockEncoder, VIntEncoder, COMPRESSION_BLOCK_SIZE};
-use crate::postings::skip::SkipSerializer;
-use crate::query::Bm25Weight;
 use crate::schema::{Field, FieldEntry, FieldType, IndexRecordOption, Schema};
 use crate::termdict::TermDictionaryBuilder;
 use crate::{DocId, Score};
@@ -46,22 +44,27 @@ use crate::{DocId, Score};
 ///
 /// A description of the serialization format is
 /// [available here](https://fulmicoton.gitbooks.io/tantivy-doc/content/inverted-index.html).
-pub struct InvertedIndexSerializer {
+pub struct InvertedIndexSerializer<C: Codec> {
     terms_write: CompositeWrite<WritePtr>,
     postings_write: CompositeWrite<WritePtr>,
     positions_write: CompositeWrite<WritePtr>,
     schema: Schema,
+    codec: C,
 }
 
-impl InvertedIndexSerializer {
+use crate::codec::postings::PostingsCodec;
+
+impl<C: Codec> InvertedIndexSerializer<C> {
     /// Open a new `InvertedIndexSerializer` for the given segment
-    pub fn open(segment: &mut Segment) -> crate::Result<InvertedIndexSerializer> {
+    pub fn open(segment: &mut Segment<C>) -> crate::Result<InvertedIndexSerializer<C>> {
         use crate::index::SegmentComponent::{Positions, Postings, Terms};
+        let codec = segment.index().codec().clone();
         let inv_index_serializer = InvertedIndexSerializer {
             terms_write: CompositeWrite::wrap(segment.open_write(Terms)?),
             postings_write: CompositeWrite::wrap(segment.open_write(Postings)?),
             positions_write: CompositeWrite::wrap(segment.open_write(Positions)?),
             schema: segment.schema(),
+            codec,
         };
         Ok(inv_index_serializer)
     }
@@ -75,7 +78,7 @@ impl InvertedIndexSerializer {
         field: Field,
         total_num_tokens: u64,
         fieldnorm_reader: Option<FieldNormReader>,
-    ) -> io::Result<FieldSerializer<'_>> {
+    ) -> io::Result<FieldSerializer<'_, C>> {
         let field_entry: &FieldEntry = self.schema.get_field_entry(field);
         let term_dictionary_write = self.terms_write.for_field(field);
         let postings_write = self.postings_write.for_field(field);
@@ -88,6 +91,7 @@ impl InvertedIndexSerializer {
             postings_write,
             positions_write,
             fieldnorm_reader,
+            &self.codec,
         )
     }
 
@@ -102,15 +106,17 @@ impl InvertedIndexSerializer {
 
 /// The field serializer is in charge of
 /// the serialization of a specific field.
-pub struct FieldSerializer<'a> {
+pub struct FieldSerializer<'a, C: Codec> {
     term_dictionary_builder: TermDictionaryBuilder<&'a mut CountingWriter<WritePtr>>,
-    postings_serializer: PostingsSerializer<&'a mut CountingWriter<WritePtr>>,
+    postings_serializer: <C::PostingsCodec as PostingsCodec>::PostingsSerializer,
     positions_serializer_opt: Option<PositionSerializer<&'a mut CountingWriter<WritePtr>>>,
     current_term_info: TermInfo,
     term_open: bool,
+    postings_write: &'a mut CountingWriter<WritePtr>,
+    postings_start_offset: u64,
 }
 
-impl<'a> FieldSerializer<'a> {
+impl<'a, C: Codec> FieldSerializer<'a, C> {
     fn create(
         field_type: &FieldType,
         total_num_tokens: u64,
@@ -118,7 +124,8 @@ impl<'a> FieldSerializer<'a> {
         postings_write: &'a mut CountingWriter<WritePtr>,
         positions_write: &'a mut CountingWriter<WritePtr>,
         fieldnorm_reader: Option<FieldNormReader>,
-    ) -> io::Result<FieldSerializer<'a>> {
+        codec: &C,
+    ) -> io::Result<FieldSerializer<'a, C>> {
         total_num_tokens.serialize(postings_write)?;
         let index_record_option = field_type
             .index_record_option()
@@ -128,8 +135,7 @@ impl<'a> FieldSerializer<'a> {
             .as_ref()
             .map(|ff_reader| total_num_tokens as Score / ff_reader.num_docs() as Score)
             .unwrap_or(0.0);
-        let postings_serializer = PostingsSerializer::new(
-            postings_write,
+        let postings_serializer = codec.postings_codec().new_serializer(
             average_fieldnorm,
             index_record_option,
             fieldnorm_reader,
@@ -140,15 +146,22 @@ impl<'a> FieldSerializer<'a> {
             None
         };
 
+        let postings_start_offset = postings_write.written_bytes();
         Ok(FieldSerializer {
             term_dictionary_builder,
             postings_serializer,
             positions_serializer_opt,
             current_term_info: TermInfo::default(),
             term_open: false,
+            postings_write,
+            postings_start_offset,
         })
     }
 
+    fn postings_offset(&self) -> usize {
+        (self.postings_write.written_bytes() - self.postings_start_offset) as usize
+    }
+
     fn current_term_info(&self) -> TermInfo {
         let positions_start =
             if let Some(positions_serializer) = self.positions_serializer_opt.as_ref() {
@@ -156,7 +169,7 @@ impl<'a> FieldSerializer<'a> {
             } else {
                 0u64
             } as usize;
-        let addr = self.postings_serializer.written_bytes() as usize;
+        let addr = self.postings_offset();
         TermInfo {
             doc_freq: 0,
             postings_range: addr..addr,
@@ -213,21 +226,22 @@ impl<'a> FieldSerializer<'a> {
         crate::fail_point!("FieldSerializer::close_term", |msg: Option<String>| {
             Err(io::Error::new(io::ErrorKind::Other, format!("{msg:?}")))
         });
-        if self.term_open {
-            self.postings_serializer
-                .close_term(self.current_term_info.doc_freq)?;
-            self.current_term_info.postings_range.end =
-                self.postings_serializer.written_bytes() as usize;
 
-            if let Some(positions_serializer) = self.positions_serializer_opt.as_mut() {
-                positions_serializer.close_term()?;
-                self.current_term_info.positions_range.end =
-                    positions_serializer.written_bytes() as usize;
-            }
-            self.term_dictionary_builder
-                .insert_value(&self.current_term_info)?;
-            self.term_open = false;
+        if !self.term_open {
+            return Ok(());
+        };
+
+        self.postings_serializer
+            .close_term(self.current_term_info.doc_freq, self.postings_write)?;
+        self.current_term_info.postings_range.end = self.postings_offset();
+        if let Some(positions_serializer) = self.positions_serializer_opt.as_mut() {
+            positions_serializer.close_term()?;
+            self.current_term_info.positions_range.end =
+                positions_serializer.written_bytes() as usize;
         }
+        self.term_dictionary_builder
+            .insert_value(&self.current_term_info)?;
+        self.term_open = false;
         Ok(())
     }
 
