discover_tools.py

#!/usr/bin/env python3
import json
from pathlib import Path
import sys

from utils.config_loader import load_config
from utils.conversation_parser import find_log_files

def discover_tools(log_files):
    """Scans log files to discover all unique tool names used."""
    unique_tool_names = set()
    total_files = len(log_files)
    print(f"Found {total_files} log files to scan.")

    for i, log_file in enumerate(log_files):
        if (i + 1) % 10 == 0 or i == total_files - 1:
            print(f"Scanning file {i + 1}/{total_files}...")
        
        with open(log_file, 'r', encoding='utf-8', errors='ignore') as f:
            for line in f:
                try:
                    turn = json.loads(line)
                    if turn.get('type') == 'assistant':
                        content_list = turn.get('message', {}).get('content', [])
                        if isinstance(content_list, list):
                            for item in content_list:
                                if item.get('type') == 'tool_use':
                                    tool_name = item.get('name')
                                    if tool_name:
                                        unique_tool_names.add(tool_name)
                except (json.JSONDecodeError, AttributeError):
                    # Ignore lines that are not valid JSON or don't have the expected structure
                    continue
    
    return unique_tool_names

def main():
    """Main function to run the discovery process."""
    config = load_config()
    io_settings = config.get('io_settings', {})
    input_dirs_conf = io_settings.get('input_log_directories')

    if not input_dirs_conf:
        print("Error: 'input_log_directories' not specified in config.yml.")
        sys.exit(1)

    input_dirs = [input_dirs_conf] if isinstance(input_dirs_conf, str) else input_dirs_conf
    exclude_patterns = io_settings.get('exclude_patterns', [])
    exclude_files = io_settings.get('exclude_files', [])

    print(f"Searching for .jsonl log files in {len(input_dirs)} directory(ies)...")
    log_files = find_log_files(input_dirs, exclude_patterns, exclude_files)
    if not log_files:
        print("No log files found in the specified directory.")
        return

    discovered_tools = sorted(list(discover_tools(log_files)))
    
    print("\n--- Discovery Complete ---")
    print(f"Total .jsonl files found: {len(log_files)}")
    print("Unique tool names discovered across all logs:")
    for tool in discovered_tools:
        print(f"- {tool}")
    print("--------------------------\n")
    print("Please review this list and update 'tool_mapping.yml' to ensure all tools are accounted for.")

if __name__ == "__main__":
    main()