review edits

meilisearch-http/src/index_controller/dump_actor/actor.rs

@@ -1,8 +1,6 @@
+use std::collections::HashMap;
+use std::path::{Path, PathBuf};
 use std::sync::Arc;
-use std::{
-    collections::HashMap,
-    path::{Path, PathBuf},
-};
 
 use async_stream::stream;
 use chrono::Utc;
@@ -24,8 +22,8 @@ pub struct DumpActor<UuidResolver, Update> {
     dump_path: PathBuf,
     lock: Arc<Mutex<()>>,
     dump_infos: Arc<RwLock<HashMap<String, DumpInfo>>>,
-    update_db_size: u64,
-    index_db_size: u64,
+    update_db_size: usize,
+    index_db_size: usize,
 }
 
 /// Generate uid from creation date
@@ -43,8 +41,8 @@ where
         uuid_resolver: UuidResolver,
         update: Update,
         dump_path: impl AsRef<Path>,
-        index_db_size: u64,
-        update_db_size: u64,
+        index_db_size: usize,
+        update_db_size: usize,
     ) -> Self {
         let dump_infos = Arc::new(RwLock::new(HashMap::new()));
         let lock = Arc::new(Mutex::new(()));

meilisearch-http/src/index_controller/dump_actor/handle_impl.rs

@@ -1,8 +1,10 @@
-use super::{DumpActor, DumpActorHandle, DumpInfo, DumpMsg, DumpResult};
-use actix_web::web::Bytes;
 use std::path::Path;
+
+use actix_web::web::Bytes;
 use tokio::sync::{mpsc, oneshot};
 
+use super::{DumpActor, DumpActorHandle, DumpInfo, DumpMsg, DumpResult};
+
 #[derive(Clone)]
 pub struct DumpActorHandleImpl {
     sender: mpsc::Sender<DumpMsg>,
@@ -30,8 +32,8 @@ impl DumpActorHandleImpl {
         path: impl AsRef<Path>,
         uuid_resolver: crate::index_controller::uuid_resolver::UuidResolverHandleImpl,
         update: crate::index_controller::update_actor::UpdateActorHandleImpl<Bytes>,
-        index_db_size: u64,
-        update_db_size: u64,
+        index_db_size: usize,
+        update_db_size: usize,
     ) -> anyhow::Result<Self> {
         let (sender, receiver) = mpsc::channel(10);
         let actor = DumpActor::new(

meilisearch-http/src/index_controller/dump_actor/loaders/v1.rs

@@ -1,22 +1,20 @@
-use std::{
-    collections::{BTreeMap, BTreeSet},
-    fs::File,
-    io::BufRead,
-    marker::PhantomData,
-    path::Path,
-    sync::Arc,
-};
+use std::collections::{BTreeMap, BTreeSet};
+use std::fs::{create_dir_all, File};
+use std::io::BufRead;
+use std::marker::PhantomData;
+use std::path::Path;
+use std::sync::Arc;
 
 use heed::EnvOpenOptions;
 use log::{error, info, warn};
-use milli::update::{IndexDocumentsMethod, UpdateBuilder, UpdateFormat};
+use milli::update::{IndexDocumentsMethod, UpdateFormat};
 use serde::{Deserialize, Serialize};
 use uuid::Uuid;
 
-use crate::{index::deserialize_some, index_controller::uuid_resolver::HeedUuidStore};
+use crate::index_controller::{self, uuid_resolver::HeedUuidStore, IndexMetadata};
 use crate::{
-    index::{Index, Unchecked},
-    index_controller::{self, IndexMetadata},
+    index::{deserialize_some, update_handler::UpdateHandler, Index, Unchecked},
+    option::IndexerOpts,
 };
 
 #[derive(Serialize, Deserialize, Debug)]
@@ -32,28 +30,33 @@ impl MetadataV1 {
         src: impl AsRef<Path>,
         dst: impl AsRef<Path>,
         size: usize,
+        indexer_options: &IndexerOpts,
     ) -> anyhow::Result<()> {
         info!(
             "Loading dump, dump database version: {}, dump version: V1",
             self.db_version
         );
 
