Update script typo and better error handling

app.py

@@ -19,23 +19,24 @@ TSV_FILE = os.path.join(DATA_ROOT, "validated.tsv")
 
 os.makedirs(AUDIO_FOLDER, exist_ok=True)
 
-# 3. SCRIPTS
+# 3. SCRIPTS (Anonymized & Improved)
 SCRIPTS = [
     "Nagerageje kohereza amafaranga kuri MoMo, ariko rezo yari mbi cyane sinabasha kubona mesaje yemeza ko yagiye.",
     "Mbere yo kwinjira muri mudasobwa, banza urebe niba sisitemu yakoze update, hanyuma ushyiremo ijambo ry'ibanga kugira ngo ufungure.",
     "Nta megabayiti zihagije mfite muri telefoni, reka nshakishe wifi hano hafi kugira ngo mbashe gukurura iyo porogaramu vuba.",
-    "Iyi sharijeri ya smartphone iragura ibihumbi bitanu, ariko niba ushaka n’ibirahure bya ecran, turaguha byose ku icumi.",
-    "Nugera kuri banki i Nyarugenge, ubwire ejenti agufashe kubikuza ayo madolari mbere y’uko ukwezi kwa Gicurasi kurangira.",
-    "Wohereze imeyili itarimo amakosa kugira ngo ubashe guhindura mot de passe, hanyuma ukande kuri linki iri bube kuri ecran yawe.",
+    "Iyi sharijeri ya smartphone iragura ibihumbi bitanu, ariko niba ushaka n'ibirahure bya ecran, turaguha byose ku icumi.",
+    "Nugera kuri banki i Nyarugenge, ubwire ejenti agufashe kubikuza ayo madolari mbere y'uko ukwezi kwa Gicurasi kurangira.",
+    "Wohereze imeli itarimo amakosa kugira ngo ubashe guhindura mot de passe, hanyuma ukande kuri linki iri bube kuri ecran yawe.",
     "Ugomba gusiba amafoto adakenewe kugira ngo ubone aho kubika izindi gigabayiti ebyiri, cyangwa ukoreshe flash.",
     "Koresha Mokash wishyure iyo fagitire, kode ni zeru gatatu rimwe, ubundi tujye Gasabo ku wa Gatanu.",
     "Koresha kibodi wandike ubutumwa, hanyuma uwohereze kuri WhatsApp kuko interineti yo mu Majyaruguru iragenda buhoro.",
-    "Nuroha gukoresha iyo apurikasiyo, wibuke gukora log out kugira ngo hatagira undi umukiriya ureba amabanga yawe."
+    "Nurangiza gukoresha iyo apurikasiyo, wibuke gukora log out kugira ngo hatagira undi umukiriya ureba amabanga yawe."
 ]
 
 # --- HELPER FUNCTIONS ---
 
 def generate_tsv():
+    """Generates the Standard Mozilla 'validated.tsv' for the backup."""
     cv_header = ["client_id", "path", "sentence", "up_votes", "down_votes", "age", "gender", "accent", "locale", "segment"]
     tsv_content = []
     tsv_content.append("\t".join(cv_header))
@@ -65,6 +66,7 @@ def generate_tsv():
         f.write("\n".join(tsv_content))
 
 def backup_to_dataset_repo(repo_name=BACKUP_REPO_NAME):
+    """Backs up data to Hugging Face Dataset."""
     if not HF_TOKEN:
         print("⚠️ Backup Skipped: HF_TOKEN missing.")
         return
@@ -89,9 +91,22 @@ def backup_to_dataset_repo(repo_name=BACKUP_REPO_NAME):
         print(f"⚠️ Backup Failed: {e}")
 
 def save_data(age, gender, region, *audios):
-    if None in audios:
-        return "<span style='color: red'>⚠️ Error: Please record ALL sentences.</span>", gr.update(visible=True), gr.update(visible=True)
-
+    """
+    Saves data locally. 
+    Includes specific error checking to tell user WHICH audio is missing.
+    """
+    # 1. Identify missing recordings
+    missing_indices = []
+    for i, audio in enumerate(audios):
+        if audio is None:
+            missing_indices.append(str(i+1)) # Store 1-based index
+            
+    if missing_indices:
+        missing_str = ", ".join(missing_indices)
+        error_msg = f"⚠️ Error: You missed Sentence(s): {missing_str}. Please record them."
+        return f"<span style='color: red'>{error_msg}</span>", gr.update(visible=True), gr.update(visible=True)
+
+    # 2. Generate Filenames & Save
     timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
     participant_id = f"user_{timestamp}"
     filenames = []
@@ -102,6 +117,7 @@ def save_data(age, gender, region, *audios):
         shutil.copy(audio_path, save_path)
         filenames.append(filename)
 
+    # 3. Append to CSV
     file_exists = os.path.exists(CSV_FILE)
     with open(CSV_FILE, mode='a', newline='', encoding='utf-8') as f:
         writer = csv.writer(f)
@@ -120,7 +136,6 @@ def transition_to_recorder(age, gender, region):
 # --- UI BUILDER ---
 
 # CSS: System Sans-Serif Stack (Fastest & Cleanest)
-# Uses "Inter" or "San Francisco" on Mac, "Segoe UI" on Windows, "Roboto" on Android.
 system_sans_css = """
 body, .gradio-container {
     font-family: system-ui, -apple-system, "Segoe UI", Roboto, "Helvetica Neue", "Noto Sans", "Liberation Sans", Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol", "Noto Color Emoji" !important;
@@ -131,7 +146,7 @@ button, input, textarea, span, div, label, p, h1, h2, h3 {
 }
 """
 
-with gr.Blocks(title="Kinyarwanda Data Tool", css=system_sans_css, theme=gr.themes.Soft()) as demo:
+with gr.Blocks(title="Improve Kinyarwanda AI", css=system_sans_css, theme=gr.themes.Soft()) as demo:
     gr.Markdown("# 🇷🇼 Improve Kinyarwanda AI")
     
     # PAGE 1: METADATA
@@ -172,12 +187,14 @@ with gr.Blocks(title="Kinyarwanda Data Tool", css=system_sans_css, theme=gr.them
     # --- WIRING ---
     next_btn.click(fn=transition_to_recorder, inputs=[age_input, gender_input, region_input], outputs=[metadata_page, recording_page, error_msg])
     
+    # 1. Save Local
     save_event = submit_btn.click(
         fn=save_data, 
         inputs=[age_input, gender_input, region_input] + audio_inputs, 
         outputs=[status_msg, recording_page, success_page]
     )
 
+    # 2. Auto-Backup to Cloud (Background)
     save_event.then(
         fn=lambda: backup_to_dataset_repo(), 
         inputs=None,