459 lines
20 KiB
Python
459 lines
20 KiB
Python
import os
|
|
import tempfile
|
|
import zipfile
|
|
import threading
|
|
import uuid
|
|
from functools import wraps
|
|
from flask import Flask, request, render_template_string, redirect, url_for, session, send_from_directory
|
|
from werkzeug.utils import secure_filename
|
|
from PyPDF2 import PdfReader
|
|
from authlib.integrations.flask_client import OAuth
|
|
|
|
app = Flask(__name__)
|
|
app.config['UPLOAD_FOLDER'] = tempfile.mkdtemp()
|
|
app.config['MAX_CONTENT_LENGTH'] = 500 * 1024 * 1024 # 500MB limit
|
|
app.secret_key = os.getenv('SECRET_KEY', 'dev-secret-key-change-in-production')
|
|
|
|
# Entra ID Configuration
|
|
TENANT_ID = os.getenv('TENANT_ID', 'your-tenant-id')
|
|
CLIENT_ID = os.getenv('CLIENT_ID', 'your-client-id')
|
|
CLIENT_SECRET = os.getenv('CLIENT_SECRET', 'your-client-secret')
|
|
|
|
# Admin Configuration
|
|
ADMIN_USERNAME = os.getenv('ADMIN_USERNAME', 'trbcit')
|
|
ADMIN_PASSWORD = os.getenv('ADMIN_PASSWORD', 'atdrmbicn')
|
|
|
|
oauth = OAuth(app)
|
|
microsoft = oauth.register(
|
|
name='microsoft',
|
|
client_id=CLIENT_ID,
|
|
client_secret=CLIENT_SECRET,
|
|
server_metadata_url=f'https://login.microsoftonline.com/{TENANT_ID}/v2.0/.well-known/openid_configuration',
|
|
client_kwargs={'scope': 'openid profile email'}
|
|
)
|
|
|
|
ALLOWED_EXTENSIONS = {'pdf', 'zip'}
|
|
|
|
# Global jobs storage (in production, use database)
|
|
job_queue = {}
|
|
|
|
def login_required(f):
|
|
@wraps(f)
|
|
def decorated_function(*args, **kwargs):
|
|
if 'user' not in session and 'admin' not in session:
|
|
return redirect(url_for('login'))
|
|
return f(*args, **kwargs)
|
|
return decorated_function
|
|
|
|
def allowed_file(filename):
|
|
return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS
|
|
|
|
def extract_names_from_pdf(pdf_path):
|
|
try:
|
|
reader = PdfReader(pdf_path)
|
|
fields = reader.get_fields()
|
|
|
|
# Extract all required fields
|
|
last_name = fields.get('Last Name', {}).get('/V', '').strip()
|
|
first_name = fields.get('First Name', {}).get('/V', '').strip()
|
|
grade = fields.get('Student Grade', {}).get('/V', '').strip()
|
|
school = fields.get('Student School', {}).get('/V', '').strip()
|
|
|
|
# Check for missing fields
|
|
missing_fields = []
|
|
if not last_name:
|
|
missing_fields.append('Last Name')
|
|
if not first_name:
|
|
missing_fields.append('First Name')
|
|
if not grade:
|
|
missing_fields.append('Student Grade')
|
|
if not school:
|
|
missing_fields.append('Student School')
|
|
|
|
if missing_fields:
|
|
return {'error': f'Missing or empty fields: {", ".join(missing_fields)}'}
|
|
|
|
return {
|
|
'filename': f"{last_name}_{first_name}.pdf",
|
|
'last_name': last_name,
|
|
'first_name': first_name,
|
|
'grade': grade,
|
|
'school': school
|
|
}
|
|
except Exception as e:
|
|
return {'error': f'PDF read error: {str(e)}'}
|
|
|
|
def process_job(job_id, files):
|
|
job_queue[job_id]['status'] = 'processing'
|
|
job_queue[job_id]['progress'] = 'Starting job processing...'
