#!/usr/bin/env python3
"""
Fork Statistics Visualizer

This script loads JSON data generated by fork_stats.py and displays
detailed statistics both as formatted lists and visual graphs.

Usage:
    python3 tools/fork_stats_visualizer.py results.json
    python3 tools/fork_stats_visualizer.py results.json --save-plots
    python3 tools/fork_stats_visualizer.py results.json --output-dir ./plots
"""

import argparse
import json
import sys
from pathlib import Path
from typing import Dict, List
from datetime import datetime

# Try to import matplotlib, but allow running without it
try:
    import matplotlib.pyplot as plt
    import matplotlib.patches as mpatches
    HAS_MATPLOTLIB = True
except ImportError:
    HAS_MATPLOTLIB = False
    print("Warning: matplotlib not installed. Graphical visualizations will be disabled.")
    print("Install with: pip install -r tools/fork_stats_visualizer_requirements.txt")
    print()

def load_json_data(filepath: str) -> Dict:
    """Load the JSON data file."""
    try:
        with open(filepath, 'r') as f:
            return json.load(f)
    except FileNotFoundError:
        print(f"Error: File '{filepath}' not found.")
        sys.exit(1)
    except json.JSONDecodeError as e:
        print(f"Error: Invalid JSON in '{filepath}': {e}")
        sys.exit(1)

def print_section_header(title: str):
    """Print a formatted section header."""
    print("\n" + "=" * 80)
    print(f"  {title}")
    print("=" * 80)

def print_repository_info(data: Dict):
    """Print repository information."""
    print_section_header("REPOSITORY INFORMATION")
    
    main_repo = data.get('main_repo', {})
    print(f"\nRepository: {main_repo.get('full_name', 'Unknown')}")
    print(f"Total Forks: {main_repo.get('forks_count', 0):,}")
    print(f"Stars: {main_repo.get('stargazers_count', 0):,}")
    print(f"Watchers: {main_repo.get('watchers_count', 0):,}")
    print(f"\nAnalyzed Forks: {data.get('total_forks', 0)}")
    
    if 'analysis_timestamp' in data:
        timestamp = data['analysis_timestamp']
        try:
            dt = datetime.fromisoformat(timestamp.replace('Z', '+00:00'))
            print(f"Analysis Date: {dt.strftime('%Y-%m-%d %H:%M:%S UTC')}")
        except:
            print(f"Analysis Date: {timestamp}")

def print_age_distribution(stats: Dict):
    """Print fork age distribution statistics."""
    print_section_header("FORK AGE DISTRIBUTION")
    
    age_dist = stats.get('age_distribution', {})
    total = stats.get('total_analyzed', 1)
    
    age_labels = {
        '1_month': 'Last updated ≤ 1 month',
        '3_months': 'Last updated ≤ 3 months',
        '6_months': 'Last updated ≤ 6 months',
        '1_year': 'Last updated ≤ 1 year',
        '2_years': 'Last updated ≤ 2 years',
        '5_plus_years': 'Last updated > 5 years'
    }
    
    print("\nAge Category                    Count    Percentage")
    print("-" * 60)
    for key, label in age_labels.items():
        count = age_dist.get(key, 0)
        pct = (count / total * 100) if total > 0 else 0
        bar = '█' * int(pct / 2)  # Visual bar
        print(f"{label:30} {count:5d}    {pct:5.1f}%  {bar}")

def print_activity_analysis(stats: Dict):
    """Print fork activity analysis."""
    print_section_header("FORK ACTIVITY ANALYSIS")
    
    total = stats.get('total_analyzed', 1)
    
    metrics = [
        ('Forks with unique branches', stats.get('forks_with_unique_branches', 0), 
         stats.get('percentage_with_unique_branches', 0)),
        ('Forks with recent main branch', stats.get('forks_with_recent_main', 0),
         stats.get('percentage_with_recent_main', 0)),
        ('Forks that contributed PRs', stats.get('forks_that_contributed_prs', 0),
         stats.get('percentage_contributed_prs', 0)),
        ('Active forks (no PR contributions)', stats.get('active_non_contributing_forks', 0),
         stats.get('percentage_active_non_contributing', 0)),
    ]
    
