Preparing for hotfix release 0.9.1

.travis.yml

@@ -29,7 +29,7 @@ addons:
 matrix:
   include:
     # Android
-    - env: TARGET=aarch64-linux-android DISABLE_TESTS=1
+    - env: TARGET=aarch64-linux-android
     #- env: TARGET=arm-linux-androideabi DISABLE_TESTS=1
     #- env: TARGET=armv7-linux-androideabi DISABLE_TESTS=1
     #- env: TARGET=i686-linux-android DISABLE_TESTS=1
@@ -68,11 +68,6 @@ cache: cargo
 before_cache:
   # Travis can't cache files that are not readable by "others"
   - chmod -R a+r $HOME/.cargo
-  - find ./target/debug -type f -maxdepth 1 -delete
-  - rm -f  ./target/.rustc_info.json
-  - rm -fr ./target/debug/{deps,.fingerprint}/tantivy*
-  - rm -r target/debug/examples/
-  - ls -1 examples/ | sed -e 's/\.rs$//' | xargs -I "{}" find target/* -name "*{}*" -type f -delete
 
 #branches:
 #  only:
@@ -82,4 +77,4 @@ before_cache:
 
 notifications:
   email:
-    on_success: never
+    on_success: never

CHANGELOG.md

@@ -1,34 +1,7 @@
-Tantivy 0.10.0
+Tantivy 0.9.1
 =====================
 
-*Tantivy 0.10.0 index format is compatible with the index format in 0.9.0.*
-
-- Added an ASCII folding filter (@drusellers)
-- Bugfix in `query.count` in presence of deletes (@pmasurel)
-
-Minor
----------
-- Small simplification of the code. 
-Calling .freq() or .doc() when .advance() has never 
-on segment postings should panic from now on.
-- Tokens exceeding `u16::max_value() - 4` chars are discarded silently instead of panicking.
-- Fast fields are now preloaded when the `SegmentReader` is created.
-
-## How to update?
-
-Your existing indexes are usable as is. Your may or may need some 
-trivial updates.
-
-### Fast fields
-
-Fast fields used to be accessed directly from the `SegmentReader`.
-The API changed, you are now required to acquire your fast field reader via the
-`segment_reader.fast_fields()`, and use one of the typed method: 
-- `.u64()`, `.i64()` if your field is single-valued ;
-- `.u64s()`, `.i64s()` if your field is multi-valued ;
-- `.bytes()` if your field is bytes fast field.
-
-
+Hotfix: The english stemmer was actually used for all languages.
 
 Tantivy 0.9.0
 =====================
@@ -47,35 +20,6 @@ previous index format.*
   for int fields. (@fulmicoton)
 - Added DateTime field (@barrotsteindev)
 - Added IndexReader. By default, index is reloaded automatically upon new commits (@fulmicoton)
-- SIMD linear search within blocks (@fulmicoton)
-
-## How to update ?
-
-tantivy 0.9 brought some API breaking change.
-To update from tantivy 0.8, you will need to go through the following steps.
-
-- `schema::INT_INDEXED` and `schema::INT_STORED`  should be replaced by `schema::INDEXED` and `schema::INT_STORED`.
-- The index now does not hold the pool of searcher anymore. You are required to create an intermediary object called 
-`IndexReader` for this. 
-    
-    ```rust
-    // create the reader. You typically need to create 1 reader for the entire
-    // lifetime of you program.
-    let reader = index.reader()?;
-    
-    // Acquire a searcher (previously `index.searcher()`) is now written:
-    let searcher = reader.searcher();
-    
-    // With the default setting of the reader, you are not required to 
-    // call `index.load_searchers()` anymore.
-    //
-    // The IndexReader will pick up that change automatically, regardless
-    // of whether the update was done in a different process or not.
-    // If this behavior is not wanted, you can create your reader with 
-    // the `ReloadPolicy::Manual`, and manually decide when to reload the index
-    // by calling `reader.reload()?`.
-  
-    ```
 
 
 Tantivy 0.8.2

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "tantivy"
-version = "0.10.0-dev"
+version = "0.9.1"
 authors = ["Paul Masurel <paul.masurel@gmail.com>"]
 license = "MIT"
 categories = ["database-implementations", "data-structures"]
@@ -23,7 +23,7 @@ snap = {version="0.2"}
 atomicwrites = {version="0.2.2", optional=true}
 tempfile = "3.0"
 log = "0.4"
-combine = ">=3.6.0,<4.0.0"
+combine = "3"
 tempdir = "0.3"
 serde = "1.0"
 serde_derive = "1.0"

README.md

@@ -17,7 +17,6 @@
 [![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/6)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/6)
 [![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/7)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/7)
 
-[![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
 
 
 **Tantivy** is a **full text search engine library** written in rust.
@@ -28,14 +27,6 @@ to build such a search engine.
 
 Tantivy is, in fact, strongly inspired by Lucene's design.
 
-# Benchmark
-
-Tantivy is typically faster than Lucene, but the results will depend on 
-the nature of the queries in your workload.
-
-The following [benchmark](https://tantivy-search.github.io/bench/) break downs 
-performance for different type of queries / collection.
-
 # Features
 
 - Full-text search
@@ -96,14 +87,6 @@ To check out and run tests, you can simply run :
 Some tests will not run with just `cargo test` because of `fail-rs`.
 To run the tests exhaustively, run `./run-tests.sh`. 
 
-# How can I support this project ?
+# Contribute
 
-There are many ways to support this project. 
-
-- If you use tantivy, tell us about your experience on [gitter](https://gitter.im/tantivy-search/tantivy) or by email (paul.masurel@gmail.com)
-- Report bugs
-- Write a blog post
-- Complete documentation
-- Contribute code (you can join [our gitter](https://gitter.im/tantivy-search/tantivy) )
-- Talk about tantivy around you
-- Drop a word on on [![Say Thanks!](https://img.shields.io/badge/Say%20Thanks-!-1EAEDB.svg)](https://saythanks.io/to/fulmicoton) or even [![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
+Send me an email (paul.masurel at gmail.com) if you want to contribute to tantivy.

examples/custom_collector.rs

@@ -18,8 +18,8 @@ use tantivy::fastfield::FastFieldReader;
 use tantivy::query::QueryParser;
 use tantivy::schema::Field;
 use tantivy::schema::{Schema, FAST, INDEXED, TEXT};
+use tantivy::Index;
 use tantivy::SegmentReader;
-use tantivy::{Index, TantivyError};
 
 #[derive(Default)]
 struct Stats {
@@ -75,18 +75,9 @@ impl Collector for StatsCollector {
     fn for_segment(
         &self,
         _segment_local_id: u32,
-        segment_reader: &SegmentReader,
+        segment: &SegmentReader,
     ) -> tantivy::Result<StatsSegmentCollector> {
-        let fast_field_reader = segment_reader
-            .fast_fields()
-            .u64(self.field)
-            .ok_or_else(|| {
-                let field_name = segment_reader.schema().get_field_name(self.field);
-                TantivyError::SchemaError(format!(
-                    "Field {:?} is not a u64 fast field.",
-                    field_name
-                ))
-            })?;
+        let fast_field_reader = segment.fast_field_reader(self.field)?;
         Ok(StatsSegmentCollector {
             fast_field_reader,
             stats: Stats::default(),

src/collector/facet_collector.rs

@@ -17,7 +17,6 @@ use Result;
 use Score;
 use SegmentLocalId;
 use SegmentReader;
-use TantivyError;
 
 struct Hit<'a> {
     count: u64,
@@ -265,10 +264,7 @@ impl Collector for FacetCollector {
         _: SegmentLocalId,
         reader: &SegmentReader,
     ) -> Result<FacetSegmentCollector> {
-        let field_name = reader.schema().get_field_name(self.field);
-        let facet_reader = reader.facet_reader(self.field).ok_or_else(|| {
-            TantivyError::SchemaError(format!("Field {:?} is not a facet field.", field_name))
-        })?;
+        let facet_reader = reader.facet_reader(self.field)?;
 
         let mut collapse_mapping = Vec::new();
         let mut counts = Vec::new();

src/collector/multi_collector.rs

@@ -2,7 +2,6 @@ use super::Collector;
 use super::SegmentCollector;
 use collector::Fruit;
 use std::marker::PhantomData;
-use std::ops::Deref;
 use DocId;
 use Result;
 use Score;
@@ -200,10 +199,7 @@ impl<'a> Collector for MultiCollector<'a> {
     }
 
     fn requires_scoring(&self) -> bool {
-        self.collector_wrappers
-            .iter()
-            .map(Deref::deref)
-            .any(Collector::requires_scoring)
+        self.collector_wrappers.iter().any(|c| c.requires_scoring())
     }
 
     fn merge_fruits(&self, segments_multifruits: Vec<MultiFruit>) -> Result<MultiFruit> {

src/collector/tests.rs

@@ -114,15 +114,11 @@ impl Collector for FastFieldTestCollector {
     fn for_segment(
         &self,
         _: SegmentLocalId,
-        segment_reader: &SegmentReader,
+        reader: &SegmentReader,
     ) -> Result<FastFieldSegmentCollector> {
-        let reader = segment_reader
-            .fast_fields()
-            .u64(self.field)
-            .expect("Requested field is not a fast field.");
         Ok(FastFieldSegmentCollector {
             vals: Vec::new(),
-            reader,
+            reader: reader.fast_field_reader(self.field)?,
         })
     }
 
@@ -174,14 +170,11 @@ impl Collector for BytesFastFieldTestCollector {
     fn for_segment(
         &self,
         _segment_local_id: u32,
-        segment_reader: &SegmentReader,
+        segment: &SegmentReader,
     ) -> Result<BytesFastFieldSegmentCollector> {
         Ok(BytesFastFieldSegmentCollector {
             vals: Vec::new(),
-            reader: segment_reader
-                .fast_fields()
-                .bytes(self.field)
-                .expect("Field is not a bytes fast field."),
+            reader: segment.bytes_fast_field_reader(self.field)?,
         })
     }
 
@@ -198,7 +191,7 @@ impl SegmentCollector for BytesFastFieldSegmentCollector {
     type Fruit = Vec<u8>;
 
     fn collect(&mut self, doc: u32, _score: f32) {
-        let data = self.reader.get_bytes(doc);
+        let data = self.reader.get_val(doc);
         self.vals.extend(data);
     }
 

src/collector/top_collector.rs

@@ -98,11 +98,11 @@ where
             .collect())
     }
 
-    pub(crate) fn for_segment<F: PartialOrd>(
+    pub(crate) fn for_segment(
         &self,
         segment_id: SegmentLocalId,
         _: &SegmentReader,
-    ) -> Result<TopSegmentCollector<F>> {
+    ) -> Result<TopSegmentCollector<T>> {
         Ok(TopSegmentCollector::new(segment_id, self.limit))
     }
 }

src/collector/top_field_collector.rs

@@ -5,12 +5,10 @@ use collector::SegmentCollector;
 use fastfield::FastFieldReader;
 use fastfield::FastValue;
 use schema::Field;
-use std::marker::PhantomData;
 use DocAddress;
 use Result;
 use SegmentLocalId;
 use SegmentReader;
-use TantivyError;
 
 /// The Top Field Collector keeps track of the K documents
 /// sorted by a fast field in the index
@@ -108,15 +106,8 @@ impl<T: FastValue + PartialOrd + Send + Sync + 'static> Collector for TopDocsByF
         reader: &SegmentReader,
     ) -> Result<TopFieldSegmentCollector<T>> {
         let collector = self.collector.for_segment(segment_local_id, reader)?;
-        let reader = reader.fast_fields().u64(self.field).ok_or_else(|| {
-            let field_name = reader.schema().get_field_name(self.field);
-            TantivyError::SchemaError(format!("Failed to find fast field reader {:?}", field_name))
-        })?;
-        Ok(TopFieldSegmentCollector {
-            collector,
-            reader,
-            _type: PhantomData,
-        })
+        let reader = reader.fast_field_reader(self.field)?;
+        Ok(TopFieldSegmentCollector { collector, reader })
     }
 
     fn requires_scoring(&self) -> bool {
@@ -131,10 +122,9 @@ impl<T: FastValue + PartialOrd + Send + Sync + 'static> Collector for TopDocsByF
     }
 }
 
-pub struct TopFieldSegmentCollector<T> {
-    collector: TopSegmentCollector<u64>,
-    reader: FastFieldReader<u64>,
-    _type: PhantomData<T>,
+pub struct TopFieldSegmentCollector<T: FastValue + PartialOrd> {
+    collector: TopSegmentCollector<T>,
+    reader: FastFieldReader<T>,
 }
 
 impl<T: FastValue + PartialOrd + Send + Sync + 'static> SegmentCollector
@@ -148,11 +138,7 @@ impl<T: FastValue + PartialOrd + Send + Sync + 'static> SegmentCollector
     }
 
     fn harvest(self) -> Vec<(T, DocAddress)> {
-        self.collector
-            .harvest()
-            .into_iter()
-            .map(|(val, doc_address)| (T::from_u64(val), doc_address))
-            .collect()
+        self.collector.harvest()
     }
 }
 
@@ -249,7 +235,7 @@ mod tests {
                 .for_segment(0, segment)
                 .map(|_| ())
                 .unwrap_err(),
-            TantivyError::SchemaError(_)
+            TantivyError::FastFieldError(_)
         );
     }
 

src/common/mod.rs

@@ -13,10 +13,7 @@ pub use self::serialize::{BinarySerializable, FixedSize};
 pub use self::vint::{read_u32_vint, serialize_vint_u32, write_u32_vint, VInt};
 pub use byteorder::LittleEndian as Endianness;
 
-/// Segment's max doc must be `< MAX_DOC_LIMIT`.
-///
-/// We do not allow segments with more than
-pub const MAX_DOC_LIMIT: u32 = 1 << 31;
+use std::io;
 
 /// Computes the number of bits that will be used for bitpacking.
 ///
@@ -55,6 +52,11 @@ pub(crate) fn is_power_of_2(n: usize) -> bool {
     (n > 0) && (n & (n - 1) == 0)
 }
 
+/// Create a default io error given a string.
+pub(crate) fn make_io_err(msg: String) -> io::Error {
+    io::Error::new(io::ErrorKind::Other, msg)
+}
+
 /// Has length trait
 pub trait HasLen {
     /// Return length
@@ -132,11 +134,4 @@ pub(crate) mod test {
         assert_eq!(compute_num_bits(256), 9u8);
         assert_eq!(compute_num_bits(5_000_000_000), 33u8);
     }
-
-    #[test]
-    fn test_max_doc() {
-        // this is the first time I write a unit test for a constant.
-        assert!(((super::MAX_DOC_LIMIT - 1) as i32) >= 0);
-        assert!((super::MAX_DOC_LIMIT as i32) < 0);
-    }
 }

src/core/index.rs

@@ -24,7 +24,6 @@ use schema::Schema;
 use serde_json;
 use std::borrow::BorrowMut;
 use std::fmt;
-#[cfg(feature = "mmap")]
 use std::path::Path;
 use std::sync::Arc;
 use tokenizer::BoxedTokenizer;
@@ -340,7 +339,7 @@ impl Index {
         Ok(self
             .searchable_segment_metas()?
             .iter()
-            .map(SegmentMeta::id)
+            .map(|segment_meta| segment_meta.id())
             .collect())
     }
 }
@@ -356,8 +355,10 @@ mod tests {
     use directory::RAMDirectory;
     use schema::Field;
     use schema::{Schema, INDEXED, TEXT};
+    use std::path::PathBuf;
     use std::thread;
     use std::time::Duration;
+    use tempdir::TempDir;
     use Index;
     use IndexReader;
     use IndexWriter;
@@ -443,69 +444,61 @@ mod tests {
         test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
     }
 