@@ -237,242 +251,8 @@ impl<'a> FieldSerializer<'a> {
         if let Some(positions_serializer) = self.positions_serializer_opt {
             positions_serializer.close()?;
         }
-        self.postings_serializer.close()?;
+        self.postings_write.flush()?;
         self.term_dictionary_builder.finish()?;
         Ok(())
     }
 }
-
-struct Block {
-    doc_ids: [DocId; COMPRESSION_BLOCK_SIZE],
-    term_freqs: [u32; COMPRESSION_BLOCK_SIZE],
-    len: usize,
-}
-
-impl Block {
-    fn new() -> Self {
-        Block {
-            doc_ids: [0u32; COMPRESSION_BLOCK_SIZE],
-            term_freqs: [0u32; COMPRESSION_BLOCK_SIZE],
-            len: 0,
-        }
-    }
-
-    fn doc_ids(&self) -> &[DocId] {
-        &self.doc_ids[..self.len]
-    }
-
-    fn term_freqs(&self) -> &[u32] {
-        &self.term_freqs[..self.len]
-    }
-
-    fn clear(&mut self) {
-        self.len = 0;
-    }
-
-    fn append_doc(&mut self, doc: DocId, term_freq: u32) {
-        let len = self.len;
-        self.doc_ids[len] = doc;
-        self.term_freqs[len] = term_freq;
-        self.len = len + 1;
-    }
-
-    fn is_full(&self) -> bool {
-        self.len == COMPRESSION_BLOCK_SIZE
-    }
-
-    fn is_empty(&self) -> bool {
-        self.len == 0
-    }
-
-    fn last_doc(&self) -> DocId {
-        assert_eq!(self.len, COMPRESSION_BLOCK_SIZE);
-        self.doc_ids[COMPRESSION_BLOCK_SIZE - 1]
-    }
-}
-
-pub struct PostingsSerializer<W: Write> {
-    output_write: CountingWriter<W>,
-    last_doc_id_encoded: u32,
-
-    block_encoder: BlockEncoder,
-    block: Box<Block>,
-
-    postings_write: Vec<u8>,
-    skip_write: SkipSerializer,
-
-    mode: IndexRecordOption,
-    fieldnorm_reader: Option<FieldNormReader>,
-
-    bm25_weight: Option<Bm25Weight>,
-    avg_fieldnorm: Score, /* Average number of term in the field for that segment.
-                           * this value is used to compute the block wand information. */
-    term_has_freq: bool,
-}
-
-impl<W: Write> PostingsSerializer<W> {
-    pub fn new(
-        write: W,
-        avg_fieldnorm: Score,
-        mode: IndexRecordOption,
-        fieldnorm_reader: Option<FieldNormReader>,
-    ) -> PostingsSerializer<W> {
-        PostingsSerializer {
-            output_write: CountingWriter::wrap(write),
-
-            block_encoder: BlockEncoder::new(),
-            block: Box::new(Block::new()),
-
-            postings_write: Vec::new(),
-            skip_write: SkipSerializer::new(),
-
-            last_doc_id_encoded: 0u32,
-            mode,
-
-            fieldnorm_reader,
-            bm25_weight: None,
-            avg_fieldnorm,
-            term_has_freq: false,
-        }
-    }
-
-    pub fn new_term(&mut self, term_doc_freq: u32, record_term_freq: bool) {
-        self.bm25_weight = None;
-
-        self.term_has_freq = self.mode.has_freq() && record_term_freq;
-        if !self.term_has_freq {
-            return;
-        }
-
-        let num_docs_in_segment: u64 =
-            if let Some(fieldnorm_reader) = self.fieldnorm_reader.as_ref() {
-                fieldnorm_reader.num_docs() as u64
-            } else {
-                return;
-            };
-
-        if num_docs_in_segment == 0 {
-            return;
-        }
-
-        self.bm25_weight = Some(Bm25Weight::for_one_term_without_explain(
-            term_doc_freq as u64,
-            num_docs_in_segment,
-            self.avg_fieldnorm,
-        ));
-    }
-
-    fn write_block(&mut self) {
-        {
-            // encode the doc ids
-            let (num_bits, block_encoded): (u8, &[u8]) = self
-                .block_encoder
-                .compress_block_sorted(self.block.doc_ids(), self.last_doc_id_encoded);
-            self.last_doc_id_encoded = self.block.last_doc();
-            self.skip_write
-                .write_doc(self.last_doc_id_encoded, num_bits);
-            // last el block 0, offset block 1,
-            self.postings_write.extend(block_encoded);
-        }
-        if self.term_has_freq {
-            let (num_bits, block_encoded): (u8, &[u8]) = self
-                .block_encoder
-                .compress_block_unsorted(self.block.term_freqs(), true);
-            self.postings_write.extend(block_encoded);
-            self.skip_write.write_term_freq(num_bits);
-            if self.mode.has_positions() {
-                // We serialize the sum of term freqs within the skip information
-                // in order to navigate through positions.
-                let sum_freq = self.block.term_freqs().iter().cloned().sum();
-                self.skip_write.write_total_term_freq(sum_freq);
-            }
-            let mut blockwand_params = (0u8, 0u32);
-            if let Some(bm25_weight) = self.bm25_weight.as_ref() {
-                if let Some(fieldnorm_reader) = self.fieldnorm_reader.as_ref() {
-                    let docs = self.block.doc_ids().iter().cloned();
-                    let term_freqs = self.block.term_freqs().iter().cloned();
-                    let fieldnorms = docs.map(|doc| fieldnorm_reader.fieldnorm_id(doc));
-                    blockwand_params = fieldnorms
-                        .zip(term_freqs)
-                        .max_by(
-                            |(left_fieldnorm_id, left_term_freq),
-                             (right_fieldnorm_id, right_term_freq)| {
-                                let left_score =
-                                    bm25_weight.tf_factor(*left_fieldnorm_id, *left_term_freq);
-                                let right_score =
-                                    bm25_weight.tf_factor(*right_fieldnorm_id, *right_term_freq);
-                                left_score
-                                    .partial_cmp(&right_score)
-                                    .unwrap_or(Ordering::Equal)
-                            },
-                        )
-                        .unwrap();
-                }
-            }
-            let (fieldnorm_id, term_freq) = blockwand_params;
-            self.skip_write.write_blockwand_max(fieldnorm_id, term_freq);
-        }
-        self.block.clear();
-    }
-
-    pub fn write_doc(&mut self, doc_id: DocId, term_freq: u32) {
-        self.block.append_doc(doc_id, term_freq);
-        if self.block.is_full() {
-            self.write_block();
-        }
-    }
-
-    fn close(mut self) -> io::Result<()> {
-        self.postings_write.flush()
-    }
-
-    pub fn close_term(&mut self, doc_freq: u32) -> io::Result<()> {
-        if !self.block.is_empty() {
-            // we have doc ids waiting to be written
-            // this happens when the number of doc ids is
-            // not a perfect multiple of our block size.
-            //
-            // In that case, the remaining part is encoded
-            // using variable int encoding.
-            {
-                let block_encoded = self
-                    .block_encoder
-                    .compress_vint_sorted(self.block.doc_ids(), self.last_doc_id_encoded);
-                self.postings_write.write_all(block_encoded)?;
-            }
-            // ... Idem for term frequencies
-            if self.term_has_freq {
-                let block_encoded = self
-                    .block_encoder
-                    .compress_vint_unsorted(self.block.term_freqs());
-                self.postings_write.write_all(block_encoded)?;
-            }
-            self.block.clear();
-        }
-        if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
-            let skip_data = self.skip_write.data();
-            VInt(skip_data.len() as u64).serialize(&mut self.output_write)?;
-            self.output_write.write_all(skip_data)?;
-        }
-        self.output_write.write_all(&self.postings_write[..])?;
-        self.skip_write.clear();
-        self.postings_write.clear();
-        self.bm25_weight = None;
-        Ok(())
-    }
-
-    /// Returns the number of bytes written in the postings write object
-    /// at this point.
-    /// When called before writing the postings of a term, this value is used as
-    /// start offset.
-    /// When called after writing the postings of a term, this value is used as a
-    /// end offset.
-    fn written_bytes(&self) -> u64 {
-        self.output_write.written_bytes()
-    }
-
-    fn clear(&mut self) {
-        self.block.clear();
-        self.last_doc_id_encoded = 0;
-    }
-}

src/query/all_query.rs

@@ -21,12 +21,7 @@ impl Query for AllQuery {
 pub struct AllWeight;
 
 impl Weight for AllWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        _seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         let all_scorer = AllScorer::new(reader.max_doc());
         if boost != 1.0 {
             Ok(Box::new(BoostScorer::new(all_scorer, boost)))
@@ -110,6 +105,7 @@ impl DocSet for AllScorer {
 }
 
 impl Scorer for AllScorer {
+    #[inline]
     fn score(&mut self) -> Score {
         1.0
     }
@@ -145,7 +141,7 @@ mod tests {
         let weight = AllQuery.weight(EnableScoring::disabled_from_schema(&index.schema()))?;
         {
             let reader = searcher.segment_reader(0);
-            let mut scorer = weight.scorer(reader, 1.0, 0)?;
+            let mut scorer = weight.scorer(reader, 1.0)?;
             assert_eq!(scorer.doc(), 0u32);
             assert_eq!(scorer.advance(), 1u32);
             assert_eq!(scorer.doc(), 1u32);
@@ -153,7 +149,7 @@ mod tests {
         }
         {
             let reader = searcher.segment_reader(1);
-            let mut scorer = weight.scorer(reader, 1.0, 0)?;
+            let mut scorer = weight.scorer(reader, 1.0)?;
             assert_eq!(scorer.doc(), 0u32);
             assert_eq!(scorer.advance(), TERMINATED);
         }
@@ -168,12 +164,12 @@ mod tests {
         let weight = AllQuery.weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
         let reader = searcher.segment_reader(0);
         {
-            let mut scorer = weight.scorer(reader, 2.0, 0)?;
+            let mut scorer = weight.scorer(reader, 2.0)?;
             assert_eq!(scorer.doc(), 0u32);
             assert_eq!(scorer.score(), 2.0);
         }
         {
-            let mut scorer = weight.scorer(reader, 1.5, 0)?;
+            let mut scorer = weight.scorer(reader, 1.5)?;
             assert_eq!(scorer.doc(), 0u32);
             assert_eq!(scorer.score(), 1.5);
         }

src/query/automaton_weight.rs

@@ -10,7 +10,7 @@ use crate::postings::TermInfo;
 use crate::query::{BitSetDocSet, ConstScorer, Explanation, Scorer, Weight};
 use crate::schema::{Field, IndexRecordOption};
 use crate::termdict::{TermDictionary, TermStreamer};
-use crate::{DocId, Score, TantivyError};
+use crate::{DocId, DocSet, Score, TantivyError};
 
 /// A weight struct for Fuzzy Term and Regex Queries
 pub struct AutomatonWeight<A> {
@@ -84,12 +84,7 @@ where
     A: Automaton + Send + Sync + 'static,
     A::State: Clone,
 {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         let max_doc = reader.max_doc();
         let mut doc_bitset = BitSet::with_max_value(max_doc);
         let inverted_index = reader.inverted_index(self.field)?;
@@ -97,22 +92,9 @@ where
         let mut term_stream = self.automaton_stream(term_dict)?;
         while term_stream.advance() {
             let term_info = term_stream.value();
-            let (mut block_segment_postings, _) = inverted_index
-                .read_block_postings_from_terminfo_with_seek(
-                    term_info,
-                    IndexRecordOption::Basic,
-                    seek_doc,
-                )?;
-            loop {
-                let docs = block_segment_postings.docs();
-                if docs.is_empty() {
-                    break;
-                }
-                for &doc in docs {
-                    doc_bitset.insert(doc);
-                }
-                block_segment_postings.advance();
-            }
+            let mut block_segment_postings =
+                inverted_index.read_postings_from_terminfo(term_info, IndexRecordOption::Basic)?;
+            block_segment_postings.fill_bitset(&mut doc_bitset);
         }
         let doc_bitset = BitSetDocSet::from(doc_bitset);
         let const_scorer = ConstScorer::new(doc_bitset, boost);
@@ -120,7 +102,7 @@ where
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let mut scorer = self.scorer(reader, 1.0, 0)?;
+        let mut scorer = self.scorer(reader, 1.0)?;
         if scorer.seek(doc) == doc {
             Ok(Explanation::new("AutomatonScorer", 1.0))
         } else {
@@ -195,7 +177,7 @@ mod tests {
         let automaton_weight = AutomatonWeight::new(field, PrefixedByA);
         let reader = index.reader()?;
         let searcher = reader.searcher();
-        let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+        let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.0)?;
         assert_eq!(scorer.doc(), 0u32);
         assert_eq!(scorer.score(), 1.0);
         assert_eq!(scorer.advance(), 2u32);
@@ -212,7 +194,7 @@ mod tests {
         let automaton_weight = AutomatonWeight::new(field, PrefixedByA);
         let reader = index.reader()?;
         let searcher = reader.searcher();
-        let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.32, 0)?;
+        let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.32)?;
         assert_eq!(scorer.doc(), 0u32);
         assert_eq!(scorer.score(), 1.32);
         Ok(())

src/query/boolean_query/block_wand.rs

@@ -4,143 +4,143 @@ use crate::query::term_query::TermScorer;
 use crate::query::Scorer;
 use crate::{DocId, DocSet, Score, TERMINATED};
 
