Spaces:

maghwa
/

privacy-annotation

Sleeping

App Files Files Community

Upload annotation_app.py

by maghwa - opened Jan 18

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+456

-0

Files changed (1) hide show

annotation_app.py +456 -0

annotation_app.py ADDED Viewed

	@@ -0,0 +1,456 @@

+"""
+Web-based Annotation Interface for Privacy Inferences
+Modified for Hugging Face Spaces deployment
+Run: python annotation_app.py
+Then open: http://localhost:7860
+"""
+from flask import Flask, render_template, request, jsonify, send_file, session, redirect, url_for
+import json
+from pathlib import Path
+from datetime import datetime
+import os
+from functools import wraps
+import zipfile
+import io
+from collections import defaultdict
+app = Flask(__name__)
+# Security configuration
+app.secret_key = os.environ.get('SECRET_KEY', 'votre-cle-secrete-tres-longue-a-changer-123456789')
+ADMIN_PASSWORD = os.environ.get('ADMIN_PASSWORD', 'antoine2025')
+# Configuration
+RESULTS_DIR = Path("results")
+ANNOTATIONS_DIR = Path("annotations")
+ANNOTATIONS_DIR.mkdir(exist_ok=True)
+MONTHS = [
+    'JANUARY', 'FEBRUARY', 'MARCH', 'APRIL', 'MAY', 'JUNE',
+    'JULY', 'AUGUST', 'SEPTEMBER', 'OCTOBER', 'NOVEMBER', 'DECEMBER'
+]
+CATEGORIES = ['health', 'religion', 'family', 'routines', 'work', 'leisure', 'economics']
+# ============================================================================
+# AUTHENTICATION
+# ============================================================================
+def login_required(f):
+    """Decorator to protect routes"""
+    @wraps(f)
+    def decorated_function(*args, **kwargs):
+        if not session.get('logged_in'):
+            return redirect(url_for('login'))
+        return f(*args, **kwargs)
+    return decorated_function
+@app.route('/login', methods=['GET', 'POST'])
+def login():
+    """Login page"""
+    if request.method == 'POST':
+        if request.form.get('password') == ADMIN_PASSWORD:
+            session['logged_in'] = True
+            return redirect(url_for('index'))
+        else:
+            return '''
+                <html>
+                <head>
+                    <title>Erreur</title>
+                    <style>
+                        body {
+                            font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
+                            max-width: 400px;
+                            margin: 100px auto;
+                            text-align: center;
+                        }
+                        .error { color: #ef4444; margin: 20px 0; }
+                        a { color: #667eea; text-decoration: none; }
+                    </style>
+                </head>
+                <body>
+                    <h2>❌ Mot de passe incorrect</h2>
+                    <p class="error">Le mot de passe saisi n'est pas valide.</p>
+                    <a href="/login">← Réessayer</a>
+                </body>
+                </html>
+            '''
+    return '''
+        <html>
+        <head>
+            <title>Connexion - Privacy Annotation</title>
+            <style>
+                body {
+                    font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif;
+                    background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+                    height: 100vh;
+                    display: flex;
+                    align-items: center;
+                    justify-content: center;
+                    margin: 0;
+                }
+                .login-box {
+                    background: white;
+                    padding: 40px;
+                    border-radius: 12px;
+                    box-shadow: 0 10px 40px rgba(0,0,0,0.2);
+                    width: 100%;
+                    max-width: 400px;
+                }
+                h2 {
+                    margin: 0 0 30px 0;
+                    color: #333;
+                    text-align: center;
+                }
+                input {
+                    width: 100%;
+                    padding: 12px;
+                    margin: 10px 0;
+                    border: 2px solid #e5e7eb;
+                    border-radius: 8px;
+                    font-size: 16px;
+                    box-sizing: border-box;
+                }
+                input:focus {
+                    outline: none;
+                    border-color: #667eea;
+                }
+                button {
+                    width: 100%;
+                    padding: 12px;
+                    margin-top: 10px;
+                    background: #667eea;
+                    color: white;
+                    border: none;
+                    border-radius: 8px;
+                    font-size: 16px;
+                    font-weight: 600;
+                    cursor: pointer;
+                    transition: background 0.2s;
+                }
+                button:hover {
