Normalize URLs when checking for duplicates (#1169)

* Normalize URLs when checking for duplicates

* Improve migration script

bookmarks/migrations/0047_populate_url_normalized_field.py

@@ -0,0 +1,38 @@
+# Generated by Django 5.2.3 on 2025-08-22 08:28
+
+from django.db import migrations, transaction
+from bookmarks.utils import normalize_url
+
+
+def populate_url_normalized(apps, schema_editor):
+    Bookmark = apps.get_model("bookmarks", "Bookmark")
+
+    batch_size = 500
+    with transaction.atomic():
+        qs = Bookmark.objects.all()
+        for start in range(0, qs.count(), batch_size):
+            batch = list(qs[start : start + batch_size])
+            for bookmark in batch:
+                bookmark.url_normalized = normalize_url(bookmark.url)
+            Bookmark.objects.bulk_update(
+                batch, ["url_normalized"], batch_size=batch_size
+            )
+
+
+def reverse_populate_url_normalized(apps, schema_editor):
+    Bookmark = apps.get_model("bookmarks", "Bookmark")
+    Bookmark.objects.all().update(url_normalized="")
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ("bookmarks", "0046_add_url_normalized_field"),
+    ]
+
+    operations = [
+        migrations.RunPython(
+            populate_url_normalized,
+            reverse_populate_url_normalized,
+        ),
+    ]