Create app.py

app.py

@@ -0,0 +1,130 @@
+"""Main Gradio application for viewing dataset with text comparison."""
+
+import gradio as gr
+import random
+from data_loader import get_dataset_size, get_sample
+from diff_utils import generate_html_diff, get_diff_stats
+
+
+def load_sample_data(sample_index: int | None = None) -> tuple[str, str, str, str]:
+    """Load and process a sample from the dataset."""
+    if sample_index is None:
+        sample_index = random.randint(0, get_dataset_size() - 1)
+    
+    # Ensure index is within bounds
+    sample_index = max(0, min(sample_index, get_dataset_size() - 1))
+    
+    original_text, cleaned_text = get_sample(sample_index)
+    
+    # Generate diff HTML
+    diff_html = generate_html_diff(original_text, cleaned_text)
+    
+    # Get statistics
+    stats = get_diff_stats(original_text, cleaned_text)
+    stats_text = f"""
+    **Sample #{sample_index}**
+    
+    - Original length: {stats['original_length']:,} characters
+    - Cleaned length: {stats['cleaned_length']:,} characters
+    - Characters removed: {stats['characters_removed']:,}
+    - Similarity: {stats['similarity_ratio']}%
+    """
+    
+    return original_text, cleaned_text, diff_html, stats_text
+
+
+def create_interface() -> gr.Blocks:
+    """Create the main Gradio interface."""
+    with gr.Blocks(
+        title="Dataset Text Comparison Viewer",
+        css="""
+        .textbox-container { 
+            max-height: 400px; 
+            overflow-y: auto; 
+            border: 1px solid #e0e0e0; 
+            border-radius: 8px; 
+            padding: 12px; 
+            background: #fafafa; 
+        }
+        .stats-box {
+            background: #f8f9fa;
+            border: 1px solid #dee2e6;
+            border-radius: 8px;
+            padding: 16px;
+            margin: 8px 0;
+        }
+        """
+    ) as interface:
+        
+        gr.Markdown("# Dataset Text Comparison Viewer")
+        gr.Markdown("Compare original and cleaned text from **sumuks/essential-web-v1.0-sample-1M-with-cleaned-text**")
+        
+        with gr.Row():
+            with gr.Column():
+                sample_input = gr.Number(
+                    label="Sample Index", 
+                    value=0, 
+                    minimum=0, 
+                    maximum=get_dataset_size() - 1,
+                    step=1
+                )
+                
+                with gr.Row():
+                    load_btn = gr.Button("Load Sample", variant="primary")
+                    random_btn = gr.Button("Random Sample", variant="secondary")
+        
+        # Statistics display
+        stats_output = gr.Markdown(label="Statistics", elem_classes=["stats-box"])
+        
+        # Text comparison
+        with gr.Row():
+            with gr.Column():
+                gr.Markdown("### Original Text")
+                original_output = gr.Textbox(
+                    label="Original",
+                    lines=15,
+                    max_lines=20,
+                    show_copy_button=True,
+                    elem_classes=["textbox-container"]
+                )
+            
+            with gr.Column():
+                gr.Markdown("### Cleaned Text")
+                cleaned_output = gr.Textbox(
+                    label="Cleaned",
+                    lines=15,
+                    max_lines=20,
+                    show_copy_button=True,
+                    elem_classes=["textbox-container"]
+                )
+        
+        # Diff visualization
+        gr.Markdown("### Diff Visualization")
+        diff_output = gr.HTML(label="Diff")
+        
+        # Event handlers
+        load_btn.click(
+            fn=load_sample_data,
+            inputs=[sample_input],
+            outputs=[original_output, cleaned_output, diff_output, stats_output]
+        )
+        
+        random_btn.click(
+            fn=lambda: load_sample_data(None),
+            inputs=[],
+            outputs=[original_output, cleaned_output, diff_output, stats_output]
+        )
+        
+        # Load initial sample
+        interface.load(
+            fn=lambda: load_sample_data(0),
+            inputs=[],
+            outputs=[original_output, cleaned_output, diff_output, stats_output]
+        )
+    
+    return interface
+
+
+if __name__ == "__main__":
+    app = create_interface()
+    app.launch(share=False, debug=True)