    print("\nMetric                              Count    Percentage")
    print("-" * 65)
    for label, count, pct in metrics:
        bar = '█' * int(pct / 2)  # Visual bar
        print(f"{label:35} {count:5d}    {pct:5.1f}%  {bar}")

def print_owner_commits(stats: Dict):
    """Print owner commit statistics."""
    print_section_header("OWNER COMMIT ANALYSIS")
    
    total_forks = stats.get('total_analyzed', 0)
    forks_with_commits = stats.get('forks_with_owner_commits', 0)
    total_commits = stats.get('total_owner_commits', 0)
    avg_commits = stats.get('avg_owner_commits_per_fork', 0)
    pct_with_commits = stats.get('percentage_with_owner_commits', 0)
    
    print(f"\nForks with owner commits:     {forks_with_commits:5d} / {total_forks:5d} ({pct_with_commits:.1f}%)")
    print(f"Total commits by fork owners: {total_commits:5d}")
    print(f"Average commits per fork:     {avg_commits:5.1f}")

def print_top_forks(data: Dict, n: int = 20):
    """Print top forks by various metrics."""
    analyzed_forks = data.get('analyzed_forks', [])
    
    if not analyzed_forks:
        print("\nNo detailed fork data available in JSON file.")
        return
    
    # Top forks by unique branches
    print_section_header(f"TOP {n} FORKS BY UNIQUE BRANCHES")
    forks_by_branches = sorted(analyzed_forks, 
                               key=lambda x: len(x.get('unique_branches', [])), 
                               reverse=True)[:n]
    
    print(f"\n{'Rank':<6} {'Fork':<45} {'Unique Branches':<20} {'Owner Commits'}")
    print("-" * 90)
    for i, fork in enumerate(forks_by_branches, 1):
        unique_count = len(fork.get('unique_branches', []))
        owner_commits = fork.get('owner_commits', 0)
        print(f"{i:<6} {fork.get('full_name', 'Unknown'):<45} {unique_count:<20} {owner_commits}")
    
    # Top forks by owner commits
    print_section_header(f"TOP {n} FORKS BY OWNER COMMITS")
    forks_by_commits = sorted(analyzed_forks,
                             key=lambda x: x.get('owner_commits', 0),
                             reverse=True)[:n]
    
    print(f"\n{'Rank':<6} {'Fork':<45} {'Owner Commits':<20} {'Active'}")
    print("-" * 90)
    for i, fork in enumerate(forks_by_commits, 1):
        commits = fork.get('owner_commits', 0)
        is_active = "Yes" if fork.get('is_active', False) else "No"
        print(f"{i:<6} {fork.get('full_name', 'Unknown'):<45} {commits:<20} {is_active}")
    
    # Active forks that haven't contributed
    print_section_header("ACTIVE FORKS WITHOUT PR CONTRIBUTIONS")
    active_no_pr = [f for f in analyzed_forks 
                   if f.get('is_active', False) and not f.get('has_contributed_prs', False)][:n]
    
    if active_no_pr:
        print(f"\n{'Fork':<45} {'Recent Commits':<20} {'Owner Commits':<20} {'Days Behind'}")
        print("-" * 110)
        for fork in active_no_pr:
            recent = fork.get('recent_commits', 0)
            owner = fork.get('owner_commits', 0)
            days = fork.get('behind_main_by_days', 0)
            print(f"{fork.get('full_name', 'Unknown'):<45} {recent:<20} {owner:<20} {days}")
    else:
        print("\nNo active forks without PR contributions found.")