-    #[cfg(feature = "mmap")]
-    mod mmap_specific {
+    #[test]
+    fn test_index_on_commit_reload_policy_mmap() {
+        let schema = throw_away_schema();
+        let field = schema.get_field("num_likes").unwrap();
+        let tempdir = TempDir::new("index").unwrap();
+        let tempdir_path = PathBuf::from(tempdir.path());
+        let index = Index::create_in_dir(&tempdir_path, schema).unwrap();
+        let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+        writer.commit().unwrap();
+        let reader = index
+            .reader_builder()
+            .reload_policy(ReloadPolicy::OnCommit)
+            .try_into()
+            .unwrap();
+        assert_eq!(reader.searcher().num_docs(), 0);
+        test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
+    }
 
-        use super::*;
-        use std::path::PathBuf;
-        use tempdir::TempDir;
+    #[test]
+    fn test_index_manual_policy_mmap() {
+        let schema = throw_away_schema();
+        let field = schema.get_field("num_likes").unwrap();
+        let index = Index::create_from_tempdir(schema).unwrap();
+        let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+        writer.commit().unwrap();
+        let reader = index
+            .reader_builder()
+            .reload_policy(ReloadPolicy::Manual)
+            .try_into()
+            .unwrap();
+        assert_eq!(reader.searcher().num_docs(), 0);
+        writer.add_document(doc!(field=>1u64));
+        writer.commit().unwrap();
+        thread::sleep(Duration::from_millis(500));
+        assert_eq!(reader.searcher().num_docs(), 0);
+        reader.reload().unwrap();
+        assert_eq!(reader.searcher().num_docs(), 1);
+    }
 
-        #[test]
-        fn test_index_on_commit_reload_policy_mmap() {
-            let schema = throw_away_schema();
-            let field = schema.get_field("num_likes").unwrap();
-            let tempdir = TempDir::new("index").unwrap();
-            let tempdir_path = PathBuf::from(tempdir.path());
-            let index = Index::create_in_dir(&tempdir_path, schema).unwrap();
-            let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            writer.commit().unwrap();
-            let reader = index
-                .reader_builder()
-                .reload_policy(ReloadPolicy::OnCommit)
-                .try_into()
-                .unwrap();
-            assert_eq!(reader.searcher().num_docs(), 0);
-            test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
-        }
-
-        #[test]
-        fn test_index_manual_policy_mmap() {
-            let schema = throw_away_schema();
-            let field = schema.get_field("num_likes").unwrap();
-            let index = Index::create_from_tempdir(schema).unwrap();
-            let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            writer.commit().unwrap();
-            let reader = index
-                .reader_builder()
-                .reload_policy(ReloadPolicy::Manual)
-                .try_into()
-                .unwrap();
-            assert_eq!(reader.searcher().num_docs(), 0);
-            writer.add_document(doc!(field=>1u64));
-            writer.commit().unwrap();
-            thread::sleep(Duration::from_millis(500));
-            assert_eq!(reader.searcher().num_docs(), 0);
-            reader.reload().unwrap();
-            assert_eq!(reader.searcher().num_docs(), 1);
-        }
-
-        #[test]
-        fn test_index_on_commit_reload_policy_different_directories() {
-            let schema = throw_away_schema();
-            let field = schema.get_field("num_likes").unwrap();
-            let tempdir = TempDir::new("index").unwrap();
-            let tempdir_path = PathBuf::from(tempdir.path());
-            let write_index = Index::create_in_dir(&tempdir_path, schema).unwrap();
-            let read_index = Index::open_in_dir(&tempdir_path).unwrap();
-            let reader = read_index
-                .reader_builder()
-                .reload_policy(ReloadPolicy::OnCommit)
-                .try_into()
-                .unwrap();
-            assert_eq!(reader.searcher().num_docs(), 0);
-            let mut writer = write_index.writer_with_num_threads(1, 3_000_000).unwrap();
-            test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
-        }
+    #[test]
+    fn test_index_on_commit_reload_policy_different_directories() {
+        let schema = throw_away_schema();
+        let field = schema.get_field("num_likes").unwrap();
+        let tempdir = TempDir::new("index").unwrap();
+        let tempdir_path = PathBuf::from(tempdir.path());
+        let write_index = Index::create_in_dir(&tempdir_path, schema).unwrap();
+        let read_index = Index::open_in_dir(&tempdir_path).unwrap();
+        let reader = read_index
+            .reader_builder()
+            .reload_policy(ReloadPolicy::OnCommit)
+            .try_into()
+            .unwrap();
+        assert_eq!(reader.searcher().num_docs(), 0);
+        let mut writer = write_index.writer_with_num_threads(1, 3_000_000).unwrap();
+        test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
     }
 
     fn test_index_on_commit_reload_policy_aux(

src/core/searcher.rs

@@ -59,7 +59,7 @@ impl Searcher {
     ) -> Searcher {
         let store_readers = segment_readers
             .iter()
-            .map(SegmentReader::get_store_reader)
+            .map(|segment_reader| segment_reader.get_store_reader())
             .collect();
         Searcher {
             schema,
@@ -218,7 +218,7 @@ impl fmt::Debug for Searcher {
         let segment_ids = self
             .segment_readers
             .iter()
-            .map(SegmentReader::segment_id)
+            .map(|segment_reader| segment_reader.segment_id())
             .collect::<Vec<_>>();
         write!(f, "Searcher({:?})", segment_ids)
     }

src/core/segment_reader.rs

@@ -5,10 +5,14 @@ use core::Segment;
 use core::SegmentComponent;
 use core::SegmentId;
 use directory::ReadOnlySource;
+use error::TantivyError;
 use fastfield::DeleteBitSet;
 use fastfield::FacetReader;
-use fastfield::FastFieldReaders;
+use fastfield::FastFieldReader;
+use fastfield::{self, FastFieldNotAvailableError};
+use fastfield::{BytesFastFieldReader, FastValue, MultiValueIntFastFieldReader};
 use fieldnorm::FieldNormReader;
+use schema::Cardinality;
 use schema::Field;
 use schema::FieldType;
 use schema::Schema;
@@ -47,7 +51,7 @@ pub struct SegmentReader {
     postings_composite: CompositeFile,
     positions_composite: CompositeFile,
     positions_idx_composite: CompositeFile,
-    fast_fields_readers: Arc<FastFieldReaders>,
+    fast_fields_composite: CompositeFile,
     fieldnorms_composite: CompositeFile,
 
     store_source: ReadOnlySource,
@@ -101,21 +105,93 @@ impl SegmentReader {
     ///
     /// # Panics
     /// May panic if the index is corrupted.
-    pub fn fast_fields(&self) -> &FastFieldReaders {
-        &self.fast_fields_readers
+    pub fn fast_field_reader<Item: FastValue>(
+        &self,
+        field: Field,
+    ) -> fastfield::Result<FastFieldReader<Item>> {
+        let field_entry = self.schema.get_field_entry(field);
+        if Item::fast_field_cardinality(field_entry.field_type()) == Some(Cardinality::SingleValue)
+        {
+            self.fast_fields_composite
+                .open_read(field)
+                .ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
+                .map(FastFieldReader::open)
+        } else {
+            Err(FastFieldNotAvailableError::new(field_entry))
+        }
+    }
+
+    pub(crate) fn fast_field_reader_with_idx<Item: FastValue>(
+        &self,
+        field: Field,
+        idx: usize,
+    ) -> fastfield::Result<FastFieldReader<Item>> {
+        if let Some(ff_source) = self.fast_fields_composite.open_read_with_idx(field, idx) {
+            Ok(FastFieldReader::open(ff_source))
+        } else {
+            let field_entry = self.schema.get_field_entry(field);
+            Err(FastFieldNotAvailableError::new(field_entry))
+        }
+    }
+
+    /// Accessor to the `MultiValueIntFastFieldReader` associated to a given `Field`.
+    /// May panick if the field is not a multivalued fastfield of the type `Item`.
+    pub fn multi_fast_field_reader<Item: FastValue>(
+        &self,
+        field: Field,
+    ) -> fastfield::Result<MultiValueIntFastFieldReader<Item>> {
+        let field_entry = self.schema.get_field_entry(field);
+        if Item::fast_field_cardinality(field_entry.field_type()) == Some(Cardinality::MultiValues)
+        {
+            let idx_reader = self.fast_field_reader_with_idx(field, 0)?;
+            let vals_reader = self.fast_field_reader_with_idx(field, 1)?;
+            Ok(MultiValueIntFastFieldReader::open(idx_reader, vals_reader))
+        } else {
+            Err(FastFieldNotAvailableError::new(field_entry))
+        }
+    }
+
+    /// Accessor to the `BytesFastFieldReader` associated to a given `Field`.
+    pub fn bytes_fast_field_reader(&self, field: Field) -> fastfield::Result<BytesFastFieldReader> {
+        let field_entry = self.schema.get_field_entry(field);
+        match *field_entry.field_type() {
+            FieldType::Bytes => {}
+            _ => return Err(FastFieldNotAvailableError::new(field_entry)),
+        }
+        let idx_reader = self
+            .fast_fields_composite
+            .open_read_with_idx(field, 0)
+            .ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
+            .map(FastFieldReader::open)?;
+        let values = self
+            .fast_fields_composite
+            .open_read_with_idx(field, 1)
+            .ok_or_else(|| FastFieldNotAvailableError::new(field_entry))?;
+        Ok(BytesFastFieldReader::open(idx_reader, values))
     }
 
     /// Accessor to the `FacetReader` associated to a given `Field`.
-    pub fn facet_reader(&self, field: Field) -> Option<FacetReader> {
+    pub fn facet_reader(&self, field: Field) -> Result<FacetReader> {
         let field_entry = self.schema.get_field_entry(field);
         if field_entry.field_type() != &FieldType::HierarchicalFacet {
-            return None;
+            return Err(TantivyError::InvalidArgument(format!(
+                "The field {:?} is not a \
+                 hierarchical facet.",
+                field_entry
+            )));
         }
-        let term_ords_reader = self.fast_fields().u64s(field)?;
-        let termdict_source = self.termdict_composite.open_read(field)?;
+        let term_ords_reader = self.multi_fast_field_reader(field)?;
+        let termdict_source = self.termdict_composite.open_read(field).ok_or_else(|| {
+            TantivyError::InvalidArgument(format!(
+                "The field \"{}\" is a hierarchical \
+                 but this segment does not seem to have the field term \
+                 dictionary.",
+                field_entry.name()
+            ))
+        })?;
         let termdict = TermDictionary::from_source(&termdict_source);
         let facet_reader = FacetReader::new(term_ords_reader, termdict);
-        Some(facet_reader)
+        Ok(facet_reader)
     }
 
     /// Accessor to the segment's `Field norms`'s reader.
@@ -171,12 +247,8 @@ impl SegmentReader {
             }
         };
 
-        let schema = segment.schema();
-
         let fast_fields_data = segment.open_read(SegmentComponent::FASTFIELDS)?;
         let fast_fields_composite = CompositeFile::open(&fast_fields_data)?;
-        let fast_field_readers =
-            Arc::new(FastFieldReaders::load_all(&schema, &fast_fields_composite)?);
 
         let fieldnorms_data = segment.open_read(SegmentComponent::FIELDNORMS)?;
         let fieldnorms_composite = CompositeFile::open(&fieldnorms_data)?;
@@ -188,13 +260,14 @@ impl SegmentReader {
             None
         };
 
+        let schema = segment.schema();
         Ok(SegmentReader {
             inv_idx_reader_cache: Arc::new(RwLock::new(HashMap::new())),
             max_doc: segment.meta().max_doc(),
             num_docs: segment.meta().num_docs(),
             termdict_composite,
             postings_composite,
-            fast_fields_readers: fast_field_readers,
+            fast_fields_composite,
             fieldnorms_composite,
             segment_id: segment.id(),
             store_source,
@@ -308,12 +381,12 @@ impl SegmentReader {
             self.postings_composite.space_usage(),
             self.positions_composite.space_usage(),
             self.positions_idx_composite.space_usage(),
-            self.fast_fields_readers.space_usage(),
+            self.fast_fields_composite.space_usage(),
             self.fieldnorms_composite.space_usage(),
             self.get_store_reader().space_usage(),
             self.delete_bitset_opt
                 .as_ref()
-                .map(DeleteBitSet::space_usage)
+                .map(|x| x.space_usage())
                 .unwrap_or(0),
         )
     }

src/directory/managed_directory.rs

@@ -260,98 +260,95 @@ impl Clone for ManagedDirectory {
 #[cfg(test)]
 mod tests {
 
+    use super::*;
     #[cfg(feature = "mmap")]
-    mod mmap_specific {
+    use directory::MmapDirectory;
+    use std::io::Write;
+    use std::path::Path;
+    use tempdir::TempDir;
 
-        use super::super::*;
-        use std::path::Path;
-        use tempdir::TempDir;
-
-        lazy_static! {
-            static ref TEST_PATH1: &'static Path = Path::new("some_path_for_test");
-            static ref TEST_PATH2: &'static Path = Path::new("some_path_for_test2");
-        }
-
-        use directory::MmapDirectory;
-        use std::io::Write;
-
-        #[test]
-        fn test_managed_directory() {
-            let tempdir = TempDir::new("index").unwrap();
-            let tempdir_path = PathBuf::from(tempdir.path());
-            {
-                let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
-                let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
-                {
-                    let mut write_file = managed_directory.open_write(*TEST_PATH1).unwrap();
-                    write_file.flush().unwrap();
-                }
-                {
-                    managed_directory
-                        .atomic_write(*TEST_PATH2, &vec![0u8, 1u8])
-                        .unwrap();
-                }
-                {
-                    assert!(managed_directory.exists(*TEST_PATH1));
-                    assert!(managed_directory.exists(*TEST_PATH2));
-                }
-                {
-                    let living_files: HashSet<PathBuf> =
-                        [TEST_PATH1.to_owned()].into_iter().cloned().collect();
-                    managed_directory.garbage_collect(|| living_files);
-                }
-                {
-                    assert!(managed_directory.exists(*TEST_PATH1));
-                    assert!(!managed_directory.exists(*TEST_PATH2));
-                }
-            }
-            {
-                let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
-                let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
-                {
-                    assert!(managed_directory.exists(*TEST_PATH1));
-                    assert!(!managed_directory.exists(*TEST_PATH2));
-                }
-                {
-                    let living_files: HashSet<PathBuf> = HashSet::new();
-                    managed_directory.garbage_collect(|| living_files);
-                }
-                {
-                    assert!(!managed_directory.exists(*TEST_PATH1));
-                    assert!(!managed_directory.exists(*TEST_PATH2));
-                }
-            }
-        }
-
-        #[test]
-        fn test_managed_directory_gc_while_mmapped() {
-            let tempdir = TempDir::new("index").unwrap();
-            let tempdir_path = PathBuf::from(tempdir.path());
-            let living_files = HashSet::new();
+    lazy_static! {
+        static ref TEST_PATH1: &'static Path = Path::new("some_path_for_test");
+        static ref TEST_PATH2: &'static Path = Path::new("some_path_for_test2");
+    }
 
