Fix test again

.github/scripts/check-release.sh

@@ -1,41 +1,24 @@
-#!/usr/bin/env bash
-set -eu -o pipefail
+#!/bin/bash
 
-check_tag() {
-    local expected=$1
-    local actual=$2
-    local filename=$3
-
-    if [[ $actual != $expected ]]; then
-        echo >&2 "Error: the current tag does not match the version in $filename: found $actual, expected $expected"
-        return 1
-    fi
+# check_tag $current_tag $file_tag $file_name
+function check_tag {
+  if [[ "$1" != "$2" ]]; then
+      echo "Error: the current tag does not match the version in Cargo.toml: found $2 - expected $1"
+      ret=1
+  fi
 }
 
-read_version() {
-    grep '^version = ' | cut -d \" -f 2
-}
-
-if [[ -z "${GITHUB_REF:-}" ]]; then
-    echo >&2 "Error: GITHUB_REF is not set"
-    exit 1
-fi
-
-if [[ ! "$GITHUB_REF" =~ ^refs/tags/v[0-9]+\.[0-9]+\.[0-9]+(-[a-z0-9]+)?$ ]]; then
-    echo >&2 "Error: GITHUB_REF is not a valid tag: $GITHUB_REF"
-    exit 1
-fi
-
-current_tag=${GITHUB_REF#refs/tags/v}
 ret=0
+current_tag=${GITHUB_REF#'refs/tags/v'}
 
-toml_tag="$(cat Cargo.toml | read_version)"
-check_tag "$current_tag" "$toml_tag" Cargo.toml || ret=1
+file_tag="$(grep '^version = ' Cargo.toml | cut -d '=' -f 2 | tr -d '"' | tr -d ' ')"
+check_tag $current_tag $file_tag
 
-lock_tag=$(grep -A 1 '^name = "meilisearch-auth"' Cargo.lock | read_version)
-check_tag "$current_tag" "$lock_tag" Cargo.lock || ret=1
+lock_file='Cargo.lock'
+lock_tag=$(grep -A 1 'name = "meilisearch-auth"' $lock_file | grep version | cut -d '=' -f 2 | tr -d '"' | tr -d ' ')
+check_tag $current_tag $lock_tag $lock_file
 
-if (( ret == 0 )); then
-    echo 'OK'
+if [[ "$ret" -eq 0 ]] ; then
+  echo 'OK'
 fi
 exit $ret

.github/workflows/fuzzer-indexing.yml

@@ -1,24 +0,0 @@
-name: Run the indexing fuzzer
-
-on:
-  push:
-    branches:
-      - main
-
-jobs:
-  fuzz:
-    name: Setup the action
-    runs-on: ubuntu-latest
-    timeout-minutes: 4320 # 72h
-    steps:
-      - uses: actions/checkout@v3
-      - uses: actions-rs/toolchain@v1
-        with:
-          profile: minimal
-          toolchain: stable
-          override: true
-
-      # Run benchmarks
-      - name: Run the fuzzer
-        run: |
-          cargo run --release --bin fuzz-indexing

.github/workflows/sdks-tests.yml

@@ -16,28 +16,13 @@ env:
   MEILI_NO_ANALYTICS: 'true'
 
 jobs:
-  define-docker-image:
-    runs-on: ubuntu-latest
-    outputs:
-      docker-image: ${{ steps.define-image.outputs.docker-image }}
-    steps:
-      - uses: actions/checkout@v3
-      - name: Define the Docker image we need to use
-        id: define-image
-        run: |
-          event=${{ github.event_name }}
-          echo "docker-image=nightly" >> $GITHUB_OUTPUT
-          if [[ $event == 'workflow_dispatch' ]]; then
-            echo "docker-image=${{ github.event.inputs.docker_image }}" >> $GITHUB_OUTPUT
-          fi
 
   meilisearch-js-tests:
-    needs: define-docker-image
     name: JS SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -67,12 +52,11 @@ jobs:
         run: yarn test:env:browser
 
   instant-meilisearch-tests:
-    needs: define-docker-image
     name: instant-meilisearch tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -94,12 +78,11 @@ jobs:
         run: yarn build
 
   meilisearch-php-tests:
-    needs: define-docker-image
     name: PHP SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -125,12 +108,11 @@ jobs:
           composer remove --dev guzzlehttp/guzzle http-interop/http-factory-guzzle
 
   meilisearch-python-tests:
-    needs: define-docker-image
     name: Python SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -150,12 +132,11 @@ jobs:
         run: pipenv run pytest
 
   meilisearch-go-tests:
-    needs: define-docker-image
     name: Go SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -180,12 +161,11 @@ jobs:
         run: go test -v ./...
 
   meilisearch-ruby-tests:
-    needs: define-docker-image
     name: Ruby SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -205,12 +185,11 @@ jobs:
         run: bundle exec rspec
 
   meilisearch-rust-tests:
-    needs: define-docker-image
     name: Rust SDK tests
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch:${{ github.event.inputs.docker_image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}

Cargo.toml

@@ -10,12 +10,10 @@ members = [
     "file-store",
     "permissive-json-pointer",
     "milli",
-    "index-stats",
     "filter-parser",
     "flatten-serde-json",
     "json-depth-checker",
-    "benchmarks",
-    "fuzzers",
+    "benchmarks"
 ]
 
 [workspace.package]

fuzzers/Cargo.toml

@@ -1,20 +0,0 @@
-[package]
-name = "fuzzers"
-publish = false
-
-version.workspace = true
-authors.workspace = true
-description.workspace = true
-homepage.workspace = true
-readme.workspace = true
-edition.workspace = true
-license.workspace = true
-
-[dependencies]
-arbitrary = { version = "1.3.0", features = ["derive"] }
-clap = { version = "4.3.0", features = ["derive"] }
-fastrand = "1.9.0"
-milli = { path = "../milli" }
-serde = { version = "1.0.160", features = ["derive"] }
-serde_json = { version = "1.0.95", features = ["preserve_order"] }
-tempfile = "3.5.0"

fuzzers/README.md

@@ -1,3 +0,0 @@
-# Fuzzers
-
-The purpose of this crate is to contains all the handmade "fuzzer" we may need.

fuzzers/src/bin/fuzz-indexing.rs

@@ -1,152 +0,0 @@
-use std::num::NonZeroUsize;
-use std::path::PathBuf;
-use std::sync::atomic::{AtomicBool, AtomicUsize, Ordering};
-use std::time::Duration;
-
-use arbitrary::{Arbitrary, Unstructured};
-use clap::Parser;
-use fuzzers::Operation;
-use milli::heed::EnvOpenOptions;
-use milli::update::{IndexDocuments, IndexDocumentsConfig, IndexerConfig};
-use milli::Index;
-use tempfile::TempDir;
-
-#[derive(Debug, Arbitrary)]
-struct Batch([Operation; 5]);
-
-#[derive(Debug, Clone, Parser)]
-struct Opt {
-    /// The number of fuzzer to run in parallel.
-    #[clap(long)]
-    par: Option<NonZeroUsize>,
-    // We need to put a lot of newlines in the following documentation or else everything gets collapsed on one line
-    /// The path in which the databases will be created.
-    /// Using a ramdisk is recommended.
-    ///
-    /// Linux:
-    ///
-    /// sudo mount -t tmpfs -o size=2g tmpfs ramdisk # to create it
-    ///
-    /// sudo umount ramdisk # to remove it
-    ///
-    /// MacOS:
-    ///
-    /// diskutil erasevolume HFS+ 'RAM Disk' `hdiutil attach -nobrowse -nomount ram://4194304 # create it
-    ///
-    /// hdiutil detach /dev/:the_disk
-    #[clap(long)]
-    path: Option<PathBuf>,
-}
-
-fn main() {
-    let opt = Opt::parse();
-    let progression: &'static AtomicUsize = Box::leak(Box::new(AtomicUsize::new(0)));
-    let stop: &'static AtomicBool = Box::leak(Box::new(AtomicBool::new(false)));
-
-    let par = opt.par.unwrap_or_else(|| std::thread::available_parallelism().unwrap()).get();
-    let mut handles = Vec::with_capacity(par);
-
-    for _ in 0..par {
-        let opt = opt.clone();
-
-        let handle = std::thread::spawn(move || {
-            let mut options = EnvOpenOptions::new();
-            options.map_size(1024 * 1024 * 1024 * 1024);
-            let tempdir = match opt.path {
-                Some(path) => TempDir::new_in(path).unwrap(),
-                None => TempDir::new().unwrap(),
-            };
-            let index = Index::new(options, tempdir.path()).unwrap();
-            let indexer_config = IndexerConfig::default();
-            let index_documents_config = IndexDocumentsConfig::default();
-
-            std::thread::scope(|s| {
-                loop {
-                    if stop.load(Ordering::Relaxed) {
-                        return;
-                    }
-                    let v: Vec<u8> =
-                        std::iter::repeat_with(|| fastrand::u8(..)).take(1000).collect();
-
-                    let mut data = Unstructured::new(&v);
-                    let batches = <[Batch; 5]>::arbitrary(&mut data).unwrap();
-                    // will be used to display the error once a thread crashes
-                    let dbg_input = format!("{:#?}", batches);
-
-                    let handle = s.spawn(|| {
-                        let mut wtxn = index.write_txn().unwrap();
-
-                        for batch in batches {
-                            let mut builder = IndexDocuments::new(
-                                &mut wtxn,
-                                &index,
-                                &indexer_config,
-                                index_documents_config.clone(),
-                                |_| (),
-                                || false,
-                            )
-                            .unwrap();
-
-                            for op in batch.0 {
-                                match op {
-                                    Operation::AddDoc(doc) => {
-                                        let documents =
-                                            milli::documents::objects_from_json_value(doc.to_d());
-                                        let documents =
-                                            milli::documents::documents_batch_reader_from_objects(
-                                                documents,
-                                            );
-                                        let (b, _added) = builder.add_documents(documents).unwrap();
-                                        builder = b;
-                                    }
-                                    Operation::DeleteDoc(id) => {
-                                        let (b, _removed) =
-                                            builder.remove_documents(vec![id.to_s()]).unwrap();
-                                        builder = b;
-                                    }
-                                }
-                            }
-                            builder.execute().unwrap();
-
-                            // after executing a batch we check if the database is corrupted
-                            let res = index.search(&wtxn).execute().unwrap();
-                            index.documents(&wtxn, res.documents_ids).unwrap();
-                            progression.fetch_add(1, Ordering::Relaxed);
-                        }
-                        wtxn.abort().unwrap();
-                    });
-                    if let err @ Err(_) = handle.join() {
-                        stop.store(true, Ordering::Relaxed);
-                        err.expect(&dbg_input);
-                    }
-                }
-            });
-        });
-        handles.push(handle);
-    }
-
-    std::thread::spawn(|| {
-        let mut last_value = 0;
-        let start = std::time::Instant::now();
-        loop {
-            let total = progression.load(Ordering::Relaxed);
-            let elapsed = start.elapsed().as_secs();
-            if elapsed > 3600 {
-                // after 1 hour, stop the fuzzer, success
-                std::process::exit(0);
-            }
-            println!(
-                "Has been running for {:?} seconds. Tested {} new values for a total of {}.",
-                elapsed,
-                total - last_value,
-                total
-            );
-            last_value = total;
-            std::thread::sleep(Duration::from_secs(1));
-        }
-    });
-
-    for handle in handles {
-        handle.join().unwrap();
-    }
-}

fuzzers/src/lib.rs

@@ -1,46 +0,0 @@
-use arbitrary::Arbitrary;
-use serde_json::{json, Value};
-
-#[derive(Debug, Arbitrary)]
-pub enum Document {
-    One,
-    Two,
-    Three,
-    Four,
-    Five,
-    Six,
-}
-
-impl Document {
-    pub fn to_d(&self) -> Value {
-        match self {
-            Document::One => json!({ "id": 0, "doggo": "bernese" }),
-            Document::Two => json!({ "id": 0, "doggo": "golden" }),
-            Document::Three => json!({ "id": 0, "catto": "jorts" }),
-            Document::Four => json!({ "id": 1, "doggo": "bernese" }),
-            Document::Five => json!({ "id": 1, "doggo": "golden" }),
-            Document::Six => json!({ "id": 1, "catto": "jorts" }),
-        }
-    }
-}
-
-#[derive(Debug, Arbitrary)]
-pub enum DocId {
-    Zero,
-    One,
-}
-
-impl DocId {
-    pub fn to_s(&self) -> String {
-        match self {
-            DocId::Zero => "0".to_string(),
-            DocId::One => "1".to_string(),
-        }
-    }
-}
-
-#[derive(Debug, Arbitrary)]
-pub enum Operation {
-    AddDoc(Document),
-    DeleteDoc(DocId),
-}

index-scheduler/src/autobatcher.rs

@@ -160,7 +160,7 @@ impl BatchKind {
 impl BatchKind {
     /// Returns a `ControlFlow::Break` if you must stop right now.
     /// The boolean tell you if an index has been created by the batched task.
-    /// To ease the writing of the code. `true` can be returned when you don't need to create an index
+    /// To ease the writting of the code. `true` can be returned when you don't need to create an index
     /// but false can't be returned if you needs to create an index.
     // TODO use an AutoBatchKind as input
     pub fn new(
@@ -214,7 +214,7 @@ impl BatchKind {
 
     /// Returns a `ControlFlow::Break` if you must stop right now.
     /// The boolean tell you if an index has been created by the batched task.
-    /// To ease the writing of the code. `true` can be returned when you don't need to create an index
+    /// To ease the writting of the code. `true` can be returned when you don't need to create an index
     /// but false can't be returned if you needs to create an index.
     #[rustfmt::skip]
     fn accumulate(self, id: TaskId, kind: AutobatchKind, index_already_exists: bool, primary_key: Option<&str>) -> ControlFlow<BatchKind, BatchKind> {
@@ -321,18 +321,9 @@ impl BatchKind {
                 })
             }
             (
-                BatchKind::DocumentOperation { method, allow_index_creation, primary_key, mut operation_ids },
+                this @ BatchKind::DocumentOperation { .. },
                 K::DocumentDeletion,
-            ) => {
-                operation_ids.push(id);
-
-                Continue(BatchKind::DocumentOperation {
-                    method,
-                    allow_index_creation,
-                    primary_key,
-                    operation_ids,
-                })
-            }
+            ) => Break(this),
             // but we can't autobatch documents if it's not the same kind
             // this match branch MUST be AFTER the previous one
             (
@@ -355,35 +346,7 @@ impl BatchKind {
                 deletion_ids.push(id);
                 Continue(BatchKind::DocumentClear { ids: deletion_ids })
             }
-            // we can autobatch the deletion and import if the index already exists
-            (
-                BatchKind::DocumentDeletion { mut deletion_ids },
-                K::DocumentImport { method, allow_index_creation, primary_key }
-            ) if index_already_exists => {
-                deletion_ids.push(id);
-
-                Continue(BatchKind::DocumentOperation {
-                    method,
-                    allow_index_creation,
-                    primary_key,
-                    operation_ids: deletion_ids,
-                })
-            }
-            // we can autobatch the deletion and import if both can't create an index
-            (
-                BatchKind::DocumentDeletion { mut deletion_ids },
-                K::DocumentImport { method, allow_index_creation, primary_key }
-            ) if !allow_index_creation => {
-                deletion_ids.push(id);
-
-                Continue(BatchKind::DocumentOperation {
-                    method,
-                    allow_index_creation,
-                    primary_key,
-                    operation_ids: deletion_ids,
-                })
-            }
-            // we can't autobatch a deletion and an import if the index does not exists but would be created by an addition
+            // we can't autobatch a deletion and an import
             (
                 this @ BatchKind::DocumentDeletion { .. },
                 K::DocumentImport { .. }
@@ -685,36 +648,36 @@ mod tests {
         debug_snapshot!(autobatch_from(false,None,  [settings(false)]), @"Some((Settings { allow_index_creation: false, settings_ids: [0] }, false))");
         debug_snapshot!(autobatch_from(false,None,  [settings(false), settings(false), settings(false)]), @"Some((Settings { allow_index_creation: false, settings_ids: [0, 1, 2] }, false))");
 
-        // We can autobatch document addition with document deletion
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, true))");
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, true))");
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, true))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, true))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, true))");
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, true))");
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, true))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, true))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        // And the other way around
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, true, None)]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, true, None)]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, false, None)]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, false, None)]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, true, Some("catto"))]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, true, Some("catto"))]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, false, Some("catto"))]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, false, Some("catto"))]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(ReplaceDocuments, false, None)]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(UpdateDocuments, false, None)]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0, 1] }, false))");
-        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(ReplaceDocuments, false, Some("catto"))]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
-        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(UpdateDocuments, false, Some("catto"))]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0, 1] }, false))"###);
+        // We can't autobatch document addition with document deletion
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0] }, true))");
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0] }, true))");
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0] }, true))"###);
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0] }, true))"###);
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(ReplaceDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0] }, false))"###);
+        debug_snapshot!(autobatch_from(true, None, [doc_imp(UpdateDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0] }, false))"###);
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0] }, true))");
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, true, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: None, operation_ids: [0] }, true))");
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, false, None), doc_del()]), @"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: None, operation_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0] }, true))"###);
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, true, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: true, primary_key: Some("catto"), operation_ids: [0] }, true))"###);
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(ReplaceDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: ReplaceDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0] }, false))"###);
+        debug_snapshot!(autobatch_from(false, None, [doc_imp(UpdateDocuments, false, Some("catto")), doc_del()]), @r###"Some((DocumentOperation { method: UpdateDocuments, allow_index_creation: false, primary_key: Some("catto"), operation_ids: [0] }, false))"###);
+        // we also can't do the only way around
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, true, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, true, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, false, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, false, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, true, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, true, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(ReplaceDocuments, false, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(true, None, [doc_del(), doc_imp(UpdateDocuments, false, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(ReplaceDocuments, false, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(UpdateDocuments, false, None)]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(ReplaceDocuments, false, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
+        debug_snapshot!(autobatch_from(false, None, [doc_del(), doc_imp(UpdateDocuments, false, Some("catto"))]), @"Some((DocumentDeletion { deletion_ids: [0] }, false))");
     }
 
     #[test]

index-scheduler/src/batch.rs

@@ -998,7 +998,7 @@ impl IndexScheduler {
                 }()
                 .unwrap_or_default();
 
-                // The write transaction is directly owned and committed inside.
+                // The write transaction is directly owned and commited inside.
                 match self.index_mapper.delete_index(wtxn, &index_uid) {
                     Ok(()) => (),
                     Err(Error::IndexNotFound(_)) if index_has_been_created => (),

index-scheduler/src/lib.rs

@@ -1785,7 +1785,7 @@ mod tests {
             assert_eq!(task.kind.as_kind(), k);
         }
 
