fix(share): keep exactly one custom address per page on alias edit (#226)

Editing an existing share alias (e.g. slug `te` -> `ted`) failed to update the displayed `/l/<alias>` link: `setAlias()` looked the requested slug up by name and, if free, INSERTed a brand-new row, leaving the page with multiple alias rows. The modal then read via `findByPageId().executeTakeFirst()` with no `ORDER BY`, so Postgres returned an arbitrary (in practice the oldest, stale) row. Every edit also spawned an orphan row that kept a live `/l/<old>` link forever. Regression of #205. Enforce the invariant "a page has EXACTLY ONE custom address": - `setAlias()` now resolves the page's current alias row and RENAMES it in place when the requested name is free (insert only when the page has none), keeps the same-name no-op and the cross-page 409 `ALIAS_REASSIGN_REQUIRED` + confirmed-retarget flow, and after any successful write DELETEs all other alias rows for the page (self-heal). Runs in one transaction so the page is never transiently empty or duplicated. - repo: add `updateAlias` (rename) and `deleteOthersForPage`; make `findByPageId` deterministic with `ORDER BY created_at DESC, id DESC`. - migration: dedup existing rows (keep newest per page) + a PARTIAL unique index `(workspace_id, page_id) WHERE page_id IS NOT NULL` so dangling aliases still coexist while live ones are one-per-page. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-27 00:56:37 +03:00
parent b6630deb32
commit 9d2bec8eb8
6 changed files with 587 additions and 56 deletions
@@ -0,0 +1,41 @@
+import { type Kysely, sql } from 'kysely';
+
+/**
+ * Enforce "a page has EXACTLY ONE custom address" at the DB level. The original
+ * `share_aliases` table only had a unique index on `(workspace_id, alias)`, so a
+ * page could accumulate several alias rows (every slug edit used to INSERT a new
+ * one), leaving orphan `/l/<old>` links live forever and making the share
+ * modal's `findByPageId` lookup nondeterministic.
+ *
+ * We first dedup any pre-existing rows (keeping the NEWEST per page — the same
+ * "current" choice the read path now makes), then add a PARTIAL unique index on
+ * `(workspace_id, page_id)`. It is partial (`WHERE page_id IS NOT NULL`) so that
+ * multiple DANGLING aliases (target page deleted -> `page_id` SET NULL) can
+ * still coexist without colliding.
+ */
+export async function up(db: Kysely<any>): Promise<void> {
+  // Reap legacy duplicates: for each (workspace_id, page_id) keep only the row
+  // with the greatest (created_at, id) — matches ShareAliasRepo.findByPageId.
+  await sql`
+    DELETE FROM share_aliases sa
+    USING share_aliases keep
+    WHERE sa.page_id IS NOT NULL
+      AND sa.workspace_id = keep.workspace_id
+      AND sa.page_id = keep.page_id
+      AND (keep.created_at, keep.id) > (sa.created_at, sa.id)
+  `.execute(db);
+
+  await db.schema
+    .createIndex('share_aliases_workspace_id_page_id_unique')
+    .on('share_aliases')
+    .columns(['workspace_id', 'page_id'])
+    .unique()
+    .where('page_id', 'is not', null)
+    .execute();
+}
+
+export async function down(db: Kysely<any>): Promise<void> {
+  await db.schema
+    .dropIndex('share_aliases_workspace_id_page_id_unique')
+    .execute();
+}