+    #[test]
+    #[cfg(feature = "mmap")]
+    fn test_managed_directory() {
+        let tempdir = TempDir::new("index").unwrap();
+        let tempdir_path = PathBuf::from(tempdir.path());
+        {
             let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
             let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
-            managed_directory
-                .atomic_write(*TEST_PATH1, &vec![0u8, 1u8])
-                .unwrap();
-            assert!(managed_directory.exists(*TEST_PATH1));
-
-            let _mmap_read = managed_directory.open_read(*TEST_PATH1).unwrap();
-            managed_directory.garbage_collect(|| living_files.clone());
-            if cfg!(target_os = "windows") {
-                // On Windows, gc should try and fail the file as it is mmapped.
+            {
+                let mut write_file = managed_directory.open_write(*TEST_PATH1).unwrap();
+                write_file.flush().unwrap();
+            }
+            {
+                managed_directory
+                    .atomic_write(*TEST_PATH2, &vec![0u8, 1u8])
+                    .unwrap();
+            }
+            {
                 assert!(managed_directory.exists(*TEST_PATH1));
-                // unmap should happen here.
-                drop(_mmap_read);
-                // The file should still be in the list of managed file and
-                // eventually be deleted once mmap is released.
+                assert!(managed_directory.exists(*TEST_PATH2));
+            }
+            {
+                let living_files: HashSet<PathBuf> =
+                    [TEST_PATH1.to_owned()].into_iter().cloned().collect();
                 managed_directory.garbage_collect(|| living_files);
-                assert!(!managed_directory.exists(*TEST_PATH1));
-            } else {
-                assert!(!managed_directory.exists(*TEST_PATH1));
+            }
+            {
+                assert!(managed_directory.exists(*TEST_PATH1));
+                assert!(!managed_directory.exists(*TEST_PATH2));
             }
         }
+        {
+            let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
+            let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
+            {
+                assert!(managed_directory.exists(*TEST_PATH1));
+                assert!(!managed_directory.exists(*TEST_PATH2));
+            }
+            {
+                let living_files: HashSet<PathBuf> = HashSet::new();
+                managed_directory.garbage_collect(|| living_files);
+            }
+            {
+                assert!(!managed_directory.exists(*TEST_PATH1));
+                assert!(!managed_directory.exists(*TEST_PATH2));
+            }
+        }
+    }
 
+    #[test]
+    #[cfg(feature = "mmap ")]
+    fn test_managed_directory_gc_while_mmapped() {
+        let tempdir = TempDir::new("index").unwrap();
+        let tempdir_path = PathBuf::from(tempdir.path());
+        let living_files = HashSet::new();
+
+        let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
+        let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
+        managed_directory
+            .atomic_write(*TEST_PATH1, &vec![0u8, 1u8])
+            .unwrap();
+        assert!(managed_directory.exists(*TEST_PATH1));
+
+        let _mmap_read = managed_directory.open_read(*TEST_PATH1).unwrap();
+        managed_directory.garbage_collect(|| living_files.clone());
+        if cfg!(target_os = "windows") {
+            // On Windows, gc should try and fail the file as it is mmapped.
+            assert!(managed_directory.exists(*TEST_PATH1));
+            // unmap should happen here.
+            drop(_mmap_read);
+            // The file should still be in the list of managed file and
+            // eventually be deleted once mmap is released.
+            managed_directory.garbage_collect(|| living_files);
+            assert!(!managed_directory.exists(*TEST_PATH1));
+        } else {
+            assert!(!managed_directory.exists(*TEST_PATH1));
+        }
     }
 
 }

src/directory/mmap_directory.rs

@@ -6,6 +6,7 @@ use self::notify::RawEvent;
 use self::notify::RecursiveMode;
 use self::notify::Watcher;
 use atomicwrites;
+use common::make_io_err;
 use core::META_FILEPATH;
 use directory::error::LockError;
 use directory::error::{DeleteError, IOError, OpenDirectoryError, OpenReadError, OpenWriteError};
@@ -36,11 +37,6 @@ use std::sync::Weak;
 use std::thread;
 use tempdir::TempDir;
 
-/// Create a default io error given a string.
-pub(crate) fn make_io_err(msg: String) -> io::Error {
-    io::Error::new(io::ErrorKind::Other, msg)
-}
-
 /// Returns None iff the file exists, can be read, but is empty (and hence
 /// cannot be mmapped)
 fn open_mmap(full_path: &Path) -> result::Result<Option<Mmap>, OpenReadError> {

src/directory/ram_directory.rs

@@ -86,7 +86,7 @@ impl InnerDirectory {
         self.fs
             .get(path)
             .ok_or_else(|| OpenReadError::FileDoesNotExist(PathBuf::from(path)))
-            .map(Clone::clone)
+            .map(|el| el.clone())
     }
 
     fn delete(&mut self, path: &Path) -> result::Result<(), DeleteError> {

src/docset.rs

@@ -1,5 +1,4 @@
 use common::BitSet;
-use fastfield::DeleteBitSet;
 use std::borrow::Borrow;
 use std::borrow::BorrowMut;
 use std::cmp::Ordering;
@@ -96,23 +95,9 @@ pub trait DocSet {
     }
 
     /// Returns the number documents matching.
-    /// Calling this method consumes the `DocSet`.
-    fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
-        let mut count = 0u32;
-        while self.advance() {
-            if !delete_bitset.is_deleted(self.doc()) {
-                count += 1u32;
-            }
-        }
-        count
-    }
-
-    /// Returns the count of documents, deleted or not.
-    /// Calling this method consumes the `DocSet`.
     ///
-    /// Of course, the result is an upper bound of the result
-    /// given by `count()`.
-    fn count_including_deleted(&mut self) -> u32 {
+    /// Calling this method consumes the `DocSet`.
+    fn count(&mut self) -> u32 {
         let mut count = 0u32;
         while self.advance() {
             count += 1u32;
@@ -142,18 +127,13 @@ impl<TDocSet: DocSet + ?Sized> DocSet for Box<TDocSet> {
         unboxed.size_hint()
     }
 
+    fn count(&mut self) -> u32 {
+        let unboxed: &mut TDocSet = self.borrow_mut();
+        unboxed.count()
+    }
+
     fn append_to_bitset(&mut self, bitset: &mut BitSet) {
         let unboxed: &mut TDocSet = self.borrow_mut();
         unboxed.append_to_bitset(bitset);
     }
-
-    fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
-        let unboxed: &mut TDocSet = self.borrow_mut();
-        unboxed.count(delete_bitset)
-    }
-
-    fn count_including_deleted(&mut self) -> u32 {
-        let unboxed: &mut TDocSet = self.borrow_mut();
-        unboxed.count_including_deleted()
-    }
 }

src/fastfield/bytes/mod.rs

@@ -23,14 +23,14 @@ mod tests {
         index_writer.add_document(doc!(field=>vec![0u8; 1000]));
         assert!(index_writer.commit().is_ok());
         let searcher = index.reader().unwrap().searcher();
-        let segment_reader = searcher.segment_reader(0);
-        let bytes_reader = segment_reader.fast_fields().bytes(field).unwrap();
+        let reader = searcher.segment_reader(0);
+        let bytes_reader = reader.bytes_fast_field_reader(field).unwrap();
 
-        assert_eq!(bytes_reader.get_bytes(0), &[0u8, 1, 2, 3]);
-        assert!(bytes_reader.get_bytes(1).is_empty());
-        assert_eq!(bytes_reader.get_bytes(2), &[255u8]);
-        assert_eq!(bytes_reader.get_bytes(3), &[1u8, 3, 5, 7, 9]);
+        assert_eq!(bytes_reader.get_val(0), &[0u8, 1, 2, 3]);
+        assert!(bytes_reader.get_val(1).is_empty());
+        assert_eq!(bytes_reader.get_val(2), &[255u8]);
+        assert_eq!(bytes_reader.get_val(3), &[1u8, 3, 5, 7, 9]);
         let long = vec![0u8; 1000];
-        assert_eq!(bytes_reader.get_bytes(4), long.as_slice());
+        assert_eq!(bytes_reader.get_val(4), long.as_slice());
     }
 }

src/fastfield/bytes/reader.rs

@@ -14,7 +14,6 @@ use DocId;
 ///
 /// Reading the value for a document is done by reading the start index for it,
 /// and the start index for the next document, and keeping the bytes in between.
-#[derive(Clone)]
 pub struct BytesFastFieldReader {
     idx_reader: FastFieldReader<u64>,
     values: OwningRef<ReadOnlySource, [u8]>,
@@ -29,20 +28,10 @@ impl BytesFastFieldReader {
         BytesFastFieldReader { idx_reader, values }
     }
 
-    fn range(&self, doc: DocId) -> (usize, usize) {
+    /// Returns the bytes associated to the given `doc`
+    pub fn get_val(&self, doc: DocId) -> &[u8] {
         let start = self.idx_reader.get(doc) as usize;
         let stop = self.idx_reader.get(doc + 1) as usize;
-        (start, stop)
-    }
-
-    /// Returns the bytes associated to the given `doc`
-    pub fn get_bytes(&self, doc: DocId) -> &[u8] {
-        let (start, stop) = self.range(doc);
         &self.values[start..stop]
     }
-
-    /// Returns the overall number of bytes in this bytes fast field.
-    pub fn total_num_bytes(&self) -> usize {
-        self.values.len()
-    }
 }

src/fastfield/delete.rs

@@ -53,18 +53,16 @@ impl DeleteBitSet {
         }
     }
 
-    /// Returns true iff the document is still "alive". In other words, if it has not been deleted.
-    pub fn is_alive(&self, doc: DocId) -> bool {
-        !self.is_deleted(doc)
-    }
-
-    /// Returns true iff the document has been marked as deleted.
-    #[inline(always)]
+    /// Returns whether the document has been marked as deleted.
     pub fn is_deleted(&self, doc: DocId) -> bool {
-        let byte_offset = doc / 8u32;
-        let b: u8 = (*self.data)[byte_offset as usize];
-        let shift = (doc & 7u32) as u8;
-        b & (1u8 << shift) != 0
+        if self.len == 0 {
+            false
+        } else {
+            let byte_offset = doc / 8u32;
+            let b: u8 = (*self.data)[byte_offset as usize];
+            let shift = (doc & 7u32) as u8;
+            b & (1u8 << shift) != 0
+        }
     }
 
     /// Summarize total space usage of this bitset.

src/fastfield/mod.rs

@@ -30,7 +30,6 @@ pub use self::error::{FastFieldNotAvailableError, Result};
 pub use self::facet_reader::FacetReader;
 pub use self::multivalued::{MultiValueIntFastFieldReader, MultiValueIntFastFieldWriter};
 pub use self::reader::FastFieldReader;
-pub use self::readers::FastFieldReaders;
 pub use self::serializer::FastFieldSerializer;
 pub use self::writer::{FastFieldsWriter, IntFastFieldWriter};
 use common;
@@ -44,7 +43,6 @@ mod error;
 mod facet_reader;
 mod multivalued;
 mod reader;
-mod readers;
 mod serializer;
 mod writer;
 
@@ -80,6 +78,10 @@ impl FastValue for u64 {
         *self
     }
 
+    fn as_u64(&self) -> u64 {
+        *self
+    }
+
     fn fast_field_cardinality(field_type: &FieldType) -> Option<Cardinality> {
         match *field_type {
             FieldType::U64(ref integer_options) => integer_options.get_fastfield_cardinality(),
@@ -87,10 +89,6 @@ impl FastValue for u64 {
             _ => None,
         }
     }
-
-    fn as_u64(&self) -> u64 {
-        *self
-    }
 }
 
 impl FastValue for i64 {

src/fastfield/multivalued/mod.rs

@@ -37,7 +37,9 @@ mod tests {
         let searcher = index.reader().unwrap().searcher();
         let segment_reader = searcher.segment_reader(0);
         let mut vals = Vec::new();
-        let multi_value_reader = segment_reader.fast_fields().u64s(field).unwrap();
+        let multi_value_reader = segment_reader
+            .multi_fast_field_reader::<u64>(field)
+            .unwrap();
         {
             multi_value_reader.get_vals(2, &mut vals);
             assert_eq!(&vals, &[4u64]);
@@ -196,9 +198,9 @@ mod tests {
         assert!(index_writer.commit().is_ok());
 
         let searcher = index.reader().unwrap().searcher();
-        let segment_reader = searcher.segment_reader(0);
+        let reader = searcher.segment_reader(0);
         let mut vals = Vec::new();
-        let multi_value_reader = segment_reader.fast_fields().i64s(field).unwrap();
+        let multi_value_reader = reader.multi_fast_field_reader::<i64>(field).unwrap();
         {
             multi_value_reader.get_vals(2, &mut vals);
             assert_eq!(&vals, &[-4i64]);

src/fastfield/multivalued/reader.rs

@@ -26,13 +26,6 @@ impl<Item: FastValue> MultiValueIntFastFieldReader<Item> {
         }
     }
 
-    pub(crate) fn into_u64s_reader(self) -> MultiValueIntFastFieldReader<u64> {
-        MultiValueIntFastFieldReader {
-            idx_reader: self.idx_reader,
-            vals_reader: self.vals_reader.into_u64_reader(),
-        }
-    }
-
     /// Returns `(start, stop)`, such that the values associated
     /// to the given document are `start..stop`.
     fn range(&self, doc: DocId) -> (u64, u64) {
@@ -48,24 +41,13 @@ impl<Item: FastValue> MultiValueIntFastFieldReader<Item> {
         vals.resize(len, Item::default());
         self.vals_reader.get_range_u64(start, &mut vals[..]);
     }
-
-    /// Returns the number of values associated with the document `DocId`.
-    pub fn num_vals(&self, doc: DocId) -> usize {
-        let (start, stop) = self.range(doc);
-        (stop - start) as usize
-    }
-
-    /// Returns the overall number of values in this field  .
-    pub fn total_num_vals(&self) -> u64 {
-        self.idx_reader.max_value()
-    }
 }
 
 #[cfg(test)]
 mod tests {
 
     use core::Index;
-    use schema::{Facet, Schema};
+    use schema::{Document, Facet, Schema};
 
     #[test]
     fn test_multifastfield_reader() {
@@ -76,12 +58,22 @@ mod tests {
         let mut index_writer = index
             .writer_with_num_threads(1, 30_000_000)
             .expect("Failed to create index writer.");
-        index_writer.add_document(doc!(
-            facet_field => Facet::from("/category/cat2"),
-            facet_field => Facet::from("/category/cat1"),
-        ));
-        index_writer.add_document(doc!(facet_field => Facet::from("/category/cat2")));
-        index_writer.add_document(doc!(facet_field => Facet::from("/category/cat3")));
+        {
+            let mut doc = Document::new();
+            doc.add_facet(facet_field, "/category/cat2");
+            doc.add_facet(facet_field, "/category/cat1");
+            index_writer.add_document(doc);
+        }
+        {
+            let mut doc = Document::new();
+            doc.add_facet(facet_field, "/category/cat2");
+            index_writer.add_document(doc);
+        }
+        {
+            let mut doc = Document::new();
+            doc.add_facet(facet_field, "/category/cat3");
+            index_writer.add_document(doc);
+        }
         index_writer.commit().expect("Commit failed");
         let searcher = index.reader().unwrap().searcher();
         let segment_reader = searcher.segment_reader(0);

src/fastfield/reader.rs

@@ -50,15 +50,6 @@ impl<Item: FastValue> FastFieldReader<Item> {
         }
     }
 
-    pub(crate) fn into_u64_reader(self) -> FastFieldReader<u64> {
-        FastFieldReader {
-            bit_unpacker: self.bit_unpacker,
-            min_value_u64: self.min_value_u64,
-            max_value_u64: self.max_value_u64,
-            _phantom: PhantomData,
-        }
-    }
-
     /// Return the value associated to the given document.
     ///
     /// This accessor should return as fast as possible.

