feat(transcribe): route dictionary_terms + initial_prompt through active profile

src-tauri/src/commands/live.rs

@@ -61,6 +61,10 @@ pub struct StartLiveTranscriptionConfig {
     /// Optional explicit microphone device name (from `list_audio_devices`).
     /// None or empty string = let `MicrophoneCapture::start` auto-select.
     pub microphone_device: Option<String>,
+    /// Optional profile id. None falls back to `DEFAULT_PROFILE_ID`. Drives
+    /// the post-processing dictionary via `profile_terms` and, when the
+    /// request's `initial_prompt` is empty, supplies the Whisper prompt.
+    pub profile_id: Option<String>,
 }
 
 #[derive(Debug, Serialize)]
@@ -131,7 +135,7 @@ struct InferenceTask {
 pub async fn start_live_transcription_session(
     state: tauri::State<'_, AppState>,
     live_state: tauri::State<'_, LiveTranscriptionState>,
-    config: StartLiveTranscriptionConfig,
+    mut config: StartLiveTranscriptionConfig,
     result_channel: Channel<LiveResultMessage>,
     status_channel: Channel<LiveStatusMessage>,
 ) -> Result<StartLiveTranscriptionResponse, String> {
@@ -142,6 +146,37 @@ pub async fn start_live_transcription_session(
         }
     }
 
+    let resolved_profile_id = config
+        .profile_id
+        .clone()
+        .unwrap_or_else(|| kon_storage::DEFAULT_PROFILE_ID.to_string());
+
+    let profile = kon_storage::database::get_profile(&state.db, &resolved_profile_id)
+        .await
+        .map_err(|e| e.to_string())?
+        .ok_or_else(|| format!("Profile {resolved_profile_id} not found"))?;
+
+    let profile_terms: Vec<String> =
+        kon_storage::database::list_profile_terms(&state.db, &resolved_profile_id)
+            .await
+            .map_err(|e| e.to_string())?
+            .into_iter()
+            .map(|t| t.term)
+            .collect();
+
+    // Collapse the effective initial_prompt on the struct so downstream
+    // `TranscriptionOptions` construction (see `maybe_dispatch_chunk`) picks
+    // up profile fallback without further plumbing.
+    let effective_prompt = match config.initial_prompt.as_deref() {
+        Some(p) if !p.is_empty() => p.to_string(),
+        _ => profile.initial_prompt.clone(),
+    };
+    config.initial_prompt = if effective_prompt.is_empty() {
+        None
+    } else {
+        Some(effective_prompt)
+    };
+
     let model_id = config
         .model_id
         .clone()
@@ -166,13 +201,7 @@ pub async fn start_live_transcription_session(
     let worker_status = status_channel.clone();
     let worker_results = result_channel.clone();
 
-    let dictionary_terms: Vec<String> =
-        kon_storage::database::list_dictionary(&state.db)
-            .await
-            .unwrap_or_default()
-            .into_iter()
-            .map(|e| e.term)
-            .collect();
+    let dictionary_terms = profile_terms.clone();
 
     let handle = tokio::task::spawn_blocking(move || {
         run_live_session(

src-tauri/src/commands/transcription.rs

@@ -36,12 +36,39 @@ pub async fn transcribe_pcm(
     british_english: bool,
     anti_hallucination: bool,
     format_mode: String,
+    profile_id: Option<String>,
 ) -> Result<(), String> {
+    let resolved_profile_id = profile_id
+        .unwrap_or_else(|| kon_storage::DEFAULT_PROFILE_ID.to_string());
+
+    let profile = kon_storage::database::get_profile(&state.db, &resolved_profile_id)
+        .await
+        .map_err(|e| e.to_string())?
+        .ok_or_else(|| format!("Profile {resolved_profile_id} not found"))?;
+
+    let profile_terms: Vec<String> =
+        kon_storage::database::list_profile_terms(&state.db, &resolved_profile_id)
+            .await
+            .map_err(|e| e.to_string())?
+            .into_iter()
+            .map(|t| t.term)
+            .collect();
+
+    let effective_prompt = if !initial_prompt.is_empty() {
+        initial_prompt
+    } else {
+        profile.initial_prompt.clone()
+    };
+
     let engine = state.whisper_engine.clone();
     let audio = kon_core::AudioSamples::mono_16khz(samples);
     let options = TranscriptionOptions {
         language: Some(language),
-        initial_prompt: Some(initial_prompt),
+        initial_prompt: if effective_prompt.is_empty() {
+            None
+        } else {
+            Some(effective_prompt)
+        },
     };
 
     let timed = tokio::task::spawn_blocking(move || {
@@ -50,12 +77,7 @@ pub async fn transcribe_pcm(
     .await
     .map_err(|e| e.to_string())??;
 
