Closes #708

Fixes a race condition in the test.

.github/FUNDING.yml

@@ -0,0 +1,12 @@
+# These are supported funding model platforms
+
+github: fulmicoton
+patreon: # Replace with a single Patreon username
+open_collective: # Replace with a single Open Collective username
+ko_fi: # Replace with a single Ko-fi username
+tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel
+community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry
+liberapay: # Replace with a single Liberapay username
+issuehunt: # Replace with a single IssueHunt username
+otechie: # Replace with a single Otechie username
+custom: # Replace with up to 4 custom sponsorship URLs e.g., ['link1', 'link2']

CHANGELOG.md

@@ -9,7 +9,11 @@ Tantivy 0.11.0
 - API change around `Box<BoxableTokenizer>`. See detail in #629
 - Avoid rebuilding Regex automaton whenever a regex query is reused. #639 (@brainlock)
 - Add footer with some metadata to index files. #605 (@fdb-hiroshima)
- 
+- TopDocs collector: ensure stable sorting on equal score. #671 (@brainlock)
+- Added handling of pre-tokenized text fields (#642), which will enable users to
+  load tokens created outside tantivy. See usage in examples/pre_tokenized_text. (@kkoziara)
+- Fix crash when committing multiple times with deleted documents. #681 (@brainlock)
+
 ## How to update?
 
 - `Box<dyn BoxableTokenizer>` has been replaced by a `BoxedTokenizer` struct.

Cargo.toml

@@ -13,7 +13,7 @@ keywords = ["search", "information", "retrieval"]
 edition = "2018"
 
 [dependencies]
-base64 = "0.10.0"
+base64 = "0.11.0"
 byteorder = "1.0"
 crc32fast = "1.2.0"
 once_cell = "1.0"
@@ -34,26 +34,25 @@ itertools = "0.8"
 levenshtein_automata = {version="0.1", features=["fst_automaton"]}
 notify = {version="4", optional=true}
 bit-set = "0.5"
-uuid = { version = "0.7.2", features = ["v4", "serde"] }
+uuid = { version = "0.8", features = ["v4", "serde"] }
 crossbeam = "0.7"
-futures = "0.1"
+futures = {version = "0.3",  features=["thread-pool"] }
-futures-cpupool = "0.1"
 owning_ref = "0.4"
 stable_deref_trait = "1.0.0"
 rust-stemmers = "1.1"
 downcast-rs = { version="1.0" }
 tantivy-query-grammar = { path="./query-grammar" }
 bitpacking = {version="0.8", default-features = false, features=["bitpacker4x"]}
-census = "0.2"
+census = "0.3" 
 fnv = "1.0.6"
 owned-read = "0.4"
 failure = "0.1"
 htmlescape = "0.3.1"
 fail = "0.3"
-scoped-pool = "1.0"
 murmurhash32 = "0.2"
 chrono = "0.4"
-smallvec = "0.6"
+smallvec = "1.0"
+rayon = "1"
 
 [target.'cfg(windows)'.dependencies]
 winapi = "0.3"
@@ -64,6 +63,10 @@ maplit = "1"
 matches = "0.1.8"
 time = "0.1.42"
 
+[dev-dependencies.fail]
+version = "0.3"
+features = ["failpoints"]
+
 [profile.release]
 opt-level = 3
 debug = false
@@ -87,10 +90,6 @@ members = ["query-grammar"]
 [badges]
 travis-ci = { repository = "tantivy-search/tantivy" }
 
-[dev-dependencies.fail]
-version = "0.3"
-features = ["failpoints"]
-
 # Following the "fail" crate best practises, we isolate
 # tests that define specific behavior in fail check points
 # in a different binary.

examples/pre_tokenized_text.rs

@@ -0,0 +1,138 @@
+// # Pre-tokenized text example
+//
+// This example shows how to use pre-tokenized text. Sometimes yout might
+// want to index and search through text which is already split into
+// tokens by some external tool.
+//
+// In this example we will:
+// - use tantivy tokenizer to create tokens and load them directly into tantivy,
+// - import tokenized text straight from json,
+// - perform a search on documents with pre-tokenized text
+
+use tantivy::tokenizer::{PreTokenizedString, SimpleTokenizer, Token, TokenStream, Tokenizer};
+
+use tantivy::collector::{Count, TopDocs};
+use tantivy::query::TermQuery;
+use tantivy::schema::*;
+use tantivy::{doc, Index, ReloadPolicy};
+use tempfile::TempDir;
+
+fn pre_tokenize_text(text: &str) -> Vec<Token> {
+    let mut token_stream = SimpleTokenizer.token_stream(text);
+    let mut tokens = vec![];
+    while token_stream.advance() {
+        tokens.push(token_stream.token().clone());
+    }
+    tokens
+}
+
+fn main() -> tantivy::Result<()> {
+    let index_path = TempDir::new()?;
+
+    let mut schema_builder = Schema::builder();
+
+    schema_builder.add_text_field("title", TEXT | STORED);
+    schema_builder.add_text_field("body", TEXT);
+
+    let schema = schema_builder.build();
+
+    let index = Index::create_in_dir(&index_path, schema.clone())?;
+
+    let mut index_writer = index.writer(50_000_000)?;
+
+    // We can create a document manually, by setting the fields
+    // one by one in a Document object.
+    let title = schema.get_field("title").unwrap();
+    let body = schema.get_field("body").unwrap();
+
+    let title_text = "The Old Man and the Sea";
+    let body_text = "He was an old man who fished alone in a skiff in the Gulf Stream";
+
+    // Content of our first document
+    // We create `PreTokenizedString` which contains original text and vector of tokens
+    let title_tok = PreTokenizedString {
+        text: String::from(title_text),
+        tokens: pre_tokenize_text(title_text),
+    };
+
+    println!(
+        "Original text: \"{}\" and tokens: {:?}",
+        title_tok.text, title_tok.tokens
+    );
+
+    let body_tok = PreTokenizedString {
+        text: String::from(body_text),
+        tokens: pre_tokenize_text(body_text),
+    };
+
+    // Now lets create a document and add our `PreTokenizedString` using
+    // `add_pre_tokenized_text` method of `Document`
+    let old_man_doc = doc!(title => title_tok, body => body_tok);
+
+    // ... now let's just add it to the IndexWriter
+    index_writer.add_document(old_man_doc);
+
+    // Pretokenized text can also be fed as JSON
+    let short_man_json = r#"{
+        "title":[{
+            "text":"The Old Man",
+            "tokens":[
+                {"offset_from":0,"offset_to":3,"position":0,"text":"The","position_length":1},
+                {"offset_from":4,"offset_to":7,"position":1,"text":"Old","position_length":1},
+                {"offset_from":8,"offset_to":11,"position":2,"text":"Man","position_length":1}
+            ]
+        }]
+    }"#;
+
+    let short_man_doc = schema.parse_document(&short_man_json)?;
+
+    index_writer.add_document(short_man_doc);
+
+    // Let's commit changes
+    index_writer.commit()?;
+
+    // ... and now is the time to query our index
+
+    let reader = index
+        .reader_builder()
+        .reload_policy(ReloadPolicy::OnCommit)
+        .try_into()?;
+
+    let searcher = reader.searcher();
+
+    // We want to get documents with token "Man", we will use TermQuery to do it
+    // Using PreTokenizedString means the tokens are stored as is avoiding stemming
+    // and lowercasing, which preserves full words in their original form
+    let query = TermQuery::new(
+        Term::from_field_text(title, "Man"),
+        IndexRecordOption::Basic,
+    );
+
+    let (top_docs, count) = searcher
+        .search(&query, &(TopDocs::with_limit(2), Count))
+        .unwrap();
+
+    assert_eq!(count, 2);
+
+    for (_score, doc_address) in top_docs {
+        let retrieved_doc = searcher.doc(doc_address)?;
+        println!("Document: {}", schema.to_json(&retrieved_doc));
+    }
+
+    // In contrary to the previous query, when we search for the "man" term we
+    // should get no results, as it's not one of the indexed tokens. SimpleTokenizer
+    // only splits text on whitespace / punctuation.
+
+    let query = TermQuery::new(
+        Term::from_field_text(title, "man"),
+        IndexRecordOption::Basic,
+    );
+
+    let (_top_docs, count) = searcher
+        .search(&query, &(TopDocs::with_limit(2), Count))
+        .unwrap();
+
+    assert_eq!(count, 0);
+
+    Ok(())
+}

src/collector/top_collector.rs

@@ -12,6 +12,9 @@ use std::collections::BinaryHeap;
 /// It has a custom implementation of `PartialOrd` that reverses the order. This is because the
 /// default Rust heap is a max heap, whereas a min heap is needed.
 ///
+/// Additionally, it guarantees stable sorting: in case of a tie on the feature, the document
+/// address is used.
+///
 /// WARNING: equality is not what you would expect here.
 /// Two elements are equal if their feature is equal, and regardless of whether `doc`
 /// is equal. This should be perfectly fine for this usage, but let's make sure this
@@ -21,29 +24,37 @@ struct ComparableDoc<T, D> {
     doc: D,
 }
 
-impl<T: PartialOrd, D> PartialOrd for ComparableDoc<T, D> {
+impl<T: PartialOrd, D: PartialOrd> PartialOrd for ComparableDoc<T, D> {
     fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
         Some(self.cmp(other))
     }
 }
 
-impl<T: PartialOrd, D> Ord for ComparableDoc<T, D> {
+impl<T: PartialOrd, D: PartialOrd> Ord for ComparableDoc<T, D> {
     #[inline]
     fn cmp(&self, other: &Self) -> Ordering {
-        other
+        // Reversed to make BinaryHeap work as a min-heap
+        let by_feature = other
             .feature
             .partial_cmp(&self.feature)
-            .unwrap_or_else(|| Ordering::Equal)
+            .unwrap_or(Ordering::Equal);
+
+        let lazy_by_doc_address = || self.doc.partial_cmp(&other.doc).unwrap_or(Ordering::Equal);
+
+        // In case of a tie on the feature, we sort by ascending
+        // `DocAddress` in order to ensure a stable sorting of the
+        // documents.
+        by_feature.then_with(lazy_by_doc_address)
     }
 }
 
-impl<T: PartialOrd, D> PartialEq for ComparableDoc<T, D> {
+impl<T: PartialOrd, D: PartialOrd> PartialEq for ComparableDoc<T, D> {
     fn eq(&self, other: &Self) -> bool {
         self.cmp(other) == Ordering::Equal
     }
 }
 
-impl<T: PartialOrd, D> Eq for ComparableDoc<T, D> {}
+impl<T: PartialOrd, D: PartialOrd> Eq for ComparableDoc<T, D> {}
 
 pub(crate) struct TopCollector<T> {
     limit: usize,
@@ -214,4 +225,94 @@ mod tests {
             ]
         );
     }
+
+    #[test]
+    fn test_top_segment_collector_stable_ordering_for_equal_feature() {
+        // given that the documents are collected in ascending doc id order,
+        // when harvesting we have to guarantee stable sorting in case of a tie
+        // on the score
+        let doc_ids_collection = [4, 5, 6];
+        let score = 3.14;
+
+        let mut top_collector_limit_2 = TopSegmentCollector::new(0, 2);
+        for id in &doc_ids_collection {
+            top_collector_limit_2.collect(*id, score);
+        }
+
+        let mut top_collector_limit_3 = TopSegmentCollector::new(0, 3);
+        for id in &doc_ids_collection {
+            top_collector_limit_3.collect(*id, score);
+        }
+
+        assert_eq!(
+            top_collector_limit_2.harvest(),
+            top_collector_limit_3.harvest()[..2].to_vec(),
+        );
+    }
+}
+
+#[cfg(all(test, feature = "unstable"))]
+mod bench {
+    use super::TopSegmentCollector;
+    use test::Bencher;
+
+    #[bench]
+    fn bench_top_segment_collector_collect_not_at_capacity(b: &mut Bencher) {
+        let mut top_collector = TopSegmentCollector::new(0, 400);
+
+        b.iter(|| {
+            for i in 0..100 {
+                top_collector.collect(i, 0.8);
+            }
+        });
+    }
+
+    #[bench]
+    fn bench_top_segment_collector_collect_at_capacity(b: &mut Bencher) {
+        let mut top_collector = TopSegmentCollector::new(0, 100);
+
+        for i in 0..100 {
+            top_collector.collect(i, 0.8);
+        }
+
+        b.iter(|| {
+            for i in 0..100 {
+                top_collector.collect(i, 0.8);
+            }
+        });
+    }
+
+    #[bench]
+    fn bench_top_segment_collector_collect_and_harvest_many_ties(b: &mut Bencher) {
+        b.iter(|| {
+            let mut top_collector = TopSegmentCollector::new(0, 100);
+
+            for i in 0..100 {
+                top_collector.collect(i, 0.8);
+            }
+
+            // it would be nice to be able to do the setup N times but still
+            // measure only harvest(). We can't since harvest() consumes
+            // the top_collector.
+            top_collector.harvest()
+        });
+    }
+
+    #[bench]
+    fn bench_top_segment_collector_collect_and_harvest_no_tie(b: &mut Bencher) {
+        b.iter(|| {
+            let mut top_collector = TopSegmentCollector::new(0, 100);
+            let mut score = 1.0;
+
+            for i in 0..100 {
+                score += 1.0;
+                top_collector.collect(i, score);
+            }
+
+            // it would be nice to be able to do the setup N times but still
+            // measure only harvest(). We can't since harvest() consumes
+            // the top_collector.
+            top_collector.harvest()
+        });
+    }
 }

src/collector/top_score_collector.rs

@@ -15,13 +15,16 @@ use crate::SegmentLocalId;
 use crate::SegmentReader;
 use std::fmt;
 
-/// The Top Score Collector keeps track of the K documents
+/// The `TopDocs` collector keeps track of the top `K` documents
 /// sorted by their score.
 ///
 /// The implementation is based on a `BinaryHeap`.
 /// The theorical complexity for collecting the top `K` out of `n` documents
 /// is `O(n log K)`.
 ///
+/// This collector guarantees a stable sorting in case of a tie on the
+/// document score. As such, it is suitable to implement pagination.
+///
 /// ```rust
 /// use tantivy::collector::TopDocs;
 /// use tantivy::query::QueryParser;
@@ -428,12 +431,13 @@ impl SegmentCollector for TopScoreSegmentCollector {
 mod tests {
     use super::TopDocs;
     use crate::collector::Collector;
-    use crate::query::{Query, QueryParser};
+    use crate::query::{AllQuery, Query, QueryParser};
     use crate::schema::{Field, Schema, FAST, STORED, TEXT};
     use crate::DocAddress;
     use crate::Index;
     use crate::IndexWriter;
     use crate::Score;
+    use itertools::Itertools;
 
     fn make_index() -> Index {
         let mut schema_builder = Schema::builder();
@@ -494,6 +498,29 @@ mod tests {
         );
     }
 
+    #[test]
+    fn test_top_collector_stable_sorting() {
+        let index = make_index();
+
+        // using AllQuery to get a constant score
+        let searcher = index.reader().unwrap().searcher();
+
+        let page_1 = searcher.search(&AllQuery, &TopDocs::with_limit(2)).unwrap();
+
+        let page_2 = searcher.search(&AllQuery, &TopDocs::with_limit(3)).unwrap();
+
+        // precondition for the test to be meaningful: we did get documents
+        // with the same score
+        assert!(page_1.iter().map(|result| result.0).all_equal());
+        assert!(page_2.iter().map(|result| result.0).all_equal());
+
+        // sanity check since we're relying on make_index()
+        assert_eq!(page_1.len(), 2);
+        assert_eq!(page_2.len(), 3);
+
+        assert_eq!(page_1, &page_2[..page_1.len()]);
+    }
+
     #[test]
     #[should_panic]
     fn test_top_0() {

src/core/executor.rs

@@ -1,6 +1,6 @@
 use crate::Result;
 use crossbeam::channel;
-use scoped_pool::{Pool, ThreadConfig};
+use rayon::{ThreadPool, ThreadPoolBuilder};
 
 /// Search executor whether search request are single thread or multithread.
 ///
@@ -11,7 +11,7 @@ use scoped_pool::{Pool, ThreadConfig};
 /// used by the client. Second, we may stop using rayon in the future.
 pub enum Executor {
     SingleThread,
-    ThreadPool(Pool),
+    ThreadPool(ThreadPool),
 }
 
 impl Executor {
@@ -21,10 +21,12 @@ impl Executor {
     }
 
     // Creates an Executor that dispatches the tasks in a thread pool.
-    pub fn multi_thread(num_threads: usize, prefix: &'static str) -> Executor {
+    pub fn multi_thread(num_threads: usize, prefix: &'static str) -> Result<Executor> {
-        let thread_config = ThreadConfig::new().prefix(prefix);
+        let pool = ThreadPoolBuilder::new()
-        let pool = Pool::with_thread_config(num_threads, thread_config);
+            .num_threads(num_threads)
-        Executor::ThreadPool(pool)
+            .thread_name(move |num| format!("{}{}", prefix, num))
+            .build()?;
+        Ok(Executor::ThreadPool(pool))
     }
 