|
|
processed_files = []
|
|
temp_extract_dirs = []
|
|
|
|
# Create base Episode directory for this job
|
|
episode_dir = os.path.join(app.config['UPLOAD_FOLDER'], f'Episode_{job_id}')
|
|
os.makedirs(episode_dir, exist_ok=True)
|
|
job_queue[job_id]['progress'] = 'Created output directory structure'
|
|
|
|
# Analyze ZIP contents before processing
|
|
zip_contents = {}
|
|
for file_info in files:
|
|
if file_info['name'].lower().endswith('.zip'):
|
|
try:
|
|
with zipfile.ZipFile(file_info['path'], 'r') as zip_ref:
|
|
pdf_files = [f for f in zip_ref.namelist() if f.lower().endswith('.pdf')]
|
|
zip_contents[file_info['name']] = pdf_files
|
|
except Exception as e:
|
|
zip_contents[file_info['name']] = [f"Error reading ZIP: {str(e)}"]
|
|
|
|
job_queue[job_id]['zip_contents'] = zip_contents
|
|
|
|
total_files = sum(1 for file_info in files if file_info['name'].lower().endswith('.pdf')) + \
|
|
sum(len(zip_contents.get(file_info['name'], []))
|
|
for file_info in files if file_info['name'].lower().endswith('.zip'))
|
|
|
|
processed_count = 0
|
|
|
|
processed_count = 0
|
|
|
|
for file_info in files:
|
|
filepath = file_info['path']
|
|
filename = file_info['name']
|
|
|
|
if filename.lower().endswith('.zip'):
|
|
job_queue[job_id]['progress'] = f'Extracting ZIP archive: {filename}'
|
|
# Extract ZIP
|
|
extract_dir = os.path.join(app.config['UPLOAD_FOLDER'], f'extract_{job_id}_{os.path.splitext(filename)[0]}')
|
|
os.makedirs(extract_dir, exist_ok=True)
|
|
temp_extract_dirs.append(extract_dir)
|
|
with zipfile.ZipFile(filepath, 'r') as zip_ref:
|
|
zip_ref.extractall(extract_dir)
|
|
job_queue[job_id]['progress'] = f'Extracted {len(zip_ref.namelist())} files from {filename}'
|
|
|
|
# Process PDFs in extract_dir
|
|
for root, dirs, files_in_zip in os.walk(extract_dir):
|
|
for f in files_in_zip:
|
|
if f.lower().endswith('.pdf'):
|
|
pdf_path = os.path.join(root, f)
|
|
job_queue[job_id]['progress'] = f'Processing PDF: {f}'
|
|
pdf_info = extract_names_from_pdf(pdf_path)
|
|
if 'error' in pdf_info:
|
|
processed_files.append(('skipped', f, pdf_info['error']))
|
|
job_queue[job_id]['progress'] = f'Skipped {f} - {pdf_info["error"]}'
|
|
else:
|
|
job_queue[job_id]['progress'] = f'Extracted data for {pdf_info["first_name"]} {pdf_info["last_name"]}'
|
|
# Create folder structure: Episode/School/Grade/
|
|
school_dir = os.path.join(episode_dir, pdf_info['school'])
|
|
grade_dir = os.path.join(school_dir, pdf_info['grade'])
|
|
os.makedirs(grade_dir, exist_ok=True)
|
|
|
|
# Move file to appropriate folder with new name
|
|
new_path = os.path.join(grade_dir, pdf_info['filename'])
|
|
counter = 1
|
|
while os.path.exists(new_path):
|
|
base, ext = os.path.splitext(pdf_info['filename'])
|
|
new_path = os.path.join(grade_dir, f"{base}_{counter}{ext}")
|
|
counter += 1
|
|
os.rename(pdf_path, new_path)
|
|
processed_files.append(('success', os.path.relpath(new_path, episode_dir), new_path))
|
|
processed_count += 1
|
|
job_queue[job_id]['progress'] = f'Organized {processed_count}/{total_files} files'
|
|
else:
|
|
# Single PDF
|
|
job_queue[job_id]['progress'] = f'Processing PDF: {filename}'
|
|
pdf_info = extract_names_from_pdf(filepath)
|
|
if 'error' in pdf_info:
|
|
processed_files.append(('skipped', filename, pdf_info['error']))
|
|
job_queue[job_id]['progress'] = f'Skipped {filename} - {pdf_info["error"]}'
|
|
else:
|
|
job_queue[job_id]['progress'] = f'Extracted data for {pdf_info["first_name"]} {pdf_info["last_name"]}'
|
|
# Create folder structure: Episode/School/Grade/
|
|
school_dir = os.path.join(episode_dir, pdf_info['school'])
|
|
grade_dir = os.path.join(school_dir, pdf_info['grade'])
|
|
os.makedirs(grade_dir, exist_ok=True)
|
|
|
|
# Move file to appropriate folder with new name
|
|
new_filepath = os.path.join(grade_dir, pdf_info['filename'])
|
|
counter = 1
|
|
while os.path.exists(new_filepath):
|
|
base, ext = os.path.splitext(pdf_info['filename'])
|
|
new_filepath = os.path.join(grade_dir, f"{base}_{counter}{ext}")
|
|
counter += 1
|
|
os.rename(filepath, new_filepath)
|
|
processed_files.append(('success', os.path.relpath(new_filepath, episode_dir), new_filepath))
|
|
processed_count += 1
|
|
job_queue[job_id]['progress'] = f'Organized {processed_count}/{total_files} files'
|
|
|
|
job_queue[job_id]['progress'] = 'Creating ZIP archive...'