src/fastfield/readers.rs

@@ -1,191 +0,0 @@
-use common::CompositeFile;
-use fastfield::BytesFastFieldReader;
-use fastfield::MultiValueIntFastFieldReader;
-use fastfield::{FastFieldNotAvailableError, FastFieldReader};
-use schema::{Cardinality, Field, FieldType, Schema};
-use space_usage::PerFieldSpaceUsage;
-use std::collections::HashMap;
-use Result;
-
-/// Provides access to all of the FastFieldReader.
-///
-/// Internally, `FastFieldReaders` have preloaded fast field readers,
-/// and just wraps several `HashMap`.
-pub struct FastFieldReaders {
-    fast_field_i64: HashMap<Field, FastFieldReader<i64>>,
-    fast_field_u64: HashMap<Field, FastFieldReader<u64>>,
-    fast_field_i64s: HashMap<Field, MultiValueIntFastFieldReader<i64>>,
-    fast_field_u64s: HashMap<Field, MultiValueIntFastFieldReader<u64>>,
-    fast_bytes: HashMap<Field, BytesFastFieldReader>,
-    fast_fields_composite: CompositeFile,
-}
-
-enum FastType {
-    I64,
-    U64,
-}
-
-fn type_and_cardinality(field_type: &FieldType) -> Option<(FastType, Cardinality)> {
-    match field_type {
-        FieldType::U64(options) => options
-            .get_fastfield_cardinality()
-            .map(|cardinality| (FastType::U64, cardinality)),
-        FieldType::I64(options) => options
-            .get_fastfield_cardinality()
-            .map(|cardinality| (FastType::I64, cardinality)),
-        FieldType::HierarchicalFacet => Some((FastType::U64, Cardinality::MultiValues)),
-        _ => None,
-    }
-}
-
-impl FastFieldReaders {
-    pub(crate) fn load_all(
-        schema: &Schema,
-        fast_fields_composite: &CompositeFile,
-    ) -> Result<FastFieldReaders> {
-        let mut fast_field_readers = FastFieldReaders {
-            fast_field_i64: Default::default(),
-            fast_field_u64: Default::default(),
-            fast_field_i64s: Default::default(),
-            fast_field_u64s: Default::default(),
-            fast_bytes: Default::default(),
-            fast_fields_composite: fast_fields_composite.clone(),
-        };
-        for (field_id, field_entry) in schema.fields().iter().enumerate() {
-            let field = Field(field_id as u32);
-            let field_type = field_entry.field_type();
-            if field_type == &FieldType::Bytes {
-                let idx_reader = fast_fields_composite
-                    .open_read_with_idx(field, 0)
-                    .ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
-                    .map(FastFieldReader::open)?;
-                let data = fast_fields_composite
-                    .open_read_with_idx(field, 1)
-                    .ok_or_else(|| FastFieldNotAvailableError::new(field_entry))?;
-                fast_field_readers
-                    .fast_bytes
-                    .insert(field, BytesFastFieldReader::open(idx_reader, data));
-            } else if let Some((fast_type, cardinality)) = type_and_cardinality(field_type) {
-                match cardinality {
-                    Cardinality::SingleValue => {
-                        if let Some(fast_field_data) = fast_fields_composite.open_read(field) {
-                            match fast_type {
-                                FastType::U64 => {
-                                    let fast_field_reader = FastFieldReader::open(fast_field_data);
-                                    fast_field_readers
-                                        .fast_field_u64
-                                        .insert(field, fast_field_reader);
-                                }
-                                FastType::I64 => {
-                                    fast_field_readers.fast_field_i64.insert(
-                                        field,
-                                        FastFieldReader::open(fast_field_data.clone()),
-                                    );
-                                }
-                            }
-                        } else {
-                            return Err(From::from(FastFieldNotAvailableError::new(field_entry)));
-                        }
-                    }
-                    Cardinality::MultiValues => {
-                        let idx_opt = fast_fields_composite.open_read_with_idx(field, 0);
-                        let data_opt = fast_fields_composite.open_read_with_idx(field, 1);
-                        if let (Some(fast_field_idx), Some(fast_field_data)) = (idx_opt, data_opt) {
-                            let idx_reader = FastFieldReader::open(fast_field_idx);
-                            match fast_type {
-                                FastType::I64 => {
-                                    let vals_reader = FastFieldReader::open(fast_field_data);
-                                    let multivalued_int_fast_field =
-                                        MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
-                                    fast_field_readers
-                                        .fast_field_i64s
-                                        .insert(field, multivalued_int_fast_field);
-                                }
-                                FastType::U64 => {
-                                    let vals_reader = FastFieldReader::open(fast_field_data);
-                                    let multivalued_int_fast_field =
-                                        MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
-                                    fast_field_readers
-                                        .fast_field_u64s
-                                        .insert(field, multivalued_int_fast_field);
-                                }
-                            }
-                        } else {
-                            return Err(From::from(FastFieldNotAvailableError::new(field_entry)));
-                        }
-                    }
-                }
-            }
-        }
-        Ok(fast_field_readers)
-    }
-
-    pub(crate) fn space_usage(&self) -> PerFieldSpaceUsage {
-        self.fast_fields_composite.space_usage()
-    }
-
-    /// Returns the `u64` fast field reader reader associated to `field`.
-    ///
-    /// If `field` is not a u64 fast field, this method returns `None`.
-    pub fn u64(&self, field: Field) -> Option<FastFieldReader<u64>> {
-        self.fast_field_u64.get(&field).cloned()
-    }
-
-    /// If the field is a u64-fast field return the associated reader.
-    /// If the field is a i64-fast field, return the associated u64 reader. Values are
-    /// mapped from i64 to u64 using a (well the, it is unique) monotonic mapping.    ///
-    ///
-    /// This method is useful when merging segment reader.
-    pub(crate) fn u64_lenient(&self, field: Field) -> Option<FastFieldReader<u64>> {
-        if let Some(u64_ff_reader) = self.u64(field) {
-            return Some(u64_ff_reader);
-        }
-        if let Some(i64_ff_reader) = self.i64(field) {
-            return Some(i64_ff_reader.into_u64_reader());
-        }
-        None
-    }
-
-    /// Returns the `i64` fast field reader reader associated to `field`.
-    ///
-    /// If `field` is not a i64 fast field, this method returns `None`.
-    pub fn i64(&self, field: Field) -> Option<FastFieldReader<i64>> {
-        self.fast_field_i64.get(&field).cloned()
-    }
-
-    /// Returns a `u64s` multi-valued fast field reader reader associated to `field`.
-    ///
-    /// If `field` is not a u64 multi-valued fast field, this method returns `None`.
-    pub fn u64s(&self, field: Field) -> Option<MultiValueIntFastFieldReader<u64>> {
-        self.fast_field_u64s.get(&field).cloned()
-    }
-
-    /// If the field is a u64s-fast field return the associated reader.
-    /// If the field is a i64s-fast field, return the associated u64s reader. Values are
-    /// mapped from i64 to u64 using a (well the, it is unique) monotonic mapping.
-    ///
-    /// This method is useful when merging segment reader.
-    pub(crate) fn u64s_lenient(&self, field: Field) -> Option<MultiValueIntFastFieldReader<u64>> {
-        if let Some(u64s_ff_reader) = self.u64s(field) {
-            return Some(u64s_ff_reader);
-        }
-        if let Some(i64s_ff_reader) = self.i64s(field) {
-            return Some(i64s_ff_reader.into_u64s_reader());
-        }
-        None
-    }
-
-    /// Returns a `i64s` multi-valued fast field reader reader associated to `field`.
-    ///
-    /// If `field` is not a i64 multi-valued fast field, this method returns `None`.
-    pub fn i64s(&self, field: Field) -> Option<MultiValueIntFastFieldReader<i64>> {
-        self.fast_field_i64s.get(&field).cloned()
-    }
-
-    /// Returns the `bytes` fast field reader associated to `field`.
-    ///
-    /// If `field` is not a bytes fast field, returns `None`.
-    pub fn bytes(&self, field: Field) -> Option<BytesFastFieldReader> {
-        self.fast_bytes.get(&field).cloned()
-    }
-}

src/functional_test.rs

@@ -13,6 +13,7 @@ fn check_index_content(searcher: &Searcher, vals: &HashSet<u64>) {
 
 #[test]
 #[ignore]
+#[cfg(feature = "mmap")]
 fn test_indexing() {
     let mut schema_builder = Schema::builder();
 

src/indexer/log_merge_policy.rs

@@ -52,7 +52,7 @@ impl MergePolicy for LogMergePolicy {
 
         let mut size_sorted_tuples = segments
             .iter()
-            .map(SegmentMeta::num_docs)
+            .map(|x| x.num_docs())
             .enumerate()
             .collect::<Vec<(usize, u32)>>();
 

src/indexer/merger.rs

@@ -1,9 +1,7 @@
-use common::MAX_DOC_LIMIT;
 use core::Segment;
 use core::SegmentReader;
 use core::SerializableSegment;
 use docset::DocSet;
-use fastfield::BytesFastFieldReader;
 use fastfield::DeleteBitSet;
 use fastfield::FastFieldReader;
 use fastfield::FastFieldSerializer;
@@ -25,7 +23,6 @@ use termdict::TermMerger;
 use termdict::TermOrdinal;
 use DocId;
 use Result;
-use TantivyError;
 
 fn compute_total_num_tokens(readers: &[SegmentReader], field: Field) -> u64 {
     let mut total_tokens = 0u64;
@@ -73,7 +70,7 @@ fn compute_min_max_val(
                 // some deleted documents,
                 // we need to recompute the max / min
                 (0..max_doc)
-                    .filter(|doc_id| delete_bitset.is_alive(*doc_id))
+                    .filter(|doc_id| !delete_bitset.is_deleted(*doc_id))
                     .map(|doc_id| u64_reader.get(doc_id))
                     .minmax()
                     .into_option()
@@ -153,14 +150,6 @@ impl IndexMerger {
                 readers.push(reader);
             }
         }
-        if max_doc >= MAX_DOC_LIMIT {
-            let err_msg = format!(
-                "The segment resulting from this merge would have {} docs,\
-                 which exceeds the limit {}.",
-                max_doc, MAX_DOC_LIMIT
-            );
-            return Err(TantivyError::InvalidArgument(err_msg));
-        }
         Ok(IndexMerger {
             schema,
             readers,
@@ -240,10 +229,7 @@ impl IndexMerger {
         let mut max_value = u64::min_value();
 
         for reader in &self.readers {
-            let u64_reader: FastFieldReader<u64> = reader
-                .fast_fields()
-                .u64_lenient(field)
-                .expect("Failed to find a reader for single fast field. This is a tantivy bug and it should never happen.");
+            let u64_reader: FastFieldReader<u64> = reader.fast_field_reader(field)?;
             if let Some((seg_min_val, seg_max_val)) =
                 compute_min_max_val(&u64_reader, reader.max_doc(), reader.delete_bitset())
             {
@@ -286,28 +272,24 @@ impl IndexMerger {
         fast_field_serializer: &mut FastFieldSerializer,
     ) -> Result<()> {
         let mut total_num_vals = 0u64;
-        let mut u64s_readers: Vec<MultiValueIntFastFieldReader<u64>> = Vec::new();
 
         // In the first pass, we compute the total number of vals.
         //
         // This is required by the bitpacker, as it needs to know
         // what should be the bit length use for bitpacking.
         for reader in &self.readers {
-            let u64s_reader = reader.fast_fields()
-                .u64s_lenient(field)
-                .expect("Failed to find index for multivalued field. This is a bug in tantivy, please report.");
-
+            let idx_reader = reader.fast_field_reader_with_idx::<u64>(field, 0)?;
             if let Some(delete_bitset) = reader.delete_bitset() {
                 for doc in 0u32..reader.max_doc() {
-                    if delete_bitset.is_alive(doc) {
-                        let num_vals = u64s_reader.num_vals(doc) as u64;
-                        total_num_vals += num_vals;
+                    if !delete_bitset.is_deleted(doc) {
+                        let start = idx_reader.get(doc);
+                        let end = idx_reader.get(doc + 1);
+                        total_num_vals += end - start;
                     }
                 }
             } else {
-                total_num_vals += u64s_reader.total_num_vals();
+                total_num_vals += idx_reader.max_value();
             }
-            u64s_readers.push(u64s_reader);
         }
 
         // We can now create our `idx` serializer, and in a second pass,
@@ -315,10 +297,13 @@ impl IndexMerger {
         let mut serialize_idx =
             fast_field_serializer.new_u64_fast_field_with_idx(field, 0, total_num_vals, 0)?;
         let mut idx = 0;
-        for (segment_reader, u64s_reader) in self.readers.iter().zip(&u64s_readers) {
-            for doc in segment_reader.doc_ids_alive() {
+        for reader in &self.readers {
+            let idx_reader = reader.fast_field_reader_with_idx::<u64>(field, 0)?;
+            for doc in reader.doc_ids_alive() {
                 serialize_idx.add_val(idx)?;
-                idx += u64s_reader.num_vals(doc) as u64;
+                let start = idx_reader.get(doc);
+                let end = idx_reader.get(doc + 1);
+                idx += end - start;
             }
         }
         serialize_idx.add_val(idx)?;
@@ -349,10 +334,8 @@ impl IndexMerger {
             for (segment_ord, segment_reader) in self.readers.iter().enumerate() {
                 let term_ordinal_mapping: &[TermOrdinal] =
                     term_ordinal_mappings.get_segment(segment_ord);
-                let ff_reader: MultiValueIntFastFieldReader<u64> = segment_reader
-                    .fast_fields()
-                    .u64s(field)
-                    .expect("Could not find multivalued u64 fast value reader.");
+                let ff_reader: MultiValueIntFastFieldReader<u64> =
+                    segment_reader.multi_fast_field_reader(field)?;
                 // TODO optimize if no deletes
                 for doc in segment_reader.doc_ids_alive() {
                     ff_reader.get_vals(doc, &mut vals);
@@ -384,8 +367,6 @@ impl IndexMerger {
 
         let mut vals = Vec::with_capacity(100);
 
-        let mut ff_readers = Vec::new();
-
         // Our values are bitpacked and we need to know what should be
         // our bitwidth and our minimum value before serializing any values.
         //
@@ -394,10 +375,7 @@ impl IndexMerger {
         // maximum value and initialize our Serializer.
         for reader in &self.readers {
             let ff_reader: MultiValueIntFastFieldReader<u64> =
-                reader.fast_fields().u64s_lenient(field).expect(
-                    "Failed to find multivalued fast field reader. This is a bug in \
-                     tantivy. Please report.",
-                );
+                reader.multi_fast_field_reader(field)?;
             for doc in reader.doc_ids_alive() {
                 ff_reader.get_vals(doc, &mut vals);
                 for &val in &vals {
@@ -405,7 +383,6 @@ impl IndexMerger {
                     max_value = cmp::max(val, max_value);
                 }
             }
-            ff_readers.push(ff_reader);
             // TODO optimize when no deletes
         }
 
