Merge pull request #6024 from meilisearch/fix-metrics-route

Limit the number of retrieved task to one in the metrics

.github/workflows/publish-docker-images.yml

@@ -14,10 +14,105 @@ on:
   workflow_dispatch:
 
 jobs:
-  docker:
+  build:
-    runs-on: docker
+    runs-on: ${{ matrix.runner }}
+
+    strategy:
+      matrix:
+        platform: [amd64, arm64]
+        edition: [community, enterprise]
+        include:
+          - platform: amd64
+            runner: ubuntu-24.04
+          - platform: arm64
+            runner: ubuntu-24.04-arm
+          - edition: community
+            registry: getmeili/meilisearch
+            feature-flag: ""
+          - edition: enterprise
+            registry: getmeili/meilisearch-enterprise
+            feature-flag: "--features enterprise"
+
+    permissions: {}
+    steps:
+      - uses: actions/checkout@v5
+
+      - name: Prepare
+        run: |
+          platform=linux/${{ matrix.platform }}
+          echo "PLATFORM_PAIR=${platform//\//-}" >> $GITHUB_ENV
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+        with:
+          platforms: linux/${{ matrix.platform }}
+          install: true
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ matrix.registry }}
+          # Prevent `latest` to be updated for each new tag pushed.
+          # We need latest and `vX.Y` tags to only be pushed for the stable Meilisearch releases.
+          flavor: latest=false
+          tags: |
+            type=ref,event=tag
+            type=raw,value=nightly,enable=${{ github.event_name != 'push' }}
+            type=semver,pattern=v{{major}}.{{minor}},enable=${{ steps.check-tag-format.outputs.stable == 'true' }}
+            type=semver,pattern=v{{major}},enable=${{ steps.check-tag-format.outputs.stable == 'true' }}
+            type=raw,value=latest,enable=${{ steps.check-tag-format.outputs.stable == 'true' && steps.check-tag-format.outputs.latest == 'true' }}
+
+      - name: Build and push by digest
+        uses: docker/build-push-action@v6
+        id: build-and-push
+        with:
+          platforms: linux/${{ matrix.platform }}
+          labels: ${{ steps.meta.outputs.labels }}
+          tags: ${{ matrix.registry }}
+          outputs: type=image,push-by-digest=true,name-canonical=true,push=true
+          build-args: |
+            COMMIT_SHA=${{ github.sha }}
+            COMMIT_DATE=${{ steps.build-metadata.outputs.date }}
+            GIT_TAG=${{ github.ref_name }}
+            EXTRA_ARGS=${{ matrix.feature-flag }}
+
+      - name: Export digest
+        run: |
+          mkdir -p ${{ runner.temp }}/digests
+          digest="${{ steps.build-and-push.outputs.digest }}"
+          touch "${{ runner.temp }}/digests/${digest#sha256:}"
+
+      - name: Upload digest
+        uses: actions/upload-artifact@v4
+        with:
+          name: digests-${{ matrix.edition }}-${{ env.PLATFORM_PAIR }}
+          path: ${{ runner.temp }}/digests/*
+          if-no-files-found: error
+          retention-days: 1
+
+  merge:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        edition: [community, enterprise]
+        include:
+          - edition: community
+            registry: getmeili/meilisearch
+          - edition: enterprise
+            registry: getmeili/meilisearch-enterprise
+    needs:
+      - build
+
     permissions:
       id-token: write # This is needed to use Cosign in keyless mode
+
     steps:
       - uses: actions/checkout@v5
 
@@ -58,26 +153,30 @@ jobs:
 
           echo "date=$commit_date" >> $GITHUB_OUTPUT
 
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3
-
       - name: Install cosign
         uses: sigstore/cosign-installer@d7543c93d881b35a8faa02e8e3605f69b7a1ce62 # tag=v3.10.0
 
+      - name: Download digests
+        uses: actions/download-artifact@v4
+        with:
+          path: ${{ runner.temp }}/digests
+          pattern: digests-${{ matrix.edition }}-*
+          merge-multiple: true
+
       - name: Login to Docker Hub
         uses: docker/login-action@v3
         with:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
 
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
       - name: Docker meta
         id: meta
         uses: docker/metadata-action@v5
         with:
-          images: getmeili/meilisearch
+          images: ${{ matrix.registry }}
           # Prevent `latest` to be updated for each new tag pushed.
           # We need latest and `vX.Y` tags to only be pushed for the stable Meilisearch releases.
           flavor: latest=false
@@ -88,55 +187,34 @@ jobs:
             type=semver,pattern=v{{major}},enable=${{ steps.check-tag-format.outputs.stable == 'true' }}
             type=raw,value=latest,enable=${{ steps.check-tag-format.outputs.stable == 'true' && steps.check-tag-format.outputs.latest == 'true' }}
 
-      - name: Build and push
+      - name: Create manifest list and push
-        uses: docker/build-push-action@v6
+        working-directory: ${{ runner.temp }}/digests
-        id: build-and-push
+        run: |
-        with:
+          docker buildx imagetools create $(jq -cr '.tags | map("-t " + .) | join(" ")' <<< "$DOCKER_METADATA_OUTPUT_JSON") \
-          push: true
+            $(printf '${{ matrix.registry }}@sha256:%s ' *)
-          platforms: linux/amd64,linux/arm64
+
-          tags: ${{ steps.meta.outputs.tags }}
+      - name: Inspect image to fetch digest to sign
-          build-args: |
+        run: |
-            COMMIT_SHA=${{ github.sha }}
+          digest=$(docker buildx imagetools inspect --format='{{ json .Manifest }}' ${{ matrix.registry }}:${{ steps.meta.outputs.version }} | jq -r '.digest')
-            COMMIT_DATE=${{ steps.build-metadata.outputs.date }}
+          echo "DIGEST=${digest}" >> $GITHUB_ENV
-            GIT_TAG=${{ github.ref_name }}
 
       - name: Sign the images with GitHub OIDC Token
         env:
-          DIGEST: ${{ steps.build-and-push.outputs.digest }}
           TAGS: ${{ steps.meta.outputs.tags }}
         run: |
           images=""
           for tag in ${TAGS}; do
-            images+="${tag}@${DIGEST} "
+            images+="${tag}@${{ env.DIGEST }} "
           done
           cosign sign --yes ${images}
 
       # /!\ Don't touch this without checking with Cloud team
       - name: Send CI information to Cloud team
         # Do not send if nightly build (i.e. 'schedule' or 'workflow_dispatch' event)
-        if: github.event_name == 'push'
+        if: ${{ (github.event_name == 'push') && (matrix.edition == 'enterprise') }}
         uses: peter-evans/repository-dispatch@v3
         with:
           token: ${{ secrets.MEILI_BOT_GH_PAT }}
           repository: meilisearch/meilisearch-cloud
           event-type: cloud-docker-build
           client-payload: '{ "meilisearch_version": "${{ github.ref_name }}", "stable": "${{ steps.check-tag-format.outputs.stable }}" }'
-
-      # Send notification to Swarmia to notify of a deployment: https://app.swarmia.com
-      # - name: 'Setup jq'
-      #   uses: dcarbone/install-jq-action
-      # - name: Send deployment to Swarmia
-      #   if: github.event_name == 'push' && success()
-      #   run: |
-      #     JSON_STRING=$( jq --null-input --compact-output \
-      #     --arg version "${{ github.ref_name }}" \
-      #     --arg appName "meilisearch" \
-      #     --arg environment "production" \
-      #     --arg commitSha "${{ github.sha }}" \
-      #     --arg repositoryFullName "${{ github.repository }}" \
-      #     '{"version": $version, "appName": $appName, "environment": $environment, "commitSha": $commitSha, "repositoryFullName": $repositoryFullName}' )
-
-      #     curl -H "Authorization: ${{ secrets.SWARMIA_DEPLOYMENTS_AUTHORIZATION }}" \
-      #       -H "Content-Type: application/json" \
-      #       -d "$JSON_STRING" \
-      #       https://hook.swarmia.com/deployments

.github/workflows/publish-release-assets.yml

@@ -32,157 +32,61 @@ jobs:
         if: github.event_name == 'release' && steps.check-tag-format.outputs.stable == 'true'
         run: bash .github/scripts/check-release.sh
 
-  publish-linux:
+  publish-binaries:
-    name: Publish binary for Linux
+    name: Publish binary for ${{ matrix.release }} ${{ matrix.edition }} edition
-    runs-on: ubuntu-latest
-    needs: check-version
-    container:
-      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:22.04
-    steps:
-      - uses: actions/checkout@v5
-      - name: Install needed dependencies
-        run: |
-          apt-get update && apt-get install -y curl
-          apt-get install build-essential -y
-      - uses: dtolnay/rust-toolchain@1.89
-      - name: Build
-        run: cargo build --release --locked
-      # No need to upload binaries for dry run (cron or workflow_dispatch)
-      - name: Upload binaries to release
-        if: github.event_name == 'release'
-        uses: svenstaro/upload-release-action@2.11.2
-        with:
-          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-          file: target/release/meilisearch
-          asset_name: meilisearch-linux-amd64
-          tag: ${{ github.ref }}
-
-  publish-macos-windows:
-    name: Publish binary for ${{ matrix.os }}
     runs-on: ${{ matrix.os }}
-    needs: check-version
     strategy:
       fail-fast: false
       matrix:
-        os: [macos-14, windows-2022]
+        edition: [community, enterprise]
+        release:
+          [macos-amd64, macos-aarch64, windows, linux-amd64, linux-aarch64]
         include:
-          - os: macos-14
+          - edition: "community"
-            artifact_name: meilisearch
+            feature-flag: ""
-            asset_name: meilisearch-macos-amd64
+            edition-suffix: ""
-          - os: windows-2022
+          - edition: "enterprise"
-            artifact_name: meilisearch.exe
+            feature-flag: "--features enterprise"
-            asset_name: meilisearch-windows-amd64.exe
+            edition-suffix: "enterprise-"
+          - release: macos-amd64
+            os: macos-15-intel
+            binary_path: release/meilisearch
+            asset_name: macos-amd64
+            extra-args: ""
+          - release: macos-aarch64
+            os: macos-14
+            binary_path: aarch64-apple-darwin/release/meilisearch
+            asset_name: macos-apple-silicon
+            extra-args: "--target aarch64-apple-darwin"
+          - release: windows
+            os: windows-2022
+            binary_path: release/meilisearch.exe
+            asset_name: windows-amd64.exe
+            extra-args: ""
+          - release: linux-amd64
+            os: ubuntu-22.04
+            binary_path: x86_64-unknown-linux-gnu/release/meilisearch
+            asset_name: linux-amd64
+            extra-args: "--target x86_64-unknown-linux-gnu"
+          - release: linux-aarch64
+            os: ubuntu-22.04-arm
+            binary_path: aarch64-unknown-linux-gnu/release/meilisearch
+            asset_name: linux-aarch64
+            extra-args: "--target aarch64-unknown-linux-gnu"
+    needs: check-version
     steps:
       - uses: actions/checkout@v5
       - uses: dtolnay/rust-toolchain@1.89
       - name: Build
-        run: cargo build --release --locked
+        run: cargo build --release --locked ${{ matrix.feature-flag }} ${{ matrix.extra-args }}
       # No need to upload binaries for dry run (cron or workflow_dispatch)
       - name: Upload binaries to release
         if: github.event_name == 'release'
         uses: svenstaro/upload-release-action@2.11.2
         with:
           repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-          file: target/release/${{ matrix.artifact_name }}
+          file: target/${{ matrix.binary_path }}
-          asset_name: ${{ matrix.asset_name }}
+          asset_name: meilisearch-${{ matrix.edition-suffix }}${{ matrix.asset_name }}
-          tag: ${{ github.ref }}
-
-  publish-macos-apple-silicon:
-    name: Publish binary for macOS silicon
-    runs-on: macos-14
-    needs: check-version
-    strategy:
-      matrix:
-        include:
-          - target: aarch64-apple-darwin
-            asset_name: meilisearch-macos-apple-silicon
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v5
-      - name: Installing Rust toolchain
-        uses: dtolnay/rust-toolchain@1.89
-        with:
-          profile: minimal
-          target: ${{ matrix.target }}
-      - name: Cargo build
-        uses: actions-rs/cargo@v1
-        with:
-          command: build
-          args: --release --target ${{ matrix.target }}
-      - name: Upload the binary to release
-        # No need to upload binaries for dry run (cron or workflow_dispatch)
-        if: github.event_name == 'release'
-        uses: svenstaro/upload-release-action@2.11.2
-        with:
-          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-          file: target/${{ matrix.target }}/release/meilisearch
-          asset_name: ${{ matrix.asset_name }}
-          tag: ${{ github.ref }}
-
-  publish-aarch64:
-    name: Publish binary for aarch64
-    runs-on: ubuntu-latest
-    needs: check-version
-    env:
-      DEBIAN_FRONTEND: noninteractive
-    container:
-      # Use ubuntu-22.04 to compile with glibc 2.35
-      image: ubuntu:22.04
-    strategy:
-      matrix:
-        include:
-          - target: aarch64-unknown-linux-gnu
-            asset_name: meilisearch-linux-aarch64
-    steps:
-      - name: Checkout repository
-        uses: actions/checkout@v5
-      - name: Install needed dependencies
-        run: |
-          apt-get update -y && apt upgrade -y
-          apt-get install -y curl build-essential gcc-aarch64-linux-gnu
-      - name: Set up Docker for cross compilation
-        run: |
-          apt-get install -y curl apt-transport-https ca-certificates software-properties-common
-          curl -fsSL https://download.docker.com/linux/ubuntu/gpg | apt-key add -
-          add-apt-repository "deb [arch=$(dpkg --print-architecture)] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"
-          apt-get update -y && apt-get install -y docker-ce
-      - name: Installing Rust toolchain
-        uses: dtolnay/rust-toolchain@1.89
-        with:
-          profile: minimal
-          target: ${{ matrix.target }}
-      - name: Configure target aarch64 GNU
-        ## Environment variable is not passed using env:
-        ## LD gold won't work with MUSL
-        # env:
-        #   JEMALLOC_SYS_WITH_LG_PAGE: 16
-        #   RUSTFLAGS: '-Clink-arg=-fuse-ld=gold'
-        run: |
-          echo '[target.aarch64-unknown-linux-gnu]' >> ~/.cargo/config
-          echo 'linker = "aarch64-linux-gnu-gcc"' >> ~/.cargo/config
-          echo 'JEMALLOC_SYS_WITH_LG_PAGE=16' >> $GITHUB_ENV
-      - name: Install a default toolchain that will be used to build cargo cross
-        run: |
-          rustup default stable
-      - name: Cargo build
-        uses: actions-rs/cargo@v1
-        with:
-          command: build
-          use-cross: true
-          args: --release --target ${{ matrix.target }}
-        env:
-          CROSS_DOCKER_IN_DOCKER: true
-      - name: List target output files
-        run: ls -lR ./target
-      - name: Upload the binary to release
-        # No need to upload binaries for dry run (cron or workflow_dispatch)
-        if: github.event_name == 'release'
-        uses: svenstaro/upload-release-action@2.11.2
-        with:
-          repo_token: ${{ secrets.MEILI_BOT_GH_PAT }}
-          file: target/${{ matrix.target }}/release/meilisearch
-          asset_name: ${{ matrix.asset_name }}
           tag: ${{ github.ref }}
 
   publish-openapi-file:

.github/workflows/sdks-tests.yml

@@ -68,7 +68,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -92,7 +92,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -122,7 +122,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -149,7 +149,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -184,7 +184,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -213,7 +213,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -238,7 +238,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -263,7 +263,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -284,7 +284,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -307,7 +307,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -338,7 +338,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}
@@ -370,7 +370,7 @@ jobs:
     runs-on: ubuntu-latest
     services:
       meilisearch:
-        image: getmeili/meilisearch:${{ needs.define-docker-image.outputs.docker-image }}
+        image: getmeili/meilisearch-enterprise:${{ needs.define-docker-image.outputs.docker-image }}
         env:
           MEILI_MASTER_KEY: ${{ env.MEILI_MASTER_KEY }}
           MEILI_NO_ANALYTICS: ${{ env.MEILI_NO_ANALYTICS }}

.github/workflows/test-suite.yml

@@ -15,8 +15,12 @@ env:
 
 jobs:
   test-linux:
-    name: Tests on ubuntu-22.04
+    name: Tests on Ubuntu
-    runs-on: ubuntu-latest
+    runs-on: ${{ matrix.runner }}
+    strategy:
+      matrix:
+        runner: [ubuntu-24.04, ubuntu-24.04-arm]
+        features: ["", "--features enterprise"]
     container:
       # Use ubuntu-22.04 to compile with glibc 2.35
       image: ubuntu:22.04
@@ -39,7 +43,7 @@ jobs:
         uses: actions-rs/cargo@v1
         with:
           command: test
-          args: --locked --release --all
+          args: --locked --release --all ${{ matrix.features }}
 
   test-others:
     name: Tests on ${{ matrix.os }}
@@ -48,6 +52,8 @@ jobs:
       fail-fast: false
       matrix:
         os: [macos-14, windows-2022]
+        features: ["", "--features enterprise"]
+    if: github.event_name == 'schedule' || github.event_name == 'workflow_dispatch' || github.event_name == 'pull_request'
     steps:
       - uses: actions/checkout@v5
       - name: Cache dependencies
@@ -62,7 +68,7 @@ jobs:
         uses: actions-rs/cargo@v1
         with:
           command: test
-          args: --locked --release --all
+          args: --locked --release --all ${{ matrix.features }}
 
   test-all-features:
     name: Tests almost all features
@@ -88,6 +94,9 @@ jobs:
   ollama-ubuntu:
     name: Test with Ollama
     runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        features: ["", "--features enterprise"]
     env:
       MEILI_TEST_OLLAMA_SERVER: "http://localhost:11434"
     steps:
@@ -115,7 +124,7 @@ jobs:
         uses: actions-rs/cargo@v1
         with:
           command: test
-          args: --locked --release --all --features test-ollama ollama
+          args: --locked --release --all --features test-ollama ollama ${{ matrix.features }}
 
   test-disabled-tokenization:
     name: Test disabled tokenization
@@ -143,28 +152,25 @@ jobs:
   # We run tests in debug also, to make sure that the debug_assertions are hit
   test-debug:
     name: Run tests in debug
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-22.04
-    container:
+    strategy:
-      # Use ubuntu-22.04 to compile with glibc 2.35
+      matrix:
-      image: ubuntu:22.04
+        features: ["", "--features enterprise"]
     steps:
       - uses: actions/checkout@v5
-      - name: Install needed dependencies
-        run: |
-          apt-get update && apt-get install -y curl
-          apt-get install build-essential -y
       - uses: dtolnay/rust-toolchain@1.89
-      - name: Cache dependencies
-        uses: Swatinem/rust-cache@v2.8.0
       - name: Run tests in debug
         uses: actions-rs/cargo@v1
         with:
           command: test
-          args: --locked --all
+          args: --locked --all ${{ matrix.features }}
 
   clippy:
     name: Run Clippy
     runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        features: ["", "--features enterprise"]
     steps:
       - uses: actions/checkout@v5
       - uses: dtolnay/rust-toolchain@1.89
@@ -177,7 +183,7 @@ jobs:
         uses: actions-rs/cargo@v1
         with:
           command: clippy
-          args: --all-targets -- --deny warnings
+          args: --all-targets ${{ matrix.features }} -- --deny warnings
 
   fmt:
     name: Run Rustfmt

Cargo.toml

@@ -23,7 +23,7 @@ members = [
 ]
 
 [workspace.package]
-version = "1.27.0"
+version = "1.28.2"
 authors = [
     "Quentin de Quelen <quentin@dequelen.me>",
     "Clément Renault <clement@meilisearch.com>",
@@ -50,3 +50,5 @@ opt-level = 3
 opt-level = 3
 [profile.dev.package.roaring]
 opt-level = 3
+[profile.dev.package.gemm-f16]
+opt-level = 3

