From fcc5b51bfdce7783621fcd5c872ffa4041023852 Mon Sep 17 00:00:00 2001
From: notsafeforgit <github@ak.codes>
Date: Fri, 20 Mar 2026 04:53:39 -0700
Subject: [PATCH] fix(sqlite): fix image duplicate detection by scanning phash
 as integer

This fixes a bug where identical image duplicates were not being detected.

The implementation was incorrectly scanning the phash BLOB into a string and then attempting to parse it as a hex string. Since phashes are stored as 64-bit integers, they were being converted to decimal strings. For phashes with the MSB set (negative when treated as int64), the resulting decimal string started with a '-', which caused the hex parser to fail and skip the image entirely.

Additionally, even for non-negative phashes, parsing a decimal string as hex yielded incorrect hash values.

Scanning directly into the utils.Phash struct (which uses int64) matches how Scene phashes are handled and ensures the hash values are correct.
---
 pkg/sqlite/image.go | 22 +++-------------------
 1 file changed, 3 insertions(+), 19 deletions(-)

diff --git a/pkg/sqlite/image.go b/pkg/sqlite/image.go
index e106167c7..780979270 100644
--- a/pkg/sqlite/image.go
+++ b/pkg/sqlite/image.go
@@ -1105,29 +1105,13 @@ func (qb *ImageStore) FindDuplicates(ctx context.Context, distance int) ([][]*mo
 
 	var hashes []*utils.Phash
 	if err := imageRepository.queryFunc(ctx, query, nil, false, func(rows *sqlx.Rows) error {
-		var sq struct {
-			ID    int     `db:"id"`
-			Phash *string `db:"phash"`
-		}
-		if err := rows.StructScan(&sq); err != nil {
-			return err
-		}
-
-		if sq.Phash == nil {
-			return nil
-		}
-
-		hashInt, err := utils.StringToPhash(*sq.Phash)
-		if err != nil {
-			return nil
-		}
-
 		phash := utils.Phash{
-			ID:       sq.ID,
-			Hash:     hashInt,
 			Bucket:   -1,
 			Duration: -1,
 		}
+		if err := rows.StructScan(&phash); err != nil {
+			return err
+		}
 
 		hashes = append(hashes, &phash)
 		return nil