Add per-user whisper settings for hotwords and initial prompt

This commit introduces the capability to configure user-specific Whisper
settings, including language, hotwords, and initial prompts. These
settings are stored in `users.toml` and are passed to the Whisper
service for more tailored transcription results.

New unit tests have been added to verify that the `transcribe` function
correctly forwards these optional settings to the Whisper client and
omits them when they are not provided.

Additionally, a new directory `tests/fixtures/dictations` has been
created to store M4A audio files, their expected transcripts, and
associated hotword files. This serves as a regression corpus for the
Whisper pipeline. A README file explains the structure and conventions
for adding new fixtures. Several new fixtures have been added, covering
various medical domains and transcription scenarios.

server/tests/transcribe_test.rs

@@ -135,6 +135,76 @@ async fn whisper_client_times_out() {
     assert!(matches!(err, WhisperError::Http(_)), "expected Http error, got {err:?}");
 }
 
+#[tokio::test]
+async fn whisper_client_forwards_hotwords_and_prompt_when_set() {
+    let server = MockServer::start().await;
+
+    // Accept any POST; inspect captured body below. language query param is
+    // still matched strictly to verify override from settings.
+    Mock::given(method("POST"))
+        .and(path("/asr"))
+        .and(query_param("language", "en"))
+        .respond_with(ResponseTemplate::new(200).set_body_string("ok"))
+        .expect(1)
+        .mount(&server)
+        .await;
+
+    let client = reqwest::Client::new();
+    let settings = WhisperUserSettings {
+        language: Some("en".into()),
+        hotwords: Some("HOCM Valsalva".into()),
+        initial_prompt: Some("Kardiologie".into()),
+    };
+    transcribe(
+        &client,
+        &server.uri(),
+        &fixture("sample.m4a"),
+        Duration::from_secs(10),
+        &settings,
+    )
+    .await
+    .expect("transcribe failed");
+
+    let received = server.received_requests().await.unwrap();
+    let body = String::from_utf8_lossy(&received[0].body);
+    assert!(body.contains("name=\"hotwords\""), "hotwords part missing");
+    assert!(body.contains("HOCM Valsalva"), "hotwords value missing: {body}");
+    assert!(body.contains("name=\"initial_prompt\""), "initial_prompt part missing");
+    assert!(body.contains("Kardiologie"), "initial_prompt value missing");
+}
+
+#[tokio::test]
+async fn whisper_client_omits_empty_optional_fields() {
+    let server = MockServer::start().await;
+
+    // Default settings = all None → only `audio_file` part, language defaults
+    // to `de`. Assert that neither optional field appears in the body.
+    Mock::given(method("POST"))
+        .and(path("/asr"))
+        .and(query_param("language", "de"))
+        .respond_with(ResponseTemplate::new(200).set_body_string("ok"))
+        .expect(1)
+        .mount(&server)
+        .await;
+
+    let client = reqwest::Client::new();
+    transcribe(
+        &client,
+        &server.uri(),
+        &fixture("sample.m4a"),
+        Duration::from_secs(10),
+        &WhisperUserSettings::default(),
+    )
+    .await
+    .expect("transcribe failed");
+
+    // Inspect the captured request to confirm absence of the optional parts.
+    let received = server.received_requests().await.unwrap();
+    let body = String::from_utf8_lossy(&received[0].body);
+    assert!(!body.contains("name=\"hotwords\""), "hotwords part leaked: {body}");
+    assert!(!body.contains("name=\"initial_prompt\""), "initial_prompt part leaked: {body}");
+}
+
 #[tokio::test]
 async fn recovery_enqueues_only_pending_recordings() {
     let tmp = tempfile::tempdir().unwrap();