Merge pull request #98 from Kerollmops/updates-with-transactions

Change updates to be handled using the RocksDB WriteBatch feature

.travis.yml

@@ -11,8 +11,8 @@ matrix:
   include:
 
   # Test crates on their minimum Rust versions.
-  - rust: 1.31.0
+  - rust: 1.32.0
-    name: "meilidb on 1.31.0"
+    name: "meilidb on 1.32.0"
     script: ./ci/meilidb.sh
 
   # Test crates on nightly Rust.

Cargo.toml

@@ -49,11 +49,6 @@ rand_xorshift = "0.1"
 structopt = "0.2"
 tempfile = "3.0"
 termcolor = "1.0"
-warp = "0.1"
-
-[dev-dependencies.chashmap]
-git = "https://gitlab.redox-os.org/redox-os/tfs.git"
-rev = "b3e7cae1"
 
 [profile.release]
 debug = true

examples/create-database.rs

@@ -11,7 +11,7 @@ use std::fs::File;
 use serde_derive::{Serialize, Deserialize};
 use structopt::StructOpt;
 
-use meilidb::database::{Database, Schema, UpdateBuilder};
+use meilidb::database::{Database, Schema};
 use meilidb::tokenizer::DefaultBuilder;
 
 #[derive(Debug, StructOpt)]
@@ -61,8 +61,7 @@ fn index(
 
     while !end_of_file {
         let tokenizer_builder = DefaultBuilder::new();
-        let update_path = tempfile::NamedTempFile::new()?;
+        let mut update = database.update()?;
-        let mut update = UpdateBuilder::new(update_path.path().to_path_buf(), schema.clone());
 
         loop {
             end_of_file = !rdr.read_record(&mut raw_record)?;
@@ -88,10 +87,8 @@ fn index(
 
         println!();
 
-        println!("building update...");
+        println!("committing update...");
-        let update = update.build()?;
+        update.commit()?;
-        println!("ingesting update...");
-        database.ingest_update_file(update)?;
     }
 
     Ok(database)

