Update app.py

app.py

@@ -13,7 +13,6 @@ from typing import List, Dict, Tuple, Union, Optional
 import requests
 import validators
 import gradio as gr
-from diskcache import Cache
 from bs4 import BeautifulSoup
 from fake_useragent import UserAgent
 from cleantext import clean
@@ -73,6 +72,7 @@ class URLProcessor:
         try:
             if not validators.url(url):
                 return {'is_valid': False, 'message': 'Invalid URL format'}
+
             response = self.session.head(url, timeout=self.timeout)
             response.raise_for_status()
             return {'is_valid': True, 'message': 'URL is valid and accessible'}
@@ -85,9 +85,11 @@ class URLProcessor:
             # Google Drive document handling
             if 'drive.google.com' in url:
                 return self._handle_google_drive(url)
+
             # Google Calendar ICS handling
             if 'calendar.google.com' in url and 'ical' in url:
                 return self._handle_google_calendar(url)
+
             # Standard HTML processing
             return self._fetch_html_content(url)
         except Exception as e:
@@ -96,14 +98,15 @@ class URLProcessor:
 
     def _handle_google_drive(self, url: str) -> Optional[Dict]:
         """Process Google Drive file links"""
-        try:
-            file_id = re.search(r'/file/d/([a-zA-Z0-9_-]+)', url)
+        try: file_id = re.search(r'/file/d/([a-zA-Z0-9_-]+)', url)
             if not file_id:
                 logger.error(f"Invalid Google Drive URL: {url}")
                 return None
+
             direct_url = f"https://drive.google.com/uc?export=download&id={file_id.group(1)}"
             response = self.session.get(direct_url, timeout=self.timeout)
             response.raise_for_status()
+
             return {
                 'content': response.text,
                 'content_type': response.headers.get('Content-Type', ''),
@@ -132,15 +135,20 @@ class URLProcessor:
         try:
             response = self.session.get(url, timeout=self.timeout)
             response.raise_for_status()
+
             soup = BeautifulSoup(response.text, 'html.parser')
+
             # Remove unwanted elements
             for element in soup(['script', 'style', 'nav', 'footer', 'header', 'meta', 'link']):
                 element.decompose()
+
             # Extract main content
             main_content = soup.find('main') or soup.find('article') or soup.body
+
             # Clean and structure content
             text_content = main_content.get_text(separator='\n', strip=True)
             cleaned_content = self.advanced_text_cleaning(text_content)
+
             return {
                 'content': cleaned_content,
                 'content_type': response.headers.get('Content-Type', ''),
@@ -207,14 +215,17 @@ class FileProcessor:
     def _process_single_file(self, file) -> List[Dict]:
         try:
             file_stat = os.stat(file.name)
+
             # For very large files, read in chunks and summarize
             if file_stat.st_size > 100 * 1024 * 1024:  # 100MB
                 logger.info(f"Processing large file: {file.name} ({file_stat.st_size} bytes)")
-                # Read first and last 1MB for extremely large files
+
+                # Read first and last  1MB for extremely large files
                 content = ""
                 with open(file.name, 'r', encoding='utf-8', errors='ignore') as f:
                     content = f.read(1 * 1024 * 1024)  # First 1MB
                     content += "\n...[Content truncated due to large file size]...\n"
+
                     # Seek to the last 1MB
                     f.seek(max(0, file_stat.st_size - 1 * 1024 * 1024))
                     content += f.read()  # Last 1MB
@@ -222,6 +233,7 @@ class FileProcessor:
                 # Regular file processing
                 with open(file.name, 'r', encoding='utf-8', errors='ignore') as f:
                     content = f.read()
+
             return [{
                 'source': 'file',
                 'filename': os.path.basename(file.name),
@@ -236,7 +248,6 @@ class FileProcessor:
             logger.error(f"File processing error: {e}")
             return []
 