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "everything_is_successfully_registered");
+        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "everything_is_succesfully_registered");
     }
 
     #[test]
@@ -2075,105 +2075,6 @@ mod tests {
         snapshot!(snapshot_index_scheduler(&index_scheduler), name: "both_task_succeeded");
     }
 
-    #[test]
-    fn document_addition_and_document_deletion() {
-        let (index_scheduler, mut handle) = IndexScheduler::test(true, vec![]);
-
-        let content = r#"[
-            { "id": 1, "doggo": "jean bob" },
-            { "id": 2, "catto": "jorts" },
-            { "id": 3, "doggo": "bork" }
-        ]"#;
-
-        let (uuid, mut file) = index_scheduler.create_update_file_with_uuid(0).unwrap();
-        let documents_count = read_json(content.as_bytes(), file.as_file_mut()).unwrap();
-        file.persist().unwrap();
-        index_scheduler
-            .register(KindWithContent::DocumentAdditionOrUpdate {
-                index_uid: S("doggos"),
-                primary_key: Some(S("id")),
-                method: ReplaceDocuments,
-                content_file: uuid,
-                documents_count,
-                allow_index_creation: true,
-            })
-            .unwrap();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "registered_the_first_task");
-        index_scheduler
-            .register(KindWithContent::DocumentDeletion {
-                index_uid: S("doggos"),
-                documents_ids: vec![S("1"), S("2")],
-            })
-            .unwrap();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "registered_the_second_task");
-
-        handle.advance_one_successful_batch(); // The addition AND deletion should've been batched together
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "after_processing_the_batch");
-
-        let index = index_scheduler.index("doggos").unwrap();
-        let rtxn = index.read_txn().unwrap();
-        let field_ids_map = index.fields_ids_map(&rtxn).unwrap();
-        let field_ids = field_ids_map.ids().collect::<Vec<_>>();
-        let documents = index
-            .all_documents(&rtxn)
-            .unwrap()
-            .map(|ret| obkv_to_json(&field_ids, &field_ids_map, ret.unwrap().1).unwrap())
-            .collect::<Vec<_>>();
-        snapshot!(serde_json::to_string_pretty(&documents).unwrap(), name: "documents");
-    }
-
-    #[test]
-    fn document_deletion_and_document_addition() {
-        let (index_scheduler, mut handle) = IndexScheduler::test(true, vec![]);
-        index_scheduler
-            .register(KindWithContent::DocumentDeletion {
-                index_uid: S("doggos"),
-                documents_ids: vec![S("1"), S("2")],
-            })
-            .unwrap();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "registered_the_first_task");
-
-        let content = r#"[
-            { "id": 1, "doggo": "jean bob" },
-            { "id": 2, "catto": "jorts" },
-            { "id": 3, "doggo": "bork" }
-        ]"#;
-
-        let (uuid, mut file) = index_scheduler.create_update_file_with_uuid(0).unwrap();
-        let documents_count = read_json(content.as_bytes(), file.as_file_mut()).unwrap();
-        file.persist().unwrap();
-        index_scheduler
-            .register(KindWithContent::DocumentAdditionOrUpdate {
-                index_uid: S("doggos"),
-                primary_key: Some(S("id")),
-                method: ReplaceDocuments,
-                content_file: uuid,
-                documents_count,
-                allow_index_creation: true,
-            })
-            .unwrap();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "registered_the_second_task");
-
-        // The deletion should have failed because it can't create an index
-        handle.advance_one_failed_batch();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "after_failing_the_deletion");
-
-        // The addition should works
-        handle.advance_one_successful_batch();
-        snapshot!(snapshot_index_scheduler(&index_scheduler), name: "after_last_successful_addition");
-
-        let index = index_scheduler.index("doggos").unwrap();
-        let rtxn = index.read_txn().unwrap();
-        let field_ids_map = index.fields_ids_map(&rtxn).unwrap();
-        let field_ids = field_ids_map.ids().collect::<Vec<_>>();
-        let documents = index
-            .all_documents(&rtxn)
-            .unwrap()
-            .map(|ret| obkv_to_json(&field_ids, &field_ids_map, ret.unwrap().1).unwrap())
-            .collect::<Vec<_>>();
-        snapshot!(serde_json::to_string_pretty(&documents).unwrap(), name: "documents");
-    }
-
     #[test]
     fn do_not_batch_task_of_different_indexes() {
         let (index_scheduler, mut handle) = IndexScheduler::test(true, vec![]);

index-scheduler/src/snapshots/lib.rs/document_addition_and_document_deletion/after_processing_the_batch.snap

@@ -1,43 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-1 {uid: 1, status: succeeded, details: { received_document_ids: 2, deleted_documents: Some(2) }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-----------------------------------------------------------------------
-### Status:
-enqueued []
-succeeded [0,1,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [0,]
-"documentDeletion" [1,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,1,]
-----------------------------------------------------------------------
-### Index Mapper:
-doggos: { number_of_documents: 1, field_distribution: {"doggo": 1, "id": 1} }
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Started At:
-[timestamp] [0,1,]
-----------------------------------------------------------------------
-### Finished At:
-[timestamp] [0,1,]
-----------------------------------------------------------------------
-### File Store:
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_addition_and_document_deletion/documents.snap

@@ -1,9 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-[
-  {
-    "id": 3,
-    "doggo": "bork"
-  }
-]

index-scheduler/src/snapshots/lib.rs/document_addition_and_document_deletion/registered_the_first_task.snap

@@ -1,37 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-----------------------------------------------------------------------
-### Status:
-enqueued [0,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [0,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,]
-----------------------------------------------------------------------
-### Index Mapper:
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-----------------------------------------------------------------------
-### Started At:
-----------------------------------------------------------------------
-### Finished At:
-----------------------------------------------------------------------
-### File Store:
-00000000-0000-0000-0000-000000000000
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_addition_and_document_deletion/registered_the_second_task.snap

@@ -1,40 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-1 {uid: 1, status: enqueued, details: { received_document_ids: 2, deleted_documents: None }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-----------------------------------------------------------------------
-### Status:
-enqueued [0,1,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [0,]
-"documentDeletion" [1,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,1,]
-----------------------------------------------------------------------
-### Index Mapper:
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Started At:
-----------------------------------------------------------------------
-### Finished At:
-----------------------------------------------------------------------
-### File Store:
-00000000-0000-0000-0000-000000000000
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_deletion_and_document_addition/after_failing_the_deletion.snap

@@ -1,43 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: failed, error: ResponseError { code: 200, message: "Index `doggos` not found.", error_code: "index_not_found", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_not_found" }, details: { received_document_ids: 2, deleted_documents: Some(0) }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-1 {uid: 1, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-----------------------------------------------------------------------
-### Status:
-enqueued [1,]
-failed [0,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [1,]
-"documentDeletion" [0,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,1,]
-----------------------------------------------------------------------
-### Index Mapper:
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Started At:
-[timestamp] [0,]
-----------------------------------------------------------------------
-### Finished At:
-[timestamp] [0,]
-----------------------------------------------------------------------
-### File Store:
-00000000-0000-0000-0000-000000000000
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_deletion_and_document_addition/after_last_successful_addition.snap

@@ -1,46 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: failed, error: ResponseError { code: 200, message: "Index `doggos` not found.", error_code: "index_not_found", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_not_found" }, details: { received_document_ids: 2, deleted_documents: Some(0) }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-1 {uid: 1, status: succeeded, details: { received_documents: 3, indexed_documents: Some(3) }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-----------------------------------------------------------------------
-### Status:
-enqueued []
-succeeded [1,]
-failed [0,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [1,]
-"documentDeletion" [0,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,1,]
-----------------------------------------------------------------------
-### Index Mapper:
-doggos: { number_of_documents: 3, field_distribution: {"catto": 1, "doggo": 2, "id": 3} }
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Started At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Finished At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### File Store:
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_deletion_and_document_addition/documents.snap

@@ -1,17 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-[
-  {
-    "id": 1,
-    "doggo": "jean bob"
-  },
-  {
-    "id": 2,
-    "catto": "jorts"
-  },
-  {
-    "id": 3,
-    "doggo": "bork"
-  }
-]

index-scheduler/src/snapshots/lib.rs/document_deletion_and_document_addition/registered_the_first_task.snap

@@ -1,36 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: enqueued, details: { received_document_ids: 2, deleted_documents: None }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-----------------------------------------------------------------------
-### Status:
-enqueued [0,]
-----------------------------------------------------------------------
-### Kind:
-"documentDeletion" [0,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,]
-----------------------------------------------------------------------
-### Index Mapper:
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-----------------------------------------------------------------------
-### Started At:
-----------------------------------------------------------------------
-### Finished At:
-----------------------------------------------------------------------
-### File Store:
-
-----------------------------------------------------------------------
-

index-scheduler/src/snapshots/lib.rs/document_deletion_and_document_addition/registered_the_second_task.snap

@@ -1,40 +0,0 @@
----
-source: index-scheduler/src/lib.rs
----
-### Autobatching Enabled = true
-### Processing Tasks:
-[]
-----------------------------------------------------------------------
-### All Tasks:
-0 {uid: 0, status: enqueued, details: { received_document_ids: 2, deleted_documents: None }, kind: DocumentDeletion { index_uid: "doggos", documents_ids: ["1", "2"] }}
-1 {uid: 1, status: enqueued, details: { received_documents: 3, indexed_documents: None }, kind: DocumentAdditionOrUpdate { index_uid: "doggos", primary_key: Some("id"), method: ReplaceDocuments, content_file: 00000000-0000-0000-0000-000000000000, documents_count: 3, allow_index_creation: true }}
-----------------------------------------------------------------------
-### Status:
-enqueued [0,1,]
-----------------------------------------------------------------------
-### Kind:
-"documentAdditionOrUpdate" [1,]
-"documentDeletion" [0,]
-----------------------------------------------------------------------
-### Index Tasks:
-doggos [0,1,]
-----------------------------------------------------------------------
-### Index Mapper:
-
-----------------------------------------------------------------------
-### Canceled By:
-
-----------------------------------------------------------------------
-### Enqueued At:
-[timestamp] [0,]
-[timestamp] [1,]
-----------------------------------------------------------------------
-### Started At:
-----------------------------------------------------------------------
-### Finished At:
-----------------------------------------------------------------------
-### File Store:
-00000000-0000-0000-0000-000000000000
-
-----------------------------------------------------------------------
-

index-stats/Cargo.toml

@@ -1,12 +0,0 @@
-[package]
-name = "index-stats"
-description = "A small program that computes internal stats of a Meilisearch index"
-version = "0.1.0"
-edition = "2021"
-publish = false
-
-[dependencies]
-anyhow = "1.0.71"
-clap = { version = "4.3.5", features = ["derive"] }
-milli = { path = "../milli" }
-piechart = "1.0.0"

index-stats/src/main.rs

@@ -1,224 +0,0 @@
-use std::cmp::Reverse;
-use std::path::PathBuf;
-
-use clap::Parser;
-use milli::heed::{types::ByteSlice, EnvOpenOptions, PolyDatabase, RoTxn};
-use milli::index::db_name::*;
-use milli::index::Index;
-use piechart::{Chart, Color, Data};
-
-/// Simple program to greet a person
-#[derive(Parser, Debug)]
-#[command(author, version, about, long_about = None)]
-struct Args {
-    /// The path to the LMDB Meilisearch index database.
-    path: PathBuf,
-
-    /// The radius of the graphs
-    #[clap(long, default_value_t = 10)]
-    graph_radius: u16,
-
-    /// The radius of the graphs
-    #[clap(long, default_value_t = 6)]
-    graph_aspect_ratio: u16,
-}
-
-fn main() -> anyhow::Result<()> {
-    let Args { path, graph_radius, graph_aspect_ratio } = Args::parse();
-    let env = EnvOpenOptions::new().max_dbs(24).open(path)?;
-
-    // TODO not sure to keep that...
-    //      if removed put the pub(crate) back in the Index struct
-    matches!(
-        Option::<Index>::None,
-        Some(Index {
-            env: _,
-            main: _,
-            word_docids: _,
-            exact_word_docids: _,
-            word_prefix_docids: _,
-            exact_word_prefix_docids: _,
-            word_pair_proximity_docids: _,
-            word_prefix_pair_proximity_docids: _,
-            prefix_word_pair_proximity_docids: _,
-            word_position_docids: _,
-            word_fid_docids: _,
-            field_id_word_count_docids: _,
-            word_prefix_position_docids: _,
-            word_prefix_fid_docids: _,
-            script_language_docids: _,
-            facet_id_exists_docids: _,
-            facet_id_is_null_docids: _,
-            facet_id_is_empty_docids: _,
-            facet_id_f64_docids: _,
-            facet_id_string_docids: _,
-            field_id_docid_facet_f64s: _,
-            field_id_docid_facet_strings: _,
-            documents: _,
-        })
-    );
-
-    let mut wtxn = env.write_txn()?;
-    let main = env.create_poly_database(&mut wtxn, Some(MAIN))?;
-    let word_docids = env.create_poly_database(&mut wtxn, Some(WORD_DOCIDS))?;
-    let exact_word_docids = env.create_poly_database(&mut wtxn, Some(EXACT_WORD_DOCIDS))?;
-    let word_prefix_docids = env.create_poly_database(&mut wtxn, Some(WORD_PREFIX_DOCIDS))?;
-    let exact_word_prefix_docids =
-        env.create_poly_database(&mut wtxn, Some(EXACT_WORD_PREFIX_DOCIDS))?;
-    let word_pair_proximity_docids =
-        env.create_poly_database(&mut wtxn, Some(WORD_PAIR_PROXIMITY_DOCIDS))?;
-    let script_language_docids =
-        env.create_poly_database(&mut wtxn, Some(SCRIPT_LANGUAGE_DOCIDS))?;
-    let word_prefix_pair_proximity_docids =
-        env.create_poly_database(&mut wtxn, Some(WORD_PREFIX_PAIR_PROXIMITY_DOCIDS))?;
-    let prefix_word_pair_proximity_docids =
-        env.create_poly_database(&mut wtxn, Some(PREFIX_WORD_PAIR_PROXIMITY_DOCIDS))?;
-    let word_position_docids = env.create_poly_database(&mut wtxn, Some(WORD_POSITION_DOCIDS))?;
-    let word_fid_docids = env.create_poly_database(&mut wtxn, Some(WORD_FIELD_ID_DOCIDS))?;
-    let field_id_word_count_docids =
-        env.create_poly_database(&mut wtxn, Some(FIELD_ID_WORD_COUNT_DOCIDS))?;
-    let word_prefix_position_docids =
-        env.create_poly_database(&mut wtxn, Some(WORD_PREFIX_POSITION_DOCIDS))?;
-    let word_prefix_fid_docids =
-        env.create_poly_database(&mut wtxn, Some(WORD_PREFIX_FIELD_ID_DOCIDS))?;
-    let facet_id_f64_docids = env.create_poly_database(&mut wtxn, Some(FACET_ID_F64_DOCIDS))?;
-    let facet_id_string_docids =
-        env.create_poly_database(&mut wtxn, Some(FACET_ID_STRING_DOCIDS))?;
-    let facet_id_exists_docids =
-        env.create_poly_database(&mut wtxn, Some(FACET_ID_EXISTS_DOCIDS))?;
-    let facet_id_is_null_docids =
-        env.create_poly_database(&mut wtxn, Some(FACET_ID_IS_NULL_DOCIDS))?;
-    let facet_id_is_empty_docids =
-        env.create_poly_database(&mut wtxn, Some(FACET_ID_IS_EMPTY_DOCIDS))?;
-    let field_id_docid_facet_f64s =
-        env.create_poly_database(&mut wtxn, Some(FIELD_ID_DOCID_FACET_F64S))?;
-    let field_id_docid_facet_strings =
-        env.create_poly_database(&mut wtxn, Some(FIELD_ID_DOCID_FACET_STRINGS))?;
-    let documents = env.create_poly_database(&mut wtxn, Some(DOCUMENTS))?;
-    wtxn.commit()?;
-
-    let list = [
-        (main, MAIN),
-        (word_docids, WORD_DOCIDS),
-        (exact_word_docids, EXACT_WORD_DOCIDS),
-        (word_prefix_docids, WORD_PREFIX_DOCIDS),
-        (exact_word_prefix_docids, EXACT_WORD_PREFIX_DOCIDS),
-        (word_pair_proximity_docids, WORD_PAIR_PROXIMITY_DOCIDS),
-        (script_language_docids, SCRIPT_LANGUAGE_DOCIDS),
-        (word_prefix_pair_proximity_docids, WORD_PREFIX_PAIR_PROXIMITY_DOCIDS),
-        (prefix_word_pair_proximity_docids, PREFIX_WORD_PAIR_PROXIMITY_DOCIDS),
-        (word_position_docids, WORD_POSITION_DOCIDS),
-        (word_fid_docids, WORD_FIELD_ID_DOCIDS),
-        (field_id_word_count_docids, FIELD_ID_WORD_COUNT_DOCIDS),
-        (word_prefix_position_docids, WORD_PREFIX_POSITION_DOCIDS),
-        (word_prefix_fid_docids, WORD_PREFIX_FIELD_ID_DOCIDS),
-        (facet_id_f64_docids, FACET_ID_F64_DOCIDS),
-        (facet_id_string_docids, FACET_ID_STRING_DOCIDS),
-        (facet_id_exists_docids, FACET_ID_EXISTS_DOCIDS),
-        (facet_id_is_null_docids, FACET_ID_IS_NULL_DOCIDS),
-        (facet_id_is_empty_docids, FACET_ID_IS_EMPTY_DOCIDS),
-        (field_id_docid_facet_f64s, FIELD_ID_DOCID_FACET_F64S),
-        (field_id_docid_facet_strings, FIELD_ID_DOCID_FACET_STRINGS),
-        (documents, DOCUMENTS),
-    ];
-
-    let rtxn = env.read_txn()?;
-    let result: Result<Vec<_>, _> =
-        list.into_iter().map(|(db, name)| compute_stats(&rtxn, db).map(|s| (s, name))).collect();
-    let mut stats = result?;
-
-    println!("{:1$} Number of Entries", "", graph_radius as usize * 2);
-    stats.sort_by_key(|(s, _)| Reverse(s.number_of_entries));
-    let data = compute_graph_data(stats.iter().map(|(s, n)| (s.number_of_entries as f32, *n)));
-    Chart::new().radius(graph_radius).aspect_ratio(graph_aspect_ratio).draw(&data);
-    display_legend(&data);
-    print!("\r\n");
-
-    println!("{:1$} Size of Entries", "", graph_radius as usize * 2);
-    stats.sort_by_key(|(s, _)| Reverse(s.size_of_entries));
-    let data = compute_graph_data(stats.iter().map(|(s, n)| (s.size_of_entries as f32, *n)));
-    Chart::new().radius(graph_radius).aspect_ratio(graph_aspect_ratio).draw(&data);
-    display_legend(&data);
-    print!("\r\n");
-
-    println!("{:1$} Size of Data", "", graph_radius as usize * 2);
-    stats.sort_by_key(|(s, _)| Reverse(s.size_of_data));
-    let data = compute_graph_data(stats.iter().map(|(s, n)| (s.size_of_data as f32, *n)));
-    Chart::new().radius(graph_radius).aspect_ratio(graph_aspect_ratio).draw(&data);
-    display_legend(&data);
-    print!("\r\n");
-
-    println!("{:1$} Size of Keys", "", graph_radius as usize * 2);
-    stats.sort_by_key(|(s, _)| Reverse(s.size_of_keys));
-    let data = compute_graph_data(stats.iter().map(|(s, n)| (s.size_of_keys as f32, *n)));
-    Chart::new().radius(graph_radius).aspect_ratio(graph_aspect_ratio).draw(&data);
-    display_legend(&data);
-
-    Ok(())
-}
-
-fn display_legend(data: &[Data]) {
-    let total: f32 = data.iter().map(|d| d.value).sum();
-    for Data { label, value, color, fill } in data {
-        println!(
-            "{} {} {:.02}%",
-            color.unwrap().paint(fill.to_string()),
-            label,
-            value / total * 100.0
-        );
-    }
-}
-
-fn compute_graph_data<'a>(stats: impl IntoIterator<Item = (f32, &'a str)>) -> Vec<Data> {
-    let mut colors = [
-        Color::Red,
-        Color::Green,
-        Color::Yellow,
-        Color::Blue,
-        Color::Purple,
-        Color::Cyan,
-        Color::White,
-    ]
-    .into_iter()
-    .cycle();
-
-    let mut characters = ['▴', '▵', '▾', '▿', '▪', '▫', '•', '◦'].into_iter().cycle();
-
-    stats
-        .into_iter()
-        .map(|(value, name)| Data {
-            label: (*name).into(),
-            value,
-            color: Some(colors.next().unwrap().into()),
-            fill: characters.next().unwrap(),
-        })
-        .collect()
-}
-
-#[derive(Debug)]
-pub struct Stats {
-    pub number_of_entries: u64,
-    pub size_of_keys: u64,
-    pub size_of_data: u64,
-    pub size_of_entries: u64,
-}
-
-fn compute_stats(rtxn: &RoTxn, db: PolyDatabase) -> anyhow::Result<Stats> {
-    let mut number_of_entries = 0;
-    let mut size_of_keys = 0;
-    let mut size_of_data = 0;
-
-    for result in db.iter::<_, ByteSlice, ByteSlice>(rtxn)? {
-        let (key, data) = result?;
-        number_of_entries += 1;
-        size_of_keys += key.len() as u64;
-        size_of_data += data.len() as u64;
-    }
-
-    Ok(Stats {
-        number_of_entries,
-        size_of_keys,
-        size_of_data,
-        size_of_entries: size_of_keys + size_of_data,
-    })
-}