examples/http-server.rs

@@ -1,431 +0,0 @@
-#[global_allocator]
-static ALLOC: jemallocator::Jemalloc = jemallocator::Jemalloc;
-
-use std::collections::{HashMap, HashSet};
-use std::error::Error;
-use std::ffi::OsStr;
-use std::fmt;
-use std::fs::{self, File};
-use std::io::{self, BufRead, BufReader};
-use std::net::SocketAddr;
-use std::path::{PathBuf, Path};
-use std::sync::Arc;
-use std::time::SystemTime;
-
-use chashmap::{CHashMap, ReadGuard};
-use elapsed::measure_time;
-use meilidb::database::{Database, UpdateBuilder};
-use meilidb::database::schema::{Schema, SchemaBuilder};
-use meilidb::tokenizer::DefaultBuilder;
-use serde_derive::{Serialize, Deserialize};
-use structopt::StructOpt;
-use warp::{Rejection, Filter};
-use log::{error, info};
-
-#[derive(Debug, StructOpt)]
-pub struct Opt {
-    /// The destination where the database must be created.
-    #[structopt(parse(from_os_str))]
-    pub database_path: PathBuf,
-
-    /// The address and port to bind the server to.
-    #[structopt(short = "l", default_value = "127.0.0.1:8080")]
-    pub listen_addr: SocketAddr,
-
-    /// The path to the list of stop words (one by line).
-    #[structopt(long = "stop-words", parse(from_os_str))]
-    pub stop_words: PathBuf,
-}
-
-//
-// ERRORS FOR THE MULTIDATABASE
-//
-
-#[derive(Debug)]
-pub enum DatabaseError {
-    AlreadyExist,
-    NotExist,
-    NotFound(String),
-    Unknown(Box<Error>),
-}
-
-impl fmt::Display for DatabaseError {
-    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
-        match self {
-            DatabaseError::AlreadyExist => write!(f, "File already exist"),
-            DatabaseError::NotExist => write!(f, "File not exist"),
-            DatabaseError::NotFound(ref name) => write!(f, "Database {} not found", name),
-            DatabaseError::Unknown(e) => write!(f, "{}", e),
-        }
-    }
-}
-
-impl Error for DatabaseError {}
-
-impl From<Box<Error>> for DatabaseError {
-    fn from(e: Box<Error>) -> DatabaseError {
-        DatabaseError::Unknown(e)
-    }
-}
-
-//
-// MULTIDATABASE DEFINITION
-//
-
-pub struct MultiDatabase {
-    databases: CHashMap<String, Database>,
-    db_path: PathBuf,
-    stop_words: HashSet<String>,
-}
-
-impl MultiDatabase {
-
-    pub fn new(path: PathBuf, stop_words: HashSet<String>) -> MultiDatabase {
-        MultiDatabase {
-            databases: CHashMap::new(),
-            db_path: path,
-            stop_words: stop_words
-        }
-    }
-
-    pub fn create(&self, name: String, schema: Schema) -> Result<(), DatabaseError> {
-        let rdb_name = format!("{}.mdb", name);
-        let database_path = self.db_path.join(rdb_name);
-
-        if database_path.exists() {
-            return Err(DatabaseError::AlreadyExist.into());
-        }
-
-        let index = Database::create(database_path, &schema)?;
-
-        self.databases.insert_new(name, index);
-
-        Ok(())
-    }
-
-    pub fn load(&self, name: String) -> Result<(), DatabaseError> {
-        let rdb_name = format!("{}.mdb", name);
-        let index_path = self.db_path.join(rdb_name);
-
-        if !index_path.exists() {
-            return Err(DatabaseError::NotExist.into());
-        }
-
-        let index = Database::open(index_path)?;
-
-        self.databases.insert_new(name, index);
-
-        Ok(())
-    }
-
-    pub fn load_existing(&self) {
-        let paths = match fs::read_dir(self.db_path.clone()){
-            Ok(p) => p,
-            Err(e) => {
-                error!("{}", e);
-                return
-            }
-        };
-
-        for path in paths {
-            let path = match path {
-                Ok(p) => p.path(),
-                Err(_) => continue
-            };
-
-            let path_str = match path.to_str() {
-                Some(p) => p,
-                None => continue
-            };
-
-            let extension = match get_extension_from_path(path_str) {
-                Some(e) => e,
-                None => continue
-            };
-
-            if extension != "mdb" {
-                continue
-            }
-
-            let name = match get_file_name_from_path(path_str) {
-                Some(f) => f,
-                None => continue
-            };
-
-            let db = match Database::open(path.clone()) {
-                Ok(db) => db,
-                Err(_) => continue
-            };
-
-            self.databases.insert_new(name.to_string(), db);
-            info!("Load database {}", name);
-        }
-    }
-
-    pub fn create_or_load(&self, name: String, schema: Schema) -> Result<(), DatabaseError> {
-        match self.create(name.clone(), schema) {
-            Err(DatabaseError::AlreadyExist) => self.load(name),
-            x => x,
-        }
-    }
-
-    pub fn get(&self, name: String) -> Result<ReadGuard<String, Database>, Box<Error>> {
-        Ok(self.databases.get(&name).ok_or(DatabaseError::NotFound(name))?)
-    }
-}
-
-fn get_extension_from_path(path: &str) -> Option<&str> {
-    Path::new(path).extension().and_then(OsStr::to_str)
-}
-
-fn get_file_name_from_path(path: &str) -> Option<&str> {
-    Path::new(path).file_stem().and_then(OsStr::to_str)
-}
-
-fn retrieve_stop_words(path: &Path) -> io::Result<HashSet<String>> {
-    let f = File::open(path)?;
-    let reader = BufReader::new(f);
-    let mut words = HashSet::new();
-
-    for line in reader.lines() {
-        let line = line?;
-        let word = line.trim().to_string();
-        words.insert(word);
-    }
-
-    Ok(words)
-}
-
-//
-// PARAMS & BODY FOR HTTPS HANDLERS
-//
-
-#[derive(Deserialize)]
-struct CreateBody {
-    name: String,
-    schema: SchemaBuilder,
-}
-
-#[derive(Deserialize)]
-struct IngestBody {
-    insert: Option<Vec<HashMap<String, String>>>,
-    delete: Option<Vec<HashMap<String, String>>>
-}
-
-#[derive(Serialize)]
-struct IngestResponse {
-    inserted: usize,
-    deleted: usize
-}
-
-#[derive(Deserialize)]
-struct SearchQuery {
-    q: String,
-    limit: Option<usize>,
-}
-
-//
-// HTTP ROUTES
-//
-
-// Create a new index.
-// The index name should be unused and the schema valid.
-//
-// POST /create
-// Body:
-//     - name: String
-//     - schema: JSON
-//     - stopwords: Vec<String>
-fn create(body: CreateBody, db: Arc<MultiDatabase>) -> Result<String, Rejection>  {
-    let schema = body.schema.build();
-
-    match db.create(body.name.clone(), schema) {
-        Ok(_) => Ok(format!("{} created ", body.name)),
-        Err(e) => {
-            error!("{:?}", e);
-            return Err(warp::reject::not_found())
-        }
-    }
-}
-
-// Ingest new document.
-// It's possible to have positive or/and negative updates.
-//
-// PUT /:name/ingest
-// Body:
-//     - insert: Option<Vec<JSON>>
-//     - delete: Option<Vec<String>>
-fn ingest(index_name: String, body: IngestBody, db: Arc<MultiDatabase>) -> Result<String, Rejection>  {
-
-    let schema = {
-        let index = match db.get(index_name.clone()){
-            Ok(i) => i,
-            Err(_) => return Err(warp::reject::not_found()),
-        };
-        let view = index.view();
-
-        view.schema().clone()
-    };
-
-    let tokenizer_builder = DefaultBuilder::new();
-    let now = match SystemTime::now().duration_since(SystemTime::UNIX_EPOCH) {
-        Ok(n) => n.as_secs(),
-        Err(_) => panic!("SystemTime before UNIX EPOCH!"),
-    };
-
-    let sst_name = format!("update-{}-{}.sst", index_name, now);
-    let sst_path = db.db_path.join(sst_name);
-
-    let mut response = IngestResponse{inserted: 0, deleted: 0};
-    let mut update = UpdateBuilder::new(sst_path, schema);
-
-    if let Some(documents) = body.delete {
-        for doc in documents {
-            if let Err(e) = update.remove_document(doc) {
-                error!("Impossible to remove document; {:?}", e);
-            } else {
-                response.deleted += 1;
-            }
-        }
-    }
-
-    let stop_words = &db.stop_words;
-    if let Some(documents) = body.insert {
-        for doc in documents {
-            if let Err(e) = update.update_document(doc, &tokenizer_builder, &stop_words) {
-                error!("Impossible to update document; {:?}", e);
-            } else {
-                response.inserted += 1;
-            }
-        }
-    }
-
-
-    let update = match update.build() {
-        Ok(u) => u,
-        Err(e) => {
-            error!("Impossible to create an update file; {:?}", e);
-            return Err(warp::reject::not_found())
-        }
-    };
-
-    {
-        let index = match db.get(index_name.clone()){
-            Ok(i) => i,
-            Err(_) => return Err(warp::reject::not_found()),
-        };
-
-        if let Err(e) = index.ingest_update_file(update) {
-            error!("Impossible to ingest sst file; {:?}", e);
-            return Err(warp::reject::not_found())
-        };
-    }
-
-    if let Ok(response) = serde_json::to_string(&response) {
-        return Ok(response);
-    };
-
-    return Err(warp::reject::not_found())
-}
-
-// Search in a specific index
-// The default limit is 20
-//
-// GET /:name/search
-// Params:
-//     - query: String
-//     - limit: Option<usize>
-fn search(index_name: String, query: SearchQuery, db: Arc<MultiDatabase>) -> Result<String, Rejection>  {
-
-    let view = {
-        let index = match db.get(index_name.clone()){
-            Ok(i) => i,
-            Err(_) => return Err(warp::reject::not_found()),
-        };
-        index.view()
-    };
-
-    let limit = query.limit.unwrap_or(20);
-
-    let query_builder = match view.query_builder() {
-        Ok(q) => q,
-        Err(_err) => return Err(warp::reject::not_found()),
-    };
-
-    let (time, responses) = measure_time(|| {
-        let docs = query_builder.query(&query.q, 0..limit);
-        let mut results: Vec<HashMap<String, String>> = Vec::with_capacity(limit);
-        for doc in docs {
-            match view.document_by_id(doc.id) {
-                Ok(val) => results.push(val),
-                Err(e) => println!("{:?}", e),
-            }
-        }
-        results
-    });
-
-    let response = match serde_json::to_string(&responses) {
-        Ok(val) => val,
-        Err(err) => format!("{:?}", err),
-    };
-
-    info!("index: {} - search: {:?} - limit: {} - time: {}", index_name, query.q, limit, time);
-    Ok(response)
-}
-
-fn start_server(listen_addr: SocketAddr, db: Arc<MultiDatabase>) {
-    let index_path = warp::path("index").and(warp::path::param::<String>());
-    let db = warp::any().map(move || db.clone());
-
-    let create_path = warp::path("create").and(warp::path::end());
-    let ingest_path = index_path.and(warp::path("ingest")).and(warp::path::end());
-    let search_path = index_path.and(warp::path("search")).and(warp::path::end());
-
-    let create = warp::post2()
-        .and(create_path)
-        .and(warp::body::json())
-        .and(db.clone())
-        .and_then(create);
-
-    let ingest = warp::put2()
-        .and(ingest_path)
-        .and(warp::body::json())
-        .and(db.clone())
-        .and_then(ingest);
-
-    let search = warp::get2()
-        .and(search_path)
-        .and(warp::query())
-        .and(db.clone())
-        .and_then(search);
-
-    let api = create
-        .or(ingest)
-        .or(search);
-
-    let logs = warp::log("server");
-    let headers = warp::reply::with::header("Content-Type", "application/json");
-
-    let routes = api.with(logs).with(headers);
-
-    info!("Server is started on {}", listen_addr);
-    warp::serve(routes).run(listen_addr);
-}
-
-fn main() {
-    env_logger::init();
-    let opt = Opt::from_args();
-
-    let stop_words = match retrieve_stop_words(&opt.stop_words) {
-        Ok(s) => s,
-        Err(_) => HashSet::new(),
-    };
-
-    let db = Arc::new(MultiDatabase::new(opt.database_path.clone(), stop_words));
-
-    db.load_existing();
-
-    start_server(opt.listen_addr, db);
-}
-
-

