fixing unit test

* Moving lock to directory/

* added fs2

* doc

* Using fs2 for locking

* Added unit test

* Fixed error message related unit test

* Fixing location of import

CHANGELOG.md

@@ -1,5 +1,20 @@
+Tantivy 0.9.0
+=====================
+- Removed most unsafe (@fulmicoton)
+- Indexer memory footprint improved. (VInt comp, inlining the first block. (@fulmicoton)
+- Stemming in other language possible (@pentlander)
+- Segments with no docs are deleted earlier (@barrotsteindev)
+
 Tantivy 0.8.1
 =====================
+Hotfix of #476.
+
+Merge was reflecting deletes before commit was passed. 
+Thanks @barrotsteindev  for reporting the bug.
+
+
+Tantivy 0.8.0
+=====================
 *No change in the index format*
 - API Breaking change in the collector API. (@jwolfe, @fulmicoton)
 - Multithreaded search (@jwolfe, @fulmicoton) 

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "tantivy"
-version = "0.8.0-dev"
+version = "0.9.0-dev"
 authors = ["Paul Masurel <paul.masurel@gmail.com>"]
 license = "MIT"
 categories = ["database-implementations", "data-structures"]
@@ -29,7 +29,8 @@ serde = "1.0"
 serde_derive = "1.0"
 serde_json = "1.0"
 num_cpus = "1.2"
-itertools = "0.7"
+fs2={version="0.4", optional=true}
+itertools = "0.8"
 levenshtein_automata = {version="0.1", features=["fst_automaton"]}
 bit-set = "0.5"
 uuid = { version = "0.7", features = ["v4", "serde"] }
@@ -39,16 +40,17 @@ futures-cpupool = "0.1"
 owning_ref = "0.4"
 stable_deref_trait = "1.0.0"
 rust-stemmers = "1"
-downcast = { version="0.9" }
+downcast-rs = { version="1.0" }
 matches = "0.1"
 bitpacking = "0.5"
-census = "0.1"
+census = "0.2"
 fnv = "1.0.6"
 owned-read = "0.4"
 failure = "0.1"
 htmlescape = "0.3.1"
 fail = "0.2"
 scoped-pool = "1.0"
+murmurhash32 = "0.2"
 
 [target.'cfg(windows)'.dependencies]
 winapi = "0.2"
@@ -69,7 +71,7 @@ overflow-checks = true
 [features]
 # by default no-fail is disabled. We manually enable it when running test.
 default = ["mmap", "no_fail"]
-mmap = ["fst/mmap", "atomicwrites"]
+mmap = ["fst/mmap", "atomicwrites", "fs2"]
 lz4-compression = ["lz4"]
 no_fail = ["fail/no_fail"]
 unstable = [] # useful for benches.

README.md

@@ -21,7 +21,7 @@
 
 **Tantivy** is a **full text search engine library** written in rust.
 
-It is closer to [Apache Lucene](https://lucene.apache.org/) than to [Elastic Search](https://www.elastic.co/products/elasticsearch) and [Apache Solr](https://lucene.apache.org/solr/) in the sense it is not
+It is closer to [Apache Lucene](https://lucene.apache.org/) than to [Elasticsearch](https://www.elastic.co/products/elasticsearch) and [Apache Solr](https://lucene.apache.org/solr/) in the sense it is not
 an off-the-shelf search engine server, but rather a crate that can be used
 to build such a search engine.
 
@@ -76,7 +76,7 @@ It will walk you through getting a wikipedia search engine up and running in a f
 Tantivy compiles on stable rust but requires `Rust >= 1.27`.
 To check out and run tests, you can simply run :
 
-    git clone git@github.com:tantivy-search/tantivy.git
+    git clone https://github.com/tantivy-search/tantivy.git
     cd tantivy
     cargo build
 

examples/custom_collector.rs

@@ -70,8 +70,6 @@ impl Collector for StatsCollector {
     // Our standard deviation will be a float.
     type Fruit = Option<Stats>;
 
-    type SegmentFruit = Self::Fruit;
-
     type Child = StatsSegmentCollector;
 
     fn for_segment(

src/collector/count_collector.rs

@@ -58,7 +58,6 @@ pub struct Count;
 
 impl Collector for Count {
     type Fruit = usize;
-    type SegmentFruit = usize;
 
     type Child = SegmentCountCollector;
 

src/collector/facet_collector.rs

@@ -197,7 +197,7 @@ fn skip<'a, I: Iterator<Item = &'a Facet>>(
 ) -> SkipResult {
     loop {
         match collapse_it.peek() {
-            Some(facet_bytes) => match facet_bytes.encoded_bytes().cmp(target) {
+            Some(facet_bytes) => match facet_bytes.encoded_str().as_bytes().cmp(target) {
                 Ordering::Less => {}
                 Ordering::Greater => {
                     return SkipResult::OverStep;
@@ -258,8 +258,6 @@ impl FacetCollector {
 impl Collector for FacetCollector {
     type Fruit = FacetCounts;
 
-    type SegmentFruit = FacetCounts;
-
     type Child = FacetSegmentCollector;
 
     fn for_segment(
@@ -371,7 +369,8 @@ impl SegmentCollector for FacetSegmentCollector {
             let mut facet = vec![];
             let facet_ord = self.collapse_facet_ords[collapsed_facet_ord];
             facet_dict.ord_to_term(facet_ord as u64, &mut facet);
-            facet_counts.insert(unsafe { Facet::from_encoded(facet) }, count);
+            // TODO
+            facet_counts.insert(Facet::from_encoded(facet).unwrap(), count);
         }
         FacetCounts { facet_counts }
     }
@@ -405,9 +404,9 @@ impl FacetCounts {
         let right_bound = if facet.is_root() {
             Bound::Unbounded
         } else {
-            let mut facet_after_bytes: Vec<u8> = facet.encoded_bytes().to_owned();
-            facet_after_bytes.push(1u8);
-            let facet_after = unsafe { Facet::from_encoded(facet_after_bytes) }; // ok logic
+            let mut facet_after_bytes: String = facet.encoded_str().to_owned();
+            facet_after_bytes.push('\u{1}');
+            let facet_after = Facet::from_encoded_string(facet_after_bytes);
             Bound::Excluded(facet_after)
         };
         let underlying: btree_map::Range<_, _> = self.facet_counts.range((left_bound, right_bound));

src/collector/int_facet_collector.rs

@@ -88,7 +88,7 @@ mod tests {
         let index = Index::create_in_ram(schema.clone());
 
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 for i in 0u64..10u64 {
                     index_writer.add_document(doc!(

src/collector/mod.rs

@@ -85,7 +85,7 @@ See the `custom_collector` example.
 
 */
 
-use downcast;
+use downcast_rs;
 use DocId;
 use Result;
 use Score;
@@ -111,9 +111,9 @@ pub use self::facet_collector::FacetCollector;
 
 /// `Fruit` is the type for the result of our collection.
 /// e.g. `usize` for the `Count` collector.
-pub trait Fruit: Send + downcast::Any {}
+pub trait Fruit: Send + downcast_rs::Downcast {}
 
-impl<T> Fruit for T where T: Send + downcast::Any {}
+impl<T> Fruit for T where T: Send + downcast_rs::Downcast {}
 
 /// Collectors are in charge of collecting and retaining relevant
 /// information from the document found and scored by the query.
@@ -136,10 +136,8 @@ pub trait Collector: Sync {
     /// e.g. `usize` for the `Count` collector.
     type Fruit: Fruit;
 
-    type SegmentFruit: Fruit;
-
     /// Type of the `SegmentCollector` associated to this collector.
-    type Child: SegmentCollector<Fruit = Self::SegmentFruit>;
+    type Child: SegmentCollector<Fruit = Self::Fruit>;
 
     /// `set_segment` is called before beginning to enumerate
     /// on this segment.
@@ -154,7 +152,7 @@ pub trait Collector: Sync {
 
     /// Combines the fruit associated to the collection of each segments
     /// into one fruit.
-    fn merge_fruits(&self, segment_fruits: Vec<Self::SegmentFruit>) -> Result<Self::Fruit>;
+    fn merge_fruits(&self, segment_fruits: Vec<Self::Fruit>) -> Result<Self::Fruit>;
 }
 
 /// The `SegmentCollector` is the trait in charge of defining the
@@ -183,9 +181,6 @@ where
     Right: Collector,
 {
     type Fruit = (Left::Fruit, Right::Fruit);
-
-    type SegmentFruit = (Left::SegmentFruit, Right::SegmentFruit);
-
     type Child = (Left::Child, Right::Child);
 
     fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
@@ -200,7 +195,7 @@ where
 
     fn merge_fruits(
         &self,
-        children: Vec<(Left::SegmentFruit, Right::SegmentFruit)>,
+        children: Vec<(Left::Fruit, Right::Fruit)>,
     ) -> Result<(Left::Fruit, Right::Fruit)> {
         let mut left_fruits = vec![];
         let mut right_fruits = vec![];
@@ -241,7 +236,6 @@ where
     Three: Collector,
 {
     type Fruit = (One::Fruit, Two::Fruit, Three::Fruit);
-    type SegmentFruit = (One::SegmentFruit, Two::SegmentFruit, Three::SegmentFruit);
     type Child = (One::Child, Two::Child, Three::Child);
 
     fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
@@ -255,7 +249,7 @@ where
         self.0.requires_scoring() || self.1.requires_scoring() || self.2.requires_scoring()
     }
 
-    fn merge_fruits(&self, children: Vec<Self::SegmentFruit>) -> Result<Self::Fruit> {
+    fn merge_fruits(&self, children: Vec<Self::Fruit>) -> Result<Self::Fruit> {
         let mut one_fruits = vec![];
         let mut two_fruits = vec![];
         let mut three_fruits = vec![];
@@ -301,7 +295,6 @@ where
     Four: Collector,
 {
     type Fruit = (One::Fruit, Two::Fruit, Three::Fruit, Four::Fruit);
-    type SegmentFruit = (One::SegmentFruit, Two::SegmentFruit, Three::SegmentFruit, Four::SegmentFruit);
     type Child = (One::Child, Two::Child, Three::Child, Four::Child);
 
     fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
@@ -319,7 +312,7 @@ where
             || self.3.requires_scoring()
     }
 
-    fn merge_fruits(&self, children: Vec<Self::SegmentFruit>) -> Result<Self::Fruit> {
+    fn merge_fruits(&self, children: Vec<Self::Fruit>) -> Result<Self::Fruit> {
         let mut one_fruits = vec![];
         let mut two_fruits = vec![];
         let mut three_fruits = vec![];
@@ -365,10 +358,7 @@ where
     }
 }
 
-#[allow(missing_docs)]
-mod downcast_impl {
-    downcast!(super::Fruit);
-}
+impl_downcast!(Fruit);
 
 #[cfg(test)]
 pub mod tests;

src/collector/multi_collector.rs

@@ -1,7 +1,6 @@
 use super::Collector;
 use super::SegmentCollector;
 use collector::Fruit;
-use downcast::Downcast;
 use std::marker::PhantomData;
 use DocId;
 use Result;
@@ -18,7 +17,6 @@ pub struct CollectorWrapper<TCollector: Collector>(TCollector);
 
 impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
     type Fruit = Box<Fruit>;
-    type SegmentFruit = Box<Fruit>;
     type Child = Box<BoxableSegmentCollector>;
 
     fn for_segment(
@@ -35,14 +33,13 @@ impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
     }
 
     fn merge_fruits(&self, children: Vec<<Self as Collector>::Fruit>) -> Result<Box<Fruit>> {
-        let typed_fruit: Vec<TCollector::SegmentFruit> = children
+        let typed_fruit: Vec<TCollector::Fruit> = children
             .into_iter()
             .map(|untyped_fruit| {
-                Downcast::<TCollector::SegmentFruit>::downcast(untyped_fruit)
+                untyped_fruit.downcast::<TCollector::Fruit>()
                     .map(|boxed_but_typed| *boxed_but_typed)
-                    .map_err(|e| {
-                        let err_msg = format!("Failed to cast child collector fruit. {:?}", e);
-                        TantivyError::InvalidArgument(err_msg)
+                    .map_err(|_| {
+                        TantivyError::InvalidArgument("Failed to cast child fruit.".to_string())
                     })
             })
             .collect::<Result<_>>()?;
@@ -90,14 +87,17 @@ pub struct FruitHandle<TFruit: Fruit> {
 impl<TFruit: Fruit> FruitHandle<TFruit> {
     pub fn extract(self, fruits: &mut MultiFruit) -> TFruit {
         let boxed_fruit = fruits.sub_fruits[self.pos].take().expect("");
-        *Downcast::<TFruit>::downcast(boxed_fruit).expect("Failed")
+        *boxed_fruit.downcast::<TFruit>().map_err(|_| ()).expect("Failed to downcast collector fruit.")
     }
 }
 
 /// Multicollector makes it possible to collect on more than one collector.
 /// It should only be used for use cases where the Collector types is unknown
 /// at compile time.
-/// If the type of the collectors is known, you should prefer to use `ChainedCollector`.
+///
+/// If the type of the collectors is known, you can just group yours collectors
+/// in a tuple. See the
+/// [Combining several collectors section of the collector documentation](./index.html#combining-several-collectors).
 ///
 /// ```rust
 /// #[macro_use]
@@ -153,7 +153,7 @@ impl<TFruit: Fruit> FruitHandle<TFruit> {
 #[derive(Default)]
 pub struct MultiCollector<'a> {
     collector_wrappers:
-        Vec<Box<Collector<Child = Box<BoxableSegmentCollector>, Fruit = Box<Fruit>, SegmentFruit = Box<Fruit>> + 'a>>,
+        Vec<Box<Collector<Child = Box<BoxableSegmentCollector>, Fruit = Box<Fruit>> + 'a>>,
 }
 
 impl<'a> MultiCollector<'a> {
@@ -178,9 +178,7 @@ impl<'a> MultiCollector<'a> {
 }
 
 impl<'a> Collector for MultiCollector<'a> {
-
     type Fruit = MultiFruit;
-    type SegmentFruit = MultiFruit;
     type Child = MultiCollectorChild;
 
     fn for_segment(

src/collector/tests.rs

@@ -40,7 +40,6 @@ impl TestFruit {
 
 impl Collector for TestCollector {
     type Fruit = TestFruit;
-    type SegmentFruit = Self::Fruit;
     type Child = TestSegmentCollector;
 
     fn for_segment(
@@ -110,8 +109,6 @@ impl FastFieldTestCollector {
 
 impl Collector for FastFieldTestCollector {
     type Fruit = Vec<u64>;
-    type SegmentFruit = Self::Fruit;
-
     type Child = FastFieldSegmentCollector;
 
     fn for_segment(
@@ -168,7 +165,6 @@ impl BytesFastFieldTestCollector {
 
 impl Collector for BytesFastFieldTestCollector {
     type Fruit = Vec<u8>;
-    type SegmentFruit = Self::Fruit;
     type Child = BytesFastFieldSegmentCollector;
 
     fn for_segment(

src/collector/top_field_collector.rs

@@ -88,7 +88,6 @@ impl<T: FastValue + PartialOrd + Clone> TopDocsByField<T> {
 
 impl<T: FastValue + PartialOrd + Send + Sync + 'static> Collector for TopDocsByField<T> {
     type Fruit = Vec<(T, DocAddress)>;
-    type SegmentFruit = Vec<(T, DocAddress)>;
 
     type Child = TopFieldSegmentCollector<T>;
 

src/collector/top_score_collector.rs

@@ -89,7 +89,6 @@ impl TopDocs {
 
 impl Collector for TopDocs {
     type Fruit = Vec<(Score, DocAddress)>;
-    type SegmentFruit = Vec<(Score, DocAddress)>;
 
     type Child = TopScoreSegmentCollector;
 
@@ -143,7 +142,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"Hello happy tax payer."));
             index_writer.add_document(doc!(text_field=>"Droopy says hello happy tax payer"));
             index_writer.add_document(doc!(text_field=>"I like Droopy"));

src/common/bitpacker.rs

@@ -1,9 +1,6 @@
-use common::serialize::BinarySerializable;
+use byteorder::{ByteOrder, LittleEndian, WriteBytesExt};
 use std::io;
-use std::io::Write;
-use std::mem;
 use std::ops::Deref;
-use std::ptr;
 
 pub(crate) struct BitPacker {
     mini_buffer: u64,
@@ -18,7 +15,7 @@ impl BitPacker {
         }
     }
 
-    pub fn write<TWrite: Write>(
+    pub fn write<TWrite: io::Write>(
         &mut self,
         val: u64,
         num_bits: u8,
@@ -28,14 +25,14 @@ impl BitPacker {
         let num_bits = num_bits as usize;
         if self.mini_buffer_written + num_bits > 64 {
             self.mini_buffer |= val_u64.wrapping_shl(self.mini_buffer_written as u32);
-            self.mini_buffer.serialize(output)?;
+            output.write_u64::<LittleEndian>(self.mini_buffer)?;
             self.mini_buffer = val_u64.wrapping_shr((64 - self.mini_buffer_written) as u32);
             self.mini_buffer_written = self.mini_buffer_written + num_bits - 64;
         } else {
             self.mini_buffer |= val_u64 << self.mini_buffer_written;
             self.mini_buffer_written += num_bits;
             if self.mini_buffer_written == 64 {
-                self.mini_buffer.serialize(output)?;
+                output.write_u64::<LittleEndian>(self.mini_buffer)?;
                 self.mini_buffer_written = 0;
                 self.mini_buffer = 0u64;
             }
@@ -43,17 +40,18 @@ impl BitPacker {
         Ok(())
     }
 
-    pub fn flush<TWrite: Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
+    pub fn flush<TWrite: io::Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
         if self.mini_buffer_written > 0 {
             let num_bytes = (self.mini_buffer_written + 7) / 8;
-            let arr: [u8; 8] = unsafe { mem::transmute::<u64, [u8; 8]>(self.mini_buffer.to_le()) };
+            let mut arr: [u8; 8] = [0u8; 8];
+            LittleEndian::write_u64(&mut arr, self.mini_buffer);
             output.write_all(&arr[..num_bytes])?;
             self.mini_buffer_written = 0;
         }
         Ok(())
     }
 
-    pub fn close<TWrite: Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
+    pub fn close<TWrite: io::Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
         self.flush(output)?;
         // Padding the write file to simplify reads.
         output.write_all(&[0u8; 7])?;
@@ -102,39 +100,10 @@ where
             addr + 8 <= data.len(),
             "The fast field field should have been padded with 7 bytes."
         );
-        #[cfg_attr(feature = "cargo-clippy", allow(clippy::cast_ptr_alignment))]
-        let val_unshifted_unmasked: u64 =
-            u64::from_le(unsafe { ptr::read_unaligned(data[addr..].as_ptr() as *const u64) });
+        let val_unshifted_unmasked: u64 = LittleEndian::read_u64(&data[addr..]);
         let val_shifted = (val_unshifted_unmasked >> bit_shift) as u64;
         val_shifted & mask
     }
-
-    /// Reads a range of values from the fast field.
-    ///
-    /// The range of values read is from
-    /// `[start..start + output.len()[`
-    pub fn get_range(&self, start: u32, output: &mut [u64]) {
-        if self.num_bits == 0 {
-            for val in output.iter_mut() {
-                *val = 0u64;
-            }
-        } else {
-            let data: &[u8] = &*self.data;
-            let num_bits = self.num_bits;
-            let mask = self.mask;
-            let mut addr_in_bits = (start as usize) * num_bits;
-            for output_val in output.iter_mut() {
-                let addr = addr_in_bits >> 3;
-                let bit_shift = addr_in_bits & 7;
-                #[cfg_attr(feature = "cargo-clippy", allow(clippy::cast_ptr_alignment))]
-                let val_unshifted_unmasked: u64 =
-                    unsafe { ptr::read_unaligned(data[addr..].as_ptr() as *const u64) };
-                let val_shifted = (val_unshifted_unmasked >> bit_shift) as u64;
-                *output_val = val_shifted & mask;
-                addr_in_bits += num_bits;
-            }
-        }
-    }
 }
 
 #[cfg(test)]
@@ -172,17 +141,4 @@ mod test {
         test_bitpacker_util(6, 14);
         test_bitpacker_util(1000, 14);
     }
-
-    #[test]
-    fn test_bitpacker_range() {
-        let (bitunpacker, vals) = create_fastfield_bitpacker(100_000, 12);
-        let buffer_len = 100;
-        let mut buffer = vec![0u64; buffer_len];
-        for start in vec![0, 10, 20, 100, 1_000] {
-            bitunpacker.get_range(start as u32, &mut buffer[..]);
-            for i in 0..buffer_len {
-                assert_eq!(buffer[i], vals[start + i]);
-            }
-        }
-    }
 }

src/common/mod.rs

@@ -10,7 +10,7 @@ pub(crate) use self::bitset::TinySet;
 pub(crate) use self::composite_file::{CompositeFile, CompositeWrite};
 pub use self::counting_writer::CountingWriter;
 pub use self::serialize::{BinarySerializable, FixedSize};
-pub use self::vint::VInt;
+pub use self::vint::{read_u32_vint, serialize_vint_u32, write_u32_vint, VInt};
 pub use byteorder::LittleEndian as Endianness;
 
 use std::io;

src/common/vint.rs

@@ -1,4 +1,5 @@
 use super::BinarySerializable;
+use byteorder::{ByteOrder, LittleEndian};
 use std::io;
 use std::io::Read;
 use std::io::Write;
@@ -9,6 +10,100 @@ pub struct VInt(pub u64);
 
 const STOP_BIT: u8 = 128;
 
+pub fn serialize_vint_u32(val: u32) -> (u64, usize) {
+    const START_2: u64 = 1 << 7;
+    const START_3: u64 = 1 << 14;
+    const START_4: u64 = 1 << 21;
+    const START_5: u64 = 1 << 28;
+
+    const STOP_1: u64 = START_2 - 1;
+    const STOP_2: u64 = START_3 - 1;
+    const STOP_3: u64 = START_4 - 1;
+    const STOP_4: u64 = START_5 - 1;
+
+    const MASK_1: u64 = 127;
+    const MASK_2: u64 = MASK_1 << 7;
+    const MASK_3: u64 = MASK_2 << 7;
+    const MASK_4: u64 = MASK_3 << 7;
+    const MASK_5: u64 = MASK_4 << 7;
+
+    let val = u64::from(val);
+    const STOP_BIT: u64 = 128u64;
+    match val {
+        0...STOP_1 => (val | STOP_BIT, 1),
+        START_2...STOP_2 => (
+            (val & MASK_1) | ((val & MASK_2) << 1) | (STOP_BIT << (8)),
+            2,
+        ),
+        START_3...STOP_3 => (
+            (val & MASK_1) | ((val & MASK_2) << 1) | ((val & MASK_3) << 2) | (STOP_BIT << (8 * 2)),
+            3,
+        ),
+        START_4...STOP_4 => (
+            (val & MASK_1)
+                | ((val & MASK_2) << 1)
+                | ((val & MASK_3) << 2)
+                | ((val & MASK_4) << 3)
+                | (STOP_BIT << (8 * 3)),
+            4,
+        ),
+        _ => (
+            (val & MASK_1)
+                | ((val & MASK_2) << 1)
+                | ((val & MASK_3) << 2)
+                | ((val & MASK_4) << 3)
+                | ((val & MASK_5) << 4)
+                | (STOP_BIT << (8 * 4)),
+            5,
+        ),
+    }
+}
+
+/// Returns the number of bytes covered by a
+/// serialized vint `u32`.
+///
+/// Expects a buffer data that starts
+/// by the serialized `vint`, scans at most 5 bytes ahead until
+/// it finds the vint final byte.
+///
+/// # May Panic
+/// If the payload does not start by a valid `vint`
+fn vint_len(data: &[u8]) -> usize {
+    for (i, &val) in data.iter().enumerate().take(5) {
+        if val >= STOP_BIT {
+            return i + 1;
+        }
+    }
+    panic!("Corrupted data. Invalid VInt 32");
+}
+
+/// Reads a vint `u32` from a buffer, and
+/// consumes its payload data.
+///
+/// # Panics
+///
+/// If the buffer does not start by a valid
+/// vint payload
+pub fn read_u32_vint(data: &mut &[u8]) -> u32 {
+    let vlen = vint_len(*data);
+    let mut result = 0u32;
+    let mut shift = 0u64;
+    for &b in &data[..vlen] {
+        result |= u32::from(b & 127u8) << shift;
+        shift += 7;
+    }
+    *data = &data[vlen..];
+    result
+}
+
+/// Write a `u32` as a vint payload.
+pub fn write_u32_vint<W: io::Write>(val: u32, writer: &mut W) -> io::Result<()> {
+    let (val, num_bytes) = serialize_vint_u32(val);
+    let mut buffer = [0u8; 8];
+    LittleEndian::write_u64(&mut buffer, val);
+    writer.write_all(&buffer[..num_bytes])
+}
+
 impl VInt {
     pub fn val(&self) -> u64 {
         self.0
@@ -24,7 +119,7 @@ impl VInt {
         output.extend(&buffer[0..num_bytes]);
     }
 
-    fn serialize_into(&self, buffer: &mut [u8; 10]) -> usize {
+    pub fn serialize_into(&self, buffer: &mut [u8; 10]) -> usize {
         let mut remaining = self.0;
         for (i, b) in buffer.iter_mut().enumerate() {
             let next_byte: u8 = (remaining % 128u64) as u8;
@@ -64,7 +159,7 @@ impl BinarySerializable for VInt {
                     return Err(io::Error::new(
                         io::ErrorKind::InvalidData,
                         "Reach end of buffer while reading VInt",
-                    ))
+                    ));
                 }
             }
         }
@@ -74,7 +169,9 @@ impl BinarySerializable for VInt {
 #[cfg(test)]
 mod tests {
 