meilisearch-types/src/error.rs

@@ -240,6 +240,8 @@ InvalidSearchOffset                   , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchPage                     , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchQ                        , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchShowMatchesPosition      , InvalidRequest       , BAD_REQUEST ;
+InvalidSearchShowRankingScore         , InvalidRequest       , BAD_REQUEST ;
+InvalidSearchShowRankingScoreDetails  , InvalidRequest       , BAD_REQUEST ;
 InvalidSearchSort                     , InvalidRequest       , BAD_REQUEST ;
 InvalidSettingsDisplayedAttributes    , InvalidRequest       , BAD_REQUEST ;
 InvalidSettingsDistinctAttribute      , InvalidRequest       , BAD_REQUEST ;

meilisearch/src/routes/indexes/search.rs

@@ -56,6 +56,10 @@ pub struct SearchQueryGet {
     sort: Option<String>,
     #[deserr(default, error = DeserrQueryParamError<InvalidSearchShowMatchesPosition>)]
     show_matches_position: Param<bool>,
+    #[deserr(default, error = DeserrQueryParamError<InvalidSearchShowRankingScore>)]
+    show_ranking_score: Param<bool>,
+    #[deserr(default, error = DeserrQueryParamError<InvalidSearchShowRankingScoreDetails>)]
+    show_ranking_score_details: Param<bool>,
     #[deserr(default, error = DeserrQueryParamError<InvalidSearchFacets>)]
     facets: Option<CS<String>>,
     #[deserr( default = DEFAULT_HIGHLIGHT_PRE_TAG(), error = DeserrQueryParamError<InvalidSearchHighlightPreTag>)]
@@ -91,6 +95,8 @@ impl From<SearchQueryGet> for SearchQuery {
             filter,
             sort: other.sort.map(|attr| fix_sort_query_parameters(&attr)),
             show_matches_position: other.show_matches_position.0,
+            show_ranking_score: other.show_ranking_score.0,
+            show_ranking_score_details: other.show_ranking_score_details.0,
             facets: other.facets.map(|o| o.into_iter().collect()),
             highlight_pre_tag: other.highlight_pre_tag,
             highlight_post_tag: other.highlight_post_tag,

meilisearch/src/search.rs

@@ -9,6 +9,7 @@ use meilisearch_auth::IndexSearchRules;
 use meilisearch_types::deserr::DeserrJsonError;
 use meilisearch_types::error::deserr_codes::*;
 use meilisearch_types::index_uid::IndexUid;
+use meilisearch_types::milli::score_details::ScoreDetails;
 use meilisearch_types::settings::DEFAULT_PAGINATION_MAX_TOTAL_HITS;
 use meilisearch_types::{milli, Document};
 use milli::tokenizer::TokenizerBuilder;
@@ -54,6 +55,10 @@ pub struct SearchQuery {
     pub attributes_to_highlight: Option<HashSet<String>>,
     #[deserr(default, error = DeserrJsonError<InvalidSearchShowMatchesPosition>, default)]
     pub show_matches_position: bool,
+    #[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScore>, default)]
+    pub show_ranking_score: bool,
+    #[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScoreDetails>, default)]
+    pub show_ranking_score_details: bool,
     #[deserr(default, error = DeserrJsonError<InvalidSearchFilter>)]
     pub filter: Option<Value>,
     #[deserr(default, error = DeserrJsonError<InvalidSearchSort>)]
@@ -103,6 +108,10 @@ pub struct SearchQueryWithIndex {
     pub crop_length: usize,
     #[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToHighlight>)]
     pub attributes_to_highlight: Option<HashSet<String>>,
+    #[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScore>, default)]
+    pub show_ranking_score: bool,
+    #[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScoreDetails>, default)]
+    pub show_ranking_score_details: bool,
     #[deserr(default, error = DeserrJsonError<InvalidSearchShowMatchesPosition>, default)]
     pub show_matches_position: bool,
     #[deserr(default, error = DeserrJsonError<InvalidSearchFilter>)]
@@ -134,6 +143,8 @@ impl SearchQueryWithIndex {
             attributes_to_crop,
             crop_length,
             attributes_to_highlight,
+            show_ranking_score,
+            show_ranking_score_details,
             show_matches_position,
             filter,
             sort,
@@ -155,6 +166,8 @@ impl SearchQueryWithIndex {
                 attributes_to_crop,
                 crop_length,
                 attributes_to_highlight,
+                show_ranking_score,
+                show_ranking_score_details,
                 show_matches_position,
                 filter,
                 sort,
@@ -194,7 +207,7 @@ impl From<MatchingStrategy> for TermsMatchingStrategy {
     }
 }
 
-#[derive(Debug, Clone, Serialize, PartialEq, Eq)]
+#[derive(Debug, Clone, Serialize, PartialEq)]
 pub struct SearchHit {
     #[serde(flatten)]
     pub document: Document,
@@ -202,6 +215,10 @@ pub struct SearchHit {
     pub formatted: Document,
     #[serde(rename = "_matchesPosition", skip_serializing_if = "Option::is_none")]
     pub matches_position: Option<MatchesPosition>,
+    #[serde(rename = "_rankingScore", skip_serializing_if = "Option::is_none")]
+    pub ranking_score: Option<u64>,
+    #[serde(rename = "_rankingScoreDetails", skip_serializing_if = "Option::is_none")]
+    pub ranking_score_details: Option<serde_json::Map<String, serde_json::Value>>,
 }
 
 #[derive(Serialize, Debug, Clone, PartialEq)]
@@ -320,7 +337,8 @@ pub fn perform_search(
         search.sort_criteria(sort);
     }
 
-    let milli::SearchResult { documents_ids, matching_words, candidates, .. } = search.execute()?;
+    let milli::SearchResult { documents_ids, matching_words, candidates, document_scores, .. } =
+        search.execute()?;
 
     let fields_ids_map = index.fields_ids_map(&rtxn).unwrap();
 
@@ -392,7 +410,7 @@ pub fn perform_search(
 
     let documents_iter = index.documents(&rtxn, documents_ids)?;
 
-    for (_id, obkv) in documents_iter {
+    for ((_id, obkv), score) in documents_iter.into_iter().zip(document_scores.into_iter()) {
         // First generate a document with all the displayed fields
         let displayed_document = make_document(&displayed_ids, &fields_ids_map, obkv)?;
 
@@ -416,7 +434,18 @@ pub fn perform_search(
             insert_geo_distance(sort, &mut document);
         }
 
-        let hit = SearchHit { document, formatted, matches_position };
+        let ranking_score =
+            query.show_ranking_score.then(|| ScoreDetails::global_score_linear_scale(score.iter()));
+        let ranking_score_details =
+            query.show_ranking_score_details.then(|| ScoreDetails::to_json_map(score.iter()));
+
+        let hit = SearchHit {
+            document,
+            formatted,
+            matches_position,
+            ranking_score_details,
+            ranking_score,
+        };
         documents.push(hit);
     }
 

meilisearch/tests/search/formatted.rs

@@ -1,3 +1,4 @@
+use insta::{allow_duplicates, assert_json_snapshot};
 use serde_json::json;
 
 use super::*;
@@ -18,30 +19,43 @@ async fn formatted_contain_wildcard() {
         |response, code|
         {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "_formatted": {
-                        "id": "852",
-                        "cattos": "<em>pésti</em>",
-                    },
-                    "_matchesPosition": {"cattos": [{"start": 0, "length": 5}]},
-                })
-            );
-        }
+            allow_duplicates! {
+              assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+              {
+                "_formatted": {
+                  "id": "852",
+                  "cattos": "<em>pésti</em>"
+                },
+                "_matchesPosition": {
+                  "cattos": [
+                    {
+                      "start": 0,
+                      "length": 5
+                    }
+                  ]
+                }
+              }
+              "###);
+            }
+    }
     )
     .await;
 
     index
         .search(json!({ "q": "pésti", "attributesToRetrieve": ["*"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                    "cattos": "pésti",
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852,
+                  "cattos": "pésti"
+                }
+                "###)
+            }
         })
         .await;
 
@@ -50,20 +64,29 @@ async fn formatted_contain_wildcard() {
             json!({ "q": "pésti", "attributesToRetrieve": ["*"], "attributesToHighlight": ["id"], "showMatchesPosition": true }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "id": 852,
-                        "cattos": "pésti",
-                        "_formatted": {
-                            "id": "852",
-                            "cattos": "pésti",
-                        },
-                        "_matchesPosition": {"cattos": [{"start": 0, "length": 5}]},
-                    })
-                );
-            }
-        )
+                allow_duplicates! {
+                  assert_json_snapshot!(response["hits"][0],
+                 { "._rankingScore" => "[score]" },
+                 @r###"
+                  {
+                    "id": 852,
+                    "cattos": "pésti",
+                    "_formatted": {
+                      "id": "852",
+                      "cattos": "pésti"
+                    },
+                    "_matchesPosition": {
+                      "cattos": [
+                        {
+                          "start": 0,
+                          "length": 5
+                        }
+                      ]
+                    }
+                  }
+                  "###)
+             }
+        })
         .await;
 
     index
@@ -71,17 +94,20 @@ async fn formatted_contain_wildcard() {
             json!({ "q": "pésti", "attributesToRetrieve": ["*"], "attributesToCrop": ["*"] }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "id": 852,
-                        "cattos": "pésti",
-                        "_formatted": {
-                            "id": "852",
-                            "cattos": "pésti",
-                        }
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "id": 852,
+                      "cattos": "pésti",
+                      "_formatted": {
+                        "id": "852",
+                        "cattos": "pésti"
+                      }
+                    }
+                    "###);
+                }
             },
         )
         .await;
@@ -89,17 +115,20 @@ async fn formatted_contain_wildcard() {
     index
         .search(json!({ "q": "pésti", "attributesToCrop": ["*"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                    "cattos": "pésti",
-                    "_formatted": {
-                        "id": "852",
-                        "cattos": "pésti",
-                    }
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852,
+                  "cattos": "pésti",
+                  "_formatted": {
+                    "id": "852",
+                    "cattos": "pésti"
+                  }
+                }
+                "###)
+            }
         })
         .await;
 }
@@ -116,21 +145,24 @@ async fn format_nested() {
     index
         .search(json!({ "q": "pésti", "attributesToRetrieve": ["doggos"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "doggos": [
-                        {
-                            "name": "bobby",
-                            "age": 2,
-                        },
-                        {
-                            "name": "buddy",
-                            "age": 4,
-                        },
-                    ],
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "doggos": [
+                    {
+                      "name": "bobby",
+                      "age": 2
+                    },
+                    {
+                      "name": "buddy",
+                      "age": 4
+                    }
+                  ]
+                }
+                "###)
+            }
         })
         .await;
 
@@ -139,19 +171,22 @@ async fn format_nested() {
             json!({ "q": "pésti", "attributesToRetrieve": ["doggos.name"] }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "doggos": [
-                            {
-                                "name": "bobby",
-                            },
-                            {
-                                "name": "buddy",
-                            },
-                        ],
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "doggos": [
+                        {
+                          "name": "bobby"
+                        },
+                        {
+                          "name": "buddy"
+                        }
+                      ]
+                    }
+                    "###)
+                }
             },
         )
         .await;
@@ -161,20 +196,30 @@ async fn format_nested() {
             json!({ "q": "bobby", "attributesToRetrieve": ["doggos.name"], "showMatchesPosition": true }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "doggos": [
-                            {
-                                "name": "bobby",
-                            },
-                            {
-                                "name": "buddy",
-                            },
-                        ],
-                        "_matchesPosition": {"doggos.name": [{"start": 0, "length": 5}]},
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "doggos": [
+                        {
+                          "name": "bobby"
+                        },
+                        {
+                          "name": "buddy"
+                        }
+                      ],
+                      "_matchesPosition": {
+                        "doggos.name": [
+                          {
+                            "start": 0,
+                            "length": 5
+                          }
+                        ]
+                      }
+                    }
+                    "###)
+                }
             }
         )
         .await;
@@ -183,21 +228,24 @@ async fn format_nested() {
         .search(json!({ "q": "pésti", "attributesToRetrieve": [], "attributesToHighlight": ["doggos.name"] }),
         |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "_formatted": {
-                        "doggos": [
-                            {
-                                "name": "bobby",
-                            },
-                            {
-                                "name": "buddy",
-                            },
-                        ],
-                    },
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "_formatted": {
+                    "doggos": [
+                      {
+                        "name": "bobby"
+                      },
+                      {
+                        "name": "buddy"
+                      }
+                    ]
+                  }
+                }
+                "###)
+            }
         })
         .await;
 
@@ -205,21 +253,24 @@ async fn format_nested() {
         .search(json!({ "q": "pésti", "attributesToRetrieve": [], "attributesToCrop": ["doggos.name"] }),
         |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "_formatted": {
-                        "doggos": [
-                            {
-                                "name": "bobby",
-                            },
-                            {
-                                "name": "buddy",
-                            },
-                        ],
-                    },
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "_formatted": {
+                    "doggos": [
+                      {
+                        "name": "bobby"
+                      },
+                      {
+                        "name": "buddy"
+                      }
+                    ]
+                  }
+                }
+                "###)
+            }
         })
         .await;
 
@@ -227,55 +278,61 @@ async fn format_nested() {
         .search(json!({ "q": "pésti", "attributesToRetrieve": ["doggos.name"], "attributesToHighlight": ["doggos.age"] }),
         |response, code| {
     assert_eq!(code, 200, "{}", response);
-    assert_eq!(
-        response["hits"][0],
-        json!({
-            "doggos": [
-                {
-                    "name": "bobby",
-                },
-                {
-                    "name": "buddy",
-                },
-            ],
-            "_formatted": {
-                "doggos": [
-                    {
-                        "name": "bobby",
-                        "age": "2",
-                    },
-                    {
-                        "name": "buddy",
-                        "age": "4",
-                    },
-                ],
+    allow_duplicates! {
+        assert_json_snapshot!(response["hits"][0],
+        { "._rankingScore" => "[score]" },
+        @r###"
+        {
+          "doggos": [
+            {
+              "name": "bobby"
             },
-        })
-    );
-        })
+            {
+              "name": "buddy"
+            }
+          ],
+          "_formatted": {
+            "doggos": [
+              {
+                "name": "bobby",
+                "age": "2"
+              },
+              {
+                "name": "buddy",
+                "age": "4"
+              }
+            ]
+          }
+        }
+        "###)
+    }
+    })
         .await;
 
     index
         .search(json!({ "q": "pésti", "attributesToRetrieve": [], "attributesToHighlight": ["doggos.age"], "attributesToCrop": ["doggos.name"] }),
         |response, code| {
                 assert_eq!(code, 200, "{}", response);
-    assert_eq!(
-        response["hits"][0],
-        json!({
-            "_formatted": {
-                "doggos": [
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
                     {
-                        "name": "bobby",
-                        "age": "2",
-                    },
-                    {
-                        "name": "buddy",
-                        "age": "4",
-                    },
-                ],
-            },
-        })
-    );
+                      "_formatted": {
+                        "doggos": [
+                          {
+                            "name": "bobby",
+                            "age": "2"
+                          },
+                          {
+                            "name": "buddy",
+                            "age": "4"
+                          }
+                        ]
+                      }
+                    }
+                    "###)
+                }
             }
         )
         .await;
@@ -297,54 +354,66 @@ async fn displayedattr_2_smol() {
         .search(json!({ "attributesToRetrieve": ["father", "id"], "attributesToHighlight": ["mother"], "attributesToCrop": ["cattos"] }),
         |response, code| {
     assert_eq!(code, 200, "{}", response);
-    assert_eq!(
-        response["hits"][0],
-        json!({
-            "id": 852,
-        })
-    );
+    allow_duplicates! {
+        assert_json_snapshot!(response["hits"][0],
+        { "._rankingScore" => "[score]" },
+        @r###"
+        {
+          "id": 852
+        }
+        "###)
+    }
         })
         .await;
 
     index
         .search(json!({ "attributesToRetrieve": ["id"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852
+                }
+                "###)
+            }
         })
         .await;
 
     index
         .search(json!({ "attributesToHighlight": ["id"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                    "_formatted": {
-                        "id": "852",
-                    }
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852,
+                  "_formatted": {
+                    "id": "852"
+                  }
+                }
+                "###)
+            }
         })
         .await;
 
     index
         .search(json!({ "attributesToCrop": ["id"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                    "_formatted": {
-                        "id": "852",
-                    }
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852,
+                  "_formatted": {
+                    "id": "852"
+                  }
+                }
+                "###)
+            }
         })
         .await;
 