Cross.toml

@@ -1,7 +0,0 @@
-[build.env]
-passthrough = [
-    "RUST_BACKTRACE",
-    "CARGO_TERM_COLOR",
-    "RUSTFLAGS",
-    "JEMALLOC_SYS_WITH_LG_PAGE"
-]

Dockerfile

@@ -8,16 +8,14 @@ WORKDIR /
 ARG     COMMIT_SHA
 ARG     COMMIT_DATE
 ARG     GIT_TAG
+ARG     EXTRA_ARGS
 ENV     VERGEN_GIT_SHA=${COMMIT_SHA} VERGEN_GIT_COMMIT_TIMESTAMP=${COMMIT_DATE} VERGEN_GIT_DESCRIBE=${GIT_TAG}
 ENV     RUSTFLAGS="-C target-feature=-crt-static"
 
 COPY    . .
 RUN     set -eux; \
         apkArch="$(apk --print-arch)"; \
-        if [ "$apkArch" = "aarch64" ]; then \
+        cargo build --release -p meilisearch -p meilitool ${EXTRA_ARGS}
-            export JEMALLOC_SYS_WITH_LG_PAGE=16; \
-        fi && \
-        cargo build --release -p meilisearch -p meilitool
 
 # Run
 FROM    alpine:3.22

crates/benchmarks/Cargo.toml

@@ -11,27 +11,27 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
-csv = "1.3.1"
+csv = "1.4.0"
-memmap2 = "0.9.7"
+memmap2 = "0.9.9"
 milli = { path = "../milli" }
-mimalloc = { version = "0.1.47", default-features = false }
+mimalloc = { version = "0.1.48", default-features = false }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
-tempfile = "3.20.0"
+tempfile = "3.23.0"
 
 [dev-dependencies]
-criterion = { version = "0.6.0", features = ["html_reports"] }
+criterion = { version = "0.7.0", features = ["html_reports"] }
 rand = "0.8.5"
 rand_chacha = "0.3.1"
 roaring = "0.10.12"
 
 [build-dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
-bytes = "1.10.1"
+bytes = "1.11.0"
-convert_case = "0.8.0"
+convert_case = "0.9.0"
-flate2 = "1.1.2"
+flate2 = "1.1.5"
-reqwest = { version = "0.12.20", features = ["blocking", "rustls-tls"], default-features = false }
+reqwest = { version = "0.12.24", features = ["blocking", "rustls-tls"], default-features = false }
 
 [features]
 default = ["milli/all-tokenizations"]

crates/build-info/Cargo.toml

@@ -11,8 +11,8 @@ license.workspace = true
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
-time = { version = "0.3.41", features = ["parsing"] }
+time = { version = "0.3.44", features = ["parsing"] }
 
 [build-dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
 vergen-git2 = "1.0.7"

crates/dump/Cargo.toml

@@ -11,24 +11,27 @@ readme.workspace = true
 license.workspace = true
 
 [dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
-flate2 = "1.1.2"
+flate2 = "1.1.5"
 http = "1.3.1"
 meilisearch-types = { path = "../meilisearch-types" }
 once_cell = "1.21.3"
-regex = "1.11.1"
+regex = "1.12.2"
 roaring = { version = "0.10.12", features = ["serde"] }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
 tar = "0.4.44"
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = ["serde-well-known", "formatting", "parsing", "macros"] }
+time = { version = "0.3.44", features = ["serde-well-known", "formatting", "parsing", "macros"] }
 tracing = "0.1.41"
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.1", features = ["serde", "v4"] }
 
 [dev-dependencies]
 big_s = "1.0.2"
 maplit = "1.0.2"
 meili-snap = { path = "../meili-snap" }
 meilisearch-types = { path = "../meilisearch-types" }
+
+[features]
+enterprise = ["meilisearch-types/enterprise"]

crates/dump/src/lib.rs

@@ -262,13 +262,13 @@ pub(crate) mod test {
     use big_s::S;
     use maplit::{btreemap, btreeset};
     use meilisearch_types::batches::{Batch, BatchEnqueuedAt, BatchStats};
-    use meilisearch_types::enterprise_edition::network::{Network, Remote};
     use meilisearch_types::facet_values_sort::FacetValuesSort;
     use meilisearch_types::features::RuntimeTogglableFeatures;
     use meilisearch_types::index_uid_pattern::IndexUidPattern;
     use meilisearch_types::keys::{Action, Key};
     use meilisearch_types::milli::update::Setting;
     use meilisearch_types::milli::{self, FilterableAttributesRule};
+    use meilisearch_types::network::{Network, Remote};
     use meilisearch_types::settings::{Checked, FacetingSettings, Settings};
     use meilisearch_types::task_view::DetailsView;
     use meilisearch_types::tasks::{BatchStopReason, Details, Kind, Status};

crates/dump/src/reader/v6/mod.rs

@@ -24,7 +24,7 @@ pub type Batch = meilisearch_types::batches::Batch;
 pub type Key = meilisearch_types::keys::Key;
 pub type ChatCompletionSettings = meilisearch_types::features::ChatCompletionSettings;
 pub type RuntimeTogglableFeatures = meilisearch_types::features::RuntimeTogglableFeatures;
-pub type Network = meilisearch_types::enterprise_edition::network::Network;
+pub type Network = meilisearch_types::network::Network;
 pub type Webhooks = meilisearch_types::webhooks::WebhooksDumpView;
 
 // ===== Other types to clarify the code of the compat module

crates/dump/src/writer.rs

@@ -5,9 +5,9 @@ use std::path::PathBuf;
 use flate2::write::GzEncoder;
 use flate2::Compression;
 use meilisearch_types::batches::Batch;
-use meilisearch_types::enterprise_edition::network::Network;
 use meilisearch_types::features::{ChatCompletionSettings, RuntimeTogglableFeatures};
 use meilisearch_types::keys::Key;
+use meilisearch_types::network::Network;
 use meilisearch_types::settings::{Checked, Settings};
 use meilisearch_types::webhooks::WebhooksDumpView;
 use serde_json::{Map, Value};

crates/file-store/Cargo.toml

@@ -11,7 +11,7 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
 tracing = "0.1.41"
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.1", features = ["serde", "v4"] }

crates/flatten-serde-json/Cargo.toml

@@ -16,7 +16,7 @@ license.workspace = true
 serde_json = "1.0"
 
 [dev-dependencies]
-criterion = { version = "0.6.0", features = ["html_reports"] }
+criterion = { version = "0.7.0", features = ["html_reports"] }
 
 [[bench]]
 name = "benchmarks"

crates/fuzzers/Cargo.toml

@@ -11,12 +11,12 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-arbitrary = { version = "1.4.1", features = ["derive"] }
+arbitrary = { version = "1.4.2", features = ["derive"] }
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
-clap = { version = "4.5.40", features = ["derive"] }
+clap = { version = "4.5.52", features = ["derive"] }
 either = "1.15.0"
 fastrand = "2.3.0"
 milli = { path = "../milli" }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
-tempfile = "3.20.0"
+tempfile = "3.23.0"

crates/index-scheduler/Cargo.toml

@@ -11,33 +11,33 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
 bincode = "1.3.3"
 byte-unit = "5.1.6"
-bytes = "1.10.1"
+bytes = "1.11.0"
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
 bumparaw-collections = "0.1.4"
-convert_case = "0.8.0"
+convert_case = "0.9.0"
-csv = "1.3.1"
+csv = "1.4.0"
 derive_builder = "0.20.2"
 dump = { path = "../dump" }
-enum-iterator = "2.1.0"
+enum-iterator = "2.3.0"
 file-store = { path = "../file-store" }
-flate2 = "1.1.2"
+flate2 = "1.1.5"
-indexmap = "2.9.0"
+indexmap = "2.12.0"
 meilisearch-auth = { path = "../meilisearch-auth" }
 meilisearch-types = { path = "../meilisearch-types" }
-memmap2 = "0.9.7"
+memmap2 = "0.9.9"
 page_size = "0.6.0"
-rayon = "1.10.0"
+rayon = "1.11.0"
 roaring = { version = "0.10.12", features = ["serde"] }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
 tar = "0.4.44"
 synchronoise = "1.0.1"
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = [
+time = { version = "0.3.44", features = [
     "serde-well-known",
     "formatting",
     "parsing",
@@ -45,11 +45,11 @@ time = { version = "0.3.41", features = [
 ] }
 tracing = "0.1.41"
 ureq = "2.12.1"
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.1", features = ["serde", "v4"] }
 backoff = "0.4.0"
-reqwest = { version = "0.12.23", features = ["rustls-tls", "http2"], default-features = false }
+reqwest = { version = "0.12.24", features = ["rustls-tls", "http2"], default-features = false }
 rusty-s3 = "0.8.1"
-tokio = { version = "1.47.1", features = ["full"] }
+tokio = { version = "1.48.0", features = ["full"] }
 
 [dev-dependencies]
 big_s = "1.0.2"

crates/index-scheduler/src/features.rs

@@ -1,9 +1,9 @@
 use std::sync::{Arc, RwLock};
 
-use meilisearch_types::enterprise_edition::network::Network;
 use meilisearch_types::features::{InstanceTogglableFeatures, RuntimeTogglableFeatures};
 use meilisearch_types::heed::types::{SerdeJson, Str};
 use meilisearch_types::heed::{Database, Env, RwTxn, WithoutTls};
+use meilisearch_types::network::Network;
 
 use crate::error::FeatureNotEnabledError;
 use crate::Result;

crates/index-scheduler/src/lib.rs

@@ -54,7 +54,6 @@ pub use features::RoFeatures;
 use flate2::bufread::GzEncoder;
 use flate2::Compression;
 use meilisearch_types::batches::Batch;
-use meilisearch_types::enterprise_edition::network::Network;
 use meilisearch_types::features::{
     ChatCompletionSettings, InstanceTogglableFeatures, RuntimeTogglableFeatures,
 };
@@ -67,6 +66,7 @@ use meilisearch_types::milli::vector::{
     Embedder, EmbedderOptions, RuntimeEmbedder, RuntimeEmbedders, RuntimeFragment,
 };
 use meilisearch_types::milli::{self, Index};
+use meilisearch_types::network::Network;
 use meilisearch_types::task_view::TaskView;
 use meilisearch_types::tasks::{KindWithContent, Task, TaskNetwork};
 use meilisearch_types::webhooks::{Webhook, WebhooksDumpView, WebhooksView};

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/after_processing_everything.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, batch_uid: 1, status: succeeded, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, batch_uid: 2, status: succeeded, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, batch_uid: 3, status: failed, error: ResponseError { code: 200, message: "Index `doggo` already exists.", error_code: "index_already_exists", error_type: "invalid_request", error_link: "https://docs.meilisearch.com/errors#index_already_exists" }, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -57,7 +57,7 @@ girafo: { number_of_documents: 0, field_distribution: {} }
 [timestamp] [4,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.28.2"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 1 {uid: 1, details: {"primaryKey":"mouse"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"catto":1}}, stop reason: "created batch containing only task with id 1 of type `indexCreation` that cannot be batched with any other task.", }
 2 {uid: 2, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 2 of type `indexCreation` that cannot be batched with any other task.", }
 3 {uid: 3, details: {"primaryKey":"bone"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"indexCreation":1},"indexUids":{"doggo":1}}, stop reason: "created batch containing only task with id 3 of type `indexCreation` that cannot be batched with any other task.", }

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/register_automatic_upgrade_task.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 ----------------------------------------------------------------------
 ### Status:
 enqueued [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/registered_a_task_while_the_upgrade_task_is_enqueued.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, status: enqueued, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 ----------------------------------------------------------------------
 ### Status:
@@ -37,7 +37,7 @@ catto [1,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.28.2"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_failed_again.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: failed, error: ResponseError { code: 200, message: "Planned failure for tests.", error_code: "internal", error_type: "internal", error_link: "https://docs.meilisearch.com/errors#internal" }, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 ----------------------------------------------------------------------
@@ -40,7 +40,7 @@ doggo [2,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.28.2"}, stats: {"totalNbTasks":1,"status":{"failed":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/scheduler/snapshots/test_failure.rs/upgrade_failure/upgrade_task_succeeded.snap

@@ -6,7 +6,7 @@ source: crates/index-scheduler/src/scheduler/test_failure.rs
 []
 ----------------------------------------------------------------------
 ### All Tasks:
-0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 27, 0) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
+0 {uid: 0, batch_uid: 0, status: succeeded, details: { from: (1, 12, 0), to: (1, 28, 2) }, kind: UpgradeDatabase { from: (1, 12, 0) }}
 1 {uid: 1, status: enqueued, details: { primary_key: Some("mouse"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "catto", primary_key: Some("mouse") }}
 2 {uid: 2, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
 3 {uid: 3, status: enqueued, details: { primary_key: Some("bone"), old_new_uid: None, new_index_uid: None }, kind: IndexCreation { index_uid: "doggo", primary_key: Some("bone") }}
@@ -43,7 +43,7 @@ doggo [2,3,]
 [timestamp] [0,]
 ----------------------------------------------------------------------
 ### All Batches:
-0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.27.0"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
+0 {uid: 0, details: {"upgradeFrom":"v1.12.0","upgradeTo":"v1.28.2"}, stats: {"totalNbTasks":1,"status":{"succeeded":1},"types":{"upgradeDatabase":1},"indexUids":{}}, stop reason: "stopped after the last task of type `upgradeDatabase` because they cannot be batched with tasks of any other type.", }
 ----------------------------------------------------------------------
 ### Batch to tasks mapping:
 0 [0,]

crates/index-scheduler/src/upgrade/mod.rs

@@ -51,6 +51,7 @@ pub fn upgrade_index_scheduler(
         (1, 25, _) => 0,
         (1, 26, _) => 0,
         (1, 27, _) => 0,
+        (1, 28, _) => 0,
         (major, minor, patch) => {
             if major > current_major
                 || (major == current_major && minor > current_minor)

crates/json-depth-checker/Cargo.toml

@@ -15,7 +15,7 @@ license.workspace = true
 serde_json = "1.0"
 
 [dev-dependencies]
-criterion = "0.6.0"
+criterion = "0.7.0"
 
 [[bench]]
 name = "depth"

crates/meili-snap/Cargo.toml

@@ -13,7 +13,7 @@ license.workspace = true
 [dependencies]
 # fixed version due to format breakages in v1.40
 insta = { version = "=1.39.0", features = ["json", "redactions"] }
-md5 = "0.7.0"
+md5 = "0.8.0"
 once_cell = "1.21"
-regex-lite = "0.1.6"
+regex-lite = "0.1.8"
-uuid = { version = "1.17.0", features = ["v4"] }
+uuid = { version = "1.18.1", features = ["v4"] }

crates/meilisearch-auth/Cargo.toml

@@ -12,15 +12,15 @@ license.workspace = true
 
 [dependencies]
 base64 = "0.22.1"
-enum-iterator = "2.1.0"
+enum-iterator = "2.3.0"
 hmac = "0.12.1"
 maplit = "1.0.2"
 meilisearch-types = { path = "../meilisearch-types" }
 rand = "0.8.5"
 roaring = { version = "0.10.12", features = ["serde"] }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
 sha2 = "0.10.9"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = ["serde-well-known", "formatting", "parsing", "macros"] }
+time = { version = "0.3.44", features = ["serde-well-known", "formatting", "parsing", "macros"] }
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.1", features = ["serde", "v4"] }

crates/meilisearch-types/Cargo.toml

@@ -11,38 +11,38 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-actix-web = { version = "4.11.0", default-features = false }
+actix-web = { version = "4.12.0", default-features = false }
-anyhow = "1.0.98"
+anyhow = "1.0.100"
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
 bumparaw-collections = "0.1.4"
 byte-unit = { version = "5.1.6", features = ["serde"] }
-convert_case = "0.8.0"
+convert_case = "0.9.0"
-csv = "1.3.1"
+csv = "1.4.0"
-deserr = { version = "0.6.3", features = ["actix-web"] }
+deserr = { version = "0.6.4", features = ["actix-web"] }
 either = { version = "1.15.0", features = ["serde"] }
-enum-iterator = "2.1.0"
+enum-iterator = "2.3.0"
 file-store = { path = "../file-store" }
-flate2 = "1.1.2"
+flate2 = "1.1.5"
 fst = "0.4.7"
-memmap2 = "0.9.7"
+memmap2 = "0.9.9"
 milli = { path = "../milli" }
 roaring = { version = "0.10.12", features = ["serde"] }
 rustc-hash = "2.1.1"
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
 serde-cs = "0.2.4"
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
 tar = "0.4.44"
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = [
+time = { version = "0.3.44", features = [
     "serde-well-known",
     "formatting",
     "parsing",
     "macros",
 ] }
-tokio = "1.45"
+tokio = "1.48"
 utoipa = { version = "5.4.0", features = ["macros"] }
-uuid = { version = "1.17.0", features = ["serde", "v4"] }
+uuid = { version = "1.18.1", features = ["serde", "v4"] }
 
 [dev-dependencies]
 # fixed version due to format breakages in v1.40
@@ -56,6 +56,9 @@ all-tokenizations = ["milli/all-tokenizations"]
 # chinese specialized tokenization
 chinese = ["milli/chinese"]
 chinese-pinyin = ["milli/chinese-pinyin"]
+
+enterprise = ["milli/enterprise"]
+
 # hebrew specialized tokenization
 hebrew = ["milli/hebrew"]
 # japanese specialized tokenization

crates/meilisearch-types/src/community_edition.rs

@@ -0,0 +1,16 @@
+pub mod network {
+    use milli::update::new::indexer::current_edition::sharding::Shards;
+
+    use crate::network::Network;
+
+    impl Network {
+        pub fn shards(&self) -> Option<Shards> {
+            None
+        }
+
+        pub fn sharding(&self) -> bool {
+            // always false in CE
+            false
+        }
+    }
+}

crates/meilisearch-types/src/enterprise_edition/network.rs

@@ -3,21 +3,9 @@
 // Use of this source code is governed by the Business Source License 1.1,
 // as found in the LICENSE-EE file or at <https://mariadb.com/bsl11>
 
-use std::collections::BTreeMap;
-
 use milli::update::new::indexer::enterprise_edition::sharding::Shards;
-use serde::{Deserialize, Serialize};
 
-#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq, Default)]
+use crate::network::Network;
-#[serde(rename_all = "camelCase")]
-pub struct Network {
-    #[serde(default, rename = "self")]
-    pub local: Option<String>,
-    #[serde(default)]
-    pub remotes: BTreeMap<String, Remote>,
-    #[serde(default)]
-    pub sharding: bool,
-}
 
 impl Network {
     pub fn shards(&self) -> Option<Shards> {
@@ -34,14 +22,8 @@ impl Network {
             None
         }
     }
-}
 