     // Perform a map in the thread pool.
@@ -48,9 +50,9 @@ impl Executor {
                 let num_fruits = args_with_indices.len();
                 let fruit_receiver = {
                     let (fruit_sender, fruit_receiver) = channel::unbounded();
-                    pool.scoped(|scope| {
+                    pool.scope(|scope| {
                         for arg_with_idx in args_with_indices {
-                            scope.execute(|| {
+                            scope.spawn(|_| {
                                 let (idx, arg) = arg_with_idx;
                                 let fruit = f(arg);
                                 if let Err(err) = fruit_sender.send((idx, fruit)) {
@@ -103,6 +105,7 @@ mod tests {
     #[should_panic] //< unfortunately the panic message is not propagated
     fn test_panic_propagates_multi_thread() {
         let _result: Vec<usize> = Executor::multi_thread(1, "search-test")
+            .unwrap()
             .map(
                 |_| {
                     panic!("panic should propagate");
@@ -126,6 +129,7 @@ mod tests {
     #[test]
     fn test_map_multithread() {
         let result: Vec<usize> = Executor::multi_thread(3, "search-test")
+            .unwrap()
             .map(|i| Ok(i * 2), 0..10)
             .unwrap();
         assert_eq!(result.len(), 10);

src/core/index.rs

@@ -73,15 +73,16 @@ impl Index {
 
     /// Replace the default single thread search executor pool
     /// by a thread pool with a given number of threads.
-    pub fn set_multithread_executor(&mut self, num_threads: usize) {
+    pub fn set_multithread_executor(&mut self, num_threads: usize) -> Result<()> {
-        self.executor = Arc::new(Executor::multi_thread(num_threads, "thrd-tantivy-search-"));
+        self.executor = Arc::new(Executor::multi_thread(num_threads, "thrd-tantivy-search-")?);
+        Ok(())
     }
 
     /// Replace the default single thread search executor pool
     /// by a thread pool with a given number of threads.
-    pub fn set_default_multithread_executor(&mut self) {
+    pub fn set_default_multithread_executor(&mut self) -> Result<()> {
         let default_num_threads = num_cpus::get();
-        self.set_multithread_executor(default_num_threads);
+        self.set_multithread_executor(default_num_threads)
     }
 
     /// Creates a new index using the `RAMDirectory`.
@@ -103,23 +104,21 @@ impl Index {
         if Index::exists(&mmap_directory) {
             return Err(TantivyError::IndexAlreadyExists);
         }
-
         Index::create(mmap_directory, schema)
     }
 
     /// Opens or creates a new index in the provided directory
     pub fn open_or_create<Dir: Directory>(dir: Dir, schema: Schema) -> Result<Index> {
-        if Index::exists(&dir) {
+        if !Index::exists(&dir) {
-            let index = Index::open(dir)?;
+            return Index::create(dir, schema);
-            if index.schema() == schema {
+        }
-                Ok(index)
+        let index = Index::open(dir)?;
-            } else {
+        if index.schema() == schema {
-                Err(TantivyError::SchemaError(
+            Ok(index)
-                    "An index exists but the schema does not match.".to_string(),
-                ))
-            }
         } else {
-            Index::create(dir, schema)
+            Err(TantivyError::SchemaError(
+                "An index exists but the schema does not match.".to_string(),
+            ))
         }
     }
 
@@ -387,12 +386,9 @@ mod tests {
     use crate::directory::RAMDirectory;
     use crate::schema::Field;
     use crate::schema::{Schema, INDEXED, TEXT};
-    use crate::Index;
     use crate::IndexReader;
-    use crate::IndexWriter;
     use crate::ReloadPolicy;
-    use std::thread;
+    use crate::{Directory, Index};
-    use std::time::Duration;
 
     #[test]
     fn test_indexer_for_field() {
@@ -470,14 +466,14 @@ mod tests {
             .try_into()
             .unwrap();
         assert_eq!(reader.searcher().num_docs(), 0);
-        let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+        test_index_on_commit_reload_policy_aux(field, &index, &reader);
-        test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
     }
 
     #[cfg(feature = "mmap")]
     mod mmap_specific {
 
         use super::*;
+        use crate::Directory;
         use std::path::PathBuf;
         use tempfile::TempDir;
 
@@ -488,22 +484,20 @@ mod tests {
             let tempdir = TempDir::new().unwrap();
             let tempdir_path = PathBuf::from(tempdir.path());
             let index = Index::create_in_dir(&tempdir_path, schema).unwrap();
-            let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            writer.commit().unwrap();
             let reader = index
                 .reader_builder()
                 .reload_policy(ReloadPolicy::OnCommit)
                 .try_into()
                 .unwrap();
             assert_eq!(reader.searcher().num_docs(), 0);
-            test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
+            test_index_on_commit_reload_policy_aux(field, &index, &reader);
         }
 
         #[test]
         fn test_index_manual_policy_mmap() {
             let schema = throw_away_schema();
             let field = schema.get_field("num_likes").unwrap();
-            let index = Index::create_from_tempdir(schema).unwrap();
+            let mut index = Index::create_from_tempdir(schema).unwrap();
             let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
             writer.commit().unwrap();
             let reader = index
@@ -513,8 +507,12 @@ mod tests {
                 .unwrap();
             assert_eq!(reader.searcher().num_docs(), 0);
             writer.add_document(doc!(field=>1u64));
+            let (sender, receiver) = crossbeam::channel::unbounded();
+            let _handle = index.directory_mut().watch(Box::new(move || {
+                let _ = sender.send(());
+            }));
             writer.commit().unwrap();
-            thread::sleep(Duration::from_millis(500));
+            assert!(receiver.recv().is_ok());
             assert_eq!(reader.searcher().num_docs(), 0);
             reader.reload().unwrap();
             assert_eq!(reader.searcher().num_docs(), 1);
@@ -534,39 +532,26 @@ mod tests {
                 .try_into()
                 .unwrap();
             assert_eq!(reader.searcher().num_docs(), 0);
-            let mut writer = write_index.writer_with_num_threads(1, 3_000_000).unwrap();
+            test_index_on_commit_reload_policy_aux(field, &write_index, &reader);
-            test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
         }
     }
 
-    fn test_index_on_commit_reload_policy_aux(
+    fn test_index_on_commit_reload_policy_aux(field: Field, index: &Index, reader: &IndexReader) {
-        field: Field,
+        let mut reader_index = reader.index();
-        writer: &mut IndexWriter,
+        let (sender, receiver) = crossbeam::channel::unbounded();
-        reader: &IndexReader,
+        let _watch_handle = reader_index.directory_mut().watch(Box::new(move || {
-    ) {
+            let _ = sender.send(());
+        }));
+        let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
         assert_eq!(reader.searcher().num_docs(), 0);
         writer.add_document(doc!(field=>1u64));
         writer.commit().unwrap();
-        let mut count = 0;
+        assert!(receiver.recv().is_ok());
-        for _ in 0..100 {
+        assert_eq!(reader.searcher().num_docs(), 1);
-            count = reader.searcher().num_docs();
-            if count > 0 {
-                break;
-            }
-            thread::sleep(Duration::from_millis(100));
-        }
-        assert_eq!(count, 1);
         writer.add_document(doc!(field=>2u64));
         writer.commit().unwrap();
-        let mut count = 0;
+        assert!(receiver.recv().is_ok());
-        for _ in 0..10 {
+        assert_eq!(reader.searcher().num_docs(), 2);
-            count = reader.searcher().num_docs();
-            if count > 1 {
-                break;
-            }
-            thread::sleep(Duration::from_millis(100));
-        }
-        assert_eq!(count, 2);
     }
 
     // This test will not pass on windows, because windows
@@ -583,9 +568,13 @@ mod tests {
         for i in 0u64..8_000u64 {
             writer.add_document(doc!(field => i));
         }
+        let (sender, receiver) = crossbeam::channel::unbounded();
+        let _handle = directory.watch(Box::new(move || {
+            let _ = sender.send(());
+        }));
         writer.commit().unwrap();
         let mem_right_after_commit = directory.total_mem_usage();
-        thread::sleep(Duration::from_millis(1_000));
+        assert!(receiver.recv().is_ok());
         let reader = index
             .reader_builder()
             .reload_policy(ReloadPolicy::Manual)
@@ -599,6 +588,11 @@ mod tests {
         reader.reload().unwrap();
         let searcher = reader.searcher();
         assert_eq!(searcher.num_docs(), 8_000);
-        assert!(mem_right_after_merge_finished < mem_right_after_commit);
+        assert!(
+            mem_right_after_merge_finished < mem_right_after_commit,
+            "(mem after merge){} is expected < (mem before merge){}",
+            mem_right_after_merge_finished,
+            mem_right_after_commit
+        );
     }
 }

src/core/index_meta.rs

@@ -150,6 +150,21 @@ impl SegmentMeta {
         self.num_deleted_docs() > 0
     }
 
+    /// Updates the max_doc value from the `SegmentMeta`.
+    ///
+    /// This method is only used when updating `max_doc` from 0
+    /// as we finalize a fresh new segment.
+    pub(crate) fn with_max_doc(self, max_doc: u32) -> SegmentMeta {
+        assert_eq!(self.tracked.max_doc, 0);
+        assert!(self.tracked.deletes.is_none());
+        let tracked = self.tracked.map(move |inner_meta| InnerSegmentMeta {
+            segment_id: inner_meta.segment_id,
+            max_doc,
+            deletes: None,
+        });
+        SegmentMeta { tracked }
+    }
+
     #[doc(hidden)]
     pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> SegmentMeta {
         let delete_meta = DeleteMeta {
@@ -285,6 +300,9 @@ mod tests {
             payload: None,
         };
         let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
-        assert_eq!(json, r#"{"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","tokenizer":"default"},"stored":false}}],"opstamp":0}"#);
+        assert_eq!(
+            json,
+            r#"{"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","tokenizer":"default"},"stored":false}}],"opstamp":0}"#
+        );
     }
 }

src/core/segment.rs

@@ -50,6 +50,17 @@ impl Segment {
         &self.meta
     }
 
+    /// Updates the max_doc value from the `SegmentMeta`.
+    ///
+    /// This method is only used when updating `max_doc` from 0
+    /// as we finalize a fresh new segment.
+    pub(crate) fn with_max_doc(self, max_doc: u32) -> Segment {
+        Segment {
+            index: self.index,
+            meta: self.meta.with_max_doc(max_doc),
+        }
+    }
+
     #[doc(hidden)]
     pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment {
         Segment {

src/core/segment_id.rs

@@ -76,7 +76,7 @@ impl SegmentId {
 }
 
 /// Error type used when parsing a `SegmentId` from a string fails.
-pub struct SegmentIdParseError(uuid::parser::ParseError);
+pub struct SegmentIdParseError(uuid::Error);
 
 impl Error for SegmentIdParseError {}
 

src/directory/footer.rs

@@ -7,6 +7,8 @@ use std::io::Write;
 
 const COMMON_FOOTER_SIZE: usize = 4 * 5;
 
+type CrcHashU32 = u32;
+
 #[derive(Debug, Clone, PartialEq)]
 pub struct Footer {
     pub tantivy_version: (u32, u32, u32),
@@ -24,7 +26,7 @@ impl Footer {
         Footer {
             tantivy_version,
             meta: format!(
-                "tantivy {}.{}.{}, index v{}",
+                "tantivy v{}.{}.{}, index_format v{}",
                 tantivy_version.0,
                 tantivy_version.1,
                 tantivy_version.2,
@@ -34,6 +36,9 @@ impl Footer {
         }
     }
 
+    /// Serialises the footer to a byte-array
+    /// [      versioned_footer     |     meta      |    common_footer ]
+    /// [           0..8            |     8..32     |        32..52    ]
     pub fn to_bytes(&self) -> Vec<u8> {
         let mut res = self.versioned_footer.to_bytes();
         res.extend_from_slice(self.meta.as_bytes());
@@ -64,20 +69,27 @@ impl Footer {
             return Err(io::Error::new(
                 io::ErrorKind::UnexpectedEof,
                 format!(
-                    "File corrupted. The footer len is {}, while the entire file len is {}",
+                    "The footer len is {}, while the entire file len is {}. \
+                     Your index is either corrupted or was built using a tantivy version\
+                     anterior to 0.11.",
                     size, len
                 ),
             ));
         }
         let footer = &data[len - size as usize..];
-        let meta_len = LittleEndian::read_u32(&footer[size - 20..]) as usize;
+        let meta_len = LittleEndian::read_u32(&footer[size - COMMON_FOOTER_SIZE..]) as usize;
         let tantivy_major = LittleEndian::read_u32(&footer[size - 16..]);
         let tantivy_minor = LittleEndian::read_u32(&footer[size - 12..]);
         let tantivy_patch = LittleEndian::read_u32(&footer[size - 8..]);
         Ok(Footer {
             tantivy_version: (tantivy_major, tantivy_minor, tantivy_patch),
-            meta: String::from_utf8_lossy(&footer[size - meta_len - 20..size - 20]).into_owned(),
+            meta: String::from_utf8_lossy(
-            versioned_footer: VersionedFooter::from_bytes(&footer[..size - meta_len - 20])?,
+                &footer[size - meta_len - COMMON_FOOTER_SIZE..size - COMMON_FOOTER_SIZE],
+            )
+            .into_owned(),
+            versioned_footer: VersionedFooter::from_bytes(
+                &footer[..size - meta_len - COMMON_FOOTER_SIZE],
+            )?,
         })
     }
 
@@ -88,24 +100,28 @@ impl Footer {
     }
 
     pub fn size(&self) -> usize {
-        self.versioned_footer.size() as usize + self.meta.len() + 20
+        self.versioned_footer.size() as usize + self.meta.len() + COMMON_FOOTER_SIZE
     }
 }
 
+/// Footer that includes a crc32 hash that enables us to checksum files in the index
 #[derive(Debug, Clone, PartialEq)]
 pub enum VersionedFooter {
     UnknownVersion { version: u32, size: u32 },
-    V0(u32), // crc
+    V0(CrcHashU32), // crc
 }
 
 impl VersionedFooter {
+    /// Serializes a valid `VersionedFooter` or panics if the version is unknown
+    /// [   version    |   crc_hash  ]
+    /// [      0..4    |     4..8    ]
     pub fn to_bytes(&self) -> Vec<u8> {
         match self {
             VersionedFooter::V0(crc) => {
-                let mut res = vec![0; 8];
+                let mut buf = [0u8; 8];
-                LittleEndian::write_u32(&mut res, 0);
+                LittleEndian::write_u32(&mut buf[0..4], 0);
-                LittleEndian::write_u32(&mut res[4..], *crc);
+                LittleEndian::write_u32(&mut buf[4..8], *crc);
-                res
+                buf.to_vec()
             }
             VersionedFooter::UnknownVersion { .. } => {
                 panic!("Unsupported index should never get serialized");
@@ -115,20 +131,26 @@ impl VersionedFooter {
 
     pub fn from_bytes(footer: &[u8]) -> Result<Self, io::Error> {
         assert!(footer.len() >= 4);
+        if footer.len() < 4 {
+            return Err(io::Error::new(
+                io::ErrorKind::InvalidData,
+                "Footer should be more than 4 bytes.",
+            ));
+        }
         let version = LittleEndian::read_u32(footer);
         match version {
+            // the first 4 bytes should be zeroed out thus returning a `0`
             0 => {
-                if footer.len() == 8 {
+                if footer.len() != 8 {
-                    Ok(VersionedFooter::V0(LittleEndian::read_u32(&footer[4..])))
+                    return Err(io::Error::new(
-                } else {
-                    Err(io::Error::new(
                         io::ErrorKind::UnexpectedEof,
                         format!(
                             "File corrupted. The versioned footer len is {}, while it should be 8",
                             footer.len()
                         ),
-                    ))
+                    ));
                 }
+                Ok(VersionedFooter::V0(LittleEndian::read_u32(&footer[4..])))
             }
             version => Ok(VersionedFooter::UnknownVersion {
                 version,
@@ -151,7 +173,7 @@ impl VersionedFooter {
         }
     }
 
-    pub fn crc(&self) -> Option<u32> {
+    pub fn crc(&self) -> Option<CrcHashU32> {
         match self {
             VersionedFooter::V0(crc) => Some(*crc),
             VersionedFooter::UnknownVersion { .. } => None,
@@ -190,7 +212,6 @@ impl<W: TerminatingWrite> Write for FooterProxy<W> {
 impl<W: TerminatingWrite> TerminatingWrite for FooterProxy<W> {
     fn terminate_ref(&mut self, _: AntiCallToken) -> io::Result<()> {
         let crc = self.hasher.take().unwrap().finalize();
-
         let footer = Footer::new(VersionedFooter::V0(crc)).to_bytes();
         let mut writer = self.writer.take().unwrap();
         writer.write_all(&footer)?;
@@ -200,14 +221,57 @@ impl<W: TerminatingWrite> TerminatingWrite for FooterProxy<W> {
 