-/// Takes a term_scorers sorted by their current doc() and a threshold and returns
-/// Returns (pivot_len, pivot_ord) defined as follows:
-/// - `pivot_doc` lowest document that has a chance of exceeding (>) the threshold score.
-/// - `before_pivot_len` number of term_scorers such that term_scorer.doc() < pivot.
-/// - `pivot_len` number of term_scorers such that term_scorer.doc() <= pivot.
-///
-/// We always have `before_pivot_len` < `pivot_len`.
-///
-/// `None` is returned if we establish that no document can exceed the threshold.
-fn find_pivot_doc(
-    term_scorers: &[TermScorerWithMaxScore],
-    threshold: Score,
-) -> Option<(usize, usize, DocId)> {
-    let mut max_score = 0.0;
-    let mut before_pivot_len = 0;
-    let mut pivot_doc = TERMINATED;
-    while before_pivot_len < term_scorers.len() {
-        let term_scorer = &term_scorers[before_pivot_len];
-        max_score += term_scorer.max_score;
-        if max_score > threshold {
-            pivot_doc = term_scorer.doc();
-            break;
-        }
-        before_pivot_len += 1;
-    }
-    if pivot_doc == TERMINATED {
-        return None;
-    }
-    // Right now i is an ordinal, we want a len.
-    let mut pivot_len = before_pivot_len + 1;
-    // Some other term_scorer may be positioned on the same document.
-    pivot_len += term_scorers[pivot_len..]
-        .iter()
-        .take_while(|term_scorer| term_scorer.doc() == pivot_doc)
-        .count();
-    Some((before_pivot_len, pivot_len, pivot_doc))
-}
+// /// Takes a term_scorers sorted by their current doc() and a threshold and returns
+// /// Returns (pivot_len, pivot_ord) defined as follows:
+// /// - `pivot_doc` lowest document that has a chance of exceeding (>) the threshold score.
+// /// - `before_pivot_len` number of term_scorers such that term_scorer.doc() < pivot.
+// /// - `pivot_len` number of term_scorers such that term_scorer.doc() <= pivot.
+// ///
+// /// We always have `before_pivot_len` < `pivot_len`.
+// ///
+// /// `None` is returned if we establish that no document can exceed the threshold.
+// fn find_pivot_doc(
+//     term_scorers: &[TermScorerWithMaxScore],
+//     threshold: Score,
+// ) -> Option<(usize, usize, DocId)> {
+//     let mut max_score = 0.0;
+//     let mut before_pivot_len = 0;
+//     let mut pivot_doc = TERMINATED;
+//     while before_pivot_len < term_scorers.len() {
+//         let term_scorer = &term_scorers[before_pivot_len];
+//         max_score += term_scorer.max_score;
+//         if max_score > threshold {
+//             pivot_doc = term_scorer.doc();
+//             break;
+//         }
+//         before_pivot_len += 1;
+//     }
+//     if pivot_doc == TERMINATED {
+//         return None;
+//     }
+//     // Right now i is an ordinal, we want a len.
+//     let mut pivot_len = before_pivot_len + 1;
+//     // Some other term_scorer may be positioned on the same document.
+//     pivot_len += term_scorers[pivot_len..]
+//         .iter()
+//         .take_while(|term_scorer| term_scorer.doc() == pivot_doc)
+//         .count();
+//     Some((before_pivot_len, pivot_len, pivot_doc))
+// }
 
-/// Advance the scorer with best score among the scorers[..pivot_len] to
-/// the next doc candidate defined by the min of `last_doc_in_block + 1` for
-/// scorer in scorers[..pivot_len] and `scorer.doc()` for scorer in scorers[pivot_len..].
-/// Note: before and after calling this method, scorers need to be sorted by their `.doc()`.
-fn block_max_was_too_low_advance_one_scorer(
-    scorers: &mut [TermScorerWithMaxScore],
-    pivot_len: usize,
-) {
-    debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
-    let mut scorer_to_seek = pivot_len - 1;
-    let mut global_max_score = scorers[scorer_to_seek].max_score;
-    let mut doc_to_seek_after = scorers[scorer_to_seek].last_doc_in_block();
-    for scorer_ord in (0..pivot_len - 1).rev() {
-        let scorer = &scorers[scorer_ord];
-        if scorer.last_doc_in_block() <= doc_to_seek_after {
-            doc_to_seek_after = scorer.last_doc_in_block();
-        }
-        if scorers[scorer_ord].max_score > global_max_score {
-            global_max_score = scorers[scorer_ord].max_score;
-            scorer_to_seek = scorer_ord;
-        }
-    }
-    // Add +1 to go to the next block unless we are already at the end.
-    if doc_to_seek_after != TERMINATED {
-        doc_to_seek_after += 1;
-    }
-    for scorer in &scorers[pivot_len..] {
-        if scorer.doc() <= doc_to_seek_after {
-            doc_to_seek_after = scorer.doc();
-        }
-    }
-    scorers[scorer_to_seek].seek(doc_to_seek_after);
+// /// Advance the scorer with best score among the scorers[..pivot_len] to
+// /// the next doc candidate defined by the min of `last_doc_in_block + 1` for
+// /// scorer in scorers[..pivot_len] and `scorer.doc()` for scorer in scorers[pivot_len..].
+// /// Note: before and after calling this method, scorers need to be sorted by their `.doc()`.
+// fn block_max_was_too_low_advance_one_scorer(
+//     scorers: &mut [TermScorerWithMaxScore],
+//     pivot_len: usize,
+// ) {
+//     debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
+//     let mut scorer_to_seek = pivot_len - 1;
+//     let mut global_max_score = scorers[scorer_to_seek].max_score;
+//     let mut doc_to_seek_after = scorers[scorer_to_seek].last_doc_in_block();
+//     for scorer_ord in (0..pivot_len - 1).rev() {
+//         let scorer = &scorers[scorer_ord];
+//         if scorer.last_doc_in_block() <= doc_to_seek_after {
+//             doc_to_seek_after = scorer.last_doc_in_block();
+//         }
+//         if scorers[scorer_ord].max_score > global_max_score {
+//             global_max_score = scorers[scorer_ord].max_score;
+//             scorer_to_seek = scorer_ord;
+//         }
+//     }
+//     // Add +1 to go to the next block unless we are already at the end.
+//     if doc_to_seek_after != TERMINATED {
+//         doc_to_seek_after += 1;
+//     }
+//     for scorer in &scorers[pivot_len..] {
+//         if scorer.doc() <= doc_to_seek_after {
+//             doc_to_seek_after = scorer.doc();
+//         }
+//     }
+//     scorers[scorer_to_seek].seek(doc_to_seek_after);
 
-    restore_ordering(scorers, scorer_to_seek);
-    debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
-}
+//     restore_ordering(scorers, scorer_to_seek);
+//     debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
+// }
 
-// Given a list of term_scorers and a `ord` and assuming that `term_scorers[ord]` is sorted
-// except term_scorers[ord] that might be in advance compared to its ranks,
-// bubble up term_scorers[ord] in order to restore the ordering.
-fn restore_ordering(term_scorers: &mut [TermScorerWithMaxScore], ord: usize) {
-    let doc = term_scorers[ord].doc();
-    for i in ord + 1..term_scorers.len() {
-        if term_scorers[i].doc() >= doc {
-            break;
-        }
-        term_scorers.swap(i, i - 1);
-    }
-    debug_assert!(is_sorted(term_scorers.iter().map(|scorer| scorer.doc())));
-}
+// // Given a list of term_scorers and a `ord` and assuming that `term_scorers[ord]` is sorted
+// // except term_scorers[ord] that might be in advance compared to its ranks,
+// // bubble up term_scorers[ord] in order to restore the ordering.
+// fn restore_ordering(term_scorers: &mut [TermScorerWithMaxScore], ord: usize) {
+//     let doc = term_scorers[ord].doc();
+//     for i in ord + 1..term_scorers.len() {
+//         if term_scorers[i].doc() >= doc {
+//             break;
+//         }
+//         term_scorers.swap(i, i - 1);
+//     }
+//     debug_assert!(is_sorted(term_scorers.iter().map(|scorer| scorer.doc())));
+// }
 
-// Attempts to advance all term_scorers between `&term_scorers[0..before_len]` to the pivot.
-// If this works, return true.
-// If this fails (ie: one of the term_scorer does not contain `pivot_doc` and seek goes past the
-// pivot), reorder the term_scorers to ensure the list is still sorted and returns `false`.
-// If a term_scorer reach TERMINATED in the process return false remove the term_scorer and return.
-fn align_scorers(
-    term_scorers: &mut Vec<TermScorerWithMaxScore>,
-    pivot_doc: DocId,
-    before_pivot_len: usize,
-) -> bool {
-    debug_assert_ne!(pivot_doc, TERMINATED);
-    for i in (0..before_pivot_len).rev() {
-        let new_doc = term_scorers[i].seek(pivot_doc);
-        if new_doc != pivot_doc {
-            if new_doc == TERMINATED {
-                term_scorers.swap_remove(i);
-            }
-            // We went past the pivot.
-            // We just go through the outer loop mechanic (Note that pivot is
-            // still a possible candidate).
-            //
-            // Termination is still guaranteed since we can only consider the same
-            // pivot at most term_scorers.len() - 1 times.
-            restore_ordering(term_scorers, i);
-            return false;
-        }
-    }
-    true
-}
+// // Attempts to advance all term_scorers between `&term_scorers[0..before_len]` to the pivot.
+// // If this works, return true.
+// // If this fails (ie: one of the term_scorer does not contain `pivot_doc` and seek goes past the
+// // pivot), reorder the term_scorers to ensure the list is still sorted and returns `false`.
+// // If a term_scorer reach TERMINATED in the process return false remove the term_scorer and
+// return. fn align_scorers(
+//     term_scorers: &mut Vec<TermScorerWithMaxScore>,
+//     pivot_doc: DocId,
+//     before_pivot_len: usize,
+// ) -> bool {
+//     debug_assert_ne!(pivot_doc, TERMINATED);
+//     for i in (0..before_pivot_len).rev() {
+//         let new_doc = term_scorers[i].seek(pivot_doc);
+//         if new_doc != pivot_doc {
+//             if new_doc == TERMINATED {
+//                 term_scorers.swap_remove(i);
+//             }
+//             // We went past the pivot.
+//             // We just go through the outer loop mechanic (Note that pivot is
+//             // still a possible candidate).
+//             //
+//             // Termination is still guaranteed since we can only consider the same
+//             // pivot at most term_scorers.len() - 1 times.
+//             restore_ordering(term_scorers, i);
+//             return false;
+//         }
+//     }
+//     true
+// }
 