-#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
+    pub fn sharding(&self) -> bool {
-#[serde(rename_all = "camelCase")]
+        self.sharding
-pub struct Remote {
+    }
-    pub url: String,
-    #[serde(default)]
-    pub search_api_key: Option<String>,
-    #[serde(default)]
-    pub write_api_key: Option<String>,
 }

crates/meilisearch-types/src/error.rs

@@ -433,6 +433,7 @@ InvalidChatCompletionSearchQueryParamPrompt    , InvalidRequest       , BAD_REQU
 InvalidChatCompletionSearchFilterParamPrompt   , InvalidRequest       , BAD_REQUEST ;
 InvalidChatCompletionSearchIndexUidParamPrompt , InvalidRequest       , BAD_REQUEST ;
 InvalidChatCompletionPreQueryPrompt            , InvalidRequest       , BAD_REQUEST ;
+RequiresEnterpriseEdition                      , InvalidRequest       , UNAVAILABLE_FOR_LEGAL_REASONS ;
 // Webhooks
 InvalidWebhooks                                , InvalidRequest       , BAD_REQUEST ;
 InvalidWebhookUrl                              , InvalidRequest       , BAD_REQUEST ;

crates/meilisearch-types/src/lib.rs

@@ -2,10 +2,17 @@
 
 pub mod batch_view;
 pub mod batches;
+#[cfg(not(feature = "enterprise"))]
+pub mod community_edition;
 pub mod compression;
 pub mod deserr;
 pub mod document_formats;
+#[cfg(feature = "enterprise")]
 pub mod enterprise_edition;
+#[cfg(not(feature = "enterprise"))]
+pub use community_edition as current_edition;
+#[cfg(feature = "enterprise")]
+pub use enterprise_edition as current_edition;
 pub mod error;
 pub mod facet_values_sort;
 pub mod features;
@@ -13,6 +20,7 @@ pub mod index_uid;
 pub mod index_uid_pattern;
 pub mod keys;
 pub mod locales;
+pub mod network;
 pub mod settings;
 pub mod star_or;
 pub mod task_view;

crates/meilisearch-types/src/network.rs

@@ -0,0 +1,23 @@
+use serde::{Deserialize, Serialize};
+use std::collections::BTreeMap;
+
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq, Default)]
+#[serde(rename_all = "camelCase")]
+pub struct Network {
+    #[serde(default, rename = "self")]
+    pub local: Option<String>,
+    #[serde(default)]
+    pub remotes: BTreeMap<String, Remote>,
+    #[serde(default)]
+    pub sharding: bool,
+}
+
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq)]
+#[serde(rename_all = "camelCase")]
+pub struct Remote {
+    pub url: String,
+    #[serde(default)]
+    pub search_api_key: Option<String>,
+    #[serde(default)]
+    pub write_api_key: Option<String>,
+}

crates/meilisearch/Cargo.toml

@@ -14,91 +14,91 @@ default-run = "meilisearch"
 
 [dependencies]
 actix-cors = "0.7.1"
-actix-http = { version = "3.11.0", default-features = false, features = [
+actix-http = { version = "3.11.2", default-features = false, features = [
     "compress-brotli",
     "compress-gzip",
     "rustls-0_23",
 ] }
 actix-utils = "3.0.1"
-actix-web = { version = "4.11.0", default-features = false, features = [
+actix-web = { version = "4.12.0", default-features = false, features = [
     "macros",
     "compress-brotli",
     "compress-gzip",
     "cookies",
     "rustls-0_23",
 ] }
-anyhow = { version = "1.0.98", features = ["backtrace"] }
+anyhow = { version = "1.0.100", features = ["backtrace"] }
-bstr = "1.12.0"
+bstr = "1.12.1"
 byte-unit = { version = "5.1.6", features = ["serde"] }
-bytes = "1.10.1"
+bytes = "1.11.0"
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
-clap = { version = "4.5.40", features = ["derive", "env"] }
+clap = { version = "4.5.52", features = ["derive", "env"] }
 crossbeam-channel = "0.5.15"
-deserr = { version = "0.6.3", features = ["actix-web"] }
+deserr = { version = "0.6.4", features = ["actix-web"] }
 dump = { path = "../dump" }
 either = "1.15.0"
 file-store = { path = "../file-store" }
-flate2 = "1.1.2"
+flate2 = "1.1.5"
 fst = "0.4.7"
 futures = "0.3.31"
 futures-util = "0.3.31"
 index-scheduler = { path = "../index-scheduler" }
-indexmap = { version = "2.9.0", features = ["serde"] }
+indexmap = { version = "2.12.0", features = ["serde"] }
-is-terminal = "0.4.16"
+is-terminal = "0.4.17"
 itertools = "0.14.0"
 jsonwebtoken = "9.3.1"
 lazy_static = "1.5.0"
 meilisearch-auth = { path = "../meilisearch-auth" }
 meilisearch-types = { path = "../meilisearch-types" }
-memmap2 = "0.9.7"
+memmap2 = "0.9.9"
-mimalloc = { version = "0.1.47", default-features = false }
+mimalloc = { version = "0.1.48", default-features = false }
 mime = "0.3.17"
 num_cpus = "1.17.0"
 obkv = "0.3.0"
 once_cell = "1.21.3"
-ordered-float = "5.0.0"
+ordered-float = "5.1.0"
-parking_lot = "0.12.4"
+parking_lot = "0.12.5"
 permissive-json-pointer = { path = "../permissive-json-pointer" }
 pin-project-lite = "0.2.16"
 platform-dirs = "0.3.0"
 prometheus = { version = "0.14.0", features = ["process"] }
 rand = "0.8.5"
-rayon = "1.10.0"
+rayon = "1.11.0"
-regex = "1.11.1"
+regex = "1.12.2"
-reqwest = { version = "0.12.20", features = [
+reqwest = { version = "0.12.24", features = [
     "rustls-tls",
     "json",
 ], default-features = false }
-rustls = { version = "0.23.28", features = ["ring"], default-features = false }
+rustls = { version = "0.23.35", features = ["ring"], default-features = false }
-rustls-pki-types = { version = "1.12.0", features = ["alloc"] }
+rustls-pki-types = { version = "1.13.0", features = ["alloc"] }
 rustls-pemfile = "2.2.0"
 segment = { version = "0.2.6" }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
 sha2 = "0.10.9"
 siphasher = "1.0.1"
 slice-group-by = "0.3.1"
-static-files = { version = "0.2.5", optional = true }
+static-files = { version = "0.3.1", optional = true }
-sysinfo = "0.35.2"
+sysinfo = "0.37.2"
 tar = "0.4.44"
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = [
+time = { version = "0.3.44", features = [
     "serde-well-known",
     "formatting",
     "parsing",
     "macros",
 ] }
-tokio = { version = "1.45.1", features = ["full"] }
+tokio = { version = "1.48.0", features = ["full"] }
-toml = "0.8.23"
+toml = "0.9.8"
-uuid = { version = "1.18.0", features = ["serde", "v4", "v7"] }
+uuid = { version = "1.18.1", features = ["serde", "v4", "v7"] }
 serde_urlencoded = "0.7.1"
 termcolor = "1.4.1"
-url = { version = "2.5.4", features = ["serde"] }
+url = { version = "2.5.7", features = ["serde"] }
 tracing = "0.1.41"
 tracing-subscriber = { version = "0.3.20", features = ["json"] }
 tracing-trace = { version = "0.1.0", path = "../tracing-trace" }
-tracing-actix-web = "0.7.18"
+tracing-actix-web = "0.7.19"
 build-info = { version = "1.7.0", path = "../build-info" }
 roaring = "0.10.12"
 mopa-maintained = "0.2.3"
@@ -114,35 +114,35 @@ utoipa = { version = "5.4.0", features = [
 utoipa-scalar = { version = "0.3.0", optional = true, features = ["actix-web"] }
 async-openai = { git = "https://github.com/meilisearch/async-openai", branch = "better-error-handling" }
 secrecy = "0.10.3"
-actix-web-lab = { version = "0.24.1", default-features = false }
+actix-web-lab = { version = "0.24.3", default-features = false }
 urlencoding = "2.1.3"
 backoff = { version = "0.4.0", features = ["tokio"] }
-
+humantime = { version = "2.3.0", default-features = false }
 
 [dev-dependencies]
-actix-rt = "2.10.0"
+actix-rt = "2.11.0"
-brotli = "8.0.1"
+brotli = "8.0.2"
 # fixed version due to format breakages in v1.40
 insta = { version = "=1.39.0", features = ["redactions"] }
 manifest-dir-macros = "0.1.18"
 maplit = "1.0.2"
 meili-snap = { path = "../meili-snap" }
 temp-env = "0.3.6"
-wiremock = "0.6.3"
+wiremock = "0.6.5"
 yaup = "0.3.1"
 
 [build-dependencies]
-anyhow = { version = "1.0.98", optional = true }
+anyhow = { version = "1.0.100", optional = true }
-cargo_toml = { version = "0.22.1", optional = true }
+cargo_toml = { version = "0.22.3", optional = true }
 hex = { version = "0.4.3", optional = true }
-reqwest = { version = "0.12.20", features = [
+reqwest = { version = "0.12.24", features = [
     "blocking",
     "rustls-tls",
 ], default-features = false, optional = true }
 sha-1 = { version = "0.10.1", optional = true }
-static-files = { version = "0.2.5", optional = true }
+static-files = { version = "0.3.1", optional = true }
-tempfile = { version = "3.20.0", optional = true }
+tempfile = { version = "3.23.0", optional = true }
-zip = { version = "4.1.0", optional = true }
+zip = { version = "6.0.0", optional = true }
 
 [features]
 default = ["meilisearch-types/all-tokenizations", "mini-dashboard"]
@@ -160,6 +160,7 @@ mini-dashboard = [
 ]
 chinese = ["meilisearch-types/chinese"]
 chinese-pinyin = ["meilisearch-types/chinese-pinyin"]
+enterprise = ["meilisearch-types/enterprise"]
 hebrew = ["meilisearch-types/hebrew"]
 japanese = ["meilisearch-types/japanese"]
 korean = ["meilisearch-types/korean"]

crates/meilisearch/src/metrics.rs

@@ -1,7 +1,8 @@
 use lazy_static::lazy_static;
 use prometheus::{
-    opts, register_gauge, register_histogram_vec, register_int_counter_vec, register_int_gauge,
+    opts, register_gauge, register_gauge_vec, register_histogram_vec, register_int_counter_vec,
-    register_int_gauge_vec, Gauge, HistogramVec, IntCounterVec, IntGauge, IntGaugeVec,
+    register_int_gauge, register_int_gauge_vec, Gauge, GaugeVec, HistogramVec, IntCounterVec,
+    IntGauge, IntGaugeVec,
 };
 
 lazy_static! {
@@ -73,6 +74,20 @@ lazy_static! {
         &["kind", "value"]
     )
     .expect("Can't create a metric");
+    pub static ref MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE: GaugeVec = register_gauge_vec!(
+        opts!("meilisearch_batch_running_progress_trace", "The currently running progress trace"),
+        &["batch_uid", "step_name"]
+    )
+    .expect("Can't create a metric");
+    pub static ref MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS: IntGaugeVec =
+        register_int_gauge_vec!(
+            opts!(
+                "meilisearch_last_finished_batches_progress_trace_ms",
+                "The last few batches progress trace in milliseconds"
+            ),
+            &["batch_uid", "step_name"]
+        )
+        .expect("Can't create a metric");
     pub static ref MEILISEARCH_LAST_UPDATE: IntGauge =
         register_int_gauge!(opts!("meilisearch_last_update", "Meilisearch Last Update"))
             .expect("Can't create a metric");

crates/meilisearch/src/routes/indexes/community_edition.rs

@@ -0,0 +1,39 @@
+pub mod proxy {
+
+    use std::fs::File;
+
+    use actix_web::HttpRequest;
+    use index_scheduler::IndexScheduler;
+
+    use crate::error::MeilisearchHttpError;
+
+    pub enum Body<T: serde::Serialize> {
+        NdJsonPayload,
+        Inline(T),
+        None,
+    }
+
+    impl Body<()> {
+        pub fn with_ndjson_payload(_file: File) -> Self {
+            Self::NdJsonPayload
+        }
+
+        pub fn none() -> Self {
+            Self::None
+        }
+    }
+
+    pub const PROXY_ORIGIN_REMOTE_HEADER: &str = "Meili-Proxy-Origin-Remote";
+    pub const PROXY_ORIGIN_TASK_UID_HEADER: &str = "Meili-Proxy-Origin-TaskUid";
+
+    pub async fn proxy<T: serde::Serialize>(
+        _index_scheduler: &IndexScheduler,
+        _index_uid: &str,
+        _req: &HttpRequest,
+        _network: meilisearch_types::network::Network,
+        _body: Body<T>,
+        _task: &meilisearch_types::tasks::Task,
+    ) -> Result<(), MeilisearchHttpError> {
+        Ok(())
+    }
+}

crates/meilisearch/src/routes/indexes/documents.rs

@@ -45,7 +45,7 @@ use crate::extractors::authentication::policies::*;
 use crate::extractors::authentication::GuardedData;
 use crate::extractors::payload::Payload;
 use crate::extractors::sequential_extractor::SeqHandler;
-use crate::routes::indexes::enterprise_edition::proxy::{proxy, Body};
+use crate::routes::indexes::current_edition::proxy::{proxy, Body};
 use crate::routes::indexes::search::fix_sort_query_parameters;
 use crate::routes::{
     get_task_id, is_dry_run, PaginationView, SummarizedTaskView, PAGINATION_DEFAULT_LIMIT,
@@ -367,7 +367,7 @@ pub async fn delete_document(
         .await??
     };
 
-    if network.sharding && !dry_run {
+    if network.sharding() && !dry_run {
         proxy(&index_scheduler, &index_uid, &req, network, Body::none(), &task).await?;
     }
 
@@ -1098,7 +1098,7 @@ async fn document_addition(
         }
     };
 
-    if network.sharding {
+    if network.sharding() {
         if let Some(file) = file {
             proxy(
                 &index_scheduler,
@@ -1222,7 +1222,7 @@ pub async fn delete_documents_batch(
         .await??
     };
 
-    if network.sharding && !dry_run {
+    if network.sharding() && !dry_run {
         proxy(&index_scheduler, &index_uid, &req, network, Body::Inline(body), &task).await?;
     }
 
@@ -1320,7 +1320,7 @@ pub async fn delete_documents_by_filter(
         .await??
     };
 
-    if network.sharding && !dry_run {
+    if network.sharding() && !dry_run {
         proxy(&index_scheduler, &index_uid, &req, network, Body::Inline(filter), &task).await?;
     }
 
@@ -1475,7 +1475,7 @@ pub async fn edit_documents_by_function(
         .await??
     };
 
-    if network.sharding && !dry_run {
+    if network.sharding() && !dry_run {
         proxy(&index_scheduler, &index_uid, &req, network, Body::Inline(body), &task).await?;
     }
 
@@ -1549,7 +1549,7 @@ pub async fn clear_all_documents(
         .await??
     };
 
-    if network.sharding && !dry_run {
+    if network.sharding() && !dry_run {
         proxy(&index_scheduler, &index_uid, &req, network, Body::none(), &task).await?;
     }
 

crates/meilisearch/src/routes/indexes/enterprise_edition/proxy.rs

@@ -52,7 +52,7 @@ pub async fn proxy<T: serde::Serialize>(
     index_scheduler: &IndexScheduler,
     index_uid: &str,
     req: &HttpRequest,
-    network: meilisearch_types::enterprise_edition::network::Network,
+    network: meilisearch_types::network::Network,
     body: Body<T>,
     task: &meilisearch_types::tasks::Task,
 ) -> Result<(), MeilisearchHttpError> {

crates/meilisearch/src/routes/indexes/mod.rs

@@ -30,7 +30,16 @@ use crate::Opt;
 
 pub mod compact;
 pub mod documents;
+
+#[cfg(not(feature = "enterprise"))]
+mod community_edition;
+#[cfg(feature = "enterprise")]
 mod enterprise_edition;
+#[cfg(not(feature = "enterprise"))]
+use community_edition as current_edition;
+#[cfg(feature = "enterprise")]
+use enterprise_edition as current_edition;
+
 pub mod facet_search;
 pub mod search;
 mod search_analytics;
@@ -41,7 +50,7 @@ mod settings_analytics;
 pub mod similar;
 mod similar_analytics;
 
-pub use enterprise_edition::proxy::{PROXY_ORIGIN_REMOTE_HEADER, PROXY_ORIGIN_TASK_UID_HEADER};
+pub use current_edition::proxy::{PROXY_ORIGIN_REMOTE_HEADER, PROXY_ORIGIN_TASK_UID_HEADER};
 
 #[derive(OpenApi)]
 #[openapi(

crates/meilisearch/src/routes/metrics.rs

@@ -4,6 +4,7 @@ use index_scheduler::{IndexScheduler, Query};
 use meilisearch_auth::AuthController;
 use meilisearch_types::error::ResponseError;
 use meilisearch_types::keys::actions;
+use meilisearch_types::milli::progress::ProgressStepView;
 use meilisearch_types::tasks::Status;
 use prometheus::{Encoder, TextEncoder};
 use time::OffsetDateTime;
@@ -38,6 +39,12 @@ pub fn configure(config: &mut web::ServiceConfig) {
 # HELP meilisearch_db_size_bytes Meilisearch DB Size In Bytes
 # TYPE meilisearch_db_size_bytes gauge
 meilisearch_db_size_bytes 1130496
+# HELP meilisearch_batch_running_progress_trace The currently running progress trace
+# TYPE meilisearch_batch_running_progress_trace gauge
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="document"} 0.710618582519409
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="extracting word proximity"} 0.2222222222222222
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="indexing"} 0.6666666666666666
+meilisearch_batch_running_progress_trace{batch_uid="0",step_name="processing tasks"} 0
 # HELP meilisearch_http_requests_total Meilisearch HTTP requests total
 # TYPE meilisearch_http_requests_total counter
 meilisearch_http_requests_total{method="GET",path="/metrics",status="400"} 1
@@ -61,6 +68,13 @@ meilisearch_http_response_time_seconds_bucket{method="GET",path="/metrics",le="1
 meilisearch_http_response_time_seconds_bucket{method="GET",path="/metrics",le="+Inf"} 0
 meilisearch_http_response_time_seconds_sum{method="GET",path="/metrics"} 0
 meilisearch_http_response_time_seconds_count{method="GET",path="/metrics"} 0
+# HELP meilisearch_last_finished_batches_progress_trace_ms The last few batches progress trace in milliseconds
+# TYPE meilisearch_last_finished_batches_progress_trace_ms gauge
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks"} 19360
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes"} 368
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes > preparing payloads"} 367
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > computing document changes > preparing payloads > payload"} 367
+meilisearch_last_finished_batches_progress_trace_ms{batch_uid="0",step_name="processing tasks > indexing"} 18970
 # HELP meilisearch_index_count Meilisearch Index Count
 # TYPE meilisearch_index_count gauge
 meilisearch_index_count 1
@@ -148,6 +162,50 @@ pub async fn get_metrics(
         }
     }
 
+    // Fetch and expose the current progressing step
+    crate::metrics::MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE.reset();
+    let (batches, _total) = index_scheduler.get_batches_from_authorized_indexes(
+        &Query { statuses: Some(vec![Status::Processing]), ..Query::default() },
+        auth_filters,
+    )?;
+    if let Some(batch) = batches.into_iter().next() {
+        let batch_uid = batch.uid.to_string();
+        if let Some(progress) = batch.progress {
+            for ProgressStepView { current_step, finished, total } in progress.steps {
+                crate::metrics::MEILISEARCH_BATCH_RUNNING_PROGRESS_TRACE
+                    .with_label_values(&[batch_uid.as_str(), current_step.as_ref()])
+                    // We return the completion ratio of the current step
+                    .set(finished as f64 / total as f64);
+            }
+        }
+    }
+
+    crate::metrics::MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS.reset();
+    let (batches, _total) = index_scheduler.get_batches_from_authorized_indexes(
+        // Fetch the finished batches...
+        &Query {
+            statuses: Some(vec![Status::Succeeded, Status::Failed]),
+            limit: Some(1),
+            ..Query::default()
+        },
+        auth_filters,
+    )?;
+    // ...and get the last batch only.
+    if let Some(batch) = batches.into_iter().next() {
+        let batch_uid = batch.uid.to_string();
+        for (step_name, duration_str) in batch.stats.progress_trace {
+            let Some(duration_str) = duration_str.as_str() else { continue };
+            match humantime::parse_duration(duration_str) {
+                Ok(duration) => {
+                    crate::metrics::MEILISEARCH_LAST_FINISHED_BATCHES_PROGRESS_TRACE_MS
+                        .with_label_values(&[&batch_uid, &step_name])
+                        .set(duration.as_millis() as i64);
+                }
+                Err(e) => tracing::error!("Failed to parse duration: {e}"),
+            }
+        }
+    }
+
     if let Some(last_update) = response.last_update {
         crate::metrics::MEILISEARCH_LAST_UPDATE.set(last_update.unix_timestamp());
     }

crates/meilisearch/src/routes/network.rs

@@ -7,7 +7,6 @@ use deserr::Deserr;
 use index_scheduler::IndexScheduler;
 use itertools::{EitherOrBoth, Itertools};
 use meilisearch_types::deserr::DeserrJsonError;
-use meilisearch_types::enterprise_edition::network::{Network as DbNetwork, Remote as DbRemote};
 use meilisearch_types::error::deserr_codes::{
     InvalidNetworkRemotes, InvalidNetworkSearchApiKey, InvalidNetworkSelf, InvalidNetworkSharding,
     InvalidNetworkUrl, InvalidNetworkWriteApiKey,
@@ -15,6 +14,7 @@ use meilisearch_types::error::deserr_codes::{
 use meilisearch_types::error::ResponseError;
 use meilisearch_types::keys::actions;
 use meilisearch_types::milli::update::Setting;
+use meilisearch_types::network::{Network as DbNetwork, Remote as DbRemote};
 use serde::Serialize;
 use tracing::debug;
 use utoipa::{OpenApi, ToSchema};
@@ -211,6 +211,16 @@ async fn patch_network(
     let old_network = index_scheduler.network();
     debug!(parameters = ?new_network, "Patch network");
 