src/database/document_key.rs

@@ -38,6 +38,10 @@ impl DocumentKey {
         DocumentKeyAttr::new(self.document_id(), attr)
     }
 
+    pub fn with_attribute_min(&self) -> DocumentKeyAttr {
+        DocumentKeyAttr::new(self.document_id(), SchemaAttr::min())
+    }
+
     pub fn with_attribute_max(&self) -> DocumentKeyAttr {
         DocumentKeyAttr::new(self.document_id(), SchemaAttr::max())
     }

src/database/mod.rs

@@ -1,17 +1,17 @@
-use std::sync::{Arc, Mutex};
 use std::error::Error;
-use std::ops::Deref;
 use std::path::Path;
+use std::ops::Deref;
+use std::sync::Arc;
 
-use rocksdb::rocksdb_options::{DBOptions, IngestExternalFileOptions, ColumnFamilyOptions};
+use rocksdb::rocksdb_options::{DBOptions, ColumnFamilyOptions};
 use rocksdb::rocksdb::{Writable, Snapshot};
-use rocksdb::{DB, DBVector, MergeOperands};
+use rocksdb::{DB, MergeOperands};
 use crossbeam::atomic::ArcCell;
 use log::info;
 
 pub use self::document_key::{DocumentKey, DocumentKeyAttr};
 pub use self::view::{DatabaseView, DocumentIter};
-pub use self::update::{Update, UpdateBuilder};
+pub use self::update::Update;
 pub use self::serde::SerializerError;
 pub use self::schema::Schema;
 pub use self::index::Index;
@@ -78,11 +78,7 @@ fn merge_indexes(key: &[u8], existing: Option<&[u8]>, operands: &mut MergeOperan
 }
 
 pub struct Database {
-    // DB is under a Mutex to sync update ingestions and separate DB update locking
+    db: Arc<DB>,
-    // and DatabaseView acquiring locking in other words:
-    // "Block readers the minimum possible amount of time"
-    db: Mutex<Arc<DB>>,
-
     // This view is updated each time the DB ingests an update
     view: ArcCell<DatabaseView<Arc<DB>>>,
 }
@@ -113,7 +109,7 @@ impl Database {
         let snapshot = Snapshot::new(db.clone());
         let view = ArcCell::new(Arc::new(DatabaseView::new(snapshot)?));
 
-        Ok(Database { db: Mutex::new(db), view })
+        Ok(Database { db, view })
     }
 
     pub fn open<P: AsRef<Path>>(path: P) -> Result<Database, Box<Error>> {
@@ -137,49 +133,16 @@ impl Database {
         let snapshot = Snapshot::new(db.clone());
         let view = ArcCell::new(Arc::new(DatabaseView::new(snapshot)?));
 
-        Ok(Database { db: Mutex::new(db), view })
+        Ok(Database { db, view })
     }
 
-    pub fn ingest_update_file(&self, update: Update) -> Result<Arc<DatabaseView<Arc<DB>>>, Box<Error>> {
+    pub fn update(&self) -> Result<Update, Box<Error>> {
-        let snapshot = {
+        let schema = match self.db.get(DATA_SCHEMA)? {
-            // We must have a mutex here to ensure that update ingestions and compactions
+            Some(value) => Schema::read_from_bin(&*value)?,
-            // are done atomatically and in the right order.
+            None => panic!("Database does not contain a schema"),
-            // This way update ingestions will block other update ingestions without blocking view
-            // creations while doing the "data-index" compaction
-            let db = match self.db.lock() {
-                Ok(db) => db,
-                Err(e) => return Err(e.to_string().into()),
         };
 