-// Assumes terms_scorers[..pivot_len] are positioned on the same doc (pivot_doc).
-// Advance term_scorers[..pivot_len] and out of these removes the terminated scores.
-// Restores the ordering of term_scorers.
-fn advance_all_scorers_on_pivot(term_scorers: &mut Vec<TermScorerWithMaxScore>, pivot_len: usize) {
-    for term_scorer in &mut term_scorers[..pivot_len] {
-        term_scorer.advance();
-    }
-    // TODO use drain_filter when available.
-    let mut i = 0;
-    while i != term_scorers.len() {
-        if term_scorers[i].doc() == TERMINATED {
-            term_scorers.swap_remove(i);
-        } else {
-            i += 1;
-        }
-    }
-    term_scorers.sort_by_key(|scorer| scorer.doc());
-}
+// // Assumes terms_scorers[..pivot_len] are positioned on the same doc (pivot_doc).
+// // Advance term_scorers[..pivot_len] and out of these removes the terminated scores.
+// // Restores the ordering of term_scorers.
+// fn advance_all_scorers_on_pivot(term_scorers: &mut Vec<TermScorerWithMaxScore>, pivot_len: usize)
+// {     for term_scorer in &mut term_scorers[..pivot_len] {
+//         term_scorer.advance();
+//     }
+//     // TODO use drain_filter when available.
+//     let mut i = 0;
+//     while i != term_scorers.len() {
+//         if term_scorers[i].doc() == TERMINATED {
+//             term_scorers.swap_remove(i);
+//         } else {
+//             i += 1;
+//         }
+//     }
+//     term_scorers.sort_by_key(|scorer| scorer.doc());
+// }
 
 /// Implements the WAND (Weak AND) algorithm for dynamic pruning
 /// described in the paper "Faster Top-k Document Retrieval Using Block-Max Indexes".
@@ -150,65 +150,63 @@ pub fn block_wand(
     mut threshold: Score,
     callback: &mut dyn FnMut(u32, Score) -> Score,
 ) {
-    let mut scorers: Vec<TermScorerWithMaxScore> = scorers
-        .iter_mut()
-        .map(TermScorerWithMaxScore::from)
-        .collect();
-    scorers.sort_by_key(|scorer| scorer.doc());
-    // At this point we need to ensure that the scorers are sorted!
-    debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
-    while let Some((before_pivot_len, pivot_len, pivot_doc)) =
-        find_pivot_doc(&scorers[..], threshold)
-    {
-        debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
-        debug_assert_ne!(pivot_doc, TERMINATED);
-        debug_assert!(before_pivot_len < pivot_len);
+    // let mut scorers: Vec<TermScorerWithMaxScore> = scorers
+    //     .iter_mut()
+    //     .map(TermScorerWithMaxScore::from)
+    //     .collect();
+    // scorers.sort_by_key(|scorer| scorer.doc());
+    // // At this point we need to ensure that the scorers are sorted!
+    // debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
+    // while let Some((before_pivot_len, pivot_len, pivot_doc)) =
+    //     find_pivot_doc(&scorers[..], threshold)
+    // {
+    //     debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
+    //     debug_assert_ne!(pivot_doc, TERMINATED);
+    //     debug_assert!(before_pivot_len < pivot_len);
 
-        let block_max_score_upperbound: Score = scorers[..pivot_len]
-            .iter_mut()
-            .map(|scorer| {
-                scorer.seek_block(pivot_doc);
-                scorer.block_max_score()
-            })
-            .sum();
+    //     let block_max_score_upperbound: Score = scorers[..pivot_len]
+    //         .iter_mut()
+    //         .map(|scorer| scorer.seek_block(pivot_doc))
+    //         .sum();
 
-        // Beware after shallow advance, skip readers can be in advance compared to
-        // the segment posting lists.
-        //
-        // `block_segment_postings.load_block()` need to be called separately.
-        if block_max_score_upperbound <= threshold {
-            // Block max condition was not reached
-            // We could get away by simply advancing the scorers to DocId + 1 but it would
-            // be inefficient. The optimization requires proper explanation and was
-            // isolated in a different function.
-            block_max_was_too_low_advance_one_scorer(&mut scorers, pivot_len);
-            continue;
-        }
+    //     // Beware after shallow advance, skip readers can be in advance compared to
+    //     // the segment posting lists.
+    //     //
+    //     // `block_segment_postings.load_block()` need to be called separately.
+    //     if block_max_score_upperbound <= threshold {
+    //         // Block max condition was not reached
+    //         // We could get away by simply advancing the scorers to DocId + 1 but it would
+    //         // be inefficient. The optimization requires proper explanation and was
+    //         // isolated in a different function.
+    //         block_max_was_too_low_advance_one_scorer(&mut scorers, pivot_len);
+    //         continue;
+    //     }
 
-        // Block max condition is observed.
-        //
-        // Let's try and advance all scorers before the pivot to the pivot.
-        if !align_scorers(&mut scorers, pivot_doc, before_pivot_len) {
-            // At least of the scorer does not contain the pivot.
-            //
-            // Let's stop scoring this pivot and go through the pivot selection again.
-            // Note that the current pivot is not necessarily a bad candidate and it
-            // may be picked again.
-            continue;
-        }
+    //     // Block max condition is observed.
+    //     //
+    //     // Let's try and advance all scorers before the pivot to the pivot.
+    //     if !align_scorers(&mut scorers, pivot_doc, before_pivot_len) {
+    //         // At least of the scorer does not contain the pivot.
+    //         //
+    //         // Let's stop scoring this pivot and go through the pivot selection again.
+    //         // Note that the current pivot is not necessarily a bad candidate and it
+    //         // may be picked again.
+    //         continue;
+    //     }
 
-        // At this point, all scorers are positioned on the doc.
-        let score = scorers[..pivot_len]
-            .iter_mut()
-            .map(|scorer| scorer.score())
-            .sum();
+    //     // At this point, all scorers are positioned on the doc.
+    //     let score = scorers[..pivot_len]
+    //         .iter_mut()
+    //         .map(|scorer| scorer.score())
+    //         .sum();
 
-        if score > threshold {
-            threshold = callback(pivot_doc, score);
-        }
-        // let's advance all of the scorers that are currently positioned on the pivot.
-        advance_all_scorers_on_pivot(&mut scorers, pivot_len);
-    }
+    //     if score > threshold {
+    //         threshold = callback(pivot_doc, score);
+    //     }
+    //     // let's advance all of the scorers that are currently positioned on the pivot.
+    //     advance_all_scorers_on_pivot(&mut scorers, pivot_len);
+    // }
+    todo!();
 }
 
 /// Specialized version of [`block_wand`] for a single scorer.
@@ -224,40 +222,42 @@ pub fn block_wand_single_scorer(
     mut threshold: Score,
     callback: &mut dyn FnMut(u32, Score) -> Score,
 ) {
-    let mut doc = scorer.doc();
-    loop {
-        // We position the scorer on a block that can reach
-        // the threshold.
-        while scorer.block_max_score() < threshold {
-            let last_doc_in_block = scorer.last_doc_in_block();
-            if last_doc_in_block == TERMINATED {
-                return;
-            }
-            doc = last_doc_in_block + 1;
-            scorer.seek_block(doc);
-        }
-        // Seek will effectively load that block.
-        doc = scorer.seek(doc);
-        if doc == TERMINATED {
-            break;
-        }
-        loop {
-            let score = scorer.score();
-            if score > threshold {
-                threshold = callback(doc, score);
-            }
-            debug_assert!(doc <= scorer.last_doc_in_block());
-            if doc == scorer.last_doc_in_block() {
-                break;
-            }
-            doc = scorer.advance();
-            if doc == TERMINATED {
-                return;
-            }
-        }
-        doc += 1;
-        scorer.seek_block(doc);
-    }
+    todo!();
+    // let mut doc = scorer.doc();
+    // let mut block_max_score = scorer.seek_block(doc);
+    // loop {
+    //     // We position the scorer on a block that can reach
+    //     // the threshold.
+    //     while block_max_score < threshold {
+    //         let last_doc_in_block = scorer.last_doc_in_block();
+    //         if last_doc_in_block == TERMINATED {
+    //             return;
+    //         }
+    //         doc = last_doc_in_block + 1;
+    //         block_max_score = scorer.seek_block(doc);
+    //     }
+    //     // Seek will effectively load that block.
+    //     doc = scorer.seek(doc);
+    //     if doc == TERMINATED {
+    //         break;
+    //     }
+    //     loop {
+    //         let score = scorer.score();
+    //         if score > threshold {
+    //             threshold = callback(doc, score);
+    //         }
+    //         debug_assert!(doc <= scorer.last_doc_in_block());
+    //         if doc == scorer.last_doc_in_block() {
+    //             break;
+    //         }
+    //         doc = scorer.advance();
+    //         if doc == TERMINATED {
+    //             return;
+    //         }
+    //     }
+    //     doc += 1;
+    //     block_max_score = scorer.seek_block(doc);
+    // }
 }
 
 struct TermScorerWithMaxScore<'a> {
@@ -305,6 +305,7 @@ mod tests {
 
     use proptest::prelude::*;
 
+    use crate::postings::SegmentPostings;
     use crate::query::score_combiner::SumCombiner;
     use crate::query::term_query::TermScorer;
     use crate::query::{Bm25Weight, BufferedUnionScorer, Scorer};

src/query/boolean_query/boolean_weight.rs

@@ -12,7 +12,7 @@ use crate::query::{
     intersect_scorers, AllScorer, BufferedUnionScorer, EmptyScorer, Exclude, Explanation, Occur,
     RequiredOptionalScorer, Scorer, Weight,
 };
-use crate::{DocId, Score, TERMINATED};
+use crate::{DocId, Score};
 
 enum SpecializedScorer {
     TermUnion(Vec<TermScorer>),
@@ -156,19 +156,6 @@ fn effective_should_scorer_for_union<TScoreCombiner: ScoreCombiner>(
     }
 }
 