+    use super::serialize_vint_u32;
     use super::VInt;
+    use byteorder::{ByteOrder, LittleEndian};
     use common::BinarySerializable;
 
     fn aux_test_vint(val: u64) {
@@ -108,4 +205,28 @@ mod tests {
         }
         aux_test_vint(10);
     }
+
+    fn aux_test_serialize_vint_u32(val: u32) {
+        let mut buffer = [0u8; 10];
+        let mut buffer2 = [0u8; 10];
+        let len_vint = VInt(val as u64).serialize_into(&mut buffer);
+        let (vint, len) = serialize_vint_u32(val);
+        assert_eq!(len, len_vint, "len wrong for val {}", val);
+        LittleEndian::write_u64(&mut buffer2, vint);
+        assert_eq!(&buffer[..len], &buffer2[..len], "array wrong for {}", val);
+    }
+
+    #[test]
+    fn test_vint_u32() {
+        aux_test_serialize_vint_u32(0);
+        aux_test_serialize_vint_u32(1);
+        aux_test_serialize_vint_u32(5);
+        for i in 1..3 {
+            let power_of_128 = 1u32 << (7 * i);
+            aux_test_serialize_vint_u32(power_of_128 - 1u32);
+            aux_test_serialize_vint_u32(power_of_128);
+            aux_test_serialize_vint_u32(power_of_128 + 1u32);
+        }
+        aux_test_serialize_vint_u32(u32::max_value());
+    }
 }

src/core/executor.rs

@@ -64,17 +64,18 @@ impl Executor {
                     // This is important as it makes it possible for the fruit_receiver iteration to
                     // terminate.
                 };
-                let mut results = Vec::with_capacity(num_fruits);
-                unsafe { results.set_len(num_fruits) };
-                let mut num_items = 0;
+                // This is lame, but safe.
+                let mut results_with_position = Vec::with_capacity(num_fruits);
                 for (pos, fruit_res) in fruit_receiver {
-                    results[pos] = fruit_res?;
-                    num_items += 1;
+                    let fruit = fruit_res?;
+                    results_with_position.push((pos, fruit));
                 }
-                // this checks ensures that we filled of this
-                // uninitialized memory.
-                assert_eq!(num_items, results.len());
-                Ok(results)
+                results_with_position.sort_by_key(|(pos, _)| *pos);
+                assert_eq!(results_with_position.len(), num_fruits);
+                Ok(results_with_position
+                    .into_iter()
+                    .map(|(_, fruit)| fruit)
+                    .collect::<Vec<_>>())
             }
         }
     }

src/core/index.rs

@@ -12,12 +12,14 @@ use core::META_FILEPATH;
 use directory::ManagedDirectory;
 #[cfg(feature = "mmap")]
 use directory::MmapDirectory;
+use directory::INDEX_WRITER_LOCK;
+use directory::META_LOCK;
 use directory::{Directory, RAMDirectory};
+use error::DataCorruption;
 use error::TantivyError;
 use indexer::index_writer::open_index_writer;
 use indexer::index_writer::HEAP_SIZE_MIN;
 use indexer::segment_updater::save_new_metas;
-use indexer::LockType;
 use num_cpus;
 use schema::Field;
 use schema::FieldType;
@@ -37,7 +39,13 @@ fn load_metas(directory: &Directory) -> Result<IndexMeta> {
     let meta_data = directory.atomic_read(&META_FILEPATH)?;
     let meta_string = String::from_utf8_lossy(&meta_data);
     serde_json::from_str(&meta_string)
-        .map_err(|_| TantivyError::CorruptedFile(META_FILEPATH.clone()))
+        .map_err(|e| {
+            DataCorruption::new(
+                META_FILEPATH.clone(),
+                format!("Meta file cannot be deserialized. {:?}.", e),
+            )
+        })
+        .map_err(From::from)
 }
 
 /// Search Index
@@ -143,7 +151,7 @@ impl Index {
     ///
     /// This will overwrite existing meta.json
     fn from_directory(mut directory: ManagedDirectory, schema: Schema) -> Result<Index> {
-        save_new_metas(schema.clone(), 0, directory.borrow_mut())?;
+        save_new_metas(schema.clone(), directory.borrow_mut())?;
         let metas = IndexMeta::with_schema(schema);
         Index::create_from_metas(directory, &metas)
     }
@@ -225,7 +233,8 @@ impl Index {
     /// Each thread will receive a budget of  `overall_heap_size_in_bytes / num_threads`.
     ///
     /// # Errors
-    /// If the lockfile already exists, returns `Error::FileAlreadyExists`.
+    /// If the lockfile already exists, returns `Error::DirectoryLockBusy` or an `Error::IOError`.
+    ///
     /// # Panics
     /// If the heap size per thread is too small, panics.
     pub fn writer_with_num_threads(
@@ -233,7 +242,21 @@ impl Index {
         num_threads: usize,
         overall_heap_size_in_bytes: usize,
     ) -> Result<IndexWriter> {
-        let directory_lock = LockType::IndexWriterLock.acquire_lock(&self.directory)?;
+        let directory_lock = self
+            .directory
+            .acquire_lock(&INDEX_WRITER_LOCK)
+            .map_err(|err| {
+                TantivyError::LockFailure(
+                    err,
+                    Some(
+                        "Failed to acquire index lock. If you are using\
+                         a regular directory, this means there is already an \
+                         `IndexWriter` working on this `Directory`, in this process \
+                         or in a different process."
+                            .to_string(),
+                    ),
+                )
+            })?;
         let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
         open_index_writer(
             self,
@@ -332,7 +355,7 @@ impl Index {
     /// get the freshest `index` at all time, is to watch `meta.json` and
     /// call `load_searchers` whenever a changes happen.
     pub fn load_searchers(&self) -> Result<()> {
-        let _meta_lock = LockType::MetaLock.acquire_lock(self.directory())?;
+        let _meta_lock = self.directory().acquire_lock(&META_LOCK)?;
         let searchable_segments = self.searchable_segments()?;
         let segment_readers: Vec<SegmentReader> = searchable_segments
             .iter()

src/core/searcher.rs

@@ -23,7 +23,7 @@ fn collect_segment<C: Collector>(
     weight: &Weight,
     segment_ord: u32,
     segment_reader: &SegmentReader,
-) -> Result<C::SegmentFruit> {
+) -> Result<C::Fruit> {
     let mut scorer = weight.scorer(segment_reader)?;
     let mut segment_collector = collector.for_segment(segment_ord as u32, segment_reader)?;
     if let Some(delete_bitset) = segment_reader.delete_bitset() {

src/core/segment_component.rs

@@ -41,6 +41,6 @@ impl SegmentComponent {
             SegmentComponent::STORE,
             SegmentComponent::DELETE,
         ];
-        SEGMENT_COMPONENTS.into_iter()
+        SEGMENT_COMPONENTS.iter()
     }
 }

src/directory/directory.rs

@@ -1,11 +1,102 @@
+use directory::directory_lock::Lock;
+use directory::error::LockError;
 use directory::error::{DeleteError, OpenReadError, OpenWriteError};
 use directory::{ReadOnlySource, WritePtr};
 use std::fmt;
 use std::io;
+use std::io::Write;
 use std::marker::Send;
 use std::marker::Sync;
 use std::path::Path;
+use std::path::PathBuf;
 use std::result;
+use std::thread;
+use std::time::Duration;
+
+/// Retry the logic of acquiring locks is pretty simple.
+/// We just retry `n` times after a given `duratio`, both
+/// depending on the type of lock.
+struct RetryPolicy {
+    num_retries: usize,
+    wait_in_ms: u64,
+}
+
+impl RetryPolicy {
+    fn no_retry() -> RetryPolicy {
+        RetryPolicy {
+            num_retries: 0,
+            wait_in_ms: 0,
+        }
+    }
+
+    fn wait_and_retry(&mut self) -> bool {
+        if self.num_retries == 0 {
+            false
+        } else {
+            self.num_retries -= 1;
+            let wait_duration = Duration::from_millis(self.wait_in_ms);
+            thread::sleep(wait_duration);
+            true
+        }
+    }
+}
+
+/// The `DirectoryLock` is an object that represents a file lock.
+/// See  [`LockType`](struct.LockType.html)
+///
+/// It is transparently associated to a lock file, that gets deleted
+/// on `Drop.` The lock is released automatically on `Drop`.
+pub struct DirectoryLock(Box<Drop + Send + 'static>);
+
+struct DirectoryLockGuard {
+    directory: Box<Directory>,
+    path: PathBuf,
+}
+
+impl<T: Drop + Send + 'static> From<Box<T>> for DirectoryLock {
+    fn from(underlying: Box<T>) -> Self {
+        DirectoryLock(underlying)
+    }
+}
+
+impl Drop for DirectoryLockGuard {
+    fn drop(&mut self) {
+        if let Err(e) = self.directory.delete(&*self.path) {
+            error!("Failed to remove the lock file. {:?}", e);
+        }
+    }
+}
+
+enum TryAcquireLockError {
+    FileExists,
+    IOError(io::Error),
+}
+
+fn try_acquire_lock(
+    filepath: &Path,
+    directory: &mut Directory,
+) -> Result<DirectoryLock, TryAcquireLockError> {
+    let mut write = directory.open_write(filepath).map_err(|e| match e {
+        OpenWriteError::FileAlreadyExists(_) => TryAcquireLockError::FileExists,
+        OpenWriteError::IOError(io_error) => TryAcquireLockError::IOError(io_error.into()),
+    })?;
+    write.flush().map_err(TryAcquireLockError::IOError)?;
+    Ok(DirectoryLock::from(Box::new(DirectoryLockGuard {
+        directory: directory.box_clone(),
+        path: filepath.to_owned(),
+    })))
+}
+
+fn retry_policy(is_blocking: bool) -> RetryPolicy {
+    if is_blocking {
+        RetryPolicy {
+            num_retries: 100,
+            wait_in_ms: 100,
+        }
+    } else {
+        RetryPolicy::no_retry()
+    }
+}
 
 /// Write-once read many (WORM) abstraction for where
 /// tantivy's data should be stored.
@@ -73,6 +164,29 @@ pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
     ///
     /// The file may or may not previously exist.
     fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()>;
+
+    /// Acquire a lock in the given directory.
+    ///
+    /// The method is blocking or not depending on the `Lock` object.
+    fn acquire_lock(&self, lock: &Lock) -> Result<DirectoryLock, LockError> {
+        let mut box_directory = self.box_clone();
+        let mut retry_policy = retry_policy(lock.is_blocking);
+        loop {
+            match try_acquire_lock(&lock.filepath, &mut *box_directory) {
+                Ok(result) => {
+                    return Ok(result);
+                }
+                Err(TryAcquireLockError::FileExists) => {
+                    if !retry_policy.wait_and_retry() {
+                        return Err(LockError::LockBusy);
+                    }
+                }
+                Err(TryAcquireLockError::IOError(io_error)) => {
+                    return Err(LockError::IOError(io_error));
+                }
+            }
+        }
+    }
 }
 
 /// DirectoryClone

src/directory/directory_lock.rs

@@ -0,0 +1,56 @@
+use std::path::PathBuf;
+
+/// A directory lock.
+///
+/// A lock is associated to a specific path and some
+/// [`LockParams`](./enum.LockParams.html).
+/// Tantivy itself uses only two locks but client application
+/// can use the directory facility to define their own locks.
+/// - [INDEX_WRITER_LOCK](./struct.INDEX_WRITER_LOCK.html)
+/// - [META_LOCK](./struct.META_LOCK.html)
+///
+/// Check out these locks documentation for more information.
+///
+#[derive(Debug)]
+pub struct Lock {
+    /// The lock needs to be associated with its own file `path`.
+    /// Depending on the platform, the lock might rely on the creation
+    /// and deletion of this filepath.
+    pub filepath: PathBuf,
+    /// `lock_params` describes whether acquiring the lock is meant
+    /// to be a blocking operation or a non-blocking.
+    ///
+    /// Acquiring a blocking lock blocks until the lock is
+    /// available.
+    /// Acquiring a blocking lock returns rapidly, either successfully
+    /// or with an error signifying that someone is already holding
+    /// the lock.
+    pub is_blocking: bool,
+}
+
+lazy_static! {
+     /// Only one process should be able to write tantivy's index at a time.
+    /// This lock file, when present, is in charge of preventing other processes to open an IndexWriter.
+    ///
+    /// If the process is killed and this file remains, it is safe to remove it manually.
+    ///
+    /// Failing to acquire this lock usually means a misuse of tantivy's API,
+    /// (creating more than one instance of the `IndexWriter`), are a spurious
+    /// lock file remaining after a crash. In the latter case, removing the file after
+    /// checking no process running tantivy is running is safe.
+    pub static ref INDEX_WRITER_LOCK: Lock = Lock {
+        filepath: PathBuf::from(".tantivy-writer.lock"),
+        is_blocking: false
+    };
+    /// The meta lock file is here to protect the segment files being opened by
+    /// `.load_searchers()` from being garbage collected.
+    /// It makes it possible for another process to safely consume
+    /// our index in-writing. Ideally, we may have prefered `RWLock` semantics
+    /// here, but it is difficult to achieve on Windows.
+    ///
+    /// Opening segment readers is a very fast process.
+    pub static ref META_LOCK: Lock = Lock {
+        filepath: PathBuf::from(".tantivy-meta.lock"),
+        is_blocking: true
+    };
+}

src/directory/error.rs

@@ -3,6 +3,22 @@ use std::fmt;
 use std::io;
 use std::path::PathBuf;
 
+/// Error while trying to acquire a directory lock.
+#[derive(Debug, Fail)]
+pub enum LockError {
+    /// Failed to acquired a lock as it is already hold by another
+    /// client.
+    /// - In the context of a blocking lock, this means the lock was not released within some `timeout` period.
+    /// - In the context of a non-blocking lock, this means the lock was busy at the moment of the call.
+    #[fail(
+        display = "Could not acquire lock as it is already held, possibly by a different process."
+    )]
+    LockBusy,
+    /// Trying to acquire a lock failed with an `IOError`
+    #[fail(display = "Failed to acquire the lock due to an io:Error.")]
+    IOError(io::Error),
+}
+
 /// General IO error with an optional path to the offending file.
 #[derive(Debug)]
 pub struct IOError {
@@ -10,6 +26,12 @@ pub struct IOError {
     err: io::Error,
 }
 
+impl Into<io::Error> for IOError {
+    fn into(self) -> io::Error {
+        self.err
+    }
+}
+
 impl fmt::Display for IOError {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
         match self.path {

src/directory/managed_directory.rs

@@ -1,8 +1,10 @@
 use core::MANAGED_FILEPATH;
-use directory::error::{DeleteError, IOError, OpenReadError, OpenWriteError};
+use directory::error::{DeleteError, IOError, LockError, OpenReadError, OpenWriteError};
+use directory::DirectoryLock;
+use directory::Lock;
+use directory::META_LOCK;
 use directory::{ReadOnlySource, WritePtr};
-use error::TantivyError;
-use indexer::LockType;
+use error::DataCorruption;
 use serde_json;
 use std::collections::HashSet;
 use std::io;
@@ -64,7 +66,12 @@ impl ManagedDirectory {
             Ok(data) => {
                 let managed_files_json = String::from_utf8_lossy(&data);
                 let managed_files: HashSet<PathBuf> = serde_json::from_str(&managed_files_json)
-                    .map_err(|_| TantivyError::CorruptedFile(MANAGED_FILEPATH.clone()))?;
+                    .map_err(|e| {
+                        DataCorruption::new(
+                            MANAGED_FILEPATH.clone(),
+                            format!("Managed file cannot be deserialized: {:?}. ", e),
+                        )
+                    })?;
                 Ok(ManagedDirectory {
                     directory: Box::new(directory),
                     meta_informations: Arc::new(RwLock::new(MetaInformation {
@@ -87,6 +94,9 @@ impl ManagedDirectory {
     ///
     /// * `living_files` - List of files that are still used by the index.
     ///
+    /// The use a callback ensures that the list of living_files is computed
+    /// while we hold the lock on meta.
+    ///
     /// This method does not panick nor returns errors.
     /// If a file cannot be deleted (for permission reasons for instance)
     /// an error is simply logged, and the file remains in the list of managed
@@ -117,7 +127,7 @@ impl ManagedDirectory {
             // 2) writer change meta.json (for instance after a merge or a commit)
             // 3) gc kicks in.
             // 4) gc removes a file that was useful for process B, before process B opened it.
-            if let Ok(_meta_lock) = LockType::MetaLock.acquire_lock(self) {
+            if let Ok(_meta_lock) = self.acquire_lock(&META_LOCK) {
                 let living_files = get_living_files();
                 for managed_path in &meta_informations_rlock.managed_paths {
                     if !living_files.contains(managed_path) {
@@ -227,6 +237,10 @@ impl Directory for ManagedDirectory {
     fn exists(&self, path: &Path) -> bool {
         self.directory.exists(path)
     }
+
+    fn acquire_lock(&self, lock: &Lock) -> result::Result<DirectoryLock, LockError> {
+        self.directory.acquire_lock(lock)
+    }
 }
 
 impl Clone for ManagedDirectory {

src/directory/mmap_directory.rs

@@ -1,8 +1,14 @@
+extern crate fs2;
+
+use self::fs2::FileExt;
 use atomicwrites;
 use common::make_io_err;
+use directory::error::LockError;
 use directory::error::{DeleteError, IOError, OpenDirectoryError, OpenReadError, OpenWriteError};
 use directory::shared_vec_slice::SharedVecSlice;
 use directory::Directory;
+use directory::DirectoryLock;
+use directory::Lock;
 use directory::ReadOnlySource;
 use directory::WritePtr;
 use fst::raw::MmapReadOnly;
@@ -115,6 +121,14 @@ impl MmapCache {
 ///
 /// The Mmap object are cached to limit the
 /// system calls.
+///
+/// In the `MmapDirectory`, locks are implemented using the `fs2` crate definition of locks.
+///
+/// On MacOS & linux, it relies on `flock` (aka `BSD Lock`). These locks solve most of the
+/// problems related to POSIX Locks, but may their contract may not be respected on `NFS`
+/// depending on the implementation.
+///
+/// On Windows the semantics are again different.
 #[derive(Clone)]
 pub struct MmapDirectory {
     root_path: PathBuf,
@@ -213,6 +227,21 @@ impl MmapDirectory {
     }
 }
 
+/// We rely on fs2 for file locking. On Windows & MacOS this
+/// uses BSD locks (`flock`). The lock is actually released when
+/// the `File` object is dropped and its associated file descriptor
+/// is closed.
+struct ReleaseLockFile {
+    _file: File,
+    path: PathBuf,
+}
+
+impl Drop for ReleaseLockFile {
+    fn drop(&mut self) {
+        debug!("Releasing lock {:?}", self.path);
+    }
+}
+
 /// This Write wraps a File, but has the specificity of
 /// call `sync_all` on flush.
 struct SafeFileWriter(File);
@@ -354,6 +383,26 @@ impl Directory for MmapDirectory {
         meta_file.write(|f| f.write_all(data))?;
         Ok(())
     }
+
+    fn acquire_lock(&self, lock: &Lock) -> Result<DirectoryLock, LockError> {
+        let full_path = self.resolve_path(&lock.filepath);
+        // We make sure that the file exists.
+        let file: File = OpenOptions::new()
+            .write(true)
+            .create(true) //< if the file does not exist yet, create it.
+            .open(&full_path)
+            .map_err(LockError::IOError)?;
+        if lock.is_blocking {
+            file.lock_exclusive().map_err(LockError::IOError)?;
+        } else {
+            file.try_lock_exclusive().map_err(|_| LockError::LockBusy)?
+        }
+        // dropping the file handle will release the lock.
+        Ok(DirectoryLock::from(Box::new(ReleaseLockFile {
+            path: lock.filepath.clone(),
+            _file: file,
+        })))
+    }
 }
 
 #[cfg(test)]

src/directory/mod.rs

@@ -8,6 +8,7 @@ WORM directory abstraction.
 mod mmap_directory;
 
 mod directory;
+mod directory_lock;
 mod managed_directory;
 mod ram_directory;
 mod read_only_source;
@@ -16,11 +17,12 @@ mod shared_vec_slice;
 /// Errors specific to the directory module.
 pub mod error;
 
-use std::io::{BufWriter, Seek, Write};
-
+pub use self::directory::DirectoryLock;
 pub use self::directory::{Directory, DirectoryClone};
+pub use self::directory_lock::{Lock, INDEX_WRITER_LOCK, META_LOCK};
 pub use self::ram_directory::RAMDirectory;
 pub use self::read_only_source::ReadOnlySource;
+use std::io::{BufWriter, Seek, Write};
 
 #[cfg(feature = "mmap")]
 pub use self::mmap_directory::MmapDirectory;
@@ -38,128 +40,4 @@ impl<T: Seek + Write> SeekableWrite for T {}
 pub type WritePtr = BufWriter<Box<SeekableWrite>>;
 
 #[cfg(test)]
-mod tests {
-
-    use super::*;
-    use std::io::{Seek, SeekFrom, Write};
-    use std::path::Path;
-
-    lazy_static! {
-        static ref TEST_PATH: &'static Path = Path::new("some_path_for_test");
-    }
-
-    #[test]
-    fn test_ram_directory() {
-        let mut ram_directory = RAMDirectory::create();
-        test_directory(&mut ram_directory);
-    }
-
-    #[test]
-    #[cfg(feature = "mmap")]
-    fn test_mmap_directory() {
-        let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
-        test_directory(&mut mmap_directory);
-    }
-
-    #[test]
-    #[should_panic]
-    fn ram_directory_panics_if_flush_forgotten() {
-        let mut ram_directory = RAMDirectory::create();
-        let mut write_file = ram_directory.open_write(*TEST_PATH).unwrap();
-        assert!(write_file.write_all(&[4]).is_ok());
-    }
-
-    fn test_simple(directory: &mut Directory) {
-        {
-            {
-                let mut write_file = directory.open_write(*TEST_PATH).unwrap();
-                assert!(directory.exists(*TEST_PATH));
-                write_file.write_all(&[4]).unwrap();
-                write_file.write_all(&[3]).unwrap();
-                write_file.write_all(&[7, 3, 5]).unwrap();
-                write_file.flush().unwrap();
-            }
-            let read_file = directory.open_read(*TEST_PATH).unwrap();
-            let data: &[u8] = &*read_file;
-            assert_eq!(data, &[4u8, 3u8, 7u8, 3u8, 5u8]);
-        }
-
-        assert!(directory.delete(*TEST_PATH).is_ok());
-        assert!(!directory.exists(*TEST_PATH));
-    }
-
-    fn test_seek(directory: &mut Directory) {
-        {
-            {
-                let mut write_file = directory.open_write(*TEST_PATH).unwrap();
-                write_file.write_all(&[4, 3, 7, 3, 5]).unwrap();
-                write_file.seek(SeekFrom::Start(0)).unwrap();
-                write_file.write_all(&[3, 1]).unwrap();
-                write_file.flush().unwrap();
-            }
-            let read_file = directory.open_read(*TEST_PATH).unwrap();
-            let data: &[u8] = &*read_file;
-            assert_eq!(data, &[3u8, 1u8, 7u8, 3u8, 5u8]);
-        }
-
-        assert!(directory.delete(*TEST_PATH).is_ok());
-    }
-
-    fn test_rewrite_forbidden(directory: &mut Directory) {
-        {
-            directory.open_write(*TEST_PATH).unwrap();
-            assert!(directory.exists(*TEST_PATH));
-        }
-        {
-            assert!(directory.open_write(*TEST_PATH).is_err());
-        }
-        assert!(directory.delete(*TEST_PATH).is_ok());
-    }
-
-    fn test_write_create_the_file(directory: &mut Directory) {
-        {
-            assert!(directory.open_read(*TEST_PATH).is_err());
-            let _w = directory.open_write(*TEST_PATH).unwrap();
-            assert!(directory.exists(*TEST_PATH));
-            assert!(directory.open_read(*TEST_PATH).is_ok());
-            assert!(directory.delete(*TEST_PATH).is_ok());
-        }
-    }
-
-    fn test_directory_delete(directory: &mut Directory) {
-        assert!(directory.open_read(*TEST_PATH).is_err());
-        let mut write_file = directory.open_write(*TEST_PATH).unwrap();
-        write_file.write_all(&[1, 2, 3, 4]).unwrap();
-        write_file.flush().unwrap();
-        {
-            let read_handle = directory.open_read(*TEST_PATH).unwrap();
-            {
-                assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
-
-                // Mapped files can't be deleted on Windows
-                if !cfg!(windows) {
-                    assert!(directory.delete(*TEST_PATH).is_ok());
-                    assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
-                }
-
-                assert!(directory.delete(Path::new("SomeOtherPath")).is_err());
-            }
-        }
-
-        if cfg!(windows) {
-            assert!(directory.delete(*TEST_PATH).is_ok());
-        }
-
-        assert!(directory.open_read(*TEST_PATH).is_err());
-        assert!(directory.delete(*TEST_PATH).is_err());
-    }
-
-    fn test_directory(directory: &mut Directory) {
-        test_simple(directory);
-        test_seek(directory);
-        test_rewrite_forbidden(directory);
-        test_write_create_the_file(directory);
-        test_directory_delete(directory);
-    }
-
-}
+mod tests;