@@ -353,15 +422,18 @@ async fn displayedattr_2_smol() {
             json!({ "attributesToHighlight": ["id"], "attributesToCrop": ["id"] }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "id": 852,
-                        "_formatted": {
-                            "id": "852",
-                        }
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "id": 852,
+                      "_formatted": {
+                        "id": "852"
+                      }
+                    }
+                    "###)
+                }
             },
         )
         .await;
@@ -369,31 +441,41 @@ async fn displayedattr_2_smol() {
     index
         .search(json!({ "attributesToHighlight": ["cattos"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852
+                }
+                "###)
+            }
         })
         .await;
 
     index
         .search(json!({ "attributesToCrop": ["cattos"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(
-                response["hits"][0],
-                json!({
-                    "id": 852,
-                })
-            );
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @r###"
+                {
+                  "id": 852
+                }
+                "###)
+            }
         })
         .await;
 
     index
         .search(json!({ "attributesToRetrieve": ["cattos"] }), |response, code| {
             assert_eq!(code, 200, "{}", response);
-            assert_eq!(response["hits"][0], json!({}));
+            allow_duplicates! {
+                assert_json_snapshot!(response["hits"][0],
+                { "._rankingScore" => "[score]" },
+                @"{}")
+            }
         })
         .await;
 
@@ -402,7 +484,11 @@ async fn displayedattr_2_smol() {
             json!({ "attributesToRetrieve": ["cattos"], "attributesToHighlight": ["cattos"], "attributesToCrop": ["cattos"] }),
             |response, code| {
     assert_eq!(code, 200, "{}", response);
-    assert_eq!(response["hits"][0], json!({}));
+    allow_duplicates! {
+        assert_json_snapshot!(response["hits"][0],
+        { "._rankingScore" => "[score]" },
+        @"{}")
+    }
 
             }
         )
@@ -413,14 +499,17 @@ async fn displayedattr_2_smol() {
             json!({ "attributesToRetrieve": ["cattos"], "attributesToHighlight": ["id"] }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "_formatted": {
-                            "id": "852",
-                        }
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "_formatted": {
+                        "id": "852"
+                      }
+                    }
+                    "###)
+                }
             },
         )
         .await;
@@ -430,14 +519,17 @@ async fn displayedattr_2_smol() {
             json!({ "attributesToRetrieve": ["cattos"], "attributesToCrop": ["id"] }),
             |response, code| {
                 assert_eq!(code, 200, "{}", response);
-                assert_eq!(
-                    response["hits"][0],
-                    json!({
-                        "_formatted": {
-                            "id": "852",
-                        }
-                    })
-                );
+                allow_duplicates! {
+                    assert_json_snapshot!(response["hits"][0],
+                    { "._rankingScore" => "[score]" },
+                    @r###"
+                    {
+                      "_formatted": {
+                        "id": "852"
+                      }
+                    }
+                    "###)
+                }
             },
         )
         .await;

meilisearch/tests/search/multi.rs

@@ -65,7 +65,7 @@ async fn simple_search_single_index() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    insta::assert_json_snapshot!(response["results"], { "[].processingTimeMs" => "[time]" }, @r###"
+    insta::assert_json_snapshot!(response["results"], { "[].processingTimeMs" => "[time]", ".**._rankingScore" => "[score]" }, @r###"
     [
       {
         "indexUid": "test",
@@ -170,7 +170,7 @@ async fn simple_search_two_indexes() {
         ]}))
         .await;
     snapshot!(code, @"200 OK");
-    insta::assert_json_snapshot!(response["results"], { "[].processingTimeMs" => "[time]" }, @r###"
+    insta::assert_json_snapshot!(response["results"], { "[].processingTimeMs" => "[time]", ".**._rankingScore" => "[score]" }, @r###"
     [
       {
         "indexUid": "test",

milli/Cargo.toml

@@ -75,6 +75,9 @@ maplit = "1.0.2"
 md5 = "0.7.0"
 rand = { version = "0.8.5", features = ["small_rng"] }
 
+[target.'cfg(fuzzing)'.dev-dependencies]
+fuzzcheck = "0.12.1"
+
 [features]
 all-tokenizations = ["charabia/default"]
 

milli/src/documents/mod.rs

@@ -111,6 +111,7 @@ pub enum Error {
     Io(#[from] io::Error),
 }
 
+#[cfg(test)]
 pub fn objects_from_json_value(json: serde_json::Value) -> Vec<crate::Object> {
     let documents = match json {
         object @ serde_json::Value::Object(_) => vec![object],
@@ -140,6 +141,7 @@ macro_rules! documents {
     }};
 }
 
+#[cfg(test)]
 pub fn documents_batch_reader_from_objects(
     objects: impl IntoIterator<Item = Object>,
 ) -> DocumentsBatchReader<std::io::Cursor<Vec<u8>>> {

milli/src/external_documents_ids.rs

@@ -106,30 +106,22 @@ impl<'a> ExternalDocumentsIds<'a> {
         map
     }
 
-    /// Return an fst of the combined hard and soft deleted ID.
-    pub fn to_fst<'b>(&'b self) -> fst::Result<Cow<'b, fst::Map<Cow<'a, [u8]>>>> {
-        if self.soft.is_empty() {
-            return Ok(Cow::Borrowed(&self.hard));
-        }
-        let union_op = self.hard.op().add(&self.soft).r#union();
-
-        let mut iter = union_op.into_stream();
-        let mut new_hard_builder = fst::MapBuilder::memory();
-        while let Some((external_id, marked_docids)) = iter.next() {
-            let value = indexed_last_value(marked_docids).unwrap();
-            if value != DELETED_ID {
-                new_hard_builder.insert(external_id, value)?;
-            }
-        }
-
-        drop(iter);
-
-        Ok(Cow::Owned(new_hard_builder.into_map().map_data(Cow::Owned)?))
-    }
-
     fn merge_soft_into_hard(&mut self) -> fst::Result<()> {
         if self.soft.len() >= self.hard.len() / 2 {
-            self.hard = self.to_fst()?.into_owned();
+            let union_op = self.hard.op().add(&self.soft).r#union();
+
+            let mut iter = union_op.into_stream();
+            let mut new_hard_builder = fst::MapBuilder::memory();
+            while let Some((external_id, marked_docids)) = iter.next() {
+                let value = indexed_last_value(marked_docids).unwrap();
+                if value != DELETED_ID {
+                    new_hard_builder.insert(external_id, value)?;
+                }
+            }
+
+            drop(iter);
+
+            self.hard = new_hard_builder.into_map().map_data(Cow::Owned)?;
             self.soft = fst::Map::default().map_data(Cow::Owned)?;
         }
 

milli/src/index.rs

@@ -93,10 +93,10 @@ pub mod db_name {
 #[derive(Clone)]
 pub struct Index {
     /// The LMDB environment which this index is associated with.
-    pub env: heed::Env,
+    pub(crate) env: heed::Env,
 
     /// Contains many different types (e.g. the fields ids map).
-    pub main: PolyDatabase,
+    pub(crate) main: PolyDatabase,
 
     /// A word and all the documents ids containing the word.
     pub word_docids: Database<Str, RoaringBitmapCodec>,
@@ -150,7 +150,7 @@ pub struct Index {
     pub field_id_docid_facet_strings: Database<FieldDocIdFacetStringCodec, Str>,
 
     /// Maps the document id to the document as an obkv store.
-    pub documents: Database<OwnedType<BEU32>, ObkvCodec>,
+    pub(crate) documents: Database<OwnedType<BEU32>, ObkvCodec>,
 }
 
 impl Index {
@@ -1466,9 +1466,9 @@ pub(crate) mod tests {
 
         db_snap!(index, field_distribution,
             @r###"
-        age              1      |
-        id               2      |
-        name             2      |
+        age              1     
+        id               2     
+        name             2     
         "###
         );
 
@@ -1486,9 +1486,9 @@ pub(crate) mod tests {
 
         db_snap!(index, field_distribution,
             @r###"
-        age              1      |
-        id               2      |
-        name             2      |
+        age              1     
+        id               2     
+        name             2     
         "###
         );
 
@@ -1502,9 +1502,9 @@ pub(crate) mod tests {
 
         db_snap!(index, field_distribution,
             @r###"
-        has_dog          1      |
-        id               2      |
-        name             2      |
+        has_dog          1     
+        id               2     
+        name             2     
         "###
         );
     }
@@ -2488,8 +2488,12 @@ pub(crate) mod tests {
 
         let rtxn = index.read_txn().unwrap();
         let search = Search::new(&rtxn, &index);
-        let SearchResult { matching_words: _, candidates: _, mut documents_ids } =
-            search.execute().unwrap();
+        let SearchResult {
+            matching_words: _,
+            candidates: _,
+            document_scores: _,
+            mut documents_ids,
+        } = search.execute().unwrap();
         let primary_key_id = index.fields_ids_map(&rtxn).unwrap().id("primary_key").unwrap();
         documents_ids.sort_unstable();
         let docs = index.documents(&rtxn, documents_ids).unwrap();

milli/src/lib.rs

@@ -17,6 +17,7 @@ mod fields_ids_map;
 pub mod heed_codec;
 pub mod index;
 pub mod proximity;
+pub mod score_details;
 mod search;
 pub mod update;
 

milli/src/score_details.rs

@@ -0,0 +1,295 @@
+use serde::Serialize;
+
+use crate::distance_between_two_points;
+
+#[derive(Debug, Clone, PartialEq)]
+pub enum ScoreDetails {
+    Words(Words),
+    Typo(Typo),
+    Proximity(Rank),
+    Fid(Rank),
+    Position(Rank),
+    ExactAttribute(ExactAttribute),
+    Exactness(Rank),
+    Sort(Sort),
+    GeoSort(GeoSort),
+}
+
+impl ScoreDetails {
+    pub fn local_score(&self) -> Option<f64> {
+        self.rank().map(Rank::local_score)
+    }
+
+    pub fn rank(&self) -> Option<Rank> {
+        match self {
+            ScoreDetails::Words(details) => Some(details.rank()),
+            ScoreDetails::Typo(details) => Some(details.rank()),
+            ScoreDetails::Proximity(details) => Some(*details),
+            ScoreDetails::Fid(details) => Some(*details),
+            ScoreDetails::Position(details) => Some(*details),
+            ScoreDetails::ExactAttribute(details) => Some(details.rank()),
+            ScoreDetails::Exactness(details) => Some(*details),
+            ScoreDetails::Sort(_) => None,
+            ScoreDetails::GeoSort(_) => None,
+        }
+    }
+
+    pub fn global_score<'a>(details: impl Iterator<Item = &'a Self>) -> f64 {
+        Rank::global_score(details.filter_map(Self::rank))
+    }
+
+    pub fn global_score_linear_scale<'a>(details: impl Iterator<Item = &'a Self>) -> u64 {
+        (Self::global_score(details) * LINEAR_SCALE_FACTOR).round() as u64
+    }
+
+    /// Panics
+    ///
+    /// - If Position is not preceded by Fid
+    /// - If Exactness is not preceded by ExactAttribute
+    /// - If a sort fid is not contained in the passed `fields_ids_map`.
+    pub fn to_json_map<'a>(
+        details: impl Iterator<Item = &'a Self>,
+    ) -> serde_json::Map<String, serde_json::Value> {
+        let mut order = 0;
+        let mut details_map = serde_json::Map::default();
+        for details in details {
+            match details {
+                ScoreDetails::Words(words) => {
+                    let words_details = serde_json::json!({
+                            "order": order,
+                            "matchingWords": words.matching_words,
+                            "maxMatchingWords": words.max_matching_words,
+                            "score": words.rank().local_score_linear_scale(),
+                    });
+                    details_map.insert("words".into(), words_details);
+                    order += 1;
+                }
+                ScoreDetails::Typo(typo) => {
+                    let typo_details = serde_json::json!({
+                        "order": order,
+                        "typoCount": typo.typo_count,
+                        "maxTypoCount": typo.max_typo_count,
+                        "score": typo.rank().local_score_linear_scale(),
+                    });
+                    details_map.insert("typo".into(), typo_details);
+                    order += 1;
+                }
+                ScoreDetails::Proximity(proximity) => {
+                    let proximity_details = serde_json::json!({
+                        "order": order,
+                        "score": proximity.local_score_linear_scale(),
+                    });
+                    details_map.insert("proximity".into(), proximity_details);
+                    order += 1;
+                }
+                ScoreDetails::Fid(fid) => {
+                    // For now, fid is a virtual rule always followed by the "position" rule
+                    let fid_details = serde_json::json!({
+                        "order": order,
+                        "attributes_ranking_order": fid.local_score_linear_scale(),
+                    });
+                    details_map.insert("attribute".into(), fid_details);
+                    order += 1;
+                }
+                ScoreDetails::Position(position) => {
+                    // For now, position is a virtual rule always preceded by the "fid" rule
+                    let attribute_details = details_map
+                        .get_mut("attribute")
+                        .expect("position not preceded by attribute");
+                    let attribute_details = attribute_details
+                        .as_object_mut()
+                        .expect("attribute details was not an object");
+                    attribute_details.insert(
+                        "attributes_query_word_order".into(),
+                        position.local_score_linear_scale().into(),
+                    );
+                    // do not update the order since this was already done by fid
+                }
+                ScoreDetails::ExactAttribute(exact_attribute) => {
+                    let exactness_details = serde_json::json!({
+                        "order": order,
+                        "exactIn": exact_attribute,
+                        "score": exact_attribute.rank().local_score_linear_scale(),
+                    });
+                    details_map.insert("exactness".into(), exactness_details);
+                    order += 1;
+                }
+                ScoreDetails::Exactness(details) => {
+                    // For now, exactness is a virtual rule always preceded by the "ExactAttribute" rule
+                    let exactness_details = details_map
+                        .get_mut("exactness")
+                        .expect("Exactness not preceded by exactAttribute");
+                    let exactness_details = exactness_details
+                        .as_object_mut()
+                        .expect("exactness details was not an object");
+                    if exactness_details.get("exactIn").expect("missing 'exactIn'")
+                        == &serde_json::json!(ExactAttribute::NoExactMatch)
+                    {
+                        let score = Rank::global_score_linear_scale(
+                            [ExactAttribute::NoExactMatch.rank(), *details].iter().copied(),
+                        );
+                        *exactness_details.get_mut("score").expect("missing score") = score.into();
+                    }
+                    // do not update the order since this was already done by exactAttribute
+                }
+                ScoreDetails::Sort(details) => {
+                    let sort = format!(
+                        "{}:{}",
+                        details.field_name,
+                        if details.ascending { "asc" } else { "desc" }
+                    );
+                    let sort_details = serde_json::json!({
+                        "order": order,
+                        "value": details.value,
+                    });
+                    details_map.insert(sort, sort_details);
+                    order += 1;
+                }
+                ScoreDetails::GeoSort(details) => {
+                    let sort = format!(
+                        "_geoPoint({}, {}):{}",
+                        details.target_point[0],
+                        details.target_point[1],
+                        if details.ascending { "asc" } else { "desc" }
+                    );
+                    let point = if let Some(value) = details.value {
+                        serde_json::json!({ "lat": value[0], "lng": value[1]})
+                    } else {
+                        serde_json::Value::Null
+                    };
+                    let sort_details = serde_json::json!({
+                        "order": order,
+                        "value": point,
+                        "distance": details.distance(),
+                    });
+                    details_map.insert(sort, sort_details);
+                    order += 1;
+                }
+            }
+        }
+        details_map
+    }
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
+pub struct Words {
+    pub matching_words: u32,
+    pub max_matching_words: u32,
+}
+
+impl Words {
+    pub fn rank(&self) -> Rank {
+        Rank { rank: self.matching_words, max_rank: self.max_matching_words }
+    }
+
+    pub(crate) fn from_rank(rank: Rank) -> Words {
+        Words { matching_words: rank.rank, max_matching_words: rank.max_rank }
+    }
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
+pub struct Typo {
+    pub typo_count: u32,
+    pub max_typo_count: u32,
+}
+
+impl Typo {
+    pub fn rank(&self) -> Rank {
+        Rank {
+            rank: self.max_typo_count - self.typo_count + 1,
+            max_rank: (self.max_typo_count + 1),
+        }
+    }
+
+    // max_rank = max_typo + 1
+    // max_typo = max_rank - 1
+    //
+    // rank = max_typo - typo + 1
+    // rank = max_rank - 1 - typo + 1
+    // rank + typo = max_rank
+    // typo = max_rank - rank
+    pub fn from_rank(rank: Rank) -> Typo {
+        Typo { typo_count: rank.max_rank - rank.rank, max_typo_count: rank.max_rank - 1 }
+    }
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
+pub struct Rank {
+    /// The ordinal rank, such that `max_rank` is the first rank, and 0 is the last rank.
+    ///
+    /// The higher the better. Documents with a rank of 0 have a score of 0 and are typically never returned
+    /// (they don't match the query).
+    pub rank: u32,
+    /// The maximum possible rank. Documents with this rank have a score of 1.
+    ///
+    /// The max rank should not be 0.
+    pub max_rank: u32,
+}
+
+impl Rank {
+    pub fn local_score(self) -> f64 {
+        self.rank as f64 / self.max_rank as f64
+    }
+
+    pub fn local_score_linear_scale(self) -> u64 {
+        (self.local_score() * LINEAR_SCALE_FACTOR).round() as u64
+    }
+
+    pub fn global_score(details: impl Iterator<Item = Self>) -> f64 {
+        let mut rank = Rank { rank: 1, max_rank: 1 };
+        for inner_rank in details {
+            rank.rank -= 1;
+
+            rank.rank *= inner_rank.max_rank;
+            rank.max_rank *= inner_rank.max_rank;
+
+            rank.rank += inner_rank.rank;
+        }
+        rank.local_score()
+    }
+
+    pub fn global_score_linear_scale(details: impl Iterator<Item = Self>) -> u64 {
+        (Self::global_score(details) * LINEAR_SCALE_FACTOR).round() as u64
+    }
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash, Serialize)]
+#[serde(rename_all = "camelCase")]
+pub enum ExactAttribute {
+    MatchesFull,
+    MatchesStart,
+    NoExactMatch,
+}
+
+impl ExactAttribute {
+    pub fn rank(&self) -> Rank {
+        let rank = match self {
+            ExactAttribute::MatchesFull => 3,
+            ExactAttribute::MatchesStart => 2,
+            ExactAttribute::NoExactMatch => 1,
+        };
+        Rank { rank, max_rank: 3 }
+    }
+}
+
+#[derive(Debug, Clone, PartialEq)]
+pub struct Sort {
+    pub field_name: String,
+    pub ascending: bool,
+    pub value: serde_json::Value,
+}
+
+#[derive(Debug, Clone, Copy, PartialEq, PartialOrd)]
+pub struct GeoSort {
+    pub target_point: [f64; 2],
+    pub ascending: bool,
+    pub value: Option<[f64; 2]>,
+}
+
+impl GeoSort {
+    pub fn distance(&self) -> Option<f64> {
+        self.value.map(|value| distance_between_two_points(&self.target_point, &value))
+    }
+}
+
+const LINEAR_SCALE_FACTOR: f64 = 1000.0;

milli/src/search/mod.rs

@@ -7,6 +7,7 @@ use roaring::bitmap::RoaringBitmap;
 pub use self::facet::{FacetDistribution, Filter, DEFAULT_VALUES_PER_FACET};
 pub use self::new::matches::{FormatOptions, MatchBounds, Matcher, MatcherBuilder, MatchingWords};
 use self::new::PartialSearchResult;
+use crate::score_details::ScoreDetails;
 use crate::{
     execute_search, AscDesc, DefaultSearchLogger, DocumentId, Index, Result, SearchContext,
 };
@@ -93,7 +94,7 @@ impl<'a> Search<'a> {
         self
     }
 
-    /// Force the search to exhastivelly compute the number of candidates,
+    /// Forces the search to exhaustively compute the number of candidates,
     /// this will increase the search time but allows finite pagination.
     pub fn exhaustive_number_hits(&mut self, exhaustive_number_hits: bool) -> &mut Search<'a> {
         self.exhaustive_number_hits = exhaustive_number_hits;
@@ -102,7 +103,7 @@ impl<'a> Search<'a> {
 
     pub fn execute(&self) -> Result<SearchResult> {
         let mut ctx = SearchContext::new(self.index, self.rtxn);
-        let PartialSearchResult { located_query_terms, candidates, documents_ids } =
+        let PartialSearchResult { located_query_terms, candidates, documents_ids, document_scores } =
             execute_search(
                 &mut ctx,
                 &self.query,
@@ -124,7 +125,7 @@ impl<'a> Search<'a> {
             None => MatchingWords::default(),
         };
 
-        Ok(SearchResult { matching_words, candidates, documents_ids })
+        Ok(SearchResult { matching_words, candidates, document_scores, documents_ids })
     }
 }
 
@@ -160,8 +161,8 @@ impl fmt::Debug for Search<'_> {
 pub struct SearchResult {
     pub matching_words: MatchingWords,
     pub candidates: RoaringBitmap,
-    // TODO those documents ids should be associated with their criteria scores.
     pub documents_ids: Vec<DocumentId>,
+    pub document_scores: Vec<Vec<ScoreDetails>>,
 }
 
 #[derive(Debug, Clone, Copy, PartialEq, Eq)]

milli/src/search/new/bucket_sort.rs

@@ -3,11 +3,13 @@ use roaring::RoaringBitmap;
 use super::logger::SearchLogger;
 use super::ranking_rules::{BoxRankingRule, RankingRuleQueryTrait};
 use super::SearchContext;
+use crate::score_details::ScoreDetails;
 use crate::search::new::distinct::{apply_distinct_rule, distinct_single_docid, DistinctOutput};
 use crate::Result;
 
 pub struct BucketSortOutput {
     pub docids: Vec<u32>,
+    pub scores: Vec<Vec<ScoreDetails>>,
     pub all_candidates: RoaringBitmap,
 }
 
@@ -31,7 +33,11 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
     };
 
     if universe.len() < from as u64 {
-        return Ok(BucketSortOutput { docids: vec![], all_candidates: universe.clone() });
+        return Ok(BucketSortOutput {
+            docids: vec![],
+            scores: vec![],
+            all_candidates: universe.clone(),
+        });
     }
     if ranking_rules.is_empty() {
         if let Some(distinct_fid) = distinct_fid {
@@ -49,22 +55,32 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
             }
             let mut all_candidates = universe - excluded;
             all_candidates.extend(results.iter().copied());
-            return Ok(BucketSortOutput { docids: results, all_candidates });
+            return Ok(BucketSortOutput {
+                scores: vec![Default::default(); results.len()],
+                docids: results,
+                all_candidates,
+            });
         } else {
-            let docids = universe.iter().skip(from).take(length).collect();
-            return Ok(BucketSortOutput { docids, all_candidates: universe.clone() });
+            let docids: Vec<u32> = universe.iter().skip(from).take(length).collect();
+            return Ok(BucketSortOutput {
+                scores: vec![Default::default(); docids.len()],
+                docids,
+                all_candidates: universe.clone(),
+            });
         };
     }
 
     let ranking_rules_len = ranking_rules.len();
 
     logger.start_iteration_ranking_rule(0, ranking_rules[0].as_ref(), query, universe);
+
     ranking_rules[0].start_iteration(ctx, logger, universe, query)?;
 
+    let mut ranking_rule_scores: Vec<ScoreDetails> = vec![];
+
     let mut ranking_rule_universes: Vec<RoaringBitmap> =
         vec![RoaringBitmap::default(); ranking_rules_len];
     ranking_rule_universes[0] = universe.clone();
-
     let mut cur_ranking_rule_index = 0;
 
     /// Finish iterating over the current ranking rule, yielding
@@ -89,11 +105,16 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
             } else {
                 cur_ranking_rule_index -= 1;
             }
+            // FIXME: check off by one
+            if ranking_rule_scores.len() > cur_ranking_rule_index {
+                ranking_rule_scores.pop();
+            }
         };
     }
 
     let mut all_candidates = universe.clone();
     let mut valid_docids = vec![];
+    let mut valid_scores = vec![];
     let mut cur_offset = 0usize;
 
     macro_rules! maybe_add_to_results {
@@ -104,23 +125,23 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
                 length,
                 logger,
                 &mut valid_docids,
+                &mut valid_scores,
                 &mut all_candidates,
                 &mut ranking_rule_universes,
                 &mut ranking_rules,
                 cur_ranking_rule_index,
                 &mut cur_offset,
                 distinct_fid,
+                &ranking_rule_scores,
                 $candidates,
             )?;
         };
     }
 
     while valid_docids.len() < length {
-        // The universe for this bucket is zero or one element, so we don't need to sort
-        // anything, just extend the results and go back to the parent ranking rule.
-        if ranking_rule_universes[cur_ranking_rule_index].len() <= 1 {
-            let bucket = std::mem::take(&mut ranking_rule_universes[cur_ranking_rule_index]);
-            maybe_add_to_results!(bucket);
+        // The universe for this bucket is zero, so we don't need to sort
+        // anything, just go back to the parent ranking rule.
+        if ranking_rule_universes[cur_ranking_rule_index].is_empty() {
             back!();
             continue;
         }
@@ -130,6 +151,8 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
             continue;
         };
 
+        ranking_rule_scores.push(next_bucket.score);
+
         logger.next_bucket_ranking_rule(
             cur_ranking_rule_index,
             ranking_rules[cur_ranking_rule_index].as_ref(),
@@ -143,10 +166,11 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
         ranking_rule_universes[cur_ranking_rule_index] -= &next_bucket.candidates;
 
         if cur_ranking_rule_index == ranking_rules_len - 1
-            || next_bucket.candidates.len() <= 1
             || cur_offset + (next_bucket.candidates.len() as usize) < from
         {
             maybe_add_to_results!(next_bucket.candidates);
+            // FIXME: use index based logic like all the other rules so that you don't have to maintain the pop/push?
+            ranking_rule_scores.pop();
             continue;
         }
 
@@ -166,7 +190,7 @@ pub fn bucket_sort<'ctx, Q: RankingRuleQueryTrait>(
         )?;
     }
 
-    Ok(BucketSortOutput { docids: valid_docids, all_candidates })
+    Ok(BucketSortOutput { docids: valid_docids, scores: valid_scores, all_candidates })
 }
 
 /// Add the candidates to the results. Take `distinct`, `from`, `length`, and `cur_offset`
@@ -179,14 +203,18 @@ fn maybe_add_to_results<'ctx, Q: RankingRuleQueryTrait>(
     logger: &mut dyn SearchLogger<Q>,
 
     valid_docids: &mut Vec<u32>,
+    valid_scores: &mut Vec<Vec<ScoreDetails>>,
     all_candidates: &mut RoaringBitmap,
 
     ranking_rule_universes: &mut [RoaringBitmap],
     ranking_rules: &mut [BoxRankingRule<'ctx, Q>],
+
     cur_ranking_rule_index: usize,
 
     cur_offset: &mut usize,
+
     distinct_fid: Option<u16>,
+    ranking_rule_scores: &[ScoreDetails],
     candidates: RoaringBitmap,
 ) -> Result<()> {
     // First apply the distinct rule on the candidates, reducing the universes if necessary
@@ -231,13 +259,17 @@ fn maybe_add_to_results<'ctx, Q: RankingRuleQueryTrait>(
             let candidates =
                 candidates.iter().take(length - valid_docids.len()).copied().collect::<Vec<_>>();
             logger.add_to_results(&candidates);
-            valid_docids.extend(&candidates);
+            valid_docids.extend_from_slice(&candidates);
+            valid_scores
+                .extend(std::iter::repeat(ranking_rule_scores.to_owned()).take(candidates.len()));
         }
     } else {
         // if we have passed the offset already, add some of the documents (up to the limit)
         let candidates = candidates.iter().take(length - valid_docids.len()).collect::<Vec<u32>>();
         logger.add_to_results(&candidates);
-        valid_docids.extend(&candidates);
+        valid_docids.extend_from_slice(&candidates);
+        valid_scores
+            .extend(std::iter::repeat(ranking_rule_scores.to_owned()).take(candidates.len()));
     }
 
     *cur_offset += candidates.len() as usize;

milli/src/search/new/distinct.rs

@@ -26,6 +26,7 @@ pub fn apply_distinct_rule(
     ctx: &mut SearchContext,
     field_id: u16,
     candidates: &RoaringBitmap,
+    // TODO: add a universe here, such that the `excluded` are a subset of the universe?
 ) -> Result<DistinctOutput> {
     let mut excluded = RoaringBitmap::new();
     let mut remaining = RoaringBitmap::new();

milli/src/search/new/exact_attribute.rs

@@ -2,6 +2,7 @@ use roaring::{MultiOps, RoaringBitmap};
 
 use super::query_graph::QueryGraph;
 use super::ranking_rules::{RankingRule, RankingRuleOutput};
+use crate::score_details::{self, ScoreDetails};
 use crate::search::new::query_graph::QueryNodeData;
 use crate::search::new::query_term::ExactTerm;
 use crate::{Result, SearchContext, SearchLogger};
@@ -206,7 +207,7 @@ impl State {
             )?;
             intersection &= &candidates;
             if !intersection.is_empty() {
-                // Although not really worth it in terms of performance,
+                // TODO: although not really worth it in terms of performance,
                 // if would be good to put this in cache for the sake of consistency
                 let candidates_with_exact_word_count = if count_all_positions < u8::MAX as usize {
                     ctx.index
@@ -244,7 +245,13 @@ impl State {
                 candidates &= universe;
                 (
                     State::AttributeStarts(query_graph.clone(), candidates_per_attribute),
-                    Some(RankingRuleOutput { query: query_graph, candidates }),
+                    Some(RankingRuleOutput {
+                        query: query_graph,
+                        candidates,
+                        score: ScoreDetails::ExactAttribute(
+                            score_details::ExactAttribute::MatchesFull,
+                        ),
+                    }),
                 )
             }
             State::AttributeStarts(query_graph, candidates_per_attribute) => {
@@ -257,12 +264,24 @@ impl State {
                 candidates &= universe;
                 (
                     State::Empty(query_graph.clone()),
-                    Some(RankingRuleOutput { query: query_graph, candidates }),
+                    Some(RankingRuleOutput {
+                        query: query_graph,
+                        candidates,
+                        score: ScoreDetails::ExactAttribute(
+                            score_details::ExactAttribute::MatchesStart,
+                        ),
+                    }),
                 )
             }
             State::Empty(query_graph) => (
                 State::Empty(query_graph.clone()),
-                Some(RankingRuleOutput { query: query_graph, candidates: universe.clone() }),
+                Some(RankingRuleOutput {
+                    query: query_graph,
+                    candidates: universe.clone(),
+                    score: ScoreDetails::ExactAttribute(
+                        score_details::ExactAttribute::NoExactMatch,
+                    ),
+                }),
             ),
         };
         (state, output)

milli/src/search/new/geo_sort.rs

@@ -8,6 +8,7 @@ use rstar::RTree;
 
 use super::ranking_rules::{RankingRule, RankingRuleOutput, RankingRuleQueryTrait};
 use crate::heed_codec::facet::{FieldDocIdFacetCodec, OrderedF64Codec};
+use crate::score_details::{self, ScoreDetails};
 use crate::{
     distance_between_two_points, lat_lng_to_xyz, GeoPoint, Index, Result, SearchContext,
     SearchLogger,
@@ -80,7 +81,7 @@ pub struct GeoSort<Q: RankingRuleQueryTrait> {
     field_ids: Option<[u16; 2]>,
     rtree: Option<RTree<GeoPoint>>,
 
-    cached_sorted_docids: VecDeque<u32>,
+    cached_sorted_docids: VecDeque<(u32, [f64; 2])>,
     geo_candidates: RoaringBitmap,
 }
 
@@ -130,7 +131,7 @@ impl<Q: RankingRuleQueryTrait> GeoSort<Q> {
                 let point = lat_lng_to_xyz(&self.point);
                 for point in rtree.nearest_neighbor_iter(&point) {
                     if self.geo_candidates.contains(point.data.0) {
-                        self.cached_sorted_docids.push_back(point.data.0);
+                        self.cached_sorted_docids.push_back(point.data);
                         if self.cached_sorted_docids.len() >= cache_size {
                             break;
                         }
@@ -142,7 +143,7 @@ impl<Q: RankingRuleQueryTrait> GeoSort<Q> {
                 let point = lat_lng_to_xyz(&opposite_of(self.point));
                 for point in rtree.nearest_neighbor_iter(&point) {
                     if self.geo_candidates.contains(point.data.0) {
-                        self.cached_sorted_docids.push_front(point.data.0);
+                        self.cached_sorted_docids.push_front(point.data);
                         if self.cached_sorted_docids.len() >= cache_size {
                             break;
                         }
@@ -177,7 +178,7 @@ impl<Q: RankingRuleQueryTrait> GeoSort<Q> {
             // computing the distance between two points is expensive thus we cache the result
             documents
                 .sort_by_cached_key(|(_, p)| distance_between_two_points(&self.point, p) as usize);
-            self.cached_sorted_docids.extend(documents.into_iter().map(|(doc_id, _)| doc_id));
+            self.cached_sorted_docids.extend(documents.into_iter());
         };
 
         Ok(())
@@ -220,12 +221,19 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for GeoSort<Q> {
         logger: &mut dyn SearchLogger<Q>,
         universe: &RoaringBitmap,
     ) -> Result<Option<RankingRuleOutput<Q>>> {
-        assert!(universe.len() > 1);
         let query = self.query.as_ref().unwrap().clone();
         self.geo_candidates &= universe;
 
         if self.geo_candidates.is_empty() {
-            return Ok(Some(RankingRuleOutput { query, candidates: universe.clone() }));
+            return Ok(Some(RankingRuleOutput {
+                query,
+                candidates: universe.clone(),
+                score: ScoreDetails::GeoSort(score_details::GeoSort {
+                    target_point: self.point,
+                    ascending: self.ascending,
+                    value: None,
+                }),
+            }));
         }
 
         let ascending = self.ascending;
@@ -236,11 +244,16 @@ impl<'ctx, Q: RankingRuleQueryTrait> RankingRule<'ctx, Q> for GeoSort<Q> {
                 cache.pop_back()
             }
         };
-        while let Some(id) = next(&mut self.cached_sorted_docids) {
+        while let Some((id, point)) = next(&mut self.cached_sorted_docids) {
             if self.geo_candidates.contains(id) {
                 return Ok(Some(RankingRuleOutput {
                     query,
                     candidates: RoaringBitmap::from_iter([id]),
+                    score: ScoreDetails::GeoSort(score_details::GeoSort {
+                        target_point: self.point,
+                        ascending: self.ascending,
+                        value: Some(point),
+                    }),
                 }));
             }
         }

milli/src/search/new/graph_based_ranking_rule.rs

@@ -50,6 +50,7 @@ use super::ranking_rule_graph::{
 };
 use super::small_bitmap::SmallBitmap;
 use super::{QueryGraph, RankingRule, RankingRuleOutput, SearchContext};
+use crate::score_details::Rank;
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::ranking_rule_graph::PathVisitor;
 use crate::{Result, TermsMatchingStrategy};
@@ -118,6 +119,8 @@ pub struct GraphBasedRankingRuleState<G: RankingRuleGraphTrait> {
     all_costs: MappedInterner<QueryNode, Vec<u64>>,
     /// An index in the first element of `all_distances`, giving the cost of the next bucket
     cur_cost: u64,
+    /// One above the highest possible cost for this rule
+    next_max_cost: u64,
 }
 
 impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBasedRankingRule<G> {
@@ -139,13 +142,12 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
                     let mut forbidden_nodes =
                         SmallBitmap::for_interned_values_in(&query_graph.nodes);
                     let mut costs = query_graph.nodes.map(|_| None);
-                    let mut cost = 100;
+                    // FIXME: this works because only words uses termsmatchingstrategy at the moment.
                     for ns in removal_order {
                         for n in ns.iter() {
-                            *costs.get_mut(n) = Some((cost, forbidden_nodes.clone()));
+                            *costs.get_mut(n) = Some((1, forbidden_nodes.clone()));
                         }
                         forbidden_nodes.union(&ns);
-                        cost += 100;
                     }
                     costs
                 }
@@ -162,12 +164,16 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
         // Then pre-compute the cost of all paths from each node to the end node
         let all_costs = graph.find_all_costs_to_end();
 
+        let next_max_cost =
+            all_costs.get(graph.query_graph.root_node).iter().copied().max().unwrap_or(0) + 1;
+
         let state = GraphBasedRankingRuleState {
             graph,
             conditions_cache: condition_docids_cache,
             dead_ends_cache,
             all_costs,
             cur_cost: 0,
+            next_max_cost,
         };
 
         self.state = Some(state);
@@ -181,17 +187,13 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
         logger: &mut dyn SearchLogger<QueryGraph>,
         universe: &RoaringBitmap,
     ) -> Result<Option<RankingRuleOutput<QueryGraph>>> {
-        // If universe.len() <= 1, the bucket sort algorithm
-        // should not have called this function.
-        assert!(universe.len() > 1);
         // Will crash if `next_bucket` is called before `start_iteration` or after `end_iteration`,
         // should never happen
         let mut state = self.state.take().unwrap();
 
+        let all_costs = state.all_costs.get(state.graph.query_graph.root_node);
         // Retrieve the cost of the paths to compute
-        let Some(&cost) = state
-            .all_costs
-            .get(state.graph.query_graph.root_node)
+        let Some(&cost) = all_costs
             .iter()
             .find(|c| **c >= state.cur_cost) else {
                 self.state = None;
@@ -207,8 +209,12 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
             dead_ends_cache,
             all_costs,
             cur_cost: _,
+            next_max_cost,
         } = &mut state;
 
+        let rank = *next_max_cost - cost;
+        let score = G::rank_to_score(Rank { rank: rank as u32, max_rank: *next_max_cost as u32 });
+
         let mut universe = universe.clone();
 
         let mut used_conditions = SmallBitmap::for_interned_values_in(&graph.conditions_interner);
@@ -325,7 +331,7 @@ impl<'ctx, G: RankingRuleGraphTrait> RankingRule<'ctx, QueryGraph> for GraphBase
 
         self.state = Some(state);
 
-        Ok(Some(RankingRuleOutput { query: next_query_graph, candidates: bucket }))
+        Ok(Some(RankingRuleOutput { query: next_query_graph, candidates: bucket, score }))
     }
 
     fn end_iteration(

milli/src/search/new/interner.rs

@@ -32,7 +32,7 @@ impl<T> Interned<T> {
 #[derive(Clone)]
 pub struct DedupInterner<T> {
     stable_store: Vec<T>,
-    lookup: FxHashMap<T, Interned<T>>,
+    lookup: FxHashMap<T, Interned<T>>, // TODO: Arc
 }
 impl<T> Default for DedupInterner<T> {
     fn default() -> Self {

milli/src/search/new/limits.rs

@@ -1,4 +1,5 @@
 /// Maximum number of tokens we consider in a single search.
+// TODO: Loic, find proper value here so we don't overflow the interner.
 pub const MAX_TOKEN_COUNT: usize = 1_000;
 
 /// Maximum number of prefixes that can be derived from a single word.

milli/src/search/new/mod.rs

@@ -44,6 +44,7 @@ use self::geo_sort::GeoSort;
 pub use self::geo_sort::Strategy as GeoSortStrategy;
 use self::graph_based_ranking_rule::Words;
 use self::interner::Interned;
+use crate::score_details::ScoreDetails;
 use crate::search::new::distinct::apply_distinct_rule;
 use crate::{AscDesc, DocumentId, Filter, Index, Member, Result, TermsMatchingStrategy, UserError};
 
@@ -426,13 +427,15 @@ pub fn execute_search(
         )?
     };
 
-    let BucketSortOutput { docids, mut all_candidates } = bucket_sort_output;
+    let BucketSortOutput { docids, scores, mut all_candidates } = bucket_sort_output;
+
+    let fields_ids_map = ctx.index.fields_ids_map(ctx.txn)?;
 
     // The candidates is the universe unless the exhaustive number of hits
     // is requested and a distinct attribute is set.
     if exhaustive_number_hits {
         if let Some(f) = ctx.index.distinct_field(ctx.txn)? {
-            if let Some(distinct_fid) = ctx.index.fields_ids_map(ctx.txn)?.id(f) {
+            if let Some(distinct_fid) = fields_ids_map.id(f) {
                 all_candidates = apply_distinct_rule(ctx, distinct_fid, &all_candidates)?.remaining;
             }
         }
@@ -440,6 +443,7 @@ pub fn execute_search(
 
     Ok(PartialSearchResult {
         candidates: all_candidates,
+        document_scores: scores,
         documents_ids: docids,
         located_query_terms,
     })
@@ -491,4 +495,5 @@ pub struct PartialSearchResult {
     pub located_query_terms: Option<Vec<LocatedQueryTerm>>,
     pub candidates: RoaringBitmap,
     pub documents_ids: Vec<DocumentId>,
+    pub document_scores: Vec<Vec<ScoreDetails>>,
 }

milli/src/search/new/query_graph.rs

@@ -92,7 +92,7 @@ impl QueryGraph {
     /// which contains ngrams.
     pub fn from_query(
         ctx: &mut SearchContext,
-        // The terms here must be consecutive
+        // NOTE: the terms here must be consecutive
         terms: &[LocatedQueryTerm],
     ) -> Result<(QueryGraph, Vec<LocatedQueryTerm>)> {
         let mut new_located_query_terms = terms.to_vec();
@@ -103,7 +103,7 @@ impl QueryGraph {
         let root_node = 0;
         let end_node = 1;
 
-        // Ee could consider generalizing to 4,5,6,7,etc. ngrams
+        // TODO: we could consider generalizing to 4,5,6,7,etc. ngrams
         let (mut prev2, mut prev1, mut prev0): (Vec<u16>, Vec<u16>, Vec<u16>) =
             (vec![], vec![], vec![root_node]);
 

milli/src/search/new/query_term/mod.rs

@@ -132,6 +132,7 @@ impl QueryTermSubset {
         if full_query_term.ngram_words.is_some() {
             return None;
         }
+        // TODO: included in subset
         if let Some(phrase) = full_query_term.zero_typo.phrase {
             self.zero_typo_subset.contains_phrase(phrase).then_some(ExactTerm::Phrase(phrase))
         } else if let Some(word) = full_query_term.zero_typo.exact {
@@ -181,6 +182,7 @@ impl QueryTermSubset {
         let word = match &self.zero_typo_subset {
             NTypoTermSubset::All => Some(use_prefix_db),
             NTypoTermSubset::Subset { words, phrases: _ } => {
+                // TODO: use a subset of prefix words instead
                 if words.contains(&use_prefix_db) {
                     Some(use_prefix_db)
                 } else {
@@ -202,6 +204,7 @@ impl QueryTermSubset {
         ctx: &mut SearchContext,
     ) -> Result<BTreeSet<Word>> {
         let mut result = BTreeSet::default();
+        // TODO: a compute_partially funtion
         if !self.one_typo_subset.is_empty() || !self.two_typo_subset.is_empty() {
             self.original.compute_fully_if_needed(ctx)?;
         }
@@ -297,6 +300,7 @@ impl QueryTermSubset {
         let mut result = BTreeSet::default();
 
         if !self.one_typo_subset.is_empty() {
+            // TODO: compute less than fully if possible
             self.original.compute_fully_if_needed(ctx)?;
         }
         let original = ctx.term_interner.get_mut(self.original);

milli/src/search/new/query_term/parse_query.rs

@@ -139,6 +139,7 @@ pub fn number_of_typos_allowed<'ctx>(
     let min_len_one_typo = ctx.index.min_word_len_one_typo(ctx.txn)?;
     let min_len_two_typos = ctx.index.min_word_len_two_typos(ctx.txn)?;
 
+    // TODO: should `exact_words` also disable prefix search, ngrams, split words, or synonyms?
     let exact_words = ctx.index.exact_words(ctx.txn)?;
 
     Ok(Box::new(move |word: &str| {
@@ -249,6 +250,8 @@ impl PhraseBuilder {
         } else {
             // token has kind Word
             let word = ctx.word_interner.insert(token.lemma().to_string());
+            // TODO: in a phrase, check that every word exists
+            // otherwise return an empty term
             self.words.push(Some(word));
         }
     }

milli/src/search/new/ranking_rule_graph/build.rs

@@ -49,10 +49,15 @@ impl<G: RankingRuleGraphTrait> RankingRuleGraph<G> {
                 if let Some((cost_of_ignoring, forbidden_nodes)) =
                     cost_of_ignoring_node.get(dest_idx)
                 {
+                    let dest = graph_nodes.get(dest_idx);
+                    let dest_size = match &dest.data {
+                        QueryNodeData::Term(term) => term.term_ids.len(),
+                        _ => panic!(),
+                    };
                     let new_edge_id = edges_store.insert(Some(Edge {
                         source_node: source_id,
                         dest_node: dest_idx,
-                        cost: *cost_of_ignoring,
+                        cost: *cost_of_ignoring * dest_size as u32,
                         condition: None,
                         nodes_to_skip: forbidden_nodes.clone(),
                     }));

milli/src/search/new/ranking_rule_graph/cheapest_paths.rs

@@ -1,48 +1,5 @@
-/** Implements a "PathVisitor" which finds all paths of a certain cost
-from the START to END node of a ranking rule graph.
+#![allow(clippy::too_many_arguments)]
 
-A path is a list of conditions. A condition is the data associated with
-an edge, given by the ranking rule. Some edges don't have a condition associated
-with them, they are "unconditional". These kinds of edges are used to "skip" a node.
-
-The algorithm uses a depth-first search. It benefits from two main optimisations:
-- The list of all possible costs to go from any node to the END node is precomputed
-- The `DeadEndsCache` reduces the number of valid paths drastically, by making some edges
-untraversable depending on what other edges were selected.
-
-These two optimisations are meant to avoid traversing edges that wouldn't lead
-to a valid path. In practically all cases, we avoid the exponential complexity
-that is inherent to depth-first search in a large ranking rule graph.
-
-The DeadEndsCache is a sort of prefix tree which associates a list of forbidden
-conditions to a list of traversed conditions.
-For example, the DeadEndsCache could say the following:
-- Immediately, from the start, the conditions `[a,b]` are forbidden
-    - if we take the condition `c`, then the conditions `[e]` are also forbidden
-        - and if after that, we take `f`, then `[h,i]` are also forbidden
-            - etc.
-    - if we take `g`, then `[f]` is also forbidden
-        - etc.
-    - etc.
-As we traverse the graph, we also traverse the `DeadEndsCache` and keep a list of forbidden
-conditions in memory. Then, we know to avoid all edges which have a condition that is forbidden.
-
-When a path is found from START to END, we give it to the `visit` closure.
-This closure takes a mutable reference to the `DeadEndsCache`. This means that
-the caller can update this cache. Therefore, we must handle the case where the
-DeadEndsCache has been updated. This means potentially backtracking up to the point
-where the traversed conditions are all allowed by the new DeadEndsCache.
-
-The algorithm also implements the `TermsMatchingStrategy` logic.
-Some edges are augmented with a list of "nodes_to_skip". Skipping
-a node means "reaching this node through an unconditional edge". If we have
-already traversed (ie. not skipped) a node that is in this list, then we know that we
-can't traverse this edge. Otherwise, we traverse the edge but make sure to skip any
-future node that was present in the "nodes_to_skip" list.
-
-The caller can decide to stop the path finding algorithm
-by returning a `ControlFlow::Break` from the `visit` closure.
-*/
 use std::collections::{BTreeSet, VecDeque};
 use std::iter::FromIterator;
 use std::ops::ControlFlow;
@@ -55,41 +12,30 @@ use crate::search::new::query_graph::QueryNode;
 use crate::search::new::small_bitmap::SmallBitmap;
 use crate::Result;
 
-/// Closure which processes a path found by the `PathVisitor`
 type VisitFn<'f, G> = &'f mut dyn FnMut(
-    // the path as a list of conditions
     &[Interned<<G as RankingRuleGraphTrait>::Condition>],
     &mut RankingRuleGraph<G>,
-    // a mutable reference to the DeadEndsCache, to update it in case the given
-    // path doesn't resolve to any valid document ids
     &mut DeadEndsCache<<G as RankingRuleGraphTrait>::Condition>,
 ) -> Result<ControlFlow<()>>;
 
-/// A structure which is kept but not updated during the traversal of the graph.
-/// It can however be updated by the `visit` closure once a valid path has been found.
 struct VisitorContext<'a, G: RankingRuleGraphTrait> {
     graph: &'a mut RankingRuleGraph<G>,
     all_costs_from_node: &'a MappedInterner<QueryNode, Vec<u64>>,
     dead_ends_cache: &'a mut DeadEndsCache<G::Condition>,
 }
 
-/// The internal state of the traversal algorithm
 struct VisitorState<G: RankingRuleGraphTrait> {
-    /// Budget from the current node to the end node
     remaining_cost: u64,
-    /// Previously visited conditions, in order.
+
     path: Vec<Interned<G::Condition>>,
-    /// Previously visited conditions, as an efficient and compact set.
+
     visited_conditions: SmallBitmap<G::Condition>,
-    /// Previously visited (ie not skipped) nodes, as an efficient and compact set.
     visited_nodes: SmallBitmap<QueryNode>,
-    /// The conditions that cannot be visited anymore
+
     forbidden_conditions: SmallBitmap<G::Condition>,
-    /// The nodes that cannot be visited anymore (they must be skipped)
-    nodes_to_skip: SmallBitmap<QueryNode>,
+    forbidden_conditions_to_nodes: SmallBitmap<QueryNode>,
 }
 
-/// See module documentation
 pub struct PathVisitor<'a, G: RankingRuleGraphTrait> {
     state: VisitorState<G>,
     ctx: VisitorContext<'a, G>,
@@ -110,13 +56,14 @@ impl<'a, G: RankingRuleGraphTrait> PathVisitor<'a, G> {
                 forbidden_conditions: SmallBitmap::for_interned_values_in(
                     &graph.conditions_interner,
                 ),
-                nodes_to_skip: SmallBitmap::for_interned_values_in(&graph.query_graph.nodes),
+                forbidden_conditions_to_nodes: SmallBitmap::for_interned_values_in(
+                    &graph.query_graph.nodes,
+                ),
             },
             ctx: VisitorContext { graph, all_costs_from_node, dead_ends_cache },
         }
     }
 
-    /// See module documentation
     pub fn visit_paths(mut self, visit: VisitFn<G>) -> Result<()> {
         let _ =
             self.state.visit_node(self.ctx.graph.query_graph.root_node, visit, &mut self.ctx)?;
@@ -125,31 +72,22 @@ impl<'a, G: RankingRuleGraphTrait> PathVisitor<'a, G> {
 }
 
 impl<G: RankingRuleGraphTrait> VisitorState<G> {
-    /// Visits a node: traverse all its valid conditional and unconditional edges.
-    ///
-    /// Returns ControlFlow::Break if the path finding algorithm should stop.
-    /// Returns whether a valid path was found from this node otherwise.
     fn visit_node(
         &mut self,
         from_node: Interned<QueryNode>,
         visit: VisitFn<G>,
         ctx: &mut VisitorContext<G>,
     ) -> Result<ControlFlow<(), bool>> {
-        // any valid path will be found from this point
-        // if a valid path was found, then we know that the DeadEndsCache may have been updated,
-        // and we will need to do more work to potentially backtrack
         let mut any_valid = false;
 
         let edges = ctx.graph.edges_of_node.get(from_node).clone();
         for edge_idx in edges.iter() {
-            // could be none if the edge was deleted
             let Some(edge) = ctx.graph.edges_store.get(edge_idx).clone() else { continue };
 
             if self.remaining_cost < edge.cost as u64 {
                 continue;
             }
             self.remaining_cost -= edge.cost as u64;
-
             let cf = match edge.condition {
                 Some(condition) => self.visit_condition(
                     condition,
@@ -181,10 +119,6 @@ impl<G: RankingRuleGraphTrait> VisitorState<G> {
         Ok(ControlFlow::Continue(any_valid))
     }
 
-    /// Visits an unconditional edge.
-    ///
-    /// Returns ControlFlow::Break if the path finding algorithm should stop.
-    /// Returns whether a valid path was found from this node otherwise.
     fn visit_no_condition(
         &mut self,
         dest_node: Interned<QueryNode>,
@@ -200,29 +134,20 @@ impl<G: RankingRuleGraphTrait> VisitorState<G> {
         {
             return Ok(ControlFlow::Continue(false));
         }
-        // We've reached the END node!
         if dest_node == ctx.graph.query_graph.end_node {
             let control_flow = visit(&self.path, ctx.graph, ctx.dead_ends_cache)?;
-            // We could change the return type of the visit closure such that the caller
-            // tells us whether the dead ends cache was updated or not.
-            // Alternatively, maybe the DeadEndsCache should have a generation number
-            // to it, so that we don't need to play with these booleans at all.
             match control_flow {
                 ControlFlow::Continue(_) => Ok(ControlFlow::Continue(true)),
                 ControlFlow::Break(_) => Ok(ControlFlow::Break(())),
             }
         } else {
-            let old_fbct = self.nodes_to_skip.clone();
-            self.nodes_to_skip.union(edge_new_nodes_to_skip);
+            let old_fbct = self.forbidden_conditions_to_nodes.clone();
+            self.forbidden_conditions_to_nodes.union(edge_new_nodes_to_skip);
             let cf = self.visit_node(dest_node, visit, ctx)?;
-            self.nodes_to_skip = old_fbct;
+            self.forbidden_conditions_to_nodes = old_fbct;
             Ok(cf)
         }
     }
-    /// Visits a conditional edge.
-    ///
-    /// Returns ControlFlow::Break if the path finding algorithm should stop.
-    /// Returns whether a valid path was found from this node otherwise.
     fn visit_condition(
         &mut self,
         condition: Interned<G::Condition>,
@@ -234,7 +159,7 @@ impl<G: RankingRuleGraphTrait> VisitorState<G> {
         assert!(dest_node != ctx.graph.query_graph.end_node);
 
         if self.forbidden_conditions.contains(condition)
-            || self.nodes_to_skip.contains(dest_node)
+            || self.forbidden_conditions_to_nodes.contains(dest_node)
             || edge_new_nodes_to_skip.intersects(&self.visited_nodes)
         {
             return Ok(ControlFlow::Continue(false));
@@ -255,19 +180,19 @@ impl<G: RankingRuleGraphTrait> VisitorState<G> {
         self.visited_nodes.insert(dest_node);
         self.visited_conditions.insert(condition);
 