-fn create_scorer(
-    weight: &dyn Weight,
-    reader: &SegmentReader,
-    boost: Score,
-    target_doc: DocId,
-) -> crate::Result<Box<dyn Scorer>> {
-    if target_doc >= reader.max_doc() {
-        Ok(Box::new(EmptyScorer))
-    } else {
-        weight.scorer(reader, boost, target_doc)
-    }
-}
-
 enum ShouldScorersCombinationMethod {
     // Should scorers are irrelevant.
     Ignored,
@@ -220,29 +207,10 @@ impl<TScoreCombiner: ScoreCombiner> BooleanWeight<TScoreCombiner> {
         &self,
         reader: &SegmentReader,
         boost: Score,
-        mut seek_first_doc: DocId,
     ) -> crate::Result<HashMap<Occur, Vec<Box<dyn Scorer>>>> {
         let mut per_occur_scorers: HashMap<Occur, Vec<Box<dyn Scorer>>> = HashMap::new();
-        let (mut must_weights, other_weights): (Vec<(Occur, _)>, Vec<(Occur, _)>) = self
-            .weights
-            .iter()
-            .map(|(occur, weight)| (*occur, weight))
-            .partition(|(occur, _weight)| *occur == Occur::Must);
-        // We start by must weights in order to get the best "seek_first_doc" so that we
-        // can skip the first few documents of the other scorers.
-        must_weights.sort_by_key(|weight| weight.1.intersection_priority());
-        for (_, must_sub_weight) in must_weights {
-            let sub_scorer: Box<dyn Scorer> =
-                create_scorer(must_sub_weight.as_ref(), reader, boost, seek_first_doc)?;
-            seek_first_doc = seek_first_doc.max(sub_scorer.doc());
-            per_occur_scorers
-                .entry(Occur::Must)
-                .or_default()
-                .push(sub_scorer);
-        }
-        for (occur, sub_weight) in &other_weights {
-            let sub_scorer: Box<dyn Scorer> =
-                create_scorer(sub_weight.as_ref(), reader, boost, seek_first_doc)?;
+        for (occur, subweight) in &self.weights {
+            let sub_scorer: Box<dyn Scorer> = subweight.scorer(reader, boost)?;
             per_occur_scorers
                 .entry(*occur)
                 .or_default()
@@ -256,10 +224,9 @@ impl<TScoreCombiner: ScoreCombiner> BooleanWeight<TScoreCombiner> {
         reader: &SegmentReader,
         boost: Score,
         score_combiner_fn: impl Fn() -> TComplexScoreCombiner,
-        seek_doc: u32,
     ) -> crate::Result<SpecializedScorer> {
         let num_docs = reader.num_docs();
-        let mut per_occur_scorers = self.per_occur_scorers(reader, boost, seek_doc)?;
+        let mut per_occur_scorers = self.per_occur_scorers(reader, boost)?;
 
         // Indicate how should clauses are combined with must clauses.
         let mut must_scorers: Vec<Box<dyn Scorer>> =
@@ -440,7 +407,7 @@ fn remove_and_count_all_and_empty_scorers(
         if scorer.is::<AllScorer>() {
             counts.num_all_scorers += 1;
             false
-        } else if scorer.doc() == TERMINATED {
+        } else if scorer.is::<EmptyScorer>() {
             counts.num_empty_scorers += 1;
             false
         } else {
@@ -451,12 +418,7 @@ fn remove_and_count_all_and_empty_scorers(
 }
 
 impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombiner> {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         let num_docs = reader.num_docs();
         if self.weights.is_empty() {
             Ok(Box::new(EmptyScorer))
@@ -465,15 +427,15 @@ impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombin
             if occur == Occur::MustNot {
                 Ok(Box::new(EmptyScorer))
             } else {
-                weight.scorer(reader, boost, seek_doc)
+                weight.scorer(reader, boost)
             }
         } else if self.scoring_enabled {
-            self.complex_scorer(reader, boost, &self.score_combiner_fn, seek_doc)
+            self.complex_scorer(reader, boost, &self.score_combiner_fn)
                 .map(|specialized_scorer| {
                     into_box_scorer(specialized_scorer, &self.score_combiner_fn, num_docs)
                 })
         } else {
-            self.complex_scorer(reader, boost, DoNothingCombiner::default, seek_doc)
+            self.complex_scorer(reader, boost, DoNothingCombiner::default)
                 .map(|specialized_scorer| {
                     into_box_scorer(specialized_scorer, DoNothingCombiner::default, num_docs)
                 })
@@ -481,7 +443,7 @@ impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombin
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let mut scorer = self.scorer(reader, 1.0, 0)?;
+        let mut scorer = self.scorer(reader, 1.0)?;
         if scorer.seek(doc) != doc {
             return Err(does_not_match(doc));
         }
@@ -505,7 +467,7 @@ impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombin
         reader: &SegmentReader,
         callback: &mut dyn FnMut(DocId, Score),
     ) -> crate::Result<()> {
-        let scorer = self.complex_scorer(reader, 1.0, &self.score_combiner_fn, 0)?;
+        let scorer = self.complex_scorer(reader, 1.0, &self.score_combiner_fn)?;
         match scorer {
             SpecializedScorer::TermUnion(term_scorers) => {
                 let mut union_scorer = BufferedUnionScorer::build(
@@ -527,7 +489,7 @@ impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombin
         reader: &SegmentReader,
         callback: &mut dyn FnMut(&[DocId]),
     ) -> crate::Result<()> {
-        let scorer = self.complex_scorer(reader, 1.0, || DoNothingCombiner, 0u32)?;
+        let scorer = self.complex_scorer(reader, 1.0, || DoNothingCombiner)?;
         let mut buffer = [0u32; COLLECT_BLOCK_BUFFER_LEN];
 
         match scorer {
@@ -562,7 +524,7 @@ impl<TScoreCombiner: ScoreCombiner + Sync> Weight for BooleanWeight<TScoreCombin
         reader: &SegmentReader,
         callback: &mut dyn FnMut(DocId, Score) -> Score,
     ) -> crate::Result<()> {
-        let scorer = self.complex_scorer(reader, 1.0, &self.score_combiner_fn, 0u32)?;
+        let scorer = self.complex_scorer(reader, 1.0, &self.score_combiner_fn)?;
         match scorer {
             SpecializedScorer::TermUnion(term_scorers) => {
                 super::block_wand(term_scorers, threshold, callback);

src/query/boolean_query/mod.rs

@@ -57,7 +57,7 @@ mod tests {
         let query = query_parser.parse_query("+a")?;
         let searcher = index.reader()?.searcher();
         let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-        let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+        let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
         assert!(scorer.is::<TermScorer>());
         Ok(())
     }
@@ -70,13 +70,13 @@ mod tests {
         {
             let query = query_parser.parse_query("+a +b +c")?;
             let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
             assert!(scorer.is::<Intersection<TermScorer>>());
         }
         {
             let query = query_parser.parse_query("+a +(b c)")?;
             let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
             assert!(scorer.is::<Intersection<Box<dyn Scorer>>>());
         }
         Ok(())
@@ -90,14 +90,14 @@ mod tests {
         {
             let query = query_parser.parse_query("+a b")?;
             let weight = query.weight(EnableScoring::enabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
             assert!(scorer
                 .is::<RequiredOptionalScorer<Box<dyn Scorer>, Box<dyn Scorer>, SumCombiner>>());
         }
         {
             let query = query_parser.parse_query("+a b")?;
             let weight = query.weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
             assert!(scorer.is::<TermScorer>());
         }
         Ok(())
@@ -244,14 +244,12 @@ mod tests {
             .weight(EnableScoring::enabled_from_searcher(&searcher))
             .unwrap();
         {
-            let mut boolean_scorer =
-                boolean_weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+            let mut boolean_scorer = boolean_weight.scorer(searcher.segment_reader(0u32), 1.0)?;
             assert_eq!(boolean_scorer.doc(), 0u32);
             assert_nearly_equals!(boolean_scorer.score(), 0.84163445);
         }
         {
-            let mut boolean_scorer =
-                boolean_weight.scorer(searcher.segment_reader(0u32), 2.0, 0)?;
+            let mut boolean_scorer = boolean_weight.scorer(searcher.segment_reader(0u32), 2.0)?;
             assert_eq!(boolean_scorer.doc(), 0u32);
             assert_nearly_equals!(boolean_scorer.score(), 1.6832689);
         }
@@ -345,7 +343,7 @@ mod tests {
                 (Occur::Must, term_match_some.box_clone()),
             ]);
             let weight = query.weight(EnableScoring::disabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32)?;
             assert!(scorer.is::<TermScorer>());
         }
         {
@@ -355,7 +353,7 @@ mod tests {
                 (Occur::Must, term_match_none.box_clone()),
             ]);
             let weight = query.weight(EnableScoring::disabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32)?;
             assert!(scorer.is::<EmptyScorer>());
         }
         {
@@ -364,7 +362,7 @@ mod tests {
                 (Occur::Should, term_match_none.box_clone()),
             ]);
             let weight = query.weight(EnableScoring::disabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32)?;
             assert!(scorer.is::<AllScorer>());
         }
         {
@@ -373,7 +371,7 @@ mod tests {
                 (Occur::Should, term_match_none.box_clone()),
             ]);
             let weight = query.weight(EnableScoring::disabled_from_searcher(&searcher))?;
-            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32, 0)?;
+            let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0f32)?;
             assert!(scorer.is::<TermScorer>());
         }
         Ok(())
@@ -613,134 +611,6 @@ mod tests {
         Ok(())
     }
 