src/directory/tests.rs

@@ -0,0 +1,182 @@
+use super::*;
+use std::io::{Seek, SeekFrom, Write};
+use std::path::{Path, PathBuf};
+use std::time;
+
+lazy_static! {
+    static ref TEST_PATH: &'static Path = Path::new("some_path_for_test");
+}
+
+#[test]
+fn test_ram_directory() {
+    let mut ram_directory = RAMDirectory::create();
+    test_directory(&mut ram_directory);
+}
+
+#[test]
+#[cfg(feature = "mmap")]
+fn test_mmap_directory() {
+    let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
+    test_directory(&mut mmap_directory);
+}
+
+#[test]
+#[should_panic]
+fn ram_directory_panics_if_flush_forgotten() {
+    let mut ram_directory = RAMDirectory::create();
+    let mut write_file = ram_directory.open_write(*TEST_PATH).unwrap();
+    assert!(write_file.write_all(&[4]).is_ok());
+}
+
+fn test_simple(directory: &mut Directory) {
+    {
+        {
+            let mut write_file = directory.open_write(*TEST_PATH).unwrap();
+            assert!(directory.exists(*TEST_PATH));
+            write_file.write_all(&[4]).unwrap();
+            write_file.write_all(&[3]).unwrap();
+            write_file.write_all(&[7, 3, 5]).unwrap();
+            write_file.flush().unwrap();
+        }
+        let read_file = directory.open_read(*TEST_PATH).unwrap();
+        let data: &[u8] = &*read_file;
+        assert_eq!(data, &[4u8, 3u8, 7u8, 3u8, 5u8]);
+    }
+
+    assert!(directory.delete(*TEST_PATH).is_ok());
+    assert!(!directory.exists(*TEST_PATH));
+}
+
+fn test_seek(directory: &mut Directory) {
+    {
+        {
+            let mut write_file = directory.open_write(*TEST_PATH).unwrap();
+            write_file.write_all(&[4, 3, 7, 3, 5]).unwrap();
+            write_file.seek(SeekFrom::Start(0)).unwrap();
+            write_file.write_all(&[3, 1]).unwrap();
+            write_file.flush().unwrap();
+        }
+        let read_file = directory.open_read(*TEST_PATH).unwrap();
+        let data: &[u8] = &*read_file;
+        assert_eq!(data, &[3u8, 1u8, 7u8, 3u8, 5u8]);
+    }
+
+    assert!(directory.delete(*TEST_PATH).is_ok());
+}
+
+fn test_rewrite_forbidden(directory: &mut Directory) {
+    {
+        directory.open_write(*TEST_PATH).unwrap();
+        assert!(directory.exists(*TEST_PATH));
+    }
+    {
+        assert!(directory.open_write(*TEST_PATH).is_err());
+    }
+    assert!(directory.delete(*TEST_PATH).is_ok());
+}
+
+fn test_write_create_the_file(directory: &mut Directory) {
+    {
+        assert!(directory.open_read(*TEST_PATH).is_err());
+        let _w = directory.open_write(*TEST_PATH).unwrap();
+        assert!(directory.exists(*TEST_PATH));
+        assert!(directory.open_read(*TEST_PATH).is_ok());
+        assert!(directory.delete(*TEST_PATH).is_ok());
+    }
+}
+
+fn test_directory_delete(directory: &mut Directory) {
+    assert!(directory.open_read(*TEST_PATH).is_err());
+    let mut write_file = directory.open_write(*TEST_PATH).unwrap();
+    write_file.write_all(&[1, 2, 3, 4]).unwrap();
+    write_file.flush().unwrap();
+    {
+        let read_handle = directory.open_read(*TEST_PATH).unwrap();
+        {
+            assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
+
+            // Mapped files can't be deleted on Windows
+            if !cfg!(windows) {
+                assert!(directory.delete(*TEST_PATH).is_ok());
+                assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
+            }
+
+            assert!(directory.delete(Path::new("SomeOtherPath")).is_err());
+        }
+    }
+
+    if cfg!(windows) {
+        assert!(directory.delete(*TEST_PATH).is_ok());
+    }
+
+    assert!(directory.open_read(*TEST_PATH).is_err());
+    assert!(directory.delete(*TEST_PATH).is_err());
+}
+
+fn test_directory(directory: &mut Directory) {
+    test_simple(directory);
+    test_seek(directory);
+    test_rewrite_forbidden(directory);
+    test_write_create_the_file(directory);
+    test_directory_delete(directory);
+    test_lock_non_blocking(directory);
+    test_lock_blocking(directory);
+}
+
+fn test_lock_non_blocking(directory: &mut Directory) {
+    {
+        let lock_a_res = directory.acquire_lock(&Lock {
+            filepath: PathBuf::from("a.lock"),
+            is_blocking: false,
+        });
+        assert!(lock_a_res.is_ok());
+        let lock_b_res = directory.acquire_lock(&Lock {
+            filepath: PathBuf::from("b.lock"),
+            is_blocking: false,
+        });
+        assert!(lock_b_res.is_ok());
+        let lock_a_res2 = directory.acquire_lock(&Lock {
+            filepath: PathBuf::from("a.lock"),
+            is_blocking: false,
+        });
+        assert!(lock_a_res2.is_err());
+    }
+    let lock_a_res = directory.acquire_lock(&Lock {
+        filepath: PathBuf::from("a.lock"),
+        is_blocking: false,
+    });
+    assert!(lock_a_res.is_ok());
+}
+
+fn test_lock_blocking(directory: &mut Directory) {
+    let lock_a_res = directory.acquire_lock(&Lock {
+        filepath: PathBuf::from("a.lock"),
+        is_blocking: true,
+    });
+    assert!(lock_a_res.is_ok());
+    std::thread::spawn(move || {
+        //< lock_a_res is sent to the thread.
+        std::thread::sleep(time::Duration::from_millis(10));
+        // explicitely droping lock_a_res. It would have been sufficient to just force it
+        // to be part of the move, but the intent seems clearer that way.
+        drop(lock_a_res);
+    });
+    {
+        // A non-blocking call should fail, as the thread is running and holding the lock.
+        let lock_a_res = directory.acquire_lock(&Lock {
+            filepath: PathBuf::from("a.lock"),
+            is_blocking: false,
+        });
+        assert!(lock_a_res.is_err());
+    }
+    {
+        // the blocking call should wait for at least 10ms.
+        let start = time::Instant::now();
+        let lock_a_res = directory.acquire_lock(&Lock {
+            filepath: PathBuf::from("a.lock"),
+            is_blocking: true,
+        });
+        assert!(lock_a_res.is_ok());
+        assert!(start.elapsed().subsec_millis() >= 10);
+    }
+}

src/error.rs

@@ -2,15 +2,48 @@
 
 use std::io;
 
+use directory::error::LockError;
 use directory::error::{IOError, OpenDirectoryError, OpenReadError, OpenWriteError};
 use fastfield::FastFieldNotAvailableError;
-use indexer::LockType;
 use query;
 use schema;
 use serde_json;
+use std::fmt;
 use std::path::PathBuf;
 use std::sync::PoisonError;
 
+pub struct DataCorruption {
+    filepath: Option<PathBuf>,
+    comment: String,
+}
+
+impl DataCorruption {
+    pub fn new(filepath: PathBuf, comment: String) -> DataCorruption {
+        DataCorruption {
+            filepath: Some(filepath),
+            comment,
+        }
+    }
+
+    pub fn comment_only(comment: String) -> DataCorruption {
+        DataCorruption {
+            filepath: None,
+            comment,
+        }
+    }
+}
+
+impl fmt::Debug for DataCorruption {
+    fn fmt(&self, f: &mut fmt::Formatter) -> Result<(), fmt::Error> {
+        write!(f, "Data corruption: ")?;
+        if let Some(ref filepath) = &self.filepath {
+            write!(f, "(in file `{:?}`)", filepath)?;
+        }
+        write!(f, ": {}.", self.comment)?;
+        Ok(())
+    }
+}
+
 /// The library's failure based error enum
 #[derive(Debug, Fail)]
 pub enum TantivyError {
@@ -24,17 +57,14 @@ pub enum TantivyError {
     #[fail(display = "Index already exists")]
     IndexAlreadyExists,
     /// Failed to acquire file lock
-    #[fail(
-        display = "Failed to acquire Lockfile: {:?}. Possible causes: another IndexWriter instance or panic during previous lock drop.",
-        _0
-    )]
-    LockFailure(LockType),
+    #[fail(display = "Failed to acquire Lockfile: {:?}. {:?}", _0, _1)]
+    LockFailure(LockError, Option<String>),
     /// IO Error.
     #[fail(display = "An IO error occurred: '{}'", _0)]
     IOError(#[cause] IOError),
     /// Data corruption.
-    #[fail(display = "File contains corrupted data: '{:?}'", _0)]
-    CorruptedFile(PathBuf),
+    #[fail(display = "{:?}", _0)]
+    DataCorruption(DataCorruption),
     /// A thread holding the locked panicked and poisoned the lock.
     #[fail(display = "A thread holding the locked panicked and poisoned the lock")]
     Poisoned,
@@ -55,12 +85,24 @@ pub enum TantivyError {
     SystemError(String),
 }
 
+impl From<DataCorruption> for TantivyError {
+    fn from(data_corruption: DataCorruption) -> TantivyError {
+        TantivyError::DataCorruption(data_corruption)
+    }
+}
+
 impl From<FastFieldNotAvailableError> for TantivyError {
     fn from(fastfield_error: FastFieldNotAvailableError) -> TantivyError {
         TantivyError::FastFieldError(fastfield_error)
     }
 }
 
+impl From<LockError> for TantivyError {
+    fn from(lock_error: LockError) -> TantivyError {
+        TantivyError::LockFailure(lock_error, None)
+    }
+}
+
 impl From<IOError> for TantivyError {
     fn from(io_error: IOError) -> TantivyError {
         TantivyError::IOError(io_error)

src/fastfield/facet_reader.rs

@@ -1,5 +1,6 @@
 use super::MultiValueIntFastFieldReader;
 use schema::Facet;
+use std::str;
 use termdict::TermDictionary;
 use termdict::TermOrdinal;
 use DocId;
@@ -20,6 +21,7 @@ use DocId;
 pub struct FacetReader {
     term_ords: MultiValueIntFastFieldReader<u64>,
     term_dict: TermDictionary,
+    buffer: Vec<u8>,
 }
 
 impl FacetReader {
@@ -37,6 +39,7 @@ impl FacetReader {
         FacetReader {
             term_ords,
             term_dict,
+            buffer: vec![],
         }
     }
 
@@ -55,11 +58,18 @@ impl FacetReader {
     }
 
     /// Given a term ordinal returns the term associated to it.
-    pub fn facet_from_ord(&self, facet_ord: TermOrdinal, output: &mut Facet) {
+    pub fn facet_from_ord(
+        &mut self,
+        facet_ord: TermOrdinal,
+        output: &mut Facet,
+    ) -> Result<(), str::Utf8Error> {
         let found_term = self
             .term_dict
-            .ord_to_term(facet_ord as u64, output.inner_buffer_mut());
+            .ord_to_term(facet_ord as u64, &mut self.buffer);
         assert!(found_term, "Term ordinal {} no found.", facet_ord);
+        let facet_str = str::from_utf8(&self.buffer[..])?;
+        output.set_facet_str(facet_str);
+        Ok(())
     }
 
     /// Return the list of facet ordinals associated to a document.

src/fastfield/multivalued/reader.rs

@@ -82,20 +82,20 @@ mod tests {
 
         let mut facet = Facet::root();
         {
-            facet_reader.facet_from_ord(1, &mut facet);
+            facet_reader.facet_from_ord(1, &mut facet).unwrap();
             assert_eq!(facet, Facet::from("/category"));
         }
         {
-            facet_reader.facet_from_ord(2, &mut facet);
+            facet_reader.facet_from_ord(2, &mut facet).unwrap();
             assert_eq!(facet, Facet::from("/category/cat1"));
         }
         {
-            facet_reader.facet_from_ord(3, &mut facet);
+            facet_reader.facet_from_ord(3, &mut facet).unwrap();
             assert_eq!(format!("{}", facet), "/category/cat2");
             assert_eq!(facet, Facet::from("/category/cat2"));
         }
         {
-            facet_reader.facet_from_ord(4, &mut facet);
+            facet_reader.facet_from_ord(4, &mut facet).unwrap();
             assert_eq!(facet, Facet::from("/category/cat3"));
         }
 

src/fastfield/reader.rs

@@ -79,11 +79,8 @@ impl<Item: FastValue> FastFieldReader<Item> {
     // TODO change start to `u64`.
     // For multifastfield, start is an index in a second fastfield, not a `DocId`
     pub fn get_range(&self, start: u32, output: &mut [Item]) {
-        // ok: Item is either `u64` or `i64`
-        let output_u64: &mut [u64] = unsafe { &mut *(output as *mut [Item] as *mut [u64]) };
-        self.bit_unpacker.get_range(start, output_u64);
-        for out in output_u64.iter_mut() {
-            *out = Item::from_u64(*out + self.min_value_u64).as_u64();
+        for (i, out) in output.iter_mut().enumerate() {
+            *out = self.get(start + i as u32);
         }
     }
 

src/indexer/directory_lock.rs

@@ -1,131 +0,0 @@
-use directory::error::OpenWriteError;
-use std::io::Write;
-use std::path::{Path, PathBuf};
-use std::thread;
-use std::time::Duration;
-use Directory;
-use TantivyError;
-
-#[derive(Debug, Clone, Copy)]
-pub enum LockType {
-    /// Only one process should be able to write tantivy's index at a time.
-    /// This lock file, when present, is in charge of preventing other processes to open an IndexWriter.
-    ///
-    /// If the process is killed and this file remains, it is safe to remove it manually.
-    ///
-    /// Failing to acquire this lock usually means a misuse of tantivy's API,
-    /// (creating more than one instance of the `IndexWriter`), are a spurious
-    /// lock file remaining after a crash. In the latter case, removing the file after
-    /// checking no process running tantivy is running is safe.
-    IndexWriterLock,
-    /// The meta lock file is here to protect the segment files being opened by
-    /// `.load_searchers()` from being garbage collected.
-    /// It makes it possible for another process to safely consume
-    /// our index in-writing. Ideally, we may have prefered `RWLock` semantics
-    /// here, but it is difficult to achieve on Windows.
-    ///
-    /// Opening segment readers is a very fast process.
-    /// Right now if the lock cannot be acquire on the first attempt, the logic
-    /// is very simplistic. We retry after `100ms` until we effectively
-    /// acquire the lock.
-    /// This lock should not have much contention in normal usage.
-    MetaLock,
-}
-
-/// Retry the logic of acquiring locks is pretty simple.
-/// We just retry `n` times after a given `duratio`, both
-/// depending on the type of lock.
-struct RetryPolicy {
-    num_retries: usize,
-    wait_in_ms: u64,
-}
-
-impl RetryPolicy {
-    fn no_retry() -> RetryPolicy {
-        RetryPolicy {
-            num_retries: 0,
-            wait_in_ms: 0,
-        }
-    }
-
-    fn wait_and_retry(&mut self) -> bool {
-        if self.num_retries == 0 {
-            false
-        } else {
-            self.num_retries -= 1;
-            let wait_duration = Duration::from_millis(self.wait_in_ms);
-            thread::sleep(wait_duration);
-            true
-        }
-    }
-}
-
-impl LockType {
-    fn retry_policy(self) -> RetryPolicy {
-        match self {
-            LockType::IndexWriterLock => RetryPolicy::no_retry(),
-            LockType::MetaLock => RetryPolicy {
-                num_retries: 100,
-                wait_in_ms: 100,
-            },
-        }
-    }
-
-    fn try_acquire_lock(self, directory: &mut Directory) -> Result<DirectoryLock, TantivyError> {
-        let path = self.filename();
-        let mut write = directory.open_write(path).map_err(|e| match e {
-            OpenWriteError::FileAlreadyExists(_) => TantivyError::LockFailure(self),
-            OpenWriteError::IOError(io_error) => TantivyError::IOError(io_error),
-        })?;
-        write.flush()?;
-        Ok(DirectoryLock {
-            directory: directory.box_clone(),
-            path: path.to_owned(),
-        })
-    }
-
-    /// Acquire a lock in the given directory.
-    pub fn acquire_lock(self, directory: &Directory) -> Result<DirectoryLock, TantivyError> {
-        let mut box_directory = directory.box_clone();
-        let mut retry_policy = self.retry_policy();
-        loop {
-            let lock_result = self.try_acquire_lock(&mut *box_directory);
-            match lock_result {
-                Ok(result) => {
-                    return Ok(result);
-                }
-                Err(TantivyError::LockFailure(ref filepath)) => {
-                    if !retry_policy.wait_and_retry() {
-                        return Err(TantivyError::LockFailure(filepath.to_owned()));
-                    }
-                }
-                Err(_) => {}
-            }
-        }
-    }
-
-    fn filename(&self) -> &Path {
-        match *self {
-            LockType::MetaLock => Path::new(".tantivy-meta.lock"),
-            LockType::IndexWriterLock => Path::new(".tantivy-indexer.lock"),
-        }
-    }
-}
-
-/// The `DirectoryLock` is an object that represents a file lock.
-/// See  [`LockType`](struct.LockType.html)
-///
-/// It is transparently associated to a lock file, that gets deleted
-/// on `Drop.` The lock is release automatically on `Drop`.
-pub struct DirectoryLock {
-    directory: Box<Directory>,
-    path: PathBuf,
-}
-
-impl Drop for DirectoryLock {
-    fn drop(&mut self) {
-        if let Err(e) = self.directory.delete(&*self.path) {
-            error!("Failed to remove the lock file. {:?}", e);
-        }
-    }
-}

src/indexer/index_writer.rs

@@ -9,15 +9,15 @@ use core::SegmentId;
 use core::SegmentMeta;
 use core::SegmentReader;
 use crossbeam::channel;
+use directory::DirectoryLock;
 use docset::DocSet;
 use error::TantivyError;
 use fastfield::write_delete_bitset;
-use futures::sync::oneshot::Receiver;
+use futures::{Canceled, Future};
 use indexer::delete_queue::{DeleteCursor, DeleteQueue};
 use indexer::doc_opstamp_mapping::DocToOpstampMapping;
 use indexer::operation::DeleteOperation;
 use indexer::stamper::Stamper;
-use indexer::DirectoryLock;
 use indexer::MergePolicy;
 use indexer::SegmentEntry;
 use indexer::SegmentWriter;
@@ -26,7 +26,7 @@ use schema::Document;
 use schema::IndexRecordOption;
 use schema::Term;
 use std::mem;
-use std::mem::swap;
+use std::sync::Arc;
 use std::thread;
 use std::thread::JoinHandle;
 use Result;
@@ -52,17 +52,19 @@ type DocumentReceiver = channel::Receiver<AddOperation>;
 ///
 /// Returns (the heap size in bytes, the hash table size in number of bits)
 fn initial_table_size(per_thread_memory_budget: usize) -> usize {
+    assert!(per_thread_memory_budget > 1_000);
     let table_size_limit: usize = per_thread_memory_budget / 3;
-    (1..)
+    if let Some(limit) = (1..)
         .take_while(|num_bits: &usize| compute_table_size(*num_bits) < table_size_limit)
         .last()
-        .unwrap_or_else(|| {
-            panic!(
-                "Per thread memory is too small: {}",
-                per_thread_memory_budget
-            )
-        })
-        .min(19) // we cap it at 512K
+    {
+        limit.min(19) // we cap it at 2^19 = 512K.
+    } else {
+        unreachable!(
+            "Per thread memory is too small: {}",
+            per_thread_memory_budget
+        );
+    }
 }
 
 /// `IndexWriter` is the user entry-point to add document to an index.
@@ -302,7 +304,7 @@ fn index_documents(
 
     let last_docstamp: u64 = *(doc_opstamps.last().unwrap());
 
-    let segment_entry: SegmentEntry = if delete_cursor.get().is_some() {
+    let delete_bitset_opt = if delete_cursor.get().is_some() {
         let doc_to_opstamps = DocToOpstampMapping::from(doc_opstamps);
         let segment_reader = SegmentReader::open(segment)?;
         let mut deleted_bitset = BitSet::with_capacity(num_docs as usize);
@@ -313,18 +315,17 @@ fn index_documents(
             &doc_to_opstamps,
             last_docstamp,
         )?;
-        SegmentEntry::new(segment_meta, delete_cursor, {
-            if may_have_deletes {
-                Some(deleted_bitset)
-            } else {
-                None
-            }
-        })
+        if may_have_deletes {
+            Some(deleted_bitset)
+        } else {
+            None
+        }
     } else {
         // if there are no delete operation in the queue, no need
         // to even open the segment.
-        SegmentEntry::new(segment_meta, delete_cursor, None)
+        None
     };
+    let segment_entry = SegmentEntry::new(segment_meta, delete_cursor, delete_bitset_opt);
     Ok(segment_updater.add_segment(generation, segment_entry))
 }
 
@@ -366,13 +367,16 @@ impl IndexWriter {
             .add_segment(self.generation, segment_entry);
     }
 
-    /// *Experimental & Advanced API* Creates a new segment.
-    /// and marks it as currently in write.
+    /// Creates a new segment.
     ///
     /// This method is useful only for users trying to do complex
     /// operations, like converting an index format to another.
+    ///
+    /// It is safe to start writing file associated to the new `Segment`.
+    /// These will not be garbage collected as long as an instance object of
+    /// `SegmentMeta` object associated to the new `Segment` is "alive".
     pub fn new_segment(&self) -> Segment {
-        self.segment_updater.new_segment()
+        self.index.new_segment()
     }
 
     /// Spawns a new worker thread for indexing.
@@ -387,6 +391,7 @@ impl IndexWriter {
         let mut delete_cursor = self.delete_queue.cursor();
 
         let mem_budget = self.heap_size_in_bytes_per_thread;
+        let index = self.index.clone();
         let join_handle: JoinHandle<Result<()>> = thread::Builder::new()
             .name(format!(
                 "thrd-tantivy-index{}-gen{}",
@@ -412,7 +417,7 @@ impl IndexWriter {
                         // was dropped.
                         return Ok(());
                     }
-                    let segment = segment_updater.new_segment();
+                    let segment = index.new_segment();
                     index_documents(
                         mem_budget,
                         &segment,
@@ -429,7 +434,7 @@ impl IndexWriter {
     }
 
     /// Accessor to the merge policy.
-    pub fn get_merge_policy(&self) -> Box<MergePolicy> {
+    pub fn get_merge_policy(&self) -> Arc<Box<MergePolicy>> {
         self.segment_updater.get_merge_policy()
     }
 
@@ -454,7 +459,10 @@ impl IndexWriter {
     /// Merges a given list of segments
     ///
     /// `segment_ids` is required to be non-empty.
-    pub fn merge(&mut self, segment_ids: &[SegmentId]) -> Result<Receiver<SegmentMeta>> {
+    pub fn merge(
+        &mut self,
+        segment_ids: &[SegmentId],
+    ) -> Result<impl Future<Item = SegmentMeta, Error = Canceled>> {
         self.segment_updater.start_merge(segment_ids)
     }
 
@@ -467,11 +475,10 @@ impl IndexWriter {
     ///
     /// Returns the former segment_ready channel.
     fn recreate_document_channel(&mut self) -> DocumentReceiver {
-        let (mut document_sender, mut document_receiver): (DocumentSender, DocumentReceiver) =
+        let (document_sender, document_receiver): (DocumentSender, DocumentReceiver) =
             channel::bounded(PIPELINE_MAX_SIZE_IN_DOCS);
-        swap(&mut self.document_sender, &mut document_sender);
-        swap(&mut self.document_receiver, &mut document_receiver);
-        document_receiver
+        mem::replace(&mut self.document_sender, document_sender);
+        mem::replace(&mut self.document_receiver, document_receiver)
     }
 
     /// Rollback to the last commit
@@ -558,17 +565,12 @@ impl IndexWriter {
         // and recreate a new one channels.
         self.recreate_document_channel();
 
-        let mut former_workers_join_handle = Vec::new();
-        swap(
-            &mut former_workers_join_handle,
-            &mut self.workers_join_handle,
-        );
+        let former_workers_join_handle = mem::replace(&mut self.workers_join_handle, Vec::new());
 
         for worker_handle in former_workers_join_handle {
             let indexing_worker_result = worker_handle
                 .join()
                 .map_err(|e| TantivyError::ErrorInThread(format!("{:?}", e)))?;
-
             indexing_worker_result?;
             // add a new worker for the next generation.
             self.add_indexing_worker()?;
@@ -653,6 +655,7 @@ impl IndexWriter {
 mod tests {
 
     use super::initial_table_size;
+    use directory::error::LockError;
     use error::*;
     use indexer::NoMergePolicy;
     use schema::{self, Document};
@@ -663,10 +666,10 @@ mod tests {
     fn test_lockfile_stops_duplicates() {
         let schema_builder = schema::Schema::builder();
         let index = Index::create_in_ram(schema_builder.build());
-        let _index_writer = index.writer(40_000_000).unwrap();
-        match index.writer(40_000_000) {
-            Err(TantivyError::LockFailure(_)) => {}
-            _ => panic!("Expected FileAlreadyExists error"),
+        let _index_writer = index.writer(3_000_000).unwrap();
+        match index.writer(3_000_000) {
+            Err(TantivyError::LockFailure(LockError::LockBusy, _)) => {}
+            _ => panic!("Expected a `LockFailure` error"),
         }
     }
 
@@ -678,8 +681,7 @@ mod tests {
         match index.writer_with_num_threads(1, 3_000_000) {
             Err(err) => {
                 let err_msg = err.to_string();
-                assert!(err_msg.contains("Lockfile"));
-                assert!(err_msg.contains("Possible causes:"))
+                assert!(err_msg.contains("already an `IndexWriter`"));
             }
             _ => panic!("Expected LockfileAlreadyExists error"),
         }
@@ -689,7 +691,7 @@ mod tests {
     fn test_set_merge_policy() {
         let schema_builder = schema::Schema::builder();
         let index = Index::create_in_ram(schema_builder.build());
-        let index_writer = index.writer(40_000_000).unwrap();
+        let index_writer = index.writer(3_000_000).unwrap();
         assert_eq!(
             format!("{:?}", index_writer.get_merge_policy()),
             "LogMergePolicy { min_merge_size: 8, min_layer_size: 10000, \
@@ -708,11 +710,11 @@ mod tests {
         let schema_builder = schema::Schema::builder();
         let index = Index::create_in_ram(schema_builder.build());
         {
-            let _index_writer = index.writer(40_000_000).unwrap();
+            let _index_writer = index.writer(3_000_000).unwrap();
             // the lock should be released when the
             // index_writer leaves the scope.
         }
-        let _index_writer_two = index.writer(40_000_000).unwrap();
+        let _index_writer_two = index.writer(3_000_000).unwrap();
     }
 
     #[test]
@@ -739,7 +741,7 @@ mod tests {
                 index_writer.add_document(doc!(text_field=>"b"));
                 index_writer.add_document(doc!(text_field=>"c"));
             }
-            assert_eq!(index_writer.commit().unwrap(), 2u64);
+            assert!(index_writer.commit().is_ok());
             index.load_searchers().unwrap();
             assert_eq!(num_docs_containing("a"), 0);
             assert_eq!(num_docs_containing("b"), 1);
@@ -802,7 +804,6 @@ mod tests {
             {
                 let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
                 prepared_commit.set_payload("first commit");
-                assert_eq!(prepared_commit.opstamp(), 100);
                 prepared_commit.commit().expect("commit failed");
             }
             {
@@ -836,7 +837,6 @@ mod tests {
             {
                 let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
                 prepared_commit.set_payload("first commit");
-                assert_eq!(prepared_commit.opstamp(), 100);
                 prepared_commit.abort().expect("commit failed");
             }
             {

src/indexer/merge_operation.rs

@@ -0,0 +1,64 @@
+use census::{Inventory, TrackedObject};
+use std::collections::HashSet;
+use SegmentId;
+
+#[derive(Default)]
+pub struct MergeOperationInventory(Inventory<InnerMergeOperation>);
+
+impl MergeOperationInventory {
+    pub fn segment_in_merge(&self) -> HashSet<SegmentId> {
+        let mut segment_in_merge = HashSet::default();
+        for merge_op in self.0.list() {
+            for &segment_id in &merge_op.segment_ids {
+                segment_in_merge.insert(segment_id);
+            }
+        }
+        segment_in_merge
+    }
+}
+
+/// A `MergeOperation` has two role.
+/// It carries all of the information required to describe a merge :
+/// - `target_opstamp` is the opstamp up to which we want to consume the
+/// delete queue and reflect their deletes.
+/// - `segment_ids` is the list of segment to be merged.
+///
+/// The second role is to ensure keep track of the fact that these
+/// segments are in merge and avoid starting a merge operation that
+/// may conflict with this one.
+///
+/// This works by tracking merge operations. When considering computing
+/// merge candidates, we simply list tracked merge operations and remove
+/// their segments from possible merge candidates.
+pub struct MergeOperation {
+    inner: TrackedObject<InnerMergeOperation>,
+}
+
+struct InnerMergeOperation {
+    target_opstamp: u64,
+    segment_ids: Vec<SegmentId>,
+}
+
+impl MergeOperation {
+    pub fn new(
+        inventory: &MergeOperationInventory,
+        target_opstamp: u64,
+        segment_ids: Vec<SegmentId>,
+    ) -> MergeOperation {
+        let inner_merge_operation = InnerMergeOperation {
+            target_opstamp,
+            segment_ids,
+        };
+        MergeOperation {
+            inner: inventory.0.track(inner_merge_operation),
+        }
+    }
+
+    pub fn target_opstamp(&self) -> u64 {
+        self.inner.target_opstamp
+    }
+
+    pub fn segment_ids(&self) -> &[SegmentId] {
+        &self.inner.segment_ids[..]
+    }
+}

src/indexer/merge_policy.rs

@@ -11,7 +11,7 @@ pub struct MergeCandidate(pub Vec<SegmentId>);
 ///
 /// Every time a the list of segments changes, the segment updater
 /// asks the merge policy if some segments should be merged.
-pub trait MergePolicy: MergePolicyClone + marker::Send + marker::Sync + Debug {
+pub trait MergePolicy: marker::Send + marker::Sync + Debug {
     /// Given the list of segment metas, returns the list of merge candidates.
     ///
     /// This call happens on the segment updater thread, and will block
@@ -19,21 +19,6 @@ pub trait MergePolicy: MergePolicyClone + marker::Send + marker::Sync + Debug {
     fn compute_merge_candidates(&self, segments: &[SegmentMeta]) -> Vec<MergeCandidate>;
 }
 