def create_age_distribution_chart(stats: Dict, save_path: str = None):
    """Create a pie chart for fork age distribution."""
    if not HAS_MATPLOTLIB:
        print("Skipping age distribution chart (matplotlib not available)")
        return
    
    age_dist = stats.get('age_distribution', {})
    
    labels = ['≤ 1 month', '≤ 3 months', '≤ 6 months', '≤ 1 year', '≤ 2 years', '> 5 years']
    sizes = [
        age_dist.get('1_month', 0),
        age_dist.get('3_months', 0),
        age_dist.get('6_months', 0),
        age_dist.get('1_year', 0),
        age_dist.get('2_years', 0),
        age_dist.get('5_plus_years', 0)
    ]
    
    colors = ['#2ecc71', '#27ae60', '#f39c12', '#e67e22', '#e74c3c', '#95a5a6']
    
    fig, ax = plt.subplots(figsize=(10, 8))
    wedges, texts, autotexts = ax.pie(sizes, labels=labels, autopct='%1.1f%%',
                                       colors=colors, startangle=90)
    
    ax.set_title('Fork Age Distribution (Last Update)', fontsize=16, fontweight='bold', pad=20)
    
    # Make percentage text more readable
    for autotext in autotexts:
        autotext.set_color('white')
        autotext.set_fontsize(10)
        autotext.set_fontweight('bold')
    
    plt.tight_layout()
    
    if save_path:
        plt.savefig(save_path, dpi=300, bbox_inches='tight')
        print(f"Saved: {save_path}")
    else:
        plt.show()
    
    plt.close()

def create_activity_bar_chart(stats: Dict, save_path: str = None):
    """Create a bar chart for fork activity metrics."""
    if not HAS_MATPLOTLIB:
        print("Skipping activity bar chart (matplotlib not available)")
        return
    
    metrics = [
        'Unique\nBranches',
        'Recent\nMain',
        'Contributed\nPRs',
        'Active\nNo PRs',
        'Owner\nCommits'
    ]
    
    values = [
        stats.get('percentage_with_unique_branches', 0),
        stats.get('percentage_with_recent_main', 0),
        stats.get('percentage_contributed_prs', 0),
        stats.get('percentage_active_non_contributing', 0),
        stats.get('percentage_with_owner_commits', 0)
    ]
    
    colors = ['#3498db', '#2ecc71', '#9b59b6', '#e67e22', '#e74c3c']
    
    fig, ax = plt.subplots(figsize=(12, 7))
    bars = ax.bar(metrics, values, color=colors, alpha=0.8, edgecolor='black', linewidth=1.5)
    
    ax.set_ylabel('Percentage of Forks (%)', fontsize=12, fontweight='bold')
    ax.set_title('Fork Activity Metrics', fontsize=16, fontweight='bold', pad=20)
    ax.set_ylim(0, 100)
    ax.grid(axis='y', alpha=0.3, linestyle='--')
    
    # Add value labels on bars
    for bar in bars:
        height = bar.get_height()
        ax.text(bar.get_x() + bar.get_width()/2., height + 1,
                f'{height:.1f}%', ha='center', va='bottom', fontweight='bold')
    
    plt.tight_layout()
    
    if save_path:
        plt.savefig(save_path, dpi=300, bbox_inches='tight')
        print(f"Saved: {save_path}")
    else:
        plt.show()
    
    plt.close()

def create_owner_commits_distribution(data: Dict, save_path: str = None):
    """Create a histogram of owner commits distribution."""
    if not HAS_MATPLOTLIB:
        print("Skipping owner commits distribution chart (matplotlib not available)")
        return
    
    analyzed_forks = data.get('analyzed_forks', [])
    
    if not analyzed_forks:
        print("No detailed fork data for commits distribution chart.")
        return
    
    commits = [fork.get('owner_commits', 0) for fork in analyzed_forks if fork.get('owner_commits', 0) > 0]
    
    if not commits:
        print("No owner commits data available.")
        return
    
    fig, ax = plt.subplots(figsize=(12, 7))
    
    # Create histogram with bins
    n, bins, patches = ax.hist(commits, bins=20, color='#3498db', alpha=0.7, edgecolor='black')
    