@@ -418,7 +395,9 @@ impl IndexMerger {
         {
             let mut serialize_vals = fast_field_serializer
                 .new_u64_fast_field_with_idx(field, min_value, max_value, 1)?;
-            for (reader, ff_reader) in self.readers.iter().zip(ff_readers) {
+            for reader in &self.readers {
+                let ff_reader: MultiValueIntFastFieldReader<u64> =
+                    reader.multi_fast_field_reader(field)?;
                 // TODO optimize if no deletes
                 for doc in reader.doc_ids_alive() {
                     ff_reader.get_vals(doc, &mut vals);
@@ -437,53 +416,19 @@ impl IndexMerger {
         field: Field,
         fast_field_serializer: &mut FastFieldSerializer,
     ) -> Result<()> {
-        let mut total_num_vals = 0u64;
-        let mut bytes_readers: Vec<BytesFastFieldReader> = Vec::new();
-
-        for reader in &self.readers {
-            let bytes_reader = reader.fast_fields().bytes(field).expect(
-                "Failed to find bytes fast field reader. This is a bug in tantivy, please report.",
-            );
-            if let Some(delete_bitset) = reader.delete_bitset() {
-                for doc in 0u32..reader.max_doc() {
-                    if delete_bitset.is_alive(doc) {
-                        let num_vals = bytes_reader.get_bytes(doc).len() as u64;
-                        total_num_vals += num_vals;
-                    }
-                }
-            } else {
-                total_num_vals += bytes_reader.total_num_bytes() as u64;
-            }
-            bytes_readers.push(bytes_reader);
-        }
-
-        {
-            // We can now create our `idx` serializer, and in a second pass,
-            // can effectively push the different indexes.
-            let mut serialize_idx =
-                fast_field_serializer.new_u64_fast_field_with_idx(field, 0, total_num_vals, 0)?;
-            let mut idx = 0;
-            for (segment_reader, bytes_reader) in self.readers.iter().zip(&bytes_readers) {
-                for doc in segment_reader.doc_ids_alive() {
-                    serialize_idx.add_val(idx)?;
-                    idx += bytes_reader.get_bytes(doc).len() as u64;
-                }
-            }
-            serialize_idx.add_val(idx)?;
-            serialize_idx.close_field()?;
-        }
+        self.write_fast_field_idx(field, fast_field_serializer)?;
 
         let mut serialize_vals = fast_field_serializer.new_bytes_fast_field_with_idx(field, 1)?;
-        for segment_reader in &self.readers {
-            let bytes_reader = segment_reader.fast_fields().bytes(field)
-                .expect("Failed to find bytes field in fast field reader. This is a bug in tantivy. Please report.");
+        for reader in &self.readers {
+            let bytes_reader = reader.bytes_fast_field_reader(field)?;
             // TODO: optimize if no deletes
-            for doc in segment_reader.doc_ids_alive() {
-                let val = bytes_reader.get_bytes(doc);
+            for doc in reader.doc_ids_alive() {
+                let val = bytes_reader.get_val(doc);
                 serialize_vals.write_all(val)?;
             }
         }
         serialize_vals.flush()?;
+
         Ok(())
     }
 
@@ -1024,16 +969,14 @@ mod tests {
 
             let score_field_reader = searcher
                 .segment_reader(0)
-                .fast_fields()
-                .u64(score_field)
+                .fast_field_reader::<u64>(score_field)
                 .unwrap();
             assert_eq!(score_field_reader.min_value(), 4000);
             assert_eq!(score_field_reader.max_value(), 7000);
 
             let score_field_reader = searcher
                 .segment_reader(1)
-                .fast_fields()
-                .u64(score_field)
+                .fast_field_reader::<u64>(score_field)
                 .unwrap();
             assert_eq!(score_field_reader.min_value(), 1);
             assert_eq!(score_field_reader.max_value(), 3);
@@ -1084,8 +1027,7 @@ mod tests {
             );
             let score_field_reader = searcher
                 .segment_reader(0)
-                .fast_fields()
-                .u64(score_field)
+                .fast_field_reader::<u64>(score_field)
                 .unwrap();
             assert_eq!(score_field_reader.min_value(), 3);
             assert_eq!(score_field_reader.max_value(), 7000);
@@ -1131,8 +1073,7 @@ mod tests {
             );
             let score_field_reader = searcher
                 .segment_reader(0)
-                .fast_fields()
-                .u64(score_field)
+                .fast_field_reader::<u64>(score_field)
                 .unwrap();
             assert_eq!(score_field_reader.min_value(), 3);
             assert_eq!(score_field_reader.max_value(), 7000);
@@ -1184,8 +1125,7 @@ mod tests {
             );
             let score_field_reader = searcher
                 .segment_reader(0)
-                .fast_fields()
-                .u64(score_field)
+                .fast_field_reader::<u64>(score_field)
                 .unwrap();
             assert_eq!(score_field_reader.min_value(), 6000);
             assert_eq!(score_field_reader.max_value(), 7000);
@@ -1431,7 +1371,7 @@ mod tests {
 
         {
             let segment = searcher.segment_reader(0u32);
-            let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
+            let ff_reader = segment.multi_fast_field_reader(int_field).unwrap();
 
             ff_reader.get_vals(0, &mut vals);
             assert_eq!(&vals, &[1, 2]);
@@ -1466,7 +1406,7 @@ mod tests {
 
         {
             let segment = searcher.segment_reader(1u32);
-            let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
+            let ff_reader = segment.multi_fast_field_reader(int_field).unwrap();
             ff_reader.get_vals(0, &mut vals);
             assert_eq!(&vals, &[28, 27]);
 
@@ -1476,7 +1416,7 @@ mod tests {
 
         {
             let segment = searcher.segment_reader(2u32);
-            let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
+            let ff_reader = segment.multi_fast_field_reader(int_field).unwrap();
             ff_reader.get_vals(0, &mut vals);
             assert_eq!(&vals, &[20]);
         }
@@ -1509,7 +1449,7 @@ mod tests {
                     .collect::<Vec<_>>()
             );
             let segment = searcher.segment_reader(0u32);
-            let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
+            let ff_reader = segment.multi_fast_field_reader(int_field).unwrap();
 
             ff_reader.get_vals(0, &mut vals);
             assert_eq!(&vals, &[1, 2]);

src/indexer/segment_register.rs

@@ -56,7 +56,7 @@ impl SegmentRegister {
             .values()
             .map(|segment_entry| segment_entry.meta().clone())
             .collect();
-        segment_ids.sort_by_key(SegmentMeta::id);
+        segment_ids.sort_by_key(|meta| meta.id());
         segment_ids
     }
 

src/indexer/segment_updater.rs

@@ -420,7 +420,6 @@ impl SegmentUpdater {
             })
             .collect::<Vec<_>>();
         merge_candidates.extend(committed_merge_candidates.into_iter());
-
         for merge_operation in merge_candidates {
             match self.start_merge_impl(merge_operation) {
                 Ok(merge_future) => {

src/indexer/segment_writer.rs

@@ -5,7 +5,6 @@ use fastfield::FastFieldsWriter;
 use fieldnorm::FieldNormsWriter;
 use indexer::segment_serializer::SegmentSerializer;
 use postings::MultiFieldPostingsWriter;
-use schema::FieldEntry;
 use schema::FieldType;
 use schema::Schema;
 use schema::Term;
@@ -54,7 +53,7 @@ impl SegmentWriter {
             schema
                 .fields()
                 .iter()
-                .map(FieldEntry::field_type)
+                .map(|field_entry| field_entry.field_type())
                 .map(|field_type| match *field_type {
                     FieldType::Str(ref text_options) => text_options
                         .get_indexing_options()

src/lib.rs

@@ -174,7 +174,6 @@ extern crate downcast_rs;
 #[macro_use]
 extern crate fail;
 
-#[cfg(feature = "mmap")]
 #[cfg(test)]
 mod functional_test;
 
@@ -876,28 +875,28 @@ mod tests {
         let searcher = reader.searcher();
         let segment_reader: &SegmentReader = searcher.segment_reader(0);
         {
-            let fast_field_reader_opt = segment_reader.fast_fields().u64(text_field);
-            assert!(fast_field_reader_opt.is_none());
+            let fast_field_reader_res = segment_reader.fast_field_reader::<u64>(text_field);
+            assert!(fast_field_reader_res.is_err());
         }
         {
-            let fast_field_reader_opt = segment_reader.fast_fields().u64(stored_int_field);
-            assert!(fast_field_reader_opt.is_none());
+            let fast_field_reader_res = segment_reader.fast_field_reader::<u64>(stored_int_field);
+            assert!(fast_field_reader_res.is_err());
         }
         {
-            let fast_field_reader_opt = segment_reader.fast_fields().u64(fast_field_signed);
-            assert!(fast_field_reader_opt.is_none());
+            let fast_field_reader_res = segment_reader.fast_field_reader::<u64>(fast_field_signed);
+            assert!(fast_field_reader_res.is_err());
         }
         {
-            let fast_field_reader_opt = segment_reader.fast_fields().i64(fast_field_signed);
-            assert!(fast_field_reader_opt.is_some());
-            let fast_field_reader = fast_field_reader_opt.unwrap();
+            let fast_field_reader_res = segment_reader.fast_field_reader::<i64>(fast_field_signed);
+            assert!(fast_field_reader_res.is_ok());
+            let fast_field_reader = fast_field_reader_res.unwrap();
             assert_eq!(fast_field_reader.get(0), 4i64)
         }
 
         {
-            let fast_field_reader_opt = segment_reader.fast_fields().i64(fast_field_signed);
-            assert!(fast_field_reader_opt.is_some());
-            let fast_field_reader = fast_field_reader_opt.unwrap();
+            let fast_field_reader_res = segment_reader.fast_field_reader::<i64>(fast_field_signed);
+            assert!(fast_field_reader_res.is_ok());
+            let fast_field_reader = fast_field_reader_res.unwrap();
             assert_eq!(fast_field_reader.get(0), 4i64)
         }
     }

src/postings/block_search.rs

@@ -1,249 +0,0 @@
-use postings::compression::AlignedBuffer;
-
-/// This modules define the logic used to search for a doc in a given
-/// block. (at most 128 docs)
-///
-/// Searching within a block is a hotspot when running intersection.
-/// so it was worth defining it in its own module.
-
-#[cfg(target_arch = "x86_64")]
-mod sse2 {
-    use postings::compression::{AlignedBuffer, COMPRESSION_BLOCK_SIZE};
-    use std::arch::x86_64::__m128i as DataType;
-    use std::arch::x86_64::_mm_add_epi32 as op_add;
-    use std::arch::x86_64::_mm_cmplt_epi32 as op_lt;
-    use std::arch::x86_64::_mm_load_si128 as op_load; // requires 128-bits alignment
-    use std::arch::x86_64::_mm_set1_epi32 as set1;
-    use std::arch::x86_64::_mm_setzero_si128 as set0;
-    use std::arch::x86_64::_mm_sub_epi32 as op_sub;
-    use std::arch::x86_64::{_mm_cvtsi128_si32, _mm_shuffle_epi32};
-
-    const MASK1: i32 = 78;
-    const MASK2: i32 = 177;
-
-    /// Performs an exhaustive linear search over the
-    ///
-    /// There is no early exit here. We simply count the
-    /// number of elements that are `< target`.
-    pub(crate) fn linear_search_sse2_128(arr: &AlignedBuffer, target: u32) -> usize {
-        unsafe {
-            let ptr = arr as *const AlignedBuffer as *const DataType;
-            let vkey = set1(target as i32);
-            let mut cnt = set0();
-            // We work over 4 `__m128i` at a time.
-            // A single `__m128i` actual contains 4 `u32`.
-            for i in 0..(COMPRESSION_BLOCK_SIZE as isize) / (4 * 4) {
-                let cmp1 = op_lt(op_load(ptr.offset(i * 4)), vkey);
-                let cmp2 = op_lt(op_load(ptr.offset(i * 4 + 1)), vkey);
-                let cmp3 = op_lt(op_load(ptr.offset(i * 4 + 2)), vkey);
-                let cmp4 = op_lt(op_load(ptr.offset(i * 4 + 3)), vkey);
-                let sum = op_add(op_add(cmp1, cmp2), op_add(cmp3, cmp4));
-                cnt = op_sub(cnt, sum);
-            }
-            cnt = op_add(cnt, _mm_shuffle_epi32(cnt, MASK1));
-            cnt = op_add(cnt, _mm_shuffle_epi32(cnt, MASK2));
-            _mm_cvtsi128_si32(cnt) as usize
-        }
-    }
-
-    #[cfg(test)]
-    mod test {
-        use super::linear_search_sse2_128;
-        use postings::compression::{AlignedBuffer, COMPRESSION_BLOCK_SIZE};
-
-        #[test]
-        fn test_linear_search_sse2_128_u32() {
-            let mut block = [0u32; COMPRESSION_BLOCK_SIZE];
-            for el in 0u32..128u32 {
-                block[el as usize] = el * 2 + 1 << 18;
-            }
-            let target = block[64] + 1;
-            assert_eq!(linear_search_sse2_128(&AlignedBuffer(block), target), 65);
-        }
-    }
-}
-
-/// This `linear search` browser exhaustively through the array.
-/// but the early exit is very difficult to predict.
-///
-/// Coupled with `exponential search` this function is likely
-/// to be called with the same `len`
-fn linear_search(arr: &[u32], target: u32) -> usize {
-    arr.iter().map(|&el| if el < target { 1 } else { 0 }).sum()
-}
-
-fn exponential_search(arr: &[u32], target: u32) -> (usize, usize) {
-    let end = arr.len();
-    let mut begin = 0;
-    for &pivot in &[1, 3, 7, 15, 31, 63] {
-        if pivot >= end {
-            break;
-        }
-        if arr[pivot] > target {
-            return (begin, pivot);
-        }
-        begin = pivot;
-    }
-    (begin, end)
-}
-
-fn galloping(block_docs: &[u32], target: u32) -> usize {
-    let (start, end) = exponential_search(&block_docs, target);
-    start + linear_search(&block_docs[start..end], target)
-}
-
-/// Tantivy may rely on SIMD instructions to search for a specific document within
-/// a given block.
-#[derive(Clone, Copy, PartialEq)]
-pub enum BlockSearcher {
-    #[cfg(target_arch = "x86_64")]
-    SSE2,
-    Scalar,
-}
-
-impl BlockSearcher {
-    /// Search the first index containing an element greater or equal to
-    /// the target.
-    ///
-    /// The results should be equivalent to
-    /// ```ignore
-    /// block[..]
-    //       .iter()
-    //       .take_while(|&&val| val < target)
-    //       .count()
-    /// ```
-    ///
-    /// The `start` argument is just used to hint that the response is
-    /// greater than beyond `start`. The implementation may or may not use
-    /// it for optimization.
-    ///
-    /// # Assumption
-    ///
-    /// The array len is > start.
-    /// The block is sorted
-    /// The target is assumed greater or equal to the `arr[start]`.
-    /// The target is assumed smaller or equal to the last element of the block.
-    ///
-    /// Currently the scalar implementation starts by an exponential search, and
-    /// then operates a linear search in the result subarray.
-    ///
-    /// If SSE2 instructions are available in the `(platform, running CPU)`,
-    /// then we use a different implementation that does an exhaustive linear search over
-    /// the full block whenever the block is full (`len == 128`). It is surprisingly faster, most likely because of the lack
-    /// of branch.
-    pub(crate) fn search_in_block(
-        self,
-        block_docs: &AlignedBuffer,
-        len: usize,
-        start: usize,
-        target: u32,
-    ) -> usize {
-        #[cfg(target_arch = "x86_64")]
-        {
-            use postings::compression::COMPRESSION_BLOCK_SIZE;
-            if self == BlockSearcher::SSE2 && len == COMPRESSION_BLOCK_SIZE {
-                return sse2::linear_search_sse2_128(block_docs, target);
-            }
-        }
-        start + galloping(&block_docs.0[start..len], target)
-    }
-}
-
-impl Default for BlockSearcher {
-    fn default() -> BlockSearcher {
-        #[cfg(target_arch = "x86_64")]
-        {
-            if is_x86_feature_detected!("sse2") {
-                return BlockSearcher::SSE2;
-            }
-        }
-        BlockSearcher::Scalar
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::exponential_search;
-    use super::linear_search;
-    use super::BlockSearcher;
-    use postings::compression::{AlignedBuffer, COMPRESSION_BLOCK_SIZE};
-
-    #[test]
-    fn test_linear_search() {
-        let len: usize = 50;
-        let arr: Vec<u32> = (0..len).map(|el| 1u32 + (el as u32) * 2).collect();
-        for target in 1..*arr.last().unwrap() {
-            let res = linear_search(&arr[..], target);
-            if res > 0 {
-                assert!(arr[res - 1] < target);
-            }
-            if res < len {
-                assert!(arr[res] >= target);
-            }
-        }
-    }
-
-    #[test]
-    fn test_exponentiel_search() {
-        assert_eq!(exponential_search(&[1, 2], 0), (0, 1));
-        assert_eq!(exponential_search(&[1, 2], 1), (0, 1));
-        assert_eq!(
-            exponential_search(&[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 7),
-            (3, 7)
-        );
-    }
-
-    fn util_test_search_in_block(block_searcher: BlockSearcher, block: &[u32], target: u32) {
-        let cursor = search_in_block_trivial_but_slow(block, target);
-        assert!(block.len() < COMPRESSION_BLOCK_SIZE);
-        let mut output_buffer = [u32::max_value(); COMPRESSION_BLOCK_SIZE];
-        output_buffer[..block.len()].copy_from_slice(block);
-        for i in 0..cursor {
-            assert_eq!(
-                block_searcher.search_in_block(
-                    &AlignedBuffer(output_buffer),
-                    block.len(),
-                    i,
-                    target
-                ),
-                cursor
-            );
-        }
-    }
-
-    fn util_test_search_in_block_all(block_searcher: BlockSearcher, block: &[u32]) {
-        use std::collections::HashSet;
-        let mut targets = HashSet::new();
-        for (i, val) in block.iter().cloned().enumerate() {
-            if i > 0 {
-                targets.insert(val - 1);
-            }
-            targets.insert(val);
-        }
-        for target in targets {
-            util_test_search_in_block(block_searcher, block, target);
-        }
-    }
-
-    fn search_in_block_trivial_but_slow(block: &[u32], target: u32) -> usize {
-        block.iter().take_while(|&&val| val < target).count()
-    }
-
-    fn test_search_in_block_util(block_searcher: BlockSearcher) {
-        for len in 1u32..128u32 {
-            let v: Vec<u32> = (0..len).map(|i| i * 2).collect();
-            util_test_search_in_block_all(block_searcher, &v[..]);
-        }
-    }
-
-    #[test]
-    fn test_search_in_block_scalar() {
-        test_search_in_block_util(BlockSearcher::Scalar);
-    }
-
-    #[cfg(target_arch = "x86_64")]
-    #[test]
-    fn test_search_in_block_sse2() {
-        test_search_in_block_util(BlockSearcher::SSE2);
-    }
-}