-            let path = update.path().to_string_lossy();
+        Ok(Update::new(self, schema))
-            let options = IngestExternalFileOptions::new();
-            // options.move_files(move_update);
-
-            let (elapsed, result) = elapsed::measure_time(|| {
-                let cf_handle = db.cf_handle("default").expect("\"default\" column family not found");
-                db.ingest_external_file_optimized(&cf_handle, &options, &[&path])
-            });
-            let _ = result?;
-            info!("ingesting update file took {}", elapsed);
-
-            Snapshot::new(db.clone())
-        };
-
-        let view = Arc::new(DatabaseView::new(snapshot)?);
-        self.view.set(view.clone());
-
-        Ok(view)
-    }
-
-    pub fn get(&self, key: &[u8]) -> Result<Option<DBVector>, Box<Error>> {
-        self.view().get(key)
-    }
-
-    pub fn flush(&self) -> Result<(), Box<Error>> {
-        match self.db.lock() {
-            Ok(db) => Ok(db.flush(true)?),
-            Err(e) => Err(e.to_string().into()),
-        }
     }
 
     pub fn view(&self) -> Arc<DatabaseView<Arc<DB>>> {
@@ -193,20 +156,18 @@ mod tests {
     use std::error::Error;
 
     use serde_derive::{Serialize, Deserialize};
-    use tempfile::tempdir;
 
     use crate::database::schema::{SchemaBuilder, STORED, INDEXED};
-    use crate::database::update::UpdateBuilder;
     use crate::tokenizer::DefaultBuilder;
 
     use super::*;
 
     #[test]
-    fn ingest_one_update_file() -> Result<(), Box<Error>> {
+    fn ingest_one_easy_update() -> Result<(), Box<Error>> {
-        let dir = tempdir()?;
+        let dir = tempfile::tempdir()?;
         let stop_words = HashSet::new();
 
-        let rocksdb_path = dir.path().join("rocksdb.rdb");
+        let meilidb_path = dir.path().join("meilidb.mdb");
 
         #[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
         struct SimpleDoc {
@@ -225,9 +186,7 @@ mod tests {
             builder.build()
         };
 
-        let database = Database::create(&rocksdb_path, &schema)?;
+        let database = Database::create(&meilidb_path, &schema)?;
-
-        let update_path = dir.path().join("update.sst");
 
         let doc0 = SimpleDoc {
             id: 0,
@@ -242,20 +201,13 @@ mod tests {
             timestamp: 7654321,
         };
 
-        let docid0;
-        let docid1;
-        let update = {
         let tokenizer_builder = DefaultBuilder::new();
-            let mut builder = UpdateBuilder::new(update_path, schema);
+        let mut builder = database.update()?;
 
-            docid0 = builder.update_document(&doc0, &tokenizer_builder, &stop_words)?;
+        let docid0 = builder.update_document(&doc0, &tokenizer_builder, &stop_words)?;
-            docid1 = builder.update_document(&doc1, &tokenizer_builder, &stop_words)?;
+        let docid1 = builder.update_document(&doc1, &tokenizer_builder, &stop_words)?;
 
-            builder.build()?
+        let view = builder.commit()?;
-        };
-
-        database.ingest_update_file(update)?;
-        let view = database.view();
 
         let de_doc0: SimpleDoc = view.document_by_id(docid0)?;
         let de_doc1: SimpleDoc = view.document_by_id(docid1)?;
@@ -267,11 +219,11 @@ mod tests {
     }
 
     #[test]
-    fn ingest_two_update_files() -> Result<(), Box<Error>> {
+    fn ingest_two_easy_updates() -> Result<(), Box<Error>> {
-        let dir = tempdir()?;
+        let dir = tempfile::tempdir()?;
         let stop_words = HashSet::new();
 
-        let rocksdb_path = dir.path().join("rocksdb.rdb");
+        let meilidb_path = dir.path().join("meilidb.mdb");
 
         #[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
         struct SimpleDoc {
@@ -290,7 +242,7 @@ mod tests {
             builder.build()
         };
 
-        let database = Database::create(&rocksdb_path, &schema)?;
+        let database = Database::create(&meilidb_path, &schema)?;
 
         let doc0 = SimpleDoc {
             id: 0,
@@ -317,36 +269,17 @@ mod tests {
             timestamp: 7654321,
         };
 
-        let docid0;
-        let docid1;
-        let update1 = {
         let tokenizer_builder = DefaultBuilder::new();
-            let update_path = dir.path().join("update-000.sst");
-            let mut builder = UpdateBuilder::new(update_path, schema.clone());
 
-            docid0 = builder.update_document(&doc0, &tokenizer_builder, &stop_words)?;
+        let mut builder = database.update()?;
-            docid1 = builder.update_document(&doc1, &tokenizer_builder, &stop_words)?;
+        let docid0 = builder.update_document(&doc0, &tokenizer_builder, &stop_words)?;
+        let docid1 = builder.update_document(&doc1, &tokenizer_builder, &stop_words)?;
+        builder.commit()?;
 
-            builder.build()?
+        let mut builder = database.update()?;
-        };
+        let docid2 = builder.update_document(&doc2, &tokenizer_builder, &stop_words)?;
-
+        let docid3 = builder.update_document(&doc3, &tokenizer_builder, &stop_words)?;
-        let docid2;
+        let view = builder.commit()?;
-        let docid3;
-        let update2 = {
-            let tokenizer_builder = DefaultBuilder::new();
-            let update_path = dir.path().join("update-001.sst");
-            let mut builder = UpdateBuilder::new(update_path, schema);
-
-            docid2 = builder.update_document(&doc2, &tokenizer_builder, &stop_words)?;
-            docid3 = builder.update_document(&doc3, &tokenizer_builder, &stop_words)?;
-
-            builder.build()?
-        };
-
-        database.ingest_update_file(update1)?;
-        database.ingest_update_file(update2)?;
-
-        let view = database.view();
 
         let de_doc0: SimpleDoc = view.document_by_id(docid0)?;
         let de_doc1: SimpleDoc = view.document_by_id(docid1)?;
@@ -380,7 +313,6 @@ mod bench {
     use rand::seq::SliceRandom;
 
     use crate::tokenizer::DefaultBuilder;
-    use crate::database::update::UpdateBuilder;
     use crate::database::schema::*;
 
     use super::*;
@@ -425,9 +357,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..300 {
@@ -439,8 +370,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        builder.commit()?;
-        database.ingest_update_file(update)?;
 
         drop(database);
 