+    #[cfg(not(feature = "enterprise"))]
+    if new_network.sharding.set().is_some() {
+        use meilisearch_types::error::Code;
+
+        return Err(ResponseError::from_msg(
+            "Meilisearch Enterprise Edition is required to set `network.sharding`".into(),
+            Code::RequiresEnterpriseEdition,
+        ));
+    }
+
     let merged_self = match new_network.local {
         Setting::Set(new_self) => Some(new_self),
         Setting::Reset => None,
@@ -312,6 +322,7 @@ async fn patch_network(
 
     let merged_network =
         DbNetwork { local: merged_self, remotes: merged_remotes, sharding: merged_sharding };
+
     index_scheduler.put_network(merged_network.clone())?;
     debug!(returns = ?merged_network, "Patch network");
     Ok(HttpResponse::Ok().json(merged_network))

crates/meilisearch/src/search/federated/perform.rs

@@ -9,12 +9,12 @@ use std::vec::{IntoIter, Vec};
 use actix_http::StatusCode;
 use index_scheduler::{IndexScheduler, RoFeatures};
 use itertools::Itertools;
-use meilisearch_types::enterprise_edition::network::{Network, Remote};
 use meilisearch_types::error::ResponseError;
 use meilisearch_types::milli::order_by_map::OrderByMap;
 use meilisearch_types::milli::score_details::{ScoreDetails, WeightedScoreValue};
 use meilisearch_types::milli::vector::Embedding;
 use meilisearch_types::milli::{self, DocumentId, OrderBy, TimeBudget, DEFAULT_VALUES_PER_FACET};
+use meilisearch_types::network::{Network, Remote};
 use roaring::RoaringBitmap;
 use tokio::task::JoinHandle;
 use uuid::Uuid;

crates/meilisearch/src/search/federated/proxy.rs

@@ -1,6 +1,6 @@
 pub use error::ProxySearchError;
 use error::ReqwestErrorWithoutUrl;
-use meilisearch_types::enterprise_edition::network::Remote;
+use meilisearch_types::network::Remote;
 use rand::Rng as _;
 use reqwest::{Client, Response, StatusCode};
 use serde::de::DeserializeOwned;

crates/meilisearch/tests/documents/get_documents.rs

@@ -1453,3 +1453,152 @@ async fn test_fetch_documents_pagination_with_sorting() {
     ]
     "###);
 }
+
+// <https://github.com/meilisearch/meilisearch/issues/5998>
+#[actix_rt::test]
+async fn get_document_sort_field_not_in_any_document() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (task, _code) = index.update_settings_sortable_attributes(json!(["created_at"])).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let documents = json!([
+        { "id": 1, "name": "Document 1" },
+        { "id": 2, "name": "Document 2" }
+    ]);
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (response, code) = index
+        .fetch_documents(json!({
+            "sort": ["created_at:asc"]
+        }))
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response), @r###"
+    {
+      "results": [
+        {
+          "id": 1,
+          "name": "Document 1"
+        },
+        {
+          "id": 2,
+          "name": "Document 2"
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 2
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn get_document_sort_includes_docs_without_field() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (task, _code) = index.update_settings_sortable_attributes(json!(["created_at"])).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let documents = json!([
+        { "id": 1, "name": "Doc without created_at" },
+        { "id": 2, "name": "Doc with created_at", "created_at": "2025-01-15" },
+        { "id": 3, "name": "Another doc without created_at" },
+        { "id": 4, "name": "Another doc with created_at", "created_at": "2025-01-10" }
+    ]);
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (response, code) = index
+        .fetch_documents(json!({
+            "sort": ["created_at:asc"]
+        }))
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response), @r###"
+    {
+      "results": [
+        {
+          "id": 4,
+          "name": "Another doc with created_at",
+          "created_at": "2025-01-10"
+        },
+        {
+          "id": 2,
+          "name": "Doc with created_at",
+          "created_at": "2025-01-15"
+        },
+        {
+          "id": 1,
+          "name": "Doc without created_at"
+        },
+        {
+          "id": 3,
+          "name": "Another doc without created_at"
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 4
+    }
+    "###);
+}
+
+#[actix_rt::test]
+async fn get_document_sort_desc_includes_docs_without_field() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+    let (task, _code) = index.create(None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (task, _code) = index.update_settings_sortable_attributes(json!(["priority"])).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let documents = json!([
+        { "id": 1, "name": "Low priority", "priority": 1 },
+        { "id": 2, "name": "No priority" },
+        { "id": 3, "name": "High priority", "priority": 10 }
+    ]);
+    let (task, _code) = index.add_documents(documents, None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (response, code) = index
+        .fetch_documents(json!({
+            "sort": ["priority:desc"]
+        }))
+        .await;
+
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response), @r###"
+    {
+      "results": [
+        {
+          "id": 3,
+          "name": "High priority",
+          "priority": 10
+        },
+        {
+          "id": 1,
+          "name": "Low priority",
+          "priority": 1
+        },
+        {
+          "id": 2,
+          "name": "No priority"
+        }
+      ],
+      "offset": 0,
+      "limit": 20,
+      "total": 3
+    }
+    "###);
+}

crates/meilisearch/tests/search/multi/proxy.rs

@@ -3142,6 +3142,7 @@ fn fail(override_response_body: Option<&str>) -> ResponseTemplate {
     }
 }
 
+#[cfg(feature = "enterprise")]
 #[actix_rt::test]
 async fn remote_auto_sharding() {
     let ms0 = Server::new().await;
@@ -3161,7 +3162,6 @@ async fn remote_auto_sharding() {
     snapshot!(json_string!(response["network"]), @"true");
 
     // set self & sharding
-
     let (response, code) = ms0.set_network(json!({"self": "ms0", "sharding": true})).await;
     snapshot!(code, @"200 OK");
     snapshot!(json_string!(response), @r###"
@@ -3462,6 +3462,30 @@ async fn remote_auto_sharding() {
     "###);
 }
 
+#[cfg(not(feature = "enterprise"))]
+#[actix_rt::test]
+async fn sharding_not_enterprise() {
+    let ms0 = Server::new().await;
+
+    // enable feature
+
+    let (response, code) = ms0.set_features(json!({"network": true})).await;
+    snapshot!(code, @"200 OK");
+    snapshot!(json_string!(response["network"]), @"true");
+
+    let (response, code) = ms0.set_network(json!({"self": "ms0", "sharding": true})).await;
+    snapshot!(code, @"451 Unavailable For Legal Reasons");
+    snapshot!(json_string!(response), @r###"
+    {
+      "message": "Meilisearch Enterprise Edition is required to set `network.sharding`",
+      "code": "requires_enterprise_edition",
+      "type": "invalid_request",
+      "link": "https://docs.meilisearch.com/errors#requires_enterprise_edition"
+    }
+    "###);
+}
+
+#[cfg(feature = "enterprise")]
 #[actix_rt::test]
 async fn remote_auto_sharding_with_custom_metadata() {
     let ms0 = Server::new().await;

crates/meilisearch/tests/settings/mod.rs

@@ -2,6 +2,7 @@ mod chat;
 mod distinct;
 mod errors;
 mod get_settings;
+mod parent_seachable_fields;
 mod prefix_search_settings;
 mod proximity_settings;
 mod tokenizer_customization;

crates/meilisearch/tests/settings/parent_seachable_fields.rs

@@ -0,0 +1,114 @@
+use meili_snap::{json_string, snapshot};
+use once_cell::sync::Lazy;
+
+use crate::common::Server;
+use crate::json;
+
+static DOCUMENTS: Lazy<crate::common::Value> = Lazy::new(|| {
+    json!([
+        {
+            "id": 1,
+            "meta": {
+                "title": "Soup of the day",
+                "description": "many the fish",
+            }
+        },
+        {
+            "id": 2,
+            "meta": {
+                "title": "Soup of day",
+                "description": "many the lazy fish",
+            }
+        },
+        {
+            "id": 3,
+            "meta": {
+                "title": "the Soup of day",
+                "description": "many the fish",
+            }
+        },
+    ])
+});
+
+#[actix_rt::test]
+async fn nested_field_becomes_searchable() {
+    let server = Server::new_shared();
+    let index = server.unique_index();
+
+    let (task, _status_code) = index.add_documents(DOCUMENTS.clone(), None).await;
+    server.wait_task(task.uid()).await.succeeded();
+
+    let (response, code) = index
+        .update_settings(json!({
+            "searchableAttributes": ["meta.title"]
+        }))
+        .await;
+    assert_eq!("202", code.as_str(), "{response:?}");
+    server.wait_task(response.uid()).await.succeeded();
+
+    // We expect no documents when searching for
+    // a nested non-searchable field
+    index
+        .search(json!({"q": "many fish"}), |response, code| {
+            snapshot!(code, @"200 OK");
+            snapshot!(json_string!(response["hits"]), @r###"[]"###);
+        })
+        .await;
+
+    let (response, code) = index
+        .update_settings(json!({
+            "searchableAttributes": ["meta.title", "meta.description"]
+        }))
+        .await;
+    assert_eq!("202", code.as_str(), "{response:?}");
+    server.wait_task(response.uid()).await.succeeded();
+
+    // We expect all the documents when the nested field becomes searchable
+    index
+        .search(json!({"q": "many fish"}), |response, code| {
+            snapshot!(code, @"200 OK");
+            snapshot!(json_string!(response["hits"]), @r###"
+            [
+              {
+                "id": 1,
+                "meta": {
+                  "title": "Soup of the day",
+                  "description": "many the fish"
+                }
+              },
+              {
+                "id": 3,
+                "meta": {
+                  "title": "the Soup of day",
+                  "description": "many the fish"
+                }
+              },
+              {
+                "id": 2,
+                "meta": {
+                  "title": "Soup of day",
+                  "description": "many the lazy fish"
+                }
+              }
+            ]
+            "###);
+        })
+        .await;
+
+    let (response, code) = index
+        .update_settings(json!({
+            "searchableAttributes": ["meta.title"]
+        }))
+        .await;
+    assert_eq!("202", code.as_str(), "{response:?}");
+    server.wait_task(response.uid()).await.succeeded();
+
+    // We expect no documents when searching for
+    // a nested non-searchable field
+    index
+        .search(json!({"q": "many fish"}), |response, code| {
+            snapshot!(code, @"200 OK");
+            snapshot!(json_string!(response["hits"]), @r###"[]"###);
+        })
+        .await;
+}

crates/meilisearch/tests/upgrade/mod.rs

@@ -43,7 +43,7 @@ async fn version_too_old() {
     std::fs::write(db_path.join("VERSION"), "1.11.9999").unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.27.0");
+    snapshot!(err, @"Database version 1.11.9999 is too old for the experimental dumpless upgrade feature. Please generate a dump using the v1.11.9999 and import it in the v1.28.2");
 }
 
 #[actix_rt::test]
@@ -58,7 +58,7 @@ async fn version_requires_downgrade() {
     std::fs::write(db_path.join("VERSION"), format!("{major}.{minor}.{patch}")).unwrap();
     let options = Opt { experimental_dumpless_upgrade: true, ..default_settings };
     let err = Server::new_with_options(options).await.map(|_| ()).unwrap_err();
-    snapshot!(err, @"Database version 1.27.1 is higher than the Meilisearch version 1.27.0. Downgrade is not supported");
+    snapshot!(err, @"Database version 1.28.3 is higher than the Meilisearch version 1.28.2. Downgrade is not supported");
 }
 
 #[actix_rt::test]

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/batches_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterEnqueuedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterFinishedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/tasks_filter_afterStartedAt_equal_2025-01-16T16_47_41.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "error": null,
       "duration": "[duration]",

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_batch_queue_once_everything_has_been_processed.snap

@@ -8,7 +8,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "progress": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "stats": {
         "totalNbTasks": 1,

crates/meilisearch/tests/upgrade/v1_12/snapshots/v1_12_0.rs/check_the_index_scheduler/the_whole_task_queue_once_everything_has_been_processed.snap

@@ -12,7 +12,7 @@ source: crates/meilisearch/tests/upgrade/v1_12/v1_12_0.rs
       "canceledBy": null,
       "details": {
         "upgradeFrom": "v1.12.0",
-        "upgradeTo": "v1.27.0"
+        "upgradeTo": "v1.28.2"
       },
       "error": null,
       "duration": "[duration]",

crates/meilitool/Cargo.toml

@@ -9,15 +9,15 @@ edition.workspace = true
 license.workspace = true
 
 [dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
-clap = { version = "4.5.40", features = ["derive"] }
+clap = { version = "4.5.52", features = ["derive"] }
 dump = { path = "../dump" }
 file-store = { path = "../file-store" }
-indexmap = { version = "2.9.0", features = ["serde"] }
+indexmap = { version = "2.12.0", features = ["serde"] }
 meilisearch-auth = { path = "../meilisearch-auth" }
 meilisearch-types = { path = "../meilisearch-types" }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order"] }
+serde_json = { version = "1.0.145", features = ["preserve_order"] }
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-time = { version = "0.3.41", features = ["formatting", "parsing", "alloc"] }
+time = { version = "0.3.44", features = ["formatting", "parsing", "alloc"] }
-uuid = { version = "1.17.0", features = ["v4"], default-features = false }
+uuid = { version = "1.18.1", features = ["v4"], default-features = false }

crates/milli/Cargo.toml

@@ -15,15 +15,15 @@ license.workspace = true
 big_s = "1.0.2"
 bimap = { version = "0.6.3", features = ["serde"] }
 bincode = "1.3.3"
-bstr = "1.12.0"
+bstr = "1.12.1"
-bytemuck = { version = "1.23.1", features = ["extern_crate_alloc"] }
+bytemuck = { version = "1.24.0", features = ["extern_crate_alloc"] }
 byteorder = "1.5.0"
-charabia = { version = "0.9.8", default-features = false }
+charabia = { version = "0.9.9", default-features = false }
 cellulite = "0.3.1-nested-rtxns-2"
 concat-arrays = "0.1.2"
-convert_case = "0.8.0"
+convert_case = "0.9.0"
 crossbeam-channel = "0.5.15"
-deserr = "0.6.3"
+deserr = "0.6.4"
 either = { version = "1.15.0", features = ["serde"] }
 flatten-serde-json = { path = "../flatten-serde-json" }
 fst = "0.4.7"
@@ -38,39 +38,39 @@ heed = { version = "0.22.1-nested-rtxns-6", default-features = false, features =
     "serde-json",
     "serde-bincode",
 ] }
-indexmap = { version = "2.9.0", features = ["serde"] }
+indexmap = { version = "2.12.0", features = ["serde"] }
 json-depth-checker = { path = "../json-depth-checker" }
 levenshtein_automata = { version = "0.2.1", features = ["fst_automaton"] }
-memchr = "2.7.5"
+memchr = "2.7.6"
-memmap2 = "0.9.7"
+memmap2 = "0.9.9"
 obkv = "0.3.0"
 once_cell = "1.21.3"
-ordered-float = "5.0.0"
+ordered-float = "5.1.0"
-rayon = "1.10.0"
+rayon = "1.11.0"
 roaring = { version = "0.10.12", features = ["serde"] }
 rstar = { version = "0.12.2", features = ["serde"] }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = { version = "1.0.140", features = ["preserve_order", "raw_value"] }
+serde_json = { version = "1.0.145", features = ["preserve_order", "raw_value"] }
 slice-group-by = "0.3.1"
-smallstr = { version = "0.3.0", features = ["serde"] }
+smallstr = { version = "0.3.1", features = ["serde"] }
 smallvec = "1.15.1"
 smartstring = "1.0.1"
-tempfile = "3.20.0"
+tempfile = "3.23.0"
-thiserror = "2.0.12"
+thiserror = "2.0.17"
-time = { version = "0.3.41", features = [
+time = { version = "0.3.44", features = [
     "serde-well-known",
     "formatting",
     "parsing",
     "macros",
 ] }
-uuid = { version = "1.17.0", features = ["v4"] }
+uuid = { version = "1.18.1", features = ["v4"] }
 
 filter-parser = { path = "../filter-parser" }
 
 # documents words self-join
 itertools = "0.14.0"
 
-csv = "1.3.1"
+csv = "1.4.0"
 candle-core = { version = "0.9.1" }
 candle-transformers = { version = "0.9.1" }
 candle-nn = { version = "0.9.1" }
@@ -81,9 +81,9 @@ hf-hub = { git = "https://github.com/dureuill/hf-hub.git", branch = "rust_tls",
     "online",
 ] }
 safetensors = "0.6.2"
-tiktoken-rs = "0.7.0"
+tiktoken-rs = "0.9.1"
 liquid = "0.26.11"
-rhai = { version = "1.22.2", features = [
+rhai = { version = "1.23.6", features = [
     "serde",
     "no_module",
     "no_custom_syntax",
@@ -95,14 +95,14 @@ hannoy = { version = "0.0.9-nested-rtxns-2", features = ["arroy"] }
 rand = "0.8.5"
 tracing = "0.1.41"
 ureq = { version = "2.12.1", features = ["json"] }
-url = "2.5.4"
+url = "2.5.7"
-hashbrown = "0.15.4"
+hashbrown = "0.15.5"
-bumpalo = "3.18.1"
+bumpalo = "3.19.0"
 bumparaw-collections = "0.1.4"
 steppe = { version = "0.4", default-features = false }
 thread_local = "1.1.9"
 rustc-hash = "2.1.1"
-enum-iterator = "2.1.0"
+enum-iterator = "2.3.0"
 bbqueue = { git = "https://github.com/meilisearch/bbqueue" }
 flume = { version = "0.11.1", default-features = false }
 utoipa = { version = "5.4.0", features = [
@@ -112,21 +112,21 @@ utoipa = { version = "5.4.0", features = [
     "time",
     "openapi_extensions",
 ] }
-lru = "0.14.0"
+lru = "0.16.2"
-twox-hash = { version = "2.1.1", default-features = false, features = [
+twox-hash = { version = "2.1.2", default-features = false, features = [
     "std",
     "xxhash3_64",
     "xxhash64",
 ] }
-geo-types = "0.7.16"
+geo-types = "0.7.17"
 zerometry = "0.3.0"
 
 [dev-dependencies]
-mimalloc = { version = "0.1.47", default-features = false }
+mimalloc = { version = "0.1.48", default-features = false }
 # fixed version due to format breakages in v1.40
 insta = "=1.39.0"
 maplit = "1.0.2"
-md5 = "0.7.0"
+md5 = "0.8.0"
 meili-snap = { path = "../meili-snap" }
 rand = { version = "0.8.5", features = ["small_rng"] }
 
@@ -141,6 +141,8 @@ lmdb-posix-sem = ["heed/posix-sem"]
 chinese = ["charabia/chinese"]
 chinese-pinyin = ["chinese", "charabia/chinese-normalization-pinyin"]
 