src/postings/compression/mod.rs

@@ -43,14 +43,9 @@ impl BlockEncoder {
     }
 }
 
-/// We ensure that the OutputBuffer is align on 128 bits
-/// in order to run SSE2 linear search on it.
-#[repr(align(128))]
-pub(crate) struct AlignedBuffer(pub [u32; COMPRESSION_BLOCK_SIZE]);
-
 pub struct BlockDecoder {
     bitpacker: BitPacker4x,
-    output: AlignedBuffer,
+    pub output: [u32; COMPRESSION_BLOCK_SIZE + 1],
     pub output_len: usize,
 }
 
@@ -60,9 +55,11 @@ impl BlockDecoder {
     }
 
     pub fn with_val(val: u32) -> BlockDecoder {
+        let mut output = [val; COMPRESSION_BLOCK_SIZE + 1];
+        output[COMPRESSION_BLOCK_SIZE] = 0u32;
         BlockDecoder {
             bitpacker: BitPacker4x::new(),
-            output: AlignedBuffer([val; COMPRESSION_BLOCK_SIZE]),
+            output,
             output_len: 0,
         }
     }
@@ -75,28 +72,23 @@ impl BlockDecoder {
     ) -> usize {
         self.output_len = COMPRESSION_BLOCK_SIZE;
         self.bitpacker
-            .decompress_sorted(offset, &compressed_data, &mut self.output.0, num_bits)
+            .decompress_sorted(offset, &compressed_data, &mut self.output, num_bits)
     }
 
     pub fn uncompress_block_unsorted(&mut self, compressed_data: &[u8], num_bits: u8) -> usize {
         self.output_len = COMPRESSION_BLOCK_SIZE;
         self.bitpacker
-            .decompress(&compressed_data, &mut self.output.0, num_bits)
+            .decompress(&compressed_data, &mut self.output, num_bits)
     }
 
     #[inline]
     pub fn output_array(&self) -> &[u32] {
-        &self.output.0[..self.output_len]
-    }
-
-    #[inline]
-    pub(crate) fn output_aligned(&self) -> (&AlignedBuffer, usize) {
-        (&self.output, self.output_len)
+        &self.output[..self.output_len]
     }
 
     #[inline]
     pub fn output(&self, idx: usize) -> u32 {
-        self.output.0[idx]
+        self.output[idx]
     }
 }
 
@@ -167,12 +159,12 @@ impl VIntDecoder for BlockDecoder {
         num_els: usize,
     ) -> usize {
         self.output_len = num_els;
-        vint::uncompress_sorted(compressed_data, &mut self.output.0[..num_els], offset)
+        vint::uncompress_sorted(compressed_data, &mut self.output[..num_els], offset)
     }
 
     fn uncompress_vint_unsorted<'a>(&mut self, compressed_data: &'a [u8], num_els: usize) -> usize {
         self.output_len = num_els;
-        vint::uncompress_unsorted(compressed_data, &mut self.output.0[..num_els])
+        vint::uncompress_unsorted(compressed_data, &mut self.output[..num_els])
     }
 }
 

src/postings/mod.rs

@@ -2,7 +2,6 @@
 Postings module (also called inverted index)
 */
 
-mod block_search;
 pub(crate) mod compression;
 /// Postings module
 ///
@@ -17,8 +16,6 @@ mod skip;
 mod stacker;
 mod term_info;
 
-pub(crate) use self::block_search::BlockSearcher;
-
 pub(crate) use self::postings_writer::MultiFieldPostingsWriter;
 pub use self::serializer::{FieldSerializer, InvertedIndexSerializer};
 
@@ -55,15 +52,13 @@ pub mod tests {
     use fieldnorm::FieldNormReader;
     use indexer::operation::AddOperation;
     use indexer::SegmentWriter;
-    use merge_policy::NoMergePolicy;
     use query::Scorer;
     use rand::rngs::StdRng;
     use rand::{Rng, SeedableRng};
+    use schema::Field;
+    use schema::IndexRecordOption;
     use schema::{Document, Schema, Term, INDEXED, STRING, TEXT};
-    use schema::{Field, TextOptions};
-    use schema::{IndexRecordOption, TextFieldIndexing};
     use std::iter;
-    use tokenizer::{SimpleTokenizer, MAX_TOKEN_LEN};
     use DocId;
     use Score;
 
@@ -109,7 +104,9 @@ pub mod tests {
         let searcher = index.reader().unwrap().searcher();
         let inverted_index = searcher.segment_reader(0u32).inverted_index(title);
         let term = Term::from_field_text(title, "abc");
+
         let mut positions = Vec::new();
+
         {
             let mut postings = inverted_index
                 .read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
@@ -162,52 +159,6 @@ pub mod tests {
         }
     }
 
-    #[test]
-    pub fn test_drop_token_that_are_too_long() {
-        let ok_token_text: String = iter::repeat('A').take(MAX_TOKEN_LEN).collect();
-        let mut exceeding_token_text: String = iter::repeat('A').take(MAX_TOKEN_LEN + 1).collect();
-        exceeding_token_text.push_str(" hello");
-        let mut schema_builder = Schema::builder();
-        let text_options = TextOptions::default().set_indexing_options(
-            TextFieldIndexing::default()
-                .set_index_option(IndexRecordOption::WithFreqsAndPositions)
-                .set_tokenizer("simple_no_truncation"),
-        );
-        let text_field = schema_builder.add_text_field("text", text_options);
-        let schema = schema_builder.build();
-        let index = Index::create_in_ram(schema.clone());
-        index
-            .tokenizers()
-            .register("simple_no_truncation", SimpleTokenizer);
-        let reader = index.reader().unwrap();
-        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-        index_writer.set_merge_policy(Box::new(NoMergePolicy));
-        {
-            index_writer.add_document(doc!(text_field=>exceeding_token_text));
-            index_writer.commit().unwrap();
-            reader.reload().unwrap();
-            let searcher = reader.searcher();
-            let segment_reader = searcher.segment_reader(0u32);
-            let inverted_index = segment_reader.inverted_index(text_field);
-            assert_eq!(inverted_index.terms().num_terms(), 1);
-            let mut bytes = vec![];
-            assert!(inverted_index.terms().ord_to_term(0, &mut bytes));
-            assert_eq!(&bytes, b"hello");
-        }
-        {
-            index_writer.add_document(doc!(text_field=>ok_token_text.clone()));
-            index_writer.commit().unwrap();
-            reader.reload().unwrap();
-            let searcher = reader.searcher();
-            let segment_reader = searcher.segment_reader(1u32);
-            let inverted_index = segment_reader.inverted_index(text_field);
-            assert_eq!(inverted_index.terms().num_terms(), 1);
-            let mut bytes = vec![];
-            assert!(inverted_index.terms().ord_to_term(0, &mut bytes));
-            assert_eq!(&bytes[..], ok_token_text.as_bytes());
-        }
-    }
-
     #[test]
     pub fn test_position_and_fieldnorm1() {
         let mut positions = Vec::new();

src/postings/postings_writer.rs

@@ -12,8 +12,8 @@ use std::io;
 use std::marker::PhantomData;
 use std::ops::DerefMut;
 use termdict::TermOrdinal;
+use tokenizer::Token;
 use tokenizer::TokenStream;
-use tokenizer::{Token, MAX_TOKEN_LEN};
 use DocId;
 use Result;
 
@@ -210,18 +210,8 @@ pub trait PostingsWriter {
     ) -> u32 {
         let mut term = Term::for_field(field);
         let mut sink = |token: &Token| {
-            // We skip all tokens with a len greater than u16.
-            if token.text.len() <= MAX_TOKEN_LEN {
-                term.set_text(token.text.as_str());
-                self.subscribe(term_index, doc_id, token.position as u32, &term, heap);
-            } else {
-                info!(
-                    "A token exceeding MAX_TOKEN_LEN ({}>{}) was dropped. Search for \
-                     MAX_TOKEN_LEN in the documentation for more information.",
-                    token.text.len(),
-                    MAX_TOKEN_LEN
-                );
-            }
+            term.set_text(token.text.as_str());
+            self.subscribe(term_index, doc_id, token.position as u32, &term, heap);
         };
         token_stream.process(&mut sink)
     }

src/postings/segment_postings.rs

@@ -4,10 +4,9 @@ use common::{BinarySerializable, VInt};
 use docset::{DocSet, SkipResult};
 use owned_read::OwnedRead;
 use positions::PositionReader;
-use postings::compression::{compressed_block_size, AlignedBuffer};
+use postings::compression::compressed_block_size;
 use postings::compression::{BlockDecoder, VIntDecoder, COMPRESSION_BLOCK_SIZE};
 use postings::serializer::PostingsSerializer;
-use postings::BlockSearcher;
 use postings::FreqReadingOption;
 use postings::Postings;
 use postings::SkipReader;
@@ -61,7 +60,6 @@ pub struct SegmentPostings {
     block_cursor: BlockSegmentPostings,
     cur: usize,
     position_computer: Option<PositionComputer>,
-    block_searcher: BlockSearcher,
 }
 
 impl SegmentPostings {
@@ -72,7 +70,6 @@ impl SegmentPostings {
             block_cursor: empty_block_cursor,
             cur: COMPRESSION_BLOCK_SIZE,
             position_computer: None,
-            block_searcher: BlockSearcher::default(),
         }
     }
 
@@ -120,33 +117,42 @@ impl SegmentPostings {
             block_cursor: segment_block_postings,
             cur: COMPRESSION_BLOCK_SIZE, // cursor within the block
             position_computer: positions_stream_opt.map(PositionComputer::new),
-            block_searcher: BlockSearcher::default(),
         }
     }
 }
 
-impl DocSet for SegmentPostings {
-    // goes to the next element.
-    // next needs to be called a first time to point to the correct element.
-    #[inline]
-    fn advance(&mut self) -> bool {
-        if self.position_computer.is_some() && self.cur < COMPRESSION_BLOCK_SIZE {
-            let term_freq = self.term_freq() as usize;
-            if let Some(position_computer) = self.position_computer.as_mut() {
-                position_computer.add_skip(term_freq);
-            }
-        }
-        self.cur += 1;
-        if self.cur >= self.block_cursor.block_len() {
-            self.cur = 0;
-            if !self.block_cursor.advance() {
-                self.cur = COMPRESSION_BLOCK_SIZE;
-                return false;
-            }
-        }
-        true
-    }
+fn linear_search(arr: &[u32], target: u32) -> usize {
+    arr.iter().map(|&el| if el < target { 1 } else { 0 }).sum()
+}
 
+fn exponential_search(arr: &[u32], target: u32) -> (usize, usize) {
+    let end = arr.len();
+    let mut begin = 0;
+    for &pivot in &[1, 3, 7, 15, 31, 63] {
+        if pivot >= end {
+            break;
+        }
+        if arr[pivot] > target {
+            return (begin, pivot);
+        }
+        begin = pivot;
+    }
+    (begin, end)
+}
+
+/// Search the first index containing an element greater or equal to the target.
+///
+/// # Assumption
+///
+/// The array is assumed non empty.
+/// The target is assumed greater or equal to the first element.
+/// The target is assumed smaller or equal to the last element.
+fn search_within_block(block_docs: &[u32], target: u32) -> usize {
+    let (start, end) = exponential_search(block_docs, target);
+    start + linear_search(&block_docs[start..end], target)
+}
+
+impl DocSet for SegmentPostings {
     fn skip_next(&mut self, target: DocId) -> SkipResult {
         if !self.advance() {
             return SkipResult::End;
@@ -169,6 +175,7 @@ impl DocSet for SegmentPostings {
 
         // skip blocks until one that might contain the target
         // check if we need to go to the next block
+        let need_positions = self.position_computer.is_some();
         let mut sum_freqs_skipped: u32 = 0;
         if !self
             .block_cursor
@@ -182,7 +189,7 @@ impl DocSet for SegmentPostings {
             // we are not in the right block.
             //
             // First compute all of the freqs skipped from the current block.
-            if self.position_computer.is_some() {
+            if need_positions {
                 sum_freqs_skipped = self.block_cursor.freqs()[self.cur..].iter().sum();
                 match self.block_cursor.skip_to(target) {
                     BlockSegmentPostingsSkipResult::Success(block_skip_freqs) => {
@@ -201,21 +208,25 @@ impl DocSet for SegmentPostings {
             self.cur = 0;
         }
 
-        let cur = self.cur;
-
         // we're in the right block now, start with an exponential search
-        let (output, len) = self.block_cursor.docs_aligned();
+        let block_docs = self.block_cursor.docs();
         let new_cur = self
-            .block_searcher
-            .search_in_block(&output, len, cur, target);
-        if let Some(position_computer) = self.position_computer.as_mut() {
-            sum_freqs_skipped += self.block_cursor.freqs()[cur..new_cur].iter().sum::<u32>();
-            position_computer.add_skip(sum_freqs_skipped as usize);
+            .cur
+            .wrapping_add(search_within_block(&block_docs[self.cur..], target));
+
+        if need_positions {
+            sum_freqs_skipped += self.block_cursor.freqs()[self.cur..new_cur]
+                .iter()
+                .sum::<u32>();
+            self.position_computer
+                .as_mut()
+                .unwrap()
+                .add_skip(sum_freqs_skipped as usize);
         }
         self.cur = new_cur;
 
         // `doc` is now the first element >= `target`
-        let doc = output.0[new_cur];
+        let doc = block_docs[new_cur];
         debug_assert!(doc >= target);
         if doc == target {
             SkipResult::Reached
@@ -224,25 +235,40 @@ impl DocSet for SegmentPostings {
         }
     }
 