    # Color code the bins
    cm = plt.cm.RdYlGn_r
    bin_centers = 0.5 * (bins[:-1] + bins[1:])
    col = bin_centers - min(bin_centers)
    col /= max(col)
    
    for c, p in zip(col, patches):
        plt.setp(p, 'facecolor', cm(c))
    
    ax.set_xlabel('Number of Owner Commits', fontsize=12, fontweight='bold')
    ax.set_ylabel('Number of Forks', fontsize=12, fontweight='bold')
    ax.set_title('Distribution of Owner Commits Across Forks', fontsize=16, fontweight='bold', pad=20)
    ax.grid(axis='y', alpha=0.3, linestyle='--')
    
    # Add statistics text
    stats_text = f'Total Forks: {len(commits)}\nMean: {sum(commits)/len(commits):.1f}\nMax: {max(commits)}'
    ax.text(0.95, 0.95, stats_text, transform=ax.transAxes,
            verticalalignment='top', horizontalalignment='right',
            bbox=dict(boxstyle='round', facecolor='wheat', alpha=0.5),
            fontsize=10, fontweight='bold')
    
    plt.tight_layout()
    
    if save_path:
        plt.savefig(save_path, dpi=300, bbox_inches='tight')
        print(f"Saved: {save_path}")
    else:
        plt.show()
    
    plt.close()

def create_combined_dashboard(stats: Dict, data: Dict, save_path: str = None):
    """Create a combined dashboard with multiple charts."""
    if not HAS_MATPLOTLIB:
        print("Skipping combined dashboard (matplotlib not available)")
        return
    
    fig = plt.figure(figsize=(16, 10))
    
    # Age distribution pie chart
    ax1 = plt.subplot(2, 2, 1)
    age_dist = stats.get('age_distribution', {})
    labels = ['≤1mo', '≤3mo', '≤6mo', '≤1yr', '≤2yr', '>5yr']
    sizes = [
        age_dist.get('1_month', 0),
        age_dist.get('3_months', 0),
        age_dist.get('6_months', 0),
        age_dist.get('1_year', 0),
        age_dist.get('2_years', 0),
        age_dist.get('5_plus_years', 0)
    ]
    colors = ['#2ecc71', '#27ae60', '#f39c12', '#e67e22', '#e74c3c', '#95a5a6']
    ax1.pie(sizes, labels=labels, autopct='%1.1f%%', colors=colors, startangle=90)
    ax1.set_title('Fork Age Distribution', fontweight='bold')
    
    # Activity metrics bar chart
    ax2 = plt.subplot(2, 2, 2)
    metrics = ['Unique\nBranches', 'Recent\nMain', 'PRs', 'Active\nNo PRs', 'Owner\nCommits']
    values = [
        stats.get('percentage_with_unique_branches', 0),
        stats.get('percentage_with_recent_main', 0),
        stats.get('percentage_contributed_prs', 0),
        stats.get('percentage_active_non_contributing', 0),
        stats.get('percentage_with_owner_commits', 0)
    ]
    colors_bar = ['#3498db', '#2ecc71', '#9b59b6', '#e67e22', '#e74c3c']
    bars = ax2.bar(metrics, values, color=colors_bar, alpha=0.8)
    ax2.set_ylabel('Percentage (%)')
    ax2.set_title('Activity Metrics', fontweight='bold')
    ax2.set_ylim(0, 100)
    ax2.grid(axis='y', alpha=0.3)
    
    # Owner commits histogram
    ax3 = plt.subplot(2, 2, 3)
    analyzed_forks = data.get('analyzed_forks', [])
    commits = [fork.get('owner_commits', 0) for fork in analyzed_forks if fork.get('owner_commits', 0) > 0]
    if commits:
        ax3.hist(commits, bins=15, color='#3498db', alpha=0.7, edgecolor='black')
        ax3.set_xlabel('Owner Commits')
        ax3.set_ylabel('Frequency')
        ax3.set_title('Owner Commits Distribution', fontweight='bold')
        ax3.grid(axis='y', alpha=0.3)
    
    # Summary statistics
    ax4 = plt.subplot(2, 2, 4)
    ax4.axis('off')
    
    main_repo = data.get('main_repo', {})
    summary_text = f"""
    REPOSITORY STATISTICS
    {'='*35}
    
    Repository: {main_repo.get('full_name', 'Unknown')}
    Total Forks: {main_repo.get('forks_count', 0):,}
    Analyzed: {stats.get('total_analyzed', 0)}
    
    KEY METRICS:
    • Unique Branches: {stats.get('forks_with_unique_branches', 0)} ({stats.get('percentage_with_unique_branches', 0):.1f}%)
    • PR Contributors: {stats.get('forks_that_contributed_prs', 0)} ({stats.get('percentage_contributed_prs', 0):.1f}%)
    • Owner Commits: {stats.get('total_owner_commits', 0):,}
    • Avg Commits/Fork: {stats.get('avg_owner_commits_per_fork', 0):.1f}
    
    INSIGHTS:
    • Recent Forks: {age_dist.get('1_month', 0) + age_dist.get('3_months', 0)}
    • Very Old (>5yr): {age_dist.get('5_plus_years', 0)}
    • Active No PRs: {stats.get('active_non_contributing_forks', 0)}
    """
    
    ax4.text(0.1, 0.9, summary_text, transform=ax4.transAxes,
             verticalalignment='top', fontsize=11, fontfamily='monospace',
             bbox=dict(boxstyle='round', facecolor='lightblue', alpha=0.3))
    
    fig.suptitle('Fork Statistics Dashboard', fontsize=18, fontweight='bold', y=0.98)
    plt.tight_layout(rect=[0, 0.03, 1, 0.96])
    
    if save_path:
        plt.savefig(save_path, dpi=300, bbox_inches='tight')
        print(f"Saved: {save_path}")
    else:
        plt.show()
    
    plt.close()

def main():
    parser = argparse.ArgumentParser(description='Visualize fork statistics data')
    parser.add_argument('json_file', help='Path to the JSON file with fork statistics')
    parser.add_argument('--save-plots', action='store_true', help='Save plots to files instead of displaying')
    parser.add_argument('--output-dir', default='./fork_plots', help='Directory to save plots (default: ./fork_plots)')
    parser.add_argument('--top-n', type=int, default=20, help='Number of top forks to display (default: 20)')
    parser.add_argument('--no-graphs', action='store_true', help='Skip graph generation, only show text statistics')
    
    args = parser.parse_args()
    
    # Load data
    data = load_json_data(args.json_file)
    stats = data.get('statistics', {})
    
    # Print text statistics
    print_repository_info(data)
    print_age_distribution(stats)
    print_activity_analysis(stats)
    print_owner_commits(stats)
    print_top_forks(data, args.top_n)
    
    # Generate graphs if not disabled
    if not args.no_graphs:
        print_section_header("GENERATING VISUAL CHARTS")
        
        if args.save_plots:
            output_dir = Path(args.output_dir)
            output_dir.mkdir(parents=True, exist_ok=True)
            print(f"\nSaving plots to: {output_dir}")
            
            create_age_distribution_chart(stats, output_dir / "age_distribution.png")
            create_activity_bar_chart(stats, output_dir / "activity_metrics.png")
            create_owner_commits_distribution(data, output_dir / "owner_commits_distribution.png")
            create_combined_dashboard(stats, data, output_dir / "dashboard.png")
            
            print(f"\nAll plots saved to: {output_dir}")
        else:
            print("\nDisplaying interactive plots...")
            print("Close each plot window to see the next one.")
            
            create_age_distribution_chart(stats)
            create_activity_bar_chart(stats)
            create_owner_commits_distribution(data)
            create_combined_dashboard(stats, data)
    
    print_section_header("VISUALIZATION COMPLETE")
    print(f"\nData source: {args.json_file}")
    print(f"Total forks analyzed: {stats.get('total_analyzed', 0)}")

if __name__ == '__main__':
    main()