+enterprise = []
+
 # allow hebrew specialized tokenization
 hebrew = ["charabia/hebrew"]
 

crates/milli/src/documents/sort.rs

@@ -240,15 +240,25 @@ impl<'ctx> SortedDocumentsIteratorBuilder<'ctx> {
     ) -> crate::Result<SortedDocumentsIterator<'ctx>> {
         let size = candidates.len() as usize;
 
+        // Get documents that have this facet field
+        let faceted_candidates = index.exists_faceted_documents_ids(rtxn, field_id)?;
+        // Documents that don't have this facet field should be returned at the end
+        let not_faceted_candidates = &candidates - &faceted_candidates;
+        // Only sort candidates that have the facet field
+        let faceted_candidates = candidates & faceted_candidates;
+        let mut not_faceted_candidates = Some(not_faceted_candidates);
+
         // Perform the sort on the first field
         let (number_iter, string_iter) = if ascending {
-            let number_iter = ascending_facet_sort(rtxn, number_db, field_id, candidates.clone())?;
+            let number_iter =
-            let string_iter = ascending_facet_sort(rtxn, string_db, field_id, candidates)?;
+                ascending_facet_sort(rtxn, number_db, field_id, faceted_candidates.clone())?;
+            let string_iter = ascending_facet_sort(rtxn, string_db, field_id, faceted_candidates)?;
 
             (itertools::Either::Left(number_iter), itertools::Either::Left(string_iter))
         } else {
-            let number_iter = descending_facet_sort(rtxn, number_db, field_id, candidates.clone())?;
+            let number_iter =
-            let string_iter = descending_facet_sort(rtxn, string_db, field_id, candidates)?;
+                descending_facet_sort(rtxn, number_db, field_id, faceted_candidates.clone())?;
+            let string_iter = descending_facet_sort(rtxn, string_db, field_id, faceted_candidates)?;
 
             (itertools::Either::Right(number_iter), itertools::Either::Right(string_iter))
         };
@@ -256,17 +266,37 @@ impl<'ctx> SortedDocumentsIteratorBuilder<'ctx> {
         // Create builders for the next level of the tree
         let number_iter = number_iter.map(|r| r.map(|(d, _)| d));
         let string_iter = string_iter.map(|r| r.map(|(d, _)| d));
-        let next_children = number_iter.chain(string_iter).map(move |r| {
+        // Chain faceted documents with non-faceted documents at the end
-            Ok(SortedDocumentsIteratorBuilder {
+        let next_children = number_iter
-                index,
+            .chain(string_iter)
-                rtxn,
+            .map(move |r| {
-                number_db,
+                Ok(SortedDocumentsIteratorBuilder {
-                string_db,
+                    index,
-                fields: next_fields,
+                    rtxn,
-                candidates: r?,
+                    number_db,
-                geo_candidates,
+                    string_db,
+                    fields: next_fields,
+                    candidates: r?,
+                    geo_candidates,
+                })
             })
-        });
+            .chain(std::iter::from_fn(move || {
+                // Once all faceted candidates have been processed, return the non-faceted ones
+                if let Some(not_faceted) = not_faceted_candidates.take() {
+                    if !not_faceted.is_empty() {
+                        return Some(Ok(SortedDocumentsIteratorBuilder {
+                            index,
+                            rtxn,
+                            number_db,
+                            string_db,
+                            fields: next_fields,
+                            candidates: not_faceted,
+                            geo_candidates,
+                        }));
+                    }
+                }
+                None
+            }));
 
         Ok(SortedDocumentsIterator::Branch {
             current_child: None,
@@ -398,10 +428,14 @@ pub fn recursive_sort<'ctx>(
         };
         if let Some((field, ascending)) = field {
             if is_faceted(&field, &sortable_fields) {
+                // The field may be in sortable_fields but not in fields_ids_map if no document
+                // has ever contained this field. In that case, we just skip this sort criterion
+                // since there are no values to sort by. Documents will be returned in their
+                // default order for this field.
                 if let Some(field_id) = fields_ids_map.id(&field) {
                     fields.push(AscDescId::Facet { field_id, ascending });
-                    continue;
                 }
+                continue;
             }
             return Err(UserError::InvalidDocumentSortableAttribute {
                 field: field.to_string(),

crates/milli/src/fields_ids_map/metadata.rs

@@ -18,6 +18,8 @@ use crate::{
 pub struct Metadata {
     /// The weight as defined in the FieldidsWeightsMap of the searchable attribute if it is searchable.
     pub searchable: Option<Weight>,
+    /// The field is part of the exact attributes.
+    pub exact: bool,
     /// The field is part of the sortable attributes.
     pub sortable: bool,
     /// The field is defined as the distinct attribute.
@@ -209,6 +211,7 @@ impl Metadata {
 #[derive(Debug, Clone)]
 pub struct MetadataBuilder {
     searchable_attributes: Option<Vec<String>>,
+    exact_searchable_attributes: Vec<String>,
     filterable_attributes: Vec<FilterableAttributesRule>,
     sortable_attributes: HashSet<String>,
     localized_attributes: Option<Vec<LocalizedAttributesRule>>,
@@ -220,15 +223,18 @@ impl MetadataBuilder {
     pub fn from_index(index: &Index, rtxn: &RoTxn) -> Result<Self> {
         let searchable_attributes = index
             .user_defined_searchable_fields(rtxn)?
-            .map(|fields| fields.into_iter().map(|s| s.to_string()).collect());
+            .map(|fields| fields.into_iter().map(String::from).collect());
+        let exact_searchable_attributes =
+            index.exact_attributes(rtxn)?.into_iter().map(String::from).collect();
         let filterable_attributes = index.filterable_attributes_rules(rtxn)?;
         let sortable_attributes = index.sortable_fields(rtxn)?;
         let localized_attributes = index.localized_attributes_rules(rtxn)?;
-        let distinct_attribute = index.distinct_field(rtxn)?.map(|s| s.to_string());
+        let distinct_attribute = index.distinct_field(rtxn)?.map(String::from);
         let asc_desc_attributes = index.asc_desc_fields(rtxn)?;
 
         Ok(Self::new(
             searchable_attributes,
+            exact_searchable_attributes,
             filterable_attributes,
             sortable_attributes,
             localized_attributes,
@@ -242,6 +248,7 @@ impl MetadataBuilder {
     /// This is used for testing, prefer using `MetadataBuilder::from_index` instead.
     pub fn new(
         searchable_attributes: Option<Vec<String>>,
+        exact_searchable_attributes: Vec<String>,
         filterable_attributes: Vec<FilterableAttributesRule>,
         sortable_attributes: HashSet<String>,
         localized_attributes: Option<Vec<LocalizedAttributesRule>>,
@@ -256,6 +263,7 @@ impl MetadataBuilder {
 
         Self {
             searchable_attributes,
+            exact_searchable_attributes,
             filterable_attributes,
             sortable_attributes,
             localized_attributes,
@@ -269,6 +277,7 @@ impl MetadataBuilder {
             // Vectors fields are not searchable, filterable, distinct or asc_desc
             return Metadata {
                 searchable: None,
+                exact: false,
                 sortable: false,
                 distinct: false,
                 asc_desc: false,
@@ -296,6 +305,7 @@ impl MetadataBuilder {
             // Geo fields are not searchable, distinct or asc_desc
             return Metadata {
                 searchable: None,
+                exact: false,
                 sortable,
                 distinct: false,
                 asc_desc: false,
@@ -309,6 +319,7 @@ impl MetadataBuilder {
             debug_assert!(!sortable, "geojson fields should not be sortable");
             return Metadata {
                 searchable: None,
+                exact: false,
                 sortable,
                 distinct: false,
                 asc_desc: false,
@@ -329,6 +340,8 @@ impl MetadataBuilder {
             None => Some(0),
         };
 
+        let exact = self.exact_searchable_attributes.iter().any(|attr| is_faceted_by(field, attr));
+
         let distinct =
             self.distinct_attribute.as_ref().is_some_and(|distinct_field| field == distinct_field);
         let asc_desc = self.asc_desc_attributes.contains(field);
@@ -343,6 +356,7 @@ impl MetadataBuilder {
 
         Metadata {
             searchable,
+            exact,
             sortable,
             distinct,
             asc_desc,

crates/milli/src/update/new/extract/searchable/extract_word_docids.rs

@@ -8,17 +8,26 @@ use bumpalo::Bump;
 
 use super::match_searchable_field;
 use super::tokenize_document::{tokenizer_builder, DocumentTokenizer};
+use crate::fields_ids_map::metadata::Metadata;
 use crate::update::new::document::DocumentContext;
 use crate::update::new::extract::cache::BalancedCaches;
 use crate::update::new::extract::perm_json_p::contained_in;
+use crate::update::new::extract::searchable::has_searchable_children;
 use crate::update::new::indexer::document_changes::{
     extract, DocumentChanges, Extractor, IndexingContext,
 };
+use crate::update::new::indexer::settings_changes::{
+    settings_change_extract, DocumentsIndentifiers, SettingsChangeExtractor,
+};
 use crate::update::new::ref_cell_ext::RefCellExt as _;
 use crate::update::new::steps::IndexingStep;
 use crate::update::new::thread_local::{FullySend, MostlySend, ThreadLocal};
-use crate::update::new::DocumentChange;
+use crate::update::new::{DocumentChange, DocumentIdentifiers};
-use crate::{bucketed_position, DocumentId, FieldId, Result, MAX_POSITION_PER_ATTRIBUTE};
+use crate::update::settings::SettingsDelta;
+use crate::{
+    bucketed_position, DocumentId, FieldId, PatternMatch, Result, UserError,
+    MAX_POSITION_PER_ATTRIBUTE,
+};
 
 const MAX_COUNTED_WORDS: usize = 30;
 
@@ -34,6 +43,15 @@ pub struct WordDocidsBalancedCaches<'extractor> {
 
 unsafe impl MostlySend for WordDocidsBalancedCaches<'_> {}
 
+/// Whether to extract or skip fields during word extraction.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+enum FieldDbExtraction {
+    /// Extract the word and put it in to the fid-based databases.
+    Extract,
+    /// Do not store the word in the fid-based databases.
+    Skip,
+}
+
 impl<'extractor> WordDocidsBalancedCaches<'extractor> {
     pub fn new_in(buckets: usize, max_memory: Option<usize>, alloc: &'extractor Bump) -> Self {
         Self {
@@ -47,12 +65,14 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
         }
     }
 
+    #[allow(clippy::too_many_arguments)]
     fn insert_add_u32(
         &mut self,
         field_id: FieldId,
         position: u16,
         word: &str,
         exact: bool,
+        field_db_extraction: FieldDbExtraction,
         docid: u32,
         bump: &Bump,
     ) -> Result<()> {
@@ -66,11 +86,13 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
         let buffer_size = word_bytes.len() + 1 + size_of::<FieldId>();
         let mut buffer = BumpVec::with_capacity_in(buffer_size, bump);
 
-        buffer.clear();
+        if field_db_extraction == FieldDbExtraction::Extract {
-        buffer.extend_from_slice(word_bytes);
+            buffer.clear();
-        buffer.push(0);
+            buffer.extend_from_slice(word_bytes);
-        buffer.extend_from_slice(&field_id.to_be_bytes());
+            buffer.push(0);
-        self.word_fid_docids.insert_add_u32(&buffer, docid)?;
+            buffer.extend_from_slice(&field_id.to_be_bytes());
+            self.word_fid_docids.insert_add_u32(&buffer, docid)?;
+        }
 
         let position = bucketed_position(position);
         buffer.clear();
@@ -83,21 +105,26 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
             self.flush_fid_word_count(&mut buffer)?;
         }
 
-        self.fid_word_count
+        if field_db_extraction == FieldDbExtraction::Extract {
-            .entry(field_id)
+            self.fid_word_count
-            .and_modify(|(_current_count, new_count)| *new_count.get_or_insert(0) += 1)
+                .entry(field_id)
-            .or_insert((None, Some(1)));
+                .and_modify(|(_current_count, new_count)| *new_count.get_or_insert(0) += 1)
+                .or_insert((None, Some(1)));
+        }
+
         self.current_docid = Some(docid);
 
         Ok(())
     }
 
+    #[allow(clippy::too_many_arguments)]
     fn insert_del_u32(
         &mut self,
         field_id: FieldId,
         position: u16,
         word: &str,
         exact: bool,
+        field_db_extraction: FieldDbExtraction,
         docid: u32,
         bump: &Bump,
     ) -> Result<()> {
@@ -111,11 +138,13 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
         let buffer_size = word_bytes.len() + 1 + size_of::<FieldId>();
         let mut buffer = BumpVec::with_capacity_in(buffer_size, bump);
 
-        buffer.clear();
+        if field_db_extraction == FieldDbExtraction::Extract {
-        buffer.extend_from_slice(word_bytes);
+            buffer.clear();
-        buffer.push(0);
+            buffer.extend_from_slice(word_bytes);
-        buffer.extend_from_slice(&field_id.to_be_bytes());
+            buffer.push(0);
-        self.word_fid_docids.insert_del_u32(&buffer, docid)?;
+            buffer.extend_from_slice(&field_id.to_be_bytes());
+            self.word_fid_docids.insert_del_u32(&buffer, docid)?;
+        }
 
         let position = bucketed_position(position);
         buffer.clear();
@@ -128,10 +157,12 @@ impl<'extractor> WordDocidsBalancedCaches<'extractor> {
             self.flush_fid_word_count(&mut buffer)?;
         }
 
-        self.fid_word_count
+        if field_db_extraction == FieldDbExtraction::Extract {
-            .entry(field_id)
+            self.fid_word_count
-            .and_modify(|(current_count, _new_count)| *current_count.get_or_insert(0) += 1)
+                .entry(field_id)
-            .or_insert((Some(1), None));
+                .and_modify(|(current_count, _new_count)| *current_count.get_or_insert(0) += 1)
+                .or_insert((Some(1), None));
+        }
 
         self.current_docid = Some(docid);
 
@@ -325,6 +356,24 @@ impl WordDocidsExtractors {
             exact_attributes.iter().any(|attr| contained_in(fname, attr))
                 || disabled_typos_terms.is_exact(word)
         };
+
+        let mut should_tokenize = |field_name: &str| {
+            let Some((field_id, meta)) = new_fields_ids_map.id_with_metadata_or_insert(field_name)
+            else {
+                return Err(UserError::AttributeLimitReached.into());
+            };
+
+            let pattern_match = if meta.is_searchable() {
+                PatternMatch::Match
+            } else {
+                // TODO: should be a match on the field_name using `match_field_legacy` function,
+                //       but for legacy reasons we iterate over all the fields to fill the field_id_map.
+                PatternMatch::Parent
+            };
+
+            Ok((field_id, pattern_match))
+        };
+
         match document_change {
             DocumentChange::Deletion(inner) => {
                 let mut token_fn = |fname: &str, fid, pos, word: &str| {
@@ -333,13 +382,14 @@ impl WordDocidsExtractors {
                         pos,
                         word,
                         is_exact(fname, word),
+                        FieldDbExtraction::Extract,
                         inner.docid(),
                         doc_alloc,
                     )
                 };
                 document_tokenizer.tokenize_document(
                     inner.current(rtxn, index, context.db_fields_ids_map)?,
-                    new_fields_ids_map,
+                    &mut should_tokenize,
                     &mut token_fn,
                 )?;
             }
@@ -361,13 +411,14 @@ impl WordDocidsExtractors {
                         pos,
                         word,
                         is_exact(fname, word),
+                        FieldDbExtraction::Extract,
                         inner.docid(),
                         doc_alloc,
                     )
                 };
                 document_tokenizer.tokenize_document(
                     inner.current(rtxn, index, context.db_fields_ids_map)?,
-                    new_fields_ids_map,
+                    &mut should_tokenize,
                     &mut token_fn,
                 )?;
 
