tooluniverse 0.2.0__py3-none-any.whl → 1.0.0__py3-none-any.whl

tooluniverse/scripts/generate_tool_graph.py

@@ -0,0 +1,408 @@
+#!/usr/bin/env python3
+"""
+Tool Graph Generator Script
+
+This script generates a compatibility graph for all tools in ToolUniverse.
+It uses the ToolCompatibilityAnalyzer and ToolGraphComposer to analyze tool relationships.
+
+Usage:
+    python generate_tool_graph.py [options]
+
+Examples:
+    # Quick analysis with limited tools
+    python generate_tool_graph.py --quick --max-tools 5
+
+    # Detailed analysis of specific categories
+    python generate_tool_graph.py --categories pubchem,chembl --analysis-depth detailed
+
+    # Full comprehensive analysis (may take hours)
+    python generate_tool_graph.py --analysis-depth comprehensive --force-rebuild
+"""
+
+import argparse
+import json
+import os
+import sys
+import time
+from datetime import datetime
+from pathlib import Path
+
+# Add parent directories to path for imports
+script_dir = Path(__file__).parent
+tooluniverse_dir = script_dir.parent
+src_dir = tooluniverse_dir.parent
+sys.path.insert(0, str(src_dir))
+
+from tooluniverse import ToolUniverse  # noqa: E402
+
+
+def parse_arguments():
+    """Parse command line arguments."""
+
+    parser = argparse.ArgumentParser(
+        description="Generate tool compatibility graph for ToolUniverse",
+        formatter_class=argparse.RawDescriptionHelpFormatter,
+        epilog="""
+Examples:
+  %(prog)s --quick --max-tools 10
+  %(prog)s --categories pubchem,chembl,europepmc
+  %(prog)s --analysis-depth comprehensive --output ./full_graph
+  %(prog)s --force-rebuild --min-score 80
+        """,
+    )
+
+    # Output options
+    parser.add_argument(
+        "--output",
+        "-o",
+        type=str,
+        default="./tool_composition_graph",
+        help="Output path for graph files (default: ./tool_composition_graph)",
+    )
+
+    parser.add_argument(
+        "--force-rebuild",
+        "-f",
+        action="store_true",
+        help="Force rebuild even if cached graph exists",
+    )
+
+    # Analysis options
+    parser.add_argument(
+        "--analysis-depth",
+        "-d",
+        choices=["quick", "detailed", "comprehensive"],
+        default="detailed",
+        help="Analysis depth (default: detailed)",
+    )
+
+    parser.add_argument(
+        "--min-score",
+        "-s",
+        type=int,
+        default=60,
+        help="Minimum compatibility score for edges (0-100, default: 60)",
+    )
+
+    # Tool selection options
+    parser.add_argument(
+        "--categories",
+        "-c",
+        type=str,
+        help="Comma-separated list of tool categories to include (e.g., 'pubchem,chembl')",
+    )
+
+    parser.add_argument(
+        "--exclude-categories",
+        "-e",
+        type=str,
+        default="tool_finder,special_tools",
+        help="Comma-separated list of categories to exclude (default: tool_finder,special_tools)",
+    )
+
+    parser.add_argument(
+        "--max-tools",
+        "-m",
+        type=int,
+        default=50,
+        help="Maximum tools per category (default: 50)",
+    )
+
+    # Convenience options
+    parser.add_argument(
+        "--quick",
+        "-q",
+        action="store_true",
+        help="Quick mode: use quick analysis and limit to 10 tools per category",
+    )
+
+    parser.add_argument("--verbose", "-v", action="store_true", help="Verbose output")
+
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="Show what would be analyzed without actually running",
+    )
+
+    return parser.parse_args()
+
+
+def setup_analysis_parameters(args):
+    """Setup analysis parameters based on arguments."""