-        dbg!("here");
-
         let uuid_store = HeedUuidStore::new(&dst)?;
-        dbg!("here");
         for index in self.indexes {
             let uuid = Uuid::new_v4();
             uuid_store.insert(index.uid.clone(), uuid)?;
             let src = src.as_ref().join(index.uid);
-            load_index(&src, &dst, uuid, index.meta.primary_key.as_deref(), size)?;
+            load_index(
+                &src,
+                &dst,
+                uuid,
+                index.meta.primary_key.as_deref(),
+                size,
+                indexer_options,
+            )?;
         }
 
         Ok(())
     }
 }
 
-//This is the settings used in the last version of meilisearch exporting dump in V1
+// These are the settings used in legacy meilisearch (<v0.21.0).
 #[derive(Default, Clone, Serialize, Deserialize, Debug)]
 #[serde(rename_all = "camelCase", deny_unknown_fields)]
 struct Settings {
@@ -87,10 +90,11 @@ fn load_index(
     uuid: Uuid,
     primary_key: Option<&str>,
     size: usize,
+    indexer_options: &IndexerOpts,
 ) -> anyhow::Result<()> {
     let index_path = dst.as_ref().join(&format!("indexes/index-{}", uuid));
 
-    std::fs::create_dir_all(&index_path)?;
+    create_dir_all(&index_path)?;
     let mut options = EnvOpenOptions::new();
     options.map_size(size);
     let index = milli::Index::new(options, index_path)?;
@@ -99,31 +103,37 @@ fn load_index(
     // extract `settings.json` file and import content
     let settings = import_settings(&src)?;
     let settings: index_controller::Settings<Unchecked> = settings.into();
-    let update_builder = UpdateBuilder::new(0);
-    index.update_settings(&settings.check(), update_builder)?;
 
-    let update_builder = UpdateBuilder::new(0);
+    let mut txn = index.write_txn()?;
+
+    let handler = UpdateHandler::new(&indexer_options)?;
+
+    index.update_settings_txn(&mut txn, &settings.check(), handler.update_builder(0))?;
+
     let file = File::open(&src.as_ref().join("documents.jsonl"))?;
     let mut reader = std::io::BufReader::new(file);
     reader.fill_buf()?;
     if !reader.buffer().is_empty() {
-        index.update_documents(
+        index.update_documents_txn(
+            &mut txn,
             UpdateFormat::JsonStream,
             IndexDocumentsMethod::ReplaceDocuments,
             Some(reader),
-            update_builder,
+            handler.update_builder(0),
             primary_key,
         )?;
     }
 
-    // the last step: we extract the original milli::Index and close it
+    txn.commit()?;
+
+    // Finaly, we extract the original milli::Index and close it
     Arc::try_unwrap(index.0)
-        .map_err(|_e| "[dumps] At this point no one is supposed to have a reference on the index")
+        .map_err(|_e| "Couln't close index properly")
         .unwrap()
         .prepare_for_closing()
         .wait();
 
-    // Ignore updates in v1.
+    // Updates are ignored in dumps V1.
 
     Ok(())
 }
@@ -172,7 +182,7 @@ impl From<Settings> for index_controller::Settings<Unchecked> {
 
 /// Extract Settings from `settings.json` file present at provided `dir_path`
 fn import_settings(dir_path: impl AsRef<Path>) -> anyhow::Result<Settings> {
-    let path = dbg!(dir_path.as_ref().join("settings.json"));
+    let path = dir_path.as_ref().join("settings.json");
     let file = File::open(path)?;
     let reader = std::io::BufReader::new(file);
     let metadata = serde_json::from_reader(reader)?;

meilisearch-http/src/index_controller/dump_actor/loaders/v2.rs

@@ -4,23 +4,21 @@ use chrono::{DateTime, Utc};
 use log::info;
 use serde::{Deserialize, Serialize};
 
-use crate::{
-    index::Index,
-    index_controller::{update_actor::UpdateStore, uuid_resolver::HeedUuidStore},
-    option::IndexerOpts,
-};
+use crate::index::Index;
+use crate::index_controller::{update_actor::UpdateStore, uuid_resolver::HeedUuidStore};
+use crate::option::IndexerOpts;
 