-/// MergePolicyClone
-pub trait MergePolicyClone {
-    /// Returns a boxed clone of the MergePolicy.
-    fn box_clone(&self) -> Box<MergePolicy>;
-}
-
-impl<T> MergePolicyClone for T
-where
-    T: 'static + MergePolicy + Clone,
-{
-    fn box_clone(&self) -> Box<MergePolicy> {
-        Box::new(self.clone())
-    }
-}
-
 /// Never merge segments.
 #[derive(Debug, Clone)]
 pub struct NoMergePolicy;

src/indexer/merger.rs

@@ -654,6 +654,7 @@ mod tests {
     use schema::IntOptions;
     use schema::Term;
     use schema::TextFieldIndexing;
+    use schema::INT_INDEXED;
     use std::io::Cursor;
     use DocAddress;
     use IndexWriter;
@@ -835,7 +836,7 @@ mod tests {
         let score_field = schema_builder.add_u64_field("score", score_fieldtype);
         let bytes_score_field = schema_builder.add_bytes_field("score_bytes");
         let index = Index::create_in_ram(schema_builder.build());
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
 
         let search_term = |searcher: &Searcher, term: Term| {
             let collector = FastFieldTestCollector::for_field(score_field);
@@ -983,7 +984,7 @@ mod tests {
                 .wait()
                 .expect("Merging failed");
             index.load_searchers().unwrap();
-            let ref searcher = *index.searcher();
+            let searcher = index.searcher();
             assert_eq!(searcher.segment_readers().len(), 1);
             assert_eq!(searcher.num_docs(), 3);
             assert_eq!(searcher.segment_readers()[0].num_docs(), 3);
@@ -1029,7 +1030,7 @@ mod tests {
             index_writer.commit().unwrap();
 
             index.load_searchers().unwrap();
-            let ref searcher = *index.searcher();
+            let searcher = index.searcher();
             assert_eq!(searcher.segment_readers().len(), 1);
             assert_eq!(searcher.num_docs(), 2);
             assert_eq!(searcher.segment_readers()[0].num_docs(), 2);
@@ -1125,18 +1126,15 @@ mod tests {
         {
             // Test removing all docs
             index_writer.delete_term(Term::from_field_text(text_field, "g"));
+            index_writer.commit().unwrap();
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            index_writer
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
             index.load_searchers().unwrap();
 
             let ref searcher = *index.searcher();
-            assert_eq!(searcher.segment_readers().len(), 1);
+            assert!(segment_ids.is_empty());
+            assert!(searcher.segment_readers().is_empty());
             assert_eq!(searcher.num_docs(), 0);
         }
     }
@@ -1147,7 +1145,7 @@ mod tests {
         let facet_field = schema_builder.add_facet_field("facet");
         let index = Index::create_in_ram(schema_builder.build());
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             let index_doc = |index_writer: &mut IndexWriter, doc_facets: &[&str]| {
                 let mut doc = Document::default();
                 for facet in doc_facets {
@@ -1212,7 +1210,7 @@ mod tests {
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer
                 .merge(&segment_ids)
                 .expect("Failed to initiate merge")
@@ -1235,7 +1233,7 @@ mod tests {
 
         // Deleting one term
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             let facet = Facet::from_path(vec!["top", "a", "firstdoc"]);
             let facet_term = Term::from_facet(facet_field, &facet);
             index_writer.delete_term(facet_term);
@@ -1255,6 +1253,34 @@ mod tests {
         }
     }
 
+    #[test]
+    fn test_bug_merge() {
+        let mut schema_builder = schema::Schema::builder();
+        let int_field = schema_builder.add_u64_field("intvals", INT_INDEXED);
+        let index = Index::create_in_ram(schema_builder.build());
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+        index_writer.add_document(doc!(int_field => 1u64));
+        index_writer.commit().expect("commit failed");
+        index_writer.add_document(doc!(int_field => 1u64));
+        index_writer.commit().expect("commit failed");
+        index.load_searchers().unwrap();
+        let searcher = index.searcher();
+        assert_eq!(searcher.num_docs(), 2);
+        index_writer.delete_term(Term::from_field_u64(int_field, 1));
+        let segment_ids = index
+            .searchable_segment_ids()
+            .expect("Searchable segments failed.");
+        index_writer
+            .merge(&segment_ids)
+            .expect("Failed to initiate merge")
+            .wait()
+            .expect("Merging failed");
+        index.load_searchers().unwrap();
+        // commit has not been called yet. The document should still be
+        // there.
+        assert_eq!(index.searcher().num_docs(), 2);
+    }
+
     #[test]
     fn test_merge_multivalued_int_fields_all_deleted() {
         let mut schema_builder = schema::Schema::builder();
@@ -1265,7 +1291,7 @@ mod tests {
         let index = Index::create_in_ram(schema_builder.build());
 
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             let mut doc = Document::default();
             doc.add_u64(int_field, 1);
             index_writer.add_document(doc.clone());
@@ -1273,24 +1299,26 @@ mod tests {
             index_writer.add_document(doc);
             index_writer.commit().expect("commit failed");
             index_writer.delete_term(Term::from_field_u64(int_field, 1));
-            index_writer.commit().expect("commit failed");
-        }
-        index.load_searchers().unwrap();
-        let searcher = index.searcher();
-        assert_eq!(searcher.num_docs(), 0);
-        // Merging the segments
-        {
+
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
             index_writer
                 .merge(&segment_ids)
                 .expect("Failed to initiate merge")
                 .wait()
                 .expect("Merging failed");
+
+            // assert delete has not been committed
+            index.load_searchers().unwrap();
+            let searcher = index.searcher();
+            assert_eq!(searcher.num_docs(), 2);
+
+            index_writer.commit().unwrap();
+
             index_writer.wait_merging_threads().unwrap();
         }
+
         index.load_searchers().unwrap();
         let searcher = index.searcher();
         assert_eq!(searcher.num_docs(), 0);
@@ -1306,7 +1334,7 @@ mod tests {
         let index = Index::create_in_ram(schema_builder.build());
 
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             let index_doc = |index_writer: &mut IndexWriter, int_vals: &[u64]| {
                 let mut doc = Document::default();
                 for &val in int_vals {
@@ -1395,7 +1423,7 @@ mod tests {
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer
                 .merge(&segment_ids)
                 .expect("Failed to initiate merge")

src/indexer/mod.rs

@@ -1,8 +1,9 @@
 pub mod delete_queue;
-mod directory_lock;
+
 mod doc_opstamp_mapping;
 pub mod index_writer;
 mod log_merge_policy;
+mod merge_operation;
 pub mod merge_policy;
 pub mod merger;
 pub mod operation;
@@ -15,14 +16,12 @@ pub mod segment_updater;
 mod segment_writer;
 mod stamper;
 
-pub(crate) use self::directory_lock::DirectoryLock;
-pub use self::directory_lock::LockType;
-
 pub use self::index_writer::IndexWriter;
 pub use self::log_merge_policy::LogMergePolicy;
+pub use self::merge_operation::{MergeOperation, MergeOperationInventory};
 pub use self::merge_policy::{MergeCandidate, MergePolicy, NoMergePolicy};
 pub use self::prepared_commit::PreparedCommit;
-pub use self::segment_entry::{SegmentEntry, SegmentState};
+pub use self::segment_entry::SegmentEntry;
 pub use self::segment_manager::SegmentManager;
 pub use self::segment_serializer::SegmentSerializer;
 pub use self::segment_writer::SegmentWriter;

src/indexer/segment_entry.rs

@@ -4,21 +4,6 @@ use core::SegmentMeta;
 use indexer::delete_queue::DeleteCursor;
 use std::fmt;
 
-#[derive(Clone, Copy, PartialEq, Eq, Debug)]
-pub enum SegmentState {
-    Ready,
-    InMerge,
-}
-
-impl SegmentState {
-    pub fn letter_code(self) -> char {
-        match self {
-            SegmentState::InMerge => 'M',
-            SegmentState::Ready => 'R',
-        }
-    }
-}
-
 /// A segment entry describes the state of
 /// a given segment, at a given instant.
 ///
@@ -35,7 +20,6 @@ impl SegmentState {
 #[derive(Clone)]
 pub struct SegmentEntry {
     meta: SegmentMeta,
-    state: SegmentState,
     delete_bitset: Option<BitSet>,
     delete_cursor: DeleteCursor,
 }
@@ -49,7 +33,6 @@ impl SegmentEntry {
     ) -> SegmentEntry {
         SegmentEntry {
             meta: segment_meta,
-            state: SegmentState::Ready,
             delete_bitset,
             delete_cursor,
         }
@@ -72,14 +55,6 @@ impl SegmentEntry {
         &mut self.delete_cursor
     }
 
-    /// Return the `SegmentEntry`.
-    ///
-    /// The state describes whether the segment is available for
-    /// a merge or not.
-    pub fn state(&self) -> SegmentState {
-        self.state
-    }
-
     /// Returns the segment id.
     pub fn segment_id(&self) -> SegmentId {
         self.meta.id()
@@ -89,33 +64,10 @@ impl SegmentEntry {
     pub fn meta(&self) -> &SegmentMeta {
         &self.meta
     }
-
-    /// Mark the `SegmentEntry` as in merge.
-    ///
-    /// Only segments that are not already
-    /// in a merge are elligible for future merge.
-    pub fn start_merge(&mut self) {
-        self.state = SegmentState::InMerge;
-    }
-
-    /// Cancel a merge
-    ///
-    /// If a merge fails, it is important to switch
-    /// the segment back to a idle state, so that it
-    /// may be elligible for future merges.
-    pub fn cancel_merge(&mut self) {
-        self.state = SegmentState::Ready;
-    }
-
-    /// Returns true iff a segment should
-    /// be considered for a merge.
-    pub fn is_ready(&self) -> bool {
-        self.state == SegmentState::Ready
-    }
 }
 
 impl fmt::Debug for SegmentEntry {
     fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
-        write!(formatter, "SegmentEntry({:?}, {:?})", self.meta, self.state)
+        write!(formatter, "SegmentEntry({:?})", self.meta)
     }
 }

src/indexer/segment_manager.rs

@@ -16,7 +16,6 @@ use Result as TantivyResult;
 struct SegmentRegisters {
     uncommitted: SegmentRegister,
     committed: SegmentRegister,
-    writing: HashSet<SegmentId>,
 }
 
 /// The segment manager stores the list of segments
@@ -41,12 +40,17 @@ impl Debug for SegmentManager {
 }
 
 pub fn get_mergeable_segments(
+    in_merge_segment_ids: &HashSet<SegmentId>,
     segment_manager: &SegmentManager,
 ) -> (Vec<SegmentMeta>, Vec<SegmentMeta>) {
     let registers_lock = segment_manager.read();
     (
-        registers_lock.committed.get_mergeable_segments(),
-        registers_lock.uncommitted.get_mergeable_segments(),
+        registers_lock
+            .committed
+            .get_mergeable_segments(in_merge_segment_ids),
+        registers_lock
+            .uncommitted
+            .get_mergeable_segments(in_merge_segment_ids),
     )
 }
 
@@ -59,7 +63,6 @@ impl SegmentManager {
             registers: RwLock::new(SegmentRegisters {
                 uncommitted: SegmentRegister::default(),
                 committed: SegmentRegister::new(segment_metas, delete_cursor),
-                writing: HashSet::new(),
             }),
         }
     }
@@ -72,12 +75,6 @@ impl SegmentManager {
         segment_entries
     }
 
-    /// Returns the overall number of segments in the `SegmentManager`
-    pub fn num_segments(&self) -> usize {
-        let registers_lock = self.read();
-        registers_lock.committed.len() + registers_lock.uncommitted.len()
-    }
-
     /// List the files that are useful to the index.
     ///
     /// This does not include lock files, or files that are obsolete
@@ -106,6 +103,21 @@ impl SegmentManager {
             .expect("Failed to acquire write lock on SegmentManager.")
     }
 
+    /// Deletes all empty segments
+    fn remove_empty_segments(&self) {
+        let mut registers_lock = self.write();
+        registers_lock
+            .committed
+            .segment_entries()
+            .iter()
+            .filter(|segment| segment.meta().num_docs() == 0)
+            .for_each(|segment| {
+                registers_lock
+                    .committed
+                    .remove_segment(&segment.segment_id())
+            });
+    }
+
     pub fn commit(&self, segment_entries: Vec<SegmentEntry>) {
         let mut registers_lock = self.write();
         registers_lock.committed.clear();
@@ -121,25 +133,22 @@ impl SegmentManager {
     /// the `segment_ids` are not either all committed or all
     /// uncommitted.
     pub fn start_merge(&self, segment_ids: &[SegmentId]) -> TantivyResult<Vec<SegmentEntry>> {
-        let mut registers_lock = self.write();
+        let registers_lock = self.read();
         let mut segment_entries = vec![];
         if registers_lock.uncommitted.contains_all(segment_ids) {
             for segment_id in segment_ids {
                 let segment_entry = registers_lock.uncommitted
-                    .start_merge(segment_id)
+                    .get(segment_id)
                     .expect("Segment id not found {}. Should never happen because of the contains all if-block.");
                 segment_entries.push(segment_entry);
             }
         } else if registers_lock.committed.contains_all(segment_ids) {
             for segment_id in segment_ids {
                 let segment_entry = registers_lock.committed
-                    .start_merge(segment_id)
+                    .get(segment_id)
                     .expect("Segment id not found {}. Should never happen because of the contains all if-block.");
                 segment_entries.push(segment_entry);
             }
-            for segment_id in segment_ids {
-                registers_lock.committed.start_merge(segment_id);
-            }
         } else {
             let error_msg = "Merge operation sent for segments that are not \
                              all uncommited or commited."
@@ -149,50 +158,8 @@ impl SegmentManager {
         Ok(segment_entries)
     }
 
-    pub fn cancel_merge(
-        &self,
-        before_merge_segment_ids: &[SegmentId],
-        after_merge_segment_id: SegmentId,
-    ) {
-        let mut registers_lock = self.write();
-
-        // we mark all segments are ready for merge.
-        {
-            let target_segment_register: &mut SegmentRegister;
-            target_segment_register = {
-                if registers_lock
-                    .uncommitted
-                    .contains_all(before_merge_segment_ids)
-                {
-                    &mut registers_lock.uncommitted
-                } else if registers_lock
-                    .committed
-                    .contains_all(before_merge_segment_ids)
-                {
-                    &mut registers_lock.committed
-                } else {
-                    warn!("couldn't find segment in SegmentManager");
-                    return;
-                }
-            };
-            for segment_id in before_merge_segment_ids {
-                target_segment_register.cancel_merge(segment_id);
-            }
-        }
-
-        // ... and we make sure the target segment entry
-        // can be garbage collected.
-        registers_lock.writing.remove(&after_merge_segment_id);
-    }
-
-    pub fn write_segment(&self, segment_id: SegmentId) {
-        let mut registers_lock = self.write();
-        registers_lock.writing.insert(segment_id);
-    }
-
     pub fn add_segment(&self, segment_entry: SegmentEntry) {
         let mut registers_lock = self.write();
-        registers_lock.writing.remove(&segment_entry.segment_id());
         registers_lock.uncommitted.add_segment_entry(segment_entry);
     }
 
@@ -202,10 +169,6 @@ impl SegmentManager {
         after_merge_segment_entry: SegmentEntry,
     ) {
         let mut registers_lock = self.write();
-        registers_lock
-            .writing
-            .remove(&after_merge_segment_entry.segment_id());
-
         let target_register: &mut SegmentRegister = {
             if registers_lock
                 .uncommitted
@@ -229,6 +192,7 @@ impl SegmentManager {
     }
 
     pub fn committed_segment_metas(&self) -> Vec<SegmentMeta> {
+        self.remove_empty_segments();
         let registers_lock = self.read();
         registers_lock.committed.segment_metas()
     }

src/indexer/segment_register.rs

@@ -3,6 +3,7 @@ use core::SegmentMeta;
 use indexer::delete_queue::DeleteCursor;
 use indexer::segment_entry::SegmentEntry;
 use std::collections::HashMap;
+use std::collections::HashSet;
 use std::fmt::{self, Debug, Formatter};
 
 /// The segment register keeps track
@@ -21,8 +22,8 @@ pub struct SegmentRegister {
 impl Debug for SegmentRegister {
     fn fmt(&self, f: &mut Formatter) -> Result<(), fmt::Error> {
         write!(f, "SegmentRegister(")?;
-        for (k, v) in &self.segment_states {
-            write!(f, "{}:{}, ", k.short_uuid_string(), v.state().letter_code())?;
+        for k in self.segment_states.keys() {
+            write!(f, "{}, ", k.short_uuid_string())?;
         }
         write!(f, ")")?;
         Ok(())
@@ -34,14 +35,13 @@ impl SegmentRegister {
         self.segment_states.clear();
     }
 
-    pub fn len(&self) -> usize {
-        self.segment_states.len()
-    }
-
-    pub fn get_mergeable_segments(&self) -> Vec<SegmentMeta> {
+    pub fn get_mergeable_segments(
+        &self,
+        in_merge_segment_ids: &HashSet<SegmentId>,
+    ) -> Vec<SegmentMeta> {
         self.segment_states
             .values()
-            .filter(|segment_entry| segment_entry.is_ready())
+            .filter(|segment_entry| !in_merge_segment_ids.contains(&segment_entry.segment_id()))
             .map(|segment_entry| segment_entry.meta().clone())
             .collect()
     }
@@ -60,7 +60,7 @@ impl SegmentRegister {
         segment_ids
     }
 
-    pub fn contains_all(&mut self, segment_ids: &[SegmentId]) -> bool {
+    pub fn contains_all(&self, segment_ids: &[SegmentId]) -> bool {
         segment_ids
             .iter()
             .all(|segment_id| self.segment_states.contains_key(segment_id))
@@ -75,20 +75,8 @@ impl SegmentRegister {
         self.segment_states.remove(segment_id);
     }
 
-    pub fn cancel_merge(&mut self, segment_id: &SegmentId) {
-        self.segment_states
-            .get_mut(segment_id)
-            .expect("Received a merge notification for a segment that is not registered")
-            .cancel_merge();
-    }
-
-    pub fn start_merge(&mut self, segment_id: &SegmentId) -> Option<SegmentEntry> {
-        if let Some(segment_entry) = self.segment_states.get_mut(segment_id) {
-            segment_entry.start_merge();
-            Some(segment_entry.clone())
-        } else {
-            None
-        }
+    pub fn get(&self, segment_id: &SegmentId) -> Option<SegmentEntry> {
+        self.segment_states.get(segment_id).cloned()
     }
 
     pub fn new(segment_metas: Vec<SegmentMeta>, delete_cursor: &DeleteCursor) -> SegmentRegister {
@@ -100,11 +88,6 @@ impl SegmentRegister {
         }
         SegmentRegister { segment_states }
     }
-
-    #[cfg(test)]
-    pub fn segment_entry(&self, segment_id: &SegmentId) -> Option<SegmentEntry> {
-        self.segment_states.get(segment_id).cloned()
-    }
 }
 
 #[cfg(test)]
@@ -113,7 +96,6 @@ mod tests {
     use core::SegmentId;
     use core::SegmentMeta;
     use indexer::delete_queue::*;
-    use indexer::SegmentState;
 
     fn segment_ids(segment_register: &SegmentRegister) -> Vec<SegmentId> {
         segment_register
@@ -137,42 +119,12 @@ mod tests {
             let segment_entry = SegmentEntry::new(segment_meta, delete_queue.cursor(), None);
             segment_register.add_segment_entry(segment_entry);
         }
-        assert_eq!(
-            segment_register
-                .segment_entry(&segment_id_a)
-                .unwrap()
-                .state(),
-            SegmentState::Ready
-        );
         assert_eq!(segment_ids(&segment_register), vec![segment_id_a]);
         {
             let segment_meta = SegmentMeta::new(segment_id_b, 0u32);
             let segment_entry = SegmentEntry::new(segment_meta, delete_queue.cursor(), None);
             segment_register.add_segment_entry(segment_entry);
         }
-        assert_eq!(
-            segment_register
-                .segment_entry(&segment_id_b)
-                .unwrap()
-                .state(),
-            SegmentState::Ready
-        );
-        segment_register.start_merge(&segment_id_a);
-        segment_register.start_merge(&segment_id_b);
-        assert_eq!(
-            segment_register
-                .segment_entry(&segment_id_a)
-                .unwrap()
-                .state(),
-            SegmentState::InMerge
-        );
-        assert_eq!(
-            segment_register
-                .segment_entry(&segment_id_b)
-                .unwrap()
-                .state(),
-            SegmentState::InMerge
-        );
         segment_register.remove_segment(&segment_id_a);
         segment_register.remove_segment(&segment_id_b);
         {

src/indexer/segment_updater.rs

@@ -16,9 +16,10 @@ use futures_cpupool::CpuFuture;
 use futures_cpupool::CpuPool;
 use indexer::delete_queue::DeleteCursor;
 use indexer::index_writer::advance_deletes;
+use indexer::merge_operation::MergeOperationInventory;
 use indexer::merger::IndexMerger;
 use indexer::stamper::Stamper;
-use indexer::MergeCandidate;
+use indexer::MergeOperation;
 use indexer::SegmentEntry;
 use indexer::SegmentSerializer;
 use indexer::{DefaultMergePolicy, MergePolicy};
@@ -26,6 +27,7 @@ use schema::Schema;
 use serde_json;
 use std::borrow::BorrowMut;
 use std::collections::HashMap;
+use std::collections::HashSet;
 use std::io::Write;
 use std::mem;
 use std::ops::DerefMut;
@@ -45,8 +47,16 @@ use Result;
 /// and flushed.
 ///
 /// This method is not part of tantivy's public API
-pub fn save_new_metas(schema: Schema, opstamp: u64, directory: &mut Directory) -> Result<()> {
-    save_metas(vec![], schema, opstamp, None, directory)
+pub fn save_new_metas(schema: Schema, directory: &mut Directory) -> Result<()> {
+    save_metas(
+        &IndexMeta {
+            segments: Vec::new(),
+            schema,
+            opstamp: 0u64,
+            payload: None,
+        },
+        directory,
+    )
 }
 
 /// Save the index meta file.
@@ -58,20 +68,9 @@ pub fn save_new_metas(schema: Schema, opstamp: u64, directory: &mut Directory) -
 /// and flushed.
 ///
 /// This method is not part of tantivy's public API
-pub fn save_metas(
-    segment_metas: Vec<SegmentMeta>,
-    schema: Schema,
-    opstamp: u64,
-    payload: Option<String>,
-    directory: &mut Directory,
-) -> Result<()> {
-    let metas = IndexMeta {
-        segments: segment_metas,
-        schema,
-        opstamp,
-        payload,
-    };
-    let mut buffer = serde_json::to_vec_pretty(&metas)?;
+fn save_metas(metas: &IndexMeta, directory: &mut Directory) -> Result<()> {
+    let mut buffer = serde_json::to_vec_pretty(metas)?;
+    // Just adding a new line at the end of the buffer.
     writeln!(&mut buffer)?;
     directory.atomic_write(&META_FILEPATH, &buffer[..])?;
     debug!("Saved metas {:?}", serde_json::to_string_pretty(&metas));
@@ -83,16 +82,21 @@ pub fn save_metas(
 //
 // All this processing happens on a single thread
 // consuming a common queue.
+//
+// We voluntarily pass a merge_operation ref to guarantee that
+// the merge_operation is alive during the process
 #[derive(Clone)]
 pub struct SegmentUpdater(Arc<InnerSegmentUpdater>);
 
 fn perform_merge(
+    merge_operation: &MergeOperation,
     index: &Index,
     mut segment_entries: Vec<SegmentEntry>,
-    mut merged_segment: Segment,
-    target_opstamp: u64,
 ) -> Result<SegmentEntry> {
+    let target_opstamp = merge_operation.target_opstamp();
+
     // first we need to apply deletes to our segment.
+    let mut merged_segment = index.new_segment();
 
     // TODO add logging
     let schema = index.schema();
@@ -126,15 +130,23 @@ fn perform_merge(
 }
 
 struct InnerSegmentUpdater {
+    // we keep a copy of the current active IndexMeta to
+    // avoid loading the file everytime we need it in the
+    // `SegmentUpdater`.
+    //
+    // This should be up to date as all update happen through
+    // the unique active `SegmentUpdater`.
+    active_metas: RwLock<Arc<IndexMeta>>,
     pool: CpuPool,
     index: Index,
     segment_manager: SegmentManager,
-    merge_policy: RwLock<Box<MergePolicy>>,
+    merge_policy: RwLock<Arc<Box<MergePolicy>>>,
     merging_thread_id: AtomicUsize,
     merging_threads: RwLock<HashMap<usize, JoinHandle<Result<()>>>>,
     generation: AtomicUsize,
     killed: AtomicBool,
     stamper: Stamper,
+    merge_operations: MergeOperationInventory,
 }
 
 impl SegmentUpdater {
@@ -149,32 +161,29 @@ impl SegmentUpdater {
             .name_prefix("segment_updater")
             .pool_size(1)
             .create();
+        let index_meta = index.load_metas()?;
         Ok(SegmentUpdater(Arc::new(InnerSegmentUpdater {
+            active_metas: RwLock::new(Arc::new(index_meta)),
             pool,
             index,
             segment_manager,
-            merge_policy: RwLock::new(Box::new(DefaultMergePolicy::default())),
+            merge_policy: RwLock::new(Arc::new(Box::new(DefaultMergePolicy::default()))),
             merging_thread_id: AtomicUsize::default(),
             merging_threads: RwLock::new(HashMap::new()),
             generation: AtomicUsize::default(),
             killed: AtomicBool::new(false),
             stamper,
+            merge_operations: Default::default(),
         })))
     }
 