+
+    # Handle quick mode
+    if args.quick:
+        analysis_depth = "quick"
+        max_tools = 10
+        min_score = 70  # Higher threshold for quick mode
+    else:
+        analysis_depth = args.analysis_depth
+        max_tools = args.max_tools
+        min_score = args.min_score
+
+    # Handle category selection
+    if args.categories:
+        include_categories = [cat.strip() for cat in args.categories.split(",")]
+        exclude_categories = []
+    else:
+        include_categories = None
+        exclude_categories = [cat.strip() for cat in args.exclude_categories.split(",")]
+
+    return {
+        "output_path": args.output,
+        "analysis_depth": analysis_depth,
+        "min_compatibility_score": min_score,
+        "exclude_categories": exclude_categories,
+        "include_categories": include_categories,
+        "max_tools_per_category": max_tools,
+        "force_rebuild": args.force_rebuild,
+    }
+
+
+def preview_analysis(tooluniverse, parameters):
+    """Show what tools would be analyzed without running the analysis."""
+
+    print("Analysis Preview")
+    print("=" * 50)
+    print(f"Analysis Depth: {parameters['analysis_depth']}")
+    print(f"Min Compatibility Score: {parameters['min_compatibility_score']}")
+    print(f"Max Tools per Category: {parameters['max_tools_per_category']}")
+    print(f"Output Path: {parameters['output_path']}")
+    print()
+
+    # Show categories and tool counts
+    categories = tooluniverse.tool_category_dicts
+    exclude_set = set(parameters.get("exclude_categories", []))
+    include_set = (
+        set(parameters.get("include_categories", []))
+        if parameters.get("include_categories")
+        else None
+    )
+
+    total_tools = 0
+    selected_categories = []
+
+    for category, tools in categories.items():
+        # Apply category filtering
+        if include_set and category not in include_set:
+            continue
+        if category in exclude_set:
+            continue
+
+        tool_count = min(len(tools), parameters["max_tools_per_category"])
+        total_tools += tool_count
+        selected_categories.append((category, tool_count, len(tools)))
+
+    print("Selected Categories:")
+    for category, selected_count, total_count in selected_categories:
+        status = (
+            f"({selected_count}/{total_count})"
+            if selected_count < total_count
+            else f"({total_count})"
+        )
+        print(f"  {category}: {status}")
+
+    print(f"\nTotal tools to analyze: {total_tools}")
+    print(f"Estimated tool pairs: {total_tools * (total_tools - 1)}")
+
+    # Time estimates
+    if parameters["analysis_depth"] == "quick":
+        est_time_per_pair = 2  # seconds
+    elif parameters["analysis_depth"] == "detailed":
+        est_time_per_pair = 5  # seconds
+    else:  # comprehensive
+        est_time_per_pair = 10  # seconds
+
+    total_pairs = total_tools * (total_tools - 1)
+    est_total_time = total_pairs * est_time_per_pair
+
+    print("\nEstimated Analysis Time:")
+    print(f"  Per pair: ~{est_time_per_pair} seconds")
+    print(f"  Total: ~{est_total_time // 3600}h {(est_total_time % 3600) // 60}m")
+
+    return total_tools > 0
+
+
+def run_graph_generation(tooluniverse, parameters, verbose=False):
+    """Run the graph generation process."""
+
+    print("Starting Tool Graph Generation")
+    print("=" * 50)
+
+    start_time = time.time()
+
+    try:
+        # Prepare arguments for ToolGraphComposer
+        composer_args = {
+            "output_path": parameters["output_path"],
+            "analysis_depth": parameters["analysis_depth"],
+            "min_compatibility_score": parameters["min_compatibility_score"],
+            "exclude_categories": parameters["exclude_categories"],
+            "max_tools_per_category": parameters["max_tools_per_category"],
+            "force_rebuild": parameters["force_rebuild"],
+        }
+
+        # Add include_categories if specified
+        if parameters.get("include_categories"):
+            # For include_categories, we need to modify exclude_categories
+            all_categories = set(tooluniverse.tool_category_dicts.keys())
+            include_set = set(parameters["include_categories"])
+            composer_args["exclude_categories"] = list(all_categories - include_set)
+
+        if verbose:
+            print("Composer arguments:")
+            print(json.dumps(composer_args, indent=2))
+            print()
+
+        # Run the composition
+        result = tooluniverse.run(
+            json.dumps([{"name": "ToolGraphComposer", "arguments": composer_args}])
+        )
+
+        # Handle result formatting
+        if isinstance(result, list) and len(result) > 0:
+            result = result[0]
+
+        end_time = time.time()
+        duration = end_time - start_time
+
+        # Display results
+        print("Graph Generation Complete!")