@@ -377,13 +428,14 @@ impl WordDocidsExtractors {
                         pos,
                         word,
                         is_exact(fname, word),
+                        FieldDbExtraction::Extract,
                         inner.docid(),
                         doc_alloc,
                     )
                 };
                 document_tokenizer.tokenize_document(
                     inner.merged(rtxn, index, context.db_fields_ids_map)?,
-                    new_fields_ids_map,
+                    &mut should_tokenize,
                     &mut token_fn,
                 )?;
             }
@@ -394,13 +446,14 @@ impl WordDocidsExtractors {
                         pos,
                         word,
                         is_exact(fname, word),
+                        FieldDbExtraction::Extract,
                         inner.docid(),
                         doc_alloc,
                     )
                 };
                 document_tokenizer.tokenize_document(
                     inner.inserted(),
-                    new_fields_ids_map,
+                    &mut should_tokenize,
                     &mut token_fn,
                 )?;
             }
@@ -411,3 +464,292 @@ impl WordDocidsExtractors {
         cached_sorter.flush_fid_word_count(&mut buffer)
     }
 }
+
+pub struct WordDocidsSettingsExtractorsData<'a, SD> {
+    tokenizer: DocumentTokenizer<'a>,
+    max_memory_by_thread: Option<usize>,
+    buckets: usize,
+    settings_delta: &'a SD,
+}
+
+impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
+    for WordDocidsSettingsExtractorsData<'_, SD>
+{
+    type Data = RefCell<Option<WordDocidsBalancedCaches<'extractor>>>;
+
+    fn init_data<'doc>(&'doc self, extractor_alloc: &'extractor Bump) -> crate::Result<Self::Data> {
+        Ok(RefCell::new(Some(WordDocidsBalancedCaches::new_in(
+            self.buckets,
+            self.max_memory_by_thread,
+            extractor_alloc,
+        ))))
+    }
+
+    fn process<'doc>(
+        &'doc self,
+        documents: impl Iterator<Item = crate::Result<DocumentIdentifiers<'doc>>>,
+        context: &'doc DocumentContext<Self::Data>,
+    ) -> crate::Result<()> {
+        for document in documents {
+            let document = document?;
+            SettingsChangeWordDocidsExtractors::extract_document_from_settings_change(
+                document,
+                context,
+                &self.tokenizer,
+                self.settings_delta,
+            )?;
+        }
+        Ok(())
+    }
+}
+
+pub struct SettingsChangeWordDocidsExtractors;
+
+impl SettingsChangeWordDocidsExtractors {
+    pub fn run_extraction<'fid, 'indexer, 'index, 'extractor, SD, MSP>(
+        settings_delta: &SD,
+        documents: &'indexer DocumentsIndentifiers<'indexer>,
+        indexing_context: IndexingContext<'fid, 'indexer, 'index, MSP>,
+        extractor_allocs: &'extractor mut ThreadLocal<FullySend<Bump>>,
+        step: IndexingStep,
+    ) -> Result<WordDocidsCaches<'extractor>>
+    where
+        SD: SettingsDelta + Sync,
+        MSP: Fn() -> bool + Sync,
+    {
+        // Warning: this is duplicated code from extract_word_pair_proximity_docids.rs
+        // TODO we need to read the new AND old settings to support changing global parameters
+        let rtxn = indexing_context.index.read_txn()?;
+        let stop_words = indexing_context.index.stop_words(&rtxn)?;
+        let allowed_separators = indexing_context.index.allowed_separators(&rtxn)?;
+        let allowed_separators: Option<Vec<_>> =
+            allowed_separators.as_ref().map(|s| s.iter().map(String::as_str).collect());
+        let dictionary = indexing_context.index.dictionary(&rtxn)?;
+        let dictionary: Option<Vec<_>> =
+            dictionary.as_ref().map(|s| s.iter().map(String::as_str).collect());
+        let mut builder = tokenizer_builder(
+            stop_words.as_ref(),
+            allowed_separators.as_deref(),
+            dictionary.as_deref(),
+        );
+        let tokenizer = builder.build();
+        let localized_attributes_rules =
+            indexing_context.index.localized_attributes_rules(&rtxn)?.unwrap_or_default();
+        let document_tokenizer = DocumentTokenizer {
+            tokenizer: &tokenizer,
+            localized_attributes_rules: &localized_attributes_rules,
+            max_positions_per_attributes: MAX_POSITION_PER_ATTRIBUTE,
+        };
+        let extractor_data = WordDocidsSettingsExtractorsData {
+            tokenizer: document_tokenizer,
+            max_memory_by_thread: indexing_context.grenad_parameters.max_memory_by_thread(),
+            buckets: rayon::current_num_threads(),
+            settings_delta,
+        };
+        let datastore = ThreadLocal::new();
+        {
+            let span = tracing::debug_span!(target: "indexing::documents::extract", "vectors");
+            let _entered = span.enter();
+
+            settings_change_extract(
+                documents,
+                &extractor_data,
+                indexing_context,
+                extractor_allocs,
+                &datastore,
+                step,
+            )?;
+        }
+
+        let mut merger = WordDocidsCaches::new();
+        for cache in datastore.into_iter().flat_map(RefCell::into_inner) {
+            merger.push(cache)?;
+        }
+
+        Ok(merger)
+    }
+
+    /// Extracts document words from a settings change.
+    fn extract_document_from_settings_change<SD: SettingsDelta>(
+        document: DocumentIdentifiers<'_>,
+        context: &DocumentContext<RefCell<Option<WordDocidsBalancedCaches>>>,
+        document_tokenizer: &DocumentTokenizer,
+        settings_delta: &SD,
+    ) -> Result<()> {
+        let mut cached_sorter_ref = context.data.borrow_mut_or_yield();
+        let cached_sorter = cached_sorter_ref.as_mut().unwrap();
+        let doc_alloc = &context.doc_alloc;
+
+        let new_fields_ids_map = settings_delta.new_fields_ids_map();
+        let old_fields_ids_map = context.index.fields_ids_map_with_metadata(&context.rtxn)?;
+        let old_searchable = settings_delta.old_searchable_attributes().as_ref();
+        let new_searchable = settings_delta.new_searchable_attributes().as_ref();
+
+        let current_document = document.current(
+            &context.rtxn,
+            context.index,
+            old_fields_ids_map.as_fields_ids_map(),
+        )?;
+
+        #[derive(Debug, Clone, Copy, PartialEq)]
+        enum ActionToOperate {
+            ReindexAllFields,
+            // TODO improve by listing field prefixes
+            IndexAddedFields,
+            SkipDocument,
+        }
+
+        let mut action = ActionToOperate::SkipDocument;
+        // Here we do a preliminary check to determine the action to take.
+        // This check doesn't trigger the tokenizer as we never return
+        // PatternMatch::Match.
+        document_tokenizer.tokenize_document(
+            current_document,
+            &mut |field_name| {
+                let fid = new_fields_ids_map.id(field_name).expect("All fields IDs must exist");
+
+                // If the document must be reindexed, early return NoMatch to stop the scanning process.
+                if action == ActionToOperate::ReindexAllFields {
+                    return Ok((fid, PatternMatch::NoMatch));
+                }
+
+                let old_field_metadata = old_fields_ids_map.metadata(fid).unwrap();
+                let new_field_metadata = new_fields_ids_map.metadata(fid).unwrap();
+
+                action = match (old_field_metadata, new_field_metadata) {
+                    // At least one field is added or removed from the exact fields => ReindexAllFields
+                    (Metadata { exact: old_exact, .. }, Metadata { exact: new_exact, .. })
+                        if old_exact != new_exact =>
+                    {
+                        ActionToOperate::ReindexAllFields
+                    }
+                    // At least one field is removed from the searchable fields => ReindexAllFields
+                    (Metadata { searchable: Some(_), .. }, Metadata { searchable: None, .. }) => {
+                        ActionToOperate::ReindexAllFields
+                    }
+                    // At least one field is added in the searchable fields => IndexAddedFields
+                    (Metadata { searchable: None, .. }, Metadata { searchable: Some(_), .. }) => {
+                        // We can safely overwrite the action, because we early return when action is ReindexAllFields.
+                        ActionToOperate::IndexAddedFields
+                    }
+                    _ => action,
+                };
+
+                Ok((fid, PatternMatch::Parent))
+            },
+            &mut |_, _, _, _| Ok(()),
+        )?;
+
+        // Early return when we don't need to index the document
+        if action == ActionToOperate::SkipDocument {
+            return Ok(());
+        }
+
+        let mut should_tokenize = |field_name: &str| {
+            let field_id = new_fields_ids_map.id(field_name).expect("All fields IDs must exist");
+            let old_field_metadata = old_fields_ids_map.metadata(field_id).unwrap();
+            let new_field_metadata = new_fields_ids_map.metadata(field_id).unwrap();
+
+            let pattern_match = match action {
+                ActionToOperate::ReindexAllFields => {
+                    if old_field_metadata.is_searchable() || new_field_metadata.is_searchable() {
+                        PatternMatch::Match
+                    // If any old or new field is searchable then we need to iterate over all fields
+                    // else if any field matches we need to iterate over all fields
+                    } else if has_searchable_children(
+                        field_name,
+                        old_searchable.zip(new_searchable).map(|(old, new)| old.iter().chain(new)),
+                    ) {
+                        PatternMatch::Parent
+                    } else {
+                        PatternMatch::NoMatch
+                    }
+                }
+                ActionToOperate::IndexAddedFields => {
+                    // Was not searchable but now is
+                    if !old_field_metadata.is_searchable() && new_field_metadata.is_searchable() {
+                        PatternMatch::Match
+                    // If the field is now a parent of a searchable field
+                    } else if has_searchable_children(field_name, new_searchable) {
+                        PatternMatch::Parent
+                    } else {
+                        PatternMatch::NoMatch
+                    }
+                }
+                ActionToOperate::SkipDocument => unreachable!(),
+            };
+
+            Ok((field_id, pattern_match))
+        };
+
+        let old_disabled_typos_terms = settings_delta.old_disabled_typos_terms();
+        let new_disabled_typos_terms = settings_delta.new_disabled_typos_terms();
+        let mut token_fn = |_field_name: &str, field_id, pos, word: &str| {
+            let old_field_metadata = old_fields_ids_map.metadata(field_id).unwrap();
+            let new_field_metadata = new_fields_ids_map.metadata(field_id).unwrap();
+
+            match (old_field_metadata, new_field_metadata) {
+                (
+                    Metadata { searchable: Some(_), exact: old_exact, .. },
+                    Metadata { searchable: None, .. },
+                ) => cached_sorter.insert_del_u32(
+                    field_id,
+                    pos,
+                    word,
+                    old_exact || old_disabled_typos_terms.is_exact(word),
+                    // We deleted the field globally
+                    FieldDbExtraction::Skip,
+                    document.docid(),
+                    doc_alloc,
+                ),
+                (
+                    Metadata { searchable: None, .. },
+                    Metadata { searchable: Some(_), exact: new_exact, .. },
+                ) => cached_sorter.insert_add_u32(
+                    field_id,
+                    pos,
+                    word,
+                    new_exact || new_disabled_typos_terms.is_exact(word),
+                    FieldDbExtraction::Extract,
+                    document.docid(),
+                    doc_alloc,
+                ),
+                (Metadata { searchable: None, .. }, Metadata { searchable: None, .. }) => {
+                    unreachable!()
+                }
+                (Metadata { exact: old_exact, .. }, Metadata { exact: new_exact, .. }) => {
+                    cached_sorter.insert_del_u32(
+                        field_id,
+                        pos,
+                        word,
+                        old_exact || old_disabled_typos_terms.is_exact(word),
+                        // The field has already been extracted
+                        FieldDbExtraction::Skip,
+                        document.docid(),
+                        doc_alloc,
+                    )?;
+                    cached_sorter.insert_add_u32(
+                        field_id,
+                        pos,
+                        word,
+                        new_exact || new_disabled_typos_terms.is_exact(word),
+                        // The field has already been extracted
+                        FieldDbExtraction::Skip,
+                        document.docid(),
+                        doc_alloc,
+                    )
+                }
+            }
+        };
+
+        // TODO we must tokenize twice when we change global parameters like stop words,
+        //      the language settings, dictionary, separators, non-separators...
+        document_tokenizer.tokenize_document(
+            current_document,
+            &mut should_tokenize,
+            &mut token_fn,
+        )?;
+
+        Ok(())
+    }
+}

crates/milli/src/update/new/extract/searchable/extract_word_pair_proximity_docids.rs

@@ -6,17 +6,24 @@ use bumpalo::Bump;
 
 use super::match_searchable_field;
 use super::tokenize_document::{tokenizer_builder, DocumentTokenizer};
+use crate::fields_ids_map::metadata::Metadata;
+use crate::proximity::ProximityPrecision::*;
 use crate::proximity::{index_proximity, MAX_DISTANCE};
 use crate::update::new::document::{Document, DocumentContext};
 use crate::update::new::extract::cache::BalancedCaches;
 use crate::update::new::indexer::document_changes::{
     extract, DocumentChanges, Extractor, IndexingContext,
 };
+use crate::update::new::indexer::settings_change_extract;
+use crate::update::new::indexer::settings_changes::{
+    DocumentsIndentifiers, SettingsChangeExtractor,
+};
 use crate::update::new::ref_cell_ext::RefCellExt as _;
 use crate::update::new::steps::IndexingStep;
 use crate::update::new::thread_local::{FullySend, ThreadLocal};
-use crate::update::new::DocumentChange;
+use crate::update::new::{DocumentChange, DocumentIdentifiers};
-use crate::{FieldId, GlobalFieldsIdsMap, Result, MAX_POSITION_PER_ATTRIBUTE};
+use crate::update::settings::SettingsDelta;
+use crate::{FieldId, PatternMatch, Result, UserError, MAX_POSITION_PER_ATTRIBUTE};
 
 pub struct WordPairProximityDocidsExtractorData<'a> {
     tokenizer: DocumentTokenizer<'a>,
@@ -116,7 +123,7 @@ impl WordPairProximityDocidsExtractor {
     // and to store the docids of the documents that have a number of words in a given field
     // equal to or under than MAX_COUNTED_WORDS.
     fn extract_document_change(
-        context: &DocumentContext<RefCell<BalancedCaches>>,
+        context: &DocumentContext<RefCell<BalancedCaches<'_>>>,
         document_tokenizer: &DocumentTokenizer,
         searchable_attributes: Option<&[&str]>,
         document_change: DocumentChange,
@@ -147,8 +154,12 @@ impl WordPairProximityDocidsExtractor {
                 process_document_tokens(
                     document,
                     document_tokenizer,
-                    new_fields_ids_map,
                     &mut word_positions,
+                    &mut |field_name| {
+                        new_fields_ids_map
+                            .id_with_metadata_or_insert(field_name)
+                            .ok_or(UserError::AttributeLimitReached.into())
+                    },
                     &mut |(w1, w2), prox| {
                         del_word_pair_proximity.push(((w1, w2), prox));
                     },
@@ -170,8 +181,12 @@ impl WordPairProximityDocidsExtractor {
                 process_document_tokens(
                     document,
                     document_tokenizer,
-                    new_fields_ids_map,
                     &mut word_positions,
+                    &mut |field_name| {
+                        new_fields_ids_map
+                            .id_with_metadata_or_insert(field_name)
+                            .ok_or(UserError::AttributeLimitReached.into())
+                    },
                     &mut |(w1, w2), prox| {
                         del_word_pair_proximity.push(((w1, w2), prox));
                     },
@@ -180,8 +195,12 @@ impl WordPairProximityDocidsExtractor {
                 process_document_tokens(
                     document,
                     document_tokenizer,
-                    new_fields_ids_map,
                     &mut word_positions,
+                    &mut |field_name| {
+                        new_fields_ids_map
+                            .id_with_metadata_or_insert(field_name)
+                            .ok_or(UserError::AttributeLimitReached.into())
+                    },
                     &mut |(w1, w2), prox| {
                         add_word_pair_proximity.push(((w1, w2), prox));
                     },
@@ -192,8 +211,12 @@ impl WordPairProximityDocidsExtractor {
                 process_document_tokens(
                     document,
                     document_tokenizer,
-                    new_fields_ids_map,
                     &mut word_positions,
+                    &mut |field_name| {
+                        new_fields_ids_map
+                            .id_with_metadata_or_insert(field_name)
+                            .ok_or(UserError::AttributeLimitReached.into())
+                    },
                     &mut |(w1, w2), prox| {
                         add_word_pair_proximity.push(((w1, w2), prox));
                     },
@@ -257,8 +280,8 @@ fn drain_word_positions(
 fn process_document_tokens<'doc>(
     document: impl Document<'doc>,
     document_tokenizer: &DocumentTokenizer,
-    fields_ids_map: &mut GlobalFieldsIdsMap,
     word_positions: &mut VecDeque<(Rc<str>, u16)>,
+    field_id_and_metadata: &mut impl FnMut(&str) -> Result<(FieldId, Metadata)>,
     word_pair_proximity: &mut impl FnMut((Rc<str>, Rc<str>), u8),
 ) -> Result<()> {
     let mut field_id = None;
@@ -279,8 +302,248 @@ fn process_document_tokens<'doc>(
         word_positions.push_back((Rc::from(word), pos));
         Ok(())
     };
-    document_tokenizer.tokenize_document(document, fields_ids_map, &mut token_fn)?;
+
+    let mut should_tokenize = |field_name: &str| {
+        let (field_id, meta) = field_id_and_metadata(field_name)?;
+
+        let pattern_match = if meta.is_searchable() {
+            PatternMatch::Match
+        } else {
+            // TODO: should be a match on the field_name using `match_field_legacy` function,
+            //       but for legacy reasons we iterate over all the fields to fill the field_id_map.
+            PatternMatch::Parent
+        };
+
+        Ok((field_id, pattern_match))
+    };
+
+    document_tokenizer.tokenize_document(document, &mut should_tokenize, &mut token_fn)?;
 
     drain_word_positions(word_positions, word_pair_proximity);
     Ok(())
 }
+
+pub struct WordPairProximityDocidsSettingsExtractorsData<'a, SD> {
+    tokenizer: DocumentTokenizer<'a>,
+    max_memory_by_thread: Option<usize>,
+    buckets: usize,
+    settings_delta: &'a SD,
+}
+
+impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
+    for WordPairProximityDocidsSettingsExtractorsData<'_, SD>
+{
+    type Data = RefCell<BalancedCaches<'extractor>>;
+
+    fn init_data<'doc>(&'doc self, extractor_alloc: &'extractor Bump) -> crate::Result<Self::Data> {
+        Ok(RefCell::new(BalancedCaches::new_in(
+            self.buckets,
+            self.max_memory_by_thread,
+            extractor_alloc,
+        )))
+    }
+
+    fn process<'doc>(
+        &'doc self,
+        documents: impl Iterator<Item = crate::Result<DocumentIdentifiers<'doc>>>,
+        context: &'doc DocumentContext<Self::Data>,
+    ) -> crate::Result<()> {
+        for document in documents {
+            let document = document?;
+            SettingsChangeWordPairProximityDocidsExtractors::extract_document_from_settings_change(
+                document,
+                context,
+                &self.tokenizer,
+                self.settings_delta,
+            )?;
+        }
+        Ok(())
+    }
+}
+
+pub struct SettingsChangeWordPairProximityDocidsExtractors;
+
+impl SettingsChangeWordPairProximityDocidsExtractors {
+    pub fn run_extraction<'fid, 'indexer, 'index, 'extractor, SD, MSP>(
+        settings_delta: &SD,
+        documents: &'indexer DocumentsIndentifiers<'indexer>,
+        indexing_context: IndexingContext<'fid, 'indexer, 'index, MSP>,
+        extractor_allocs: &'extractor mut ThreadLocal<FullySend<Bump>>,
+        step: IndexingStep,
+    ) -> Result<Vec<BalancedCaches<'extractor>>>
+    where
+        SD: SettingsDelta + Sync,
+        MSP: Fn() -> bool + Sync,
+    {
+        // Warning: this is duplicated code from extract_word_docids.rs
+        let rtxn = indexing_context.index.read_txn()?;
+        let stop_words = indexing_context.index.stop_words(&rtxn)?;
+        let allowed_separators = indexing_context.index.allowed_separators(&rtxn)?;
+        let allowed_separators: Option<Vec<_>> =
+            allowed_separators.as_ref().map(|s| s.iter().map(String::as_str).collect());
+        let dictionary = indexing_context.index.dictionary(&rtxn)?;
+        let dictionary: Option<Vec<_>> =
+            dictionary.as_ref().map(|s| s.iter().map(String::as_str).collect());
+        let mut builder = tokenizer_builder(
+            stop_words.as_ref(),
+            allowed_separators.as_deref(),
+            dictionary.as_deref(),
+        );
+        let tokenizer = builder.build();
+        let localized_attributes_rules =
+            indexing_context.index.localized_attributes_rules(&rtxn)?.unwrap_or_default();
+        let document_tokenizer = DocumentTokenizer {
+            tokenizer: &tokenizer,
+            localized_attributes_rules: &localized_attributes_rules,
+            max_positions_per_attributes: MAX_POSITION_PER_ATTRIBUTE,
+        };
+        let extractor_data = WordPairProximityDocidsSettingsExtractorsData {
+            tokenizer: document_tokenizer,
+            max_memory_by_thread: indexing_context.grenad_parameters.max_memory_by_thread(),
+            buckets: rayon::current_num_threads(),
+            settings_delta,
+        };
+        let datastore = ThreadLocal::new();
+        {
+            let span = tracing::trace_span!(target: "indexing::documents::extract", "word_pair_proximity_docids_extraction");
+            let _entered = span.enter();
+
+            settings_change_extract(
+                documents,
+                &extractor_data,
+                indexing_context,
+                extractor_allocs,
+                &datastore,
+                step,
+            )?;
+        }
+
+        Ok(datastore.into_iter().map(RefCell::into_inner).collect())
+    }
+
+    /// Extracts document words from a settings change.
+    fn extract_document_from_settings_change<SD: SettingsDelta>(
+        document: DocumentIdentifiers<'_>,
+        context: &DocumentContext<RefCell<BalancedCaches<'_>>>,
+        document_tokenizer: &DocumentTokenizer,
+        settings_delta: &SD,
+    ) -> Result<()> {
+        let mut cached_sorter = context.data.borrow_mut_or_yield();
+        let doc_alloc = &context.doc_alloc;
+
+        let new_fields_ids_map = settings_delta.new_fields_ids_map();
+        let old_fields_ids_map = settings_delta.old_fields_ids_map();
+        let old_proximity_precision = *settings_delta.old_proximity_precision();
+        let new_proximity_precision = *settings_delta.new_proximity_precision();
+
+        let current_document = document.current(
+            &context.rtxn,
+            context.index,
+            old_fields_ids_map.as_fields_ids_map(),
+        )?;
+
+        #[derive(Debug, Clone, Copy, PartialEq)]
+        enum ActionToOperate {
+            ReindexAllFields,
+            SkipDocument,
+        }
+
+        // TODO prefix_fid delete_old_fid_based_databases
+        let mut action = match (old_proximity_precision, new_proximity_precision) {
+            (ByAttribute, ByWord) => ActionToOperate::ReindexAllFields,
+            (_, _) => ActionToOperate::SkipDocument,
+        };
+
+        // Here we do a preliminary check to determine the action to take.
+        // This check doesn't trigger the tokenizer as we never return
+        // PatternMatch::Match.
+        if action != ActionToOperate::ReindexAllFields {
+            document_tokenizer.tokenize_document(
+                current_document,
+                &mut |field_name| {
+                    let fid = new_fields_ids_map.id(field_name).expect("All fields IDs must exist");
+
+                    // If the document must be reindexed, early return NoMatch to stop the scanning process.
+                    if action == ActionToOperate::ReindexAllFields {
+                        return Ok((fid, PatternMatch::NoMatch));
+                    }
+
+                    let old_field_metadata = old_fields_ids_map.metadata(fid).unwrap();
+                    let new_field_metadata = new_fields_ids_map.metadata(fid).unwrap();
+
+                    action = match (old_field_metadata, new_field_metadata) {
+                        // At least one field is removed or added from the searchable fields
+                        (
+                            Metadata { searchable: Some(_), .. },
+                            Metadata { searchable: None, .. },
+                        )
+                        | (
+                            Metadata { searchable: None, .. },
+                            Metadata { searchable: Some(_), .. },
+                        ) => ActionToOperate::ReindexAllFields,
+                        _ => action,
+                    };
+
+                    Ok((fid, PatternMatch::Parent))
+                },
+                &mut |_, _, _, _| Ok(()),
+            )?;
+        }
+
+        // Early return when we don't need to index the document
+        if action == ActionToOperate::SkipDocument {
+            return Ok(());
+        }
+
+        let mut del_word_pair_proximity = bumpalo::collections::Vec::new_in(doc_alloc);
+        let mut add_word_pair_proximity = bumpalo::collections::Vec::new_in(doc_alloc);
+
+        // is a vecdequeue, and will be smol, so can stay on the heap for now
+        let mut word_positions: VecDeque<(Rc<str>, u16)> =
+            VecDeque::with_capacity(MAX_DISTANCE as usize);
+
+        process_document_tokens(
+            current_document,
+            // TODO Tokenize must be based on old settings
+            document_tokenizer,
+            &mut word_positions,
+            &mut |field_name| {
+                Ok(old_fields_ids_map.id_with_metadata(field_name).expect("All fields must exist"))
+            },
+            &mut |(w1, w2), prox| {
+                del_word_pair_proximity.push(((w1, w2), prox));
+            },
+        )?;
+
+        process_document_tokens(
+            current_document,
+            // TODO Tokenize must be based on new settings
+            document_tokenizer,
+            &mut word_positions,
+            &mut |field_name| {
+                Ok(new_fields_ids_map.id_with_metadata(field_name).expect("All fields must exist"))
+            },
+            &mut |(w1, w2), prox| {
+                add_word_pair_proximity.push(((w1, w2), prox));
+            },
+        )?;
+
+        let mut key_buffer = bumpalo::collections::Vec::new_in(doc_alloc);
+
+        del_word_pair_proximity.sort_unstable();
+        del_word_pair_proximity.dedup_by(|(k1, _), (k2, _)| k1 == k2);
+        for ((w1, w2), prox) in del_word_pair_proximity.iter() {
+            let key = build_key(*prox, w1, w2, &mut key_buffer);
+            cached_sorter.insert_del_u32(key, document.docid())?;
+        }
+
+        add_word_pair_proximity.sort_unstable();
+        add_word_pair_proximity.dedup_by(|(k1, _), (k2, _)| k1 == k2);
+        for ((w1, w2), prox) in add_word_pair_proximity.iter() {
+            let key = build_key(*prox, w1, w2, &mut key_buffer);
+            cached_sorter.insert_add_u32(key, document.docid())?;
+        }
+
+        Ok(())
+    }
+}

