refactor(handlers): split urls.rs into workers, pipeline and startup repair

urls.rs carried five reasons to change: the job channel and its worker pool,
the single-flight fetch, URL parsing, the per-URL pipeline, and the startup
repair of queued retries. The two with their own lifecycle move out:

- url_workers.rs: the bounded channel, its supervised pool and
  start/stop_url_workers (the pipeline stays in urls.rs, which the workers
  call).
- repair.rs: needs_refetch/apply_refresh/refetch/repair_lost_local_media with
  their tests, moved whole (the live one keeps its #[ignore]).

Also folds the two byte-identical render_fields -> CachedPost mappings
(urls.rs and repair.rs) into urls::cached_snapshot, and moves the shared
permanent_error test fixture into ctx::test_support.

crates/xmedia-bot/src/ctx.rs

@@ -65,6 +65,13 @@ pub(crate) mod test_support {
         RequestError::Api(ApiError::Unknown(message.to_string()))
     }
 
+    /// The API error a caption edit that changes nothing answers with — what
+    /// the mocks script for a permanent send failure. A `fn` pointer, so it can
+    /// be handed to `MockSender::scripted` as-is.
+    pub(crate) fn permanent_error() -> RequestError {
+        api_error("Bad Request: message is not modified")
+    }
+
     /// One photo payload item: `media` in the two flags the tests vary (no
     /// smaller variant, since that is the field most tests leave alone —
     /// `send`'s own tests build that case directly).