-    /// Test that the seek_doc parameter correctly skips documents in BooleanWeight::scorer.
-    ///
-    /// When seek_doc is provided, the scorer should start from that document (or the first
-    /// matching document >= seek_doc), skipping earlier documents.
-    #[test]
-    pub fn test_boolean_weight_seek_doc() -> crate::Result<()> {
-        let mut schema_builder = Schema::builder();
-        let text_field = schema_builder.add_text_field("text", TEXT);
-        let value_field = schema_builder.add_u64_field("value", FAST);
-        let schema = schema_builder.build();
-        let index = Index::create_in_ram(schema);
-        let mut index_writer: IndexWriter = index.writer_for_tests()?;
-
-        // Create 11 documents:
-        // doc 0: value=0
-        // doc 1: value=10
-        // doc 2: value=20
-        // ...
-        // doc 9: value=90
-        // doc 10: value=50 (matches range 30-70)
-        for i in 0..10 {
-            index_writer.add_document(doc!(
-                text_field => "hello",
-                value_field => (i * 10) as u64
-            ))?;
-        }
-        index_writer.add_document(doc!(
-            text_field => "hello",
-            value_field => 50u64
-        ))?;
-        index_writer.commit()?;
-
-        let searcher = index.reader()?.searcher();
-        let segment_reader = searcher.segment_reader(0);
-
-        // Create a Boolean query: MUST(term "hello") AND MUST(range 30..=70)
-        // This should match docs with value in [30, 70]: docs 3, 4, 5, 6, 7, 10
-        let term_query: Box<dyn Query> = Box::new(TermQuery::new(
-            Term::from_field_text(text_field, "hello"),
-            IndexRecordOption::Basic,
-        ));
-        let range_query: Box<dyn Query> = Box::new(RangeQuery::new(
-            Bound::Included(Term::from_field_u64(value_field, 30)),
-            Bound::Included(Term::from_field_u64(value_field, 70)),
-        ));
-
-        let boolean_query =
-            BooleanQuery::new(vec![(Occur::Must, term_query), (Occur::Must, range_query)]);
-
-        let weight =
-            boolean_query.weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
-
-        let doc_when_seeking_from = |seek_from: DocId| {
-            let scorer = weight.scorer(segment_reader, 1.0f32, seek_from).unwrap();
-            crate::docset::docset_to_doc_vec(scorer)
-        };
-
-        // Expected matching docs: 3, 4, 5, 6, 7, 10 (values 30, 40, 50, 60, 70, 50)
-        assert_eq!(doc_when_seeking_from(0), vec![3, 4, 5, 6, 7, 10]);
-        assert_eq!(doc_when_seeking_from(1), vec![3, 4, 5, 6, 7, 10]);
-        assert_eq!(doc_when_seeking_from(3), vec![3, 4, 5, 6, 7, 10]);
-        assert_eq!(doc_when_seeking_from(4), vec![4, 5, 6, 7, 10]);
-        assert_eq!(doc_when_seeking_from(7), vec![7, 10]);
-        assert_eq!(doc_when_seeking_from(8), vec![10]);
-        assert_eq!(doc_when_seeking_from(10), vec![10]);
-        assert_eq!(doc_when_seeking_from(11), Vec::<DocId>::new());
-
-        Ok(())
-    }
-
-    /// Test that the seek_doc parameter works correctly with SHOULD clauses.
-    #[test]
-    pub fn test_boolean_weight_seek_doc_with_should() -> crate::Result<()> {
-        let mut schema_builder = Schema::builder();
-        let text_field = schema_builder.add_text_field("text", TEXT);
-        let schema = schema_builder.build();
-        let index = Index::create_in_ram(schema);
-        let mut index_writer: IndexWriter = index.writer_for_tests()?;
-
-        // Create documents:
-        // doc 0: "a b"
-        // doc 1: "a"
-        // doc 2: "b"
-        // doc 3: "c"
-        // doc 4: "a b c"
-        index_writer.add_document(doc!(text_field => "a b"))?;
-        index_writer.add_document(doc!(text_field => "a"))?;
-        index_writer.add_document(doc!(text_field => "b"))?;
-        index_writer.add_document(doc!(text_field => "c"))?;
-        index_writer.add_document(doc!(text_field => "a b c"))?;
-        index_writer.commit()?;
-
-        let searcher = index.reader()?.searcher();
-        let segment_reader = searcher.segment_reader(0);
-
-        // Create a Boolean query: SHOULD(term "a") OR SHOULD(term "b")
-        // This should match docs 0, 1, 2, 4
-        let term_a: Box<dyn Query> = Box::new(TermQuery::new(
-            Term::from_field_text(text_field, "a"),
-            IndexRecordOption::Basic,
-        ));
-        let term_b: Box<dyn Query> = Box::new(TermQuery::new(
-            Term::from_field_text(text_field, "b"),
-            IndexRecordOption::Basic,
-        ));
-
-        let boolean_query =
-            BooleanQuery::new(vec![(Occur::Should, term_a), (Occur::Should, term_b)]);
-
-        let weight =
-            boolean_query.weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
-
-        let doc_when_seeking_from = |seek_from: DocId| {
-            let scorer = weight.scorer(segment_reader, 1.0f32, seek_from).unwrap();
-            crate::docset::docset_to_doc_vec(scorer)
-        };
-
-        // Expected matching docs: 0, 1, 2, 4
-        assert_eq!(doc_when_seeking_from(0), vec![0, 1, 2, 4]);
-        assert_eq!(doc_when_seeking_from(1), vec![1, 2, 4]);
-        assert_eq!(doc_when_seeking_from(2), vec![2, 4]);
-        assert_eq!(doc_when_seeking_from(3), vec![4]);
-        assert_eq!(doc_when_seeking_from(4), vec![4]);
-        assert_eq!(doc_when_seeking_from(5), Vec::<DocId>::new());
-
-        Ok(())
-    }
-
     /// Test multiple AllScorer instances in different clause types.
     ///
     /// Verifies correct behavior when AllScorers appear in multiple positions.

src/query/boost_query.rs

@@ -67,13 +67,8 @@ impl BoostWeight {
 }
 
 impl Weight for BoostWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
-        self.weight.scorer(reader, boost * self.boost, seek_doc)
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
+        self.weight.scorer(reader, boost * self.boost)
     }
 
     fn explain(&self, reader: &SegmentReader, doc: u32) -> crate::Result<Explanation> {
@@ -88,10 +83,6 @@ impl Weight for BoostWeight {
     fn count(&self, reader: &SegmentReader) -> crate::Result<u32> {
         self.weight.count(reader)
     }
-
-    fn intersection_priority(&self) -> u32 {
-        self.weight.intersection_priority()
-    }
 }
 
 pub(crate) struct BoostScorer<S: Scorer> {
@@ -143,6 +134,7 @@ impl<S: Scorer> DocSet for BoostScorer<S> {
 }
 
 impl<S: Scorer> Scorer for BoostScorer<S> {
+    #[inline]
     fn score(&mut self) -> Score {
         self.underlying.score() * self.boost
     }

src/query/const_score_query.rs

@@ -63,18 +63,13 @@ impl ConstWeight {
 }
 
 impl Weight for ConstWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
-        let inner_scorer = self.weight.scorer(reader, boost, seek_doc)?;
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
+        let inner_scorer = self.weight.scorer(reader, boost)?;
         Ok(Box::new(ConstScorer::new(inner_scorer, boost * self.score)))
     }
 
     fn explain(&self, reader: &SegmentReader, doc: u32) -> crate::Result<Explanation> {
-        let mut scorer = self.scorer(reader, 1.0, 0)?;
+        let mut scorer = self.scorer(reader, 1.0)?;
         if scorer.seek(doc) != doc {
             return Err(TantivyError::InvalidArgument(format!(
                 "Document #({doc}) does not match"
@@ -89,10 +84,6 @@ impl Weight for ConstWeight {
     fn count(&self, reader: &SegmentReader) -> crate::Result<u32> {
         self.weight.count(reader)
     }
-
-    fn intersection_priority(&self) -> u32 {
-        self.weight.intersection_priority()
-    }
 }
 
 /// Wraps a `DocSet` and simply returns a constant `Scorer`.
@@ -146,6 +137,7 @@ impl<TDocSet: DocSet> DocSet for ConstScorer<TDocSet> {
 }
 
 impl<TDocSet: DocSet + 'static> Scorer for ConstScorer<TDocSet> {
+    #[inline]
     fn score(&mut self) -> Score {
         self.score
     }

src/query/disjunction.rs

@@ -173,6 +173,7 @@ impl<TScorer: Scorer, TScoreCombiner: ScoreCombiner> DocSet
 impl<TScorer: Scorer, TScoreCombiner: ScoreCombiner> Scorer
     for Disjunction<TScorer, TScoreCombiner>
 {
+    #[inline]
     fn score(&mut self) -> Score {
         self.current_score
     }
@@ -307,6 +308,7 @@ mod tests {
     }
 
     impl Scorer for DummyScorer {
+        #[inline]
         fn score(&mut self) -> Score {
             self.foo.get(self.cursor).map(|x| x.1).unwrap_or(0.0)
         }

src/query/empty_query.rs

@@ -26,24 +26,13 @@ impl Query for EmptyQuery {
 /// It is useful for tests and handling edge cases.
 pub struct EmptyWeight;
 impl Weight for EmptyWeight {
-    fn scorer(
-        &self,
-        _reader: &SegmentReader,
-        _boost: Score,
-        _seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, _reader: &SegmentReader, _boost: Score) -> crate::Result<Box<dyn Scorer>> {
         Ok(Box::new(EmptyScorer))
     }
 
     fn explain(&self, _reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
         Err(does_not_match(doc))
     }
-
-    /// Returns a priority number used to sort weights when running an
-    /// intersection.
-    fn intersection_priority(&self) -> u32 {
-        0u32
-    }
 }
 
 /// `EmptyScorer` is a dummy `Scorer` in which no document matches.
@@ -66,6 +55,7 @@ impl DocSet for EmptyScorer {
 }
 
 impl Scorer for EmptyScorer {
+    #[inline]
     fn score(&mut self) -> Score {
         0.0
     }

src/query/exclude.rs

@@ -84,6 +84,7 @@ where
     TScorer: Scorer,
     TDocSetExclude: DocSet + 'static,
 {
+    #[inline]
     fn score(&mut self) -> Score {
         self.underlying_docset.score()
     }

src/query/exist_query.rs

@@ -98,12 +98,7 @@ pub struct ExistsWeight {
 }
 
 impl Weight for ExistsWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        _seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         let fast_field_reader = reader.fast_fields();
         let mut column_handles = fast_field_reader.dynamic_column_handles(&self.field_name)?;
         if self.field_type == Type::Json && self.json_subpaths {
@@ -171,7 +166,7 @@ impl Weight for ExistsWeight {
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let mut scorer = self.scorer(reader, 1.0, 0)?;
+        let mut scorer = self.scorer(reader, 1.0)?;
         if scorer.seek(doc) != doc {
             return Err(does_not_match(doc));
         }

src/query/intersection.rs

@@ -105,6 +105,7 @@ impl<TDocSet: DocSet> Intersection<TDocSet, TDocSet> {
 }
 
 impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOtherDocSet> {
+    #[inline]
     fn advance(&mut self) -> DocId {
         let (left, right) = (&mut self.left, &mut self.right);
         let mut candidate = left.advance();
@@ -174,6 +175,7 @@ impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOt
                 .all(|docset| docset.seek_into_the_danger_zone(target))
     }
 
+    #[inline]
     fn doc(&self) -> DocId {
         self.left.doc()
     }
@@ -200,6 +202,7 @@ where
     TScorer: Scorer,
     TOtherScorer: Scorer,
 {
+    #[inline]
     fn score(&mut self) -> Score {
         self.left.score()
             + self.right.score()

src/query/mod.rs

@@ -24,7 +24,7 @@ mod reqopt_scorer;
 mod scorer;
 mod set_query;
 mod size_hint;
-mod term_query;
+pub(crate) mod term_query;
 mod union;
 mod weight;
 

src/query/phrase_prefix_query/phrase_prefix_scorer.rs

@@ -81,6 +81,7 @@ impl<TPostings: Postings> DocSet for PhraseKind<TPostings> {
 }
 
 impl<TPostings: Postings> Scorer for PhraseKind<TPostings> {
+    #[inline]
     fn score(&mut self) -> Score {
         match self {
             PhraseKind::SinglePrefix { positions, .. } => {
@@ -215,6 +216,7 @@ impl<TPostings: Postings> DocSet for PhrasePrefixScorer<TPostings> {
 }
 
 impl<TPostings: Postings> Scorer for PhrasePrefixScorer<TPostings> {
+    #[inline]
     fn score(&mut self) -> Score {
         // TODO modify score??
         self.phrase_scorer.score()

src/query/phrase_prefix_query/phrase_prefix_weight.rs

@@ -1,7 +1,7 @@
 use super::{prefix_end, PhrasePrefixScorer};
 use crate::fieldnorm::FieldNormReader;
 use crate::index::SegmentReader;
-use crate::postings::SegmentPostings;
+use crate::postings::Postings;
 use crate::query::bm25::Bm25Weight;
 use crate::query::explanation::does_not_match;
 use crate::query::{EmptyScorer, Explanation, Scorer, Weight};
@@ -42,29 +42,26 @@ impl PhrasePrefixWeight {
         Ok(FieldNormReader::constant(reader.max_doc(), 1))
     }
 