+                    background: #5568d3;
+                }
+                .info {
+                    text-align: center;
+                    color: #6b7280;
+                    font-size: 14px;
+                    margin-top: 20px;
+                }
+            </style>
+        </head>
+        <body>
+            <div class="login-box">
+                <h2>🔒 Privacy Annotation</h2>
+                <form method="post">
+                    <input type="password" name="password" placeholder="Mot de passe" required autofocus>
+                    <button type="submit">Se connecter</button>
+                </form>
+                <p class="info">Interface d'annotation pour évaluation des modèles LLM</p>
+            </div>
+        </body>
+        </html>
+    '''
+@app.route('/logout')
+def logout():
+    """Logout"""
+    session.pop('logged_in', None)
+    return redirect(url_for('login'))
+# ============================================================================
+# HELPER FUNCTIONS
+# ============================================================================
+def get_available_models():
+    """Get list of available model directories."""
+    if not RESULTS_DIR.exists():
+        return []
+    return [d.name for d in RESULTS_DIR.iterdir() if d.is_dir() and not d.name.startswith('.')]
+def load_result(model_name, month):
+    """Load a specific result file."""
+    filepath = RESULTS_DIR / model_name / "2017" / f"2017_{month}_P1.json"
+    if not filepath.exists():
+        return None
+    with open(filepath, 'r') as f:
+        return json.load(f)
+def get_annotation_status():
+    """Get status of all annotations."""
+    status = {}
+    models = get_available_models()
+    for model in models:
+        status[model] = {}
+        for month in MONTHS:
+            ann_file = ANNOTATIONS_DIR / f"{model}_{month}_annotations.json"
+            status[model][month] = ann_file.exists()
+    return status
+def extract_inferences(response, category):
+    """Extract inferences for a specific category."""
+    inferences = []
+    lines = response.split('\n')
+    in_category = False
+    for line in lines:
+        # Check if entering this category
+        if category.lower() in line.lower() and (':' in line or category in line):
+            in_category = True
+            continue
+        # Check if entering new category
+        if any(cat.lower() in line.lower() for cat in CATEGORIES if cat != category):
+            in_category = False
+        # Extract inference
+        if in_category and line.strip():
+            cleaned = line.strip().lstrip('•-*0123456789.) ')
+            if len(cleaned) > 15:
+                inferences.append(cleaned)
+    return inferences
+# ============================================================================
+# MAIN ROUTES
+# ============================================================================
+@app.route('/')
+@login_required
+def index():
+    """Main page - show annotation dashboard."""
+    models = get_available_models()
+    status = get_annotation_status()
+    return render_template('index.html',
+                         models=models,
+                         months=MONTHS,
+                         status=status)
+@app.route('/annotate/<model>/<month>')
+@login_required
+def annotate(model, month):
+    """Annotation page for specific model and month."""
+    result = load_result(model, month)
+    if not result:
+        return f"Result not found: {model}/{month}", 404
+    # Check if already annotated
+    ann_file = ANNOTATIONS_DIR / f"{model}_{month}_annotations.json"
+    existing_annotation = None
+    if ann_file.exists():
+        with open(ann_file, 'r') as f:
+            existing_annotation = json.load(f)
+    # Extract inferences by category
+    category_inferences = {}
+    for category in CATEGORIES:
+        category_inferences[category] = extract_inferences(result['response'], category)
+    return render_template('annotate.html',
+                         model=model,
+                         month=month,
+                         response=result['response'],
+                         categories=CATEGORIES,
+                         category_inferences=category_inferences,
+                         existing_annotation=existing_annotation,
+                         trajectory_stats=result.get('metadata', {}).get('trajectory_stats', {}))
+@app.route('/api/save_annotation', methods=['POST'])
+@login_required
+def save_annotation():
+    """Save annotation via API."""
+    data = request.json
+    model = data['model']
+    month = data['month']
+    annotations = data['annotations']
+    # Load original result
+    result = load_result(model, month)
+    if not result:
+        return jsonify({'success': False, 'error': 'Result not found'}), 404
+    # Create annotation data
+    ann_data = {
+        'model_name': model,