 #[cfg(test)]
 mod tests {
+
     use crate::directory::footer::{Footer, VersionedFooter};
+    use regex::Regex;
 
     #[test]
     fn test_serialize_deserialize_footer() {
         let crc = 123456;
         let footer = Footer::new(VersionedFooter::V0(crc));
         let footer_bytes = footer.to_bytes();
-
         assert_eq!(Footer::from_bytes(&footer_bytes).unwrap(), footer);
     }
+
+    #[test]
+    fn footer_length() {
+        // test to make sure the ascii art in the doc-strings is correct
+        let crc = 1111111 as u32;
+        let versioned_footer = VersionedFooter::V0(crc);
+        assert_eq!(versioned_footer.size(), 8);
+        let footer = Footer::new(versioned_footer);
+        let regex_ptn = Regex::new(
+            "tantivy v[0-9]{1,3}\\.[0-9]{1,3}\\.[0-9]{1,3}\\.{0,10}, index_format v[0-9]{1,5}",
+        )
+        .unwrap();
+        assert!(regex_ptn.find(&footer.meta).is_some());
+    }
+
+    #[test]
+    fn versioned_footer_from_bytes() {
+        use byteorder::{ByteOrder, LittleEndian};
+        let v_footer_bytes = vec![0, 0, 0, 0, 12, 35, 89, 18];
+        let versioned_footer = VersionedFooter::from_bytes(&v_footer_bytes).unwrap();
+        let expected_versioned_footer =
+            VersionedFooter::V0(LittleEndian::read_u32(&[12, 35, 89, 18]));
+        assert_eq!(versioned_footer, expected_versioned_footer);
+
+        assert_eq!(versioned_footer.to_bytes(), v_footer_bytes);
+    }
+
+    #[should_panic(expected = "Unsupported index should never get serialized")]
+    #[test]
+    fn versioned_footer_panic() {
+        use byteorder::{ByteOrder, LittleEndian};
+        let v_footer_bytes = vec![1; 8];
+        let versioned_footer = VersionedFooter::from_bytes(&v_footer_bytes).unwrap();
+        let expected_version = LittleEndian::read_u32(&[1, 1, 1, 1]);
+        let expected_versioned_footer = VersionedFooter::UnknownVersion {
+            version: expected_version,
+            size: v_footer_bytes.len() as u32,
+        };
+        assert_eq!(versioned_footer, expected_versioned_footer);
+
+        versioned_footer.to_bytes();
+    }
 }

src/directory/managed_directory.rs

@@ -2,6 +2,7 @@ use crate::core::MANAGED_FILEPATH;
 use crate::directory::error::{DeleteError, IOError, LockError, OpenReadError, OpenWriteError};
 use crate::directory::footer::{Footer, FooterProxy};
 use crate::directory::DirectoryLock;
+use crate::directory::GarbageCollectionResult;
 use crate::directory::Lock;
 use crate::directory::META_LOCK;
 use crate::directory::{ReadOnlySource, WritePtr};
@@ -104,7 +105,10 @@ impl ManagedDirectory {
     /// 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
     /// files.
-    pub fn garbage_collect<L: FnOnce() -> HashSet<PathBuf>>(&mut self, get_living_files: L) {
+    pub fn garbage_collect<L: FnOnce() -> HashSet<PathBuf>>(
+        &mut self,
+        get_living_files: L,
+    ) -> crate::Result<GarbageCollectionResult> {
         info!("Garbage collect");
         let mut files_to_delete = vec![];
 
@@ -130,19 +134,25 @@ 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) = self.acquire_lock(&META_LOCK) {
+            match self.acquire_lock(&META_LOCK) {
-                let living_files = get_living_files();
+                Ok(_meta_lock) => {
-                for managed_path in &meta_informations_rlock.managed_paths {
+                    let living_files = get_living_files();
-                    if !living_files.contains(managed_path) {
+                    for managed_path in &meta_informations_rlock.managed_paths {
-                        files_to_delete.push(managed_path.clone());
+                        if !living_files.contains(managed_path) {
+                            files_to_delete.push(managed_path.clone());
+                        }
                     }
                 }
-            } else {
+                Err(err) => {
-                error!("Failed to acquire lock for GC");
+                    error!("Failed to acquire lock for GC");
+                    return Err(crate::Error::from(err));
+                }
             }
         }
 
+        let mut failed_to_delete_files = vec![];
         let mut deleted_files = vec![];
+
         for file_to_delete in files_to_delete {
             match self.delete(&file_to_delete) {
                 Ok(_) => {
@@ -152,9 +162,10 @@ impl ManagedDirectory {
                 Err(file_error) => {
                     match file_error {
                         DeleteError::FileDoesNotExist(_) => {
-                            deleted_files.push(file_to_delete);
+                            deleted_files.push(file_to_delete.clone());
                         }
                         DeleteError::IOError(_) => {
+                            failed_to_delete_files.push(file_to_delete.clone());
                             if !cfg!(target_os = "windows") {
                                 // On windows, delete is expected to fail if the file
                                 // is mmapped.
@@ -177,10 +188,13 @@ impl ManagedDirectory {
             for delete_file in &deleted_files {
                 managed_paths_write.remove(delete_file);
             }
-            if save_managed_paths(self.directory.as_mut(), &meta_informations_wlock).is_err() {
+            save_managed_paths(self.directory.as_mut(), &meta_informations_wlock)?;
-                error!("Failed to save the list of managed files.");
-            }
         }
+
+        Ok(GarbageCollectionResult {
+            deleted_files,
+            failed_to_delete_files,
+        })
     }
 
     /// Registers a file as managed
@@ -327,9 +341,8 @@ mod tests_mmap_specific {
                 .unwrap();
             assert!(managed_directory.exists(test_path1));
             assert!(managed_directory.exists(test_path2));
-            let living_files: HashSet<PathBuf> =
+            let living_files: HashSet<PathBuf> = [test_path1.to_owned()].iter().cloned().collect();
-                [test_path1.to_owned()].into_iter().cloned().collect();
+            assert!(managed_directory.garbage_collect(|| living_files).is_ok());
-            managed_directory.garbage_collect(|| living_files);
             assert!(managed_directory.exists(test_path1));
             assert!(!managed_directory.exists(test_path2));
         }
@@ -339,7 +352,7 @@ mod tests_mmap_specific {
             assert!(managed_directory.exists(test_path1));
             assert!(!managed_directory.exists(test_path2));
             let living_files: HashSet<PathBuf> = HashSet::new();
-            managed_directory.garbage_collect(|| living_files);
+            assert!(managed_directory.garbage_collect(|| living_files).is_ok());
             assert!(!managed_directory.exists(test_path1));
             assert!(!managed_directory.exists(test_path2));
         }
@@ -361,7 +374,9 @@ mod tests_mmap_specific {
         assert!(managed_directory.exists(test_path1));
 
         let _mmap_read = managed_directory.open_read(test_path1).unwrap();
-        managed_directory.garbage_collect(|| living_files.clone());
+        assert!(managed_directory
+            .garbage_collect(|| living_files.clone())
+            .is_ok());
         if cfg!(target_os = "windows") {
             // On Windows, gc should try and fail the file as it is mmapped.
             assert!(managed_directory.exists(test_path1));
@@ -369,7 +384,7 @@ mod tests_mmap_specific {
             drop(_mmap_read);
             // The file should still be in the list of managed file and
             // eventually be deleted once mmap is released.
-            managed_directory.garbage_collect(|| living_files);
+            assert!(managed_directory.garbage_collect(|| living_files).is_ok());
             assert!(!managed_directory.exists(test_path1));
         } else {
             assert!(!managed_directory.exists(test_path1));

src/directory/mmap_directory.rs

@@ -174,7 +174,7 @@ impl WatcherWrapper {
                             // We might want to be more accurate than this at one point.
                             if let Some(filename) = changed_path.file_name() {
                                 if filename == *META_FILEPATH {
-                                    watcher_router_clone.broadcast();
+                                    let _ = watcher_router_clone.broadcast();
                                 }
                             }
                         }
@@ -538,16 +538,15 @@ mod tests {
     // The following tests are specific to the MmapDirectory
 
     use super::*;
+    use crate::indexer::LogMergePolicy;
     use crate::schema::{Schema, SchemaBuilder, TEXT};
     use crate::Index;
     use crate::ReloadPolicy;
     use std::fs;
     use std::sync::atomic::{AtomicUsize, Ordering};
-    use std::thread;
-    use std::time::Duration;
 
     #[test]
-    fn test_open_non_existant_path() {
+    fn test_open_non_existent_path() {
         assert!(MmapDirectory::open(PathBuf::from("./nowhere")).is_err());
     }
 
@@ -640,13 +639,18 @@ mod tests {
         let tmp_dir = tempfile::TempDir::new().unwrap();
         let tmp_dirpath = tmp_dir.path().to_owned();
         let mut watch_wrapper = WatcherWrapper::new(&tmp_dirpath).unwrap();
-        let tmp_file = tmp_dirpath.join("coucou");
+        let tmp_file = tmp_dirpath.join(*META_FILEPATH);
         let _handle = watch_wrapper.watch(Box::new(move || {
             counter_clone.fetch_add(1, Ordering::SeqCst);
         }));
+        let (sender, receiver) = crossbeam::channel::unbounded();
+        let _handle2 = watch_wrapper.watch(Box::new(move || {
+            let _ = sender.send(());
+        }));
         assert_eq!(counter.load(Ordering::SeqCst), 0);
         fs::write(&tmp_file, b"whateverwilldo").unwrap();
-        thread::sleep(Duration::new(0, 1_000u32));
+        assert!(receiver.recv().is_ok());
+        assert!(counter.load(Ordering::SeqCst) >= 1);
     }
 
     #[test]
@@ -655,34 +659,42 @@ mod tests {
         let mut schema_builder: SchemaBuilder = Schema::builder();
         let text_field = schema_builder.add_text_field("text", TEXT);
         let schema = schema_builder.build();
+
         {
             let index = Index::create(mmap_directory.clone(), schema).unwrap();
+
             let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            for _num_commits in 0..16 {
+            let mut log_merge_policy = LogMergePolicy::default();
+            log_merge_policy.set_min_merge_size(3);
+            index_writer.set_merge_policy(Box::new(log_merge_policy));
+            for _num_commits in 0..10 {
                 for _ in 0..10 {
                     index_writer.add_document(doc!(text_field=>"abc"));
                 }
                 index_writer.commit().unwrap();
             }
+
             let reader = index
                 .reader_builder()
                 .reload_policy(ReloadPolicy::Manual)
                 .try_into()
                 .unwrap();
-            for _ in 0..30 {
+
+            for _ in 0..4 {
                 index_writer.add_document(doc!(text_field=>"abc"));
                 index_writer.commit().unwrap();
                 reader.reload().unwrap();
             }
             index_writer.wait_merging_threads().unwrap();
+
             reader.reload().unwrap();
             let num_segments = reader.searcher().segment_readers().len();
-            assert_eq!(num_segments, 4);
+            assert!(num_segments <= 4);
             assert_eq!(
                 num_segments * 7,
                 mmap_directory.get_cache_info().mmapped.len()
             );
         }
-        assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 0);
+        assert!(mmap_directory.get_cache_info().mmapped.is_empty());
     }
 }

src/directory/mod.rs

@@ -26,6 +26,21 @@ pub use self::read_only_source::ReadOnlySource;
 pub(crate) use self::watch_event_router::WatchCallbackList;
 pub use self::watch_event_router::{WatchCallback, WatchHandle};
 use std::io::{self, BufWriter, Write};
+use std::path::PathBuf;
+
+/// Outcome of the Garbage collection
+pub struct GarbageCollectionResult {
+    /// List of files that were deleted in this cycle
+    pub deleted_files: Vec<PathBuf>,
+    /// List of files that were schedule to be deleted in this cycle,
+    /// but deletion did not work. This typically happens on windows,
+    /// as deleting a memory mapped file is forbidden.
+    ///
+    /// If a searcher is still held, a file cannot be deleted.
+    /// This is not considered a bug, the file will simply be deleted
+    /// in the next GC.
+    pub failed_to_delete_files: Vec<PathBuf>,
+}
 
 #[cfg(feature = "mmap")]
 pub use self::mmap_directory::MmapDirectory;

src/directory/ram_directory.rs

@@ -195,7 +195,7 @@ impl Directory for RAMDirectory {
         vec_writer.write_all(data)?;
         vec_writer.flush()?;
         if path == Path::new(&*META_FILEPATH) {
-            self.fs.write().unwrap().watch_router.broadcast();
+            let _ = self.fs.write().unwrap().watch_router.broadcast();
         }
         Ok(())
     }

src/directory/tests.rs

@@ -1,25 +1,117 @@
 use super::*;
+use futures::channel::oneshot;
+use futures::executor::block_on;
 use std::io::Write;
 use std::mem;
 use std::path::{Path, PathBuf};
-use std::sync::atomic::AtomicUsize;
+use std::sync::atomic::Ordering::SeqCst;
-use std::sync::atomic::Ordering;
+use std::sync::atomic::{AtomicBool, AtomicUsize};
 use std::sync::Arc;
-use std::thread;
-use std::time;
 use std::time::Duration;
 
-#[test]
+#[cfg(feature = "mmap")]
-fn test_ram_directory() {
+mod mmap_directory_tests {
-    let mut ram_directory = RAMDirectory::create();
+    use crate::directory::MmapDirectory;
-    test_directory(&mut ram_directory);
+
+    type DirectoryImpl = MmapDirectory;
+
+    fn make_directory() -> DirectoryImpl {
+        MmapDirectory::create_from_tempdir().unwrap()
+    }
+
+    #[test]
+    fn test_simple() {
+        let mut directory = make_directory();
+        super::test_simple(&mut directory);
+    }
+
+    #[test]
+    fn test_write_create_the_file() {
+        let mut directory = make_directory();
+        super::test_write_create_the_file(&mut directory);
+    }
+
+    #[test]
+    fn test_rewrite_forbidden() {
+        let mut directory = make_directory();
+        super::test_rewrite_forbidden(&mut directory);
+    }
+
+    #[test]
+    fn test_directory_delete() {
+        let mut directory = make_directory();
+        super::test_directory_delete(&mut directory);
+    }
+
+    #[test]
+    fn test_lock_non_blocking() {
+        let mut directory = make_directory();
+        super::test_lock_non_blocking(&mut directory);
+    }
+
+    #[test]
+    fn test_lock_blocking() {
+        let mut directory = make_directory();
+        super::test_lock_blocking(&mut directory);
+    }
+
+    #[test]
+    fn test_watch() {
+        let mut directory = make_directory();
+        super::test_watch(&mut directory);
+    }
 }
 
-#[test]
+mod ram_directory_tests {
-#[cfg(feature = "mmap")]
+    use crate::directory::RAMDirectory;
-fn test_mmap_directory() {
+
-    let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
+    type DirectoryImpl = RAMDirectory;
-    test_directory(&mut mmap_directory);
+
+    fn make_directory() -> DirectoryImpl {
+        RAMDirectory::default()
+    }
+
+    #[test]
+    fn test_simple() {
+        let mut directory = make_directory();
+        super::test_simple(&mut directory);
+    }
+
+    #[test]
+    fn test_write_create_the_file() {
+        let mut directory = make_directory();
+        super::test_write_create_the_file(&mut directory);
+    }
+
+    #[test]
+    fn test_rewrite_forbidden() {
+        let mut directory = make_directory();
+        super::test_rewrite_forbidden(&mut directory);
+    }
+
+    #[test]
+    fn test_directory_delete() {
+        let mut directory = make_directory();
+        super::test_directory_delete(&mut directory);
+    }
+
+    #[test]
+    fn test_lock_non_blocking() {
+        let mut directory = make_directory();
+        super::test_lock_non_blocking(&mut directory);
+    }
+
+    #[test]
+    fn test_lock_blocking() {
+        let mut directory = make_directory();
+        super::test_lock_blocking(&mut directory);
+    }
+
+    #[test]
+    fn test_watch() {
+        let mut directory = make_directory();
+        super::test_watch(&mut directory);
+    }
 }
 
 #[test]
@@ -99,48 +191,39 @@ fn test_directory_delete(directory: &mut dyn Directory) {
     assert!(directory.delete(&test_path).is_err());
 }
 
