Update app.py

app.py

@@ -40,7 +40,7 @@ STOPWORDS = {
 POS_HINTS = {"good","great","love","like","enjoy","awesome","amazing","nice","positive","best","fantastic","excellent"}
 NEG_HINTS = {"bad","hate","dislike","worst","awful","terrible","negative","poor","meh","gross","unsafe","hard","difficult"}
 
-punct_re = re.compile(f"[{re.escape('!"#$%&\'()*+,-./:;<=>?@[\\]^_`{|}~')}]")
+punct_re = re.compile(r"[{}]".format(re.escape("""!"#$%&'()*+,-./:;<=>?@[\]^_`{|}~""")))
 
 def normalize(text: str) -> str:
     return punct_re.sub(" ", (text or "").lower())
@@ -62,50 +62,56 @@ def infer_sentiment(user_text: str) -> str:
 # Retrieval
 # -----------------------------
 def best_match_quote(user_text: str) -> str:
-    """Search across all categories with fuzzy matching and return best quote."""
-    max_score = 0
+    """Search all categories, return best fuzzy match, fallback to random from any category."""
+    best_score = 0
     best_quote = None
 
-    for cat_quotes in QUOTES.values():
-        for q_obj in cat_quotes:
-            q_text = q_obj.get("quote", "")
-            score = fuzz.token_set_ratio(user_text.lower(), q_text.lower())
-            if score > max_score:
-                max_score = score
-                best_quote = q_text
-
-    if max_score < 30:  # threshold; anything below treated as unknown
+    for cat, quotes_list in QUOTES.items():
+        for quote_entry in quotes_list:
+            quote = quote_entry.get("quote", "")
+            if not quote.strip():
+                continue
+            score = fuzz.partial_ratio(user_text.lower(), quote.lower())
+            if score > best_score:
+                best_score = score
+                best_quote = quote
+
+    # Threshold for unknown
+    if best_score < 30 or best_quote is None:
         return f"No data about '{user_text}'"
     return best_quote
 
-# -----------------------------
-# 3-fold response generation
-# -----------------------------
-def generate_three_fold_response(user_text: str):
-    quote = best_match_quote(user_text)
-    if quote.startswith("No data"):
-        return [quote, "", ""]
-
-    # Very simple 3-fold split
-    sentences = [s.strip() for s in quote.split('.') if s.strip()]
-    first = sentences[0] if len(sentences) > 0 else ""
-    second = " ".join(sentences[1:3]) if len(sentences) > 2 else (sentences[1] if len(sentences)>1 else "")
-    third = " ".join(sentences[3:]) if len(sentences) > 3 else ""
-    return [first, second, third]
-
 # -----------------------------
 # Gradio callbacks
 # -----------------------------
 def respond(message, history, category):
-    if not message:
+    if not QUOTES:
+        bot = "No dataset loaded. Please upload a JSON file first."
+        history.append({"role": "user", "content": message})
+        history.append({"role": "assistant", "content": bot})
+        return "", history
+
+    if not category:
+        bot = "Please select a category."
+        history.append({"role": "user", "content": message})
+        history.append({"role": "assistant", "content": bot})
         return "", history
-    responses = generate_three_fold_response(message)
-    bot_response = [
-        {"label": "Summary", "text": responses[0]},
-        {"label": "Details", "text": responses[1]},
-        {"label": "What real people say", "text": responses[2]}
-    ]
-    history.append((message, bot_response))
+
+    quote = best_match_quote(message)
+
+    # 3-fold response
+    summary = quote.split(". ")[0] + "." if "." in quote else quote
+    detail = quote
+    unknown = ""
+    if "No data about" in quote:
+        unknown = quote
+
+    bot_text = f"Summary:\n{summary}\n\nWhat real people say:\n{detail}"
+    if unknown:
+        bot_text += f"\n\n{unknown}"
+
+    history.append({"role": "user", "content": message})
+    history.append({"role": "assistant", "content": bot_text})
     return "", history
 
 def clear_chat():
@@ -127,34 +133,26 @@ def upload_json(filepath):
         return f"Error loading file: {e}", gr.update(choices=[])
 
 def download_current_json():
-    tmp = DATA_PATH or "quotes_export.json"
-    with open(tmp, "w", encoding="utf-8") as f:
+    tmp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".json")
+    with open(tmp_file.name, "w", encoding="utf-8") as f:
         json.dump(QUOTES, f, indent=2, ensure_ascii=False)
-    return tmp
+    return tmp_file.name
 
 def download_conversation_csv(history):
-    if not history:
-        tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".csv")
-        tmp.close()
-        return tmp.name
-    tmp = tempfile.NamedTemporaryFile(delete=False, suffix=".csv", mode='w', newline='', encoding='utf-8')
-    writer = csv.writer(tmp)
-    writer.writerow(["User Message", "Summary", "Details", "What real people say"])
-    for msg, bot_resp in history:
-        summary = bot_resp[0]['text']
-        details = bot_resp[1]['text']
-        real_people = bot_resp[2]['text']
-        writer.writerow([msg, summary, details, real_people])
-    tmp.close()
-    return tmp.name
+    tmp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".csv")
+    with open(tmp_file.name, "w", newline="", encoding="utf-8") as f:
+        writer = csv.writer(f)
+        writer.writerow(["role", "message"])
+        for msg in history:
+            writer.writerow([msg.get("role"), msg.get("content")])
+    return tmp_file.name
 
 # -----------------------------
 # UI
 # -----------------------------
 with gr.Blocks() as demo:
-    gr.Markdown("## 🎓 College Life Chatbot — Category-Aware, 3-Fold Responses")
+    gr.Markdown("## 🎓 College Life Chatbot — Category-Aware, Fuzzy Matching")
 
-    # Category list from loaded data (may be empty until upload)
     initial_categories = sorted(list(QUOTES.keys()))
 
     with gr.Row():
@@ -170,27 +168,27 @@ with gr.Blocks() as demo:
     clear = gr.Button("Clear")
 
     with gr.Row():
-        uploader = gr.File(label="Upload dataset (.json)", file_types=[".json"], type="filepath")
+        upload_btn = gr.File(label="Upload dataset (.json)", file_types=[".json"], type="filepath")
         upload_status = gr.Textbox(label="Upload status", interactive=False)
-        download_json_btn = gr.File(label="Download dataset")
+        download_json_btn = gr.File(label="Download current dataset (.json)")
         download_csv_btn = gr.File(label="Export conversation to CSV")
 
-    # Wire events
+    # Events
     msg.submit(respond, [msg, chatbot, category], [msg, chatbot])
     send.click(respond, [msg, chatbot, category], [msg, chatbot])
     clear.click(clear_chat, None, chatbot, queue=False)
 
-    uploader.upload(upload_json, uploader, [upload_status, category])
+    upload_btn.upload(upload_json, upload_btn, [upload_status, category])
     download_json_btn.download(download_current_json)
-    download_csv_btn.click(download_conversation_csv, chatbot, download_csv_btn)
+    download_csv_btn.download(lambda: download_conversation_csv(chatbot.value))
 
 # -----------------------------
 # Startup log
 # -----------------------------
 print(f"===== Application Startup at {datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')} =====")
 if QUOTES:
-    for cat, qlist in QUOTES.items():
-        print(f" - {cat}: {len(qlist)} entries")
+    for cat, entries in QUOTES.items():
+        print(f" - {cat}: {len(entries)} entries")
 
 if __name__ == "__main__":
     demo.launch(server_name="0.0.0.0", server_port=7860)