-        let old_forb_cond = self.forbidden_conditions.clone();
+        let old_fc = self.forbidden_conditions.clone();
         if let Some(next_forbidden) =
             ctx.dead_ends_cache.forbidden_conditions_after_prefix(self.path.iter().copied())
         {
             self.forbidden_conditions.union(&next_forbidden);
         }
-        let old_nodes_to_skip = self.nodes_to_skip.clone();
-        self.nodes_to_skip.union(edge_new_nodes_to_skip);
+        let old_fctn = self.forbidden_conditions_to_nodes.clone();
+        self.forbidden_conditions_to_nodes.union(edge_new_nodes_to_skip);
 
         let cf = self.visit_node(dest_node, visit, ctx)?;
 
-        self.nodes_to_skip = old_nodes_to_skip;
-        self.forbidden_conditions = old_forb_cond;
+        self.forbidden_conditions_to_nodes = old_fctn;
+        self.forbidden_conditions = old_fc;
 
         self.visited_conditions.remove(condition);
         self.visited_nodes.remove(dest_node);

milli/src/search/new/ranking_rule_graph/condition_docids_cache.rs

@@ -9,8 +9,12 @@ use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::SearchContext;
 use crate::Result;
 
+// TODO: give a generation to each universe, then be able to get the exact
+// delta of docids between two universes of different generations!
+
 /// A cache storing the document ids associated with each ranking rule edge
 pub struct ConditionDocIdsCache<G: RankingRuleGraphTrait> {
+    // TOOD: should be a mapped interner?
     pub cache: FxHashMap<Interned<G::Condition>, ComputedCondition>,
     _phantom: PhantomData<G>,
 }
@@ -50,7 +54,7 @@ impl<G: RankingRuleGraphTrait> ConditionDocIdsCache<G> {
         }
         let condition = graph.conditions_interner.get_mut(interned_condition);
         let computed = G::resolve_condition(ctx, condition, universe)?;
-        // Can we put an assert here for computed.universe_len == universe.len() ?
+        // TODO: if computed.universe_len != universe.len() ?
         let _ = self.cache.insert(interned_condition, computed);
         let computed = &self.cache[&interned_condition];
         Ok(computed)

milli/src/search/new/ranking_rule_graph/dead_ends_cache.rs

@@ -2,7 +2,6 @@ use crate::search::new::interner::{FixedSizeInterner, Interned};
 use crate::search::new::small_bitmap::SmallBitmap;
 
 pub struct DeadEndsCache<T> {
-    // conditions and next could/should be part of the same vector
     conditions: Vec<Interned<T>>,
     next: Vec<Self>,
     pub forbidden: SmallBitmap<T>,
@@ -28,7 +27,7 @@ impl<T> DeadEndsCache<T> {
         self.forbidden.insert(condition);
     }
 
-    fn advance(&mut self, condition: Interned<T>) -> Option<&mut Self> {
+    pub fn advance(&mut self, condition: Interned<T>) -> Option<&mut Self> {
         if let Some(idx) = self.conditions.iter().position(|c| *c == condition) {
             Some(&mut self.next[idx])
         } else {

milli/src/search/new/ranking_rule_graph/exactness/mod.rs

@@ -1,6 +1,7 @@
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::{ExactTerm, LocatedQueryTermSubset};
 use crate::search::new::resolve_query_graph::compute_query_term_subset_docids;
@@ -84,4 +85,8 @@ impl RankingRuleGraphTrait for ExactnessGraph {
 
         Ok(vec![(0, exact_condition), (dest_node.term_ids.len() as u32, skip_condition)])
     }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Exactness(rank)
+    }
 }

milli/src/search/new/ranking_rule_graph/fid/mod.rs

@@ -2,6 +2,7 @@ use fxhash::FxHashSet;
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::resolve_query_graph::compute_query_term_subset_docids_within_field_id;
@@ -68,13 +69,47 @@ impl RankingRuleGraphTrait for FidGraph {
         }
 
         let mut edges = vec![];
-        for fid in all_fields {
+        for fid in all_fields.iter().copied() {
+            // TODO: We can improve performances and relevancy by storing
+            //       the term subsets associated to each field ids fetched.
             edges.push((
-                fid as u32 * term.term_ids.len() as u32,
-                conditions_interner.insert(FidCondition { term: term.clone(), fid }),
+                fid as u32 * term.term_ids.len() as u32, // TODO improve the fid score i.e. fid^10.
+                conditions_interner.insert(FidCondition {
+                    term: term.clone(), // TODO remove this ugly clone
+                    fid,
+                }),
             ));
         }
 
+        // always lookup the max_fid if we don't already and add an artificial condition for max scoring
+        let max_fid: Option<u16> = {
+            if let Some(max_fid) = ctx
+                .index
+                .searchable_fields_ids(ctx.txn)?
+                .map(|field_ids| field_ids.into_iter().max())
+            {
+                max_fid
+            } else {
+                ctx.index.fields_ids_map(ctx.txn)?.ids().max()
+            }
+        };
+
+        if let Some(max_fid) = max_fid {
+            if !all_fields.contains(&max_fid) {
+                edges.push((
+                    max_fid as u32 * term.term_ids.len() as u32, // TODO improve the fid score i.e. fid^10.
+                    conditions_interner.insert(FidCondition {
+                        term: term.clone(), // TODO remove this ugly clone
+                        fid: max_fid,
+                    }),
+                ));
+            }
+        }
+
         Ok(edges)
     }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Fid(rank)
+    }
 }

milli/src/search/new/ranking_rule_graph/mod.rs

@@ -41,6 +41,7 @@ use super::interner::{DedupInterner, FixedSizeInterner, Interned, MappedInterner
 use super::query_term::LocatedQueryTermSubset;
 use super::small_bitmap::SmallBitmap;
 use super::{QueryGraph, QueryNode, SearchContext};
+use crate::score_details::{Rank, ScoreDetails};
 use crate::Result;
 
 pub struct ComputedCondition {
@@ -110,6 +111,9 @@ pub trait RankingRuleGraphTrait: Sized + 'static {
         source_node: Option<&LocatedQueryTermSubset>,
         dest_node: &LocatedQueryTermSubset,
     ) -> Result<Vec<(u32, Interned<Self::Condition>)>>;
+
+    /// Convert the rank of a path to its corresponding score for the ranking rule
+    fn rank_to_score(rank: Rank) -> ScoreDetails;
 }
 
 /// The graph used by graph-based ranking rules.

milli/src/search/new/ranking_rule_graph/position/mod.rs

@@ -2,6 +2,7 @@ use fxhash::{FxHashMap, FxHashSet};
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::resolve_query_graph::compute_query_term_subset_docids_within_position;
@@ -94,14 +95,31 @@ impl RankingRuleGraphTrait for PositionGraph {
         let mut edges = vec![];
 
         for (cost, positions) in positions_for_costs {
+            // TODO: We can improve performances and relevancy by storing
+            //       the term subsets associated to each position fetched
             edges.push((
                 cost,
-                conditions_interner.insert(PositionCondition { term: term.clone(), positions }),
+                conditions_interner.insert(PositionCondition {
+                    term: term.clone(), // TODO remove this ugly clone
+                    positions,
+                }),
             ));
         }
 
+        // artificial empty condition for computing max cost
+        let max_cost = term.term_ids.len() as u32 * 10;
+        edges.push((
+            max_cost,
+            conditions_interner
+                .insert(PositionCondition { term: term.clone(), positions: Vec::default() }),
+        ));
+
         Ok(edges)
     }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Position(rank)
+    }
 }
 
 fn cost_from_position(sum_positions: u32) -> u32 {

milli/src/search/new/ranking_rule_graph/proximity/compute_docids.rs

@@ -65,6 +65,13 @@ pub fn compute_docids(
         }
     }
 
+    // TODO: add safeguard in case the cartesian product is too large!
+    // even if we restrict the word derivations to a maximum of 100, the size of the
+    // caterisan product could reach a maximum of 10_000 derivations, which is way too much.
+    // Maybe prioritise the product of zero typo derivations, then the product of zero-typo/one-typo
+    // + one-typo/zero-typo, then one-typo/one-typo, then ... until an arbitrary limit has been
+    // reached
+
     for (left_phrase, left_word) in last_words_of_term_derivations(ctx, &left_term.term_subset)? {
         // Before computing the edges, check that the left word and left phrase
         // aren't disjoint with the universe, but only do it if there is more than
@@ -104,6 +111,8 @@ pub fn compute_docids(
     Ok(ComputedCondition {
         docids,
         universe_len: universe.len(),
+        // TODO: think about whether we want to reduce the subset,
+        // we probably should!
         start_term_subset: Some(left_term.clone()),
         end_term_subset: right_term.clone(),
     })
@@ -194,7 +203,12 @@ fn compute_non_prefix_edges(
             *docids |= new_docids;
         }
     }
-    if backward_proximity >= 1 && left_phrase.is_none() && right_phrase.is_none() {
+    if backward_proximity >= 1
+            // TODO: for now, we don't do any swapping when either term is a phrase
+            // but maybe we should. We'd need to look at the first/last word of the phrase
+            // depending on the context.
+            && left_phrase.is_none() && right_phrase.is_none()
+    {
         if let Some(new_docids) =
             ctx.get_db_word_pair_proximity_docids(word2, word1, backward_proximity)?
         {

milli/src/search/new/ranking_rule_graph/proximity/mod.rs

@@ -4,6 +4,7 @@ pub mod compute_docids;
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::SearchContext;
@@ -36,4 +37,8 @@ impl RankingRuleGraphTrait for ProximityGraph {
     ) -> Result<Vec<(u32, Interned<Self::Condition>)>> {
         build::build_edges(ctx, conditions_interner, source_term, dest_term)
     }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Proximity(rank)
+    }
 }

milli/src/search/new/ranking_rule_graph/typo/mod.rs

@@ -1,6 +1,7 @@
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{self, Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::resolve_query_graph::compute_query_term_subset_docids;
@@ -75,4 +76,8 @@ impl RankingRuleGraphTrait for TypoGraph {
         }
         Ok(edges)
     }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Typo(score_details::Typo::from_rank(rank))
+    }
 }

milli/src/search/new/ranking_rule_graph/words/mod.rs

@@ -1,6 +1,7 @@
 use roaring::RoaringBitmap;
 
 use super::{ComputedCondition, RankingRuleGraphTrait};
+use crate::score_details::{self, Rank, ScoreDetails};
 use crate::search::new::interner::{DedupInterner, Interned};
 use crate::search::new::query_term::LocatedQueryTermSubset;
 use crate::search::new::resolve_query_graph::compute_query_term_subset_docids;
@@ -41,9 +42,10 @@ impl RankingRuleGraphTrait for WordsGraph {
         _from: Option<&LocatedQueryTermSubset>,
         to_term: &LocatedQueryTermSubset,
     ) -> Result<Vec<(u32, Interned<Self::Condition>)>> {
-        Ok(vec![(
-            to_term.term_ids.len() as u32,
-            conditions_interner.insert(WordsCondition { term: to_term.clone() }),
-        )])
+        Ok(vec![(0, conditions_interner.insert(WordsCondition { term: to_term.clone() }))])
+    }
+
+    fn rank_to_score(rank: Rank) -> ScoreDetails {
+        ScoreDetails::Words(score_details::Words::from_rank(rank))
     }
 }

milli/src/search/new/ranking_rules.rs

@@ -2,6 +2,7 @@ use roaring::RoaringBitmap;
 
 use super::logger::SearchLogger;
 use super::{QueryGraph, SearchContext};
+use crate::score_details::ScoreDetails;
 use crate::Result;
 
 /// An internal trait implemented by only [`PlaceholderQuery`] and [`QueryGraph`]
@@ -66,4 +67,6 @@ pub struct RankingRuleOutput<Q> {
     pub query: Q,
     /// The allowed candidates for the child ranking rule
     pub candidates: RoaringBitmap,
+    /// The score for the candidates of the current bucket
+    pub score: ScoreDetails,
 }

milli/src/search/new/resolve_query_graph.rs

@@ -33,6 +33,8 @@ pub fn compute_query_term_subset_docids(
     ctx: &mut SearchContext,
     term: &QueryTermSubset,
 ) -> Result<RoaringBitmap> {
+    // TODO Use the roaring::MultiOps trait
+
     let mut docids = RoaringBitmap::new();
     for word in term.all_single_words_except_prefix_db(ctx)? {
         if let Some(word_docids) = ctx.word_docids(word)? {
@@ -57,6 +59,8 @@ pub fn compute_query_term_subset_docids_within_field_id(
     term: &QueryTermSubset,
     fid: u16,
 ) -> Result<RoaringBitmap> {
+    // TODO Use the roaring::MultiOps trait
+
     let mut docids = RoaringBitmap::new();
     for word in term.all_single_words_except_prefix_db(ctx)? {
         if let Some(word_fid_docids) = ctx.get_db_word_fid_docids(word.interned(), fid)? {
@@ -67,6 +71,7 @@ pub fn compute_query_term_subset_docids_within_field_id(
     for phrase in term.all_phrases(ctx)? {
         // There may be false positives when resolving a phrase, so we're not
         // guaranteed that all of its words are within a single fid.
+        // TODO: fix this?
         if let Some(word) = phrase.words(ctx).iter().flatten().next() {
             if let Some(word_fid_docids) = ctx.get_db_word_fid_docids(*word, fid)? {
                 docids |= ctx.get_phrase_docids(phrase)? & word_fid_docids;
@@ -90,6 +95,7 @@ pub fn compute_query_term_subset_docids_within_position(
     term: &QueryTermSubset,
     position: u16,
 ) -> Result<RoaringBitmap> {
+    // TODO Use the roaring::MultiOps trait
     let mut docids = RoaringBitmap::new();
     for word in term.all_single_words_except_prefix_db(ctx)? {
         if let Some(word_position_docids) =
@@ -102,6 +108,7 @@ pub fn compute_query_term_subset_docids_within_position(
     for phrase in term.all_phrases(ctx)? {
         // It's difficult to know the expected position of the words in the phrase,
         // so instead we just check the first one.
+        // TODO: fix this?
         if let Some(word) = phrase.words(ctx).iter().flatten().next() {
             if let Some(word_position_docids) = ctx.get_db_word_position_docids(*word, position)? {
                 docids |= ctx.get_phrase_docids(phrase)? & word_position_docids
@@ -125,6 +132,9 @@ pub fn compute_query_graph_docids(
     q: &QueryGraph,
     universe: &RoaringBitmap,
 ) -> Result<RoaringBitmap> {
+    // TODO: there must be a faster way to compute this big
+    // roaring bitmap expression
+
     let mut nodes_resolved = SmallBitmap::for_interned_values_in(&q.nodes);
     let mut path_nodes_docids = q.nodes.map(|_| RoaringBitmap::new());
 

milli/src/search/new/sort.rs

@@ -1,9 +1,11 @@
+use heed::BytesDecode;
 use roaring::RoaringBitmap;
 
 use super::logger::SearchLogger;
 use super::{RankingRule, RankingRuleOutput, RankingRuleQueryTrait, SearchContext};
-use crate::heed_codec::facet::FacetGroupKeyCodec;
-use crate::heed_codec::ByteSliceRefCodec;
+use crate::heed_codec::facet::{FacetGroupKeyCodec, OrderedF64Codec};
+use crate::heed_codec::{ByteSliceRefCodec, StrRefCodec};
+use crate::score_details::{self, ScoreDetails};
 use crate::search::facet::{ascending_facet_sort, descending_facet_sort};
 use crate::{FieldId, Index, Result};
 
@@ -67,7 +69,7 @@ impl<'ctx, Query> Sort<'ctx, Query> {
 impl<'ctx, Query: RankingRuleQueryTrait> RankingRule<'ctx, Query> for Sort<'ctx, Query> {
     fn id(&self) -> String {
         let Self { field_name, is_ascending, .. } = self;
-        format!("{field_name}:{}", if *is_ascending { "asc" } else { "desc " })
+        format!("{field_name}:{}", if *is_ascending { "asc" } else { "desc" })
     }
     fn start_iteration(
         &mut self,
@@ -118,12 +120,43 @@ impl<'ctx, Query: RankingRuleQueryTrait> RankingRule<'ctx, Query> for Sort<'ctx,
 
