Update rust Docker tag to v1.95.0

Dockerfile.nidx

@@ -3,7 +3,7 @@
 # For a version that uses pre-built binaries (outside Docker) see `Dockerfile.nidx_prebuild`
 #
 
-FROM rust:1.94.0-trixie AS builder
+FROM rust:1.95.0-trixie AS builder
 RUN apt update && apt install -y protobuf-compiler
 COPY nucliadb_protos/*proto /app/nucliadb_protos/
 COPY nidx /app/nidx

nidx/nidx_paragraph/src/reader.rs

@@ -26,8 +26,8 @@ use nidx_protos::{OrderBy, ParagraphItem, ParagraphSearchResponse, StreamRequest
 use nidx_types::prefilter::PrefilterResult;
 use tantivy::collector::{Collector, Count, FacetCollector, TopDocs};
 use tantivy::query::{AllQuery, Query};
-use tantivy::{DateTime, Order, schema::*};
-use tantivy::{DocAddress, Index, IndexReader};
+use tantivy::schema::{Facet, Field, Value};
+use tantivy::{DateTime, DocAddress, Index, IndexReader, Order, TantivyDocument};
 use tracing::*;
 
 use super::schema::ParagraphSchema;

nidx/nidx_vector/src/data_store/v2.rs

@@ -65,7 +65,7 @@ impl DataStoreV2 {
             None
         };
 
-        for (idx, elem) in (0..).zip(entries.into_iter()) {
+        for (idx, elem) in (0..).zip(entries) {
             let (first_vector, _) = vectors.write(idx, elem.vectors.iter().map(|v| config.vector_type.encode(v)))?;
             if let Some(quantized) = &mut quantized {
                 for v in &elem.vectors {

nidx/nidx_vector/src/hnsw/search.rs

@@ -207,11 +207,7 @@ impl<'a, DR: DataRetriever> HnswSearcher<'a, DR> {
 
         let mut preloaded = 0;
 
-        loop {
-            let Some(Cnx(candidate, candidate_similarity)) = candidates.pop() else {
-                break;
-            };
-
+        while let Some(Cnx(candidate, candidate_similarity)) = candidates.pop() {
             if candidate_similarity < self.retriever.min_score() {
                 break;
             }

nidx/nidx_vector/src/indexer.rs

@@ -65,29 +65,26 @@ impl<'a> ResourceWrapper<'a> {
 
     pub fn fields(&self) -> impl Iterator<Item = (&String, impl Iterator<Item = ParagraphVectors<'_>>)> {
         self.resource.paragraphs.iter().map(|(field_id, paragraphs_wrapper)| {
-            let sentences_iterator = paragraphs_wrapper
-                .paragraphs
-                .iter()
-                .filter_map(|(_paragraph_id, paragraph)| {
-                    let sentences = if let Some(vectorset) = &self.vectorset {
-                        // indexing a vectorset, we should return only paragraphs from this vectorset.
-                        // If vectorset is not found, we'll skip this paragraph
-                        if let Some(vectorset_sentences) = paragraph.vectorsets_sentences.get(vectorset) {
-                            Some(&vectorset_sentences.sentences)
-                        } else if self.fallback_to_default_vectorset {
-                            Some(&paragraph.sentences)
-                        } else {
-                            None
-                        }
-                    } else {
-                        // Default vectors index (no vectorset)
+            let sentences_iterator = paragraphs_wrapper.paragraphs.values().filter_map(|paragraph| {
+                let sentences = if let Some(vectorset) = &self.vectorset {
+                    // indexing a vectorset, we should return only paragraphs from this vectorset.
+                    // If vectorset is not found, we'll skip this paragraph
+                    if let Some(vectorset_sentences) = paragraph.vectorsets_sentences.get(vectorset) {
+                        Some(&vectorset_sentences.sentences)
+                    } else if self.fallback_to_default_vectorset {
                         Some(&paragraph.sentences)
-                    };
-                    sentences.map(|s| ParagraphVectors {
-                        vectors: s,
-                        labels: &paragraph.labels,
-                    })
-                });
+                    } else {
+                        None
+                    }
+                } else {
+                    // Default vectors index (no vectorset)
+                    Some(&paragraph.sentences)
+                };
+                sentences.map(|s| ParagraphVectors {
+                    vectors: s,
+                    labels: &paragraph.labels,
+                })
+            });
             (field_id, sentences_iterator)
         })
     }

nidx/nidx_vector/src/segment.rs

@@ -823,7 +823,7 @@ mod test {
         let path_merged = tempdir()?;
         let merged_dp = merge(path_merged.path(), work, &config)?;
 
-        for (i, (elem, mut labels)) in elems1.into_iter().chain(elems2.into_iter()).enumerate() {
+        for (i, (elem, mut labels)) in elems1.into_iter().chain(elems2).enumerate() {
             let vector = merged_dp.data_store.get_vector(VectorAddr(i as u32));
             assert_eq!(config.vector_type.encode(&elem.vectors[0]), vector.vector());
 

nidx/src/searcher.rs

@@ -76,7 +76,7 @@ async fn refresher_task(mut rx: Receiver<(IndexId, bool)>, index_cache: Arc<Inde
         }
 
         let mut unique_indexes = HashMap::new();
-        for (index_id, force) in try_later.drain(std::ops::RangeFull).chain(recv_buf.into_iter()) {
+        for (index_id, force) in try_later.drain(std::ops::RangeFull).chain(recv_buf) {
             unique_indexes
                 .entry(index_id)
                 .and_modify(|f| *f = *f || force)

nidx/src/searcher/sync.rs

@@ -114,7 +114,7 @@ pub async fn run_sync(
 
             // Remove deleted indexes
             let deleted = Index::marked_to_delete(&meta.pool).await?;
-            for (shard_id, index_id) in deleted.into_iter().map(|i| (i.shard_id, i.id)).chain(indexes_to_delete.into_iter()) {
+            for (shard_id, index_id) in deleted.into_iter().map(|i| (i.shard_id, i.id)).chain(indexes_to_delete) {
                 if shutdown.is_cancelled() {
                     break;
                 }
@@ -140,7 +140,7 @@ pub async fn run_sync(
 
             let sync_semaphore = Arc::new(Semaphore::new(settings.parallel_index_syncs));
             let mut tasks = JoinSet::new();
-            for index in indexes.into_iter().chain(retry_indexes.into_iter()).chain(indexes_to_sync.into_iter()) {
+            for index in indexes.into_iter().chain(retry_indexes).chain(indexes_to_sync) {
                 let index_id = index.id;
                 let meta2 = meta.clone();
                 let index_metadata2 = Arc::clone(&index_metadata);

rust-toolchain.toml

@@ -1,3 +1,3 @@
 [toolchain]
 # if this changes, also change the version in the Dockerfiles
-channel = "1.94.0"
+channel = "1.95.0"