-fn test_directory(directory: &mut dyn Directory) {
-    test_simple(directory);
-    test_rewrite_forbidden(directory);
-    test_write_create_the_file(directory);
-    test_directory_delete(directory);
-    test_lock_non_blocking(directory);
-    test_lock_blocking(directory);
-    test_watch(directory);
-}
-
 fn test_watch(directory: &mut dyn Directory) {
+    let num_progress: Arc<AtomicUsize> = Default::default();
     let counter: Arc<AtomicUsize> = Default::default();
     let counter_clone = counter.clone();
+    let (sender, receiver) = crossbeam::channel::unbounded();
     let watch_callback = Box::new(move || {
-        counter_clone.fetch_add(1, Ordering::SeqCst);
+        counter_clone.fetch_add(1, SeqCst);
     });
-    assert!(directory
+    // This callback is used to synchronize watching in our unit test.
-        .atomic_write(Path::new("meta.json"), b"random_test_data")
+    // We bind it to a variable because the callback is removed when that
-        .is_ok());
+    // handle is dropped.
-    thread::sleep(Duration::new(0, 10_000));
-    assert_eq!(0, counter.load(Ordering::SeqCst));
-
     let watch_handle = directory.watch(watch_callback).unwrap();
+    let _progress_listener = directory
+        .watch(Box::new(move || {
+            let val = num_progress.fetch_add(1, SeqCst);
+            let _ = sender.send(val);
+        }))
+        .unwrap();
+
     for i in 0..10 {
-        assert_eq!(i, counter.load(Ordering::SeqCst));
+        assert_eq!(i, counter.load(SeqCst));
         assert!(directory
             .atomic_write(Path::new("meta.json"), b"random_test_data_2")
             .is_ok());
-        for _ in 0..1_000 {
+        assert_eq!(receiver.recv_timeout(Duration::from_millis(500)), Ok(i));
-            if counter.load(Ordering::SeqCst) > i {
+        assert_eq!(i + 1, counter.load(SeqCst));
-                break;
-            }
-            thread::sleep(Duration::from_millis(10));
-        }
-        assert_eq!(i + 1, counter.load(Ordering::SeqCst));
     }
     mem::drop(watch_handle);
     assert!(directory
         .atomic_write(Path::new("meta.json"), b"random_test_data")
         .is_ok());
-    thread::sleep(Duration::from_millis(200));
+    assert!(receiver.recv_timeout(Duration::from_millis(500)).is_ok());
-    assert_eq!(10, counter.load(Ordering::SeqCst));
+    assert_eq!(10, counter.load(SeqCst));
 }
 
 fn test_lock_non_blocking(directory: &mut dyn Directory) {
@@ -174,9 +257,13 @@ fn test_lock_blocking(directory: &mut dyn Directory) {
         is_blocking: true,
     });
     assert!(lock_a_res.is_ok());
+    let in_thread = Arc::new(AtomicBool::default());
+    let in_thread_clone = in_thread.clone();
+    let (sender, receiver) = oneshot::channel();
     std::thread::spawn(move || {
         //< lock_a_res is sent to the thread.
-        std::thread::sleep(time::Duration::from_millis(10));
+        in_thread_clone.store(true, SeqCst);
+        let _just_sync = block_on(receiver);
         // 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);
@@ -189,14 +276,18 @@ fn test_lock_blocking(directory: &mut dyn Directory) {
         });
         assert!(lock_a_res.is_err());
     }
-    {
+    let directory_clone = directory.box_clone();
-        // the blocking call should wait for at least 10ms.
+    let (sender2, receiver2) = oneshot::channel();
-        let start = time::Instant::now();
+    let join_handle = std::thread::spawn(move || {
-        let lock_a_res = directory.acquire_lock(&Lock {
+        assert!(sender2.send(()).is_ok());
+        let lock_a_res = directory_clone.acquire_lock(&Lock {
             filepath: PathBuf::from("a.lock"),
             is_blocking: true,
         });
+        assert!(in_thread.load(SeqCst));
         assert!(lock_a_res.is_ok());
-        assert!(start.elapsed().subsec_millis() >= 10);
+    });
-    }
+    assert!(block_on(receiver2).is_ok());
+    assert!(sender.send(()).is_ok());
+    assert!(join_handle.join().is_ok());
 }

src/directory/watch_event_router.rs

@@ -1,3 +1,5 @@
+use futures::channel::oneshot;
+use futures::{Future, TryFutureExt};
 use std::sync::Arc;
 use std::sync::RwLock;
 use std::sync::Weak;
@@ -47,14 +49,21 @@ impl WatchCallbackList {
     }
 
     /// Triggers all callbacks
-    pub fn broadcast(&self) {
+    pub fn broadcast(&self) -> impl Future<Output = ()> {
         let callbacks = self.list_callback();
+        let (sender, receiver) = oneshot::channel();
+        let result = receiver.unwrap_or_else(|_| ());
+        if callbacks.is_empty() {
+            let _ = sender.send(());
+            return result;
+        }
         let spawn_res = std::thread::Builder::new()
             .name("watch-callbacks".to_string())
             .spawn(move || {
                 for callback in callbacks {
                     callback();
                 }
+                let _ = sender.send(());
             });
         if let Err(err) = spawn_res {
             error!(
@@ -62,19 +71,17 @@ impl WatchCallbackList {
                 err
             );
         }
+        result
     }
 }
 
 #[cfg(test)]
 mod tests {
     use crate::directory::WatchCallbackList;
+    use futures::executor::block_on;
     use std::mem;
     use std::sync::atomic::{AtomicUsize, Ordering};
     use std::sync::Arc;
-    use std::thread;
-    use std::time::Duration;
-
-    const WAIT_TIME: u64 = 20;
 
     #[test]
     fn test_watch_event_router_simple() {
@@ -84,22 +91,22 @@ mod tests {
         let inc_callback = Box::new(move || {
             counter_clone.fetch_add(1, Ordering::SeqCst);
         });
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
         assert_eq!(0, counter.load(Ordering::SeqCst));
         let handle_a = watch_event_router.subscribe(inc_callback);
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(0, counter.load(Ordering::SeqCst));
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(1, counter.load(Ordering::SeqCst));
-        watch_event_router.broadcast();
+        block_on(async {
-        watch_event_router.broadcast();
+            (
-        watch_event_router.broadcast();
+                watch_event_router.broadcast().await,
-        thread::sleep(Duration::from_millis(WAIT_TIME));
+                watch_event_router.broadcast().await,
+                watch_event_router.broadcast().await,
+            )
+        });
         assert_eq!(4, counter.load(Ordering::SeqCst));
         mem::drop(handle_a);
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(4, counter.load(Ordering::SeqCst));
     }
 
@@ -115,20 +122,20 @@ mod tests {
         };
         let handle_a = watch_event_router.subscribe(inc_callback(1));
         let handle_a2 = watch_event_router.subscribe(inc_callback(10));
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(0, counter.load(Ordering::SeqCst));
-        watch_event_router.broadcast();
+        block_on(async {
-        watch_event_router.broadcast();
+            futures::join!(
-        thread::sleep(Duration::from_millis(WAIT_TIME));
+                watch_event_router.broadcast(),
+                watch_event_router.broadcast()
+            )
+        });
         assert_eq!(22, counter.load(Ordering::SeqCst));
         mem::drop(handle_a);
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(32, counter.load(Ordering::SeqCst));
         mem::drop(handle_a2);
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
-        watch_event_router.broadcast();
+        block_on(watch_event_router.broadcast());
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         assert_eq!(32, counter.load(Ordering::SeqCst));
     }
 
@@ -142,14 +149,15 @@ mod tests {
         });
         let handle_a = watch_event_router.subscribe(inc_callback);
         assert_eq!(0, counter.load(Ordering::SeqCst));
-        watch_event_router.broadcast();
+        block_on(async {
-        watch_event_router.broadcast();
+            let future1 = watch_event_router.broadcast();
-        thread::sleep(Duration::from_millis(WAIT_TIME));
+            let future2 = watch_event_router.broadcast();
+            futures::join!(future1, future2)
+        });
         assert_eq!(2, counter.load(Ordering::SeqCst));
-        thread::sleep(Duration::from_millis(WAIT_TIME));
         mem::drop(handle_a);
-        watch_event_router.broadcast();
+        let _ = watch_event_router.broadcast();
-        thread::sleep(Duration::from_millis(WAIT_TIME));
+        block_on(watch_event_router.broadcast());
         assert_eq!(2, counter.load(Ordering::SeqCst));
     }
 }

src/error.rs

@@ -170,3 +170,9 @@ impl From<serde_json::Error> for TantivyError {
         TantivyError::IOError(io_err.into())
     }
 }
+
+impl From<rayon::ThreadPoolBuildError> for TantivyError {
+    fn from(error: rayon::ThreadPoolBuildError) -> TantivyError {
+        TantivyError::SystemError(error.to_string())
+    }
+}

src/fastfield/delete.rs

@@ -10,11 +10,14 @@ use std::io::Write;
 /// Write a delete `BitSet`
 ///
 /// where `delete_bitset` is the set of deleted `DocId`.
-pub fn write_delete_bitset(delete_bitset: &BitSet, writer: &mut WritePtr) -> io::Result<()> {
+pub fn write_delete_bitset(
-    let max_doc = delete_bitset.capacity();
+    delete_bitset: &BitSet,
+    max_doc: u32,
+    writer: &mut WritePtr,
+) -> io::Result<()> {
     let mut byte = 0u8;
     let mut shift = 0u8;
-    for doc in 0..max_doc {
+    for doc in 0..(max_doc as usize) {
         if delete_bitset.contains(doc) {
             byte |= 1 << shift;
         }
@@ -86,18 +89,17 @@ mod tests {
     use bit_set::BitSet;
     use std::path::PathBuf;
 
-    fn test_delete_bitset_helper(bitset: &BitSet) {
+    fn test_delete_bitset_helper(bitset: &BitSet, max_doc: u32) {
         let test_path = PathBuf::from("test");
         let mut directory = RAMDirectory::create();
         {
             let mut writer = directory.open_write(&*test_path).unwrap();
-            write_delete_bitset(bitset, &mut writer).unwrap();
+            write_delete_bitset(bitset, max_doc, &mut writer).unwrap();
         }
         {
             let source = directory.open_read(&test_path).unwrap();
             let delete_bitset = DeleteBitSet::open(source);
-            let n = bitset.capacity();
+            for doc in 0..max_doc as usize {
-            for doc in 0..n {
                 assert_eq!(bitset.contains(doc), delete_bitset.is_deleted(doc as DocId));
             }
             assert_eq!(delete_bitset.len(), bitset.len());
@@ -110,7 +112,7 @@ mod tests {
             let mut bitset = BitSet::with_capacity(10);
             bitset.insert(1);
             bitset.insert(9);
-            test_delete_bitset_helper(&bitset);
+            test_delete_bitset_helper(&bitset, 10);
         }
         {
             let mut bitset = BitSet::with_capacity(8);
@@ -119,7 +121,7 @@ mod tests {
             bitset.insert(3);
             bitset.insert(5);
             bitset.insert(7);
-            test_delete_bitset_helper(&bitset);
+            test_delete_bitset_helper(&bitset, 8);
         }
     }
 }

src/indexer/index_writer.rs

@@ -7,8 +7,8 @@ use crate::core::SegmentComponent;
 use crate::core::SegmentId;
 use crate::core::SegmentMeta;
 use crate::core::SegmentReader;
-use crate::directory::DirectoryLock;
 use crate::directory::TerminatingWrite;
+use crate::directory::{DirectoryLock, GarbageCollectionResult};
 use crate::docset::DocSet;
 use crate::error::TantivyError;
 use crate::fastfield::write_delete_bitset;
@@ -23,10 +23,10 @@ use crate::schema::Document;
 use crate::schema::IndexRecordOption;
 use crate::schema::Term;
 use crate::Opstamp;
-use crate::Result;
 use bit_set::BitSet;
 use crossbeam::channel;
-use futures::{Canceled, Future};
+use futures::executor::block_on;
+use futures::future::Future;
 use smallvec::smallvec;
 use smallvec::SmallVec;
 use std::mem;
@@ -72,7 +72,7 @@ pub struct IndexWriter {
 
     heap_size_in_bytes_per_thread: usize,
 
-    workers_join_handle: Vec<JoinHandle<Result<()>>>,
+    workers_join_handle: Vec<JoinHandle<crate::Result<()>>>,
 
     operation_receiver: OperationReceiver,
     operation_sender: OperationSender,
@@ -95,7 +95,7 @@ fn compute_deleted_bitset(
     delete_cursor: &mut DeleteCursor,
     doc_opstamps: &DocToOpstampMapping,
     target_opstamp: Opstamp,
-) -> Result<bool> {
+) -> crate::Result<bool> {
     let mut might_have_changed = false;
     while let Some(delete_op) = delete_cursor.get() {
         if delete_op.opstamp > target_opstamp {
@@ -132,7 +132,7 @@ pub(crate) fn advance_deletes(
     mut segment: Segment,
     segment_entry: &mut SegmentEntry,
     target_opstamp: Opstamp,
-) -> Result<()> {
+) -> crate::Result<()> {
     {
         if segment_entry.meta().delete_opstamp() == Some(target_opstamp) {
             // We are already up-to-date here.
@@ -148,7 +148,6 @@ pub(crate) fn advance_deletes(
         };
 
         let delete_cursor = segment_entry.delete_cursor();
-
         compute_deleted_bitset(
             &mut delete_bitset,
             &segment_reader,
@@ -168,7 +167,7 @@ pub(crate) fn advance_deletes(
         if num_deleted_docs > 0 {
             segment = segment.with_delete_meta(num_deleted_docs as u32, target_opstamp);
             let mut delete_file = segment.open_write(SegmentComponent::DELETE)?;
-            write_delete_bitset(&delete_bitset, &mut delete_file)?;
+            write_delete_bitset(&delete_bitset, max_doc, &mut delete_file)?;
             delete_file.terminate()?;
         }
     }
@@ -178,13 +177,13 @@ pub(crate) fn advance_deletes(
 
 fn index_documents(
     memory_budget: usize,
-    segment: &Segment,
+    segment: Segment,
     grouped_document_iterator: &mut dyn Iterator<Item = OperationGroup>,
     segment_updater: &mut SegmentUpdater,
     mut delete_cursor: DeleteCursor,
-) -> Result<bool> {
+) -> crate::Result<bool> {
     let schema = segment.schema();
-    let segment_id = segment.id();
+
     let mut segment_writer = SegmentWriter::for_segment(memory_budget, segment.clone(), &schema)?;
     for document_group in grouped_document_iterator {
         for doc in document_group {
@@ -204,22 +203,32 @@ fn index_documents(
         return Ok(false);
     }
 
-    let num_docs = segment_writer.max_doc();
+    let max_doc = segment_writer.max_doc();
 
     // this is ensured by the call to peek before starting
     // the worker thread.
-    assert!(num_docs > 0);
+    assert!(max_doc > 0);
 
     let doc_opstamps: Vec<Opstamp> = segment_writer.finalize()?;
-    let segment_meta = segment.index().new_segment_meta(segment_id, num_docs);
+
+    let segment_with_max_doc = segment.with_max_doc(max_doc);
 
     let last_docstamp: Opstamp = *(doc_opstamps.last().unwrap());
 
-    let delete_bitset_opt =
+    let delete_bitset_opt = apply_deletes(
-        apply_deletes(&segment, &mut delete_cursor, &doc_opstamps, last_docstamp)?;
+        &segment_with_max_doc,
+        &mut delete_cursor,
+        &doc_opstamps,
+        last_docstamp,
+    )?;
 
-    let segment_entry = SegmentEntry::new(segment_meta, delete_cursor, delete_bitset_opt);
+    let segment_entry = SegmentEntry::new(
-    Ok(segment_updater.add_segment(segment_entry))
+        segment_with_max_doc.meta().clone(),
+        delete_cursor,
+        delete_bitset_opt,
+    );
+    block_on(segment_updater.schedule_add_segment(segment_entry))?;
+    Ok(true)
 }
 
 fn apply_deletes(
@@ -227,7 +236,7 @@ fn apply_deletes(
     mut delete_cursor: &mut DeleteCursor,
     doc_opstamps: &[Opstamp],
     last_docstamp: Opstamp,
-) -> Result<Option<BitSet<u32>>> {
+) -> crate::Result<Option<BitSet<u32>>> {
     if delete_cursor.get().is_none() {
         // if there are no delete operation in the queue, no need
         // to even open the segment.
@@ -235,7 +244,9 @@ fn apply_deletes(
     }
     let segment_reader = SegmentReader::open(segment)?;
     let doc_to_opstamps = DocToOpstampMapping::from(doc_opstamps);
-    let mut deleted_bitset = BitSet::with_capacity(segment_reader.max_doc() as usize);
+
+    let max_doc = segment.meta().max_doc();
+    let mut deleted_bitset = BitSet::with_capacity(max_doc as usize);
     let may_have_deletes = compute_deleted_bitset(
         &mut deleted_bitset,
         &segment_reader,
@@ -270,7 +281,7 @@ impl IndexWriter {
         num_threads: usize,
         heap_size_in_bytes_per_thread: usize,
         directory_lock: DirectoryLock,
-    ) -> Result<IndexWriter> {
+    ) -> crate::Result<IndexWriter> {
         if heap_size_in_bytes_per_thread < HEAP_SIZE_MIN {
             let err_msg = format!(
                 "The heap size per thread needs to be at least {}.",
@@ -321,7 +332,7 @@ impl IndexWriter {
 
     /// If there are some merging threads, blocks until they all finish their work and
     /// then drop the `IndexWriter`.
-    pub fn wait_merging_threads(mut self) -> Result<()> {
+    pub fn wait_merging_threads(mut self) -> crate::Result<()> {
         // this will stop the indexing thread,
         // dropping the last reference to the segment_updater.
         drop(self.operation_sender);
@@ -350,10 +361,10 @@ impl IndexWriter {
     }
 