-    pub(crate) fn prefix_phrase_scorer(
+    pub(crate) fn phrase_scorer(
         &self,
         reader: &SegmentReader,
         boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Option<PhrasePrefixScorer<SegmentPostings>>> {
+    ) -> crate::Result<Option<Box<dyn Scorer>>> {
         let similarity_weight_opt = self
             .similarity_weight_opt
             .as_ref()
             .map(|similarity_weight| similarity_weight.boost_by(boost));
         let fieldnorm_reader = self.fieldnorm_reader(reader)?;
-        let mut term_postings_list = Vec::new();
+        let mut term_postings_list: Vec<(usize, Box<dyn Postings>)> = Vec::new();
         for &(offset, ref term) in &self.phrase_terms {
-            let inverted_index = reader.inverted_index(term.field())?;
-            let Some(term_info) = inverted_index.get_term_info(term)? else {
+            if let Some(postings) = reader
+                .inverted_index(term.field())?
+                .read_postings(term, IndexRecordOption::WithFreqsAndPositions)?
+            {
+                term_postings_list.push((offset, postings));
+            } else {
                 return Ok(None);
-            };
-            let postings = inverted_index.read_postings_from_terminfo(
-                &term_info,
-                IndexRecordOption::WithFreqsAndPositions,
-                seek_doc,
-            )?;
-            term_postings_list.push((offset, postings));
+            }
         }
 
         let inv_index = reader.inverted_index(self.prefix.1.field())?;
@@ -106,51 +103,44 @@ impl PhrasePrefixWeight {
             }
         }
 
-        Ok(Some(PhrasePrefixScorer::new(
+        // TODO make this specialized.
+        Ok(Some(Box::new(PhrasePrefixScorer::new(
             term_postings_list,
             similarity_weight_opt,
             fieldnorm_reader,
             suffixes,
             self.prefix.0,
-        )))
+        ))))
     }
 }
 
 impl Weight for PhrasePrefixWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
-        if let Some(scorer) = self.prefix_phrase_scorer(reader, boost, seek_doc)? {
-            Ok(Box::new(scorer))
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
+        if let Some(scorer) = self.phrase_scorer(reader, boost)? {
+            Ok(scorer)
         } else {
             Ok(Box::new(EmptyScorer))
         }
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let scorer_opt = self.prefix_phrase_scorer(reader, 1.0, doc)?;
-        if scorer_opt.is_none() {
-            return Err(does_not_match(doc));
-        }
-        let mut scorer = scorer_opt.unwrap();
-        if scorer.seek(doc) != doc {
-            return Err(does_not_match(doc));
-        }
-        let fieldnorm_reader = self.fieldnorm_reader(reader)?;
-        let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
-        let phrase_count = scorer.phrase_count();
-        let mut explanation = Explanation::new("Phrase Prefix Scorer", scorer.score());
-        if let Some(similarity_weight) = self.similarity_weight_opt.as_ref() {
-            explanation.add_detail(similarity_weight.explain(fieldnorm_id, phrase_count));
-        }
-        Ok(explanation)
-    }
-
-    fn intersection_priority(&self) -> u32 {
-        50u32
+        todo!();
+        // let scorer_opt = self.phrase_scorer(reader, 1.0)?;
+        // if scorer_opt.is_none() {
+        //     return Err(does_not_match(doc));
+        // }
+        // let mut scorer = scorer_opt.unwrap();
+        // if scorer.seek(doc) != doc {
+        //     return Err(does_not_match(doc));
+        // }
+        // let fieldnorm_reader = self.fieldnorm_reader(reader)?;
+        // let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
+        // let phrase_count = scorer.phrase_count();
+        // let mut explanation = Explanation::new("Phrase Prefix Scorer", scorer.score());
+        // if let Some(similarity_weight) = self.similarity_weight_opt.as_ref() {
+        //     explanation.add_detail(similarity_weight.explain(fieldnorm_id, phrase_count));
+        // }
+        // Ok(explanation)
     }
 }
 
@@ -199,7 +189,7 @@ mod tests {
             .unwrap()
             .unwrap();
         let mut phrase_scorer = phrase_weight
-            .prefix_phrase_scorer(searcher.segment_reader(0u32), 1.0, 0u32)?
+            .phrase_scorer(searcher.segment_reader(0u32), 1.0)?
             .unwrap();
         assert_eq!(phrase_scorer.doc(), 1);
         assert_eq!(phrase_scorer.phrase_count(), 2);
@@ -226,7 +216,7 @@ mod tests {
             .unwrap()
             .unwrap();
         let mut phrase_scorer = phrase_weight
-            .prefix_phrase_scorer(searcher.segment_reader(0u32), 1.0, 0u32)?
+            .phrase_scorer(searcher.segment_reader(0u32), 1.0)?
             .unwrap();
         assert_eq!(phrase_scorer.doc(), 1);
         assert_eq!(phrase_scorer.phrase_count(), 2);
@@ -250,7 +240,7 @@ mod tests {
             .unwrap()
             .is_none());
         let weight = phrase_query.weight(enable_scoring).unwrap();
-        let mut phrase_scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+        let mut phrase_scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
         assert_eq!(phrase_scorer.doc(), 1);
         assert_eq!(phrase_scorer.advance(), 2);
         assert_eq!(phrase_scorer.doc(), 2);
@@ -271,7 +261,7 @@ mod tests {
         ]);
         let enable_scoring = EnableScoring::enabled_from_searcher(&searcher);
         let weight = phrase_query.weight(enable_scoring).unwrap();
-        let mut phrase_scorer = weight.scorer(searcher.segment_reader(0u32), 1.0, 0)?;
+        let mut phrase_scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
         assert_eq!(phrase_scorer.advance(), TERMINATED);
         Ok(())
     }

src/query/phrase_query/mod.rs

@@ -84,7 +84,7 @@ pub(crate) mod tests {
         let phrase_query = PhraseQuery::new(terms);
         let phrase_weight =
             phrase_query.phrase_weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
-        let mut phrase_scorer = phrase_weight.scorer(searcher.segment_reader(0), 1.0, 0)?;
+        let mut phrase_scorer = phrase_weight.scorer(searcher.segment_reader(0), 1.0)?;
         assert_eq!(phrase_scorer.doc(), 1);
         assert_eq!(phrase_scorer.advance(), TERMINATED);
         Ok(())
@@ -343,43 +343,6 @@ pub(crate) mod tests {
         Ok(())
     }
 
-    #[test]
-    pub fn test_phrase_weight_seek_doc() -> crate::Result<()> {
-        // Create an index with documents where the phrase "a b" appears in some of them.
-        // Documents: 0: "c d", 1: "a b", 2: "e f", 3: "a b c", 4: "g h", 5: "a b", 6: "i j"
-        let index = create_index(&["c d", "a b", "e f", "a b c", "g h", "a b", "i j"])?;
-        let text_field = index.schema().get_field("text").unwrap();
-        let searcher = index.reader()?.searcher();
-        let segment_reader = searcher.segment_reader(0);
-
-        let phrase_query = PhraseQuery::new(vec![
-            Term::from_field_text(text_field, "a"),
-            Term::from_field_text(text_field, "b"),
-        ]);
-        let phrase_weight =
-            phrase_query.phrase_weight(EnableScoring::disabled_from_schema(searcher.schema()))?;
-
-        // Helper function to collect all docs from a scorer created with a given seek_doc
-        let docs_when_seeking_from = |seek_from: DocId| {
-            let scorer = phrase_weight
-                .scorer(segment_reader, 1.0f32, seek_from)
-                .unwrap();
-            crate::docset::docset_to_doc_vec(scorer)
-        };
-
-        // Documents with "a b": 1, 3, 5
-        assert_eq!(docs_when_seeking_from(0), vec![1, 3, 5]);
-        assert_eq!(docs_when_seeking_from(1), vec![1, 3, 5]);
-        assert_eq!(docs_when_seeking_from(2), vec![3, 5]);
-        assert_eq!(docs_when_seeking_from(3), vec![3, 5]);
-        assert_eq!(docs_when_seeking_from(4), vec![5]);
-        assert_eq!(docs_when_seeking_from(5), vec![5]);
-        assert_eq!(docs_when_seeking_from(6), Vec::<DocId>::new());
-        assert_eq!(docs_when_seeking_from(7), Vec::<DocId>::new());
-
-        Ok(())
-    }
-
     #[test]
     pub fn test_phrase_query_on_json() -> crate::Result<()> {
         let mut schema_builder = Schema::builder();
@@ -410,7 +373,7 @@ pub(crate) mod tests {
                 .weight(EnableScoring::disabled_from_schema(searcher.schema()))
                 .unwrap();
             let mut phrase_scorer = phrase_weight
-                .scorer(searcher.segment_reader(0), 1.0f32, 0)
+                .scorer(searcher.segment_reader(0), 1.0f32)
                 .unwrap();
             let mut docs = Vec::new();
             loop {

src/query/phrase_query/phrase_scorer.rs

@@ -563,6 +563,7 @@ impl<TPostings: Postings> DocSet for PhraseScorer<TPostings> {
 }
 
 impl<TPostings: Postings> Scorer for PhraseScorer<TPostings> {
+    #[inline]
     fn score(&mut self) -> Score {
         let doc = self.doc();
         let fieldnorm_id = self.fieldnorm_reader.fieldnorm_id(doc);

