fix(posts): ensure rebuild invalidates tag posts cache

src/api/posts/delete.rs

@@ -1,7 +1,7 @@
 //! 删除文章接口。
 //!
 //! 采用软删除方式，将 posts.deleted_at 设置为当前时间，
-//! 同时清空所有文章相关缓存。
+//! 并按影响范围失效相关缓存。
 //! Dioxus server function，注册在 `/api` 路径下。
 //! 仅在 `feature = "server"` 启用的服务端构建中执行删除与缓存失效。
 
@@ -18,7 +18,7 @@ use crate::db::pool::get_conn;
 /// 删除指定文章。
 ///
 /// 仅 admin 可调用；通过设置 deleted_at 实现软删除，
-/// 成功后清空全部文章缓存。
+/// 成功后按影响范围失效文章列表、标签云、统计、slug 及相关标签文章缓存。
 #[server(DeletePost, "/api")]
 pub async fn delete_post(post_id: i32) -> Result<CreatePostResponse, ServerFnError> {
     let _user = get_current_admin_user().await?;

src/api/posts/rebuild.rs

@@ -34,14 +34,14 @@ pub async fn rebuild_content_html(rebuild_all: bool) -> Result<RebuildResult, Se
     {
         let client = get_conn().await.map_err(AppError::db_conn)?;
 
-        // 根据参数构造 WHERE 条件，限制单次处理数量；同时查询 slug 用于精准失效缓存。
+        // 根据参数构造 WHERE 条件，限制单次处理数量。
         let query = if rebuild_all {
             format!(
-                "SELECT id, slug, content_md FROM posts WHERE deleted_at IS NULL ORDER BY id LIMIT {REBUILD_BATCH_LIMIT}"
+                "SELECT id, content_md FROM posts WHERE deleted_at IS NULL ORDER BY id LIMIT {REBUILD_BATCH_LIMIT}"
             )
         } else {
             format!(
-                "SELECT id, slug, content_md FROM posts WHERE deleted_at IS NULL AND content_html IS NULL ORDER BY id LIMIT {REBUILD_BATCH_LIMIT}"
+                "SELECT id, content_md FROM posts WHERE deleted_at IS NULL AND content_html IS NULL ORDER BY id LIMIT {REBUILD_BATCH_LIMIT}"
             )
         };
 
@@ -50,12 +50,10 @@ pub async fn rebuild_content_html(rebuild_all: bool) -> Result<RebuildResult, Se
         let mut rebuilt: u64 = 0;
         let mut failed: u64 = 0;
         let mut errors: Vec<String> = Vec::new();
-        let mut rebuilt_slugs: Vec<String> = Vec::with_capacity(rows.len());
 
         for row in &rows {
             let id: i32 = row.get(0);
-            let slug: String = row.get(1);
-            let content_md: String = row.get(2);
+            let content_md: String = row.get(1);
 
             // 捕获 Markdown 渲染 panic，避免单条记录导致整批失败。
             let rendered = match std::panic::catch_unwind(|| {
@@ -95,7 +93,6 @@ pub async fn rebuild_content_html(rebuild_all: bool) -> Result<RebuildResult, Se
             {
                 Ok(_) => {
                     rebuilt += 1;
-                    rebuilt_slugs.push(slug);
                 }
                 Err(_) => {
                     failed += 1;
@@ -106,14 +103,10 @@ pub async fn rebuild_content_html(rebuild_all: bool) -> Result<RebuildResult, Se
             }
         }
 
-        // 只要有文章被更新，就按影响范围失效缓存：列表、标签云、统计，以及每篇被重建文章的 slug 缓存。
+        // 重建会修改 word_count / reading_time 等列表项字段，批量影响列表、标签云、
+        // 标签文章及单篇缓存；这里使用全量失效作为务实的回退策略。
         if rebuilt > 0 {
-            crate::cache::invalidate_post_lists();
-            crate::cache::invalidate_all_tags();
-            crate::cache::invalidate_post_stats();
-            for slug in &rebuilt_slugs {
-                crate::cache::invalidate_post_by_slug(slug).await;
-            }
+            crate::cache::invalidate_all_post_caches();
         }
 
         Ok(RebuildResult {