crates/milli/src/update/new/extract/searchable/mod.rs

@@ -2,8 +2,12 @@ mod extract_word_docids;
 mod extract_word_pair_proximity_docids;
 mod tokenize_document;
 
-pub use extract_word_docids::{WordDocidsCaches, WordDocidsExtractors};
+pub use extract_word_docids::{
-pub use extract_word_pair_proximity_docids::WordPairProximityDocidsExtractor;
+    SettingsChangeWordDocidsExtractors, WordDocidsCaches, WordDocidsExtractors,
+};
+pub use extract_word_pair_proximity_docids::{
+    SettingsChangeWordPairProximityDocidsExtractors, WordPairProximityDocidsExtractor,
+};
 
 use crate::attribute_patterns::{match_field_legacy, PatternMatch};
 
@@ -27,3 +31,17 @@ pub fn match_searchable_field(
 
     selection
 }
+
+/// return `true` if the provided `field_name` is a parent of at least one of the fields contained in `searchable`,
+/// or if `searchable` is `None`.
+fn has_searchable_children<I, A>(field_name: &str, searchable: Option<I>) -> bool
+where
+    I: IntoIterator<Item = A>,
+    A: AsRef<str>,
+{
+    searchable.is_none_or(|fields| {
+        fields
+            .into_iter()
+            .any(|attr| match_field_legacy(attr.as_ref(), field_name) == PatternMatch::Parent)
+    })
+}

crates/milli/src/update/new/extract/searchable/tokenize_document.rs

@@ -8,10 +8,7 @@ use crate::update::new::document::Document;
 use crate::update::new::extract::perm_json_p::{
     seek_leaf_values_in_array, seek_leaf_values_in_object, Depth,
 };
-use crate::{
+use crate::{FieldId, InternalError, LocalizedAttributesRule, Result, MAX_WORD_LENGTH};
-    FieldId, GlobalFieldsIdsMap, InternalError, LocalizedAttributesRule, Result, UserError,
-    MAX_WORD_LENGTH,
-};
 
 // todo: should be crate::proximity::MAX_DISTANCE but it has been forgotten
 const MAX_DISTANCE: u32 = 8;
@@ -26,26 +23,25 @@ impl DocumentTokenizer<'_> {
     pub fn tokenize_document<'doc>(
         &self,
         document: impl Document<'doc>,
-        field_id_map: &mut GlobalFieldsIdsMap,
+        should_tokenize: &mut impl FnMut(&str) -> Result<(FieldId, PatternMatch)>,
         token_fn: &mut impl FnMut(&str, FieldId, u16, &str) -> Result<()>,
     ) -> Result<()> {
         let mut field_position = HashMap::new();
-        let mut tokenize_field = |field_name: &str, _depth, value: &Value| {
-            let Some((field_id, meta)) = field_id_map.id_with_metadata_or_insert(field_name) else {
-                return Err(UserError::AttributeLimitReached.into());
-            };
-
-            if meta.is_searchable() {
-                self.tokenize_field(field_id, field_name, value, token_fn, &mut field_position)?;
-            }
-
-            // todo: should be a match on the field_name using `match_field_legacy` function,
-            // but for legacy reasons we iterate over all the fields to fill the field_id_map.
-            Ok(PatternMatch::Match)
-        };
-
         for entry in document.iter_top_level_fields() {
             let (field_name, value) = entry?;
+
+            if let (_, PatternMatch::NoMatch) = should_tokenize(field_name)? {
+                continue;
+            }
+
+            let mut tokenize_field = |field_name: &str, _depth, value: &Value| {
+                let (fid, pattern_match) = should_tokenize(field_name)?;
+                if pattern_match == PatternMatch::Match {
+                    self.tokenize_field(fid, field_name, value, token_fn, &mut field_position)?;
+                }
+                Ok(pattern_match)
+            };
+
             // parse json.
             match serde_json::to_value(value).map_err(InternalError::SerdeJson)? {
                 Value::Object(object) => seek_leaf_values_in_object(
@@ -192,7 +188,7 @@ mod test {
     use super::*;
     use crate::fields_ids_map::metadata::{FieldIdMapWithMetadata, MetadataBuilder};
     use crate::update::new::document::{DocumentFromVersions, Versions};
-    use crate::FieldsIdsMap;
+    use crate::{FieldsIdsMap, GlobalFieldsIdsMap, UserError};
 
     #[test]
     fn test_tokenize_document() {
@@ -231,6 +227,7 @@ mod test {
                 Default::default(),
                 Default::default(),
                 Default::default(),
+                Default::default(),
                 None,
                 None,
                 Default::default(),
@@ -251,15 +248,19 @@ mod test {
         let document = Versions::single(document);
         let document = DocumentFromVersions::new(&document);
 
+        let mut should_tokenize = |field_name: &str| {
+            let Some(field_id) = global_fields_ids_map.id_or_insert(field_name) else {
+                return Err(UserError::AttributeLimitReached.into());
+            };
+
+            Ok((field_id, PatternMatch::Match))
+        };
+
         document_tokenizer
-            .tokenize_document(
+            .tokenize_document(document, &mut should_tokenize, &mut |_fname, fid, pos, word| {
-                document,
+                words.insert([fid, pos], word.to_string());
-                &mut global_fields_ids_map,
+                Ok(())
-                &mut |_fname, fid, pos, word| {
+            })
-                    words.insert([fid, pos], word.to_string());
-                    Ok(())
-                },
-            )
             .unwrap();
 
         snapshot!(format!("{:#?}", words), @r###"

crates/milli/src/update/new/extract/vectors/mod.rs

@@ -1,5 +1,6 @@
 use std::cell::RefCell;
 use std::fmt::Debug;
+use std::sync::RwLock;
 
 use bumpalo::collections::Vec as BVec;
 use bumpalo::Bump;
@@ -27,7 +28,10 @@ use crate::vector::extractor::{
 use crate::vector::session::{EmbedSession, Input, Metadata, OnEmbed};
 use crate::vector::settings::ReindexAction;
 use crate::vector::{Embedding, RuntimeEmbedder, RuntimeEmbedders, RuntimeFragment};
-use crate::{DocumentId, FieldDistribution, InternalError, Result, ThreadPoolNoAbort, UserError};
+use crate::{
+    DocumentId, FieldDistribution, GlobalFieldsIdsMap, InternalError, Result, ThreadPoolNoAbort,
+    UserError,
+};
 
 pub struct EmbeddingExtractor<'a, 'b> {
     embedders: &'a RuntimeEmbedders,
@@ -321,6 +325,15 @@ impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
         let old_embedders = self.settings_delta.old_embedders();
         let unused_vectors_distribution = UnusedVectorsDistributionBump::new_in(&context.doc_alloc);
 
+        // We get a reference to the new and old fields ids maps but
+        // note that those are local versions where updates to them
+        // will not be reflected in the database. It's not an issue
+        // because new settings do not generate new fields.
+        let new_fields_ids_map = RwLock::new(self.settings_delta.new_fields_ids_map().clone());
+        let new_fields_ids_map = RefCell::new(GlobalFieldsIdsMap::new(&new_fields_ids_map));
+        let old_fields_ids_map = RwLock::new(self.settings_delta.old_fields_ids_map().clone());
+        let old_fields_ids_map = RefCell::new(GlobalFieldsIdsMap::new(&old_fields_ids_map));
+
         let mut all_chunks = BVec::with_capacity_in(embedders.len(), &context.doc_alloc);
         let embedder_configs = context.index.embedding_configs();
         for (embedder_name, action) in self.settings_delta.embedder_actions().iter() {
@@ -396,6 +409,7 @@ impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
                         if !must_regenerate {
                             continue;
                         }
+
                         // we need to regenerate the prompts for the document
                         chunks.settings_change_autogenerated(
                             document.docid(),
@@ -406,7 +420,8 @@ impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
                                 context.db_fields_ids_map,
                             )?,
                             self.settings_delta,
-                            context.new_fields_ids_map,
+                            &old_fields_ids_map,
+                            &new_fields_ids_map,
                             &unused_vectors_distribution,
                             old_is_user_provided,
                             fragments_changed,
@@ -442,7 +457,8 @@ impl<'extractor, SD: SettingsDelta + Sync> SettingsChangeExtractor<'extractor>
                                     context.db_fields_ids_map,
                                 )?,
                                 self.settings_delta,
-                                context.new_fields_ids_map,
+                                &old_fields_ids_map,
+                                &new_fields_ids_map,
                                 &unused_vectors_distribution,
                                 old_is_user_provided,
                                 true,
@@ -638,7 +654,8 @@ impl<'a, 'b, 'extractor> Chunks<'a, 'b, 'extractor> {
         external_docid: &'a str,
         document: D,
         settings_delta: &SD,
-        fields_ids_map: &'a RefCell<crate::GlobalFieldsIdsMap>,
+        old_fields_ids_map: &'a RefCell<GlobalFieldsIdsMap<'a>>,
+        new_fields_ids_map: &'a RefCell<GlobalFieldsIdsMap<'a>>,
         unused_vectors_distribution: &UnusedVectorsDistributionBump<'a>,
         old_is_user_provided: bool,
         full_reindex: bool,
@@ -733,10 +750,17 @@ impl<'a, 'b, 'extractor> Chunks<'a, 'b, 'extractor> {
                     old_embedder.as_ref().map(|old_embedder| &old_embedder.document_template)
                 };
 
-                let extractor =
+                let extractor = DocumentTemplateExtractor::new(
-                    DocumentTemplateExtractor::new(document_template, doc_alloc, fields_ids_map);
+                    document_template,
+                    doc_alloc,
+                    new_fields_ids_map,
+                );
                 let old_extractor = old_document_template.map(|old_document_template| {
-                    DocumentTemplateExtractor::new(old_document_template, doc_alloc, fields_ids_map)
+                    DocumentTemplateExtractor::new(
+                        old_document_template,
+                        doc_alloc,
+                        old_fields_ids_map,
+                    )
                 });
                 let metadata =
                     Metadata { docid, external_docid, extractor_id: extractor.extractor_id() };

crates/milli/src/update/new/indexer/community_edition.rs

@@ -0,0 +1,9 @@
+pub mod sharding {
+    pub struct Shards;
+
+    impl Shards {
+        pub fn must_process(&self, _docid: &str) -> bool {
+            true
+        }
+    }
+}

crates/milli/src/update/new/indexer/document_operation.rs

@@ -17,7 +17,7 @@ use super::guess_primary_key::retrieve_or_guess_primary_key;
 use crate::documents::PrimaryKey;
 use crate::progress::{AtomicPayloadStep, Progress};
 use crate::update::new::document::{DocumentContext, Versions};
-use crate::update::new::indexer::enterprise_edition::sharding::Shards;
+use crate::update::new::indexer::current_edition::sharding::Shards;
 use crate::update::new::steps::IndexingStep;
 use crate::update::new::thread_local::MostlySend;
 use crate::update::new::{DocumentIdentifiers, Insertion, Update};

crates/milli/src/update/new/indexer/extract.rs

@@ -372,11 +372,10 @@ where
     SD: SettingsDelta + Sync,
 {
     // Create the list of document ids to extract
-    let rtxn = indexing_context.index.read_txn()?;
+    let index = indexing_context.index;
-    let all_document_ids =
+    let rtxn = index.read_txn()?;
-        indexing_context.index.documents_ids(&rtxn)?.into_iter().collect::<Vec<_>>();
+    let all_document_ids = index.documents_ids(&rtxn)?.into_iter().collect::<Vec<_>>();
-    let primary_key =
+    let primary_key = primary_key_from_db(index, &rtxn, &indexing_context.db_fields_ids_map)?;
-        primary_key_from_db(indexing_context.index, &rtxn, &indexing_context.db_fields_ids_map)?;
     let documents = DocumentsIndentifiers::new(&all_document_ids, primary_key);
 
     let span =
@@ -391,6 +390,133 @@ where
         extractor_allocs,
     )?;
 
+    {
+        let WordDocidsCaches {
+            word_docids,
+            word_fid_docids,
+            exact_word_docids,
+            word_position_docids,
+            fid_word_count_docids,
+        } = {
+            let span = tracing::trace_span!(target: "indexing::documents::extract", "word_docids");
+            let _entered = span.enter();
+            SettingsChangeWordDocidsExtractors::run_extraction(
+                settings_delta,
+                &documents,
+                indexing_context,
+                extractor_allocs,
+                IndexingStep::ExtractingWords,
+            )?
+        };
+
+        indexing_context.progress.update_progress(IndexingStep::MergingWordCaches);
+
+        {
+            let span = tracing::trace_span!(target: "indexing::documents::merge", "word_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(MergingWordCache::WordDocids);
+
+            merge_and_send_docids(
+                word_docids,
+                index.word_docids.remap_types(),
+                index,
+                extractor_sender.docids::<WordDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+
+        {
+            let span =
+                tracing::trace_span!(target: "indexing::documents::merge", "word_fid_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(MergingWordCache::WordFieldIdDocids);
+
+            merge_and_send_docids(
+                word_fid_docids,
+                index.word_fid_docids.remap_types(),
+                index,
+                extractor_sender.docids::<WordFidDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+
+        {
+            let span =
+                tracing::trace_span!(target: "indexing::documents::merge", "exact_word_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(MergingWordCache::ExactWordDocids);
+
+            merge_and_send_docids(
+                exact_word_docids,
+                index.exact_word_docids.remap_types(),
+                index,
+                extractor_sender.docids::<ExactWordDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+
+        {
+            let span =
+                tracing::trace_span!(target: "indexing::documents::merge", "word_position_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(MergingWordCache::WordPositionDocids);
+
+            merge_and_send_docids(
+                word_position_docids,
+                index.word_position_docids.remap_types(),
+                index,
+                extractor_sender.docids::<WordPositionDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+
+        {
+            let span =
+                tracing::trace_span!(target: "indexing::documents::merge", "fid_word_count_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(MergingWordCache::FieldIdWordCountDocids);
+
+            merge_and_send_docids(
+                fid_word_count_docids,
+                index.field_id_word_count_docids.remap_types(),
+                index,
+                extractor_sender.docids::<FidWordCountDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+    }
+
+    // Run the proximity extraction only if the precision is ByWord.
+    let new_proximity_precision = settings_delta.new_proximity_precision();
+    if *new_proximity_precision == ProximityPrecision::ByWord {
+        let caches = {
+            let span = tracing::trace_span!(target: "indexing::documents::extract", "word_pair_proximity_docids");
+            let _entered = span.enter();
+
+            SettingsChangeWordPairProximityDocidsExtractors::run_extraction(
+                settings_delta,
+                &documents,
+                indexing_context,
+                extractor_allocs,
+                IndexingStep::ExtractingWordProximity,
+            )?
+        };
+
+        {
+            let span = tracing::trace_span!(target: "indexing::documents::merge", "word_pair_proximity_docids");
+            let _entered = span.enter();
+            indexing_context.progress.update_progress(IndexingStep::MergingWordProximity);
+
+            merge_and_send_docids(
+                caches,
+                index.word_pair_proximity_docids.remap_types(),
+                index,
+                extractor_sender.docids::<WordPairProximityDocids>(),
+                &indexing_context.must_stop_processing,
+            )?;
+        }
+    }
+
     'vectors: {
         if settings_delta.embedder_actions().is_empty() {
             break 'vectors;

crates/milli/src/update/new/indexer/mod.rs

@@ -1,4 +1,4 @@
-use std::collections::BTreeMap;
+use std::collections::{BTreeMap, BTreeSet};
 use std::sync::atomic::AtomicBool;
 use std::sync::{Arc, Once, RwLock};
 use std::thread::{self, Builder};
@@ -8,9 +8,11 @@ use document_changes::{DocumentChanges, IndexingContext};
 pub use document_deletion::DocumentDeletion;
 pub use document_operation::{DocumentOperation, PayloadStats};
 use hashbrown::HashMap;
-use heed::{RoTxn, RwTxn};
+use heed::types::DecodeIgnore;
+use heed::{BytesDecode, Database, RoTxn, RwTxn};
 pub use partial_dump::PartialDump;
 pub use post_processing::recompute_word_fst_from_word_docids_database;
+pub use settings_changes::settings_change_extract;
 pub use update_by_function::UpdateByFunction;
 pub use write::ChannelCongestion;
 use write::{build_vectors, update_index, write_to_db};
@@ -20,18 +22,31 @@ use super::steps::IndexingStep;
 use super::thread_local::ThreadLocal;
 use crate::documents::PrimaryKey;
 use crate::fields_ids_map::metadata::{FieldIdMapWithMetadata, MetadataBuilder};
+use crate::heed_codec::StrBEU16Codec;
 use crate::progress::{EmbedderStats, Progress};
+use crate::proximity::ProximityPrecision;
+use crate::update::new::steps::SettingsIndexerStep;
+use crate::update::new::FacetFieldIdsDelta;
 use crate::update::settings::SettingsDelta;
 use crate::update::GrenadParameters;
 use crate::vector::settings::{EmbedderAction, RemoveFragments, WriteBackToDocuments};
 use crate::vector::{Embedder, RuntimeEmbedders, VectorStore};
-use crate::{FieldsIdsMap, GlobalFieldsIdsMap, Index, InternalError, Result, ThreadPoolNoAbort};
+use crate::{
+    Error, FieldsIdsMap, GlobalFieldsIdsMap, Index, InternalError, Result, ThreadPoolNoAbort,
+};
 