-    pub fn new_segment(&self) -> Segment {
-        let new_segment = self.0.index.new_segment();
-        let segment_id = new_segment.id();
-        self.0.segment_manager.write_segment(segment_id);
-        new_segment
-    }
-
-    pub fn get_merge_policy(&self) -> Box<MergePolicy> {
-        self.0.merge_policy.read().unwrap().box_clone()
+    pub fn get_merge_policy(&self) -> Arc<Box<MergePolicy>> {
+        self.0.merge_policy.read().unwrap().clone()
     }
 
     pub fn set_merge_policy(&self, merge_policy: Box<MergePolicy>) {
-        *self.0.merge_policy.write().unwrap() = merge_policy;
+        let arc_merge_policy = Arc::new(merge_policy);
+        *self.0.merge_policy.write().unwrap() = arc_merge_policy;
     }
 
     fn get_merging_thread_id(&self) -> usize {
@@ -244,14 +253,15 @@ impl SegmentUpdater {
             //
             // Segment 1 from disk 1, Segment 1 from disk 2, etc.
             commited_segment_metas.sort_by_key(|segment_meta| -(segment_meta.max_doc() as i32));
-            save_metas(
-                commited_segment_metas,
-                index.schema(),
+            let index_meta = IndexMeta {
+                segments: commited_segment_metas,
+                schema: index.schema(),
                 opstamp,
-                commit_message,
-                directory.box_clone().borrow_mut(),
-            )
-            .expect("Could not save metas.");
+                payload: commit_message,
+            };
+            save_metas(&index_meta, directory.box_clone().borrow_mut())
+                .expect("Could not save metas.");
+            self.store_meta(&index_meta);
         }
     }
 
@@ -286,51 +296,62 @@ impl SegmentUpdater {
     }
 
     pub fn start_merge(&self, segment_ids: &[SegmentId]) -> Result<Receiver<SegmentMeta>> {
-        //let future_merged_segment = */
-        let segment_ids_vec = segment_ids.to_vec();
-        self.run_async(move |segment_updater| {
-            segment_updater.start_merge_impl(&segment_ids_vec[..])
-        })
-        .wait()?
+        let commit_opstamp = self.load_metas().opstamp;
+        let merge_operation = MergeOperation::new(
+            &self.0.merge_operations,
+            commit_opstamp,
+            segment_ids.to_vec(),
+        );
+        self.run_async(move |segment_updater| segment_updater.start_merge_impl(merge_operation))
+            .wait()?
+    }
+
+    fn store_meta(&self, index_meta: &IndexMeta) {
+        *self.0.active_metas.write().unwrap() = Arc::new(index_meta.clone());
+    }
+    fn load_metas(&self) -> Arc<IndexMeta> {
+        self.0.active_metas.read().unwrap().clone()
     }
 
     // `segment_ids` is required to be non-empty.
-    fn start_merge_impl(&self, segment_ids: &[SegmentId]) -> Result<Receiver<SegmentMeta>> {
-        assert!(!segment_ids.is_empty(), "Segment_ids cannot be empty.");
+    fn start_merge_impl(&self, merge_operation: MergeOperation) -> Result<Receiver<SegmentMeta>> {
+        assert!(
+            !merge_operation.segment_ids().is_empty(),
+            "Segment_ids cannot be empty."
+        );
 
         let segment_updater_clone = self.clone();
-        let segment_entries: Vec<SegmentEntry> = self.0.segment_manager.start_merge(segment_ids)?;
+        let segment_entries: Vec<SegmentEntry> = self
+            .0
+            .segment_manager
+            .start_merge(merge_operation.segment_ids())?;
 
-        let segment_ids_vec = segment_ids.to_vec();
+        //        let segment_ids_vec = merge_operation.segment_ids.to_vec();
 
         let merging_thread_id = self.get_merging_thread_id();
         info!(
             "Starting merge thread #{} - {:?}",
-            merging_thread_id, segment_ids
+            merging_thread_id,
+            merge_operation.segment_ids()
         );
         let (merging_future_send, merging_future_recv) = oneshot();
 
-        let target_opstamp = self.0.stamper.stamp();
-
         // first we need to apply deletes to our segment.
         let merging_join_handle = thread::Builder::new()
             .name(format!("mergingthread-{}", merging_thread_id))
             .spawn(move || {
                 // first we need to apply deletes to our segment.
-                let merged_segment = segment_updater_clone.new_segment();
-                let merged_segment_id = merged_segment.id();
                 let merge_result = perform_merge(
+                    &merge_operation,
                     &segment_updater_clone.0.index,
                     segment_entries,
-                    merged_segment,
-                    target_opstamp,
                 );
 
                 match merge_result {
                     Ok(after_merge_segment_entry) => {
                         let merged_segment_meta = after_merge_segment_entry.meta().clone();
                         segment_updater_clone
-                            .end_merge(segment_ids_vec, after_merge_segment_entry)
+                            .end_merge(merge_operation, after_merge_segment_entry)
                             .expect("Segment updater thread is corrupted.");
 
                         // the future may fail if the listener of the oneshot future
@@ -341,13 +362,18 @@ impl SegmentUpdater {
                         let _merging_future_res = merging_future_send.send(merged_segment_meta);
                     }
                     Err(e) => {
-                        warn!("Merge of {:?} was cancelled: {:?}", segment_ids_vec, e);
+                        warn!(
+                            "Merge of {:?} was cancelled: {:?}",
+                            merge_operation.segment_ids(),
+                            e
+                        );
                         // ... cancel merge
                         if cfg!(test) {
                             panic!("Merge failed.");
                         }
-                        segment_updater_clone.cancel_merge(&segment_ids_vec, merged_segment_id);
-                        // merging_future_send will be dropped, sending an error to the future.
+                        // As `merge_operation` will be dropped, the segment in merge state will
+                        // be available for merge again.
+                        // `merging_future_send` will be dropped, sending an error to the future.
                     }
                 }
                 segment_updater_clone
@@ -368,16 +394,34 @@ impl SegmentUpdater {
     }
 
     fn consider_merge_options(&self) {
+        let merge_segment_ids: HashSet<SegmentId> = self.0.merge_operations.segment_in_merge();
         let (committed_segments, uncommitted_segments) =
-            get_mergeable_segments(&self.0.segment_manager);
+            get_mergeable_segments(&merge_segment_ids, &self.0.segment_manager);
+
         // Committed segments cannot be merged with uncommitted_segments.
         // We therefore consider merges using these two sets of segments independently.
         let merge_policy = self.get_merge_policy();
-        let mut merge_candidates = merge_policy.compute_merge_candidates(&uncommitted_segments);
-        let committed_merge_candidates = merge_policy.compute_merge_candidates(&committed_segments);
-        merge_candidates.extend_from_slice(&committed_merge_candidates[..]);
-        for MergeCandidate(segment_metas) in merge_candidates {
-            match self.start_merge_impl(&segment_metas) {
+
+        let current_opstamp = self.0.stamper.stamp();
+        let mut merge_candidates: Vec<MergeOperation> = merge_policy
+            .compute_merge_candidates(&uncommitted_segments)
+            .into_iter()
+            .map(|merge_candidate| {
+                MergeOperation::new(&self.0.merge_operations, current_opstamp, merge_candidate.0)
+            })
+            .collect();
+
+        let commit_opstamp = self.load_metas().opstamp;
+        let committed_merge_candidates = merge_policy
+            .compute_merge_candidates(&committed_segments)
+            .into_iter()
+            .map(|merge_candidate| {
+                MergeOperation::new(&self.0.merge_operations, commit_opstamp, merge_candidate.0)
+            })
+            .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) => {
                     if let Err(e) = merge_future.fuse().poll() {
                         error!("The merge task failed quickly after starting: {:?}", e);
@@ -393,31 +437,16 @@ impl SegmentUpdater {
         }
     }
 
-    fn cancel_merge(
-        &self,
-        before_merge_segment_ids: &[SegmentId],
-        after_merge_segment_entry: SegmentId,
-    ) {
-        self.0
-            .segment_manager
-            .cancel_merge(before_merge_segment_ids, after_merge_segment_entry);
-    }
-
     fn end_merge(
         &self,
-        before_merge_segment_ids: Vec<SegmentId>,
+        merge_operation: MergeOperation,
         mut after_merge_segment_entry: SegmentEntry,
     ) -> Result<()> {
         self.run_async(move |segment_updater| {
             info!("End merge {:?}", after_merge_segment_entry.meta());
             let mut delete_cursor = after_merge_segment_entry.delete_cursor().clone();
             if let Some(delete_operation) = delete_cursor.get() {
-                let committed_opstamp = segment_updater
-                    .0
-                    .index
-                    .load_metas()
-                    .expect("Failed to read opstamp")
-                    .opstamp;
+                let committed_opstamp = segment_updater.load_metas().opstamp;
                 if delete_operation.opstamp < committed_opstamp {
                     let index = &segment_updater.0.index;
                     let segment = index.segment(after_merge_segment_entry.meta().clone());
@@ -426,16 +455,15 @@ impl SegmentUpdater {
                     {
                         error!(
                             "Merge of {:?} was cancelled (advancing deletes failed): {:?}",
-                            before_merge_segment_ids, e
+                            merge_operation.segment_ids(),
+                            e
                         );
-                        // ... cancel merge
                         if cfg!(test) {
                             panic!("Merge failed.");
                         }
-                        segment_updater.cancel_merge(
-                            &before_merge_segment_ids,
-                            after_merge_segment_entry.segment_id(),
-                        );
+                        // ... cancel merge
+                        // `merge_operations` are tracked. As it is dropped, the
+                        // the segment_ids will be available again for merge.
                         return;
                     }
                 }
@@ -443,11 +471,11 @@ impl SegmentUpdater {
             segment_updater
                 .0
                 .segment_manager
-                .end_merge(&before_merge_segment_ids, after_merge_segment_entry);
+                .end_merge(merge_operation.segment_ids(), after_merge_segment_entry);
             segment_updater.consider_merge_options();
             info!("save metas");
-            let previous_metas = segment_updater.0.index.load_metas().unwrap();
-            segment_updater.save_metas(previous_metas.opstamp, previous_metas.payload);
+            let previous_metas = segment_updater.load_metas();
+            segment_updater.save_metas(previous_metas.opstamp, previous_metas.payload.clone());
             segment_updater.garbage_collect_files_exec();
         })
         .wait()
@@ -469,32 +497,25 @@ impl SegmentUpdater {
     /// Obsolete files will eventually be cleaned up
     /// by the directory garbage collector.
     pub fn wait_merging_thread(&self) -> Result<()> {
-        let mut num_segments: usize;
         loop {
-            num_segments = self.0.segment_manager.num_segments();
-
-            let mut new_merging_threads = HashMap::new();
-            {
+            let merging_threads: HashMap<usize, JoinHandle<Result<()>>> = {
                 let mut merging_threads = self.0.merging_threads.write().unwrap();
-                mem::swap(&mut new_merging_threads, merging_threads.deref_mut());
+                mem::replace(merging_threads.deref_mut(), HashMap::new())
+            };
+            if merging_threads.is_empty() {
+                return Ok(());
             }
-            debug!("wait merging thread {}", new_merging_threads.len());
-            for (_, merging_thread_handle) in new_merging_threads {
+            debug!("wait merging thread {}", merging_threads.len());
+            for (_, merging_thread_handle) in merging_threads {
                 merging_thread_handle
                     .join()
                     .map(|_| ())
                     .map_err(|_| TantivyError::ErrorInThread("Merging thread failed.".into()))?;
             }
-            // Our merging thread may have queued their completed
+            // Our merging thread may have queued their completed merged segment.
+            // Let's wait for that too.
             self.run_async(move |_| {}).wait()?;
-
-            let new_num_segments = self.0.segment_manager.num_segments();
-
-            if new_num_segments >= num_segments {
-                break;
-            }
         }
-        Ok(())
     }
 }
 
@@ -514,7 +535,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
 
         // writing the segment
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         index_writer.set_merge_policy(Box::new(MergeWheneverPossible));
 
         {
@@ -558,4 +579,75 @@ mod tests {
         assert_eq!(index.searcher().segment_readers().len(), 1);
         assert_eq!(index.searcher().num_docs(), 302);
     }
+
+    #[test]
+    fn delete_all_docs() {
+        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);
+
+        // writing the segment
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+
+        {
+            for _ in 0..100 {
+                index_writer.add_document(doc!(text_field=>"a"));
+                index_writer.add_document(doc!(text_field=>"b"));
+            }
+            assert!(index_writer.commit().is_ok());
+        }
+
+        {
+            for _ in 0..100 {
+                index_writer.add_document(doc!(text_field=>"c"));
+                index_writer.add_document(doc!(text_field=>"d"));
+            }
+            assert!(index_writer.commit().is_ok());
+        }
+
+        {
+            index_writer.add_document(doc!(text_field=>"e"));
+            index_writer.add_document(doc!(text_field=>"f"));
+            assert!(index_writer.commit().is_ok());
+        }
+
+        {
+            let seg_ids = index
+                .searchable_segment_ids()
+                .expect("Searchable segments failed.");
+            // docs exist, should have at least 1 segment
+            assert!(seg_ids.len() > 0);
+        }
+
+        {
+            let term_vals = vec!["a", "b", "c", "d", "e", "f"];
+            for term_val in term_vals {
+                let term = Term::from_field_text(text_field, term_val);
+                index_writer.delete_term(term);
+                assert!(index_writer.commit().is_ok());
+            }
+        }
+
+        {
+            index_writer
+                .wait_merging_threads()
+                .expect("waiting for merging threads");
+        }
+
+        index.load_searchers().unwrap();
+        assert_eq!(index.searcher().num_docs(), 0);
+
+        let seg_ids = index
+            .searchable_segment_ids()
+            .expect("Searchable segments failed.");
+        assert!(seg_ids.is_empty());
+
+        index.load_searchers().unwrap();
+        assert_eq!(index.searcher().num_docs(), 0);
+        // empty segments should be erased
+        assert!(index.searchable_segment_metas().unwrap().is_empty());
+        assert!(index.searcher().segment_readers().is_empty());
+    }
 }

src/indexer/segment_writer.rs

@@ -111,19 +111,18 @@ impl SegmentWriter {
             }
             match *field_options.field_type() {
                 FieldType::HierarchicalFacet => {
-                    let facets: Vec<&[u8]> = field_values
+                    let facets: Vec<&str> = field_values
                         .iter()
                         .flat_map(|field_value| match *field_value.value() {
-                            Value::Facet(ref facet) => Some(facet.encoded_bytes()),
+                            Value::Facet(ref facet) => Some(facet.encoded_str()),
                             _ => {
                                 panic!("Expected hierarchical facet");
                             }
                         })
                         .collect();
                     let mut term = Term::for_field(field); // we set the Term
-                    for facet_bytes in facets {
+                    for fake_str in facets {
                         let mut unordered_term_id_opt = None;
-                        let fake_str = unsafe { str::from_utf8_unchecked(facet_bytes) };
                         FacetTokenizer.token_stream(fake_str).process(&mut |token| {
                             term.set_text(&token.text);
                             let unordered_term_id =

src/indexer/stamper.rs

@@ -1,50 +1,66 @@
+use std::sync::atomic::Ordering;
+use std::sync::Arc;
+
 // AtomicU64 have not landed in stable.
 // For the moment let's just use AtomicUsize on
 // x86/64 bit platform, and a mutex on other platform.
-
-#[cfg(target = "x86_64")]
+#[cfg(target_arch = "x86_64")]
 mod archicture_impl {
 
     use std::sync::atomic::{AtomicUsize, Ordering};
-    use std::sync::Arc;
 
-    #[derive(Clone, Default)]
-    pub struct Stamper(Arc<AtomicU64>);
+    #[derive(Default)]
+    pub struct AtomicU64Ersatz(AtomicUsize);
 
-    impl Stamper {
-        pub fn new(first_opstamp: u64) -> Stamper {
-            Stamper(Arc::new(AtomicU64::new(first_opstamp)))
+    impl AtomicU64Ersatz {
+        pub fn new(first_opstamp: u64) -> AtomicU64Ersatz {
+            AtomicU64Ersatz(AtomicUsize::new(first_opstamp as usize))
         }
 
-        pub fn stamp(&self) -> u64 {
-            self.0.fetch_add(1u64, Ordering::SeqCst) as u64
+        pub fn fetch_add(&self, val: u64, order: Ordering) -> u64 {
+            self.0.fetch_add(val as usize, order) as u64
         }
     }
 }
 
-#[cfg(not(target = "x86_64"))]
+#[cfg(not(target_arch = "x86_64"))]
 mod archicture_impl {
 
-    use std::sync::{Arc, Mutex};
+    use std::sync::atomic::Ordering;
+    /// Under other architecture, we rely on a mutex.
+    use std::sync::RwLock;
 
-    #[derive(Clone, Default)]
-    pub struct Stamper(Arc<Mutex<u64>>);
+    #[derive(Default)]
+    pub struct AtomicU64Ersatz(RwLock<u64>);
 
-    impl Stamper {
-        pub fn new(first_opstamp: u64) -> Stamper {
-            Stamper(Arc::new(Mutex::new(first_opstamp)))
+    impl AtomicU64Ersatz {
+        pub fn new(first_opstamp: u64) -> AtomicU64Ersatz {
+            AtomicU64Ersatz(RwLock::new(first_opstamp))
         }
 
-        pub fn stamp(&self) -> u64 {
-            let mut guard = self.0.lock().expect("Failed to lock the stamper");
-            let previous_val = *guard;
-            *guard = previous_val + 1;
+        pub fn fetch_add(&self, incr: u64, _order: Ordering) -> u64 {
+            let mut lock = self.0.write().unwrap();
+            let previous_val = *lock;
+            *lock = previous_val + incr;
             previous_val
         }
     }
 }
 
-pub use self::archicture_impl::Stamper;
+use self::archicture_impl::AtomicU64Ersatz;
+
+#[derive(Clone, Default)]
+pub struct Stamper(Arc<AtomicU64Ersatz>);
+
+impl Stamper {
+    pub fn new(first_opstamp: u64) -> Stamper {
+        Stamper(Arc::new(AtomicU64Ersatz::new(first_opstamp)))
+    }
+
+    pub fn stamp(&self) -> u64 {
+        self.0.fetch_add(1u64, Ordering::SeqCst) as u64
+    }
+}
 
 #[cfg(test)]
 mod test {

src/lib.rs

@@ -129,10 +129,7 @@ extern crate base64;
 extern crate bit_set;
 extern crate bitpacking;
 extern crate byteorder;
-extern crate scoped_pool;
-
 extern crate combine;
-
 extern crate crossbeam;
 extern crate fnv;
 extern crate fst;
@@ -146,6 +143,7 @@ extern crate num_cpus;
 extern crate owning_ref;
 extern crate regex;
 extern crate rust_stemmers;
+extern crate scoped_pool;
 extern crate serde;
 extern crate stable_deref_trait;
 extern crate tempdir;
@@ -170,7 +168,7 @@ extern crate maplit;
 extern crate test;
 
 #[macro_use]
-extern crate downcast;
+extern crate downcast_rs;
 
 #[macro_use]
 extern crate fail;
@@ -231,11 +229,7 @@ pub use common::{i64_to_u64, u64_to_i64};
 /// Expose the current version of tantivy, as well
 /// whether it was compiled with the simd compression.
 pub fn version() -> &'static str {
-    if cfg!(feature = "simdcompression") {
-        concat!(env!("CARGO_PKG_VERSION"), "-simd")
-    } else {
-        concat!(env!("CARGO_PKG_VERSION"), "-nosimd")
-    }
+    env!("CARGO_PKG_VERSION")
 }
 
 /// Defines tantivy's merging strategy
@@ -348,7 +342,7 @@ mod tests {
         let index = Index::create_from_tempdir(schema).unwrap();
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field=>"af b");
                 index_writer.add_document(doc);
@@ -370,7 +364,7 @@ mod tests {
         let mut schema_builder = Schema::builder();
         let text_field = schema_builder.add_text_field("text", TEXT);
         let index = Index::create_in_ram(schema_builder.build());
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         {
             index_writer.add_document(doc!(text_field=>"a b c"));
             index_writer.commit().unwrap();
@@ -412,7 +406,7 @@ mod tests {
         let text_field = schema_builder.add_text_field("text", TEXT);
         let index = Index::create_in_ram(schema_builder.build());
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field=>"a b c");
                 index_writer.add_document(doc);
@@ -440,7 +434,7 @@ mod tests {
         let text_field = schema_builder.add_text_field("text", TEXT);
         let index = Index::create_in_ram(schema_builder.build());
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field=>"a b c");
                 index_writer.add_document(doc);
@@ -487,7 +481,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             // 0
             index_writer.add_document(doc!(text_field=>"a b"));
             // 1
@@ -534,7 +528,7 @@ mod tests {
         }
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             // 0
             index_writer.add_document(doc!(text_field=>"a b"));
             // 1
@@ -571,7 +565,7 @@ mod tests {
         }
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"a b"));
             index_writer.delete_term(Term::from_field_text(text_field, "c"));
             index_writer.rollback().unwrap();
@@ -620,7 +614,7 @@ mod tests {
         let schema = schema_builder.build();
 
         let index = Index::create_in_ram(schema);
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         index_writer.add_document(doc!(field=>1u64));
         index_writer.commit().unwrap();
         index.load_searchers().unwrap();
@@ -643,7 +637,7 @@ mod tests {
         let schema = schema_builder.build();
 
         let index = Index::create_in_ram(schema);
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         let negative_val = -1i64;
         index_writer.add_document(doc!(value_field => negative_val));
         index_writer.commit().unwrap();
@@ -667,7 +661,7 @@ mod tests {
         let absent_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(2, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(2, 6_000_000).unwrap();
         index_writer.add_document(doc!(text_field=>"a"));
         assert!(index_writer.commit().is_ok());
         assert!(index.load_searchers().is_ok());
@@ -684,7 +678,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
 
         // writing the segment
-        let mut index_writer = index.writer_with_num_threads(2, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(2, 6_000_000).unwrap();
 
         let add_document = |index_writer: &mut IndexWriter, val: &'static str| {
             let doc = doc!(text_field=>val);
@@ -720,7 +714,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field=>"af af af bc bc");
                 index_writer.add_document(doc);
@@ -756,7 +750,7 @@ mod tests {
 
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"af af af b"));
             index_writer.add_document(doc!(text_field=>"a b c"));
             index_writer.add_document(doc!(text_field=>"a b c d"));
@@ -809,7 +803,7 @@ mod tests {
 
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field=>"af b");
                 index_writer.add_document(doc);

src/postings/mod.rs

@@ -280,7 +280,7 @@ pub mod tests {
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let mut doc = Document::default();
                 doc.add_text(text_field, "g b b d c g c");
@@ -322,7 +322,7 @@ pub mod tests {
 
             let index = Index::create_in_ram(schema);
             {
-                let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+                let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
                 for i in 0..num_docs {
                     let mut doc = Document::default();
                     doc.add_u64(value_field, 2);
@@ -399,7 +399,7 @@ pub mod tests {
 
         // delete some of the documents
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.delete_term(term_0);
             assert!(index_writer.commit().is_ok());
         }
@@ -449,7 +449,7 @@ pub mod tests {
 
         // delete everything else
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.delete_term(term_1);
 
             assert!(index_writer.commit().is_ok());
@@ -457,25 +457,14 @@ pub mod tests {
         index.load_searchers().unwrap();
 
         let searcher = index.searcher();
-        let segment_reader = searcher.segment_reader(0);
 
         // finally, check that it's empty
         {
-            let mut segment_postings = segment_reader
-                .inverted_index(term_2.field())
-                .read_postings(&term_2, IndexRecordOption::Basic)
-                .unwrap();
-
-            assert_eq!(segment_postings.skip_next(0), SkipResult::Reached);
-            assert_eq!(segment_postings.doc(), 0);
-            assert!(segment_reader.is_deleted(0));
-
-            let mut segment_postings = segment_reader
-                .inverted_index(term_2.field())
-                .read_postings(&term_2, IndexRecordOption::Basic)
-                .unwrap();
-
-            assert_eq!(segment_postings.skip_next(num_docs), SkipResult::End);
+            let searchable_segment_ids = index
+                .searchable_segment_ids()
+                .expect("could not get index segment ids");
+            assert!(searchable_segment_ids.is_empty());
+            assert_eq!(searcher.num_docs(), 0);
         }
     }
 
@@ -506,7 +495,7 @@ pub mod tests {
             let index = Index::create_in_ram(schema);
             let posting_list_size = 1_000_000;
             {
-                let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+                let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
                 for _ in 0..posting_list_size {
                     let mut doc = Document::default();
                     if rng.gen_bool(1f64 / 15f64) {

src/postings/postings_writer.rs

@@ -1,6 +1,8 @@
 use super::stacker::{Addr, MemoryArena, TermHashMap};
 
-use postings::recorder::{NothingRecorder, Recorder, TFAndPositionRecorder, TermFrequencyRecorder};
+use postings::recorder::{
+    BufferLender, NothingRecorder, Recorder, TFAndPositionRecorder, TermFrequencyRecorder,
+};
 use postings::UnorderedTermId;
 use postings::{FieldSerializer, InvertedIndexSerializer};
 use schema::IndexRecordOption;
@@ -213,7 +215,7 @@ pub trait PostingsWriter {
 
