Update app.py

app.py

@@ -1,47 +1,85 @@
 import gradio as gr
+from transformers import pipeline
+from wordcloud import WordCloud
+import matplotlib
+matplotlib.use("Agg")  # fix for HF Spaces
+import matplotlib.pyplot as plt
+import io
 import pandas as pd
-from pipeline import process_comment
 
-def analyze_single(text):
-    sentiment, summary, wc_img = process_comment(text)
-    result = f"**Label:** {sentiment['label']} | **Score:** {sentiment['score']:.2f}"
-    return result, summary, wc_img
+# Load Hugging Face pipelines with smaller stable models
+sentiment_model = pipeline(
+    "sentiment-analysis",
+    model="distilbert-base-uncased-finetuned-sst-2-english"
+)
+summarizer = pipeline(
+    "summarization",
+    model="sshleifer/distilbart-cnn-12-6"
+)
 
-def analyze_batch(file):
+# Word cloud generator
+def generate_wordcloud(text):
+    wordcloud = WordCloud(width=800, height=400, background_color="white").generate(text)
+    img = io.BytesIO()
+    plt.figure(figsize=(8, 4))
+    plt.imshow(wordcloud, interpolation="bilinear")
+    plt.axis("off")
+    plt.savefig(img, format="png")
+    plt.close()
+    img.seek(0)
+    return img
+
+# Process single comment
+def process_comment(comment):
+    # Sentiment
+    sentiment = sentiment_model(comment)[0]
+
+    # Summary
+    try:
+        summary = summarizer(comment, max_length=60, min_length=10, do_sample=False)[0]['summary_text']
+    except Exception:
+        summary = "Summary not available for short text."
+
+    # Word cloud
+    wc_img = generate_wordcloud(comment)
+
+    return f"{sentiment['label']} ({sentiment['score']:.2f})", summary, wc_img
+
+# Batch processing from CSV
+def analyze_batch(file_path):
     df = pd.read_csv(file_path)
-    results = []
-    summaries = []
+    results, summaries = [], []
     combined_text = ""
 
-    for comment in df["comment"]:
+    for comment in df["comment"].astype(str):
         sentiment, summary, _ = process_comment(comment)
-        results.append(f"{sentiment['label']} ({sentiment['score']:.2f})")
+        results.append(sentiment)
         summaries.append(summary)
         combined_text += " " + comment
 
-    # Generate one big word cloud for all comments
-    _, _, wc_img = process_comment(combined_text)
+    # One word cloud for all comments
+    wc_img = generate_wordcloud(combined_text)
 
     df["Sentiment"] = results
     df["Summary"] = summaries
-
     return df, wc_img
 
+# Gradio UI
 with gr.Blocks() as demo:
-    gr.Markdown("# 🏛️ E-Consultation Sentiment Analysis\nAnalyze stakeholder comments with AI.")
-    
+    gr.Markdown("## 📝 E-Consultation Sentiment Analysis")
+
     with gr.Tab("Single Comment"):
-        inp = gr.Textbox(lines=5, placeholder="Enter stakeholder comment...")
+        inp = gr.Textbox(label="Enter Comment")
         out1 = gr.Textbox(label="Sentiment")
         out2 = gr.Textbox(label="Summary")
         out3 = gr.Image(label="Word Cloud")
-        btn = gr.Button("Analyze")
-        btn.click(analyze_single, inputs=inp, outputs=[out1, out2, out3])
+        inp.submit(process_comment, inp, [out1, out2, out3])
 
-    with gr.Tab("Batch Upload (CSV)"):
+    with gr.Tab("Batch Upload"):
         file_in = gr.File(label="Upload CSV with 'comment' column", type="filepath")
-        df_out = gr.Dataframe(label="Analysis Results")
-        wc_out = gr.Image(label="Word Cloud")
-        file_in.change(analyze_batch, inputs=file_in, outputs=[df_out, wc_out])
+        out_df = gr.Dataframe(label="Results")
+        out_wc = gr.Image(label="Word Cloud (All Comments)")
+        file_in.change(analyze_batch, file_in, [out_df, out_wc])
 
-demo.launch()
+if __name__ == "__main__":
+    demo.launch()