-    let dictionary_terms: Vec<String> = kon_storage::database::list_dictionary(&state.db)
-        .await
-        .unwrap_or_default()
-        .into_iter()
-        .map(|e| e.term)
-        .collect();
+    let dictionary_terms = profile_terms.clone();
 
     let mut segments: Vec<Segment> = timed.transcript.segments().to_vec();
     post_process_segments(
@@ -98,7 +120,30 @@ pub async fn transcribe_file(
     british_english: bool,
     anti_hallucination: bool,
     format_mode: String,
+    profile_id: Option<String>,
 ) -> Result<serde_json::Value, String> {
+    let resolved_profile_id = profile_id
+        .unwrap_or_else(|| kon_storage::DEFAULT_PROFILE_ID.to_string());
+
+    let profile = kon_storage::database::get_profile(&state.db, &resolved_profile_id)
+        .await
+        .map_err(|e| e.to_string())?
+        .ok_or_else(|| format!("Profile {resolved_profile_id} not found"))?;
+
+    let profile_terms: Vec<String> =
+        kon_storage::database::list_profile_terms(&state.db, &resolved_profile_id)
+            .await
+            .map_err(|e| e.to_string())?
+            .into_iter()
+            .map(|t| t.term)
+            .collect();
+
+    let effective_prompt = if !initial_prompt.is_empty() {
+        initial_prompt
+    } else {
+        profile.initial_prompt.clone()
+    };
+
     let engine_name = engine.unwrap_or_else(|| "whisper".to_string());
     let model_id = model_id
         .unwrap_or_else(|| default_model_id_for_engine(&engine_name).to_string());
@@ -107,7 +152,11 @@ pub async fn transcribe_file(
     let engine = pick_engine(&state, &engine_name)?;
     let options = TranscriptionOptions {
         language: Some(language),
-        initial_prompt: Some(initial_prompt),
+        initial_prompt: if effective_prompt.is_empty() {
+            None
+        } else {
+            Some(effective_prompt)
+        },
     };
 
     let timed = tokio::task::spawn_blocking(move || {
@@ -122,12 +171,7 @@ pub async fn transcribe_file(
     .await
     .map_err(|e| e.to_string())??;
 
-    let dictionary_terms: Vec<String> = kon_storage::database::list_dictionary(&state.db)
-        .await
-        .unwrap_or_default()
-        .into_iter()
-        .map(|e| e.term)
-        .collect();
+    let dictionary_terms = profile_terms.clone();
 
     let mut segments: Vec<Segment> = timed.transcript.segments().to_vec();
     post_process_segments(
@@ -162,7 +206,26 @@ pub async fn transcribe_pcm_parakeet(
     british_english: bool,
     anti_hallucination: bool,
     format_mode: String,
+    profile_id: Option<String>,
 ) -> Result<(), String> {
+    let resolved_profile_id = profile_id
+        .unwrap_or_else(|| kon_storage::DEFAULT_PROFILE_ID.to_string());
+
+    // Validate the profile exists so parakeet and whisper behave identically
+    // when a bogus id slips through from the frontend.
+    kon_storage::database::get_profile(&state.db, &resolved_profile_id)
+        .await
+        .map_err(|e| e.to_string())?
+        .ok_or_else(|| format!("Profile {resolved_profile_id} not found"))?;
+
+    let profile_terms: Vec<String> =
+        kon_storage::database::list_profile_terms(&state.db, &resolved_profile_id)
+            .await
+            .map_err(|e| e.to_string())?
+            .into_iter()
+            .map(|t| t.term)
+            .collect();
+
     let engine = state.parakeet_engine.clone();
     let audio = kon_core::AudioSamples::mono_16khz(samples);
     let options = TranscriptionOptions::default();
@@ -173,12 +236,7 @@ pub async fn transcribe_pcm_parakeet(
     .await
     .map_err(|e| e.to_string())??;
 
-    let dictionary_terms: Vec<String> = kon_storage::database::list_dictionary(&state.db)
-        .await
-        .unwrap_or_default()
-        .into_iter()
-        .map(|e| e.term)
-        .collect();
+    let dictionary_terms = profile_terms.clone();
 
     let mut segments: Vec<Segment> = timed.transcript.segments().to_vec();
     post_process_segments(