 /// The `SpecializedPostingsWriter` is just here to remove dynamic
 /// dispatch to the recorder information.
-pub struct SpecializedPostingsWriter<Rec: Recorder + 'static> {
+pub(crate) struct SpecializedPostingsWriter<Rec: Recorder + 'static> {
     total_num_tokens: u64,
     _recorder_type: PhantomData<Rec>,
 }
@@ -245,8 +247,7 @@ impl<Rec: Recorder + 'static> PostingsWriter for SpecializedPostingsWriter<Rec>
         debug_assert!(term.as_slice().len() >= 4);
         self.total_num_tokens += 1;
         term_index.mutate_or_create(term, |opt_recorder: Option<Rec>| {
-            if opt_recorder.is_some() {
-                let mut recorder = opt_recorder.unwrap();
+            if let Some(mut recorder) = opt_recorder {
                 let current_doc = recorder.current_doc();
                 if current_doc != doc {
                     recorder.close_doc(heap);
@@ -255,7 +256,7 @@ impl<Rec: Recorder + 'static> PostingsWriter for SpecializedPostingsWriter<Rec>
                 recorder.record_position(position, heap);
                 recorder
             } else {
-                let mut recorder = Rec::new(heap);
+                let mut recorder = Rec::new();
                 recorder.new_doc(doc, heap);
                 recorder.record_position(position, heap);
                 recorder
@@ -270,10 +271,11 @@ impl<Rec: Recorder + 'static> PostingsWriter for SpecializedPostingsWriter<Rec>
         termdict_heap: &MemoryArena,
         heap: &MemoryArena,
     ) -> io::Result<()> {
+        let mut buffer_lender = BufferLender::default();
         for &(term_bytes, addr, _) in term_addrs {
-            let recorder: Rec = unsafe { termdict_heap.read(addr) };
+            let recorder: Rec = termdict_heap.read(addr);
             serializer.new_term(&term_bytes[4..])?;
-            recorder.serialize(serializer, heap)?;
+            recorder.serialize(&mut buffer_lender, serializer, heap)?;
             serializer.close_term()?;
         }
         Ok(())

src/postings/recorder.rs

@@ -1,10 +1,51 @@
 use super::stacker::{ExpUnrolledLinkedList, MemoryArena};
+use common::{read_u32_vint, write_u32_vint};
 use postings::FieldSerializer;
-use std::{self, io};
+use std::io;
 use DocId;
 
 const EMPTY_ARRAY: [u32; 0] = [0u32; 0];
-const POSITION_END: u32 = std::u32::MAX;
+const POSITION_END: u32 = 0;
+
+#[derive(Default)]
+pub(crate) struct BufferLender {
+    buffer_u8: Vec<u8>,
+    buffer_u32: Vec<u32>,
+}
+
+impl BufferLender {
+    pub fn lend_u8(&mut self) -> &mut Vec<u8> {
+        self.buffer_u8.clear();
+        &mut self.buffer_u8
+    }
+    pub fn lend_all(&mut self) -> (&mut Vec<u8>, &mut Vec<u32>) {
+        self.buffer_u8.clear();
+        self.buffer_u32.clear();
+        (&mut self.buffer_u8, &mut self.buffer_u32)
+    }
+}
+
+pub struct VInt32Reader<'a> {
+    data: &'a [u8],
+}
+
+impl<'a> VInt32Reader<'a> {
+    fn new(data: &'a [u8]) -> VInt32Reader<'a> {
+        VInt32Reader { data }
+    }
+}
+
+impl<'a> Iterator for VInt32Reader<'a> {
+    type Item = u32;
+
+    fn next(&mut self) -> Option<u32> {
+        if self.data.is_empty() {
+            None
+        } else {
+            Some(read_u32_vint(&mut self.data))
+        }
+    }
+}
 
 /// Recorder is in charge of recording relevant information about
 /// the presence of a term in a document.
@@ -15,9 +56,9 @@ const POSITION_END: u32 = std::u32::MAX;
 ///   * the document id
 ///   * the term frequency
 ///   * the term positions
-pub trait Recorder: Copy {
+pub(crate) trait Recorder: Copy + 'static {
     ///
-    fn new(heap: &mut MemoryArena) -> Self;
+    fn new() -> Self;
     /// Returns the current document
     fn current_doc(&self) -> u32;
     /// Starts recording information about a new document
@@ -29,7 +70,12 @@ pub trait Recorder: Copy {
     /// Close the document. It will help record the term frequency.
     fn close_doc(&mut self, heap: &mut MemoryArena);
     /// Pushes the postings information to the serializer.
-    fn serialize(&self, serializer: &mut FieldSerializer, heap: &MemoryArena) -> io::Result<()>;
+    fn serialize(
+        &self,
+        buffer_lender: &mut BufferLender,
+        serializer: &mut FieldSerializer,
+        heap: &MemoryArena,
+    ) -> io::Result<()>;
 }
 
 /// Only records the doc ids
@@ -40,9 +86,9 @@ pub struct NothingRecorder {
 }
 
 impl Recorder for NothingRecorder {
-    fn new(heap: &mut MemoryArena) -> Self {
+    fn new() -> Self {
         NothingRecorder {
-            stack: ExpUnrolledLinkedList::new(heap),
+            stack: ExpUnrolledLinkedList::new(),
             current_doc: u32::max_value(),
         }
     }
@@ -53,16 +99,23 @@ impl Recorder for NothingRecorder {
 
     fn new_doc(&mut self, doc: DocId, heap: &mut MemoryArena) {
         self.current_doc = doc;
-        self.stack.push(doc, heap);
+        let _ = write_u32_vint(doc, &mut self.stack.writer(heap));
     }
 
     fn record_position(&mut self, _position: u32, _heap: &mut MemoryArena) {}
 
     fn close_doc(&mut self, _heap: &mut MemoryArena) {}
 
-    fn serialize(&self, serializer: &mut FieldSerializer, heap: &MemoryArena) -> io::Result<()> {
-        for doc in self.stack.iter(heap) {
-            serializer.write_doc(doc, 0u32, &EMPTY_ARRAY)?;
+    fn serialize(
+        &self,
+        buffer_lender: &mut BufferLender,
+        serializer: &mut FieldSerializer,
+        heap: &MemoryArena,
+    ) -> io::Result<()> {
+        let buffer = buffer_lender.lend_u8();
+        self.stack.read_to_end(heap, buffer);
+        for doc in VInt32Reader::new(&buffer[..]) {
+            serializer.write_doc(doc as u32, 0u32, &EMPTY_ARRAY)?;
         }
         Ok(())
     }
@@ -77,9 +130,9 @@ pub struct TermFrequencyRecorder {
 }
 
 impl Recorder for TermFrequencyRecorder {
-    fn new(heap: &mut MemoryArena) -> Self {
+    fn new() -> Self {
         TermFrequencyRecorder {
-            stack: ExpUnrolledLinkedList::new(heap),
+            stack: ExpUnrolledLinkedList::new(),
             current_doc: u32::max_value(),
             current_tf: 0u32,
         }
@@ -91,7 +144,7 @@ impl Recorder for TermFrequencyRecorder {
 
     fn new_doc(&mut self, doc: DocId, heap: &mut MemoryArena) {
         self.current_doc = doc;
-        self.stack.push(doc, heap);
+        let _ = write_u32_vint(doc, &mut self.stack.writer(heap));
     }
 
     fn record_position(&mut self, _position: u32, _heap: &mut MemoryArena) {
@@ -100,24 +153,24 @@ impl Recorder for TermFrequencyRecorder {
 
     fn close_doc(&mut self, heap: &mut MemoryArena) {
         debug_assert!(self.current_tf > 0);
-        self.stack.push(self.current_tf, heap);
+        let _ = write_u32_vint(self.current_tf, &mut self.stack.writer(heap));
         self.current_tf = 0;
     }
 
-    fn serialize(&self, serializer: &mut FieldSerializer, heap: &MemoryArena) -> io::Result<()> {
-        // the last document has not been closed...
-        // its term freq is self.current_tf.
-        let mut doc_iter = self
-            .stack
-            .iter(heap)
-            .chain(Some(self.current_tf).into_iter());
-
-        while let Some(doc) = doc_iter.next() {
-            let term_freq = doc_iter
-                .next()
-                .expect("The IndexWriter recorded a doc without a term freq.");
-            serializer.write_doc(doc, term_freq, &EMPTY_ARRAY)?;
+    fn serialize(
+        &self,
+        buffer_lender: &mut BufferLender,
+        serializer: &mut FieldSerializer,
+        heap: &MemoryArena,
+    ) -> io::Result<()> {
+        let buffer = buffer_lender.lend_u8();
+        self.stack.read_to_end(heap, buffer);
+        let mut u32_it = VInt32Reader::new(&buffer[..]);
+        while let Some(doc) = u32_it.next() {
+            let term_freq = u32_it.next().unwrap_or(self.current_tf);
+            serializer.write_doc(doc as u32, term_freq, &EMPTY_ARRAY)?;
         }
+
         Ok(())
     }
 }
@@ -128,11 +181,10 @@ pub struct TFAndPositionRecorder {
     stack: ExpUnrolledLinkedList,
     current_doc: DocId,
 }
-
 impl Recorder for TFAndPositionRecorder {
-    fn new(heap: &mut MemoryArena) -> Self {
+    fn new() -> Self {
         TFAndPositionRecorder {
-            stack: ExpUnrolledLinkedList::new(heap),
+            stack: ExpUnrolledLinkedList::new(),
             current_doc: u32::max_value(),
         }
     }
@@ -143,33 +195,88 @@ impl Recorder for TFAndPositionRecorder {
 
     fn new_doc(&mut self, doc: DocId, heap: &mut MemoryArena) {
         self.current_doc = doc;
-        self.stack.push(doc, heap);
+        let _ = write_u32_vint(doc, &mut self.stack.writer(heap));
     }
 
     fn record_position(&mut self, position: u32, heap: &mut MemoryArena) {
-        self.stack.push(position, heap);
+        let _ = write_u32_vint(position + 1u32, &mut self.stack.writer(heap));
     }
 
     fn close_doc(&mut self, heap: &mut MemoryArena) {
-        self.stack.push(POSITION_END, heap);
+        let _ = write_u32_vint(POSITION_END, &mut self.stack.writer(heap));
     }
 
-    fn serialize(&self, serializer: &mut FieldSerializer, heap: &MemoryArena) -> io::Result<()> {
-        let mut doc_positions = Vec::with_capacity(100);
-        let mut positions_iter = self.stack.iter(heap);
-        while let Some(doc) = positions_iter.next() {
-            let mut prev_position = 0;
-            doc_positions.clear();
-            for position in &mut positions_iter {
-                if position == POSITION_END {
-                    break;
-                } else {
-                    doc_positions.push(position - prev_position);
-                    prev_position = position;
+    fn serialize(
+        &self,
+        buffer_lender: &mut BufferLender,
+        serializer: &mut FieldSerializer,
+        heap: &MemoryArena,
+    ) -> io::Result<()> {
+        let (buffer_u8, buffer_positions) = buffer_lender.lend_all();
+        self.stack.read_to_end(heap, buffer_u8);
+        let mut u32_it = VInt32Reader::new(&buffer_u8[..]);
+        while let Some(doc) = u32_it.next() {
+            let mut prev_position_plus_one = 1u32;
+            buffer_positions.clear();
+            loop {
+                match u32_it.next() {
+                    Some(POSITION_END) | None => {
+                        break;
+                    }
+                    Some(position_plus_one) => {
+                        let delta_position = position_plus_one - prev_position_plus_one;
+                        buffer_positions.push(delta_position);
+                        prev_position_plus_one = position_plus_one;
+                    }
                 }
             }
-            serializer.write_doc(doc, doc_positions.len() as u32, &doc_positions)?;
+            serializer.write_doc(doc, buffer_positions.len() as u32, &buffer_positions)?;
         }
         Ok(())
     }
 }
+
+#[cfg(test)]
+mod tests {
+
+    use super::write_u32_vint;
+    use super::BufferLender;
+    use super::VInt32Reader;
+
+    #[test]
+    fn test_buffer_lender() {
+        let mut buffer_lender = BufferLender::default();
+        {
+            let buf = buffer_lender.lend_u8();
+            assert!(buf.is_empty());
+            buf.push(1u8);
+        }
+        {
+            let buf = buffer_lender.lend_u8();
+            assert!(buf.is_empty());
+            buf.push(1u8);
+        }
+        {
+            let (_, buf) = buffer_lender.lend_all();
+            assert!(buf.is_empty());
+            buf.push(1u32);
+        }
+        {
+            let (_, buf) = buffer_lender.lend_all();
+            assert!(buf.is_empty());
+            buf.push(1u32);
+        }
+    }
+
+    #[test]
+    fn test_vint_u32() {
+        let mut buffer = vec![];
+        let vals = [0, 1, 324_234_234, u32::max_value()];
+        for &i in &vals {
+            assert!(write_u32_vint(i, &mut buffer).is_ok());
+        }
+        assert_eq!(buffer.len(), 1 + 1 + 5 + 5);
+        let res: Vec<u32> = VInt32Reader::new(&buffer[..]).collect();
+        assert_eq!(&res[..], &vals[..]);
+    }
+}

src/postings/segment_postings.rs

@@ -124,22 +124,17 @@ impl SegmentPostings {
 }
 
 fn exponential_search(target: u32, arr: &[u32]) -> (usize, usize) {
-    let mut start = 0;
     let end = arr.len();
-    debug_assert!(target >= arr[start]);
+    debug_assert!(arr.len() <= 128);
     debug_assert!(target <= arr[end - 1]);
-    let mut jump = 1;
-    loop {
-        let new = start + jump;
-        if new >= end {
-            return (start, end);
+    let mut begin = 0;
+    for &pivot in [1,3,7,15,31,63].iter().take_while(|&&el| el < end) {
+        if arr[pivot] > target {
+            return (begin, pivot);
         }
-        if arr[new] > target {
-            return (start, new);
-        }
-        start = new;
-        jump *= 2;
+        begin = pivot;
     }
+    (begin, end)
 }
 
 /// Search the first index containing an element greater or equal to the target.
@@ -216,11 +211,10 @@ impl DocSet for SegmentPostings {
 
         // we're in the right block now, start with an exponential search
         let block_docs = self.block_cursor.docs();
-
-        debug_assert!(target >= self.doc());
         let new_cur = self
             .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()
@@ -622,6 +616,7 @@ impl<'b> Streamer<'b> for BlockSegmentPostings {
 #[cfg(test)]
 mod tests {
 
+    use super::exponential_search;
     use super::search_within_block;
     use super::BlockSegmentPostings;
     use super::BlockSegmentPostingsSkipResult;
@@ -635,6 +630,7 @@ mod tests {
     use schema::Term;
     use schema::INT_INDEXED;
     use DocId;
+    use SkipResult;
 
     #[test]
     fn test_empty_segment_postings() {
@@ -662,6 +658,16 @@ mod tests {
             .0
     }
 
+    #[test]
+    fn test_exponentiel_search() {
+        assert_eq!(exponential_search(0, &[1, 2]), (0, 1));
+        assert_eq!(exponential_search(1, &[1, 2]), (0, 1));
+        assert_eq!(
+            exponential_search(7, &[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]),
+            (3, 7)
+        );
+    }
+
     fn util_test_search_within_block(block: &[u32], target: u32) {
         assert_eq!(
             search_within_block(block, target),
@@ -693,7 +699,7 @@ mod tests {
 
     #[test]
     fn test_block_segment_postings() {
-        let mut block_segments = build_block_postings((0..100_000).collect::<Vec<u32>>());
+        let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>());
         let mut offset: u32 = 0u32;
         // checking that the block before calling advance is empty
         assert!(block_segments.docs().is_empty());
@@ -707,14 +713,44 @@ mod tests {
         }
     }
 
-    fn build_block_postings(docs: Vec<DocId>) -> BlockSegmentPostings {
+    #[test]
+    fn test_skip_right_at_new_block() {
+        let mut doc_ids = (0..128).collect::<Vec<u32>>();
+        doc_ids.push(129);
+        doc_ids.push(130);
+        {
+            let block_segments = build_block_postings(&doc_ids);
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
+            assert_eq!(docset.skip_next(128), SkipResult::OverStep);
+            assert_eq!(docset.doc(), 129);
+            assert!(docset.advance());
+            assert_eq!(docset.doc(), 130);
+            assert!(!docset.advance());
+        }
+        {
+            let block_segments = build_block_postings(&doc_ids);
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
+            assert_eq!(docset.skip_next(129), SkipResult::Reached);
+            assert_eq!(docset.doc(), 129);
+            assert!(docset.advance());
+            assert_eq!(docset.doc(), 130);
+            assert!(!docset.advance());
+        }
+        {
+            let block_segments = build_block_postings(&doc_ids);
+            let mut docset = SegmentPostings::from_block_postings(block_segments, None);
+            assert_eq!(docset.skip_next(131), SkipResult::End);
+        }
+    }
+
+    fn build_block_postings(docs: &[DocId]) -> BlockSegmentPostings {
         let mut schema_builder = Schema::builder();
         let int_field = schema_builder.add_u64_field("id", INT_INDEXED);
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         let mut last_doc = 0u32;
-        for doc in docs {
+        for &doc in docs {
             for _ in last_doc..doc {
                 index_writer.add_document(doc!(int_field=>1u64));
             }
@@ -734,7 +770,7 @@ mod tests {
     #[test]
     fn test_block_segment_postings_skip() {
         for i in 0..4 {
-            let mut block_postings = build_block_postings(vec![3]);
+            let mut block_postings = build_block_postings(&[3]);
             assert_eq!(
                 block_postings.skip_to(i),
                 BlockSegmentPostingsSkipResult::Success(0u32)
@@ -744,7 +780,7 @@ mod tests {
                 BlockSegmentPostingsSkipResult::Terminated
             );
         }
-        let mut block_postings = build_block_postings(vec![3]);
+        let mut block_postings = build_block_postings(&[3]);
         assert_eq!(
             block_postings.skip_to(4u32),
             BlockSegmentPostingsSkipResult::Terminated
@@ -757,7 +793,7 @@ mod tests {
         for i in 0..1300 {
             docs.push((i * i / 100) + i);
         }
-        let mut block_postings = build_block_postings(docs.clone());
+        let mut block_postings = build_block_postings(&docs[..]);
         for i in vec![0, 424, 10000] {
             assert_eq!(
                 block_postings.skip_to(i),
@@ -783,7 +819,7 @@ mod tests {
         let int_field = schema_builder.add_u64_field("id", INT_INDEXED);
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
-        let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         // create two postings list, one containg even number,
         // the other containing odd numbers.
         for i in 0..6 {

src/postings/stacker/expull.rs

@@ -1,28 +1,37 @@
 use super::{Addr, MemoryArena};
 
-use common::is_power_of_2;
+use postings::stacker::memory_arena::load;
+use postings::stacker::memory_arena::store;
+use std::io;
 use std::mem;
 
 const MAX_BLOCK_LEN: u32 = 1u32 << 15;
+const FIRST_BLOCK: usize = 16;
+const INLINED_BLOCK_LEN: usize = FIRST_BLOCK + mem::size_of::<Addr>();
 
-const FIRST_BLOCK: u32 = 4u32;
+enum CapacityResult {
+    Available(u32),
+    NeedAlloc(u32),
+}
 
-#[inline]
-pub fn jump_needed(len: u32) -> Option<usize> {
+fn len_to_capacity(len: u32) -> CapacityResult {
     match len {
-        0...3 => None,
-        4...MAX_BLOCK_LEN => {
-            if is_power_of_2(len as usize) {
-                Some(len as usize)
+        0...15 => CapacityResult::Available(FIRST_BLOCK as u32 - len),
+        16...MAX_BLOCK_LEN => {
+            let cap = 1 << (32u32 - (len - 1u32).leading_zeros());
+            let available = cap - len;
+            if available == 0 {
+                CapacityResult::NeedAlloc(len)
             } else {
-                None
+                CapacityResult::Available(available)
             }
         }
         n => {
-            if n % MAX_BLOCK_LEN == 0 {
-                Some(MAX_BLOCK_LEN as usize)
+            let available = n % MAX_BLOCK_LEN;
+            if available == 0 {
+                CapacityResult::NeedAlloc(MAX_BLOCK_LEN)
             } else {
-                None
+                CapacityResult::Available(MAX_BLOCK_LEN - available)
             }
         }
     }
@@ -52,82 +61,119 @@ pub fn jump_needed(len: u32) -> Option<usize> {
 #[derive(Debug, Clone, Copy)]
 pub struct ExpUnrolledLinkedList {
     len: u32,
-    head: Addr,
     tail: Addr,
+    inlined_data: [u8; INLINED_BLOCK_LEN as usize],
+}
+
+pub struct ExpUnrolledLinkedListWriter<'a> {
+    eull: &'a mut ExpUnrolledLinkedList,
+    heap: &'a mut MemoryArena,
+}
+
+fn ensure_capacity<'a>(
+    eull: &'a mut ExpUnrolledLinkedList,
+    heap: &'a mut MemoryArena,
+) -> &'a mut [u8] {
+    if eull.len <= FIRST_BLOCK as u32 {
+        // We are still hitting the inline block.
+        if eull.len < FIRST_BLOCK as u32 {
+            return &mut eull.inlined_data[eull.len as usize..FIRST_BLOCK];
+        }
+        // We need to allocate a new block!
+        let new_block_addr: Addr = heap.allocate_space(FIRST_BLOCK + mem::size_of::<Addr>());
+        store(&mut eull.inlined_data[FIRST_BLOCK..], new_block_addr);
+        eull.tail = new_block_addr;
+        return heap.slice_mut(eull.tail, FIRST_BLOCK);
+    }
+    let len = match len_to_capacity(eull.len) {
+        CapacityResult::NeedAlloc(new_block_len) => {
+            let new_block_addr: Addr =
+                heap.allocate_space(new_block_len as usize + mem::size_of::<Addr>());
+            heap.write_at(eull.tail, new_block_addr);
+            eull.tail = new_block_addr;
+            new_block_len
+        }
+        CapacityResult::Available(available) => available,
+    };
+    heap.slice_mut(eull.tail, len as usize)
+}
+
+impl<'a> ExpUnrolledLinkedListWriter<'a> {
+    pub fn extend_from_slice(&mut self, mut buf: &[u8]) {
+        if buf.is_empty() {
+            // we need to cut early, because `ensure_capacity`
+            // allocates if there is no capacity at all right now.
+            return;
+        }
+        while !buf.is_empty() {
+            let add_len: usize;
+            {
+                let output_buf = ensure_capacity(self.eull, self.heap);
+                add_len = buf.len().min(output_buf.len());
+                output_buf[..add_len].copy_from_slice(&buf[..add_len]);
+            }
+            self.eull.len += add_len as u32;
+            self.eull.tail = self.eull.tail.offset(add_len as u32);
+            buf = &buf[add_len..];
+        }
+    }
+}
+
+impl<'a> io::Write for ExpUnrolledLinkedListWriter<'a> {
+    fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
+        // There is no use case to only write the capacity.
+        // This is not IO after all, so we write the whole
+        // buffer even if the contract of `.write` is looser.
+        self.extend_from_slice(buf);
+        Ok(buf.len())
+    }
+
+    fn write_all(&mut self, buf: &[u8]) -> io::Result<()> {
+        self.extend_from_slice(buf);
+        Ok(())
+    }
+
+    fn flush(&mut self) -> io::Result<()> {
+        Ok(())
+    }
 }
 
 impl ExpUnrolledLinkedList {
-    pub fn new(heap: &mut MemoryArena) -> ExpUnrolledLinkedList {
-        let addr = heap.allocate_space((FIRST_BLOCK as usize) * mem::size_of::<u32>());
+    pub fn new() -> ExpUnrolledLinkedList {
         ExpUnrolledLinkedList {
             len: 0u32,
-            head: addr,
-            tail: addr,
+            tail: Addr::null_pointer(),
+            inlined_data: [0u8; INLINED_BLOCK_LEN as usize],
         }
     }
 
-    pub fn iter<'a>(&self, heap: &'a MemoryArena) -> ExpUnrolledLinkedListIterator<'a> {
-        ExpUnrolledLinkedListIterator {
-            heap,
-            addr: self.head,
-            len: self.len,
-            consumed: 0,
-        }
+    #[inline(always)]
+    pub fn writer<'a>(&'a mut self, heap: &'a mut MemoryArena) -> ExpUnrolledLinkedListWriter<'a> {
+        ExpUnrolledLinkedListWriter { eull: self, heap }
     }
 
-    /// Appends a new element to the current stack.
-    ///
-    /// If the current block end is reached, a new block is allocated.
-    pub fn push(&mut self, val: u32, heap: &mut MemoryArena) {
-        self.len += 1;
-        if let Some(new_block_len) = jump_needed(self.len) {
-            // We need to allocate another block.
-            // We also allocate an extra `u32` to store the pointer
-            // to the future next block.
-            let new_block_size: usize = (new_block_len + 1) * mem::size_of::<u32>();
-            let new_block_addr: Addr = heap.allocate_space(new_block_size);
-            unsafe {
-                // logic
-                heap.write(self.tail, new_block_addr)
-            };
-            self.tail = new_block_addr;
+    pub fn read_to_end(&self, heap: &MemoryArena, output: &mut Vec<u8>) {
+        let len = self.len as usize;
+        if len <= FIRST_BLOCK {
+            output.extend_from_slice(&self.inlined_data[..len]);
+            return;
         }
-        unsafe {
-            // logic
-            heap.write(self.tail, val);
-            self.tail = self.tail.offset(mem::size_of::<u32>() as u32);
-        }
-    }
-}
-
-pub struct ExpUnrolledLinkedListIterator<'a> {
-    heap: &'a MemoryArena,
-    addr: Addr,
-    len: u32,
-    consumed: u32,
-}
-
-impl<'a> Iterator for ExpUnrolledLinkedListIterator<'a> {
-    type Item = u32;
-
-    fn next(&mut self) -> Option<u32> {
-        if self.consumed == self.len {
-            None
-        } else {
-            self.consumed += 1;
-            let addr: Addr = if jump_needed(self.consumed).is_some() {
-                unsafe {
-                    // logic
-                    self.heap.read(self.addr)
-                }
-            } else {
-                self.addr
-            };
-            self.addr = addr.offset(mem::size_of::<u32>() as u32);
-            Some(unsafe {
-                // logic
-                self.heap.read(addr)
-            })
+        output.extend_from_slice(&self.inlined_data[..FIRST_BLOCK]);
+        let mut cur = FIRST_BLOCK;
+        let mut addr = load(&self.inlined_data[FIRST_BLOCK..]);
+        loop {
+            let cap = match len_to_capacity(cur as u32) {
+                CapacityResult::Available(capacity) => capacity,
+                CapacityResult::NeedAlloc(capacity) => capacity,
+            } as usize;
+            let data = heap.slice(addr, cap);
+            if cur + cap >= len {
+                output.extend_from_slice(&data[..(len - cur)]);
+                return;
+            }
+            output.extend_from_slice(data);
+            cur += cap;
+            addr = heap.read(addr.offset(cap as u32));
         }
     }
 }
@@ -136,39 +182,126 @@ impl<'a> Iterator for ExpUnrolledLinkedListIterator<'a> {
 mod tests {
 
     use super::super::MemoryArena;
-    use super::jump_needed;
+    use super::len_to_capacity;
     use super::*;
+    use byteorder::{ByteOrder, LittleEndian, WriteBytesExt};
 
