From 39d5d407f41a004335fea8c8e40cf5d2be4601a1 Mon Sep 17 00:00:00 2001
From: "google-labs-jules[bot]"
 <161369871+google-labs-jules[bot]@users.noreply.github.com>
Date: Sat, 12 Sep 2026 13:02:30 +0000
Subject: [PATCH] perf(db): replace correlated subqueries with early
 aggregation for unpaginated queries

When querying entire unpaginated tables in SQLite, using correlated scalar subqueries `(SELECT COUNT(*) FROM...)` inside the `SELECT` clause results in an N+1 execution pattern. This commit optimizes two critical full-table scans (`consolidate_duplicate_books` and `find_book_by_title_author_conn`) by refactoring them to use "early aggregation" derived tables via `LEFT JOIN`s, dramatically improving query performance from O(N log M) to O(N + M).

Co-authored-by: jspann21 <179991454+jspann21@users.noreply.github.com>
---
 .jules/bolt.md              |  3 +++
 src-tauri/src/library/db.rs | 18 +++++++++++++-----
 2 files changed, 16 insertions(+), 5 deletions(-)
 create mode 100644 .jules/bolt.md

diff --git a/.jules/bolt.md b/.jules/bolt.md
new file mode 100644
index 0000000..2681816
--- /dev/null
+++ b/.jules/bolt.md
@@ -0,0 +1,3 @@
+## 2025-01-20 - Avoid repository pollution with benchmark scripts
+**Learning:** Checking in temporary test scripts and binary dummy databases (like `benchmark.py` and `test.db`) during development is considered a blocking issue for merges as it pollutes the repository.
+**Action:** Always place temporary testing artifacts in `/tmp/` and clean them up before requesting code review or submitting changes.
diff --git a/src-tauri/src/library/db.rs b/src-tauri/src/library/db.rs
index 6040018..38f645c 100644
--- a/src-tauri/src/library/db.rs
+++ b/src-tauri/src/library/db.rs
@@ -946,6 +946,9 @@ impl Repository {
     let tx = conn.transaction()?;
 
     let mut stmt = tx.prepare(
+      // Performance optimization: Avoid correlated scalar subqueries in the SELECT clause (EXISTS, COUNT)
+      // during unpaginated full-table scans. Using Early Aggregation via derived tables prevents
+      // an N+1 execution bottleneck.
       "SELECT
          b.id,
          b.title,
@@ -953,9 +956,11 @@ impl Repository {
          b.isbn10,
          b.isbn13,
          b.updated_at,
-         EXISTS(SELECT 1 FROM manual_overrides mo WHERE mo.book_id = b.id) AS has_manual_overrides,
-         (SELECT COUNT(*) FROM book_files bf WHERE bf.book_id = b.id) AS file_count
-       FROM books b",
+         mo.has_manual IS NOT NULL AS has_manual_overrides,
+         COALESCE(bf.file_count, 0) AS file_count
+       FROM books b
+       LEFT JOIN (SELECT book_id, 1 AS has_manual FROM manual_overrides GROUP BY book_id) mo ON mo.book_id = b.id
+       LEFT JOIN (SELECT book_id, COUNT(*) AS file_count FROM book_files GROUP BY book_id) bf ON bf.book_id = b.id",
     )?;
 
     let mut candidates = Vec::new();
@@ -1339,9 +1344,12 @@ impl Repository {
          (SELECT COUNT(*) FROM book_files bf WHERE bf.book_id = lb.id) AS file_count
        FROM limited_books lb"
     } else {
+      // Performance optimization: Avoid correlated scalar subqueries (COUNT) in the SELECT
+      // clause during large unpaginated table scans. Early Aggregation prevents an N+1 bottleneck.
       "SELECT b.id, b.title, b.authors_json,
-         (SELECT COUNT(*) FROM book_files bf WHERE bf.book_id = b.id) AS file_count
-       FROM books b"
+         COALESCE(bf.file_count, 0) AS file_count
+       FROM books b
+       LEFT JOIN (SELECT book_id, COUNT(*) AS file_count FROM book_files GROUP BY book_id) bf ON bf.book_id = b.id"
     };
     let mut stmt = conn.prepare(query)?;