 #[derive(Serialize, Deserialize, Debug)]
 #[serde(rename_all = "camelCase")]
 pub struct MetadataV2 {
     db_version: String,
-    index_db_size: u64,
-    update_db_size: u64,
+    index_db_size: usize,
+    update_db_size: usize,
     dump_date: DateTime<Utc>,
 }
 
 impl MetadataV2 {
-    pub fn new(index_db_size: u64, update_db_size: u64) -> Self {
+    pub fn new(index_db_size: usize, update_db_size: usize) -> Self {
         Self {
             db_version: env!("CARGO_PKG_VERSION").to_string(),
             index_db_size,
@@ -33,9 +31,8 @@ impl MetadataV2 {
         self,
         src: impl AsRef<Path>,
         dst: impl AsRef<Path>,
-        // TODO: use these variable to test if loading the index is possible.
-        _index_db_size: u64,
-        _update_db_size: u64,
+        index_db_size: usize,
+        update_db_size: usize,
         indexing_options: &IndexerOpts,
     ) -> anyhow::Result<()> {
         info!(
@@ -47,14 +44,14 @@ impl MetadataV2 {
         HeedUuidStore::load_dump(src.as_ref(), &dst)?;
 
         info!("Loading updates.");
-        UpdateStore::load_dump(&src, &dst, self.update_db_size)?;
+        UpdateStore::load_dump(&src, &dst, update_db_size)?;
 
-        info!("Loading indexes");
+        info!("Loading indexes.");
         let indexes_path = src.as_ref().join("indexes");
         let indexes = indexes_path.read_dir()?;
         for index in indexes {
             let index = index?;
-            Index::load_dump(&index.path(), &dst, self.index_db_size, indexing_options)?;
+            Index::load_dump(&index.path(), &dst, index_db_size, indexing_options)?;
         }
 
         Ok(())

meilisearch-http/src/index_controller/dump_actor/mod.rs

@@ -8,6 +8,7 @@ use log::{error, info, warn};
 use mockall::automock;
 use serde::{Deserialize, Serialize};
 use thiserror::Error;
+use tokio::fs::create_dir_all;
 
 use loaders::v1::MetadataV1;
 use loaders::v2::MetadataV2;
@@ -15,7 +16,6 @@ use loaders::v2::MetadataV2;
 pub use actor::DumpActor;
 pub use handle_impl::*;
 pub use message::DumpMsg;
-use tokio::fs::create_dir_all;
 
 use super::{update_actor::UpdateActorHandle, uuid_resolver::UuidResolverHandle};
 use crate::{helpers::compression, option::IndexerOpts};
@@ -61,7 +61,7 @@ pub enum Metadata {
 }
 
 impl Metadata {
-    pub fn new_v2(index_db_size: u64, update_db_size: u64) -> Self {
+    pub fn new_v2(index_db_size: usize, update_db_size: usize) -> Self {
         let meta = MetadataV2::new(index_db_size, update_db_size);
         Self::V2(meta)
     }
@@ -117,8 +117,8 @@ impl DumpInfo {
 pub fn load_dump(
     dst_path: impl AsRef<Path>,
     src_path: impl AsRef<Path>,
-    index_db_size: u64,
-    update_db_size: u64,
+    index_db_size: usize,
+    update_db_size: usize,
     indexer_opts: &IndexerOpts,
 ) -> anyhow::Result<()> {
     let tmp_src = tempfile::tempdir_in(".")?;
@@ -139,7 +139,7 @@ pub fn load_dump(
 
     match meta {
         Metadata::V1(meta) => {
-            meta.load_dump(&tmp_src_path, tmp_dst.path(), index_db_size as usize)?
+            meta.load_dump(&tmp_src_path, tmp_dst.path(), index_db_size, indexer_opts)?
         }
         Metadata::V2(meta) => meta.load_dump(
             &tmp_src_path,
@@ -166,8 +166,8 @@ struct DumpTask<U, P> {
     uuid_resolver: U,
     update_handle: P,
     uid: String,
-    update_db_size: u64,
-    index_db_size: u64,
+    update_db_size: usize,
+    index_db_size: usize,
 }
 
 impl<U, P> DumpTask<U, P>