Merge #4889

4889: When `retrieveVectors` is true, retrieve `_vectors.embedder` even if … r=Kerollmops a=dureuill

…there are no vector for that embedder


backports a bug fix from v1.10.0: 82647bcded

Co-authored-by: Louis Dureuil <louis@meilisearch.com>

.github/workflows/flaky-tests.yml

@@ -1,4 +1,6 @@
 name: Look for flaky tests
+env:
+  ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
 on:
   workflow_dispatch:
   schedule:

.github/workflows/fuzzer-indexing.yml

@@ -1,5 +1,6 @@
 name: Run the indexing fuzzer
-
+env:
+  ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
 on:
   push:
     branches:

.github/workflows/publish-apt-brew-pkg.yml

@@ -15,6 +15,8 @@ jobs:
 
   debian:
     name: Publish debian packagge
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     runs-on: ubuntu-latest
     needs: check-version
     container:

.github/workflows/publish-binaries.yml

@@ -35,6 +35,8 @@ jobs:
   publish-linux:
     name: Publish binary for Linux
     runs-on: ubuntu-latest
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     needs: check-version
     container:
       # Use ubuntu-18.04 to compile with glibc 2.27
@@ -132,6 +134,8 @@ jobs:
     name: Publish binary for aarch64
     runs-on: ubuntu-latest
     needs: check-version
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     container:
       # Use ubuntu-18.04 to compile with glibc 2.27
       image: ubuntu:18.04

.github/workflows/test-suite.yml

@@ -21,6 +21,8 @@ jobs:
   test-linux:
     name: Tests on ubuntu-18.04
     runs-on: ubuntu-latest
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     container:
       # Use ubuntu-18.04 to compile with glibc 2.27, which are the production expectations
       image: ubuntu:18.04
@@ -77,6 +79,8 @@ jobs:
   test-all-features:
     name: Tests almost all features
     runs-on: ubuntu-latest
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     container:
       # Use ubuntu-18.04 to compile with glibc 2.27, which are the production expectations
       image: ubuntu:18.04
@@ -100,6 +104,8 @@ jobs:
 
   test-disabled-tokenization:
     name: Test disabled tokenization
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     runs-on: ubuntu-latest
     container:
       image: ubuntu:18.04
@@ -127,6 +133,8 @@ jobs:
   # We run tests in debug also, to make sure that the debug_assertions are hit
   test-debug:
     name: Run tests in debug
+    env:
+      ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true
     runs-on: ubuntu-latest
     container:
       # Use ubuntu-18.04 to compile with glibc 2.27, which are the production expectations

Cargo.lock

@@ -5098,9 +5098,9 @@ dependencies = [
 
 [[package]]
 name = "time"
-version = "0.3.34"
+version = "0.3.36"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c8248b6521bb14bc45b4067159b9b6ad792e2d6d754d6c41fb50e29fefe38749"
+checksum = "5dfd88e563464686c916c7e46e623e520ddc6d79fa6641390f2e3fa86e83e885"
 dependencies = [
  "deranged",
  "itoa",
@@ -5121,9 +5121,9 @@ checksum = "ef927ca75afb808a4d64dd374f00a2adf8d0fcff8e7b184af886c3c87ec4a3f3"
 
 [[package]]
 name = "time-macros"
-version = "0.2.17"
+version = "0.2.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7ba3a3ef41e6672a2f0f001392bb5dcd3ff0a9992d618ca761a11c3121547774"
+checksum = "3f252a68540fde3a3877aeea552b832b40ab9a69e318efd078774a01ddee1ccf"
 dependencies = [
  "num-conv",
  "time-core",

meilisearch/tests/documents/get_documents.rs

@@ -644,7 +644,12 @@ async fn get_document_with_vectors() {
         {
           "id": 1,
           "name": "echo",
-          "_vectors": {}
+          "_vectors": {
+            "manual": {
+              "embeddings": [],
+              "regenerate": false
+            }
+          }
         }
       ],
       "offset": 0,
@@ -700,7 +705,12 @@ async fn get_document_with_vectors() {
         },
         {
           "name": "echo",
-          "_vectors": {}
+          "_vectors": {
+            "manual": {
+              "embeddings": [],
+              "regenerate": false
+            }
+          }
         }
       ],
       "offset": 0,