@@ -472,9 +402,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..3000 {
@@ -486,8 +415,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        builder.commit()?;
-        database.ingest_update_file(update)?;
 
         drop(database);
 
@@ -520,9 +448,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..30_000 {
@@ -534,8 +461,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        builder.commit()?;
-        database.ingest_update_file(update)?;
 
         drop(database);
 
@@ -567,9 +493,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..300 {
@@ -581,8 +506,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        let view = builder.commit()?;
-        let view = database.ingest_update_file(update)?;
 
         bench.iter(|| {
             for q in &["a", "b", "c", "d", "e"] {
@@ -614,9 +538,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..3000 {
@@ -628,8 +551,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        let view = builder.commit()?;
-        let view = database.ingest_update_file(update)?;
 
         bench.iter(|| {
             for q in &["a", "b", "c", "d", "e"] {
@@ -662,9 +584,8 @@ mod bench {
             description: String,
         }
 
-        let path = dir.path().join("update-000.sst");
         let tokenizer_builder = DefaultBuilder;
-        let mut builder = UpdateBuilder::new(path, schema);
+        let mut builder = database.update()?;
         let mut rng = XorShiftRng::seed_from_u64(42);
 
         for i in 0..30_000 {
@@ -676,8 +597,7 @@ mod bench {
             builder.update_document(&document, &tokenizer_builder, &stop_words)?;
         }
 
-        let update = builder.build()?;
+        let view = builder.commit()?;
-        let view = database.ingest_update_file(update)?;
 
         bench.iter(|| {
             for q in &["a", "b", "c", "d", "e"] {

src/database/serde/indexer_serializer.rs

@@ -10,15 +10,15 @@ use crate::tokenizer::TokenizerBuilder;
 use crate::tokenizer::Token;
 use crate::{DocumentId, DocIndex};
 
-pub struct IndexerSerializer<'a, B> {
+pub struct IndexerSerializer<'a, 'b, B> {
     pub tokenizer_builder: &'a B,
-    pub update: &'a mut DocumentUpdate,
+    pub update: &'a mut DocumentUpdate<'b>,
     pub document_id: DocumentId,
     pub attribute: SchemaAttr,
     pub stop_words: &'a HashSet<String>,
 }
 
-impl<'a, B> ser::Serializer for IndexerSerializer<'a, B>
+impl<'a, 'b, B> ser::Serializer for IndexerSerializer<'a, 'b, B>
 where B: TokenizerBuilder
 {
     type Ok = ();
@@ -71,14 +71,14 @@ where B: TokenizerBuilder
                 let char_length = length;
 
                 let doc_index = DocIndex { document_id, attribute, word_index, char_index, char_length };
-                self.update.insert_doc_index(word_unidecoded.into_bytes(), doc_index);
+                self.update.insert_doc_index(word_unidecoded.into_bytes(), doc_index)?;
             }
 
             let char_index = char_index as u32;
             let char_length = length;
 
             let doc_index = DocIndex { document_id, attribute, word_index, char_index, char_length };
-            self.update.insert_doc_index(word_lower.into_bytes(), doc_index);
+            self.update.insert_doc_index(word_lower.into_bytes(), doc_index)?;
         }
         Ok(())
     }

src/database/serde/mod.rs

@@ -56,3 +56,9 @@ impl fmt::Display for SerializerError {
 }
 
 impl Error for SerializerError {}
+
+impl From<String> for SerializerError {
+    fn from(value: String) -> SerializerError {
+        SerializerError::Custom(value)
+    }
+}

src/database/serde/serializer.rs

@@ -11,15 +11,15 @@ use crate::tokenizer::TokenizerBuilder;
 use crate::database::schema::Schema;
 use crate::DocumentId;
 
-pub struct Serializer<'a, B> {
+pub struct Serializer<'a, 'b, B> {
     pub schema: &'a Schema,
-    pub update: &'a mut DocumentUpdate,
+    pub update: &'a mut DocumentUpdate<'b>,
     pub document_id: DocumentId,
     pub tokenizer_builder: &'a B,
     pub stop_words: &'a HashSet<String>,
 }
 
-impl<'a, B> ser::Serializer for Serializer<'a, B>
+impl<'a, 'b, B> ser::Serializer for Serializer<'a, 'b, B>
 where B: TokenizerBuilder
 {
     type Ok = ();
@@ -28,8 +28,8 @@ where B: TokenizerBuilder
     type SerializeTuple = ser::Impossible<Self::Ok, Self::Error>;
     type SerializeTupleStruct = ser::Impossible<Self::Ok, Self::Error>;
     type SerializeTupleVariant = ser::Impossible<Self::Ok, Self::Error>;
-    type SerializeMap = MapSerializer<'a, B>;
+    type SerializeMap = MapSerializer<'a, 'b, B>;
-    type SerializeStruct = StructSerializer<'a, B>;
+    type SerializeStruct = StructSerializer<'a, 'b, B>;
     type SerializeStructVariant = ser::Impossible<Self::Ok, Self::Error>;
 
     forward_to_unserializable_type! {
@@ -174,16 +174,16 @@ where B: TokenizerBuilder
     }
 }
 
-pub struct MapSerializer<'a, B> {
+pub struct MapSerializer<'a, 'b, B> {
     pub schema: &'a Schema,
     pub document_id: DocumentId,
-    pub update: &'a mut DocumentUpdate,
+    pub update: &'a mut DocumentUpdate<'b>,
     pub tokenizer_builder: &'a B,
     pub stop_words: &'a HashSet<String>,
     pub current_key_name: Option<String>,
 }
 
