Update app.py

app.py

@@ -72,6 +72,8 @@ if 'should_rerun' not in st.session_state:
     st.session_state['should_rerun'] = False
 if 'old_val' not in st.session_state:
     st.session_state['old_val'] = None
+if 'audio_generated' not in st.session_state:
+    st.session_state['audio_generated'] = {}
 
 # 🎨 4. Custom CSS
 st.markdown("""
@@ -179,7 +181,7 @@ def create_file(prompt, response, file_type="md"):
     """Create file with intelligent naming"""
     filename = generate_filename(prompt.strip(), response.strip(), file_type)
     with open(filename, 'w', encoding='utf-8') as f:
-        f.write(prompt + "\n\n" + response)
+        f.write(prompt.replace('\n', ' ') + "\n\n" + response)
     return filename
 
 def get_download_link(file):
@@ -347,9 +349,10 @@ def parse_arxiv_refs(ref_text: str):
             'summary': summary,
             'year': year
         })
+        if len(results) >= 20:
+            break
     return results
 
-
 def perform_ai_lookup(q, vocal_summary=True, extended_refs=False, 
                       titles_summary=True, full_audio=False, selected_voice="en-US-AriaNeural"):
     """Perform Arxiv search and generate audio summaries."""
@@ -361,12 +364,12 @@ def perform_ai_lookup(q, vocal_summary=True, extended_refs=False,
     r2 = client.predict(q,"mistralai/Mixtral-8x7B-Instruct-v0.1",True,api_name="/ask_llm")
 
     # 🎯 2) Combine for final text output
-    result = f"### 🔎 {q}\n\n{r2}\n\n{refs}"
+    result = f"### 🔎 {q.replace('\n', ' ')}\n\n{r2}\n\n{refs}"
     st.markdown(result)
 
     # 🎯 3) Generate "all at once" audio if requested
     if full_audio:
-        complete_text = f"Complete response for query: {q}. {clean_for_speech(r2)} {clean_for_speech(refs)}"
+        complete_text = f"Complete response for query: {q.replace('\n', ' ')}. {clean_for_speech(r2)} {clean_for_speech(refs)}"
         audio_file_full = speak_with_edge_tts(complete_text, selected_voice)
         st.write("### 📚 Full Audio")
         play_and_download_audio(audio_file_full)
@@ -389,35 +392,32 @@ def perform_ai_lookup(q, vocal_summary=True, extended_refs=False,
     # --------------------------------------
     parsed_refs = parse_arxiv_refs(refs)
 
-    # Sort by year descending (put None at bottom)
-    parsed_refs.sort(key=lambda x: x["year"] if x["year"] else 0, reverse=True)
-
     st.write("## Individual Papers (Most Recent First)")
     for idx, paper in enumerate(parsed_refs):
-        year_str = paper["year"] if paper["year"] else "Unknown Year"
-        st.markdown(f"**{idx+1}. {paper['title']}**  \n*Year:* {year_str}")
-        st.markdown(f"*Summary:* {paper['summary']}")
-
-        # Two new TTS buttons: Title only or Title+Summary
-        colA, colB = st.columns(2)
-        with colA:
-            if st.button(f"🔊 Title", key=f"title_{idx}"):
-                text_tts = clean_for_speech(paper['title'])
-                audio_file_title = speak_with_edge_tts(text_tts, selected_voice)
-                play_and_download_audio(audio_file_title)
-
-        with colB:
-            if st.button(f"🔊 Title+Summary", key=f"summary_{idx}"):
-                text_tts = clean_for_speech(paper['title'] + ". " + paper['summary'])
-                audio_file_title_summary = speak_with_edge_tts(text_tts, selected_voice)
-                play_and_download_audio(audio_file_title_summary)
-
-        st.write("---")
+        section_key = f"section_{idx}"
+        with st.expander(f"{idx+1}. {paper['title']} - {paper['year'] if paper['year'] else 'Unknown Year'}", expanded=False, key=section_key):
+            st.markdown(f"**Summary:** {paper['summary']}")
+            colA, colB = st.columns(2)
+            with colA:
+                if st.checkbox(f"Generate Audio for Title", key=f"gen_title_{idx}"):
+                    if f"title_audio_{idx}" not in st.session_state['audio_generated']:
+                        text_tts = clean_for_speech(paper['title'])
+                        audio_file_title = speak_with_edge_tts(text_tts, selected_voice)
+                        st.session_state['audio_generated'][f"title_audio_{idx}"] = audio_file_title
+                    play_and_download_audio(st.session_state['audio_generated'].get(f"title_audio_{idx}"))
+            with colB:
+                if st.checkbox(f"Generate Audio for Title + Summary", key=f"gen_summary_{idx}"):
+                    if f"summary_audio_{idx}" not in st.session_state['audio_generated']:
+                        text_tts = clean_for_speech(paper['title'] + ". " + paper['summary'])
+                        audio_file_title_summary = speak_with_edge_tts(text_tts, selected_voice)
+                        st.session_state['audio_generated'][f"summary_audio_{idx}"] = audio_file_title_summary
+                    play_and_download_audio(st.session_state['audio_generated'].get(f"summary_audio_{idx}"))
+            st.write("---")
 