-
 def generate_qr(json_data):
     """Generate QR code from JSON data and return the file path."""
     try:
@@ -248,7 +259,7 @@ def generate_qr(json_data):
         )
         qr.add_data(json_data)
         qr.make(fit=True)
-
+        
         img = qr.make_image(fill_color="black", back_color="white")
         temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".png")
         img.save(temp_file.name)
@@ -256,6 +267,7 @@ def generate_qr(json_data):
     except Exception as e:
         # If the data is too large for a QR code
         logger.error(f"QR generation error: {e}")
+        
         # Create a simple QR with error message
         qr = qrcode.QRCode(
             version=1,
@@ -265,211 +277,54 @@ def generate_qr(json_data):
         )
         qr.add_data("Error: Data too large for QR code")
         qr.make(fit=True)
-
+        
         img = qr.make_image(fill_color="black", back_color="white")
         temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".png")
         img.save(temp_file.name)
         return temp_file.name
 
-
 def create_interface():
-    """Create a comprehensive Gradio interface with advanced features and styling"""
+    """Create a comprehensive Gradio interface with advanced features"""
+
     css = """
-    body {
-        font-family: 'Inter', sans-serif;
-        background: linear-gradient(to bottom, #08041C, #030712); /* Dark cosmic background */
-        color: #ffffff;
-    }
-    .container {
-        max-width: 1200px;
-        margin: auto;
-        background-color: rgba(255, 255, 255, 0.06);
-        backdrop-filter: blur(12px);
-        border: 1px solid rgba(255, 255, 255, 0.1);
-        border-radius: 1rem;
-        box-shadow: 0 4px 12px rgba(0, 0, 0, 0.25);
-        padding: 2rem;
-    }
-    h1 {
-        color: #00FF00;
-        text-align: center;
-        text-shadow: 0 0 10px rgba(0, 255, 0, 0.8);
-    }
-    h2, h3, h4 {
-        color: #FF9900;
-        text-shadow: 0 0 10px rgba(255, 153, 0, 0.8);
-    }
-    .tab {
-        background-color: rgba(255, 255, 255, 0.06);
-        backdrop-filter: blur(12px);
-        border: 1px solid rgba(255, 255, 255, 0.1);
-        border-radius: 0.75rem;
-        margin-bottom: 1rem;
-        padding: 1.5rem;
-    }
-    .tab:hover {
-        background-color: rgba(255, 255, 255, 0.1);
-    }
-    .warning {
-        background-color: #fff3cd;
-        color: #856404;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        margin-bottom: 1rem;
-    }
-    .error {
-        background-color: #f8d7da;
-        color: #721c24;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        margin-bottom: 1rem;
-    }
-    input[type="text"], input[type="file"] {
-        width: 100%;
-        padding: 0.75rem;
-        border-radius: 0.5rem;
-        background-color: rgba(0, 0, 0, 0.2);
-        color: #ffffff;
-        border: 1px solid #4a5568;
-        font-size: 1rem;
-        transition: border-color 0.3s ease;
-    }
-    input[type="text"]:focus, input[type="file"]:focus {
-        outline: none;
-        border-color: #00FF00;
-        box-shadow: 0 0 5px rgba(0, 255, 0, 0.7);
-    }
-    .btn-primary {
-        padding: 0.75rem 1.5rem;
-        border-radius: 1.5rem;
-        font-weight: 600;
-        cursor: pointer;
-        transition: transform 0.2s ease, box-shadow 0.2s ease, background-image 0.3s;
-        background-image: linear-gradient(to right, #00FF00, #00A300);
-        color: #000000;
-        border: none;
-        box-shadow: 0 0 8px rgba(0, 255, 0, 0.5);
-    }
-    .btn-primary:hover {
-        transform: scale(1.05);
-        box-shadow: 0 0 12px rgba(0, 255, 0, 0.7);
-        background-image: linear-gradient(to right, #00A300, #007D00);
-    }
-    .btn-secondary {
-        padding: 0.75rem 1.5rem;
-        border-radius: 1.5rem;
-        font-weight: 600;
-        cursor: pointer;
-        transition: transform 0.2s ease, box-shadow 0.2s ease, background-image 0.3s;
-        background-image: linear-gradient(to right, #FF9900, #FF6600);
-        color: #000000;
-        border: none;
-        box-shadow: 0 0 8px rgba(255, 153, 0, 0.5);
-    }
-    .btn-secondary:hover {
-        transform: scale(1.05);
-        box-shadow: 0 0 12px rgba(255, 153, 0, 0.7);
-        background-image: linear-gradient(to right, #FF6600, #CC4700);
-    }
-    textarea {
-        width: 100%;
-        padding: 0.75rem;
-        border-radius: 0.5rem;
-        background-color: rgba(0, 0, 0, 0.2);
-        color: #ffffff;
-        border: 1px solid #4a5568;
-        font-size: 1rem;
-        transition: border-color 0.3s ease;
-        min-height: 8rem;
-    }
-    textarea:focus {
-        outline: none;
-        border-color: #00FF00;
-        box-shadow: 0 0 5px rgba(0, 255, 0, 0.7);
-    }
-    .output-box {
-        background-color: rgba(0, 0, 0, 0.2);
-        border: 1px solid #4a5568;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        overflow-x: auto;
-        color: #ffffff;
-        font-size: 1rem;
-        white-space: pre-wrap;
-    }
-    #json-editor {
-        background-color: rgba(0, 0, 0, 0.2);
-        color: #ffffff;
-        border: 1px solid #4a5568;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        font-size: 1rem;
-        min-height: 20rem;
-    }
-    #json-editor:focus {
-        outline: none;
-        border-color: #00FF00;
-        box-shadow: 0 0 5px rgba(0, 255, 0, 0.7);
-    }
-    #url-input {
-        background-color: rgba(0, 0, 0, 0.2);
-        color: #ffffff;
-        border: 1px solid #4a5568;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        font-size: 1rem;
-        min-height: 8rem;
-    }
-    #url-input:focus {
-        outline: none;
-        border-color: #00FF00;
-        box-shadow: 0 0 5px rgba(0, 255, 0, 0.7);
-    }
-    #output-text{
-        background-color: rgba(0, 0, 0, 0.2);
-        color: #ffffff;
-        border: 1px solid #4a5568;
-        border-radius: 0.5rem;
-        padding: 1rem;
-        font-size: 1rem;
-    }
+    .container { max-width: 1200px; margin: auto; }
+    .warning { background-color: #fff3cd; color: #856404; }
+    .error { background-color: #f8d7da; color: #721c24; }
     """
+
     with gr.Blocks(css=css, title="Advanced Text & URL Processor") as interface:
         gr.Markdown("# 🌐 Advanced URL & Text Processing Toolkit")
 