-impl<'a, B> ser::SerializeMap for MapSerializer<'a, B>
+impl<'a, 'b, B> ser::SerializeMap for MapSerializer<'a, 'b, B>
 where B: TokenizerBuilder
 {
     type Ok = ();
@@ -207,7 +207,7 @@ where B: TokenizerBuilder
     fn serialize_entry<K: ?Sized, V: ?Sized>(
         &mut self,
         key: &K,
-        value: &V
+        value: &V,
     ) -> Result<(), Self::Error>
     where K: Serialize, V: Serialize,
     {
@@ -217,7 +217,7 @@ where B: TokenizerBuilder
             let props = self.schema.props(attr);
             if props.is_stored() {
                 let value = bincode::serialize(value).unwrap();
-                self.update.insert_attribute_value(attr, value);
+                self.update.insert_attribute_value(attr, &value)?;
             }
             if props.is_indexed() {
                 let serializer = IndexerSerializer {
@@ -239,15 +239,15 @@ where B: TokenizerBuilder
     }
 }
 
-pub struct StructSerializer<'a, B> {
+pub struct StructSerializer<'a, 'b, B> {
     pub schema: &'a Schema,
     pub document_id: DocumentId,
-    pub update: &'a mut DocumentUpdate,
+    pub update: &'a mut DocumentUpdate<'b>,
     pub tokenizer_builder: &'a B,
     pub stop_words: &'a HashSet<String>,
 }
 
-impl<'a, B> ser::SerializeStruct for StructSerializer<'a, B>
+impl<'a, 'b, B> ser::SerializeStruct for StructSerializer<'a, 'b, B>
 where B: TokenizerBuilder
 {
     type Ok = ();
@@ -264,7 +264,7 @@ where B: TokenizerBuilder
             let props = self.schema.props(attr);
             if props.is_stored() {
                 let value = bincode::serialize(value).unwrap();
-                self.update.insert_attribute_value(attr, value);
+                self.update.insert_attribute_value(attr, &value)?;
             }
             if props.is_indexed() {
                 let serializer = IndexerSerializer {

src/database/update.rs

@@ -0,0 +1,209 @@
+use std::collections::{HashSet, BTreeMap};
+use std::error::Error;
+use std::sync::Arc;
+
+use rocksdb::rocksdb::{DB, Writable, Snapshot, WriteBatch};
+use hashbrown::hash_map::HashMap;
+use serde::Serialize;
+use fst::map::Map;
+use sdset::Set;
+
+use crate::database::{DATA_INDEX, Database, DatabaseView};
+use crate::database::index::{Positive, PositiveBuilder, Negative};
+use crate::database::document_key::{DocumentKey, DocumentKeyAttr};
+use crate::database::serde::serializer::Serializer;
+use crate::database::serde::SerializerError;
+use crate::database::schema::SchemaAttr;
+use crate::tokenizer::TokenizerBuilder;
+use crate::data::{DocIds, DocIndexes};
+use crate::database::schema::Schema;
+use crate::{DocumentId, DocIndex};
+use crate::database::index::Index;
+
+pub type Token = Vec<u8>; // TODO could be replaced by a SmallVec
+
+pub struct Update<'a> {
+    database: &'a Database,
+    schema: Schema,
+    raw_builder: RawUpdateBuilder,
+}
+
+impl<'a> Update<'a> {
+    pub(crate) fn new(database: &'a Database, schema: Schema) -> Update<'a> {
+        Update { database, schema, raw_builder: RawUpdateBuilder::new() }
+    }
+
+    pub fn update_document<T, B>(
+        &mut self,
+        document: T,
+        tokenizer_builder: &B,
+        stop_words: &HashSet<String>,
+    ) -> Result<DocumentId, SerializerError>
+    where T: Serialize,
+          B: TokenizerBuilder,
+    {
+        let document_id = self.schema.document_id(&document)?;
+
+        let serializer = Serializer {
+            schema: &self.schema,
+            document_id: document_id,
+            tokenizer_builder: tokenizer_builder,
+            update: &mut self.raw_builder.document_update(document_id)?,
+            stop_words: stop_words,
+        };
+
+        document.serialize(serializer)?;
+
+        Ok(document_id)
+    }
+
+    pub fn remove_document<T>(&mut self, document: T) -> Result<DocumentId, SerializerError>
+    where T: Serialize,
+    {
+        let document_id = self.schema.document_id(&document)?;
+        self.raw_builder.document_update(document_id)?.remove()?;
+        Ok(document_id)
+    }
+
+    pub fn commit(self) -> Result<Arc<DatabaseView<Arc<DB>>>, Box<Error>> {
+        let batch = self.raw_builder.build()?;
+        self.database.db.write(batch)?;
+
+        let snapshot = Snapshot::new(self.database.db.clone());
+        let view = Arc::new(DatabaseView::new(snapshot)?);
+        self.database.view.set(view.clone());
+
+        Ok(view)
+    }
+
+    pub fn abort(self) { }
+}
+
+#[derive(Copy, Clone, PartialEq, Eq)]
+enum UpdateType {
+    Updated,
+    Deleted,
+}
+
+use UpdateType::{Updated, Deleted};
+
+pub struct RawUpdateBuilder {
+    documents_update: HashMap<DocumentId, UpdateType>,
+    indexed_words: BTreeMap<Token, Vec<DocIndex>>,
+    batch: WriteBatch,
+}
+
+impl RawUpdateBuilder {
+    pub fn new() -> RawUpdateBuilder {
+        RawUpdateBuilder {
+            documents_update: HashMap::new(),
+            indexed_words: BTreeMap::new(),
+            batch: WriteBatch::new(),
+        }
+    }
+
+    pub fn document_update(&mut self, document_id: DocumentId) -> Result<DocumentUpdate, SerializerError> {
+        use serde::ser::Error;
+
+        match self.documents_update.get(&document_id) {
+            Some(Deleted) | None => Ok(DocumentUpdate { document_id, inner: self }),
+            Some(Updated) => Err(SerializerError::custom(
+                "This document has already been removed and cannot be updated in the same update"
+            )),
+        }
+    }
+
+    pub fn build(self) -> Result<WriteBatch, Box<Error>> {
+        let negative = {
+            let mut removed_document_ids = Vec::new();
+            for (id, update_type) in self.documents_update {
+                if update_type == Deleted {
+                    removed_document_ids.push(id);
+                }
+            }
+
+            removed_document_ids.sort_unstable();
+            let removed_document_ids = Set::new_unchecked(&removed_document_ids);
+            let doc_ids = DocIds::new(removed_document_ids);
+
+            Negative::new(doc_ids)
+        };
+
+        let positive = {
+            let mut positive_builder = PositiveBuilder::memory();
+
+            for (key, mut indexes) in self.indexed_words {
+                indexes.sort_unstable();
+                let indexes = Set::new_unchecked(&indexes);
+                positive_builder.insert(key, indexes)?;
+            }
+
+            let (map, indexes) = positive_builder.into_inner()?;
+            let map = Map::from_bytes(map)?;
+            let indexes = DocIndexes::from_bytes(indexes)?;
+
+            Positive::new(map, indexes)
+        };
+
+        let index = Index { negative, positive };
+
+        // write the data-index
+        let mut bytes = Vec::new();
+        index.write_to_bytes(&mut bytes);
+        self.batch.merge(DATA_INDEX, &bytes)?;
+
+        Ok(self.batch)
+    }
+}
+
+pub struct DocumentUpdate<'a> {
+    document_id: DocumentId,
+    inner: &'a mut RawUpdateBuilder,
+}
+
+impl<'a> DocumentUpdate<'a> {
+    pub fn remove(&mut self) -> Result<(), SerializerError> {
+        use serde::ser::Error;
+
+        if let Updated = self.inner.documents_update.entry(self.document_id).or_insert(Deleted) {
+            return Err(SerializerError::custom(
+                "This document has already been updated and cannot be removed in the same update"
+            ));
+        }
+
+        let start = DocumentKey::new(self.document_id).with_attribute_min();
+        let end = DocumentKey::new(self.document_id).with_attribute_max(); // FIXME max + 1
+        self.inner.batch.delete_range(start.as_ref(), end.as_ref())?;
+
+        Ok(())
+    }
+
+    pub fn insert_attribute_value(&mut self, attr: SchemaAttr, value: &[u8]) -> Result<(), SerializerError> {
+        use serde::ser::Error;
+
+        if let Deleted = self.inner.documents_update.entry(self.document_id).or_insert(Updated) {
+            return Err(SerializerError::custom(
+                "This document has already been deleted and cannot be updated in the same update"
+            ));
+        }
+
+        let key = DocumentKeyAttr::new(self.document_id, attr);
+        self.inner.batch.put(key.as_ref(), &value)?;
+
+        Ok(())
+    }
+
+    pub fn insert_doc_index(&mut self, token: Token, doc_index: DocIndex) -> Result<(), SerializerError> {
+        use serde::ser::Error;
+
+        if let Deleted = self.inner.documents_update.entry(self.document_id).or_insert(Updated) {
+            return Err(SerializerError::custom(
+                "This document has already been deleted and cannot be updated in the same update"
+            ));
+        }
+
+        self.inner.indexed_words.entry(token).or_insert_with(Vec::new).push(doc_index);
+
+        Ok(())
+    }
+}