meilisearch/tests/vector/mod.rs

@@ -119,7 +119,12 @@ async fn add_remove_user_provided() {
         {
           "id": 1,
           "name": "echo",
-          "_vectors": {}
+          "_vectors": {
+            "manual": {
+              "embeddings": [],
+              "regenerate": false
+            }
+          }
         }
       ],
       "offset": 0,
@@ -141,7 +146,12 @@ async fn add_remove_user_provided() {
         {
           "id": 1,
           "name": "echo",
-          "_vectors": {}
+          "_vectors": {
+            "manual": {
+              "embeddings": [],
+              "regenerate": false
+            }
+          }
         }
       ],
       "offset": 0,
@@ -577,7 +587,12 @@ async fn add_remove_one_vector_4588() {
         {
           "id": 0,
           "name": "kefir",
-          "_vectors": {}
+          "_vectors": {
+            "manual": {
+              "embeddings": [],
+              "regenerate": false
+            }
+          }
         }
       ],
       "offset": 0,

milli/Cargo.toml

@@ -141,3 +141,6 @@ swedish-recomposition = ["charabia/swedish-recomposition"]
 
 # allow CUDA support, see <https://github.com/meilisearch/meilisearch/issues/4306>
 cuda = ["candle-core/cuda"]
+
+[lints.rust]
+unexpected_cfgs = { level = "warn", check-cfg = ['cfg(fuzzing)'] }

milli/src/index.rs

@@ -1660,9 +1660,7 @@ impl Index {
                 }
             }
 
-            if !embeddings.is_empty() {
-                res.insert(embedder_name.to_owned(), embeddings);
-            }
+            res.insert(embedder_name.to_owned(), embeddings);
         }
         Ok(res)
     }

milli/src/update/index_documents/mod.rs

@@ -2180,33 +2180,6 @@ mod tests {
         index.add_documents(doc1).unwrap();
     }
 
-    #[cfg(feature = "default")]
-    #[test]
-    fn store_detected_script_and_language_per_document_during_indexing() {
-        use charabia::{Language, Script};
-        let index = TempIndex::new();
-        index
-            .add_documents(documents!([
-                { "id": 1, "title": "The quick (\"brown\") fox can't jump 32.3 feet, right? Brr, it's 29.3°F!" },
-                { "id": 2, "title": "人人生而自由﹐在尊嚴和權利上一律平等。他們賦有理性和良心﹐並應以兄弟關係的精神互相對待。" },
-                { "id": 3, "title": "הַשּׁוּעָל הַמָּהִיר (״הַחוּם״) לֹא יָכוֹל לִקְפֹּץ 9.94 מֶטְרִים, נָכוֹן? ברר, 1.5°C- בַּחוּץ!" },
-                { "id": 4, "title": "関西国際空港限定トートバッグ すもももももももものうち" },
-                { "id": 5, "title": "ภาษาไทยง่ายนิดเดียว" },
-                { "id": 6, "title": "The quick 在尊嚴和權利上一律平等。" },
-            ]))
-            .unwrap();
-
-        let rtxn = index.read_txn().unwrap();
-        let key_jpn = (Script::Cj, Language::Jpn);
-        let key_cmn = (Script::Cj, Language::Cmn);
-        let cj_jpn_docs = index.script_language_documents_ids(&rtxn, &key_jpn).unwrap().unwrap();
-        let cj_cmn_docs = index.script_language_documents_ids(&rtxn, &key_cmn).unwrap().unwrap();
-        let expected_cj_jpn_docids = [3].iter().collect();
-        assert_eq!(cj_jpn_docs, expected_cj_jpn_docids);
-        let expected_cj_cmn_docids = [1, 5].iter().collect();
-        assert_eq!(cj_cmn_docs, expected_cj_cmn_docids);
-    }
-
     #[test]
     fn add_and_delete_documents_in_single_transform() {
         let mut index = TempIndex::new();