Update app.py

app.py

@@ -17,7 +17,7 @@ urdu_model = pipeline(
 
 roman_urdu_model = pipeline(
     "sentiment-analysis",
-    model="mrgmd01/sentiment_model_FineTune_cardiffnlp"  # <- replace with your roman urdu model
+    model="mrgmd01/sentiment_model_FineTune_cardiffnlp"  # Replace with your Roman Urdu model if available
 )
 
 # File to store only sentences
@@ -33,7 +33,6 @@ def detect_language(text):
     urdu_chars = set("ابتثجحخدذرزسشصضطظعغفقکلمنوہیءآؤئۀ")
     if any(ch in urdu_chars for ch in text):
         return "Urdu"
-    # Heuristic: Roman Urdu often has "hai, kia, kaisa, bohot, acha" etc.
     roman_urdu_pattern = r"\b(hai|kia|kyun|nahi|bohot|acha|galat|sahi|parhai|ustad)\b"
     if re.search(roman_urdu_pattern, text.lower()):
         return "Roman Urdu"
@@ -49,48 +48,56 @@ def normalize_label(label):
     else:
         return "Neutral"
 
-# Prediction + Save sentence
-def predict_and_save(text):
+# Prediction function
+def analyze_single(text, lang_hint):
     if not text.strip():
-        return "Please enter a sentence.", SAVE_FILE
+        return "Please enter a sentence.", "", "", SAVE_FILE
 
-    lang = detect_language(text)
+    # If user gives hint, use it; else auto-detect
+    if lang_hint and lang_hint != "Auto":
+        lang = lang_hint
+    else:
+        lang = detect_language(text)
 
     if lang == "English":
         result = english_model(text)[0]
     elif lang == "Urdu":
         result = urdu_model(text)[0]
-    else:  # Roman Urdu
+    else:
         result = roman_urdu_model(text)[0]
 
     sentiment = normalize_label(result["label"])
     score = round(result["score"], 3)
+    polarity = "Positive" if sentiment == "Positive" else ("Negative" if sentiment == "Negative" else "Neutral")
 
-    # Save only the sentence
+    # Save only sentence
     df = pd.read_csv(SAVE_FILE)
     new_row = pd.DataFrame([[text]], columns=["Sentence"])
     df = pd.concat([df, new_row], ignore_index=True)
     df.to_csv(SAVE_FILE, index=False)
 
-    return f"Sentiment: {sentiment} (Confidence: {score}, Language: {lang})", SAVE_FILE
+    return sentiment, str(score), polarity, SAVE_FILE
+
 
 # Gradio UI
 with gr.Blocks() as demo:
-   "## 🌍 Multilingual Sentiment Analysis (Positive • Neutral • Negative)\n"
-        "**Languages:** English, Urdu, Roman Urdu  \n"
-        "Model: `siebert/sentiment-roberta-large-english (For English)`\n `mrgmd01/sentiment_model_FineTune_cardiffnlp(For Urdu & Roman Urdu)`"
+    gr.Markdown("## 🌍 Multilingual Sentiment Analysis (Positive • Neutral • Negative)")
+    gr.Markdown("**Languages:** English, Urdu, Roman Urdu  \n"
+                "Model: `cardiffnlp/twitter-roberta-base-sentiment-latest (English)`  \n"
+                "`mrgmd01/sentiment_model_FineTune_cardiffnlp (Urdu & Roman Urdu)`")
 
- with gr.Tab():
+    with gr.Tab("Sentiment Analysis"):
         user_text = gr.Textbox(label="Enter text", placeholder="Type in English, Urdu, or Roman Urdu...")
-        lang_dropdown = gr.Dropdown(["English", "Urdu", "Roman Urdu"], label="Language Hint", value="English")
+        lang_dropdown = gr.Dropdown(["Auto", "English", "Urdu", "Roman Urdu"], label="Language Hint", value="Auto")
         btn = gr.Button("Analyze")
 
         out_sent = gr.Textbox(label="Sentiment")
         out_conf = gr.Textbox(label="Confidence (0–1)")
         out_pol  = gr.Textbox(label="Polarity")
-        out_file = gr.File(label="Download logs (.xlsx)")
+        out_file = gr.File(label="Download logs (.csv)", type="filepath")
 
         btn.click(analyze_single, inputs=[user_text, lang_dropdown],
                   outputs=[out_sent, out_conf, out_pol, out_file])
+
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch()