analyze_jplag.py

#!/usr/bin/env python3

import os
import json
import time
import sys
import zipfile
from pathlib import Path

# Set the similarity threshold here
SIMILARITY_THRESHOLD = 70

def die(msg):
    print(msg, file=sys.stderr)
    sys.exit(1)

def get_jplag_jar():
    """Find JPlag JAR file in third_party directory next to this script"""
    script_dir = Path(__file__).parent
    jplag_jar = script_dir / "third_party" / "jplag.jar"
    if not jplag_jar.exists():
        die(f"Error: JPlag JAR not found at {jplag_jar}")
    return jplag_jar

def get_latest_results_dir():
    """Get the latest subdirectory under jplag_results"""
    results_dir = Path("jplag_results")
    if not results_dir.exists():
        die("Error: jplag_results directory does not exist")
    
    subdirs = [d for d in results_dir.iterdir() if d.is_dir()]
    if not subdirs:
        die("Error: no JPlag results found in jplag_results")
    
    # Sort by modification time, newest first
    subdirs.sort(key=lambda d: d.stat().st_mtime, reverse=True)
    return subdirs[0]

def analyze_jplag_results(results_dir):
    """Analyze JPlag results and suggest which reports to investigate"""
    csv_start_time = time.time()
    
    # Find all ZIP files in the results directory
    zip_files = list(results_dir.glob("*.zip"))
    if not zip_files:
        die(f"Error: no JPlag results found in {results_dir}")
    
    for zip_file in zip_files:
        lang_prob = zip_file.stem
        lang, prob = lang_prob.split('-', 1)
        
        try:
            with zipfile.ZipFile(zip_file) as zf:
                with zf.open("overview.json") as f:
                    data = json.load(f)
                    
                    # Find high similarity cases
                    high_similarity = []
                    for comparison in data['top_comparisons']:
                        similarity = comparison['similarities']['MAX'] * 100
                        if similarity >= SIMILARITY_THRESHOLD:
                            high_similarity.append((
                                comparison['first_submission'],
                                comparison['second_submission'],
                                similarity
                            ))
                    
                    if high_similarity:
                        cases_string = 'case' if len(high_similarity) == 1 else 'cases'
                        print(f"\n{lang}/{prob} ({len(high_similarity)} {cases_string}):")
                        for sub1, sub2, sim in high_similarity:
                            print(f"  • {sub1} vs {sub2}: {sim:.2f}%")
                        
                        # Suggest command to view report
                        print(f"\n  View report:\n    java -jar {get_jplag_jar()} {zip_file}")
        
        except Exception as e:
            print(f"Error processing {zip_file}: {e}")
    
    csv_time = time.time() - csv_start_time
    print(f"\nAnalysis completed in {csv_time:.1f}s")

def main():
    if len(sys.argv) > 2:
        print(f"Usage: {sys.argv[0]} [results_dir]")
        print("Example: python analyze_jplag.py jplag_results/b46e698a7")
        sys.exit(1)
    
    # Use provided directory or get latest
    results_dir = Path(sys.argv[1]) if len(sys.argv) == 2 else get_latest_results_dir()
    
    if not results_dir.exists():
        die(f"Error: results directory '{results_dir}' does not exist")
    
    analyze_jplag_results(results_dir)

if __name__ == "__main__":
    main()