src/query/phrase_query/phrase_weight.rs

@@ -1,7 +1,6 @@
 use super::PhraseScorer;
 use crate::fieldnorm::FieldNormReader;
 use crate::index::SegmentReader;
-use crate::postings::SegmentPostings;
 use crate::query::bm25::Bm25Weight;
 use crate::query::explanation::does_not_match;
 use crate::query::{EmptyScorer, Explanation, Scorer, Weight};
@@ -43,32 +42,30 @@ impl PhraseWeight {
         &self,
         reader: &SegmentReader,
         boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Option<PhraseScorer<SegmentPostings>>> {
+    ) -> crate::Result<Option<Box<dyn Scorer>>> {
         let similarity_weight_opt = self
             .similarity_weight_opt
             .as_ref()
             .map(|similarity_weight| similarity_weight.boost_by(boost));
         let fieldnorm_reader = self.fieldnorm_reader(reader)?;
         let mut term_postings_list = Vec::new();
+        // TODO make it specialized
         for &(offset, ref term) in &self.phrase_terms {
-            let inverted_index = reader.inverted_index(term.field())?;
-            let Some(term_info) = inverted_index.get_term_info(term)? else {
+            if let Some(postings) = reader
+                .inverted_index(term.field())?
+                .read_postings(term, IndexRecordOption::WithFreqsAndPositions)?
+            {
+                term_postings_list.push((offset, postings));
+            } else {
                 return Ok(None);
-            };
-            let postings = inverted_index.read_postings_from_terminfo(
-                &term_info,
-                IndexRecordOption::WithFreqsAndPositions,
-                seek_doc,
-            )?;
-            term_postings_list.push((offset, postings));
+            }
         }
-        Ok(Some(PhraseScorer::new(
+        Ok(Some(Box::new(PhraseScorer::new(
             term_postings_list,
             similarity_weight_opt,
             fieldnorm_reader,
             self.slop,
-        )))
+        ))))
     }
 
     pub fn slop(&mut self, slop: u32) {
@@ -77,40 +74,32 @@ impl PhraseWeight {
 }
 
 impl Weight for PhraseWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
-        if let Some(scorer) = self.phrase_scorer(reader, boost, seek_doc)? {
-            Ok(Box::new(scorer))
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
+        if let Some(scorer) = self.phrase_scorer(reader, boost)? {
+            Ok(scorer)
         } else {
             Ok(Box::new(EmptyScorer))
         }
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let scorer_opt = self.phrase_scorer(reader, 1.0, doc)?;
-        if scorer_opt.is_none() {
-            return Err(does_not_match(doc));
-        }
-        let mut scorer = scorer_opt.unwrap();
-        if scorer.doc() != doc {
-            return Err(does_not_match(doc));
-        }
-        let fieldnorm_reader = self.fieldnorm_reader(reader)?;
-        let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
-        let phrase_count = scorer.phrase_count();
-        let mut explanation = Explanation::new("Phrase Scorer", scorer.score());
-        if let Some(similarity_weight) = self.similarity_weight_opt.as_ref() {
-            explanation.add_detail(similarity_weight.explain(fieldnorm_id, phrase_count));
-        }
-        Ok(explanation)
-    }
-
-    fn intersection_priority(&self) -> u32 {
-        40u32
+        todo!();
+        // let scorer_opt = self.phrase_scorer(reader, 1.0)?;
+        // if scorer_opt.is_none() {
+        //     return Err(does_not_match(doc));
+        // }
+        // let mut scorer = scorer_opt.unwrap();
+        // if scorer.seek(doc) != doc {
+        //     return Err(does_not_match(doc));
+        // }
+        // let fieldnorm_reader = self.fieldnorm_reader(reader)?;
+        // let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
+        // let phrase_count = scorer.phrase_count();
+        // let mut explanation = Explanation::new("Phrase Scorer", scorer.score());
+        // if let Some(similarity_weight) = self.similarity_weight_opt.as_ref() {
+        //     explanation.add_detail(similarity_weight.explain(fieldnorm_id, phrase_count));
+        // }
+        // Ok(explanation)
     }
 }
 
@@ -134,7 +123,7 @@ mod tests {
         let enable_scoring = EnableScoring::enabled_from_searcher(&searcher);
         let phrase_weight = phrase_query.phrase_weight(enable_scoring).unwrap();
         let mut phrase_scorer = phrase_weight
-            .phrase_scorer(searcher.segment_reader(0u32), 1.0, 0)?
+            .phrase_scorer(searcher.segment_reader(0u32), 1.0)?
             .unwrap();
         assert_eq!(phrase_scorer.doc(), 1);
         assert_eq!(phrase_scorer.phrase_count(), 2);

src/query/phrase_query/regex_phrase_weight.rs

@@ -6,7 +6,7 @@ use tantivy_fst::Regex;
 use super::PhraseScorer;
 use crate::fieldnorm::FieldNormReader;
 use crate::index::SegmentReader;
-use crate::postings::{LoadedPostings, Postings, SegmentPostings, TermInfo};
+use crate::postings::{LoadedPostings, Postings, TermInfo};
 use crate::query::bm25::Bm25Weight;
 use crate::query::explanation::does_not_match;
 use crate::query::union::{BitSetPostingUnion, SimpleUnion};
@@ -103,18 +103,9 @@ impl RegexPhraseWeight {
         term_info: &TermInfo,
         doc_bitset: &mut BitSet,
     ) -> crate::Result<()> {
-        let mut block_segment_postings = inverted_index
-            .read_block_postings_from_terminfo(term_info, IndexRecordOption::Basic)?;
-        loop {
-            let docs = block_segment_postings.docs();
-            if docs.is_empty() {
-                break;
-            }
-            for &doc in docs {
-                doc_bitset.insert(doc);
-            }
-            block_segment_postings.advance();
-        }
+        let mut segment_postings =
+            inverted_index.read_postings_from_terminfo(term_info, IndexRecordOption::Basic)?;
+        segment_postings.fill_bitset(doc_bitset);
         Ok(())
     }
 
@@ -188,18 +179,15 @@ impl RegexPhraseWeight {
         // - Bucket 1: Terms appearing in 0.1% to 1% of documents
         // - Bucket 2: Terms appearing in 1% to 10% of documents
         // - Bucket 3: Terms appearing in more than 10% of documents
-        let mut buckets: Vec<(BitSet, Vec<SegmentPostings>)> = (0..4)
+        let mut buckets: Vec<(BitSet, Vec<Box<dyn Postings>>)> = (0..4)
             .map(|_| (BitSet::with_max_value(max_doc), Vec::new()))
             .collect();
 
         const SPARSE_TERM_DOC_THRESHOLD: u32 = 100;
 
         for term_info in term_infos {
-            let mut term_posting = inverted_index.read_postings_from_terminfo(
-                term_info,
-                IndexRecordOption::WithFreqsAndPositions,
-                0u32,
-            )?;
+            let mut term_posting = inverted_index
+                .read_postings_from_terminfo(term_info, IndexRecordOption::WithFreqsAndPositions)?;
             let num_docs = term_posting.doc_freq();
 
             if num_docs < SPARSE_TERM_DOC_THRESHOLD {
@@ -272,12 +260,7 @@ impl RegexPhraseWeight {
 }
 
 impl Weight for RegexPhraseWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        _seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         if let Some(scorer) = self.phrase_scorer(reader, boost)? {
             Ok(Box::new(scorer))
         } else {

src/query/range_query/fast_field_range_doc_set.rs

@@ -61,11 +61,7 @@ pub(crate) struct RangeDocSet<T> {
 
 const DEFAULT_FETCH_HORIZON: u32 = 128;
 impl<T: Send + Sync + PartialOrd + Copy + Debug + 'static> RangeDocSet<T> {
-    pub(crate) fn new(
-        value_range: RangeInclusive<T>,
-        column: Column<T>,
-        seek_first_doc: DocId,
-    ) -> Self {
+    pub(crate) fn new(value_range: RangeInclusive<T>, column: Column<T>) -> Self {
         if *value_range.start() > column.max_value() || *value_range.end() < column.min_value() {
             return Self {
                 value_range,
@@ -81,7 +77,7 @@ impl<T: Send + Sync + PartialOrd + Copy + Debug + 'static> RangeDocSet<T> {
             value_range,
             column,
             loaded_docs: VecCursor::new(),
-            next_fetch_start: seek_first_doc,
+            next_fetch_start: 0,
             fetch_horizon: DEFAULT_FETCH_HORIZON,
             last_seek_pos_opt: None,
         };

src/query/range_query/range_query.rs

@@ -11,7 +11,7 @@ use crate::query::range_query::is_type_valid_for_fastfield_range_query;
 use crate::query::{BitSetDocSet, ConstScorer, EnableScoring, Explanation, Query, Scorer, Weight};
 use crate::schema::{Field, IndexRecordOption, Term, Type};
 use crate::termdict::{TermDictionary, TermStreamer};
-use crate::{DocId, Score};
+use crate::{DocId, DocSet, Score};
 
 /// `RangeQuery` matches all documents that have at least one term within a defined range.
 ///
@@ -212,12 +212,7 @@ impl InvertedIndexRangeWeight {
 }
 
 impl Weight for InvertedIndexRangeWeight {
-    fn scorer(
-        &self,
-        reader: &SegmentReader,
-        boost: Score,
-        seek_doc: DocId,
-    ) -> crate::Result<Box<dyn Scorer>> {
+    fn scorer(&self, reader: &SegmentReader, boost: Score) -> crate::Result<Box<dyn Scorer>> {
         let max_doc = reader.max_doc();
         let mut doc_bitset = BitSet::with_max_value(max_doc);
 
@@ -233,30 +228,16 @@ impl Weight for InvertedIndexRangeWeight {
             }
             processed_count += 1;
             let term_info = term_range.value();
-            let mut block_segment_postings = inverted_index
-                .read_block_postings_from_terminfo_with_seek(
-                    term_info,
-                    IndexRecordOption::Basic,
-                    seek_doc,
-                )?
-                .0;
-            loop {
-                let docs = block_segment_postings.docs();
-                if docs.is_empty() {
-                    break;
-                }
-                for &doc in block_segment_postings.docs() {
-                    doc_bitset.insert(doc);
-                }
-                block_segment_postings.advance();
-            }
+            let mut postings =
+                inverted_index.read_postings_from_terminfo(term_info, IndexRecordOption::Basic)?;
+            postings.fill_bitset(&mut doc_bitset);
         }
         let doc_bitset = BitSetDocSet::from(doc_bitset);
         Ok(Box::new(ConstScorer::new(doc_bitset, boost)))
     }
 
     fn explain(&self, reader: &SegmentReader, doc: DocId) -> crate::Result<Explanation> {
-        let mut scorer = self.scorer(reader, 1.0, 0)?;
+        let mut scorer = self.scorer(reader, 1.0)?;
         if scorer.seek(doc) != doc {
             return Err(does_not_match(doc));
         }
@@ -696,7 +677,7 @@ mod tests {
                 .weight(EnableScoring::disabled_from_schema(&schema))
                 .unwrap();
             let range_scorer = range_weight
-                .scorer(&searcher.segment_readers()[0], 1.0f32, 0)
+                .scorer(&searcher.segment_readers()[0], 1.0f32)
                 .unwrap();
             range_scorer
         };