     #[doc(hidden)]
-    pub fn add_segment(&mut self, segment_meta: SegmentMeta) {
+    pub fn add_segment(&self, segment_meta: SegmentMeta) -> crate::Result<()> {
         let delete_cursor = self.delete_queue.cursor();
         let segment_entry = SegmentEntry::new(segment_meta, delete_cursor, None);
-        self.segment_updater.add_segment(segment_entry);
+        block_on(self.segment_updater.schedule_add_segment(segment_entry))
     }
 
     /// Creates a new segment.
@@ -370,7 +381,7 @@ impl IndexWriter {
 
     /// Spawns a new worker thread for indexing.
     /// The thread consumes documents from the pipeline.
-    fn add_indexing_worker(&mut self) -> Result<()> {
+    fn add_indexing_worker(&mut self) -> crate::Result<()> {
         let document_receiver_clone = self.operation_receiver.clone();
         let mut segment_updater = self.segment_updater.clone();
 
@@ -378,7 +389,7 @@ impl IndexWriter {
 
         let mem_budget = self.heap_size_in_bytes_per_thread;
         let index = self.index.clone();
-        let join_handle: JoinHandle<Result<()>> = thread::Builder::new()
+        let join_handle: JoinHandle<crate::Result<()>> = thread::Builder::new()
             .name(format!("thrd-tantivy-index{}", self.worker_id))
             .spawn(move || {
                 loop {
@@ -407,7 +418,7 @@ impl IndexWriter {
                     let segment = index.new_segment();
                     index_documents(
                         mem_budget,
-                        &segment,
+                        segment,
                         &mut document_iterator,
                         &mut segment_updater,
                         delete_cursor.clone(),
@@ -424,22 +435,23 @@ impl IndexWriter {
         self.segment_updater.get_merge_policy()
     }
 
-    /// Set the merge policy.
+    /// Setter for the merge policy.
     pub fn set_merge_policy(&self, merge_policy: Box<dyn MergePolicy>) {
         self.segment_updater.set_merge_policy(merge_policy);
     }
 
-    fn start_workers(&mut self) -> Result<()> {
+    fn start_workers(&mut self) -> crate::Result<()> {
         for _ in 0..self.num_threads {
             self.add_indexing_worker()?;
         }
         Ok(())
     }
 
-    /// Detects and removes the files that
+    /// Detects and removes the files that are not used by the index anymore.
-    /// are not used by the index anymore.
+    pub fn garbage_collect_files(
-    pub fn garbage_collect_files(&mut self) -> Result<()> {
+        &self,
-        self.segment_updater.garbage_collect_files().wait()
+    ) -> impl Future<Output = crate::Result<GarbageCollectionResult>> {
+        self.segment_updater.schedule_garbage_collect()
     }
 
     /// Deletes all documents from the index
@@ -478,7 +490,7 @@ impl IndexWriter {
     ///     Ok(())
     /// }
     /// ```
-    pub fn delete_all_documents(&mut self) -> Result<Opstamp> {
+    pub fn delete_all_documents(&self) -> crate::Result<Opstamp> {
         // Delete segments
         self.segment_updater.remove_all_segments();
         // Return new stamp - reverted stamp
@@ -489,11 +501,9 @@ impl IndexWriter {
     /// Merges a given list of segments
     ///
     /// `segment_ids` is required to be non-empty.
-    pub fn merge(
+    pub async fn merge(&mut self, segment_ids: &[SegmentId]) -> crate::Result<SegmentMeta> {
-        &mut self,
+        let merge_operation = self.segment_updater.make_merge_operation(segment_ids);
-        segment_ids: &[SegmentId],
+        self.segment_updater.start_merge(merge_operation)?.await
-    ) -> Result<impl Future<Item = SegmentMeta, Error = Canceled>> {
-        self.segment_updater.start_merge(segment_ids)
     }
 
     /// Closes the current document channel send.
@@ -519,13 +529,8 @@ impl IndexWriter {
     /// state as it was after the last commit.
     ///
     /// The opstamp at the last commit is returned.
-    pub fn rollback(&mut self) -> Result<Opstamp> {
+    pub fn rollback(&mut self) -> crate::Result<Opstamp> {
         info!("Rolling back to opstamp {}", self.committed_opstamp);
-        self.rollback_impl()
-    }
-
-    /// Private, implementation of rollback
-    fn rollback_impl(&mut self) -> Result<Opstamp> {
         // marks the segment updater as killed. From now on, all
         // segment updates will be ignored.
         self.segment_updater.kill();
@@ -581,7 +586,7 @@ impl IndexWriter {
     /// It is also possible to add a payload to the `commit`
     /// using this API.
     /// See [`PreparedCommit::set_payload()`](PreparedCommit.html)
-    pub fn prepare_commit(&mut self) -> Result<PreparedCommit<'_>> {
+    pub fn prepare_commit(&mut self) -> crate::Result<PreparedCommit> {
         // Here, because we join all of the worker threads,
         // all of the segment update for this commit have been
         // sent.
@@ -628,7 +633,7 @@ impl IndexWriter {
     /// Commit returns the `opstamp` of the last document
     /// that made it in the commit.
     ///
-    pub fn commit(&mut self) -> Result<Opstamp> {
+    pub fn commit(&mut self) -> crate::Result<Opstamp> {
         self.prepare_commit()?.commit()
     }
 
@@ -669,9 +674,6 @@ impl IndexWriter {
     /// The opstamp is an increasing `u64` that can
     /// be used by the client to align commits with its own
     /// document queue.
-    ///
-    /// Currently it represents the number of documents that
-    /// have been added since the creation of the index.
     pub fn add_document(&self, document: Document) -> Opstamp {
         let opstamp = self.stamper.stamp();
         let add_operation = AddOperation { opstamp, document };

src/indexer/merge_operation.rs

@@ -2,14 +2,23 @@ use crate::Opstamp;
 use crate::SegmentId;
 use census::{Inventory, TrackedObject};
 use std::collections::HashSet;
+use std::ops::Deref;
 
 #[derive(Default)]
-pub struct MergeOperationInventory(Inventory<InnerMergeOperation>);
+pub(crate) struct MergeOperationInventory(Inventory<InnerMergeOperation>);
+
+impl Deref for MergeOperationInventory {
+    type Target = Inventory<InnerMergeOperation>;
+
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
 
 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 merge_op in self.list() {
             for &segment_id in &merge_op.segment_ids {
                 segment_in_merge.insert(segment_id);
             }
@@ -35,13 +44,13 @@ pub struct MergeOperation {
     inner: TrackedObject<InnerMergeOperation>,
 }
 
-struct InnerMergeOperation {
+pub(crate) struct InnerMergeOperation {
     target_opstamp: Opstamp,
     segment_ids: Vec<SegmentId>,
 }
 
 impl MergeOperation {
-    pub fn new(
+    pub(crate) fn new(
         inventory: &MergeOperationInventory,
         target_opstamp: Opstamp,
         segment_ids: Vec<SegmentId>,
@@ -51,7 +60,7 @@ impl MergeOperation {
             segment_ids,
         };
         MergeOperation {
-            inner: inventory.0.track(inner_merge_operation),
+            inner: inventory.track(inner_merge_operation),
         }
     }
 

src/indexer/merger.rs

@@ -709,7 +709,7 @@ mod tests {
     use crate::IndexWriter;
     use crate::Searcher;
     use byteorder::{BigEndian, ReadBytesExt, WriteBytesExt};
-    use futures::Future;
+    use futures::executor::block_on;
     use std::io::Cursor;
 
     #[test]
@@ -792,11 +792,7 @@ mod tests {
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
             let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            index_writer
+            block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
             index_writer.wait_merging_threads().unwrap();
         }
         {
@@ -1040,11 +1036,7 @@ mod tests {
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            index_writer
+            block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
             reader.reload().unwrap();
             let searcher = reader.searcher();
             assert_eq!(searcher.segment_readers().len(), 1);
@@ -1139,11 +1131,7 @@ mod tests {
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            index_writer
+            block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
             reader.reload().unwrap();
 
             let searcher = reader.searcher();
@@ -1277,11 +1265,7 @@ mod tests {
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
             let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            index_writer
+            block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
             index_writer.wait_merging_threads().unwrap();
             reader.reload().unwrap();
             test_searcher(
@@ -1336,11 +1320,7 @@ mod tests {
         let segment_ids = index
             .searchable_segment_ids()
             .expect("Searchable segments failed.");
-        index_writer
+        block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
-            .merge(&segment_ids)
-            .expect("Failed to initiate merge")
-            .wait()
-            .expect("Merging failed");
         reader.reload().unwrap();
         // commit has not been called yet. The document should still be
         // there.
@@ -1361,22 +1341,18 @@ mod tests {
             let mut doc = Document::default();
             doc.add_u64(int_field, 1);
             index_writer.add_document(doc.clone());
-            index_writer.commit().expect("commit failed");
+            assert!(index_writer.commit().is_ok());
             index_writer.add_document(doc);
-            index_writer.commit().expect("commit failed");
+            assert!(index_writer.commit().is_ok());
             index_writer.delete_term(Term::from_field_u64(int_field, 1));
 
             let segment_ids = index
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
-            index_writer
+            assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
-                .merge(&segment_ids)
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
 
             // assert delete has not been committed
-            reader.reload().expect("failed to load searcher 1");
+            assert!(reader.reload().is_ok());
             let searcher = reader.searcher();
             assert_eq!(searcher.num_docs(), 2);
 
@@ -1415,12 +1391,12 @@ mod tests {
             index_doc(&mut index_writer, &[1, 5]);
             index_doc(&mut index_writer, &[3]);
             index_doc(&mut index_writer, &[17]);
-            index_writer.commit().expect("committed");
+            assert!(index_writer.commit().is_ok());
             index_doc(&mut index_writer, &[20]);
-            index_writer.commit().expect("committed");
+            assert!(index_writer.commit().is_ok());
             index_doc(&mut index_writer, &[28, 27]);
             index_doc(&mut index_writer, &[1_000]);
-            index_writer.commit().expect("committed");
+            assert!(index_writer.commit().is_ok());
         }
         let reader = index.reader().unwrap();
         let searcher = reader.searcher();
@@ -1452,15 +1428,6 @@ mod tests {
             assert_eq!(&vals, &[17]);
         }
 
-        println!(
-            "{:?}",
-            searcher
-                .segment_readers()
-                .iter()
-                .map(|reader| reader.max_doc())
-                .collect::<Vec<_>>()
-        );
-
         {
             let segment = searcher.segment_reader(1u32);
             let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
@@ -1484,27 +1451,13 @@ mod tests {
                 .searchable_segment_ids()
                 .expect("Searchable segments failed.");
             let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
-            index_writer
+            assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
-                .merge(&segment_ids)
+            assert!(index_writer.wait_merging_threads().is_ok());
-                .expect("Failed to initiate merge")
-                .wait()
-                .expect("Merging failed");
-            index_writer
-                .wait_merging_threads()
-                .expect("Wait for merging threads");
         }
-        reader.reload().expect("Load searcher");
+        assert!(reader.reload().is_ok());
 
         {
             let searcher = reader.searcher();
-            println!(
-                "{:?}",
-                searcher
-                    .segment_readers()
-                    .iter()
-                    .map(|reader| reader.max_doc())
-                    .collect::<Vec<_>>()
-            );
             let segment = searcher.segment_reader(0u32);
             let ff_reader = segment.fast_fields().u64s(int_field).unwrap();
 

src/indexer/mod.rs

@@ -18,7 +18,7 @@ mod stamper;
 
 pub use self::index_writer::IndexWriter;
 pub use self::log_merge_policy::LogMergePolicy;
-pub use self::merge_operation::{MergeOperation, MergeOperationInventory};
+pub use self::merge_operation::MergeOperation;
 pub use self::merge_policy::{MergeCandidate, MergePolicy, NoMergePolicy};
 pub use self::prepared_commit::PreparedCommit;
 pub use self::segment_entry::SegmentEntry;
@@ -28,3 +28,25 @@ pub use self::segment_writer::SegmentWriter;
 
 /// Alias for the default merge policy, which is the `LogMergePolicy`.
 pub type DefaultMergePolicy = LogMergePolicy;
+
+#[cfg(test)]
+mod tests {
+    use crate::schema::{self, Schema};
+    use crate::{Index, Term};
+    #[test]
+    fn test_advance_delete_bug() {
+        let mut schema_builder = Schema::builder();
+        let text_field = schema_builder.add_text_field("text", schema::TEXT);
+        let index = Index::create_from_tempdir(schema_builder.build()).unwrap();
+        let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
+        // there must be one deleted document in the segment
+        index_writer.add_document(doc!(text_field=>"b"));
+        index_writer.delete_term(Term::from_field_text(text_field, "b"));
+        // we need enough data to trigger the bug (at least 32 documents)
+        for _ in 0..32 {
+            index_writer.add_document(doc!(text_field=>"c"));
+        }
+        index_writer.commit().unwrap();
+        index_writer.commit().unwrap();
+    }
+}

src/indexer/prepared_commit.rs

@@ -1,6 +1,7 @@
 use super::IndexWriter;
 use crate::Opstamp;
 use crate::Result;
+use futures::executor::block_on;
 
 /// A prepared commit
 pub struct PreparedCommit<'a> {
@@ -32,9 +33,11 @@ impl<'a> PreparedCommit<'a> {
 
     pub fn commit(self) -> Result<Opstamp> {
         info!("committing {}", self.opstamp);
-        self.index_writer
+        let _ = block_on(
-            .segment_updater()
+            self.index_writer
-            .commit(self.opstamp, self.payload)?;
+                .segment_updater()
+                .schedule_commit(self.opstamp, self.payload),
+        );
         Ok(self.opstamp)
     }
 }

src/indexer/segment_manager.rs

@@ -16,6 +16,28 @@ struct SegmentRegisters {
     committed: SegmentRegister,
 }
 
+#[derive(PartialEq, Eq)]
+pub(crate) enum SegmentsStatus {
+    Committed,
+    Uncommitted,
+}
+
+impl SegmentRegisters {
+    /// Check if all the segments are committed or uncommited.
+    ///
+    /// If some segment is missing or segments are in a different state (this should not happen
+    /// if tantivy is used correctly), returns `None`.
+    fn segments_status(&self, segment_ids: &[SegmentId]) -> Option<SegmentsStatus> {
+        if self.uncommitted.contains_all(segment_ids) {
+            Some(SegmentsStatus::Uncommitted)
+        } else if self.committed.contains_all(segment_ids) {
+            Some(SegmentsStatus::Committed)
+        } else {
+            None
+        }
+    }
+}
+
 /// The segment manager stores the list of segments
 /// as well as their state.
 ///
@@ -153,33 +175,35 @@ impl SegmentManager {
         let mut registers_lock = self.write();
         registers_lock.uncommitted.add_segment_entry(segment_entry);
     }
-
+    // Replace a list of segments for their equivalent merged segment.
-    pub fn end_merge(
+    //
+    // Returns true if these segments are committed, false if the merge segments are uncommited.
+    pub(crate) fn end_merge(
         &self,
         before_merge_segment_ids: &[SegmentId],
         after_merge_segment_entry: SegmentEntry,
-    ) {
+    ) -> crate::Result<SegmentsStatus> {
         let mut registers_lock = self.write();
-        let target_register: &mut SegmentRegister = {
+        let segments_status = registers_lock
-            if registers_lock
+            .segments_status(before_merge_segment_ids)
-                .uncommitted
+            .ok_or_else(|| {
-                .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;
+                crate::Error::InvalidArgument(
-            }
+                    "The segments that were merged could not be found in the SegmentManager. \
+                     This is not necessarily a bug, and can happen after a rollback for instance."
+                        .to_string(),
+                )
+            })?;
+
+        let target_register: &mut SegmentRegister = match segments_status {
+            SegmentsStatus::Uncommitted => &mut registers_lock.uncommitted,
+            SegmentsStatus::Committed => &mut registers_lock.committed,
         };
         for segment_id in before_merge_segment_ids {
             target_register.remove_segment(segment_id);
         }
         target_register.add_segment_entry(after_merge_segment_entry);
+        Ok(segments_status)
     }
 
     pub fn committed_segment_metas(&self) -> Vec<SegmentMeta> {

src/indexer/segment_updater.rs

@@ -6,39 +6,34 @@ use crate::core::SegmentId;
 use crate::core::SegmentMeta;
 use crate::core::SerializableSegment;
 use crate::core::META_FILEPATH;
-use crate::directory::{Directory, DirectoryClone};
+use crate::directory::{Directory, DirectoryClone, GarbageCollectionResult};
-use crate::error::TantivyError;
 use crate::indexer::delete_queue::DeleteCursor;
 use crate::indexer::index_writer::advance_deletes;
 use crate::indexer::merge_operation::MergeOperationInventory;
 use crate::indexer::merger::IndexMerger;
+use crate::indexer::segment_manager::SegmentsStatus;
 use crate::indexer::stamper::Stamper;
-use crate::indexer::MergeOperation;
 use crate::indexer::SegmentEntry;
 use crate::indexer::SegmentSerializer;
 use crate::indexer::{DefaultMergePolicy, MergePolicy};
+use crate::indexer::{MergeCandidate, MergeOperation};
 use crate::schema::Schema;
 use crate::Opstamp;
-use crate::Result;
+use futures::channel::oneshot;
-use futures::oneshot;
+use futures::executor::{ThreadPool, ThreadPoolBuilder};
-use futures::sync::oneshot::Receiver;
+use futures::future::Future;
-use futures::Future;
+use futures::future::TryFutureExt;
-use futures_cpupool::Builder as CpuPoolBuilder;
-use futures_cpupool::CpuFuture;
-use futures_cpupool::CpuPool;
 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::Deref;
-use std::ops::DerefMut;
 use std::path::PathBuf;
-use std::sync::atomic::{AtomicBool, AtomicUsize, Ordering};
+use std::sync::atomic::{AtomicBool, Ordering};
 use std::sync::Arc;
 use std::sync::RwLock;
-use std::thread;
+
-use std::thread::JoinHandle;
+const NUM_MERGE_THREADS: usize = 4;
 
 /// Save the index meta file.
 /// This operation is atomic :
@@ -49,7 +44,7 @@ use std::thread::JoinHandle;
 /// and flushed.
 ///
 /// This method is not part of tantivy's public API
-pub fn save_new_metas(schema: Schema, directory: &mut dyn Directory) -> Result<()> {
+pub fn save_new_metas(schema: Schema, directory: &mut dyn Directory) -> crate::Result<()> {
     save_metas(
         &IndexMeta {
             segments: Vec::new(),
@@ -70,7 +65,7 @@ pub fn save_new_metas(schema: Schema, directory: &mut dyn Directory) -> Result<(
 /// and flushed.
 ///
 /// This method is not part of tantivy's public API
-fn save_metas(metas: &IndexMeta, directory: &mut dyn Directory) -> Result<()> {
+fn save_metas(metas: &IndexMeta, directory: &mut dyn Directory) -> crate::Result<()> {
     info!("save metas");
     let mut buffer = serde_json::to_vec_pretty(metas)?;
     // Just adding a new line at the end of the buffer.
@@ -89,21 +84,38 @@ fn save_metas(metas: &IndexMeta, directory: &mut dyn Directory) -> Result<()> {
 // 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>);
+pub(crate) struct SegmentUpdater(Arc<InnerSegmentUpdater>);
 
-fn perform_merge(
+impl Deref for SegmentUpdater {
-    merge_operation: &MergeOperation,
+    type Target = InnerSegmentUpdater;
+
+    #[inline]
+    fn deref(&self) -> &Self::Target {
+        &self.0
+    }
+}
+
+async fn garbage_collect_files(
+    segment_updater: SegmentUpdater,
+) -> crate::Result<GarbageCollectionResult> {
+    info!("Running garbage collection");
+    let mut index = segment_updater.index.clone();
+    index
+        .directory_mut()
+        .garbage_collect(move || segment_updater.list_files())
+}
+
+/// Merges a list of segments the list of segment givens in the `segment_entries`.
+/// This function happens in the calling thread and is computationally expensive.
+fn merge(
     index: &Index,
     mut segment_entries: Vec<SegmentEntry>,
-) -> Result<SegmentEntry> {
+    target_opstamp: Opstamp,
-    let target_opstamp = merge_operation.target_opstamp();
+) -> crate::Result<SegmentEntry> {
-
     // first we need to apply deletes to our segment.
     let mut merged_segment = index.new_segment();
 
-    // TODO add logging
+    // First we apply all of the delet to the merged segment, up to the target opstamp.
-    let schema = index.schema();
-
     for segment_entry in &mut segment_entries {
         let segment = index.segment(segment_entry.meta().clone());
         advance_deletes(segment, segment_entry, target_opstamp)?;
@@ -117,22 +129,19 @@ fn perform_merge(
         .collect();
 
     // An IndexMerger is like a "view" of our merged segments.
-    let merger: IndexMerger = IndexMerger::open(schema, &segments[..])?;
+    let merger: IndexMerger = IndexMerger::open(index.schema(), &segments[..])?;
-
-    // ... we just serialize this index merger in our new segment
-    // to merge the two segments.
 
+    // ... we just serialize this index merger in our new segment to merge the two segments.
     let segment_serializer = SegmentSerializer::for_segment(&mut merged_segment)?;
 
     let num_docs = merger.write(segment_serializer)?;
 
     let segment_meta = index.new_segment_meta(merged_segment.id(), num_docs);
 
-    let after_merge_segment_entry = SegmentEntry::new(segment_meta.clone(), delete_cursor, None);
+    Ok(SegmentEntry::new(segment_meta.clone(), delete_cursor, None))
-    Ok(after_merge_segment_entry)
 }
 
-struct InnerSegmentUpdater {
+pub(crate) struct InnerSegmentUpdater {
     // we keep a copy of the current active IndexMeta to
     // avoid loading the file everytime we need it in the
     // `SegmentUpdater`.
@@ -140,12 +149,12 @@ struct InnerSegmentUpdater {
     // This should be up to date as all update happen through
     // the unique active `SegmentUpdater`.
     active_metas: RwLock<Arc<IndexMeta>>,
-    pool: CpuPool,
+    pool: ThreadPool,
+    merge_thread_pool: ThreadPool,
+
     index: Index,
     segment_manager: SegmentManager,
     merge_policy: RwLock<Arc<Box<dyn MergePolicy>>>,
-    merging_thread_id: AtomicUsize,
-    merging_threads: RwLock<HashMap<usize, JoinHandle<Result<()>>>>,
     killed: AtomicBool,
     stamper: Stamper,
     merge_operations: MergeOperationInventory,
@@ -156,22 +165,31 @@ impl SegmentUpdater {
         index: Index,
         stamper: Stamper,
         delete_cursor: &DeleteCursor,
-    ) -> Result<SegmentUpdater> {
+    ) -> crate::Result<SegmentUpdater> {
         let segments = index.searchable_segment_metas()?;
         let segment_manager = SegmentManager::from_segments(segments, delete_cursor);
-        let pool = CpuPoolBuilder::new()
+        let pool = ThreadPoolBuilder::new()
             .name_prefix("segment_updater")
             .pool_size(1)
-            .create();
+            .create()
+            .map_err(|_| {
+                crate::Error::SystemError("Failed to spawn segment updater thread".to_string())
+            })?;
+        let merge_thread_pool = ThreadPoolBuilder::new()
+            .name_prefix("merge_thread")
+            .pool_size(NUM_MERGE_THREADS)
+            .create()
+            .map_err(|_| {
+                crate::Error::SystemError("Failed to spawn segment merging thread".to_string())
+            })?;
         let index_meta = index.load_metas()?;
         Ok(SegmentUpdater(Arc::new(InnerSegmentUpdater {
             active_metas: RwLock::new(Arc::new(index_meta)),
             pool,
+            merge_thread_pool,
             index,
             segment_manager,
             merge_policy: RwLock::new(Arc::new(Box::new(DefaultMergePolicy::default()))),
-            merging_thread_id: AtomicUsize::default(),
-            merging_threads: RwLock::new(HashMap::new()),
             killed: AtomicBool::new(false),
             stamper,
             merge_operations: Default::default(),
@@ -179,67 +197,76 @@ impl SegmentUpdater {
     }
 
     pub fn get_merge_policy(&self) -> Arc<Box<dyn MergePolicy>> {
-        self.0.merge_policy.read().unwrap().clone()
+        self.merge_policy.read().unwrap().clone()
     }
 
     pub fn set_merge_policy(&self, merge_policy: Box<dyn MergePolicy>) {
         let arc_merge_policy = Arc::new(merge_policy);
-        *self.0.merge_policy.write().unwrap() = arc_merge_policy;
+        *self.merge_policy.write().unwrap() = arc_merge_policy;
     }
 
-    fn get_merging_thread_id(&self) -> usize {
+    fn schedule_future<T: 'static + Send, F: Future<Output = crate::Result<T>> + 'static + Send>(
-        self.0.merging_thread_id.fetch_add(1, Ordering::SeqCst)
-    }
-
-    fn run_async<T: 'static + Send, F: 'static + Send + FnOnce(SegmentUpdater) -> T>(
         &self,
         f: F,
-    ) -> CpuFuture<T, TantivyError> {
+    ) -> impl Future<Output = crate::Result<T>> {
-        let me_clone = self.clone();
+        let (sender, receiver) = oneshot::channel();
-        self.0.pool.spawn_fn(move || Ok(f(me_clone)))
+        self.pool.spawn_ok(async move {
+            let _ = sender.send(f.await);
+        });
+        receiver.unwrap_or_else(|_| {
+            let err_msg =
+                "A segment_updater future did not success. This should never happen.".to_string();
+            Err(crate::Error::SystemError(err_msg))
+        })
     }
 
-    pub fn add_segment(&self, segment_entry: SegmentEntry) -> bool {
+    pub fn schedule_add_segment(
-        self.run_async(|segment_updater| {
+        &self,
-            segment_updater.0.segment_manager.add_segment(segment_entry);
+        segment_entry: SegmentEntry,
-            segment_updater.consider_merge_options();
+    ) -> impl Future<Output = crate::Result<()>> {
-            true
+        let segment_updater = self.clone();
+        self.schedule_future(async move {
+            segment_updater.segment_manager.add_segment(segment_entry);
+            segment_updater.consider_merge_options().await;
+            Ok(())
         })
-        .forget();
-        true
     }
 
     /// Orders `SegmentManager` to remove all segments
     pub(crate) fn remove_all_segments(&self) {
-        self.0.segment_manager.remove_all_segments();
+        self.segment_manager.remove_all_segments();
     }
 
     pub fn kill(&mut self) {
-        self.0.killed.store(true, Ordering::Release);
+        self.killed.store(true, Ordering::Release);
     }
 
     pub fn is_alive(&self) -> bool {
-        !self.0.killed.load(Ordering::Acquire)
+        !self.killed.load(Ordering::Acquire)
     }
 
     /// Apply deletes up to the target opstamp to all segments.
     ///
     /// The method returns copies of the segment entries,
     /// updated with the delete information.
-    fn purge_deletes(&self, target_opstamp: Opstamp) -> Result<Vec<SegmentEntry>> {
+    fn purge_deletes(&self, target_opstamp: Opstamp) -> crate::Result<Vec<SegmentEntry>> {
-        let mut segment_entries = self.0.segment_manager.segment_entries();
+        let mut segment_entries = self.segment_manager.segment_entries();
         for segment_entry in &mut segment_entries {
-            let segment = self.0.index.segment(segment_entry.meta().clone());
+            let segment = self.index.segment(segment_entry.meta().clone());
             advance_deletes(segment, segment_entry, target_opstamp)?;
         }
         Ok(segment_entries)
     }
 
-    pub fn save_metas(&self, opstamp: Opstamp, commit_message: Option<String>) {
+    pub fn save_metas(
+        &self,
+        opstamp: Opstamp,
+        commit_message: Option<String>,
+    ) -> crate::Result<()> {
         if self.is_alive() {
-            let index = &self.0.index;
+            let index = &self.index;
             let directory = index.directory();
-            let mut commited_segment_metas = self.0.segment_manager.committed_segment_metas();
+            let mut commited_segment_metas = self.segment_manager.committed_segment_metas();
 
             // We sort segment_readers by number of documents.
             // This is an heuristic to make multithreading more efficient.
@@ -261,16 +288,18 @@ impl SegmentUpdater {
                 opstamp,
                 payload: commit_message,
             };
-            save_metas(&index_meta, directory.box_clone().borrow_mut())
+            // TODO add context to the error.
-                .expect("Could not save metas.");
+            save_metas(&index_meta, directory.box_clone().borrow_mut())?;
             self.store_meta(&index_meta);
         }
+        Ok(())
     }
 
-    pub fn garbage_collect_files(&self) -> CpuFuture<(), TantivyError> {
+    pub fn schedule_garbage_collect(
-        self.run_async(move |segment_updater| {
+        &self,
-            segment_updater.garbage_collect_files_exec();
+    ) -> impl Future<Output = crate::Result<GarbageCollectionResult>> {
-        })
+        let garbage_collect_future = garbage_collect_files(self.clone());
+        self.schedule_future(garbage_collect_future)
     }
 
     /// List the files that are useful to the index.
@@ -278,148 +307,132 @@ impl SegmentUpdater {
     /// This does not include lock files, or files that are obsolete
     /// but have not yet been deleted by the garbage collector.
     fn list_files(&self) -> HashSet<PathBuf> {
-        let mut files = HashSet::new();
+        let mut files: HashSet<PathBuf> = self
+            .index
+            .list_all_segment_metas()
+            .into_iter()
+            .flat_map(|segment_meta| segment_meta.list_files())
+            .collect();
         files.insert(META_FILEPATH.to_path_buf());
-        for segment_meta in self.0.index.list_all_segment_metas() {
-            files.extend(segment_meta.list_files());
-        }
         files
     }
 
-    fn garbage_collect_files_exec(&self) {
+    pub fn schedule_commit(
-        info!("Running garbage collection");
+        &self,
-        let mut index = self.0.index.clone();
+        opstamp: Opstamp,
-        index.directory_mut().garbage_collect(|| self.list_files());
+        payload: Option<String>,
-    }
+    ) -> impl Future<Output = crate::Result<()>> {
-
+        let segment_updater: SegmentUpdater = self.clone();
-    pub fn commit(&self, opstamp: Opstamp, payload: Option<String>) -> Result<()> {
+        self.schedule_future(async move {
-        self.run_async(move |segment_updater| {
             if segment_updater.is_alive() {
-                let segment_entries = segment_updater
+                let segment_entries = segment_updater.purge_deletes(opstamp)?;
-                    .purge_deletes(opstamp)
+                segment_updater.segment_manager.commit(segment_entries);
-                    .expect("Failed purge deletes");
+                segment_updater.save_metas(opstamp, payload)?;
-                segment_updater.0.segment_manager.commit(segment_entries);
+                let _ = garbage_collect_files(segment_updater.clone()).await;
-                segment_updater.save_metas(opstamp, payload);
+                segment_updater.consider_merge_options().await;
-                segment_updater.garbage_collect_files_exec();
-                segment_updater.consider_merge_options();
             }
+            Ok(())
         })
-        .wait()
-    }
-
-    pub fn start_merge(&self, segment_ids: &[SegmentId]) -> Result<Receiver<SegmentMeta>> {
-        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());
+        *self.active_metas.write().unwrap() = Arc::new(index_meta.clone());
-    }
-    fn load_metas(&self) -> Arc<IndexMeta> {
-        self.0.active_metas.read().unwrap().clone()
     }
 
+    fn load_metas(&self) -> Arc<IndexMeta> {
+        self.active_metas.read().unwrap().clone()
+    }
+
+    pub(crate) fn make_merge_operation(&self, segment_ids: &[SegmentId]) -> MergeOperation {
+        let commit_opstamp = self.load_metas().opstamp;
+        MergeOperation::new(&self.merge_operations, commit_opstamp, segment_ids.to_vec())
+    }
+
+    // Starts a merge operation. This function will block until the merge operation is effectively
+    // started. Note that it does not wait for the merge to terminate.
+    // The calling thread should not be block for a long time, as this only involve waiting for the
+    // `SegmentUpdater` queue which in turns only contains lightweight operations.
+    //
+    // The merge itself happens on a different thread.
+    //
+    // When successful, this function returns a `Future` for a `Result<SegmentMeta>` that represents
+    // the actual outcome of the merge operation.
+    //
+    // It returns an error if for some reason the merge operation could not be started.
+    //
+    // At this point an error is not necessarily the sign of a malfunction.
+    // (e.g. A rollback could have happened, between the instant when the merge operaiton was
+    // suggested and the moment when it ended up being executed.)
+    //
     // `segment_ids` is required to be non-empty.
-    fn start_merge_impl(&self, merge_operation: MergeOperation) -> Result<Receiver<SegmentMeta>> {
+    pub fn start_merge(
+        &self,
+        merge_operation: MergeOperation,
+    ) -> crate::Result<impl Future<Output = crate::Result<SegmentMeta>>> {
         assert!(
             !merge_operation.segment_ids().is_empty(),
             "Segment_ids cannot be empty."
         );
 
-        let segment_updater_clone = self.clone();
+        let segment_updater = self.clone();
         let segment_entries: Vec<SegmentEntry> = self
-            .0
             .segment_manager
             .start_merge(merge_operation.segment_ids())?;
 
-        //        let segment_ids_vec = merge_operation.segment_ids.to_vec();
+        info!("Starting merge  - {:?}", merge_operation.segment_ids());
 
-        let merging_thread_id = self.get_merging_thread_id();
+        let (merging_future_send, merging_future_recv) =
-        info!(
+            oneshot::channel::<crate::Result<SegmentMeta>>();
-            "Starting merge thread #{} - {:?}",
-            merging_thread_id,
-            merge_operation.segment_ids()
-        );
-        let (merging_future_send, merging_future_recv) = oneshot();
 