+#[cfg(not(feature = "enterprise"))]
+pub mod community_edition;
 pub(crate) mod de;
 pub mod document_changes;
 mod document_deletion;
 mod document_operation;
+#[cfg(feature = "enterprise")]
 pub mod enterprise_edition;
+#[cfg(not(feature = "enterprise"))]
+pub use community_edition as current_edition;
+#[cfg(feature = "enterprise")]
+pub use enterprise_edition as current_edition;
 mod extract;
 mod guess_primary_key;
 mod partial_dump;
@@ -235,6 +250,20 @@ where
     SD: SettingsDelta + Sync,
 {
     delete_old_embedders_and_fragments(wtxn, index, settings_delta)?;
+    delete_old_fid_based_databases(wtxn, index, settings_delta, must_stop_processing, progress)?;
+
+    // Clear word_pair_proximity if byWord to byAttribute
+    let old_proximity_precision = settings_delta.old_proximity_precision();
+    let new_proximity_precision = settings_delta.new_proximity_precision();
+    if *old_proximity_precision == ProximityPrecision::ByWord
+        && *new_proximity_precision == ProximityPrecision::ByAttribute
+    {
+        index.word_pair_proximity_docids.clear(wtxn)?;
+    }
+
+    // TODO delete useless searchable databases
+    //      - Clear fid_prefix_* in the post processing
+    //      - clear the prefix + fid_prefix if setting `PrefixSearch` is enabled
 
     let mut bbbuffers = Vec::new();
     let finished_extraction = AtomicBool::new(false);
@@ -293,6 +322,8 @@ where
                 .unwrap()
             })?;
 
+        let global_fields_ids_map = GlobalFieldsIdsMap::new(&new_fields_ids_map);
+
         let new_embedders = settings_delta.new_embedders();
         let embedder_actions = settings_delta.embedder_actions();
         let index_embedder_category_ids = settings_delta.new_embedder_category_id();
@@ -327,6 +358,18 @@ where
         })
         .unwrap()?;
 
+        pool.install(|| {
+            // WARN When implementing the facets don't forget this
+            let facet_field_ids_delta = FacetFieldIdsDelta::new(0, 0);
+            post_processing::post_process(
+                indexing_context,
+                wtxn,
+                global_fields_ids_map,
+                facet_field_ids_delta,
+            )
+        })
+        .unwrap()?;
+
         indexing_context.progress.update_progress(IndexingStep::BuildingGeoJson);
         index.cellulite.build(
             wtxn,
@@ -456,6 +499,106 @@ where
     Ok(())
 }
 
+/// Deletes entries refering the provided
+/// fids from the fid-based databases.
+fn delete_old_fid_based_databases<SD, MSP>(
+    wtxn: &mut RwTxn<'_>,
+    index: &Index,
+    settings_delta: &SD,
+    must_stop_processing: &MSP,
+    progress: &Progress,
+) -> Result<()>
+where
+    SD: SettingsDelta + Sync,
+    MSP: Fn() -> bool + Sync,
+{
+    let fids_to_delete: Option<BTreeSet<_>> = {
+        let rtxn = index.read_txn()?;
+        let fields_ids_map = index.fields_ids_map(&rtxn)?;
+        let old_searchable_attributes = settings_delta.old_searchable_attributes().as_ref();
+        let new_searchable_attributes = settings_delta.new_searchable_attributes().as_ref();
+        old_searchable_attributes.zip(new_searchable_attributes).map(|(old, new)| {
+            old.iter()
+                // Ignore the field if it is not searchable anymore
+                // or if it was never referenced in any document
+                .filter_map(|name| if new.contains(name) { None } else { fields_ids_map.id(name) })
+                .collect()
+        })
+    };
+
+    let Some(fids_to_delete) = fids_to_delete else {
+        return Ok(());
+    };
+
+    progress.update_progress(SettingsIndexerStep::DeletingOldWordFidDocids);
+    delete_old_word_fid_docids(wtxn, index.word_fid_docids, must_stop_processing, &fids_to_delete)?;
+
+    progress.update_progress(SettingsIndexerStep::DeletingOldFidWordCountDocids);
+    delete_old_fid_word_count_docids(wtxn, index, must_stop_processing, &fids_to_delete)?;
+
+    progress.update_progress(SettingsIndexerStep::DeletingOldWordPrefixFidDocids);
+    delete_old_word_fid_docids(
+        wtxn,
+        index.word_prefix_fid_docids,
+        must_stop_processing,
+        &fids_to_delete,
+    )?;
+
+    Ok(())
+}
+
+fn delete_old_word_fid_docids<'txn, MSP, DC>(
+    wtxn: &mut RwTxn<'txn>,
+    database: Database<StrBEU16Codec, DC>,
+    must_stop_processing: &MSP,
+    fids_to_delete: &BTreeSet<u16>,
+) -> Result<(), Error>
+where
+    MSP: Fn() -> bool + Sync,
+    DC: BytesDecode<'txn>,
+{
+    let mut iter = database.iter_mut(wtxn)?.remap_data_type::<DecodeIgnore>();
+    while let Some(((_word, fid), ())) = iter.next().transpose()? {
+        // TODO should I call it that often?
+        if must_stop_processing() {
+            return Err(Error::InternalError(InternalError::AbortedIndexation));
+        }
+
+        if fids_to_delete.contains(&fid) {
+            // safety: We don't keep any references to the data.
+            unsafe { iter.del_current()? };
+        }
+    }
+
+    Ok(())
+}
+
+fn delete_old_fid_word_count_docids<MSP>(
+    wtxn: &mut RwTxn<'_>,
+    index: &Index,
+    must_stop_processing: &MSP,
+    fids_to_delete: &BTreeSet<u16>,
+) -> Result<(), Error>
+where
+    MSP: Fn() -> bool + Sync,
+{
+    let db = index.field_id_word_count_docids.remap_data_type::<DecodeIgnore>();
+    for &fid_to_delete in fids_to_delete {
+        if must_stop_processing() {
+            return Err(Error::InternalError(InternalError::AbortedIndexation));
+        }
+
+        let mut iter = db.prefix_iter_mut(wtxn, &(fid_to_delete, 0))?;
+        while let Some(((fid, _word_count), ())) = iter.next().transpose()? {
+            debug_assert_eq!(fid, fid_to_delete);
+            // safety: We don't keep any references to the data.
+            unsafe { iter.del_current()? };
+        }
+    }
+
+    Ok(())
+}
+
 fn indexer_memory_settings(
     current_num_threads: usize,
     grenad_parameters: GrenadParameters,

crates/milli/src/update/new/steps.rs

@@ -28,6 +28,9 @@ make_enum_progress! {
         ChangingVectorStore,
         UsingStableIndexer,
         UsingExperimentalIndexer,
+        DeletingOldWordFidDocids,
+        DeletingOldFidWordCountDocids,
+        DeletingOldWordPrefixFidDocids,
     }
 }
 

crates/milli/src/update/settings.rs

@@ -1589,33 +1589,33 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
 
         // only use the new indexer when only the embedder possibly changed
         if let Self {
-            searchable_fields: Setting::NotSet,
+            searchable_fields: _,
             displayed_fields: Setting::NotSet,
             filterable_fields: Setting::NotSet,
             sortable_fields: Setting::NotSet,
             criteria: Setting::NotSet,
-            stop_words: Setting::NotSet,
+            stop_words: Setting::NotSet, // TODO (require force reindexing of searchables)
-            non_separator_tokens: Setting::NotSet,
+            non_separator_tokens: Setting::NotSet, // TODO (require force reindexing of searchables)
-            separator_tokens: Setting::NotSet,
+            separator_tokens: Setting::NotSet, // TODO (require force reindexing of searchables)
-            dictionary: Setting::NotSet,
+            dictionary: Setting::NotSet, // TODO (require force reindexing of searchables)
             distinct_field: Setting::NotSet,
             synonyms: Setting::NotSet,
             primary_key: Setting::NotSet,
             authorize_typos: Setting::NotSet,
             min_word_len_two_typos: Setting::NotSet,
             min_word_len_one_typo: Setting::NotSet,
-            exact_words: Setting::NotSet,
+            exact_words: Setting::NotSet, // TODO (require force reindexing of searchables)
-            exact_attributes: Setting::NotSet,
+            exact_attributes: _,
             max_values_per_facet: Setting::NotSet,
             sort_facet_values_by: Setting::NotSet,
             pagination_max_total_hits: Setting::NotSet,
-            proximity_precision: Setting::NotSet,
+            proximity_precision: _,
             embedder_settings: _,
             search_cutoff: Setting::NotSet,
-            localized_attributes_rules: Setting::NotSet,
+            localized_attributes_rules: Setting::NotSet, // TODO to start with
-            prefix_search: Setting::NotSet,
+            prefix_search: Setting::NotSet,              // TODO continue with this
             facet_search: Setting::NotSet,
-            disable_on_numbers: Setting::NotSet,
+            disable_on_numbers: Setting::NotSet, // TODO (require force reindexing of searchables)
             chat: Setting::NotSet,
             vector_store: Setting::NotSet,
             wtxn: _,
@@ -1632,10 +1632,12 @@ impl<'a, 't, 'i> Settings<'a, 't, 'i> {
             // Update index settings
             let embedding_config_updates = self.update_embedding_configs()?;
             self.update_user_defined_searchable_attributes()?;
+            self.update_exact_attributes()?;
+            self.update_proximity_precision()?;
 
-            let mut new_inner_settings =
+            // Note that we don't need to update the searchables here,
-                InnerIndexSettings::from_index(self.index, self.wtxn, None)?;
+            // as it will be done after the settings update.
-            new_inner_settings.recompute_searchables(self.wtxn, self.index)?;
+            let new_inner_settings = InnerIndexSettings::from_index(self.index, self.wtxn, None)?;
 
             let primary_key_id = self
                 .index
@@ -2062,9 +2064,12 @@ impl InnerIndexSettings {
         let sortable_fields = index.sortable_fields(rtxn)?;
         let asc_desc_fields = index.asc_desc_fields(rtxn)?;
         let distinct_field = index.distinct_field(rtxn)?.map(|f| f.to_string());
-        let user_defined_searchable_attributes = index
+        let user_defined_searchable_attributes = match index.user_defined_searchable_fields(rtxn)? {
-            .user_defined_searchable_fields(rtxn)?
+            Some(fields) if fields.contains(&"*") => None,
-            .map(|fields| fields.into_iter().map(|f| f.to_string()).collect());
+            Some(fields) => Some(fields.into_iter().map(|f| f.to_string()).collect()),
+            None => None,
+        };
+
         let builder = MetadataBuilder::from_index(index, rtxn)?;
         let fields_ids_map = FieldIdMapWithMetadata::new(fields_ids_map, builder);
         let disabled_typos_terms = index.disabled_typos_terms(rtxn)?;
@@ -2578,8 +2583,20 @@ fn deserialize_sub_embedder(
 /// Implement this trait for the settings delta type.
 /// This is used in the new settings update flow and will allow to easily replace the old settings delta type: `InnerIndexSettingsDiff`.
 pub trait SettingsDelta {
-    fn new_embedders(&self) -> &RuntimeEmbedders;
+    fn old_fields_ids_map(&self) -> &FieldIdMapWithMetadata;
+    fn new_fields_ids_map(&self) -> &FieldIdMapWithMetadata;
+
+    fn old_searchable_attributes(&self) -> &Option<Vec<String>>;
+    fn new_searchable_attributes(&self) -> &Option<Vec<String>>;
+
+    fn old_disabled_typos_terms(&self) -> &DisabledTyposTerms;
+    fn new_disabled_typos_terms(&self) -> &DisabledTyposTerms;
+
+    fn old_proximity_precision(&self) -> &ProximityPrecision;
+    fn new_proximity_precision(&self) -> &ProximityPrecision;
+
     fn old_embedders(&self) -> &RuntimeEmbedders;
+    fn new_embedders(&self) -> &RuntimeEmbedders;
     fn new_embedder_category_id(&self) -> &HashMap<String, u8>;
     fn embedder_actions(&self) -> &BTreeMap<String, EmbedderAction>;
     fn try_for_each_fragment_diff<F, E>(
@@ -2589,7 +2606,6 @@ pub trait SettingsDelta {
     ) -> std::result::Result<(), E>
     where
         F: FnMut(FragmentDiff) -> std::result::Result<(), E>;
-    fn new_fields_ids_map(&self) -> &FieldIdMapWithMetadata;
 }
 
 pub struct FragmentDiff<'a> {
@@ -2598,26 +2614,47 @@ pub struct FragmentDiff<'a> {
 }
 
 impl SettingsDelta for InnerIndexSettingsDiff {
-    fn new_embedders(&self) -> &RuntimeEmbedders {
+    fn old_fields_ids_map(&self) -> &FieldIdMapWithMetadata {
-        &self.new.runtime_embedders
+        &self.old.fields_ids_map
+    }
+    fn new_fields_ids_map(&self) -> &FieldIdMapWithMetadata {
+        &self.new.fields_ids_map
+    }
+
+    fn old_searchable_attributes(&self) -> &Option<Vec<String>> {
+        &self.old.user_defined_searchable_attributes
+    }
+    fn new_searchable_attributes(&self) -> &Option<Vec<String>> {
+        &self.new.user_defined_searchable_attributes
+    }
+
+    fn old_disabled_typos_terms(&self) -> &DisabledTyposTerms {
+        &self.old.disabled_typos_terms
+    }
+    fn new_disabled_typos_terms(&self) -> &DisabledTyposTerms {
+        &self.new.disabled_typos_terms
+    }
+
+    fn old_proximity_precision(&self) -> &ProximityPrecision {
+        &self.old.proximity_precision
+    }
+    fn new_proximity_precision(&self) -> &ProximityPrecision {
+        &self.new.proximity_precision
     }
 
     fn old_embedders(&self) -> &RuntimeEmbedders {
         &self.old.runtime_embedders
     }
+    fn new_embedders(&self) -> &RuntimeEmbedders {
+        &self.new.runtime_embedders
+    }
 
     fn new_embedder_category_id(&self) -> &HashMap<String, u8> {
         &self.new.embedder_category_id
     }
-
     fn embedder_actions(&self) -> &BTreeMap<String, EmbedderAction> {
         &self.embedding_config_updates
     }
-
-    fn new_fields_ids_map(&self) -> &FieldIdMapWithMetadata {
-        &self.new.fields_ids_map
-    }
-
     fn try_for_each_fragment_diff<F, E>(
         &self,
         embedder_name: &str,

crates/milli/src/update/test_settings.rs

@@ -14,28 +14,21 @@ fn set_and_reset_searchable_fields() {
     let index = TempIndex::new();
 
     // First we send 3 documents with ids from 1 to 3.
-    let mut wtxn = index.write_txn().unwrap();
-
     index
-        .add_documents_using_wtxn(
+        .add_documents(documents!([
-            &mut wtxn,
+            { "id": 1, "name": "kevin", "age": 23 },
-            documents!([
+            { "id": 2, "name": "kevina", "age": 21},
-                { "id": 1, "name": "kevin", "age": 23 },
+            { "id": 3, "name": "benoit", "age": 34 }
-                { "id": 2, "name": "kevina", "age": 21},
+        ]))
-                { "id": 3, "name": "benoit", "age": 34 }
-            ]),
-        )
         .unwrap();
 
     // We change the searchable fields to be the "name" field only.
     index
-        .update_settings_using_wtxn(&mut wtxn, |settings| {
+        .update_settings(|settings| {
             settings.set_searchable_fields(vec!["name".into()]);
         })
         .unwrap();
 
-    wtxn.commit().unwrap();
-
     db_snap!(index, fields_ids_map, @r###"
     0   id               |
     1   name             |

crates/milli/src/update/upgrade/mod.rs

@@ -45,6 +45,7 @@ const UPGRADE_FUNCTIONS: &[&dyn UpgradeIndex] = &[
     &ToTargetNoOp { target: (1, 25, 0) },
     &ToTargetNoOp { target: (1, 26, 0) },
     &ToTargetNoOp { target: (1, 27, 0) },
+    &ToTargetNoOp { target: (1, 28, 0) },
     // This is the last upgrade function, it will be called when the index is up to date.
     // any other upgrade function should be added before this one.
     &ToCurrentNoOp {},
@@ -83,6 +84,7 @@ const fn start(from: (u32, u32, u32)) -> Option<usize> {
         (1, 25, _) => function_index!(15),
         (1, 26, _) => function_index!(16),
         (1, 27, _) => function_index!(17),
+        (1, 28, _) => function_index!(18),
         // We deliberately don't add a placeholder with (VERSION_MAJOR, VERSION_MINOR, VERSION_PATCH) here to force manually
         // considering dumpless upgrade.
         (_major, _minor, _patch) => return None,

crates/openapi-generator/Cargo.toml

@@ -7,6 +7,6 @@ publish = false
 [dependencies]
 meilisearch = { path = "../meilisearch" , default-features = false}
 serde_json = "1.0"
-clap = { version = "4.5.40", features = ["derive"] }
+clap = { version = "4.5.52", features = ["derive"] }
-anyhow = "1.0.98"
+anyhow = "1.0.100"
 utoipa = "5.4.0"

crates/tracing-trace/Cargo.toml

@@ -8,8 +8,8 @@ edition = "2021"
 [dependencies]
 color-spantrace = "0.3.0"
 fxprof-processed-profile = "0.7.0"
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = "1.0.140"
+serde_json = "1.0.145"
 tracing = "0.1.41"
 tracing-error = "0.2.1"
 tracing-subscriber = "0.3.20"
@@ -18,7 +18,7 @@ byte-unit = { version = "5.1.6", default-features = false, features = [
     "byte",
     "serde",
 ] }
-tokio = { version = "1.45.1", features = ["sync"] }
+tokio = { version = "1.48.0", features = ["sync"] }
 
 [target.'cfg(any(target_os = "linux", target_os = "macos"))'.dependencies]
-libproc = "0.14.10"
+libproc = "0.14.11"

crates/xtask/Cargo.toml

@@ -11,27 +11,27 @@ license.workspace = true
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
 
 [dependencies]
-anyhow = "1.0.98"
+anyhow = "1.0.100"
 build-info = { version = "1.7.0", path = "../build-info" }
-cargo_metadata = "0.20.0"
+cargo_metadata = "0.23.1"
-clap = { version = "4.5.40", features = ["derive"] }
+clap = { version = "4.5.52", features = ["derive"] }
 futures-core = "0.3.31"
 futures-util = "0.3.31"
-reqwest = { version = "0.12.20", features = [
+reqwest = { version = "0.12.24", features = [
     "stream",
     "json",
     "rustls-tls",
 ], default-features = false }
-serde = { version = "1.0.219", features = ["derive"] }
+serde = { version = "1.0.228", features = ["derive"] }
-serde_json = "1.0.140"
+serde_json = "1.0.145"
 sha2 = "0.10.9"
-sysinfo = "0.35.2"
+sysinfo = "0.37.2"
-time = { version = "0.3.41", features = [
+time = { version = "0.3.44", features = [
     "serde",
     "serde-human-readable",
     "macros",
 ] }
-tokio = { version = "1.45.1", features = [
+tokio = { version = "1.48.0", features = [
     "rt",
     "net",
     "time",
@@ -41,4 +41,4 @@ tokio = { version = "1.45.1", features = [
 tracing = "0.1.41"
 tracing-subscriber = "0.3.20"
 tracing-trace = { version = "0.1.0", path = "../tracing-trace" }
-uuid = { version = "1.17.0", features = ["v7", "serde"] }
+uuid = { version = "1.18.1", features = ["v7", "serde"] }