-        with gr.Tab("URL Processing") as url_tab:
+        with gr.Tab("URL Processing"):
             url_input = gr.Textbox(
                 label="Enter URLs (comma or newline separated)",
                 lines=5,
-                placeholder="https://example1.com\nhttps://example2.com",
-                interactive=True,
-                elem_id="url-input"
+                placeholder="https://example1.com\nhttps://example2.com"
             )
 
-        with gr.Tab("File Input") as file_tab:
+        with gr.Tab("File Input"):
             file_input = gr.File(
                 label="Upload text file or ZIP archive",
                 file_types=[".txt", ".zip", ".md", ".csv", ".json", ".xml"]
             )
 
-        with gr.Tab("Text Input") as text_tab:
+        with gr.Tab("Text Input"):
             text_input = gr.Textbox(
                 label="Raw Text Input",
                 lines=5,
-                placeholder="Paste your text here...",
-                interactive=True
+                placeholder="Paste your text here..."
             )
 
-        with gr.Tab("JSON Editor") as json_tab:
+        with gr.Tab("JSON Editor"):
             json_editor = gr.Textbox(
                 label="JSON Editor",
                 lines=20,
                 placeholder="View and edit your JSON data here...",
                 interactive=True,
-                elem_id="json-editor"
+                elem_id="json-editor"  # Optional: for custom styling
             )
 
-        with gr.Tab("Scratchpad") as scratchpad_tab:
+        with gr.Tab("Scratchpad"):
             scratchpad = gr.Textbox(
                 label="Scratchpad",
                 lines=10,
@@ -480,7 +335,7 @@ def create_interface():
         process_btn = gr.Button("Process Input", variant="primary")
         qr_btn = gr.Button("Generate QR Code", variant="secondary")
 