src/database/update/builder.rs

@@ -1,64 +0,0 @@
-use std::collections::HashSet;
-use std::path::PathBuf;
-use std::error::Error;
-
-use serde::Serialize;
-
-use crate::database::serde::serializer::Serializer;
-use crate::database::serde::SerializerError;
-use crate::tokenizer::TokenizerBuilder;
-use crate::database::Schema;
-
-use crate::DocumentId;
-use super::{Update, RawUpdateBuilder};
-
-pub struct UpdateBuilder {
-    schema: Schema,
-    raw_builder: RawUpdateBuilder,
-}
-
-impl UpdateBuilder {
-    pub fn new(path: PathBuf, schema: Schema) -> UpdateBuilder {
-        UpdateBuilder {
-            schema: schema,
-            raw_builder: RawUpdateBuilder::new(path),
-        }
-    }
-
-    pub fn update_document<T, B>(
-        &mut self,
-        document: T,
-        tokenizer_builder: &B,
-        stop_words: &HashSet<String>,
-    ) -> Result<DocumentId, SerializerError>
-    where T: Serialize,
-          B: TokenizerBuilder,
-    {
-        let document_id = self.schema.document_id(&document)?;
-        let update = self.raw_builder.document_update(document_id);
-
-        let serializer = Serializer {
-            schema: &self.schema,
-            document_id: document_id,
-            tokenizer_builder: tokenizer_builder,
-            update: update,
-            stop_words: stop_words,
-        };
-
-        document.serialize(serializer)?;
-
-        Ok(document_id)
-    }
-
-    pub fn remove_document<T>(&mut self, document: T) -> Result<DocumentId, SerializerError>
-    where T: Serialize,
-    {
-        let document_id = self.schema.document_id(&document)?;
-        self.raw_builder.document_update(document_id).remove();
-        Ok(document_id)
-    }
-
-    pub fn build(self) -> Result<Update, Box<Error>> {
-        self.raw_builder.build()
-    }
-}

src/database/update/mod.rs

@@ -1,17 +0,0 @@
-use std::path::{Path, PathBuf};
-
-mod builder;
-mod raw_builder;
-
-pub use self::builder::UpdateBuilder;
-pub use self::raw_builder::{RawUpdateBuilder, DocumentUpdate};
-
-pub struct Update {
-    sst_file: PathBuf,
-}
-
-impl Update {
-    pub fn path(&self) -> &Path {
-        &self.sst_file
-    }
-}

