From bab2176cbb38115c654ee95e8d64a84a9ddb3736 Mon Sep 17 00:00:00 2001
From: Paul Masurel <paul@quickwit.io>
Date: Thu, 1 Aug 2024 00:00:47 +0900
Subject: [PATCH] Avoid deserializing QueryAST for every split.

---
 quickwit/quickwit-search/src/leaf.rs | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/quickwit/quickwit-search/src/leaf.rs b/quickwit/quickwit-search/src/leaf.rs
index b225a53c13d..df563c231b0 100644
--- a/quickwit/quickwit-search/src/leaf.rs
+++ b/quickwit/quickwit-search/src/leaf.rs
@@ -345,6 +345,7 @@ fn get_leaf_resp_from_count(count: u64) -> LeafSearchResponse {
 async fn leaf_search_single_split(
     searcher_context: &SearcherContext,
     mut search_request: SearchRequest,
+    query_ast: Arc<QueryAst>,
     storage: Arc<dyn Storage>,
     split: SplitIdAndFooterOffsets,
     doc_mapper: Arc<dyn DocMapper>,
@@ -363,9 +364,6 @@ async fn leaf_search_single_split(
         return Ok(cached_answer);
     }
 
-    let query_ast: QueryAst = serde_json::from_str(search_request.query_ast.as_str())
-        .map_err(|err| SearchError::InvalidQuery(err.to_string()))?;
-
     // CanSplitDoBetter or rewrite_request may have changed the request to be a count only request
     // This may be the case for AllQuery with a sort by date and time filter, where the current
     // split can't have better results.
@@ -394,7 +392,7 @@ async fn leaf_search_single_split(
     let mut collector =
         make_collector_for_split(split_id.clone(), &search_request, aggregations_limits)?;
 
-    let (query, mut warmup_info) = doc_mapper.query(split_schema.clone(), &query_ast, false)?;
+    let (query, mut warmup_info) = doc_mapper.query(split_schema.clone(), &*query_ast, false)?;
 
     let collector_warmup_info = collector.warmup_info();
     warmup_info.merge(collector_warmup_info);
@@ -413,7 +411,7 @@ async fn leaf_search_single_split(
                 // request based on the results of the preceding searches
                 check_optimize_search_request(&mut search_request, &split, &split_filter);
                 collector.update_search_param(&search_request);
-                if is_metadata_count_request_with_ast(&query_ast, &search_request) {
+                if is_metadata_count_request_with_ast(&*query_ast, &search_request) {
                     return Ok((
                         search_request,
                         get_leaf_resp_from_count(searcher.num_docs() as u64),
@@ -1223,6 +1221,10 @@ pub async fn leaf_search(
     let incremental_merge_collector = IncrementalCollector::new(merge_collector);
     let incremental_merge_collector = Arc::new(Mutex::new(incremental_merge_collector));
 
+   let query_ast: Arc<QueryAst> = serde_json::from_str::<QueryAst>(&request.query_ast)
+       .map_err(|err| SearchError::InvalidQuery(err.to_string()))?
+       .into();
+
     for (split, mut request) in split_with_req {
         let leaf_split_search_permit = searcher_context.leaf_search_split_semaphore
             .clone()
@@ -1239,6 +1241,7 @@ pub async fn leaf_search(
         leaf_search_single_split_futures.push(tokio::spawn(
             leaf_search_single_split_wrapper(
                 request,
+                query_ast.clone(),
                 searcher_context.clone(),
                 index_storage.clone(),
                 doc_mapper.clone(),
@@ -1289,6 +1292,7 @@ pub async fn leaf_search(
 #[instrument(skip_all, fields(split_id = split.split_id))]
 async fn leaf_search_single_split_wrapper(
     request: SearchRequest,
+    query_ast: Arc<QueryAst>,
     searcher_context: Arc<SearcherContext>,
     index_storage: Arc<dyn Storage>,
     doc_mapper: Arc<dyn DocMapper>,
@@ -1305,6 +1309,7 @@ async fn leaf_search_single_split_wrapper(
     let leaf_search_single_split_res = leaf_search_single_split(
         &searcher_context,
         request,
+        query_ast,
         index_storage,
         split.clone(),
         doc_mapper,