-        // first we need to apply deletes to our segment.
+        self.merge_thread_pool.spawn_ok(async move {
-        let merging_join_handle = thread::Builder::new()
+            // The fact that `merge_operation` is moved here is important.
-            .name(format!("mergingthread-{}", merging_thread_id))
+            // Its lifetime is used to track how many merging thread are currently running,
-            .spawn(move || {
+            // as well as which segment is currently in merge and therefore should not be
-                // first we need to apply deletes to our segment.
+            // candidate for another merge.
-                let merge_result = perform_merge(
+            match merge(
-                    &merge_operation,
+                &segment_updater.index,
-                    &segment_updater_clone.0.index,
+                segment_entries,
-                    segment_entries,
+                merge_operation.target_opstamp(),
-                );
+            ) {
-
+                Ok(after_merge_segment_entry) => {
-                match merge_result {
+                    let segment_meta = segment_updater
-                    Ok(after_merge_segment_entry) => {
+                        .end_merge(merge_operation, after_merge_segment_entry)
-                        let merged_segment_meta = after_merge_segment_entry.meta().clone();
+                        .await;
-                        segment_updater_clone
+                    let _send_result = merging_future_send.send(segment_meta);
-                            .end_merge(merge_operation, after_merge_segment_entry)
+                }
-                            .expect("Segment updater thread is corrupted.");
+                Err(e) => {
-
+                    warn!(
-                        // the future may fail if the listener of the oneshot future
+                        "Merge of {:?} was cancelled: {:?}",
-                        // has been destroyed.
+                        merge_operation.segment_ids().to_vec(),
-                        //
+                        e
-                        // This is not a problem here, so we just ignore any
+                    );
-                        // possible error.
+                    // ... cancel merge
-                        let _merging_future_res = merging_future_send.send(merged_segment_meta);
+                    if cfg!(test) {
-                    }
+                        panic!("Merge failed.");
-                    Err(e) => {
-                        warn!(
-                            "Merge of {:?} was cancelled: {:?}",
-                            merge_operation.segment_ids(),
-                            e
-                        );
-                        // ... cancel merge
-                        if cfg!(test) {
-                            panic!("Merge failed.");
-                        }
-                        // 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
+            }
-                    .0
+        });
-                    .merging_threads
+
-                    .write()
+        Ok(merging_future_recv
-                    .unwrap()
+            .unwrap_or_else(|_| Err(crate::Error::SystemError("Merge failed".to_string()))))
-                    .remove(&merging_thread_id);
-                Ok(())
-            })
-            .expect("Failed to spawn a thread.");
-        self.0
-            .merging_threads
-            .write()
-            .unwrap()
-            .insert(merging_thread_id, merging_join_handle);
-        Ok(merging_future_recv)
     }
 
-    fn consider_merge_options(&self) {
+    async fn consider_merge_options(&self) {
-        let merge_segment_ids: HashSet<SegmentId> = self.0.merge_operations.segment_in_merge();
+        let merge_segment_ids: HashSet<SegmentId> = self.merge_operations.segment_in_merge();
         let (committed_segments, uncommitted_segments) =
-            get_mergeable_segments(&merge_segment_ids, &self.0.segment_manager);
+            get_mergeable_segments(&merge_segment_ids, &self.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 current_opstamp = self.0.stamper.stamp();
+        let current_opstamp = self.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)
+                MergeOperation::new(&self.merge_operations, current_opstamp, merge_candidate.0)
             })
             .collect();
 
@@ -427,25 +440,18 @@ impl SegmentUpdater {
         let committed_merge_candidates = merge_policy
             .compute_merge_candidates(&committed_segments)
             .into_iter()
-            .map(|merge_candidate| {
+            .map(|merge_candidate: MergeCandidate| {
-                MergeOperation::new(&self.0.merge_operations, commit_opstamp, merge_candidate.0)
+                MergeOperation::new(&self.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) {
+            if let Err(err) = self.start_merge(merge_operation) {
-                Ok(merge_future) => {
+                warn!(
-                    if let Err(e) = merge_future.fuse().poll() {
+                    "Starting the merge failed for the following reason. This is not fatal. {}",
-                        error!("The merge task failed quickly after starting: {:?}", e);
+                    err
-                    }
+                );
-                }
-                Err(err) => {
-                    warn!(
-                        "Starting the merge failed for the following reason. This is not fatal. {}",
-                        err
-                    );
-                }
             }
         }
     }
@@ -454,15 +460,17 @@ impl SegmentUpdater {
         &self,
         merge_operation: MergeOperation,
         mut after_merge_segment_entry: SegmentEntry,
-    ) -> Result<()> {
+    ) -> impl Future<Output = crate::Result<SegmentMeta>> {
-        self.run_async(move |segment_updater| {
+        let segment_updater = self.clone();
+        let after_merge_segment_meta = after_merge_segment_entry.meta().clone();
+        let end_merge_future = self.schedule_future(async move {
             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.load_metas().opstamp;
                     if delete_operation.opstamp < committed_opstamp {
-                        let index = &segment_updater.0.index;
+                        let index = &segment_updater.index;
                         let segment = index.segment(after_merge_segment_entry.meta().clone());
                         if let Err(e) = advance_deletes(
                             segment,
@@ -480,21 +488,26 @@ impl SegmentUpdater {
                             // ... cancel merge
                             // `merge_operations` are tracked. As it is dropped, the
                             // the segment_ids will be available again for merge.
-                            return;
+                            return Err(e);
                         }
                     }
                 }
                 let previous_metas = segment_updater.load_metas();
-                segment_updater
+                let segments_status = segment_updater
-                    .0
                     .segment_manager
-                    .end_merge(merge_operation.segment_ids(), after_merge_segment_entry);
+                    .end_merge(merge_operation.segment_ids(), after_merge_segment_entry)?;
-                segment_updater.consider_merge_options();
+
-                segment_updater.save_metas(previous_metas.opstamp, previous_metas.payload.clone());
+                if segments_status == SegmentsStatus::Committed {
+                    segment_updater
+                        .save_metas(previous_metas.opstamp, previous_metas.payload.clone())?;
+                }
+
+                segment_updater.consider_merge_options().await;
             } // we drop all possible handle to a now useless `SegmentMeta`.
-            segment_updater.garbage_collect_files_exec();
+            let _ = garbage_collect_files(segment_updater).await;
-        })
+            Ok(())
-        .wait()
+        });
+        end_merge_future.map_ok(|_| after_merge_segment_meta)
     }
 
     /// Wait for current merging threads.
@@ -512,26 +525,13 @@ impl SegmentUpdater {
     ///
     /// Obsolete files will eventually be cleaned up
     /// by the directory garbage collector.
-    pub fn wait_merging_thread(&self) -> Result<()> {
+    pub fn wait_merging_thread(&self) -> crate::Result<()> {
-        loop {
+        for merge_operation in self.merge_operations.changes_iter() {
-            let merging_threads: HashMap<usize, JoinHandle<Result<()>>> = {
+            if merge_operation.is_empty() {
-                let mut merging_threads = self.0.merging_threads.write().unwrap();
+                break;
-                mem::replace(merging_threads.deref_mut(), HashMap::new())
-            };
-            if merging_threads.is_empty() {
-                return Ok(());
             }
-            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 merged segment.
-            // Let's wait for that too.
-            self.run_async(move |_| {}).wait()?;
         }
+        Ok(())
     }
 }
 
@@ -687,7 +687,6 @@ mod tests {
         index_writer.segment_updater().remove_all_segments();
         let seg_vec = index_writer
             .segment_updater()
-            .0
             .segment_manager
             .segment_entries();
         assert!(seg_vec.is_empty());

src/indexer/segment_writer.rs

@@ -13,7 +13,8 @@ use crate::schema::Value;
 use crate::schema::{Field, FieldEntry};
 use crate::tokenizer::BoxedTokenizer;
 use crate::tokenizer::FacetTokenizer;
-use crate::tokenizer::{TokenStream, Tokenizer};
+use crate::tokenizer::PreTokenizedStream;
+use crate::tokenizer::{TokenStream, TokenStreamChain, Tokenizer};
 use crate::DocId;
 use crate::Opstamp;
 use crate::Result;
@@ -158,26 +159,44 @@ impl SegmentWriter {
                     }
                 }
                 FieldType::Str(_) => {
-                    let num_tokens = if let Some(ref mut tokenizer) =
+                    let mut token_streams: Vec<Box<dyn TokenStream>> = vec![];
-                        self.tokenizers[field.field_id() as usize]
+                    let mut offsets = vec![];
-                    {
+                    let mut total_offset = 0;
-                        let texts: Vec<&str> = field_values
+
-                            .iter()
+                    for field_value in field_values {
-                            .flat_map(|field_value| match *field_value.value() {
+                        match field_value.value() {
-                                Value::Str(ref text) => Some(text.as_str()),
+                            Value::PreTokStr(tok_str) => {
-                                _ => None,
+                                offsets.push(total_offset);
-                            })
+                                if let Some(last_token) = tok_str.tokens.last() {
-                            .collect();
+                                    total_offset += last_token.offset_to;
-                        if texts.is_empty() {
+                                }
-                            0
+
-                        } else {
+                                token_streams
-                            let mut token_stream = tokenizer.token_stream_texts(&texts[..]);
+                                    .push(Box::new(PreTokenizedStream::from(tok_str.clone())));
-                            self.multifield_postings
+                            }
-                                .index_text(doc_id, field, &mut token_stream)
+                            Value::Str(ref text) => {
+                                if let Some(ref mut tokenizer) =
+                                    self.tokenizers[field.field_id() as usize]
+                                {
+                                    offsets.push(total_offset);
+                                    total_offset += text.len();
+
+                                    token_streams.push(tokenizer.token_stream(text));
+                                }
+                            }
+                            _ => (),
                         }
-                    } else {
+                    }
+
+                    let num_tokens = if token_streams.is_empty() {
                         0
+                    } else {
+                        let mut token_stream: Box<dyn TokenStream> =
+                            Box::new(TokenStreamChain::new(offsets, token_streams));
+                        self.multifield_postings
+                            .index_text(doc_id, field, &mut token_stream)
                     };
+
                     self.fieldnorms_writer.record(doc_id, field, num_tokens);
                 }
                 FieldType::U64(ref int_option) => {

src/reader/mod.rs

@@ -162,6 +162,11 @@ pub struct IndexReader {
 }
 
 impl IndexReader {
+    #[cfg(test)]
+    pub(crate) fn index(&self) -> Index {
+        self.inner.index.clone()
+    }
+
     /// Update searchers so that they reflect the state of the last
     /// `.commit()`.
     ///

src/reader/pool.rs

@@ -167,7 +167,7 @@ mod tests {
 
     use super::Pool;
     use super::Queue;
-    use std::iter;
+    use std::{iter, mem};
 
     #[test]
     fn test_pool() {
@@ -197,33 +197,67 @@ mod tests {
     fn test_pool_dont_panic_on_empty_pop() {
         // When the object pool is exhausted, it shouldn't panic on pop()
         use std::sync::Arc;
-        use std::{thread, time};
+        use std::thread;
 
         // Wrap the pool in an Arc, same way as its used in `core/index.rs`
-        let pool = Arc::new(Pool::new());
+        let pool1 = Arc::new(Pool::new());
         // clone pools outside the move scope of each new thread
-        let pool1 = Arc::clone(&pool);
+        let pool2 = Arc::clone(&pool1);
-        let pool2 = Arc::clone(&pool);
+        let pool3 = Arc::clone(&pool1);
+
         let elements_for_pool = vec![1, 2];
-        pool.publish_new_generation(elements_for_pool);
+        pool1.publish_new_generation(elements_for_pool);
 
         let mut threads = vec![];
-        let sleep_dur = time::Duration::from_millis(10);
         // spawn one more thread than there are elements in the pool
+
+        let (start_1_send, start_1_recv) = crossbeam::bounded(0);
+        let (start_2_send, start_2_recv) = crossbeam::bounded(0);
+        let (start_3_send, start_3_recv) = crossbeam::bounded(0);
+
+        let (event_send1, event_recv) = crossbeam::unbounded();
+        let event_send2 = event_send1.clone();
+        let event_send3 = event_send1.clone();
+
         threads.push(thread::spawn(move || {
-            // leasing to make sure it's not dropped before sleep is called
+            assert_eq!(start_1_recv.recv(), Ok("start"));
-            let _leased_searcher = &pool.acquire();
-            thread::sleep(sleep_dur);
-        }));
-        threads.push(thread::spawn(move || {
-            // leasing to make sure it's not dropped before sleep is called
             let _leased_searcher = &pool1.acquire();
-            thread::sleep(sleep_dur);
+            assert!(event_send1.send("1 acquired").is_ok());
+            assert_eq!(start_1_recv.recv(), Ok("stop"));
+            assert!(event_send1.send("1 stopped").is_ok());
+            mem::drop(_leased_searcher);
         }));
+
         threads.push(thread::spawn(move || {
-            // leasing to make sure it's not dropped before sleep is called
+            assert_eq!(start_2_recv.recv(), Ok("start"));
             let _leased_searcher = &pool2.acquire();
-            thread::sleep(sleep_dur);
+            assert!(event_send2.send("2 acquired").is_ok());
+            assert_eq!(start_2_recv.recv(), Ok("stop"));
+            mem::drop(_leased_searcher);
+            assert!(event_send2.send("2 stopped").is_ok());
         }));
+
+        threads.push(thread::spawn(move || {
+            assert_eq!(start_3_recv.recv(), Ok("start"));
+            let _leased_searcher = &pool3.acquire();
+            assert!(event_send3.send("3 acquired").is_ok());
+            assert_eq!(start_3_recv.recv(), Ok("stop"));
+            mem::drop(_leased_searcher);
+            assert!(event_send3.send("3 stopped").is_ok());
+        }));
+
+        assert!(start_1_send.send("start").is_ok());
+        assert_eq!(event_recv.recv(), Ok("1 acquired"));
+        assert!(start_2_send.send("start").is_ok());
+        assert_eq!(event_recv.recv(), Ok("2 acquired"));
+        assert!(start_3_send.send("start").is_ok());
+        assert!(event_recv.try_recv().is_err());
+        assert!(start_1_send.send("stop").is_ok());
+        assert_eq!(event_recv.recv(), Ok("1 stopped"));
+        assert_eq!(event_recv.recv(), Ok("3 acquired"));
+        assert!(start_3_send.send("stop").is_ok());
+        assert_eq!(event_recv.recv(), Ok("3 stopped"));
+        assert!(start_2_send.send("stop").is_ok());
+        assert_eq!(event_recv.recv(), Ok("2 stopped"));
     }
 }

src/schema/document.rs

@@ -1,6 +1,7 @@
 use super::*;
 use crate::common::BinarySerializable;
 use crate::common::VInt;
+use crate::tokenizer::PreTokenizedString;
 use crate::DateTime;
 use itertools::Itertools;
 use std::io::{self, Read, Write};
@@ -29,8 +30,8 @@ impl From<Vec<FieldValue>> for Document {
 impl PartialEq for Document {
     fn eq(&self, other: &Document) -> bool {
         // super slow, but only here for tests
-        let mut self_field_values = self.field_values.clone();
+        let mut self_field_values: Vec<&_> = self.field_values.iter().collect();
-        let mut other_field_values = other.field_values.clone();
+        let mut other_field_values: Vec<&_> = other.field_values.iter().collect();
         self_field_values.sort();
         other_field_values.sort();
         self_field_values.eq(&other_field_values)
@@ -78,6 +79,16 @@ impl Document {
         self.add(FieldValue::new(field, value));
     }
 
+    /// Add a pre-tokenized text field.
+    pub fn add_pre_tokenized_text(
+        &mut self,
+        field: Field,
+        pre_tokenized_text: &PreTokenizedString,
+    ) {
+        let value = Value::PreTokStr(pre_tokenized_text.clone());
+        self.add(FieldValue::new(field, value));
+    }
+
     /// Add a u64 field
     pub fn add_u64(&mut self, field: Field, value: u64) {
         self.add(FieldValue::new(field, Value::U64(value)));

src/schema/field.rs

@@ -15,6 +15,7 @@ impl Field {
     }
 
     /// Returns a u32 identifying uniquely a field within a schema.
+    #[allow(clippy::trivially_copy_pass_by_ref)]
     pub fn field_id(&self) -> u32 {
         self.0
     }

src/schema/field_type.rs

@@ -1,11 +1,11 @@
 use base64::decode;
 
-use crate::schema::{IntOptions, TextOptions};
-
 use crate::schema::Facet;
 use crate::schema::IndexRecordOption;
 use crate::schema::TextFieldIndexing;
 use crate::schema::Value;
+use crate::schema::{IntOptions, TextOptions};
+use crate::tokenizer::PreTokenizedString;
 use serde_json::Value as JsonValue;
 