+        'month': month,
+        'year': 2017,
+        'annotated_at': datetime.now().isoformat(),
+        'annotation_mode': 'web_interface',
+        'original_response': result['response'],
+        'annotations': annotations,
+        'metadata': {
+            'trajectory_period': result.get('trajectory_period'),
+            'prompt_type': result.get('prompt_type'),
+        }
+    }
+    # Save to file
+    ann_file = ANNOTATIONS_DIR / f"{model}_{month}_annotations.json"
+    with open(ann_file, 'w') as f:
+        json.dump(ann_data, f, indent=2)
+    return jsonify({'success': True, 'file': str(ann_file)})
+@app.route('/api/metrics/<model>')
+@login_required
+def get_metrics(model):
+    """Calculate metrics for a model."""
+    pattern = f"{model}_*_annotations.json"
+    ann_files = list(ANNOTATIONS_DIR.glob(pattern))
+    if not ann_files:
+        return jsonify({'success': False, 'error': 'No annotations found'})
+    total_tp = 0
+    total_fp = 0
+    total_pa = 0
+    by_category = {}
+    annotated_months = []
+    for ann_file in ann_files:
+        with open(ann_file, 'r') as f:
+            data = json.load(f)
+        annotated_months.append(data['month'])
+        for category, items in data['annotations'].items():
+            if category not in by_category:
+                by_category[category] = {'TP': 0, 'FP': 0, 'PA': 0}
+            for item in items:
+                label = item['label']
+                if label == 'TP':
+                    total_tp += 1
+                    by_category[category]['TP'] += 1
+                elif label == 'FP':
+                    total_fp += 1
+                    by_category[category]['FP'] += 1
+                elif label == 'PA':
+                    total_pa += 1
+                    by_category[category]['PA'] += 1
+    # Calculate precision
+    precision = total_tp / (total_tp + total_fp) if (total_tp + total_fp) > 0 else 0
+    # Category metrics
+    category_metrics = {}
+    for category, stats in by_category.items():
+        tp = stats['TP']
+        fp = stats['FP']
+        prec = tp / (tp + fp) if (tp + fp) > 0 else 0
+        category_metrics[category] = {
+            'tp': tp,
+            'fp': fp,
+            'pa': stats['PA'],
+            'precision': round(prec, 3)
+        }
+    return jsonify({
+        'success': True,
+        'model': model,
+        'months_annotated': len(annotated_months),
+        'annotated_months': annotated_months,
+        'total_tp': total_tp,
+        'total_fp': total_fp,
+        'total_pa': total_pa,
+        'precision': round(precision, 3),
+        'by_category': category_metrics
+    })
+@app.route('/metrics')
+@login_required
+def metrics_page():
+    """Metrics dashboard page."""
+    models = get_available_models()
+    return render_template('metrics.html', models=models)
+@app.route('/download-annotations')
+@login_required
+def download_annotations():
+    """Download all annotations as ZIP file"""
+    memory_file = io.BytesIO()
+    # Create ZIP with all annotations
+    with zipfile.ZipFile(memory_file, 'w', zipfile.ZIP_DEFLATED) as zf:
+        ann_files = list(ANNOTATIONS_DIR.glob('*.json'))
+        if not ann_files:
+            return "Aucune annotation disponible pour le moment.", 404
+        for file in ann_files:
+            zf.write(file, file.name)
+    memory_file.seek(0)
+    # Filename with date
+    filename = f'annotations_antoine_{datetime.now().strftime("%Y%m%d_%H%M")}.zip'
+    return send_file(
+        memory_file,
+        mimetype='application/zip',
+        as_attachment=True,
+        download_name=filename
+    )
+# ============================================================================
+# MAIN
+# ============================================================================
+if __name__ == '__main__':
+    # Hugging Face Spaces uses port 7860 by default
+    port = int(os.environ.get('PORT', 7860))
+    # Create directories if needed
+    templates_dir = Path("templates")
+    templates_dir.mkdir(exist_ok=True)
+    RESULTS_DIR.mkdir(exist_ok=True)
+    ANNOTATIONS_DIR.mkdir(exist_ok=True)
+    print("="*70)
+    print("PRIVACY INFERENCE ANNOTATION WEB INTERFACE")
+    print("="*70)
+    print(f"\n✓ Starting server on port {port}...")
+    print(f"✓ Results directory: {RESULTS_DIR.absolute()}")
+    print(f"✓ Annotations will be saved to: {ANNOTATIONS_DIR.absolute()}")
+    # Check if running on HF
+    if os.environ.get('SPACE_ID'):
+        print(f"✓ Running on Hugging Face Spaces")
+        print(f"✓ Space ID: {os.environ.get('SPACE_ID')}")
+    else:
+        print(f"\n🌐 Open your browser to: http://localhost:{port}")
+    print("\nPress Ctrl+C to stop the server")
+    print("="*70 + "\n")
+    # IMPORTANT: host='0.0.0.0' to be accessible from outside
+    # debug=False for production
+    app.run(host='0.0.0.0', port=port, debug=False)