+    #[test]
     #[test]
     fn test_stack() {
         let mut heap = MemoryArena::new();
-        let mut stack = ExpUnrolledLinkedList::new(&mut heap);
-        stack.push(1u32, &mut heap);
-        stack.push(2u32, &mut heap);
-        stack.push(4u32, &mut heap);
-        stack.push(8u32, &mut heap);
+        let mut stack = ExpUnrolledLinkedList::new();
+        stack.writer(&mut heap).extend_from_slice(&[1u8]);
+        stack.writer(&mut heap).extend_from_slice(&[2u8]);
+        stack.writer(&mut heap).extend_from_slice(&[3u8, 4u8]);
+        stack.writer(&mut heap).extend_from_slice(&[5u8]);
         {
-            let mut it = stack.iter(&heap);
-            assert_eq!(it.next().unwrap(), 1u32);
-            assert_eq!(it.next().unwrap(), 2u32);
-            assert_eq!(it.next().unwrap(), 4u32);
-            assert_eq!(it.next().unwrap(), 8u32);
-            assert!(it.next().is_none());
+            let mut buffer = Vec::new();
+            stack.read_to_end(&heap, &mut buffer);
+            assert_eq!(&buffer[..], &[1u8, 2u8, 3u8, 4u8, 5u8]);
         }
     }
 
     #[test]
-    fn test_jump_if_needed() {
-        let mut block_len = 4u32;
-        let mut i = 0;
-        while i < 10_000_000 {
-            assert!(jump_needed(i + block_len - 1).is_none());
-            assert!(jump_needed(i + block_len + 1).is_none());
-            assert!(jump_needed(i + block_len).is_some());
-            let new_block_len = jump_needed(i + block_len).unwrap();
-            i += block_len;
-            block_len = new_block_len as u32;
+    fn test_stack_long() {
+        let mut heap = MemoryArena::new();
+        let mut stack = ExpUnrolledLinkedList::new();
+        let source: Vec<u32> = (0..100).collect();
+        for &el in &source {
+            assert!(stack
+                .writer(&mut heap)
+                .write_u32::<LittleEndian>(el)
+                .is_ok());
         }
+        let mut buffer = Vec::new();
+        stack.read_to_end(&heap, &mut buffer);
+        let mut result = vec![];
+        let mut remaining = &buffer[..];
+        while !remaining.is_empty() {
+            result.push(LittleEndian::read_u32(&remaining[..4]));
+            remaining = &remaining[4..];
+        }
+        assert_eq!(&result[..], &source[..]);
+    }
+
+    #[test]
+    fn test_stack_interlaced() {
+        let mut heap = MemoryArena::new();
+        let mut stack = ExpUnrolledLinkedList::new();
+        let mut stack2 = ExpUnrolledLinkedList::new();
+
+        let mut vec1: Vec<u8> = vec![];
+        let mut vec2: Vec<u8> = vec![];
+
+        for i in 0..9 {
+            assert!(stack.writer(&mut heap).write_u32::<LittleEndian>(i).is_ok());
+            assert!(vec1.write_u32::<LittleEndian>(i).is_ok());
+            if i % 2 == 0 {
+                assert!(stack2
+                    .writer(&mut heap)
+                    .write_u32::<LittleEndian>(i)
+                    .is_ok());
+                assert!(vec2.write_u32::<LittleEndian>(i).is_ok());
+            }
+        }
+        let mut res1 = vec![];
+        let mut res2 = vec![];
+        stack.read_to_end(&heap, &mut res1);
+        stack2.read_to_end(&heap, &mut res2);
+        assert_eq!(&vec1[..], &res1[..]);
+        assert_eq!(&vec2[..], &res2[..]);
+    }
+
+    #[test]
+    fn test_jump_if_needed() {
+        let mut available = 16u32;
+        for i in 0..10_000_000 {
+            match len_to_capacity(i) {
+                CapacityResult::NeedAlloc(cap) => {
+                    assert_eq!(available, 0, "Failed len={}: Expected 0 got {}", i, cap);
+                    available = cap;
+                }
+                CapacityResult::Available(cap) => {
+                    assert_eq!(
+                        available, cap,
+                        "Failed len={}: Expected {} Got {}",
+                        i, available, cap
+                    );
+                }
+            }
+            available -= 1;
+        }
+    }
+
+    #[test]
+    fn test_jump_if_needed_progression() {
+        let mut v = vec![];
+        for i in 0.. {
+            if v.len() >= 10 {
+                break;
+            }
+            match len_to_capacity(i) {
+                CapacityResult::NeedAlloc(cap) => {
+                    v.push((i, cap));
+                }
+                _ => {}
+            }
+        }
+        assert_eq!(
+            &v[..],
+            &[
+                (16, 16),
+                (32, 32),
+                (64, 64),
+                (128, 128),
+                (256, 256),
+                (512, 512),
+                (1024, 1024),
+                (2048, 2048),
+                (4096, 4096),
+                (8192, 8192)
+            ]
+        );
     }
 }
 
@@ -176,6 +309,7 @@ mod tests {
 mod bench {
     use super::super::MemoryArena;
     use super::ExpUnrolledLinkedList;
+    use byteorder::{NativeEndian, WriteBytesExt};
     use test::Bencher;
 
     const NUM_STACK: usize = 10_000;
@@ -203,13 +337,13 @@ mod bench {
             let mut heap = MemoryArena::new();
             let mut stacks = Vec::with_capacity(100);
             for _ in 0..NUM_STACK {
-                let mut stack = ExpUnrolledLinkedList::new(&mut heap);
+                let mut stack = ExpUnrolledLinkedList::new();
                 stacks.push(stack);
             }
             for s in 0..NUM_STACK {
                 for i in 0u32..STACK_SIZE {
                     let t = s * 392017 % NUM_STACK;
-                    stacks[t].push(i, &mut heap);
+                    let _ = stacks[t].writer(&mut heap).write_u32::<NativeEndian>(i);
                 }
             }
         });

src/postings/stacker/memory_arena.rs

@@ -37,7 +37,7 @@ const PAGE_SIZE: usize = 1 << NUM_BITS_PAGE_ADDR; // pages are 1 MB large
 /// page of memory.
 ///
 /// The last 20 bits are an address within this page of memory.
-#[derive(Clone, Copy, Debug)]
+#[derive(Copy, Clone, Debug)]
 pub struct Addr(u32);
 
 impl Addr {
@@ -69,32 +69,16 @@ impl Addr {
     }
 }
 
-/// Trait required for an object to be `storable`.
-///
-/// # Warning
-///
-/// Most of the time you should not implement this trait,
-/// and only use the `MemoryArena` with object implementing `Copy`.
-///
-/// `ArenaStorable` is used in `tantivy` to force
-/// a `Copy` object and a `slice` of data to be stored contiguously.
-pub trait ArenaStorable {
-    fn num_bytes(&self) -> usize;
-    unsafe fn write_into(self, arena: &mut MemoryArena, addr: Addr);
+pub fn store<Item: Copy + 'static>(dest: &mut [u8], val: Item) {
+    assert_eq!(dest.len(), std::mem::size_of::<Item>());
+    unsafe {
+        ptr::write_unaligned(dest.as_mut_ptr() as *mut Item, val);
+    }
 }
 
-impl<V> ArenaStorable for V
-where
-    V: Copy,
-{
-    fn num_bytes(&self) -> usize {
-        mem::size_of::<V>()
-    }
-
-    unsafe fn write_into(self, arena: &mut MemoryArena, addr: Addr) {
-        let dst_ptr = arena.get_mut_ptr(addr) as *mut V;
-        ptr::write_unaligned(dst_ptr, self);
-    }
+pub fn load<Item: Copy + 'static>(data: &[u8]) -> Item {
+    assert_eq!(data.len(), std::mem::size_of::<Item>());
+    unsafe { ptr::read_unaligned(data.as_ptr() as *const Item) }
 }
 
 /// The `MemoryArena`
@@ -126,47 +110,9 @@ impl MemoryArena {
         self.pages.len() * PAGE_SIZE
     }
 
-    /// Writes a slice at the given address, assuming the
-    /// memory was allocated beforehands.
-    ///
-    ///  # Panics
-    ///
-    /// May panic or corrupt the heap if he space was not
-    /// properly allocated beforehands.
-    pub fn write_bytes<B: AsRef<[u8]>>(&mut self, addr: Addr, data: B) {
-        let bytes = data.as_ref();
-        self.pages[addr.page_id()]
-            .get_mut_slice(addr.page_local_addr(), bytes.len())
-            .copy_from_slice(bytes);
-    }
-
-    /// Returns the `len` bytes starting at `addr`
-    ///
-    /// # Panics
-    ///
-    /// Panics if the memory has not been allocated beforehands.
-    pub fn read_slice(&self, addr: Addr, len: usize) -> &[u8] {
-        self.pages[addr.page_id()].get_slice(addr.page_local_addr(), len)
-    }
-
-    unsafe fn get_mut_ptr(&mut self, addr: Addr) -> *mut u8 {
-        self.pages[addr.page_id()].get_mut_ptr(addr.page_local_addr())
-    }
-
-    /// Stores an item's data in the heap
-    ///
-    /// It allocates the `Item` beforehands.
-    pub fn store<Item: ArenaStorable>(&mut self, val: Item) -> Addr {
-        let num_bytes = val.num_bytes();
-        let addr = self.allocate_space(num_bytes);
-        unsafe {
-            self.write(addr, val);
-        };
-        addr
-    }
-
-    pub unsafe fn write<Item: ArenaStorable>(&mut self, addr: Addr, val: Item) {
-        val.write_into(self, addr)
+    pub fn write_at<Item: Copy + 'static>(&mut self, addr: Addr, val: Item) {
+        let dest = self.slice_mut(addr, std::mem::size_of::<Item>());
+        store(dest, val);
     }
 
     /// Read an item in the heap at the given `address`.
@@ -174,9 +120,21 @@ impl MemoryArena {
     /// # Panics
     ///
     /// If the address is erroneous
-    pub unsafe fn read<Item: Copy>(&self, addr: Addr) -> Item {
-        let ptr = self.pages[addr.page_id()].get_ptr(addr.page_local_addr());
-        ptr::read_unaligned(ptr as *const Item)
+    pub fn read<Item: Copy + 'static>(&self, addr: Addr) -> Item {
+        load(self.slice(addr, mem::size_of::<Item>()))
+    }
+
+    pub fn slice(&self, addr: Addr, len: usize) -> &[u8] {
+        self.pages[addr.page_id()].slice(addr.page_local_addr(), len)
+    }
+
+    pub fn slice_from(&self, addr: Addr) -> &[u8] {
+        self.pages[addr.page_id()].slice_from(addr.page_local_addr())
+    }
+
+    #[inline(always)]
+    pub fn slice_mut(&mut self, addr: Addr, len: usize) -> &mut [u8] {
+        self.pages[addr.page_id()].slice_mut(addr.page_local_addr(), len)
     }
 
     /// Allocates `len` bytes and returns the allocated address.
@@ -197,14 +155,10 @@ struct Page {
 
 impl Page {
     fn new(page_id: usize) -> Page {
-        let mut data: Vec<u8> = Vec::with_capacity(PAGE_SIZE);
-        unsafe {
-            data.set_len(PAGE_SIZE);
-        } // avoid initializing page
         Page {
             page_id,
             len: 0,
-            data: data.into_boxed_slice(),
+            data: vec![0u8; PAGE_SIZE].into_boxed_slice(),
         }
     }
 
@@ -213,12 +167,16 @@ impl Page {
         len + self.len <= PAGE_SIZE
     }
 
-    fn get_mut_slice(&mut self, local_addr: usize, len: usize) -> &mut [u8] {
-        &mut self.data[local_addr..][..len]
+    fn slice(&self, local_addr: usize, len: usize) -> &[u8] {
+        &self.slice_from(local_addr)[..len]
     }
 
-    fn get_slice(&self, local_addr: usize, len: usize) -> &[u8] {
-        &self.data[local_addr..][..len]
+    fn slice_from(&self, local_addr: usize) -> &[u8] {
+        &self.data[local_addr..]
+    }
+
+    fn slice_mut(&mut self, local_addr: usize, len: usize) -> &mut [u8] {
+        &mut self.data[local_addr..][..len]
     }
 
     fn allocate_space(&mut self, len: usize) -> Option<Addr> {
@@ -230,16 +188,6 @@ impl Page {
             None
         }
     }
-
-    #[inline(always)]
-    pub(crate) unsafe fn get_ptr(&self, addr: usize) -> *const u8 {
-        self.data.as_ptr().add(addr)
-    }
-
-    #[inline(always)]
-    pub(crate) unsafe fn get_mut_ptr(&mut self, addr: usize) -> *mut u8 {
-        self.data.as_mut_ptr().add(addr)
-    }
 }
 
 #[cfg(test)]
@@ -254,13 +202,13 @@ mod tests {
         let b = b"happy tax payer";
 
         let addr_a = arena.allocate_space(a.len());
-        arena.write_bytes(addr_a, a);
+        arena.slice_mut(addr_a, a.len()).copy_from_slice(a);
 
         let addr_b = arena.allocate_space(b.len());
-        arena.write_bytes(addr_b, b);
+        arena.slice_mut(addr_b, b.len()).copy_from_slice(b);
 
-        assert_eq!(arena.read_slice(addr_a, a.len()), a);
-        assert_eq!(arena.read_slice(addr_b, b.len()), b);
+        assert_eq!(arena.slice(addr_a, a.len()), a);
+        assert_eq!(arena.slice(addr_b, b.len()), b);
     }
 
     #[derive(Clone, Copy, Debug, Eq, PartialEq)]
@@ -283,9 +231,15 @@ mod tests {
             b: 221,
             c: 12,
         };
-        let addr_a = arena.store(a);
-        let addr_b = arena.store(b);
-        assert_eq!(unsafe { arena.read::<MyTest>(addr_a) }, a);
-        assert_eq!(unsafe { arena.read::<MyTest>(addr_b) }, b);
+
+        let num_bytes = std::mem::size_of::<MyTest>();
+        let addr_a = arena.allocate_space(num_bytes);
+        arena.write_at(addr_a, a);
+
+        let addr_b = arena.allocate_space(num_bytes);
+        arena.write_at(addr_b, b);
+
+        assert_eq!(arena.read::<MyTest>(addr_a), a);
+        assert_eq!(arena.read::<MyTest>(addr_b), b);
     }
 }

src/postings/stacker/mod.rs

@@ -1,9 +1,7 @@
 mod expull;
 mod memory_arena;
-mod murmurhash2;
 mod term_hashmap;
 
 pub use self::expull::ExpUnrolledLinkedList;
-pub use self::memory_arena::{Addr, ArenaStorable, MemoryArena};
-use self::murmurhash2::murmurhash2;
+pub use self::memory_arena::{Addr, MemoryArena};
 pub use self::term_hashmap::{compute_table_size, TermHashMap};

src/postings/stacker/murmurhash2.rs

@@ -1,87 +0,0 @@
-use std::ptr;
-const SEED: u32 = 3_242_157_231u32;
-const M: u32 = 0x5bd1_e995;
-
-#[inline(always)]
-pub fn murmurhash2(key: &[u8]) -> u32 {
-    #[cfg_attr(feature = "cargo-clippy", allow(clippy::cast_ptr_alignment))]
-    let mut key_ptr: *const u32 = key.as_ptr() as *const u32;
-    let len = key.len() as u32;
-    let mut h: u32 = SEED ^ len;
-
-    let num_blocks = len >> 2;
-    for _ in 0..num_blocks {
-        let mut k: u32 = unsafe { ptr::read_unaligned(key_ptr) }; // ok because of num_blocks definition
-        k = k.wrapping_mul(M);
-        k ^= k >> 24;
-        k = k.wrapping_mul(M);
-        h = h.wrapping_mul(M);
-        h ^= k;
-        key_ptr = key_ptr.wrapping_offset(1);
-    }
-
-    // Handle the last few bytes of the input array
-    let remaining: &[u8] = &key[key.len() & !3..];
-    match remaining.len() {
-        3 => {
-            h ^= u32::from(remaining[2]) << 16;
-            h ^= u32::from(remaining[1]) << 8;
-            h ^= u32::from(remaining[0]);
-            h = h.wrapping_mul(M);
-        }
-        2 => {
-            h ^= u32::from(remaining[1]) << 8;
-            h ^= u32::from(remaining[0]);
-            h = h.wrapping_mul(M);
-        }
-        1 => {
-            h ^= u32::from(remaining[0]);
-            h = h.wrapping_mul(M);
-        }
-        _ => {}
-    }
-    h ^= h >> 13;
-    h = h.wrapping_mul(M);
-    h ^ (h >> 15)
-}
-
-#[cfg(test)]
-mod test {
-
-    use super::murmurhash2;
-    use std::collections::HashSet;
-
-    #[test]
-    fn test_murmur() {
-        let s1 = "abcdef";
-        let s2 = "abcdeg";
-        for i in 0..5 {
-            assert_eq!(
-                murmurhash2(&s1[i..5].as_bytes()),
-                murmurhash2(&s2[i..5].as_bytes())
-            );
-        }
-    }
-
-    #[test]
-    fn test_murmur_against_reference_impl() {
-        assert_eq!(murmurhash2("".as_bytes()), 3632506080);
-        assert_eq!(murmurhash2("a".as_bytes()), 455683869);
-        assert_eq!(murmurhash2("ab".as_bytes()), 2448092234);
-        assert_eq!(murmurhash2("abc".as_bytes()), 2066295634);
-        assert_eq!(murmurhash2("abcd".as_bytes()), 2588571162);
-        assert_eq!(murmurhash2("abcde".as_bytes()), 2988696942);
-        assert_eq!(murmurhash2("abcdefghijklmnop".as_bytes()), 2350868870);
-    }
-
-    #[test]
-    fn test_murmur_collisions() {
-        let mut set: HashSet<u32> = HashSet::default();
-        for i in 0..10_000 {
-            let s = format!("hash{}", i);
-            let hash = murmurhash2(s.as_bytes());
-            set.insert(hash);
-        }
-        assert_eq!(set.len(), 10_000);
-    }
-}

src/postings/stacker/term_hashmap.rs

@@ -1,37 +1,16 @@
-use super::murmurhash2;
-use super::{Addr, ArenaStorable, MemoryArena};
+extern crate murmurhash32;
+
+use self::murmurhash32::murmurhash2;
+
+use super::{Addr, MemoryArena};
+use byteorder::{ByteOrder, NativeEndian};
+use postings::stacker::memory_arena::store;
 use std::iter;
 use std::mem;
 use std::slice;
 
 pub type BucketId = usize;
 
-struct KeyBytesValue<'a, V> {
-    key: &'a [u8],
-    value: V,
-}
-
-impl<'a, V> KeyBytesValue<'a, V> {
-    fn new(key: &'a [u8], value: V) -> KeyBytesValue<'a, V> {
-        KeyBytesValue { key, value }
-    }
-}
-
-impl<'a, V> ArenaStorable for KeyBytesValue<'a, V>
-where
-    V: ArenaStorable,
-{
-    fn num_bytes(&self) -> usize {
-        0u16.num_bytes() + self.key.len() + self.value.num_bytes()
-    }
-
-    unsafe fn write_into(self, arena: &mut MemoryArena, addr: Addr) {
-        arena.write(addr, self.key.len() as u16);
-        arena.write_bytes(addr.offset(2), self.key);
-        arena.write(addr.offset(2 + self.key.len() as u32), self.value);
-    }
-}
-
 /// Returns the actual memory size in bytes
 /// required to create a table of size $2^num_bits$.
 pub fn compute_table_size(num_bits: usize) -> usize {
@@ -111,8 +90,7 @@ impl<'a> Iterator for Iter<'a> {
     fn next(&mut self) -> Option<Self::Item> {
         self.inner.next().cloned().map(move |bucket: usize| {
             let kv = self.hashmap.table[bucket];
-            let (key, offset): (&'a [u8], Addr) =
-                unsafe { self.hashmap.get_key_value(kv.key_value_addr) };
+            let (key, offset): (&'a [u8], Addr) = self.hashmap.get_key_value(kv.key_value_addr);
             (key, offset, bucket as BucketId)
         })
     }
@@ -143,12 +121,22 @@ impl TermHashMap {
         self.table.len() < self.occupied.len() * 3
     }
 
-    unsafe fn get_key_value(&self, addr: Addr) -> (&[u8], Addr) {
-        let key_bytes_len = self.heap.read::<u16>(addr) as usize;
-        let key_addr = addr.offset(2u32);
-        let key_bytes: &[u8] = self.heap.read_slice(key_addr, key_bytes_len);
-        let val_addr: Addr = key_addr.offset(key_bytes.len() as u32);
-        (key_bytes, val_addr)
+    #[inline(always)]
+    fn get_key_value(&self, addr: Addr) -> (&[u8], Addr) {
+        let data = self.heap.slice_from(addr);
+        let key_bytes_len = NativeEndian::read_u16(data) as usize;
+        let key_bytes: &[u8] = &data[2..][..key_bytes_len];
+        (key_bytes, addr.offset(2u32 + key_bytes_len as u32))
+    }
+
+    #[inline(always)]
+    fn get_value_addr_if_key_match(&self, target_key: &[u8], addr: Addr) -> Option<Addr> {
+        let (stored_key, value_addr) = self.get_key_value(addr);
+        if stored_key == target_key {
+            Some(value_addr)
+        } else {
+            None
+        }
     }
 
     pub fn set_bucket(&mut self, hash: u32, key_value_addr: Addr, bucket: usize) {
@@ -199,36 +187,39 @@ impl TermHashMap {
     pub fn mutate_or_create<S, V, TMutator>(&mut self, key: S, mut updater: TMutator) -> BucketId
     where
         S: AsRef<[u8]>,
-        V: Copy,
+        V: Copy + 'static,
         TMutator: FnMut(Option<V>) -> V,
     {
         if self.is_saturated() {
             self.resize();
         }
         let key_bytes: &[u8] = key.as_ref();
-        let hash = murmurhash2::murmurhash2(key.as_ref());
+        let hash = murmurhash2(key.as_ref());
         let mut probe = self.probe(hash);
         loop {
             let bucket = probe.next_probe();
             let kv: KeyValue = self.table[bucket];
             if kv.is_empty() {
                 let val = updater(None);
-                let key_addr = self.heap.store(KeyBytesValue::new(key_bytes, val));
+                let num_bytes =
+                    std::mem::size_of::<u16>() + key_bytes.len() + std::mem::size_of::<V>();
+                let key_addr = self.heap.allocate_space(num_bytes);
+                {
+                    let data = self.heap.slice_mut(key_addr, num_bytes);
+                    NativeEndian::write_u16(data, key_bytes.len() as u16);
+                    let stop = 2 + key_bytes.len();
+                    data[2..stop].copy_from_slice(key_bytes);
+                    store(&mut data[stop..], val);
+                }
                 self.set_bucket(hash, key_addr, bucket);
                 return bucket as BucketId;
             } else if kv.hash == hash {
-                let (key_matches, val_addr) = {
-                    let (stored_key, val_addr): (&[u8], Addr) =
-                        unsafe { self.get_key_value(kv.key_value_addr) };
-                    (stored_key == key_bytes, val_addr)
-                };
-                if key_matches {
-                    unsafe {
-                        // logic
-                        let v = self.heap.read(val_addr);
-                        let new_v = updater(Some(v));
-                        self.heap.write(val_addr, new_v);
-                    };
+                if let Some(val_addr) =
+                    self.get_value_addr_if_key_match(key_bytes, kv.key_value_addr)
+                {
+                    let v = self.heap.read(val_addr);
+                    let new_v = updater(Some(v));
+                    self.heap.write_at(val_addr, new_v);
                     return bucket as BucketId;
                 }
             }
@@ -236,24 +227,6 @@ impl TermHashMap {
     }
 }
 
-#[cfg(all(test, feature = "unstable"))]
-mod bench {
-    use super::murmurhash2::murmurhash2;
-    use test::Bencher;
-
-    #[bench]
-    fn bench_murmurhash2(b: &mut Bencher) {
-        let keys: [&'static str; 3] = ["wer qwe qwe qwe ", "werbq weqweqwe2 ", "weraq weqweqwe3 "];
-        b.iter(|| {
-            let mut s = 0;
-            for &key in &keys {
-                s ^= murmurhash2(key.as_bytes());
-            }
-            s
-        });
-    }
-}
-
 #[cfg(test)]
 mod tests {
 
@@ -285,10 +258,7 @@ mod tests {
         let mut vanilla_hash_map = HashMap::new();
         let mut iter_values = hash_map.iter();
         while let Some((key, addr, _)) = iter_values.next() {
-            let val: u32 = unsafe {
-                // test
-                hash_map.heap.read(addr)
-            };
+            let val: u32 = hash_map.heap.read(addr);
             vanilla_hash_map.insert(key.to_owned(), val);
         }
         assert_eq!(vanilla_hash_map.len(), 2);

src/query/boolean_query/boolean_weight.rs

@@ -1,5 +1,5 @@
 use core::SegmentReader;
-use downcast::Downcast;
+use downcast_rs::Downcast;
 use query::intersect_scorers;
 use query::score_combiner::{DoNothingCombiner, ScoreCombiner, SumWithCoordsCombiner};
 use query::term_query::TermScorer;
@@ -10,7 +10,6 @@ use query::RequiredOptionalScorer;
 use query::Scorer;
 use query::Union;
 use query::Weight;
-use std::borrow::Borrow;
 use std::collections::HashMap;
 use Result;
 
@@ -25,13 +24,12 @@ where
 
     {
         let is_all_term_queries = scorers.iter().all(|scorer| {
-            let scorer_ref: &Scorer = scorer.borrow();
-            Downcast::<TermScorer>::is_type(scorer_ref)
+            scorer.is::<TermScorer>()
         });
         if is_all_term_queries {
             let scorers: Vec<TermScorer> = scorers
                 .into_iter()
-                .map(|scorer| *Downcast::<TermScorer>::downcast(scorer).unwrap())
+                .map(|scorer| *(scorer.downcast::<TermScorer>().map_err(|_| ()).unwrap() ))
                 .collect();
             let scorer: Box<Scorer> = Box::new(Union::<TermScorer, TScoreCombiner>::from(scorers));
             return scorer;

src/query/boolean_query/mod.rs

@@ -8,7 +8,7 @@ mod tests {
 
     use super::*;
     use collector::tests::TestCollector;
-    use downcast::Downcast;
+    use downcast_rs::Downcast;
     use query::score_combiner::SumWithCoordsCombiner;
     use query::term_query::TermScorer;
     use query::Intersection;
@@ -29,7 +29,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field => "a b c");
                 index_writer.add_document(doc);
@@ -72,7 +72,7 @@ mod tests {
         let searcher = index.searcher();
         let weight = query.weight(&searcher, true).unwrap();
         let scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
-        assert!(Downcast::<TermScorer>::is_type(&*scorer));
+        assert!(scorer.is::<TermScorer>());
     }
 