|
|
# Create ZIP of the entire Episode directory structure
|
|
successful_files = [p for p in processed_files if p[0] == 'success']
|
|
if successful_files:
|
|
try:
|
|
zip_output = os.path.join(app.config['UPLOAD_FOLDER'], f'renamed_{job_id}.zip')
|
|
with zipfile.ZipFile(zip_output, 'w', zipfile.ZIP_DEFLATED) as zip_out:
|
|
for root, dirs, files_in_dir in os.walk(episode_dir):
|
|
for file in files_in_dir:
|
|
file_path = os.path.join(root, file)
|
|
arcname = os.path.relpath(file_path, episode_dir)
|
|
zip_out.write(file_path, arcname)
|
|
job_queue[job_id]['zip_path'] = zip_output
|
|
job_queue[job_id]['zip_name'] = f'renamed_{job_id}.zip'
|
|
job_queue[job_id]['progress'] = f'Created ZIP archive with {len(successful_files)} organized files'
|
|
except Exception as e:
|
|
job_queue[job_id]['progress'] = f'Error creating ZIP: {str(e)}'
|
|
job_queue[job_id]['status'] = 'error'
|
|
else:
|
|
job_queue[job_id]['progress'] = 'No files to archive - all files were skipped'
|
|
|
|
job_queue[job_id]['status'] = 'completed'
|
|
job_queue[job_id]['processed_files'] = processed_files
|
|
job_queue[job_id]['skipped'] = [(p[1], p[2]) for p in processed_files if p[0] == 'skipped']
|
|
|
|
@app.route('/login')
|
|
def login():
|
|
return render_template_string('''
|
|
<!DOCTYPE html>
|
|
<html>
|
|
<head>
|
|
<title>Login</title>
|
|
<style>
|
|
body { font-family: Arial, sans-serif; background-color: #f4f4f4; margin: 0; padding: 20px; text-align: center; }
|
|
.login-container { background-color: #fff; padding: 30px; border-radius: 10px; box-shadow: 0 0 10px rgba(0,0,0,0.1); display: inline-block; margin: 20px; }
|
|
h1 { color: #333; }
|
|
.login-options { display: flex; justify-content: space-around; flex-wrap: wrap; }
|
|
.login-option { margin: 10px; }
|
|
a { color: #007bff; text-decoration: none; padding: 15px 30px; background-color: #007bff; color: white; border-radius: 5px; display: inline-block; }
|
|
a:hover { background-color: #0056b3; }
|
|
form { margin-top: 20px; }
|
|
input[type="text"], input[type="password"] { padding: 10px; margin: 5px; width: 200px; border: 1px solid #ccc; border-radius: 5px; }
|
|
input[type="submit"] { background-color: #28a745; color: white; border: none; padding: 10px 20px; border-radius: 5px; cursor: pointer; }
|
|
input[type="submit"]:hover { background-color: #218838; }
|
|
</style>
|
|
</head>
|
|
<body>
|
|
<img src="/static/trbc-logo-full.png" alt="Logo" width="200">
|
|
<h1>PDF Renamer Login</h1>
|
|
<div class="login-options">
|
|
<div class="login-option">
|
|
<h2>Microsoft Entra ID</h2>
|
|
<a href="/entra-login">Login with Microsoft</a>
|
|
</div>
|
|
<div class="login-option">
|
|
<h2>Admin Login</h2>
|
|
<form method="post" action="/admin-login">
|
|
<input type="text" name="username" placeholder="Username" required><br>
|
|
<input type="password" name="password" placeholder="Password" required><br>
|
|
<input type="submit" value="Admin Login">
|
|
</form>
|
|
</div>
|
|
</div>
|
|
</body>
|
|
</html>
|
|
''')
|
|
|
|
@app.route('/entra-login')
|
|
def entra_login():
|
|
redirect_uri = url_for('auth', _external=True)
|
|
