Update app.py

app.py

@@ -35,33 +35,97 @@ from smolagents.agent_types import AgentText, AgentImage, AgentAudio
 from smolagents.gradio_ui import pull_messages_from_step, handle_agent_output_types
 
 
+# Constants and configurations
+"""Main application for the OpenDeepResearch Gradio interface."""
+
+import mimetypes
+import os
+import re
+import shutil
+from typing import Optional
+
+from dotenv import load_dotenv
+from huggingface_hub import login
+import gradio as gr
+
+from scripts.text_inspector_tool import TextInspectorTool
+from scripts.text_web_browser import (
+    ArchiveSearchTool,
+    FinderTool,
+    FindNextTool,
+    PageDownTool,
+    PageUpTool,
+    SimpleTextBrowser,
+    VisitTool,
+)
+from scripts.visual_qa import visualizer
+from scripts.text_cleaner_tool import TextCleanerTool
+
+from smolagents import (
+    CodeAgent,
+    HfApiModel,
+    LiteLLMModel,
+    OpenAIServerModel,
+    TransformersModel,
+    GoogleSearchTool,
+    Tool,
+)
+from smolagents.agent_types import AgentText, AgentImage, AgentAudio
+from smolagents.gradio_ui import pull_messages_from_step, handle_agent_output_types
+
+
 # Constants and configurations
 AUTHORIZED_IMPORTS = [
-    "requests",
-    "zipfile",
-    "pandas",
-    "numpy",
-    "sympy",
-    "json",
-    "bs4",
-    "pubchempy",
-    "xml",
-    "yahoo_finance",
-    "Bio",
-    "sklearn",
-    "scipy",
-    "pydub",
-    "PIL",
-    "chess",
-    "PyPDF2",
-    "pptx",
-    "torch",
-    "datetime",
-    "fractions",
-    "csv",
-    "clean-text",
+    "requests",  # Web requests (fetching data from the internet)
+    "zipfile",  # Working with ZIP archives
+    "pandas",  # Data manipulation and analysis (DataFrames)
+    "numpy",  # Numerical computing (arrays, linear algebra)
+    "sympy",  # Symbolic mathematics (algebra, calculus)
+    "json",  # JSON data serialization/deserialization
+    "bs4",  # Beautiful Soup for HTML/XML parsing
+    "pubchempy",  # Accessing PubChem chemical database
+    "xml",  # XML processing
+    "yahoo_finance",  # Fetching stock data
+    "Bio",  # Bioinformatics tools (e.g., sequence analysis)
+    "sklearn",  # Scikit-learn for machine learning
+    "scipy",  # Scientific computing (stats, optimization)
+    "pydub",  # Audio manipulation
+    "PIL",  # Pillow for image processing
+    "chess",  # Chess-related functionality
+    "PyPDF2",  # PDF manipulation
+    "pptx",  # PowerPoint file manipulation
+    "torch",  # PyTorch for neural networks
+    "datetime",  # Date and time handling
+    "fractions",  # Rational number arithmetic
+    "csv",  # CSV file reading/writing
+    "cleantext",  # Text cleaning and normalization
+    "os",  # Operating system interaction (file system, etc.) VERY IMPORTANT
+    "re",  # Regular expressions for text processing
+    "collections",  # Useful data structures (e.g., defaultdict, Counter)
+    "math",  # Basic mathematical functions
+    "random",  # Random number generation
+    "io",  # Input/output streams
+    "urllib.parse",  # URL parsing and manipulation (safe URL handling)
+    "typing",  # Support for type hints (improve code clarity)
+    "concurrent.futures",  # For parallel execution
+    "time",  # Measuring time
+    "tempfile",  # Creating temporary files and directories
+    # Data Visualization (if needed) - Consider security implications carefully
+    "matplotlib",  # Plotting library (basic charts) - Consider security implications
+    "seaborn",  # Statistical data visualization (more advanced) - Consider security implications
+    # Web Scraping (more specific/controlled) - Consider ethical implications carefully
+    "lxml",  # Faster XML/HTML processing (alternative to bs4)
+    "selenium",  # Automated browser control (for dynamic websites) - HIGH SECURITY RISK
+    # Database interaction (if needed) - Handle credentials securely!
+    "sqlite3",  # SQLite database access
+    # "psycopg2",         # PostgreSQL adapter if needed
+    # Task scheduling
+    "schedule",  # Allow the agent to schedule tasks
+    # Networking
+    # "socket",          # Networking
 ]
 
+
 USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/119.0.0.0 Safari/537.36 Edg/119.0.0.0"
 BROWSER_CONFIG = {
     "viewport_size": 1024 * 5,
@@ -170,6 +234,15 @@ class ToolRegistry:
             print(f"✗ Couldn't initialize image generation tool: {e}")
             raise
 
+    @staticmethod
+    def load_clean_text_tool():
+        """Initialize and return image generation tools."""
+        try:
+            return TextCleanerTool
+        except Exception as e:
+            print(f"✗ Couldn't initialize clean text tool: {e}")
+            raise
+
 
 def create_agent():
     """Creates a fresh agent instance with properly configured tools."""