     #[test]
@@ -84,13 +84,13 @@ mod tests {
             let query = query_parser.parse_query("+a +b +c").unwrap();
             let weight = query.weight(&searcher, true).unwrap();
             let scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
-            assert!(Downcast::<Intersection<TermScorer>>::is_type(&*scorer));
+            assert!(scorer.is::<Intersection<TermScorer>>());
         }
         {
             let query = query_parser.parse_query("+a +(b c)").unwrap();
             let weight = query.weight(&searcher, true).unwrap();
             let scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
-            assert!(Downcast::<Intersection<Box<Scorer>>>::is_type(&*scorer));
+            assert!(scorer.is::<Intersection<Box<Scorer>>>());
         }
     }
 
@@ -103,18 +103,16 @@ mod tests {
             let query = query_parser.parse_query("+a b").unwrap();
             let weight = query.weight(&searcher, true).unwrap();
             let scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
-            assert!(Downcast::<
-                RequiredOptionalScorer<Box<Scorer>, Box<Scorer>, SumWithCoordsCombiner>,
-            >::is_type(&*scorer));
+            assert!(scorer.is::<RequiredOptionalScorer<Box<Scorer>, Box<Scorer>, SumWithCoordsCombiner>>());
         }
         {
             let query = query_parser.parse_query("+a b").unwrap();
             let weight = query.weight(&searcher, false).unwrap();
             let scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
-            println!("{:?}", scorer.type_name());
-            assert!(Downcast::<TermScorer>::is_type(&*scorer));
+            assert!(scorer.is::<TermScorer>());
         }
-    }
+     }
+
 
     #[test]
     pub fn test_boolean_query() {

src/query/intersection.rs

@@ -1,11 +1,9 @@
 use docset::{DocSet, SkipResult};
-use downcast::Downcast;
-use query::term_query::TermScorer;
 use query::EmptyScorer;
 use query::Scorer;
-use std::borrow::Borrow;
 use DocId;
 use Score;
+use query::term_query::TermScorer;
 
 /// Returns the intersection scorer.
 ///
@@ -26,13 +24,12 @@ pub fn intersect_scorers(mut scorers: Vec<Box<Scorer>>) -> Box<Scorer> {
         (Some(single_docset), None) => single_docset,
         (Some(left), Some(right)) => {
             {
-                let all_term_scorers = [&left, &right].into_iter().all(|scorer| {
-                    let scorer_ref: &Scorer = (*scorer).borrow();
-                    Downcast::<TermScorer>::is_type(scorer_ref)
+                let all_term_scorers = [&left, &right].iter().all(|&scorer| {
+                    scorer.is::<TermScorer>()
                 });
                 if all_term_scorers {
-                    let left = *Downcast::<TermScorer>::downcast(left).unwrap();
-                    let right = *Downcast::<TermScorer>::downcast(right).unwrap();
+                    let left = *(left.downcast::<TermScorer>().map_err(|_| ()).unwrap());
+                    let right = *(right.downcast::<TermScorer>().map_err(|_| ()).unwrap());
                     return Box::new(Intersection {
                         left,
                         right,

src/query/phrase_query/mod.rs

@@ -24,7 +24,7 @@ mod tests {
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             for &text in texts {
                 let doc = doc!(text_field=>text);
                 index_writer.add_document(doc);
@@ -86,7 +86,7 @@ mod tests {
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"a b c"));
             assert!(index_writer.commit().is_ok());
         }
@@ -141,7 +141,7 @@ mod tests {
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"b"));
             index_writer.add_document(doc!(text_field=>"a b"));
             index_writer.add_document(doc!(text_field=>"b a"));
@@ -173,7 +173,7 @@ mod tests {
         let schema = schema_builder.build();
         let index = Index::create_in_ram(schema);
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field=>"a b c d e f g h"));
             assert!(index_writer.commit().is_ok());
         }

src/query/query.rs

@@ -1,6 +1,6 @@
 use super::Weight;
 use core::searcher::Searcher;
-use downcast;
+use downcast_rs;
 use std::collections::BTreeSet;
 use std::fmt;
 use Result;
@@ -39,7 +39,7 @@ use Term;
 ///
 /// When implementing a new type of `Query`, it is normal to implement a
 /// dedicated `Query`, `Weight` and `Scorer`.
-pub trait Query: QueryClone + downcast::Any + fmt::Debug {
+pub trait Query: QueryClone + downcast_rs::Downcast + fmt::Debug {
     /// Create the weight associated to a query.
     ///
     /// If scoring is not required, setting `scoring_enabled` to `false`
@@ -96,7 +96,4 @@ impl QueryClone for Box<Query> {
     }
 }
 
-#[allow(missing_docs)]
-mod downcast_impl {
-    downcast!(super::Query);
-}
+impl_downcast!(Query);

src/query/scorer.rs

@@ -1,6 +1,6 @@
 use common::BitSet;
 use docset::{DocSet, SkipResult};
-use downcast;
+use downcast_rs;
 use std::ops::DerefMut;
 use DocId;
 use Score;
@@ -8,7 +8,7 @@ use Score;
 /// Scored set of documents matching a query within a specific segment.
 ///
 /// See [`Query`](./trait.Query.html).
-pub trait Scorer: downcast::Any + DocSet + 'static {
+pub trait Scorer: downcast_rs::Downcast + DocSet + 'static {
     /// Returns the score.
     ///
     /// This method will perform a bit of computation and is not cached.
@@ -23,10 +23,8 @@ pub trait Scorer: downcast::Any + DocSet + 'static {
     }
 }
 
-#[allow(missing_docs)]
-mod downcast_impl {
-    downcast!(super::Scorer);
-}
+impl_downcast!(Scorer);
+
 
 impl Scorer for Box<Scorer> {
     fn score(&mut self) -> Score {

src/query/term_query/mod.rs

@@ -25,7 +25,7 @@ mod tests {
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc!(text_field => "a");
                 index_writer.add_document(doc);

src/schema/facet.rs

@@ -6,6 +6,7 @@ use std::borrow::Cow;
 use std::fmt::{self, Debug, Display, Formatter};
 use std::io::{self, Read, Write};
 use std::str;
+use std::string::FromUtf8Error;
 
 const SLASH_BYTE: u8 = b'/';
 const ESCAPE_BYTE: u8 = b'\\';
@@ -14,6 +15,10 @@ const ESCAPE_BYTE: u8 = b'\\';
 /// representation of facets.
 pub const FACET_SEP_BYTE: u8 = 0u8;
 
+/// `char` used as a level separation in the binary
+/// representation of facets. (It is the null codepoint.)
+pub const FACET_SEP_CHAR: char = '\u{0}';
+
 /// A Facet represent a point in a given hierarchy.
 ///
 /// They are typically represented similarly to a filepath.
@@ -26,18 +31,18 @@ pub const FACET_SEP_BYTE: u8 = 0u8;
 /// its facet. In the example above, `/electronics/tv_and_video/`
 /// and `/electronics`.
 #[derive(Clone, Eq, Hash, PartialEq, Ord, PartialOrd)]
-pub struct Facet(Vec<u8>);
+pub struct Facet(String);
 
 impl Facet {
     /// Returns a new instance of the "root facet"
     /// Equivalent to `/`.
     pub fn root() -> Facet {
-        Facet(vec![])
+        Facet("".to_string())
     }
 
     /// Returns true iff the facet is the root facet `/`.
     pub fn is_root(&self) -> bool {
-        self.encoded_bytes().is_empty()
+        self.encoded_str().is_empty()
     }
 
     /// Returns a binary representation of the facet.
@@ -49,13 +54,19 @@ impl Facet {
     /// This representation has the benefit of making it possible to
     /// express "being a child of a given facet" as a range over
     /// the term ordinals.
-    pub fn encoded_bytes(&self) -> &[u8] {
+    pub fn encoded_str(&self) -> &str {
         &self.0
     }
 
+    pub(crate) fn from_encoded_string(facet_string: String) -> Facet {
+        Facet(facet_string)
+    }
+
     /// Creates a `Facet` from its binary representation.
-    pub(crate) unsafe fn from_encoded(encoded_bytes: Vec<u8>) -> Facet {
-        Facet(encoded_bytes)
+    pub fn from_encoded(encoded_bytes: Vec<u8>) -> Result<Facet, FromUtf8Error> {
+        // facet bytes validation. `0u8` is used a separator but that is still legal utf-8
+        //Ok(Facet(String::from_utf8(encoded_bytes)?))
+        String::from_utf8(encoded_bytes).map(Facet)
     }
 
     /// Parse a text representation of a facet.
@@ -79,36 +90,37 @@ impl Facet {
         Path: IntoIterator,
         Path::Item: ToString,
     {
-        let mut facet_bytes: Vec<u8> = Vec::with_capacity(100);
+        let mut facet_string: String = String::with_capacity(100);
         let mut step_it = path.into_iter();
         if let Some(step) = step_it.next() {
-            facet_bytes.extend_from_slice(step.to_string().as_bytes());
+            facet_string.push_str(&step.to_string());
         }
         for step in step_it {
-            facet_bytes.push(FACET_SEP_BYTE);
-            facet_bytes.extend_from_slice(step.to_string().as_bytes());
+            facet_string.push(FACET_SEP_CHAR);
+            facet_string.push_str(&step.to_string());
         }
-        Facet(facet_bytes)
+        Facet(facet_string)
     }
 
     /// Accessor for the inner buffer of the `Facet`.
-    pub(crate) fn inner_buffer_mut(&mut self) -> &mut Vec<u8> {
-        &mut self.0
+    pub(crate) fn set_facet_str(&mut self, facet_str: &str) {
+        self.0.clear();
+        self.0.push_str(facet_str);
     }
 
     /// Returns `true` iff other is a subfacet of `self`.
     pub fn is_prefix_of(&self, other: &Facet) -> bool {
-        let self_bytes: &[u8] = self.encoded_bytes();
-        let other_bytes: &[u8] = other.encoded_bytes();
-        self_bytes.len() < other_bytes.len()
-            && other_bytes.starts_with(self_bytes)
-            && other_bytes[self_bytes.len()] == 0u8
+        let self_str = self.encoded_str();
+        let other_str = other.encoded_str();
+        self_str.len() < other_str.len()
+            && other_str.starts_with(self_str)
+            && other_str.as_bytes()[self_str.len()] == FACET_SEP_BYTE
     }
 }
 
-impl Borrow<[u8]> for Facet {
-    fn borrow(&self) -> &[u8] {
-        self.encoded_bytes()
+impl Borrow<str> for Facet {
+    fn borrow(&self) -> &str {
+        self.encoded_str()
     }
 }
 
@@ -120,45 +132,51 @@ impl<'a, T: ?Sized + AsRef<str>> From<&'a T> for Facet {
             Idle,
         }
         let path: &str = path_asref.as_ref();
-        let mut facet_encoded = Vec::new();
+        assert!(!path.is_empty());
+        assert!(path.starts_with('/'));
+        let mut facet_encoded = String::new();
         let mut state = State::Idle;
         let path_bytes = path.as_bytes();
-        for &c in &path_bytes[1..] {
+        let mut last_offset = 1;
+        for i in 1..path_bytes.len() {
+            let c = path_bytes[i];
             match (state, c) {
-                (State::Idle, ESCAPE_BYTE) => state = State::Escaped,
+                (State::Idle, ESCAPE_BYTE) => {
+                    facet_encoded.push_str(&path[last_offset..i]);
+                    last_offset = i + 1;
+                    state = State::Escaped
+                }
                 (State::Idle, SLASH_BYTE) => {
-                    facet_encoded.push(FACET_SEP_BYTE);
+                    facet_encoded.push_str(&path[last_offset..i]);
+                    facet_encoded.push(FACET_SEP_CHAR);
+                    last_offset = i + 1;
                 }
-                (State::Escaped, any_char) => {
+                (State::Escaped, _escaped_char) => {
                     state = State::Idle;
-                    facet_encoded.push(any_char);
-                }
-                (State::Idle, other_char) => {
-                    facet_encoded.push(other_char);
                 }
+                (State::Idle, _any_char) => {}
             }
         }
+        facet_encoded.push_str(&path[last_offset..]);
         Facet(facet_encoded)
     }
 }
 
 impl BinarySerializable for Facet {
     fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
-        <Vec<u8> as BinarySerializable>::serialize(&self.0, writer)
+        <String as BinarySerializable>::serialize(&self.0, writer)
     }
 
     fn deserialize<R: Read>(reader: &mut R) -> io::Result<Self> {
-        let bytes = <Vec<u8> as BinarySerializable>::deserialize(reader)?;
-        Ok(Facet(bytes))
+        Ok(Facet(<String as BinarySerializable>::deserialize(reader)?))
     }
 }
 
 impl Display for Facet {
     fn fmt(&self, f: &mut Formatter) -> fmt::Result {
-        for step in self.0.split(|&b| b == FACET_SEP_BYTE) {
+        for step in self.0.split(FACET_SEP_CHAR) {
             write!(f, "/")?;
-            let step_str = unsafe { str::from_utf8_unchecked(step) };
-            write!(f, "{}", escape_slashes(step_str))?;
+            write!(f, "{}", escape_slashes(step))?;
         }
         Ok(())
     }

src/schema/term.rs

@@ -32,7 +32,7 @@ impl Term {
 
     /// Creates a `Term` given a facet.
     pub fn from_facet(field: Field, facet: &Facet) -> Term {
-        let bytes = facet.encoded_bytes();
+        let bytes = facet.encoded_str().as_bytes();
         let buffer = Vec::with_capacity(4 + bytes.len());
         let mut term = Term(buffer);
         term.set_field(field);
@@ -68,12 +68,7 @@ impl Term {
         term
     }
 
-    /// Creates a new Term with an empty buffer,
-    /// but with a given capacity.
-    ///
-    /// It is declared unsafe, as the term content
-    /// is not initialized, and a call to `.field()`
-    /// would panic.
+    /// Creates a new Term for a given field.
     pub(crate) fn for_field(field: Field) -> Term {
         let mut term = Term(Vec::with_capacity(100));
         term.set_field(field);

src/snippet/mod.rs

@@ -523,7 +523,7 @@ Survey in 2016, 2017, and 2018."#;
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             index_writer.add_document(doc!(text_field => "a"));
             index_writer.add_document(doc!(text_field => "a"));
             index_writer.add_document(doc!(text_field => "a b"));
@@ -580,7 +580,7 @@ Survey in 2016, 2017, and 2018."#;
         let index = Index::create_in_ram(schema);
         {
             // writing the segment
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 let doc = doc ! (text_field => TEST_TEXT);
                 index_writer.add_document(doc);

src/termdict/mod.rs

@@ -133,7 +133,7 @@ mod tests {
         let text_field = schema_builder.add_text_field("text", TEXT);
         let index = Index::create_in_ram(schema_builder.build());
         {
-            let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
+            let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             {
                 {
                     let mut doc = Document::default();
@@ -167,7 +167,7 @@ mod tests {
         let mut term_string = String::new();
         while term_it.advance() {
             //let term = Term::from_bytes(term_it.key());
-            term_string.push_str(unsafe { str::from_utf8_unchecked(term_it.key()) }); // ok test
+            term_string.push_str(str::from_utf8(term_it.key()).expect("test"));
         }
         assert_eq!(&*term_string, "abcdef");
     }

src/termdict/term_info_store.rs

@@ -1,4 +1,4 @@
-use byteorder::ByteOrder;
+use byteorder::{ByteOrder, LittleEndian};
 use common::bitpacker::BitPacker;
 use common::compute_num_bits;
 use common::Endianness;
@@ -7,7 +7,6 @@ use directory::ReadOnlySource;
 use postings::TermInfo;
 use std::cmp;
 use std::io::{self, Read, Write};
-use std::ptr;
 use termdict::TermOrdinal;
 
 const BLOCK_LEN: usize = 256;
@@ -88,13 +87,17 @@ fn extract_bits(data: &[u8], addr_bits: usize, num_bits: u8) -> u64 {
     assert!(num_bits <= 56);
     let addr_byte = addr_bits / 8;
     let bit_shift = (addr_bits % 8) as u64;
-    assert!(data.len() >= addr_byte + 7);
-    let val_unshifted_unmasked: u64 = unsafe {
-        // ok because the pointer is only accessed using `ptr::read_unaligned`
-        #[cfg_attr(feature = "cargo-clippy", allow(clippy::cast_ptr_alignment))]
-        let addr = data.as_ptr().add(addr_byte) as *const u64;
-        // ok thanks to the 7 byte padding
-        ptr::read_unaligned(addr)
+    let val_unshifted_unmasked: u64 = if data.len() >= addr_byte + 8 {
+        LittleEndian::read_u64(&data[addr_byte..][..8])
+    } else {
+        // the buffer is not large enough.
+        // Let's copy the few remaining bytes to a 8 byte buffer
+        // padded with 0s.
+        let mut buf = [0u8; 8];
+        let data_to_copy = &data[addr_byte..];
+        let nbytes = data_to_copy.len();
+        buf[..nbytes].copy_from_slice(data_to_copy);
+        LittleEndian::read_u64(&buf)
     };
     let val_shifted_unmasked = val_unshifted_unmasked >> bit_shift;
     let mask = (1u64 << u64::from(num_bits)) - 1;
@@ -246,7 +249,6 @@ impl TermInfoStoreWriter {
         self.num_terms.serialize(write)?;
         write.write_all(&self.buffer_block_metas)?;
         write.write_all(&self.buffer_term_infos)?;
-        write.write_all(&[0u8; 7])?;
         Ok(())
     }
 }

src/tokenizer/facet_tokenizer.rs

@@ -1,6 +1,5 @@
 use super::{Token, TokenStream, Tokenizer};
 use schema::FACET_SEP_BYTE;
-use std::str;
 
 /// The `FacetTokenizer` process a `Facet` binary representation
 /// and emits a token for all of its parent.
@@ -57,12 +56,11 @@ impl<'a> TokenStream for FacetTokenStream<'a> {
                     .position(|b| b == FACET_SEP_BYTE)
                     .map(|pos| cursor + 1 + pos)
                 {
-                    let facet_part =
-                        unsafe { str::from_utf8_unchecked(&bytes[cursor..next_sep_pos]) };
+                    let facet_part = &self.text[cursor..next_sep_pos];
                     self.token.text.push_str(facet_part);
                     self.state = State::UpToPosition(next_sep_pos);
                 } else {
-                    let facet_part = unsafe { str::from_utf8_unchecked(&bytes[cursor..]) };
+                    let facet_part = &self.text[cursor..];
                     self.token.text.push_str(facet_part);
                     self.state = State::Terminated;
                 }
@@ -86,7 +84,6 @@ mod tests {
 
     use super::FacetTokenizer;
     use schema::Facet;
-    use std::str;
     use tokenizer::{Token, TokenStream, Tokenizer};
 
     #[test]
@@ -95,11 +92,11 @@ mod tests {
         let mut tokens = vec![];
         {
             let mut add_token = |token: &Token| {
-                let facet = unsafe { Facet::from_encoded(token.text.as_bytes().to_owned()) }; // ok test
+                let facet = Facet::from_encoded(token.text.as_bytes().to_owned()).unwrap();
                 tokens.push(format!("{}", facet));
             };
             FacetTokenizer
-                .token_stream(unsafe { str::from_utf8_unchecked(facet.encoded_bytes()) })
+                .token_stream(facet.encoded_str())
                 .process(&mut add_token);
         }
         assert_eq!(tokens.len(), 4);
@@ -115,11 +112,11 @@ mod tests {
         let mut tokens = vec![];
         {
             let mut add_token = |token: &Token| {
-                let facet = unsafe { Facet::from_encoded(token.text.as_bytes().to_owned()) }; // ok test
+                let facet = Facet::from_encoded(token.text.as_bytes().to_owned()).unwrap(); // ok test
                 tokens.push(format!("{}", facet));
             };
             FacetTokenizer
-                .token_stream(unsafe { str::from_utf8_unchecked(facet.encoded_bytes()) }) // ok test
+                .token_stream(facet.encoded_str()) // ok test
                 .process(&mut add_token);
         }
         assert_eq!(tokens.len(), 1);

src/tokenizer/lower_caser.rs

@@ -50,7 +50,6 @@ where
                 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

src/tokenizer/mod.rs

@@ -73,7 +73,7 @@
 //! let en_stem = SimpleTokenizer
 //!     .filter(RemoveLongFilter::limit(40))
 //!     .filter(LowerCaser)
-//!     .filter(Stemmer::new());
+//!     .filter(Stemmer::new(Language::English));
 //! # }
 //! ```
 //!
@@ -148,7 +148,7 @@ pub use self::ngram_tokenizer::NgramTokenizer;
 pub use self::raw_tokenizer::RawTokenizer;
 pub use self::remove_long::RemoveLongFilter;
 pub use self::simple_tokenizer::SimpleTokenizer;
-pub use self::stemmer::Stemmer;
+pub use self::stemmer::{Language, Stemmer};
 pub use self::stop_word_filter::StopWordFilter;
 pub(crate) use self::token_stream_chain::TokenStreamChain;
 pub(crate) use self::tokenizer::box_tokenizer;
@@ -159,8 +159,10 @@ pub use self::tokenizer_manager::TokenizerManager;
 
 #[cfg(test)]
 pub mod tests {
-    use super::Token;
-    use super::TokenizerManager;
+    use super::{
+        Language, LowerCaser, RemoveLongFilter, SimpleTokenizer, Stemmer, Token, Tokenizer,
+        TokenizerManager,
+    };
 
     /// This is a function that can be used in tests and doc tests
     /// to assert a token's correctness.
@@ -214,6 +216,7 @@ pub mod tests {
                 .token_stream("Hello, happy tax payer!")
                 .process(&mut add_token);
         }
+
         assert_eq!(tokens.len(), 4);
         assert_token(&tokens[0], 0, "hello", 0, 5);
         assert_token(&tokens[1], 1, "happi", 7, 12);
@@ -221,6 +224,33 @@ pub mod tests {
         assert_token(&tokens[3], 3, "payer", 17, 22);
     }
 
+    #[test]
+    fn test_non_en_tokenizer() {
+        let tokenizer_manager = TokenizerManager::default();
+        tokenizer_manager.register(
+            "es_stem",
+            SimpleTokenizer
+                .filter(RemoveLongFilter::limit(40))
+                .filter(LowerCaser)
+                .filter(Stemmer::new(Language::Spanish)),
+        );
+        let en_tokenizer = tokenizer_manager.get("es_stem").unwrap();
+        let mut tokens: Vec<Token> = vec![];
+        {
+            let mut add_token = |token: &Token| {
+                tokens.push(token.clone());
+            };
+            en_tokenizer
+                .token_stream("Hola, feliz contribuyente!")
+                .process(&mut add_token);
+        }
+
+        assert_eq!(tokens.len(), 3);
+        assert_token(&tokens[0], 0, "hola", 0, 4);
+        assert_token(&tokens[1], 1, "feliz", 6, 11);
+        assert_token(&tokens[2], 2, "contribuyent", 12, 25);
+    }
+
     #[test]
     fn test_tokenizer_empty() {
         let tokenizer_manager = TokenizerManager::default();

src/tokenizer/stemmer.rs

@@ -4,22 +4,76 @@ use super::{Token, TokenFilter, TokenStream};
 use rust_stemmers::{self, Algorithm};
 use std::sync::Arc;
 
-/// `Stemmer` token filter. Currently only English is supported.
-/// Tokens are expected to be lowercased beforehands.
+/// Available stemmer languages.
+#[derive(Debug, Serialize, Deserialize, Eq, PartialEq, Copy, Clone)]
+#[allow(missing_docs)]
+pub enum Language {
+    Arabic,
+    Danish,
+    Dutch,
+    English,
+    Finnish,
+    French,
+    German,
+    Hungarian,
+    Italian,
+    Portuguese,
+    Romanian,
+    Russian,
+    Spanish,
+    Swedish,
+    Tamil,
+    Turkish,
+}
+
+impl Language {
+    fn algorithm(self) -> Algorithm {
+        use self::Language::*;
+        match self {
+            Arabic => Algorithm::Arabic,
+            Danish => Algorithm::Danish,
+            Dutch => Algorithm::Dutch,
+            English => Algorithm::English,
+            Finnish => Algorithm::Finnish,
+            French => Algorithm::French,
+            German => Algorithm::German,
+            Hungarian => Algorithm::Hungarian,
+            Italian => Algorithm::Italian,
+            Portuguese => Algorithm::Portuguese,
+            Romanian => Algorithm::Romanian,
+            Russian => Algorithm::Russian,
+            Spanish => Algorithm::Spanish,
+            Swedish => Algorithm::Swedish,
+            Tamil => Algorithm::Tamil,
+            Turkish => Algorithm::Turkish,
+        }
+    }
+}
+
+/// `Stemmer` token filter. Several languages are supported, see `Language` for the available
+/// languages.
+/// Tokens are expected to be lowercased beforehand.
 #[derive(Clone)]
 pub struct Stemmer {
     stemmer_algorithm: Arc<Algorithm>,
 }
 
 impl Stemmer {
-    /// Creates a new Stemmer `TokenFilter`.
-    pub fn new() -> Stemmer {
+    /// Creates a new Stemmer `TokenFilter` for a given language algorithm.
+    pub fn new(language: Language) -> Stemmer {
         Stemmer {
-            stemmer_algorithm: Arc::new(Algorithm::English),
+            stemmer_algorithm: Arc::new(language.algorithm()),
         }
     }
 }
 
+impl Default for Stemmer {
+    /// Creates a new Stemmer `TokenFilter` for English.
+    fn default() -> Self {
+        Stemmer::new(Language::English)
+    }
+}
+
 impl<TailTokenStream> TokenFilter<TailTokenStream> for Stemmer
 where
     TailTokenStream: TokenStream,

src/tokenizer/tokenizer.rs

@@ -64,7 +64,7 @@ pub trait Tokenizer<'a>: Sized + Clone {
     /// let en_stem = SimpleTokenizer
     ///     .filter(RemoveLongFilter::limit(40))
     ///     .filter(LowerCaser)
-    ///     .filter(Stemmer::new());
+    ///     .filter(Stemmer::default());
     /// # }
     /// ```
     ///

src/tokenizer/tokenizer_manager.rs

@@ -1,6 +1,7 @@
 use std::collections::HashMap;
 use std::sync::{Arc, RwLock};
 use tokenizer::box_tokenizer;
+use tokenizer::stemmer::Language;
 use tokenizer::BoxedTokenizer;
 use tokenizer::LowerCaser;
 use tokenizer::RawTokenizer;
@@ -71,7 +72,7 @@ impl Default for TokenizerManager {
             SimpleTokenizer
                 .filter(RemoveLongFilter::limit(40))
                 .filter(LowerCaser)
-                .filter(Stemmer::new()),
+                .filter(Stemmer::new(Language::English)),
         );
         manager
     }