                     (itertools::Either::Right(number_iter), itertools::Either::Right(string_iter))
                 };
+                let number_iter = number_iter.map(|r| -> Result<_> {
+                    let (docids, bytes) = r?;
+                    Ok((
+                        docids,
+                        serde_json::Value::Number(
+                            serde_json::Number::from_f64(
+                                OrderedF64Codec::bytes_decode(bytes).expect("some number"),
+                            )
+                            .expect("too big float"),
+                        ),
+                    ))
+                });
+                let string_iter = string_iter.map(|r| -> Result<_> {
+                    let (docids, bytes) = r?;
+                    Ok((
+                        docids,
+                        serde_json::Value::String(
+                            StrRefCodec::bytes_decode(bytes).expect("some string").to_owned(),
+                        ),
+                    ))
+                });
 
                 let query_graph = parent_query.clone();
+                let ascending = self.is_ascending;
+                let field_name = self.field_name.clone();
                 RankingRuleOutputIterWrapper::new(Box::new(number_iter.chain(string_iter).map(
                     move |r| {
-                        let (docids, _) = r?;
-                        Ok(RankingRuleOutput { query: query_graph.clone(), candidates: docids })
+                        let (docids, value) = r?;
+                        Ok(RankingRuleOutput {
+                            query: query_graph.clone(),
+                            candidates: docids,
+                            score: ScoreDetails::Sort(score_details::Sort {
+                                field_name: field_name.clone(),
+                                ascending,
+                                value,
+                            }),
+                        })
                     },
                 )))
             }
@@ -141,12 +174,24 @@ impl<'ctx, Query: RankingRuleQueryTrait> RankingRule<'ctx, Query> for Sort<'ctx,
         universe: &RoaringBitmap,
     ) -> Result<Option<RankingRuleOutput<Query>>> {
         let iter = self.iter.as_mut().unwrap();
+        // TODO: we should make use of the universe in the function below
+        // good for correctness, but ideally iter.next_bucket would take the current universe into account,
+        // as right now it could return buckets that don't intersect with the universe, meaning we will make many
+        // unneeded calls.
         if let Some(mut bucket) = iter.next_bucket()? {
             bucket.candidates &= universe;
             Ok(Some(bucket))
         } else {
             let query = self.original_query.as_ref().unwrap().clone();
-            Ok(Some(RankingRuleOutput { query, candidates: universe.clone() }))
+            Ok(Some(RankingRuleOutput {
+                query,
+                candidates: universe.clone(),
+                score: ScoreDetails::Sort(score_details::Sort {
+                    field_name: self.field_name.clone(),
+                    ascending: self.is_ascending,
+                    value: serde_json::Value::Null,
+                }),
+            }))
         }
     }
 

milli/src/search/new/tests/distinct.rs

@@ -527,7 +527,7 @@ fn test_distinct_all_candidates() {
     let SearchResult { documents_ids, candidates, .. } = s.execute().unwrap();
     let candidates = candidates.iter().collect::<Vec<_>>();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[14, 26, 4, 7, 17, 23, 1, 19, 25, 8, 20, 24]");
-    // This is incorrect, but unfortunately impossible to do better efficiently.
+    // TODO: this is incorrect!
     insta::assert_snapshot!(format!("{candidates:?}"), @"[1, 4, 7, 8, 14, 17, 19, 20, 23, 24, 25, 26]");
 }
 

milli/src/search/new/tests/proximity.rs

@@ -122,11 +122,11 @@ fn create_edge_cases_index() -> TempIndex {
             sta stb stc ste stf stg sth sti stj stk stl stm stn sto stp stq str stst stt stu stv stw stx sty stz
             "
         },
-        // The next 5 documents lay out a trap with the split word, phrase search, or synonym `sun flower`.
-        // If the search query is "sunflower", the split word "Sun Flower" will match some documents.
+        // The next 5 documents lay out a trap with the split word, phrase search, or synonym `sun flower`. 
+        // If the search query is "sunflower", the split word "Sun Flower" will match some documents. 
         // If the query is `sunflower wilting`, then we should make sure that
-        // the proximity condition `flower wilting: sprx N` also comes with the condition
-        // `sun wilting: sprx N+1`, but this is not the exact condition we use for now.
+        // the sprximity condition `flower wilting: sprx N` also comes with the condition
+        // `sun wilting: sprx N+1`. TODO: this is not the exact condition we use for now. 
         // We only check that the phrase `sun flower` exists and `flower wilting: sprx N`, which
         // is better than nothing but not the best.
         {
@@ -139,7 +139,7 @@ fn create_edge_cases_index() -> TempIndex {
         },
         {
             "id": 3,
-            // This document matches the query `sunflower wilting`, but the sprximity condition
+            // This document matches the query `sunflower wilting`, but the sprximity condition 
             // between `sunflower` and `wilting` cannot be through the split-word `Sun Flower`
             // which would reduce to only `flower` and `wilting` being in sprximity.
             "text": "A flower wilting under the sun, unlike a sunflower"
@@ -299,7 +299,7 @@ fn test_proximity_split_word() {
     let SearchResult { documents_ids, .. } = s.execute().unwrap();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[2, 4, 5, 1, 3]");
     let texts = collect_field_values(&index, &txn, "text", &documents_ids);
-    // "2" and "4" should be swapped ideally
+    // TODO: "2" and "4" should be swapped ideally
     insta::assert_debug_snapshot!(texts, @r###"
     [
         "\"Sun Flower sounds like the title of a painting, maybe about a flower wilting under the heat.\"",
@@ -316,7 +316,7 @@ fn test_proximity_split_word() {
     let SearchResult { documents_ids, .. } = s.execute().unwrap();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[2, 4, 1]");
     let texts = collect_field_values(&index, &txn, "text", &documents_ids);
-    // "2" and "4" should be swapped ideally
+    // TODO: "2" and "4" should be swapped ideally
     insta::assert_debug_snapshot!(texts, @r###"
     [
         "\"Sun Flower sounds like the title of a painting, maybe about a flower wilting under the heat.\"",
@@ -341,7 +341,7 @@ fn test_proximity_split_word() {
     let SearchResult { documents_ids, .. } = s.execute().unwrap();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[2, 4, 1]");
     let texts = collect_field_values(&index, &txn, "text", &documents_ids);
-    // "2" and "4" should be swapped ideally
+    // TODO: "2" and "4" should be swapped ideally
     insta::assert_debug_snapshot!(texts, @r###"
     [
         "\"Sun Flower sounds like the title of a painting, maybe about a flower wilting under the heat.\"",

milli/src/search/new/tests/proximity_typo.rs

@@ -2,8 +2,9 @@
 This module tests the interactions between the proximity and typo ranking rules.
 
 The proximity ranking rule should transform the query graph such that it
-only contains the word pairs that it used to compute its bucket, but this is not currently
-implemented.
+only contains the word pairs that it used to compute its bucket.
+
+TODO: This is not currently implemented.
 */
 
 use crate::index::tests::TempIndex;
@@ -63,7 +64,7 @@ fn test_trap_basic() {
     let SearchResult { documents_ids, .. } = s.execute().unwrap();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[0, 1]");
     let texts = collect_field_values(&index, &txn, "text", &documents_ids);
-    // This is incorrect, 1 should come before 0
+    // TODO: this is incorrect, 1 should come before 0
     insta::assert_debug_snapshot!(texts, @r###"
     [
         "\"summer. holiday. sommer holidty\"",

milli/src/search/new/tests/typo.rs

@@ -571,8 +571,8 @@ fn test_typo_synonyms() {
     s.terms_matching_strategy(TermsMatchingStrategy::All);
     s.query("the fast brownish fox jumps over the lackadaisical dog");
 
-    // The interaction of ngrams + synonyms means that the multi-word synonyms end up having a typo cost.
-    // This is probably not what we want.
+    // TODO: is this correct? interaction of ngrams + synonyms means that the
+    // multi-word synonyms end up having a typo cost. This is probably not what we want.
     let SearchResult { documents_ids, .. } = s.execute().unwrap();
     insta::assert_snapshot!(format!("{documents_ids:?}"), @"[21, 0, 22]");
     let texts = collect_field_values(&index, &txn, "text", &documents_ids);

milli/src/snapshot_tests.rs

@@ -318,7 +318,7 @@ pub fn snap_field_distributions(index: &Index) -> String {
     let rtxn = index.read_txn().unwrap();
     let mut snap = String::new();
     for (field, count) in index.field_distribution(&rtxn).unwrap() {
-        writeln!(&mut snap, "{field:<16} {count:<6} |").unwrap();
+        writeln!(&mut snap, "{field:<16} {count:<6}").unwrap();
     }
     snap
 }
@@ -328,7 +328,7 @@ pub fn snap_fields_ids_map(index: &Index) -> String {
     let mut snap = String::new();
     for field_id in fields_ids_map.ids() {
         let name = fields_ids_map.name(field_id).unwrap();
-        writeln!(&mut snap, "{field_id:<3} {name:<16} |").unwrap();
+        writeln!(&mut snap, "{field_id:<3} {name:<16}").unwrap();
     }
     snap
 }

milli/src/snapshots/index.rs/initial_field_distribution/1/field_distribution.snap

@@ -1,7 +1,7 @@
 ---
 source: milli/src/index.rs
 ---
-age              1      |
-id               2      |
-name             2      |
+age              1     
+id               2     
+name             2     
 

milli/src/snapshots/index.rs/initial_field_distribution/field_distribution.snap

@@ -1,7 +1,7 @@
 ---
 source: milli/src/index.rs
 ---
-age              1      |
-id               2      |
-name             2      |
+age              1     
+id               2     
+name             2     
 

milli/src/update/delete_documents.rs

@@ -71,6 +71,7 @@ impl std::fmt::Display for DeletionStrategy {
 pub(crate) struct DetailedDocumentDeletionResult {
     pub deleted_documents: u64,
     pub remaining_documents: u64,
+    pub soft_deletion_used: bool,
 }
 
 impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
@@ -107,8 +108,11 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         Some(docid)
     }
     pub fn execute(self) -> Result<DocumentDeletionResult> {
-        let DetailedDocumentDeletionResult { deleted_documents, remaining_documents } =
-            self.execute_inner()?;
+        let DetailedDocumentDeletionResult {
+            deleted_documents,
+            remaining_documents,
+            soft_deletion_used: _,
+        } = self.execute_inner()?;
 
         Ok(DocumentDeletionResult { deleted_documents, remaining_documents })
     }
@@ -129,6 +133,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
             return Ok(DetailedDocumentDeletionResult {
                 deleted_documents: 0,
                 remaining_documents: 0,
+                soft_deletion_used: false,
             });
         }
 
@@ -144,6 +149,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
             return Ok(DetailedDocumentDeletionResult {
                 deleted_documents: current_documents_ids_len,
                 remaining_documents,
+                soft_deletion_used: false,
             });
         }
 
@@ -212,6 +218,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
             return Ok(DetailedDocumentDeletionResult {
                 deleted_documents: self.to_delete_docids.len(),
                 remaining_documents: documents_ids.len(),
+                soft_deletion_used: true,
             });
         }
 
@@ -434,6 +441,7 @@ impl<'t, 'u, 'i> DeleteDocuments<'t, 'u, 'i> {
         Ok(DetailedDocumentDeletionResult {
             deleted_documents: self.to_delete_docids.len(),
             remaining_documents: documents_ids.len(),
+            soft_deletion_used: false,
         })
     }
 

milli/src/update/index_documents/helpers/clonable_mmap.rs

@@ -2,7 +2,7 @@ use std::sync::Arc;
 
 use memmap2::Mmap;
 
-/// Wrapper around Mmap allowing to virtually clone grenad-chunks
+/// Wrapper around Mmap allowing to virtualy clone grenad-chunks
 /// in a parallel process like the indexing.
 #[derive(Debug, Clone)]
 pub struct ClonableMmap {

milli/src/update/index_documents/mod.rs

@@ -236,7 +236,7 @@ where
             primary_key,
             fields_ids_map,
             field_distribution,
-            new_external_documents_ids,
+            mut external_documents_ids,
             new_documents_ids,
             replaced_documents_ids,
             documents_count,
@@ -363,6 +363,9 @@ where
             deletion_builder.delete_documents(&replaced_documents_ids);
             let deleted_documents_result = deletion_builder.execute_inner()?;
             debug!("{} documents actually deleted", deleted_documents_result.deleted_documents);
+            if !deleted_documents_result.soft_deletion_used {
+                external_documents_ids.delete_soft_deleted_documents_ids_from_fsts()?;
+            }
         }
 
         let index_documents_ids = self.index.documents_ids(self.wtxn)?;
@@ -442,9 +445,6 @@ where
         self.index.put_primary_key(self.wtxn, &primary_key)?;
 
         // We write the external documents ids into the main database.
-        let mut external_documents_ids = self.index.external_documents_ids(self.wtxn)?;
-        external_documents_ids.insert_ids(&new_external_documents_ids)?;
-        let external_documents_ids = external_documents_ids.into_static();
         self.index.put_external_documents_ids(self.wtxn, &external_documents_ids)?;
 
         let all_documents_ids = index_documents_ids | new_documents_ids;
@@ -2514,170 +2514,4 @@ mod tests {
         db_snap!(index, word_fid_docids, 3, @"4c2e2a1832e5802796edc1638136d933");
         db_snap!(index, word_position_docids, 3, @"74f556b91d161d997a89468b4da1cb8f");
     }
-
-    #[test]
-    fn reproduce_the_bug() {
-        /*
-            [milli/examples/fuzz.rs:69] &batches = [
-            Batch(
-                [
-                    AddDoc(
-                        { "id": 1, "doggo": "bernese" }, => internal 0
-                    ),
-                ],
-            ),
-            Batch(
-                [
-                    DeleteDoc(
-                        1, => delete internal 0
-                    ),
-                    AddDoc(
-                        { "id": 0, "catto": "jorts" }, => internal 1
-                    ),
-                ],
-            ),
-            Batch(
-                [
-                    AddDoc(
-                        { "id": 1, "catto": "jorts" }, => internal 2
-                    ),
-                ],
-            ),
-        ]
-        */
-        let mut index = TempIndex::new();
-        index.index_documents_config.deletion_strategy = DeletionStrategy::AlwaysHard;
-
-        // START OF BATCH
-
-        println!("--- ENTERING BATCH 1");
-
-        let mut wtxn = index.write_txn().unwrap();
-
-        let builder = IndexDocuments::new(
-            &mut wtxn,
-            &index,
-            &index.indexer_config,
-            index.index_documents_config.clone(),
-            |_| (),
-            || false,
-        )
-        .unwrap();
-
-        // OP
-
-        let documents = documents!([
-            { "id": 1, "doggo": "bernese" },
-        ]);
-        let (builder, added) = builder.add_documents(documents).unwrap();
-        insta::assert_display_snapshot!(added.unwrap(), @"1");
-
-        // FINISHING
-        let addition = builder.execute().unwrap();
-        insta::assert_debug_snapshot!(addition, @r###"
-        DocumentAdditionResult {
-            indexed_documents: 1,
-            number_of_documents: 1,
-        }
-        "###);
-        wtxn.commit().unwrap();
-
-        db_snap!(index, documents, @r###"
-        {"id":1,"doggo":"bernese"}
-        "###);
-        db_snap!(index, external_documents_ids, @r###"
-        soft:
-        hard:
-        1                        0
-        "###);
-
-        // A first batch of documents has been inserted
-
-        // BATCH 2
-
-        println!("--- ENTERING BATCH 2");
-
-        let mut wtxn = index.write_txn().unwrap();
-
-        let builder = IndexDocuments::new(
-            &mut wtxn,
-            &index,
-            &index.indexer_config,
-            index.index_documents_config.clone(),
-            |_| (),
-            || false,
-        )
-        .unwrap();
-
-        let (builder, removed) = builder.remove_documents(vec![S("1")]).unwrap();
-        insta::assert_display_snapshot!(removed.unwrap(), @"1");
-
-        let documents = documents!([
-            { "id": 0, "catto": "jorts" },
-        ]);
-        let (builder, added) = builder.add_documents(documents).unwrap();
-        insta::assert_display_snapshot!(added.unwrap(), @"1");
-
-        let addition = builder.execute().unwrap();
-        insta::assert_debug_snapshot!(addition, @r###"
-        DocumentAdditionResult {
-            indexed_documents: 1,
-            number_of_documents: 1,
-        }
-        "###);
-        wtxn.commit().unwrap();
-
-        db_snap!(index, documents, @r###"
-        {"id":0,"catto":"jorts"}
-        "###);
-
-        db_snap!(index, external_documents_ids, @r###"
-        soft:
-        hard:
-        0                        1
-        "###);
-
-        db_snap!(index, soft_deleted_documents_ids, @"[]");
-
-        // BATCH 3
-
-        println!("--- ENTERING BATCH 3");
-
-        let mut wtxn = index.write_txn().unwrap();
-
-        let builder = IndexDocuments::new(
-            &mut wtxn,
-            &index,
-            &index.indexer_config,
-            index.index_documents_config.clone(),
-            |_| (),
-            || false,
-        )
-        .unwrap();
-
-        let documents = documents!([
-            { "id": 1, "catto": "jorts" },
-        ]);
-        let (builder, added) = builder.add_documents(documents).unwrap();
-        insta::assert_display_snapshot!(added.unwrap(), @"1");
-
-        let addition = builder.execute().unwrap();
-        insta::assert_debug_snapshot!(addition, @r###"
-        DocumentAdditionResult {
-            indexed_documents: 1,
-            number_of_documents: 2,
-        }
-        "###);
-        wtxn.commit().unwrap();
-
-        db_snap!(index, documents, @r###"
-        {"id":1,"catto":"jorts"}
-        {"id":0,"catto":"jorts"}
-        "###);
-
-        // Ensuring all the returned IDs actually exists
-        let rtxn = index.read_txn().unwrap();
-        let res = index.search(&rtxn).execute().unwrap();
-        index.documents(&rtxn, res.documents_ids).unwrap();
-    }
 }

milli/src/update/index_documents/transform.rs

@@ -21,14 +21,15 @@ use crate::error::{Error, InternalError, UserError};
 use crate::index::{db_name, main_key};
 use crate::update::{AvailableDocumentsIds, ClearDocuments, UpdateIndexingStep};
 use crate::{
-    FieldDistribution, FieldId, FieldIdMapMissingEntry, FieldsIdsMap, Index, Result, BEU32,
+    ExternalDocumentsIds, FieldDistribution, FieldId, FieldIdMapMissingEntry, FieldsIdsMap, Index,
+    Result, BEU32,
 };
 
 pub struct TransformOutput {
     pub primary_key: String,
     pub fields_ids_map: FieldsIdsMap,
     pub field_distribution: FieldDistribution,
-    pub new_external_documents_ids: fst::Map<Cow<'static, [u8]>>,
+    pub external_documents_ids: ExternalDocumentsIds<'static>,
     pub new_documents_ids: RoaringBitmap,
     pub replaced_documents_ids: RoaringBitmap,
     pub documents_count: usize,
@@ -567,6 +568,8 @@ impl<'a, 'i> Transform<'a, 'i> {
             }))?
             .to_string();
 
+        let mut external_documents_ids = self.index.external_documents_ids(wtxn)?;
+
         // We create a final writer to write the new documents in order from the sorter.
         let mut writer = create_writer(
             self.indexer_settings.chunk_compression_type,
@@ -648,12 +651,13 @@ impl<'a, 'i> Transform<'a, 'i> {
             fst_new_external_documents_ids_builder.insert(key, value)
         })?;
         let new_external_documents_ids = fst_new_external_documents_ids_builder.into_map();
+        external_documents_ids.insert_ids(&new_external_documents_ids)?;
 
         Ok(TransformOutput {
             primary_key,
             fields_ids_map: self.fields_ids_map,
             field_distribution,
-            new_external_documents_ids: new_external_documents_ids.map_data(Cow::Owned).unwrap(),
+            external_documents_ids: external_documents_ids.into_static(),
             new_documents_ids: self.new_documents_ids,
             replaced_documents_ids: self.replaced_documents_ids,
             documents_count: self.documents_count,
@@ -687,8 +691,7 @@ impl<'a, 'i> Transform<'a, 'i> {
         let new_external_documents_ids = {
             let mut external_documents_ids = self.index.external_documents_ids(wtxn)?;
             external_documents_ids.delete_soft_deleted_documents_ids_from_fsts()?;
-            // This call should be free and can't fail since the previous method merged both fsts.
-            external_documents_ids.into_static().to_fst()?.into_owned()
+            external_documents_ids
         };
 
         let documents_ids = self.index.documents_ids(wtxn)?;
@@ -773,7 +776,7 @@ impl<'a, 'i> Transform<'a, 'i> {
             primary_key,
             fields_ids_map: new_fields_ids_map,
             field_distribution,
-            new_external_documents_ids,
+            external_documents_ids: new_external_documents_ids.into_static(),
             new_documents_ids: documents_ids,
             replaced_documents_ids: RoaringBitmap::default(),
             documents_count,