return microsoft.authorize_redirect(redirect_uri)
|
|
|
|
@app.route('/admin-login', methods=['POST'])
|
|
def admin_login():
|
|
username = request.form.get('username')
|
|
password = request.form.get('password')
|
|
if username == ADMIN_USERNAME and password == ADMIN_PASSWORD:
|
|
session['admin'] = True
|
|
return redirect(url_for('upload_files'))
|
|
else:
|
|
return render_template_string('''
|
|
<!DOCTYPE html>
|
|
<html>
|
|
<head>
|
|
<title>Login Failed</title>
|
|
<style>
|
|
body { font-family: Arial, sans-serif; background-color: #f4f4f4; text-align: center; padding: 50px; }
|
|
.error { color: #dc3545; background-color: #f8d7da; padding: 20px; border-radius: 5px; border: 1px solid #f5c6cb; }
|
|
a { color: #007bff; text-decoration: none; }
|
|
</style>
|
|
</head>
|
|
<body>
|
|
<div class="error">
|
|
<h1>Login Failed</h1>
|
|
<p>Invalid username or password.</p>
|
|
<a href="/login">Try Again</a>
|
|
</div>
|
|
</body>
|
|
</html>
|
|
''')
|
|
|
|
@app.route('/auth')
|
|
def auth():
|
|
token = microsoft.authorize_access_token()
|
|
user = microsoft.parse_id_token(token)
|
|
session['user'] = user
|
|
return redirect(url_for('upload_files'))
|
|
|
|
@app.route('/logout')
|
|
@login_required
|
|
def logout():
|
|
session.pop('user', None)
|
|
session.pop('admin', None)
|
|
return redirect(url_for('login'))
|
|
|
|
@app.route('/', methods=['GET', 'POST'])
|
|
@login_required
|
|
def upload_files():
|
|
if request.method == 'POST':
|
|
files = request.files.getlist('files')
|
|
if not files or all(not f.filename for f in files):
|
|
return redirect(url_for('upload_files'))
|
|
|
|
job_id = str(uuid.uuid4())
|
|
job_queue[job_id] = {
|
|
'status': 'queued',
|
|
'files': [],
|
|
'zip_path': None,
|
|
'zip_name': None,
|
|
'processed_files': [],
|
|
'skipped': []
|
|
}
|
|
|
|
file_list = []
|
|
for file in files:
|
|
if file and allowed_file(file.filename):
|
|
filename = secure_filename(file.filename)
|
|
filepath = os.path.join(app.config['UPLOAD_FOLDER'], f"{job_id}_{filename}")
|
|
file.save(filepath)
|
|
file_list.append({'name': filename, 'path': filepath})
|
|
|
|
job_queue[job_id]['files'] = file_list
|
|
|
|
# Start processing in background
|
|
thread = threading.Thread(target=process_job, args=(job_id, file_list))
|
|
thread.daemon = True
|
|
thread.start()
|
|
|
|
return redirect(url_for('jobs'))
|
|
|
|
return render_template_string('''
|
|
<!DOCTYPE html>
|
|
<html>
|
|
<head>
|
|
<title>PDF Renamer</title>
|
|
<style>
|
|
body { font-family: Arial, sans-serif; background-color: #f4f4f4; margin: 0; padding: 20px; text-align: center; }
|
|
img { margin-bottom: 20px; }
|
|
h1 { color: #333; }
|
|
form { background-color: #fff; padding: 20px; border-radius: 10px; box-shadow: 0 0 10px rgba(0,0,0,0.1); display: inline-block; }
|
|
input[type="file"] { margin: 10px 0; }
|
|
input[type="submit"] { background-color: #28a745; color: white; border: none; padding: 10px 20px; border-radius: 5px; cursor: pointer; }
|
|
input[type="submit"]:hover { background-color: #218838; }
|
|
.logout { position: absolute; top: 20px; right: 20px; background-color: #dc3545; color: white; text-decoration: none; padding: 10px 20px; border-radius: 5px; }