-        output_text = gr.Textbox(label="Processing Results", interactive=False, elem_id="output-text")
+        output_text = gr.Textbox(label="Processing Results", interactive=False)
         output_file = gr.File(label="Processed Output")
         qr_output = gr.Image(label="QR Code", type="filepath")  # To display the generated QR code
 
@@ -490,10 +345,11 @@ def create_interface():
                 processor = URLProcessor()
                 file_processor = FileProcessor()
                 results = []
-                # Process URLs
-                if urls:
+
+                # Process URLsif urls:
                     url_list = re.split(r'[,\n]', urls)
                     url_list = [url.strip() for url in url_list if url.strip()]
+
                     for url in url_list:
                         validation = processor.validate_url(url)
                         if validation.get('is_valid'):
@@ -505,9 +361,11 @@ def create_interface():
                                     'content': content,
                                     'timestamp': datetime.now().isoformat()
                                 })
+
                 # Process files
                 if file:
                     results.extend(file_processor.process_file(file))
+
                 # Process text input
                 if text:
                     cleaned_text = processor.advanced_text_cleaning(text)
@@ -516,54 +374,62 @@ def create_interface():
                         'content': cleaned_text,
                         'timestamp': datetime.now().isoformat()
                     })
+
                 # Generate output
                 if results:
                     output_dir = Path('output') / datetime.now().strftime('%Y-%m-%d')
                     output_dir.mkdir(parents=True, exist_ok=True)
                     output_path = output_dir / f'processed_{int(time.time())}.json'
+
                     with open(output_path, 'w', encoding='utf-8') as f:
                         json.dump(results, f, ensure_ascii=False, indent=2)
+
                     summary = f"Processed {len(results)} items successfully!"
                     json_data = json.dumps(results, indent=2)  # Prepare JSON for QR code
                     return str(output_path), summary, json_data  # Return JSON for editor
                 else:
                     return None, "No valid content to process.", ""
+
             except Exception as e:
                 logger.error(f"Processing error: {e}")
                 return None, f"Error: {str(e)}", ""
 
         def generate_qr_code(json_data):
-            """Generate QR code from JSON data."""
-            if not json_data:
-                return "No data to encode."
-            qr_file = generate_qr(json_data)
-            return qr_file
+            """Generate QR code from JSON data and return the file path."""
+            if json_data:
+                return generate_qr(json_data)
+            return None
 
         process_btn.click(
             process_all_inputs,
             inputs=[url_input, file_input, text_input, scratchpad],
-            outputs=[output_file, output_text, json_editor]
+            outputs=[output_file, output_text, json_editor]  # Update outputs to include JSON editor
+        )
+
+        qr_btn.click(
+            generate_qr_code,
+            inputs=json_editor,
+            outputs=qr_output
         )
-        qr_btn.click(generate_qr_code, inputs=[json_editor], outputs=[qr_output])
 
         gr.Markdown("""
-        ### Usage Guidelines
-        - **URL Processing**: Enter valid HTTP/HTTPS URLs, separated by commas or newlines.
-        - **File Input**: Upload text files or ZIP archives containing text files.
-        - **Text Input**: Paste text directly for processing.
-        - **JSON Editor**: View the processed data in JSON format. This is automatically updated after processing.
-        - **Scratchpad**: Use this area for temporary notes or text snippets.
-        - Click "Process Input" to analyze the data. The results will be available for download and in the JSON Editor.
-        - Click "Generate QR Code" to create a QR code from the JSON data.
-        """)
+    ### Usage Guidelines
+    - **URL Processing**: Enter valid HTTP/HTTPS URLs
+    - **File Input**: Upload text files or ZIP archives
+    - **Text Input**: Direct text processing
+    - **JSON Editor**: View and edit your JSON data
+    - **Scratchpad**: Quick notes or text collections
+    - Advanced cleaning and validation included
+    """)
     return interface
 
-
 def main():
     # Configure system settings
     mimetypes.init()
+
     # Create and launch interface
     interface = create_interface()
+
     # Launch with proper configuration
     interface.launch(
         server_name="0.0.0.0",
@@ -574,6 +440,5 @@ def main():
         debug=True
     )
 
-
 if __name__ == "__main__":
     main()