+    // goes to the next element.
+    // next needs to be called a first time to point to the correct element.
+    #[inline]
+    fn advance(&mut self) -> bool {
+        if self.position_computer.is_some() {
+            let term_freq = self.term_freq() as usize;
+            self.position_computer.as_mut().unwrap().add_skip(term_freq);
+        }
+        self.cur += 1;
+        if self.cur >= self.block_cursor.block_len() {
+            self.cur = 0;
+            if !self.block_cursor.advance() {
+                self.cur = COMPRESSION_BLOCK_SIZE;
+                return false;
+            }
+        }
+        true
+    }
+
+    fn size_hint(&self) -> u32 {
+        self.len() as u32
+    }
+
     /// Return the current document's `DocId`.
-    ///
-    /// # Panics
-    ///
-    /// Will panics if called without having called advance before.
     #[inline]
     fn doc(&self) -> DocId {
         let docs = self.block_cursor.docs();
         debug_assert!(
             self.cur < docs.len(),
-            "Have you forgotten to call `.advance()` at least once before calling `.doc()`                                      ."
+            "Have you forgotten to call `.advance()` at least once before calling .doc()."
         );
         docs[self.cur]
     }
 
-    fn size_hint(&self) -> u32 {
-        self.len() as u32
-    }
-
     fn append_to_bitset(&mut self, bitset: &mut BitSet) {
         // finish the current block
         if self.advance() {
@@ -266,33 +292,17 @@ impl HasLen for SegmentPostings {
 }
 
 impl Postings for SegmentPostings {
-    /// Returns the frequency associated to the current document.
-    /// If the schema is set up so that no frequency have been encoded,
-    /// this method should always return 1.
-    ///
-    /// # Panics
-    ///
-    /// Will panics if called without having called advance before.
     fn term_freq(&self) -> u32 {
-        debug_assert!(
-            // Here we do not use the len of `freqs()`
-            // because it is actually ok to request for the freq of doc
-            // even if no frequency were encoded for the field.
-            //
-            // In that case we hit the block just as if the frequency had been
-            // decoded. The block is simply prefilled by the value 1.
-            self.cur < COMPRESSION_BLOCK_SIZE,
-            "Have you forgotten to call `.advance()` at least once before calling \
-             `.term_freq()`."
-        );
         self.block_cursor.freq(self.cur)
     }
 
     fn positions_with_offset(&mut self, offset: u32, output: &mut Vec<u32>) {
-        let term_freq = self.term_freq() as usize;
-        if let Some(position_comp) = self.position_computer.as_mut() {
-            output.resize(term_freq, 0u32);
-            position_comp.positions_with_offset(offset, &mut output[..]);
+        if self.position_computer.is_some() {
+            output.resize(self.term_freq() as usize, 0u32);
+            self.position_computer
+                .as_mut()
+                .unwrap()
+                .positions_with_offset(offset, &mut output[..])
         } else {
             output.clear();
         }
@@ -414,10 +424,6 @@ impl BlockSegmentPostings {
         self.doc_decoder.output_array()
     }
 
-    pub(crate) fn docs_aligned(&self) -> (&AlignedBuffer, usize) {
-        self.doc_decoder.output_aligned()
-    }
-
     /// Return the document at index `idx` of the block.
     #[inline]
     pub fn doc(&self, idx: usize) -> u32 {
@@ -608,13 +614,16 @@ impl<'b> Streamer<'b> for BlockSegmentPostings {
 
 #[cfg(test)]
 mod tests {
+
+    use super::exponential_search;
+    use super::linear_search;
+    use super::search_within_block;
     use super::BlockSegmentPostings;
     use super::BlockSegmentPostingsSkipResult;
     use super::SegmentPostings;
     use common::HasLen;
     use core::Index;
     use docset::DocSet;
-    use postings::postings::Postings;
     use schema::IndexRecordOption;
     use schema::Schema;
     use schema::Term;
@@ -623,6 +632,21 @@ mod tests {
     use DocId;
     use SkipResult;
 
+    #[test]
+    fn test_linear_search() {
+        let len: usize = 50;
+        let arr: Vec<u32> = (0..len).map(|el| 1u32 + (el as u32) * 2).collect();
+        for target in 1..*arr.last().unwrap() {
+            let res = linear_search(&arr[..], target);
+            if res > 0 {
+                assert!(arr[res - 1] < target);
+            }
+            if res < len {
+                assert!(arr[res] >= target);
+            }
+        }
+    }
+
     #[test]
     fn test_empty_segment_postings() {
         let mut postings = SegmentPostings::empty();
@@ -631,18 +655,6 @@ mod tests {
         assert_eq!(postings.len(), 0);
     }
 
-    #[test]
-    #[should_panic(expected = "Have you forgotten to call `.advance()`")]
-    fn test_panic_if_doc_called_before_advance() {
-        SegmentPostings::empty().doc();
-    }
-
-    #[test]
-    #[should_panic(expected = "Have you forgotten to call `.advance()`")]
-    fn test_panic_if_freq_called_before_advance() {
-        SegmentPostings::empty().term_freq();
-    }
-
     #[test]
     fn test_empty_block_segment_postings() {
         let mut postings = BlockSegmentPostings::empty();
@@ -650,6 +662,56 @@ mod tests {
         assert_eq!(postings.doc_freq(), 0);
     }
 
+    fn search_within_block_trivial_but_slow(block: &[u32], target: u32) -> usize {
+        block
+            .iter()
+            .cloned()
+            .enumerate()
+            .filter(|&(_, ref val)| *val >= target)
+            .next()
+            .unwrap()
+            .0
+    }
+
+    #[test]
+    fn test_exponentiel_search() {
+        assert_eq!(exponential_search(&[1, 2], 0), (0, 1));
+        assert_eq!(exponential_search(&[1, 2], 1), (0, 1));
+        assert_eq!(
+            exponential_search(&[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 7),
+            (3, 7)
+        );
+    }
+
+    fn util_test_search_within_block(block: &[u32], target: u32) {
+        assert_eq!(
+            search_within_block(block, target),
+            search_within_block_trivial_but_slow(block, target)
+        );
+    }
+
+    fn util_test_search_within_block_all(block: &[u32]) {
+        use std::collections::HashSet;
+        let mut targets = HashSet::new();
+        for (i, val) in block.iter().cloned().enumerate() {
+            if i > 0 {
+                targets.insert(val - 1);
+            }
+            targets.insert(val);
+        }
+        for target in targets {
+            util_test_search_within_block(block, target);
+        }
+    }
+
+    #[test]
+    fn test_search_within_block() {
+        for len in 1u32..128u32 {
+            let v: Vec<u32> = (0..len).map(|i| i * 2).collect();
+            util_test_search_within_block_all(&v[..]);
+        }
+    }
+
     #[test]
     fn test_block_segment_postings() {
         let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>());

src/postings/serializer.rs

@@ -14,7 +14,7 @@ use termdict::{TermDictionaryBuilder, TermOrdinal};
 use DocId;
 use Result;
 
-/// `InvertedIndexSerializer` is in charge of serializing
+/// `PostingsSerializer` is in charge of serializing
 /// postings on disk, in the
 /// * `.idx` (inverted index)
 /// * `.pos` (positions file)
@@ -54,7 +54,7 @@ pub struct InvertedIndexSerializer {
 }
 
 impl InvertedIndexSerializer {
-    /// Open a new `InvertedIndexSerializer` for the given segment
+    /// Open a new `PostingsSerializer` for the given segment
     fn create(
         terms_write: CompositeWrite<WritePtr>,
         postings_write: CompositeWrite<WritePtr>,
@@ -175,7 +175,7 @@ impl<'a> FieldSerializer<'a> {
         let positions_idx = self
             .positions_serializer_opt
             .as_ref()
-            .map(PositionSerializer::positions_idx)
+            .map(|positions_serializer| positions_serializer.positions_idx())
             .unwrap_or(0u64);
         TermInfo {
             doc_freq: 0,

src/query/intersection.rs

@@ -14,35 +14,41 @@ use Score;
 /// specialized implementation if the two
 /// shortest scorers are `TermScorer`s.
 pub fn intersect_scorers(mut scorers: Vec<Box<Scorer>>) -> Box<Scorer> {
-    if scorers.is_empty() {
-        return Box::new(EmptyScorer);
-    }
-    if scorers.len() == 1 {
-        return scorers.pop().unwrap();
-    }
-    // We know that we have at least 2 elements.
     let num_docsets = scorers.len();
     scorers.sort_by(|left, right| right.size_hint().cmp(&left.size_hint()));
-    let left = scorers.pop().unwrap();
-    let right = scorers.pop().unwrap();
+    let rarest_opt = scorers.pop();
+    let second_rarest_opt = scorers.pop();
     scorers.reverse();
-    let all_term_scorers = [&left, &right]
-        .iter()
-        .all(|&scorer| scorer.is::<TermScorer>());
-    if all_term_scorers {
-        return Box::new(Intersection {
-            left: *(left.downcast::<TermScorer>().map_err(|_| ()).unwrap()),
-            right: *(right.downcast::<TermScorer>().map_err(|_| ()).unwrap()),
-            others: scorers,
-            num_docsets,
-        });
+    match (rarest_opt, second_rarest_opt) {
+        (None, None) => Box::new(EmptyScorer),
+        (Some(single_docset), None) => single_docset,
+        (Some(left), Some(right)) => {
+            {
+                let all_term_scorers = [&left, &right]
+                    .iter()
+                    .all(|&scorer| scorer.is::<TermScorer>());
+                if all_term_scorers {
+                    let left = *(left.downcast::<TermScorer>().map_err(|_| ()).unwrap());
+                    let right = *(right.downcast::<TermScorer>().map_err(|_| ()).unwrap());
+                    return Box::new(Intersection {
+                        left,
+                        right,
+                        others: scorers,
+                        num_docsets,
+                    });
+                }
+            }
+            Box::new(Intersection {
+                left,
+                right,
+                others: scorers,
+                num_docsets,
+            })
+        }
+        _ => {
+            unreachable!();
+        }
     }
-    Box::new(Intersection {
-        left,
-        right,
-        others: scorers,
-        num_docsets,
-    })
 }
 
 /// Creates a `DocSet` that iterator through the intersection of two `DocSet`s.
@@ -118,6 +124,7 @@ impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOt
                         return false;
                     }
                 }
+
                 match left.skip_next(candidate) {
                     SkipResult::Reached => {
                         break;
@@ -133,36 +140,35 @@ impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOt
             }
             // test the remaining scorers;
             for (ord, docset) in self.others.iter_mut().enumerate() {
-                if ord == other_candidate_ord {
-                    continue;
-                }
-                // `candidate_ord` is already at the
-                // right position.
-                //
-                // Calling `skip_next` would advance this docset
-                // and miss it.
-                match docset.skip_next(candidate) {
-                    SkipResult::Reached => {}
-                    SkipResult::OverStep => {
-                        // this is not in the intersection,
-                        // let's update our candidate.
-                        candidate = docset.doc();
-                        match left.skip_next(candidate) {
-                            SkipResult::Reached => {
-                                other_candidate_ord = ord;
-                            }
-                            SkipResult::OverStep => {
-                                candidate = left.doc();
-                                other_candidate_ord = usize::max_value();
-                            }
-                            SkipResult::End => {
-                                return false;
+                if ord != other_candidate_ord {
+                    // `candidate_ord` is already at the
+                    // right position.
+                    //
+                    // Calling `skip_next` would advance this docset
+                    // and miss it.
+                    match docset.skip_next(candidate) {
+                        SkipResult::Reached => {}
+                        SkipResult::OverStep => {
+                            // this is not in the intersection,
+                            // let's update our candidate.
+                            candidate = docset.doc();
+                            match left.skip_next(candidate) {
+                                SkipResult::Reached => {
+                                    other_candidate_ord = ord;
+                                }
+                                SkipResult::OverStep => {
+                                    candidate = left.doc();
+                                    other_candidate_ord = usize::max_value();
+                                }
+                                SkipResult::End => {
+                                    return false;
+                                }
                             }
+                            continue 'outer;
+                        }
+                        SkipResult::End => {
+                            return false;
                         }
-                        continue 'outer;
-                    }
-                    SkipResult::End => {
-                        return false;
                     }
                 }
             }
@@ -214,102 +220,6 @@ impl<TDocSet: DocSet, TOtherDocSet: DocSet> DocSet for Intersection<TDocSet, TOt
     }
 }
 
-// `ahead` is assumed to be initialized (ahead.advance() has been called at least once,
-// and this returned true).
-//
-// If behind is either uninitialized or `ahead.doc() > behind.doc()`.
-fn next_in_intersection<'a, TScorer: Scorer>(
-    ahead: &'a mut TScorer,
-    behind: &'a mut TScorer,
-) -> Option<DocId> {
-    let candidate = ahead.doc();
-    match behind.skip_next(candidate) {
-        SkipResult::Reached => Some(candidate),
-        SkipResult::OverStep => {
-            // yeah for tail-recursion
-            next_in_intersection(behind, ahead)
-        }
-        SkipResult::End => None,
-    }
-}
-
-enum SkipResultComplex {
-    Reached,
-    Overstep { other_ord: usize, candidate: DocId },
-    End,
-}
-
-fn skip_several_scorers<TDocSet: DocSet>(
-    others: &mut [TDocSet],
-    except_candidate_ord: usize,
-    target: DocId,
-) -> SkipResultComplex {
-    for (ord, docset) in others.iter_mut().enumerate() {
-        // `candidate_ord` is already at the
-        // right position.
-        //
-        // Calling `skip_next` would advance this docset
-        // and miss it.
-        if ord == except_candidate_ord {
-            continue;
-        }
-        match docset.skip_next(target) {
-            SkipResult::Reached => {}
-            SkipResult::OverStep => {
-                return SkipResultComplex::Overstep {
-                    other_ord: ord,
-                    candidate: docset.doc(),
-                };
-            }
-            SkipResult::End => {
-                return SkipResultComplex::End;
-            }
-        }
-    }
-    SkipResultComplex::Reached
-}
-
-fn for_each<'a, TScorer: Scorer, TOtherscorer: Scorer>(
-    left: &'a mut TScorer,
-    right: &'a mut TScorer,
-    others: &'a mut [TOtherscorer],
-    callback: &mut FnMut(DocId, Score),
-) {
-    let mut other_candidate_ord: usize = usize::max_value();
-    if !left.advance() {
-        return;
-    }
-    while let Some(candidate) = next_in_intersection(left, right) {
-        // test the remaining scorers
-        match skip_several_scorers(others, other_candidate_ord, candidate) {
-            SkipResultComplex::Reached => {
-                let intersection_score: Score = left.score()
-                    + right.score()
-                    + others.iter_mut().map(|other| other.score()).sum::<Score>();
-                callback(candidate, intersection_score);
-                if !left.advance() {
-                    return;
-                }
-            }
-            SkipResultComplex::Overstep {
-                other_ord,
-                candidate,
-            } => match left.skip_next(candidate) {
-                SkipResult::End => {
-                    return;
-                }
-                SkipResult::Reached => {
-                    other_candidate_ord = other_ord;
-                }
-                SkipResult::OverStep => other_candidate_ord = usize::max_value(),
-            },
-            SkipResultComplex::End => {
-                return;
-            }
-        }
-    }
-}
-
 impl<TScorer, TOtherScorer> Scorer for Intersection<TScorer, TOtherScorer>
 where
     TScorer: Scorer,
@@ -320,10 +230,6 @@ where
             + self.right.score()
             + self.others.iter_mut().map(Scorer::score).sum::<Score>()
     }
-
-    fn for_each(&mut self, callback: &mut FnMut(DocId, Score)) {
-        for_each(&mut self.left, &mut self.right, &mut self.others, callback);
-    }
 }
 
 #[cfg(test)]

src/query/phrase_query/phrase_query.rs

@@ -4,7 +4,6 @@ use error::TantivyError;
 use query::bm25::BM25Weight;
 use query::Query;
 use query::Weight;
-use schema::IndexRecordOption;
 use schema::{Field, Term};
 use std::collections::BTreeSet;
 use Result;
@@ -84,7 +83,7 @@ impl Query for PhraseQuery {
         let has_positions = field_entry
             .field_type()
             .get_index_record_option()
-            .map(IndexRecordOption::has_positions)
+            .map(|index_record_option| index_record_option.has_positions())
             .unwrap_or(false);
         if !has_positions {
             let field_name = field_entry.name();

src/query/query_parser/query_grammar.rs

@@ -1,7 +1,6 @@
 #![cfg_attr(feature = "cargo-clippy", allow(clippy::unneeded_field_pattern))]
 #![cfg_attr(feature = "cargo-clippy", allow(clippy::toplevel_ref_arg))]
 
-use super::query_grammar;
 use super::user_input_ast::*;
 use combine::char::*;
 use combine::error::StreamError;
@@ -23,7 +22,7 @@ parser! {
 parser! {
     fn word[I]()(I) -> String
     where [I: Stream<Item = char>] {
-        many1(satisfy(char::is_alphanumeric))
+        many1(satisfy(|c: char| c.is_alphanumeric()))
                .and_then(|s: String| {
                    match s.as_str() {
                      "OR" => Err(StreamErrorFor::<I>::unexpected_static_message("OR")),
@@ -63,7 +62,7 @@ parser! {
     fn negative_number[I]()(I) -> String
     where [I: Stream<Item = char>]
     {
-            (char('-'), many1(satisfy(char::is_numeric)))
+            (char('-'), many1(satisfy(|c: char| c.is_numeric())))
                 .map(|(s1, s2): (char, String)| format!("{}{}", s1, s2))
     }
 }
@@ -185,7 +184,7 @@ parser! {
                         }
                     )
                 )
-                .map(query_grammar::Element::into_dnf)
+                .map(|el| el.into_dnf())
                 .map(|fnd| {
                     if fnd.len() == 1 {
                         UserInputAST::and(fnd.into_iter().next().unwrap()) //< safe

src/query/scorer.rs

@@ -16,9 +16,6 @@ pub trait Scorer: downcast_rs::Downcast + DocSet + 'static {
 
     /// Iterates through all of the document matched by the DocSet
     /// `DocSet` and push the scored documents to the collector.
-    ///
-    /// This method assumes that the Scorer is brand new, and `.advance()`
-    /// and `.skip()` haven't been called yet.
     fn for_each(&mut self, callback: &mut FnMut(DocId, Score)) {
         while self.advance() {
             callback(self.doc(), self.score());

src/query/term_query/mod.rs

@@ -98,20 +98,4 @@ mod tests {
         }
     }
 
-    #[test]
-    fn test_term_query_count_when_there_are_deletes() {
-        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 = index.writer_with_num_threads(1, 5_000_000).unwrap();
-        index_writer.add_document(doc!(text_field=>"a b"));
-        index_writer.add_document(doc!(text_field=>"a c"));
-        index_writer.delete_term(Term::from_field_text(text_field, "b"));
-        index_writer.commit().unwrap();
-        let term_a = Term::from_field_text(text_field, "a");
-        let term_query = TermQuery::new(term_a, IndexRecordOption::Basic);
-        let reader = index.reader().unwrap();
-        assert_eq!(term_query.count(&*reader.searcher()).unwrap(), 1);
-    }
 }

src/query/term_query/term_weight.rs

@@ -39,15 +39,15 @@ impl Weight for TermWeight {
     }
 
     fn count(&self, reader: &SegmentReader) -> Result<u32> {
-        if let Some(delete_bitset) = reader.delete_bitset() {
-            Ok(self.scorer(reader)?.count(delete_bitset))
-        } else {
+        if reader.num_deleted_docs() == 0 {
             let field = self.term.field();
             Ok(reader
                 .inverted_index(field)
                 .get_term_info(&self.term)
                 .map(|term_info| term_info.doc_freq)
                 .unwrap_or(0))
+        } else {
+            Ok(self.scorer(reader)?.count())
         }
     }
 }

src/query/union.rs

@@ -96,7 +96,7 @@ fn refill<TScorer: Scorer, TScoreCombiner: ScoreCombiner>(
 
 impl<TScorer: Scorer, TScoreCombiner: ScoreCombiner> Union<TScorer, TScoreCombiner> {
     fn refill(&mut self) -> bool {
-        if let Some(min_doc) = self.docsets.iter().map(DocSet::doc).min() {
+        if let Some(min_doc) = self.docsets.iter_mut().map(|docset| docset.doc()).min() {
             self.offset = min_doc;
             self.cursor = 0;
             refill(
@@ -145,7 +145,7 @@ where
         }
     }
 
-    fn count_including_deleted(&mut self) -> u32 {
+    fn count(&mut self) -> u32 {
         let mut count = self.bitsets[self.cursor..HORIZON_NUM_TINYBITSETS]
             .iter()
             .map(|bitset| bitset.len())
@@ -163,8 +163,6 @@ where
         count
     }
 
-    // TODO implement `count` efficiently.
-
     fn skip_next(&mut self, target: DocId) -> SkipResult {
         if !self.advance() {
             return SkipResult::End;
@@ -260,23 +258,6 @@ where
     fn score(&mut self) -> Score {
         self.score
     }
-
-    fn for_each(&mut self, callback: &mut FnMut(DocId, Score)) {
-        // TODO how do we deal with the fact that people may have called .advance() before.
-        while self.refill() {
-            let offset = self.offset;
-            for cursor in 0..HORIZON_NUM_TINYBITSETS {
-                while let Some(val) = self.bitsets[cursor].pop_lowest() {
-                    let delta = val + (cursor as u32) * 64;
-                    let doc = offset + delta;
-                    let score_combiner = &mut self.scores[delta as usize];
-                    let score = score_combiner.score();
-                    callback(doc, score);
-                    score_combiner.clear();
-                }
-            }
-        }
-    }
 }
 
 #[cfg(test)]
@@ -319,7 +300,7 @@ mod tests {
             count += 1;
         }
         assert!(!union_expected.advance());
-        assert_eq!(count, make_union().count_including_deleted());
+        assert_eq!(count, make_union().count());
     }
 
     #[test]

src/query/weight.rs

@@ -13,11 +13,6 @@ pub trait Weight: Send + Sync + 'static {
 
     /// Returns the number documents within the given `SegmentReader`.
     fn count(&self, reader: &SegmentReader) -> Result<u32> {
-        let mut scorer = self.scorer(reader)?;
-        if let Some(delete_bitset) = reader.delete_bitset() {
-            Ok(scorer.count(delete_bitset))
-        } else {
-            Ok(scorer.count_including_deleted())
-        }
+        Ok(self.scorer(reader)?.count())
     }
 }

src/schema/document.rs

@@ -128,7 +128,7 @@ impl Document {
         self.field_values
             .iter()
             .filter(|field_value| field_value.field() == field)
-            .map(FieldValue::value)
+            .map(|field_value| field_value.value())
             .collect()
     }
 
@@ -137,7 +137,7 @@ impl Document {
         self.field_values
             .iter()
             .find(|field_value| field_value.field() == field)
-            .map(FieldValue::value)
+            .map(|field_value| field_value.value())
     }
 }
 

src/schema/field_type.rs

@@ -4,7 +4,6 @@ use schema::{IntOptions, TextOptions};
 
 use schema::Facet;
 use schema::IndexRecordOption;
-use schema::TextFieldIndexing;
 use schema::Value;
 use serde_json::Value as JsonValue;
 
@@ -95,7 +94,7 @@ impl FieldType {
         match *self {
             FieldType::Str(ref text_options) => text_options
                 .get_indexing_options()
-                .map(TextFieldIndexing::index_option),
+                .map(|indexing_options| indexing_options.index_option()),
             FieldType::U64(ref int_options)
             | FieldType::I64(ref int_options)
             | FieldType::Date(ref int_options) => {

src/schema/schema.rs

@@ -130,16 +130,7 @@ impl SchemaBuilder {
         self.add_field(field_entry)
     }
 
-    /// Adds a fast bytes field to the schema.
-    ///
-    /// Bytes field are not searchable and are only used
-    /// as fast field, to associate any kind of payload
-    /// to a document.
-    ///
-    /// For instance, learning-to-rank often requires to access
-    /// some document features at scoring time.
-    /// These can be serializing and stored as a bytes field to
-    /// get access rapidly when scoring each document.
+    /// Adds a fast bytes field to the schema
     pub fn add_bytes_field(&mut self, field_name: &str) -> Field {
         let field_entry = FieldEntry::new_bytes(field_name.to_string());
         self.add_field(field_entry)
@@ -233,7 +224,7 @@ impl Schema {
             let field_name = self.get_field_name(field);
             let values: Vec<Value> = field_values
                 .into_iter()
-                .map(FieldValue::value)
+                .map(|field_val| field_val.value())
                 .cloned()
                 .collect();
             field_map.insert(field_name.to_string(), values);

src/snippet/mod.rs

@@ -1,7 +1,6 @@
 use htmlescape::encode_minimal;
 use query::Query;
 use schema::Field;
-use schema::Value;
 use std::cmp::Ordering;
 use std::collections::BTreeMap;
 use std::collections::BTreeSet;
@@ -304,7 +303,7 @@ impl SnippetGenerator {
         let text: String = doc
             .get_all(self.field)
             .into_iter()
-            .flat_map(Value::text)
+            .flat_map(|val| val.text())
             .collect::<Vec<&str>>()
             .join(" ");
         self.snippet(&text)

src/space_usage/mod.rs

@@ -227,7 +227,7 @@ pub struct PerFieldSpaceUsage {
 
 impl PerFieldSpaceUsage {
     pub(crate) fn new(fields: HashMap<Field, FieldUsage>) -> PerFieldSpaceUsage {
-        let total = fields.values().map(FieldUsage::total).sum();
+        let total = fields.values().map(|x| x.total()).sum();
         PerFieldSpaceUsage { fields, total }
     }
 

src/tokenizer/lower_caser.rs

@@ -44,17 +44,18 @@ where
     }
 
     fn advance(&mut self) -> bool {
-        if !self.tail.advance() {
-            return false;
-        }
-        if self.token_mut().text.is_ascii() {
-            // fast track for ascii.
-            self.token_mut().text.make_ascii_lowercase();
+        if self.tail.advance() {
+            if self.token_mut().text.is_ascii() {
+                // fast track for ascii.
+                self.token_mut().text.make_ascii_lowercase();
+            } else {
+                to_lowercase_unicode(&mut self.tail.token_mut().text, &mut self.buffer);
+                mem::swap(&mut self.tail.token_mut().text, &mut self.buffer);
+            }
+            true
         } else {
-            to_lowercase_unicode(&mut self.tail.token_mut().text, &mut self.buffer);
-            mem::swap(&mut self.tail.token_mut().text, &mut self.buffer);
+            false
         }
-        true
     }
 }
 

src/tokenizer/mod.rs

@@ -97,8 +97,6 @@
 //! If you built your schema programmatically, a complete example
 //! could like this for instance.
 //!
-//! Note that tokens with a len greater or equal to [`MAX_TOKEN_LEN`](./constant.MAX_TOKEN_LEN.html).
-//!
 //! # Example
 //!
 //! ```
@@ -131,7 +129,6 @@
 //! ```
 //!
 mod alphanum_only;
-mod ascii_folding_filter;
 mod facet_tokenizer;
 mod lower_caser;
 mod ngram_tokenizer;
@@ -145,7 +142,6 @@ mod tokenizer;
 mod tokenizer_manager;
 
 pub use self::alphanum_only::AlphaNumOnlyFilter;
-pub use self::ascii_folding_filter::AsciiFoldingFilter;
 pub use self::facet_tokenizer::FacetTokenizer;
 pub use self::lower_caser::LowerCaser;
 pub use self::ngram_tokenizer::NgramTokenizer;
@@ -161,13 +157,6 @@ pub use self::tokenizer::BoxedTokenizer;
 pub use self::tokenizer::{Token, TokenFilter, TokenStream, Tokenizer};
 pub use self::tokenizer_manager::TokenizerManager;
 
-/// Maximum authorized len (in bytes) for a token.
-///
-/// Tokenizer are in charge of not emitting tokens larger than this value.
-/// Currently, if a faulty tokenizer implementation emits tokens with a length larger than
-/// `2^16 - 1 - 4`, the token will simply be ignored downstream.
-pub const MAX_TOKEN_LEN: usize = u16::max_value() as usize - 4;
-
 #[cfg(test)]
 pub mod tests {
     use super::{

src/tokenizer/raw_tokenizer.rs

@@ -29,9 +29,12 @@ impl<'a> Tokenizer<'a> for RawTokenizer {
 
 impl TokenStream for RawTokenStream {
     fn advance(&mut self) -> bool {
-        let result = self.has_token;
-        self.has_token = false;
-        result
+        if self.has_token {
+            self.has_token = false;
+            true
+        } else {
+            false
+        }
     }
 
     fn token(&self) -> &Token {

src/tokenizer/remove_long.rs

@@ -91,6 +91,7 @@ where
                 return true;
             }
         }
+
         false
     }
 }

src/tokenizer/simple_tokenizer.rs

@@ -38,16 +38,23 @@ impl<'a> TokenStream for SimpleTokenStream<'a> {
     fn advance(&mut self) -> bool {
         self.token.text.clear();
         self.token.position = self.token.position.wrapping_add(1);
-        while let Some((offset_from, c)) = self.chars.next() {
-            if c.is_alphanumeric() {
-                let offset_to = self.search_token_end();
-                self.token.offset_from = offset_from;
-                self.token.offset_to = offset_to;
-                self.token.text.push_str(&self.text[offset_from..offset_to]);
-                return true;
+
+        loop {
+            match self.chars.next() {
+                Some((offset_from, c)) => {
+                    if c.is_alphanumeric() {
+                        let offset_to = self.search_token_end();
+                        self.token.offset_from = offset_from;
+                        self.token.offset_to = offset_to;
+                        self.token.text.push_str(&self.text[offset_from..offset_to]);
+                        return true;
+                    }
+                }
+                None => {
+                    return false;
+                }
             }
         }
-        false
     }
 
     fn token(&self) -> &Token {

src/tokenizer/stemmer.rs

@@ -108,14 +108,15 @@ where
     }
 
     fn advance(&mut self) -> bool {
-        if !self.tail.advance() {
-            return false;
+        if self.tail.advance() {
+            // TODO remove allocation
+            let stemmed_str: String = self.stemmer.stem(&self.token().text).into_owned();
+            self.token_mut().text.clear();
+            self.token_mut().text.push_str(&stemmed_str);
+            true
+        } else {
+            false
         }
-        // TODO remove allocation
-        let stemmed_str: String = self.stemmer.stem(&self.token().text).into_owned();
-        self.token_mut().text.clear();
-        self.token_mut().text.push_str(&stemmed_str);
-        true
     }
 }
 

src/tokenizer/stop_word_filter.rs

@@ -104,6 +104,7 @@ where
                 return true;
             }
         }
+
         false
     }
 }

src/tokenizer/tokenizer_manager.rs

@@ -1,5 +1,4 @@
 use std::collections::HashMap;
-use std::ops::Deref;
 use std::sync::{Arc, RwLock};
 use tokenizer::box_tokenizer;
 use tokenizer::stemmer::Language;
@@ -47,8 +46,7 @@ impl TokenizerManager {
             .read()
             .expect("Acquiring the lock should never fail")
             .get(tokenizer_name)
-            .map(Deref::deref)
-            .map(BoxedTokenizer::boxed_clone)
+            .map(|boxed_tokenizer| boxed_tokenizer.boxed_clone())
     }
 }