src/database/update/raw_builder.rs

@@ -1,170 +0,0 @@
-use std::collections::btree_map::{BTreeMap, Entry};
-use std::path::PathBuf;
-use std::error::Error;
-
-use rocksdb::rocksdb_options;
-use hashbrown::HashMap;
-use fst::map::Map;
-use sdset::Set;
-use log::warn;
-
-use crate::database::index::{Index, Positive, PositiveBuilder, Negative};
-use crate::database::{DATA_INDEX, DocumentKeyAttr};
-use crate::database::schema::SchemaAttr;
-use crate::data::{DocIds, DocIndexes};
-use crate::{DocumentId, DocIndex};
-use super::Update;
-
-type Token = Vec<u8>; // TODO could be replaced by a SmallVec
-type Value = Vec<u8>;
-
-pub struct RawUpdateBuilder {
-    sst_file: PathBuf,
-    document_updates: BTreeMap<DocumentId, DocumentUpdate>,
-}
-
-pub struct DocumentUpdate {
-    cleared: bool,
-    words_indexes: HashMap<Token, Vec<DocIndex>>,
-    attributes: BTreeMap<SchemaAttr, Value>,
-}
-
-impl DocumentUpdate {
-    pub fn new() -> DocumentUpdate {
-        DocumentUpdate {
-            cleared: false,
-            words_indexes: HashMap::new(),
-            attributes: BTreeMap::new(),
-        }
-    }
-
-    pub fn remove(&mut self) {
-        self.cleared = true;
-        self.clear();
-    }
-
-    pub fn clear(&mut self) {
-        self.words_indexes.clear();
-        self.attributes.clear();
-    }
-
-    pub fn insert_attribute_value(&mut self, attr: SchemaAttr, value: Vec<u8>) {
-        self.attributes.insert(attr, value);
-    }
-
-    pub fn insert_doc_index(&mut self, token: Vec<u8>, doc_index: DocIndex) {
-        self.words_indexes.entry(token).or_insert_with(Vec::new).push(doc_index)
-    }
-}
-
-impl RawUpdateBuilder {
-    pub fn new(path: PathBuf) -> RawUpdateBuilder {
-        RawUpdateBuilder {
-            sst_file: path,
-            document_updates: BTreeMap::new(),
-        }
-    }
-
-    pub fn document_update(&mut self, document_id: DocumentId) -> &mut DocumentUpdate {
-        match self.document_updates.entry(document_id) {
-            Entry::Occupied(mut occupied) => {
-                warn!("Already updated document {:?}, clearing it", document_id);
-                occupied.get_mut().clear();
-                occupied.into_mut()
-            },
-            Entry::Vacant(vacant) => vacant.insert(DocumentUpdate::new()),
-        }
-    }
-
-    pub fn build(mut self) -> Result<Update, Box<Error>> {
-        let mut removed_document_ids = Vec::new();
-        let mut words_indexes = BTreeMap::new();
-
-        for (&id, update) in self.document_updates.iter_mut() {
-            if update.cleared { removed_document_ids.push(id) }
-
-            for (token, indexes) in &update.words_indexes {
-                words_indexes.entry(token).or_insert_with(Vec::new).extend_from_slice(indexes)
-            }
-        }
-
-        let negative = {
-            let removed_document_ids = Set::new_unchecked(&removed_document_ids);
-            let doc_ids = DocIds::new(removed_document_ids);
-            Negative::new(doc_ids)
-        };
-
-        let positive = {
-            let mut positive_builder = PositiveBuilder::memory();
-
-            for (key, mut indexes) in words_indexes {
-                indexes.sort_unstable();
-                let indexes = Set::new_unchecked(&indexes);
-                positive_builder.insert(key, indexes)?;
-            }
-
-            let (map, indexes) = positive_builder.into_inner()?;
-            let map = Map::from_bytes(map)?;
-            let indexes = DocIndexes::from_bytes(indexes)?;
-            Positive::new(map, indexes)
-        };
-
-        let index = Index { negative, positive };
-
-        let env_options = rocksdb_options::EnvOptions::new();
-        let column_family_options = rocksdb_options::ColumnFamilyOptions::new();
-        let mut file_writer = rocksdb::SstFileWriter::new(env_options, column_family_options);
-        file_writer.open(&self.sst_file.to_string_lossy())?;
-
-        // write the data-index
-        let mut bytes = Vec::new();
-        index.write_to_bytes(&mut bytes);
-        file_writer.merge(DATA_INDEX, &bytes)?;
-
-        // write all the documents attributes updates
-        for (id, update) in self.document_updates {
-
-            let mut last_attr: Option<SchemaAttr> = None;
-            for (attr, value) in update.attributes {
-
-                if update.cleared {
-                    // if there is no last attribute, remove from the first attribute
-                    let start_attr = match last_attr {
-                        Some(attr) => attr.next(),
-                        None       => Some(SchemaAttr::min())
-                    };
-                    let start = start_attr.map(|a| DocumentKeyAttr::new(id, a));
-                    let end = attr.prev().map(|a| DocumentKeyAttr::new(id, a));
-
-                    // delete_range between (last_attr + 1) and (attr - 1)
-                    if let (Some(start), Some(end)) = (start, end) {
-                        file_writer.delete_range(start.as_ref(), end.as_ref())?;
-                    }
-                }
-
-                let key = DocumentKeyAttr::new(id, attr);
-                file_writer.put(key.as_ref(), &value)?;
-                last_attr = Some(attr);
-            }
-
-            if update.cleared {
-                // if there is no last attribute, remove from the first attribute
-                let start_attr = match last_attr {
-                    Some(attr) => attr.next(),
-                    None       => Some(SchemaAttr::min())
-                };
-                let start = start_attr.map(|a| DocumentKeyAttr::new(id, a));
-                let end = DocumentKeyAttr::with_attribute_max(id);
-
-                // delete_range between (last_attr + 1) and attr_max
-                if let Some(start) = start {
-                    file_writer.delete_range(start.as_ref(), end.as_ref())?;
-                }
-            }
-        }
-
-        file_writer.finish()?;
-
-        Ok(Update { sst_file: self.sst_file })
-    }
-}