 /// Possible error that may occur while parsing a field value
@@ -169,6 +169,28 @@ impl FieldType {
                     Err(ValueParsingError::TypeError(msg))
                 }
             },
+            JsonValue::Object(_) => match *self {
+                FieldType::Str(_) => {
+                    if let Ok(tok_str_val) =
+                        serde_json::from_value::<PreTokenizedString>(json.clone())
+                    {
+                        Ok(Value::PreTokStr(tok_str_val))
+                    } else {
+                        let msg = format!(
+                            "Json value {:?} cannot be translated to PreTokenizedString.",
+                            json
+                        );
+                        Err(ValueParsingError::TypeError(msg))
+                    }
+                }
+                _ => {
+                    let msg = format!(
+                        "Json value not supported error {:?}. Expected {:?}",
+                        json, self
+                    );
+                    Err(ValueParsingError::TypeError(msg))
+                }
+            },
             _ => {
                 let msg = format!(
                     "Json value not supported error {:?}. Expected {:?}",
@@ -184,7 +206,9 @@ impl FieldType {
 mod tests {
     use super::FieldType;
     use crate::schema::field_type::ValueParsingError;
+    use crate::schema::TextOptions;
     use crate::schema::Value;
+    use crate::tokenizer::{PreTokenizedString, Token};
 
     #[test]
     fn test_bytes_value_from_json() {
@@ -205,4 +229,71 @@ mod tests {
             _ => panic!("Expected parse failure for invalid base64"),
         }
     }
+
+    #[test]
+    fn test_pre_tok_str_value_from_json() {
+        let pre_tokenized_string_json = r#"{
+  "text": "The Old Man",
+  "tokens": [
+    {
+      "offset_from": 0,
+      "offset_to": 3,
+      "position": 0,
+      "text": "The",
+      "position_length": 1
+    },
+    {
+      "offset_from": 4,
+      "offset_to": 7,
+      "position": 1,
+      "text": "Old",
+      "position_length": 1
+    },
+    {
+      "offset_from": 8,
+      "offset_to": 11,
+      "position": 2,
+      "text": "Man",
+      "position_length": 1
+    }
+  ]
+}"#;
+
+        let expected_value = Value::PreTokStr(PreTokenizedString {
+            text: String::from("The Old Man"),
+            tokens: vec![
+                Token {
+                    offset_from: 0,
+                    offset_to: 3,
+                    position: 0,
+                    text: String::from("The"),
+                    position_length: 1,
+                },
+                Token {
+                    offset_from: 4,
+                    offset_to: 7,
+                    position: 1,
+                    text: String::from("Old"),
+                    position_length: 1,
+                },
+                Token {
+                    offset_from: 8,
+                    offset_to: 11,
+                    position: 2,
+                    text: String::from("Man"),
+                    position_length: 1,
+                },
+            ],
+        });
+
+        let deserialized_value = FieldType::Str(TextOptions::default())
+            .value_from_json(&serde_json::from_str(pre_tokenized_string_json).unwrap())
+            .unwrap();
+
+        assert_eq!(deserialized_value, expected_value);
+
+        let serialized_value_json = serde_json::to_string_pretty(&expected_value).unwrap();
+
+        assert_eq!(serialized_value_json, pre_tokenized_string_json);
+    }
 }

src/schema/value.rs

@@ -1,4 +1,5 @@
 use crate::schema::Facet;
+use crate::tokenizer::PreTokenizedString;
 use crate::DateTime;
 use serde::de::Visitor;
 use serde::{Deserialize, Deserializer, Serialize, Serializer};
@@ -10,6 +11,8 @@ use std::{cmp::Ordering, fmt};
 pub enum Value {
     /// The str type is used for any text information.
     Str(String),
+    /// Pre-tokenized str type,
+    PreTokStr(PreTokenizedString),
     /// Unsigned 64-bits Integer `u64`
     U64(u64),
     /// Signed 64-bits Integer `i64`
@@ -29,6 +32,7 @@ impl Ord for Value {
     fn cmp(&self, other: &Self) -> Ordering {
         match (self, other) {
             (Value::Str(l), Value::Str(r)) => l.cmp(r),
+            (Value::PreTokStr(l), Value::PreTokStr(r)) => l.cmp(r),
             (Value::U64(l), Value::U64(r)) => l.cmp(r),
             (Value::I64(l), Value::I64(r)) => l.cmp(r),
             (Value::Date(l), Value::Date(r)) => l.cmp(r),
@@ -44,6 +48,8 @@ impl Ord for Value {
             }
             (Value::Str(_), _) => Ordering::Less,
             (_, Value::Str(_)) => Ordering::Greater,
+            (Value::PreTokStr(_), _) => Ordering::Less,
+            (_, Value::PreTokStr(_)) => Ordering::Greater,
             (Value::U64(_), _) => Ordering::Less,
             (_, Value::U64(_)) => Ordering::Greater,
             (Value::I64(_), _) => Ordering::Less,
@@ -65,6 +71,7 @@ impl Serialize for Value {
     {
         match *self {
             Value::Str(ref v) => serializer.serialize_str(v),
+            Value::PreTokStr(ref v) => v.serialize(serializer),
             Value::U64(u) => serializer.serialize_u64(u),
             Value::I64(u) => serializer.serialize_i64(u),
             Value::F64(u) => serializer.serialize_f64(u),
@@ -124,6 +131,15 @@ impl Value {
         }
     }
 
+    /// Returns the tokenized text, provided the value is of the `PreTokStr` type.
+    /// (Returns None if the value is not of the `PreTokStr` type).
+    pub fn tokenized_text(&self) -> Option<&PreTokenizedString> {
+        match *self {
+            Value::PreTokStr(ref tok_text) => Some(tok_text),
+            _ => None,
+        }
+    }
+
     /// Returns the u64-value, provided the value is of the `U64` type.
     ///
     /// # Panics
@@ -217,10 +233,17 @@ impl From<Vec<u8>> for Value {
     }
 }
 
+impl From<PreTokenizedString> for Value {
+    fn from(pretokenized_string: PreTokenizedString) -> Value {
+        Value::PreTokStr(pretokenized_string)
+    }
+}
+
 mod binary_serialize {
     use super::Value;
     use crate::common::{f64_to_u64, u64_to_f64, BinarySerializable};
     use crate::schema::Facet;
+    use crate::tokenizer::PreTokenizedString;
     use chrono::{TimeZone, Utc};
     use std::io::{self, Read, Write};
 
@@ -231,6 +254,11 @@ mod binary_serialize {
     const BYTES_CODE: u8 = 4;
     const DATE_CODE: u8 = 5;
     const F64_CODE: u8 = 6;
+    const EXT_CODE: u8 = 7;
+
+    // extended types
+
+    const TOK_STR_CODE: u8 = 0;
 
     impl BinarySerializable for Value {
         fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
@@ -239,6 +267,18 @@ mod binary_serialize {
                     TEXT_CODE.serialize(writer)?;
                     text.serialize(writer)
                 }
+                Value::PreTokStr(ref tok_str) => {
+                    EXT_CODE.serialize(writer)?;
+                    TOK_STR_CODE.serialize(writer)?;
+                    if let Ok(text) = serde_json::to_string(tok_str) {
+                        text.serialize(writer)
+                    } else {
+                        Err(io::Error::new(
+                            io::ErrorKind::Other,
+                            "Failed to dump Value::PreTokStr(_) to json.",
+                        ))
+                    }
+                }
                 Value::U64(ref val) => {
                     U64_CODE.serialize(writer)?;
                     val.serialize(writer)
@@ -290,6 +330,30 @@ mod binary_serialize {
                 }
                 HIERARCHICAL_FACET_CODE => Ok(Value::Facet(Facet::deserialize(reader)?)),
                 BYTES_CODE => Ok(Value::Bytes(Vec::<u8>::deserialize(reader)?)),
+                EXT_CODE => {
+                    let ext_type_code = u8::deserialize(reader)?;
+                    match ext_type_code {
+                        TOK_STR_CODE => {
+                            let str_val = String::deserialize(reader)?;
+                            if let Ok(value) = serde_json::from_str::<PreTokenizedString>(&str_val)
+                            {
+                                Ok(Value::PreTokStr(value))
+                            } else {
+                                Err(io::Error::new(
+                                    io::ErrorKind::Other,
+                                    "Failed to parse string data as Value::PreTokStr(_).",
+                                ))
+                            }
+                        }
+                        _ => Err(io::Error::new(
+                            io::ErrorKind::InvalidData,
+                            format!(
+                                "No extened field type is associated with code {:?}",
+                                ext_type_code
+                            ),
+                        )),
+                    }
+                }
                 _ => Err(io::Error::new(
                     io::ErrorKind::InvalidData,
                     format!("No field type is associated with code {:?}", type_code),

src/snippet/mod.rs

@@ -331,9 +331,8 @@ mod tests {
     use std::collections::BTreeMap;
     use std::iter::Iterator;
 
-    const TEST_TEXT: &'static str =
+    const TEST_TEXT: &'static str = r#"Rust is a systems programming language sponsored by
-        r#"Rust is a systems programming language sponsored by Mozilla which
+Mozilla which describes it as a "safe, concurrent, practical language", supporting functional and
-describes it as a "safe, concurrent, practical language", supporting functional and
 imperative-procedural paradigms. Rust is syntactically similar to C++[according to whom?],
 but its designers intend it to provide better memory safety while still maintaining
 performance.
@@ -363,13 +362,13 @@ Survey in 2016, 2017, and 2018."#;
         let snippet = select_best_fragment_combination(&fragments[..], &TEST_TEXT);
         assert_eq!(
             snippet.fragments,
-            "Rust is a systems programming language sponsored by \
+            "Rust is a systems programming language sponsored by\n\
-             Mozilla which\ndescribes it as a \"safe"
+             Mozilla which describes it as a \"safe"
         );
         assert_eq!(
             snippet.to_html(),
             "<b>Rust</b> is a systems programming <b>language</b> \
-             sponsored by Mozilla which\ndescribes it as a &quot;safe"
+             sponsored by\nMozilla which describes it as a &quot;safe"
         )
     }
 

src/tokenizer/mod.rs

@@ -136,6 +136,7 @@ mod simple_tokenizer;
 mod stemmer;
 mod stop_word_filter;
 mod token_stream_chain;
+mod tokenized_string;
 mod tokenizer;
 mod tokenizer_manager;
 
@@ -152,7 +153,9 @@ pub use self::stop_word_filter::StopWordFilter;
 pub(crate) use self::token_stream_chain::TokenStreamChain;
 pub use self::tokenizer::BoxedTokenizer;
 
+pub use self::tokenized_string::{PreTokenizedStream, PreTokenizedString};
 pub use self::tokenizer::{Token, TokenFilter, TokenStream, Tokenizer};
+
 pub use self::tokenizer_manager::TokenizerManager;
 
 /// Maximum authorized len (in bytes) for a token.

src/tokenizer/tokenized_string.rs

@@ -0,0 +1,189 @@
+use crate::tokenizer::{Token, TokenStream, TokenStreamChain};
+use std::cmp::Ordering;
+
+/// Struct representing pre-tokenized text
+#[derive(Debug, Clone, Serialize, Deserialize, Eq, PartialEq)]
+pub struct PreTokenizedString {
+    /// Original text
+    pub text: String,
+    /// Tokens derived from the text
+    pub tokens: Vec<Token>,
+}
+
+impl Ord for PreTokenizedString {
+    fn cmp(&self, other: &Self) -> Ordering {
+        self.text.cmp(&other.text)
+    }
+}
+
+impl PartialOrd for PreTokenizedString {
+    fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
+        Some(self.cmp(other))
+    }
+}
+
+/// TokenStream implementation which wraps PreTokenizedString
+pub struct PreTokenizedStream {
+    tokenized_string: PreTokenizedString,
+    current_token: i64,
+}
+
+impl From<PreTokenizedString> for PreTokenizedStream {
+    fn from(s: PreTokenizedString) -> PreTokenizedStream {
+        PreTokenizedStream {
+            tokenized_string: s,
+            current_token: -1,
+        }
+    }
+}
+
+impl PreTokenizedStream {
+    /// Creates a TokenStream from PreTokenizedString array
+    pub fn chain_tokenized_strings<'a>(
+        tok_strings: &'a [&'a PreTokenizedString],
+    ) -> Box<dyn TokenStream + 'a> {
+        if tok_strings.len() == 1 {
+            Box::new(PreTokenizedStream::from((*tok_strings[0]).clone()))
+        } else {
+            let mut offsets = vec![];
+            let mut total_offset = 0;
+            for &tok_string in tok_strings {
+                offsets.push(total_offset);
+                if let Some(last_token) = tok_string.tokens.last() {
+                    total_offset += last_token.offset_to;
+                }
+            }
+            let token_streams: Vec<_> = tok_strings
+                .iter()
+                .map(|tok_string| PreTokenizedStream::from((*tok_string).clone()))
+                .collect();
+            Box::new(TokenStreamChain::new(offsets, token_streams))
+        }
+    }
+}
+
+impl TokenStream for PreTokenizedStream {
+    fn advance(&mut self) -> bool {
+        self.current_token += 1;
+        self.current_token < self.tokenized_string.tokens.len() as i64
+    }
+
+    fn token(&self) -> &Token {
+        assert!(
+            self.current_token >= 0,
+            "TokenStream not initialized. You should call advance() at least once."
+        );
+        &self.tokenized_string.tokens[self.current_token as usize]
+    }
+
+    fn token_mut(&mut self) -> &mut Token {
+        assert!(
+            self.current_token >= 0,
+            "TokenStream not initialized. You should call advance() at least once."
+        );
+        &mut self.tokenized_string.tokens[self.current_token as usize]
+    }
+}
+
+#[cfg(test)]
+mod tests {
+
+    use super::*;
+
+    use crate::tokenizer::Token;
+
+    #[test]
+    fn test_tokenized_stream() {
+        let tok_text = PreTokenizedString {
+            text: String::from("A a"),
+            tokens: vec![
+                Token {
+                    offset_from: 0,
+                    offset_to: 1,
+                    position: 0,
+                    text: String::from("A"),
+                    position_length: 1,
+                },
+                Token {
+                    offset_from: 2,
+                    offset_to: 3,
+                    position: 1,
+                    text: String::from("a"),
+                    position_length: 1,
+                },
+            ],
+        };
+
+        let mut token_stream = PreTokenizedStream::from(tok_text.clone());
+
+        for expected_token in tok_text.tokens {
+            assert!(token_stream.advance());
+            assert_eq!(token_stream.token(), &expected_token);
+        }
+        assert!(!token_stream.advance());
+    }
+
+    #[test]
+    fn test_chain_tokenized_strings() {
+        let tok_text = PreTokenizedString {
+            text: String::from("A a"),
+            tokens: vec![
+                Token {
+                    offset_from: 0,
+                    offset_to: 1,
+                    position: 0,
+                    text: String::from("A"),
+                    position_length: 1,
+                },
+                Token {
+                    offset_from: 2,
+                    offset_to: 3,
+                    position: 1,
+                    text: String::from("a"),
+                    position_length: 1,
+                },
+            ],
+        };
+
+        let chain_parts = vec![&tok_text, &tok_text];
+
+        let mut token_stream = PreTokenizedStream::chain_tokenized_strings(&chain_parts[..]);
+
+        let expected_tokens = vec![
+            Token {
+                offset_from: 0,
+                offset_to: 1,
+                position: 0,
+                text: String::from("A"),
+                position_length: 1,
+            },
+            Token {
+                offset_from: 2,
+                offset_to: 3,
+                position: 1,
+                text: String::from("a"),
+                position_length: 1,
+            },
+            Token {
+                offset_from: 3,
+                offset_to: 4,
+                position: 3,
+                text: String::from("A"),
+                position_length: 1,
+            },
+            Token {
+                offset_from: 5,
+                offset_to: 6,
+                position: 4,
+                text: String::from("a"),
+                position_length: 1,
+            },
+        ];
+
+        for expected_token in expected_tokens {
+            assert!(token_stream.advance());
+            assert_eq!(token_stream.token(), &expected_token);
+        }
+        assert!(!token_stream.advance());
+    }
+}

src/tokenizer/tokenizer.rs

@@ -4,7 +4,7 @@ use crate::tokenizer::TokenStreamChain;
 use std::borrow::{Borrow, BorrowMut};
 
 /// Token
-#[derive(Debug, Clone)]
+#[derive(Debug, Clone, Serialize, Deserialize, Eq, PartialEq)]
 pub struct Token {
     /// Offset (byte index) of the first character of the token.
     /// Offsets shall not be modified by token filters.

tests/failpoints/mod.rs

@@ -28,11 +28,11 @@ fn test_failpoints_managed_directory_gc_if_delete_fails() {
     // The initial 1*off is there to allow for the removal of the
     // lock file.
     fail::cfg("RAMDirectory::delete", "1*off->1*return").unwrap();
-    managed_directory.garbage_collect(Default::default);
+    assert!(managed_directory.garbage_collect(Default::default).is_ok());
     assert!(managed_directory.exists(test_path));
 
     // running the gc a second time should remove the file.
-    managed_directory.garbage_collect(Default::default);
+    assert!(managed_directory.garbage_collect(Default::default).is_ok());
     assert!(
         !managed_directory.exists(test_path),
         "The file should have been deleted"