     # Keep your original block for "Titles Only" if you want:
     if titles_summary:
         titles = []
-        for line in refs.split('\n'):
+        for line in refs.split('\n')[:20]:
             m = re.search(r"\[([^\]]+)\]", line)
             if m:
                 titles.append(m.group(1))
@@ -432,7 +432,7 @@ def perform_ai_lookup(q, vocal_summary=True, extended_refs=False,
     st.write(f"**Total Elapsed:** {elapsed:.2f} s")
 
     # Always create a file with the result
-    create_file(q, result, "md")
+    create_file(q.replace('\n', ' '), result, "md")
 
     return result
 
@@ -588,7 +588,7 @@ def display_file_manager_sidebar(groups, sorted_prefixes):
 # 🎯 11. Main Application
 async def get_available_voices():
     voices = await edge_tts.list_voices()
-    return [voice["ShortName"] for voice in voices]
+    return [voice["ShortName"] for voice in voices if voice["Locale"].startswith("en")]
 
 @st.cache_resource
 def fetch_voices():
@@ -661,7 +661,7 @@ def main():
 
     if tab_main == "🔍 ArXiv":
         st.subheader("🔍 Query ArXiv")
-        q = st.text_input("🔍 Query:")
+        q = st.text_input("🔍 Query:").replace('\n', ' ')
 
         st.markdown("### 🎛 Options")
         vocal_summary = st.checkbox("🎙ShortAudio", value=True)
@@ -679,7 +679,7 @@ def main():
                 save_full_transcript(q, result)
 
         st.markdown("### Change Prompt & Re-Run")
-        q_new = st.text_input("🔄 Modify Query:")
+        q_new = st.text_input("🔄 Modify Query:").replace('\n', ' ')
         if q_new and st.button("🔄 Re-Run with Modified Query"):
             result = perform_ai_lookup(q_new, vocal_summary=vocal_summary, extended_refs=extended_refs, 
                                        titles_summary=titles_summary, full_audio=full_audio, selected_voice=selected_voice)
@@ -711,7 +711,7 @@ def main():
             if imgs:
                 c = st.slider("Cols",1,5,3)
                 cols = st.columns(c)
-                for i,f in enumerate(imgs):
+                for i,f in enumerate(imgs[:20]):
                     with cols[i%c]:
                         st.image(Image.open(f),use_container_width=True)
                         if st.button(f"👀 Analyze {os.path.basename(f)}", key=f"analyze_{f}"):
@@ -720,7 +720,7 @@ def main():
             else:
                 st.write("No images found.")
         with tabs[1]:
-            vids = glob.glob("*.mp4")
+            vids = glob.glob("*.mp4")[:20]
             if vids:
                 for v in vids:
                     with st.expander(f"🎥 {os.path.basename(v)}"):