perf(cache): use Arc<str> for etag/content_type in ContentCachePort

Replace String with Arc<str> for etag and content_type fields in the
content cache. String::clone() allocates and copies the full string on
every cache hit (O(n)), while Arc<str>::clone() is O(1) — just an atomic
ref-count increment.

This eliminates 2 heap allocations per cache hit on the hottest download
path. At 1000 req/s that is 2000 fewer alloc/dealloc cycles per second.

Changed files:
- cache_ports.rs: trait signatures String → Arc<str>
- file_content_cache.rs: CacheEntry fields, get/put methods, tests
- stubs.rs: StubContentCachePort signatures
- file_retrieval_service.rs: caller creates Arc<str> before put()

src/application/ports/cache_ports.rs

@@ -12,6 +12,7 @@ use crate::common::errors::DomainError;
 use async_trait::async_trait;
 use bytes::Bytes;
 use std::path::{Path, PathBuf};
+use std::sync::Arc;
 
 /// Statistics for monitoring write-behind cache status.
 #[derive(Debug, Clone, Default)]
@@ -111,10 +112,13 @@ pub trait ContentCachePort: Send + Sync + 'static {
     fn should_cache(&self, size: usize) -> bool;
 
     /// Get cached content. Returns `(content, etag, content_type)` on hit.
-    async fn get(&self, file_id: &str) -> Option<(Bytes, String, String)>;
+    ///
+    /// `etag` and `content_type` are `Arc<str>` so cloning on cache hit is O(1)
+    /// (atomic ref-count increment) instead of O(n) heap allocation.
+    async fn get(&self, file_id: &str) -> Option<(Bytes, Arc<str>, Arc<str>)>;
 
     /// Store content in the cache (may evict older entries).
-    async fn put(&self, file_id: String, content: Bytes, etag: String, content_type: String);
+    async fn put(&self, file_id: String, content: Bytes, etag: Arc<str>, content_type: Arc<str>);
 
     /// Remove a file from the cache (e.g. on delete/update).
     async fn invalidate(&self, file_id: &str);