+        print("-" * 30)
+        print(
+            f"Duration: {duration // 3600:.0f}h {(duration % 3600) // 60:.0f}m {duration % 60:.1f}s"
+        )
+
+        if isinstance(result, dict):
+            if result.get("status") == "success":
+                print("✅ Status: Success")
+
+                # Show files created
+                if "graph_files" in result:
+                    print("\nGenerated Files:")
+                    for format_type, file_path in result["graph_files"].items():
+                        if os.path.exists(file_path):
+                            file_size = os.path.getsize(file_path)
+                            print(
+                                f"  {format_type.upper()}: {file_path} ({file_size:,} bytes)"
+                            )
+
+                # Show statistics
+                if "statistics" in result:
+                    stats = result["statistics"]
+                    print("\nGraph Statistics:")
+                    print(f"  Tools Analyzed: {result.get('tools_analyzed', 'N/A')}")
+                    print(f"  Nodes: {stats.get('total_nodes', 'N/A')}")
+                    print(f"  Edges: {stats.get('total_edges', 'N/A')}")
+                    print(f"  Edge Density: {stats.get('edge_density', 0):.3f}")
+
+                    if "compatibility_scores" in stats:
+                        scores = stats["compatibility_scores"]
+                        print(
+                            f"  Avg Compatibility Score: {scores.get('average', 0):.1f}"
+                        )
+                        print(f"  High-Quality Edges: {scores.get('count_high', 0)}")
+
+                    if "automation_ready_percentage" in stats:
+                        print(
+                            f"  Automation Ready: {stats['automation_ready_percentage']:.1f}%"
+                        )
+
+                return True
+            else:
+                print("❌ Status: Failed")
+                print(f"Error: {result.get('message', 'Unknown error')}")
+                return False
+        else:
+            print("❌ Unexpected result format")
+            print(result)
+            return False
+
+    except Exception as e:
+        print(f"❌ Error during graph generation: {e}")
+        if verbose:
+            import traceback
+
+            traceback.print_exc()
+        return False
+
+
+def main():
+    """Main function."""
+
+    args = parse_arguments()
+
+    print("ToolUniverse Graph Generator")
+    print("=" * 60)
+    print(f"Timestamp: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
+    print()
+
+    # Initialize ToolUniverse
+    if args.verbose:
+        print("Initializing ToolUniverse...")
+
+    try:
+        tu = ToolUniverse()
+        tu.load_tools()
+
+        if args.verbose:
+            print(f"Loaded {len(tu.all_tool_dict)} tools")
+            print(f"Categories: {len(tu.tool_category_dicts)}")
+            print()
+
+    except Exception as e:
+        print(f"❌ Failed to initialize ToolUniverse: {e}")
+        return 1
+
+    # Setup parameters
+    parameters = setup_analysis_parameters(args)
+
+    # Show preview
+    if not preview_analysis(tu, parameters):
+        print("❌ No tools selected for analysis. Check your category filters.")
+        return 1
+
+    # Dry run mode
+    if args.dry_run:
+        print(
+            "\n✅ Dry run complete. Use --force-rebuild to actually generate the graph."
+        )
+        return 0
+
+    # Confirmation for large analyses
+    total_categories = len(
+        [
+            cat
+            for cat in tu.tool_category_dicts.keys()
+            if cat not in parameters.get("exclude_categories", [])
+        ]
+    )
+
+    if (
+        not args.quick
+        and total_categories > 5
+        and parameters["max_tools_per_category"] > 20
+    ):
+        response = input("\nThis may take a very long time. Continue? (y/N): ")
+        if response.lower() != "y":
+            print("Cancelled.")
+            return 0
+
+    print("\n" + "=" * 60)
+
+    # Run the generation
+    success = run_graph_generation(tu, parameters, verbose=args.verbose)
+
+    if success:
+        print("\n🎉 Graph generation completed successfully!")
+        print(f"Files saved to: {parameters['output_path']}.*")
+        print("\nTo visualize the graph, run:")
+        print(f"python visualize_tool_graph.py {parameters['output_path']}.json")
+        return 0
+    else:
+        print("\n❌ Graph generation failed.")
+        return 1
+
+
+if __name__ == "__main__":
+    sys.exit(main())