|
|
.logout:hover { background-color: #c82333; }
|
|
.view-jobs-btn { background-color: #007bff; color: white; text-decoration: none; padding: 10px 20px; border-radius: 5px; display: inline-block; margin: 10px 0; }
|
|
.view-jobs-btn:hover { background-color: #0056b3; }
|
|
</style>
|
|
</head>
|
|
<body>
|
|
<a href="/logout" class="logout">Logout</a>
|
|
<img src="/static/trbc-logo-full.png" alt="Logo" width="200">
|
|
<h1>Upload PDF Files or ZIP to Rename</h1>
|
|
<a href="/jobs" class="view-jobs-btn">View Jobs</a>
|
|
<br><br>
|
|
<form method="post" enctype="multipart/form-data">
|
|
<input type="file" name="files" multiple accept=".pdf,.zip">
|
|
<br><br>
|
|
<input type="submit" value="Process PDFs">
|
|
</form>
|
|
</body>
|
|
</html>
|
|
''')
|
|
|
|
@app.route('/jobs')
|
|
@login_required
|
|
def jobs():
|
|
return render_template_string('''
|
|
<!DOCTYPE html>
|
|
<html>
|
|
<head>
|
|
<title>Jobs</title>
|
|
<style>
|
|
body { font-family: Arial, sans-serif; background-color: #f4f4f4; margin: 0; padding: 20px; text-align: center; }
|
|
img { margin-bottom: 20px; }
|
|
h1 { color: #333; }
|
|
.job { background-color: #fff; margin: 10px auto; padding: 15px; border-radius: 10px; box-shadow: 0 0 10px rgba(0,0,0,0.1); max-width: 800px; text-align: left; }
|
|
.status { padding: 5px 10px; border-radius: 5px; color: white; font-weight: bold; }
|
|
.status.queued { background-color: #ffc107; }
|
|
.status.processing { background-color: #17a2b8; }
|
|
.status.completed { background-color: #28a745; }
|
|
.status.error { background-color: #dc3545; }
|
|
.progress-text { color: #17a2b8; font-style: italic; }
|
|
.download-btn { background-color: #007bff; color: white; text-decoration: none; padding: 8px 16px; border-radius: 5px; margin-left: 10px; }
|
|
.download-btn:hover { background-color: #0056b3; }
|
|
.logout { position: absolute; top: 20px; right: 20px; background-color: #dc3545; color: white; text-decoration: none; padding: 10px 20px; border-radius: 5px; }
|
|
.logout:hover { background-color: #c82333; }
|
|
.zip-content { background-color: #f8f9fa; padding: 10px; margin: 5px 0; border-radius: 5px; border-left: 4px solid #007bff; }
|
|
ul { list-style-type: none; padding: 0; }
|
|
li { background-color: #ffffff; margin: 2px 0; padding: 5px; border-radius: 3px; }
|
|
</style>
|
|
</head>
|
|
<body>
|
|
<a href="/logout" class="logout">Logout</a>
|
|
<img src="/static/trbc-logo-full.png" alt="Logo" width="200">
|
|
<h1>Processing Jobs</h1>
|
|
{% for job_id, job in jobs.items() %}
|
|
<div class="job">
|
|
<h3>Job {{ job_id[:8] }}</h3>
|
|
<p><strong>Status:</strong> <span class="status {{ job.status }}">{{ job.status.title() }}</span></p>
|
|
{% if job.status == 'processing' and job.progress %}
|
|
<p><strong>Progress:</strong> <span class="progress-text">{{ job.progress }}</span></p>
|
|
{% endif %}
|
|
<p><strong>Files:</strong> {{ job.files|length }}</p>
|
|
{% if job.zip_contents %}
|
|
<p><strong>ZIP Contents:</strong></p>
|
|
{% for zip_name, pdf_files in job.zip_contents.items() %}
|
|
<div class="zip-content">
|
|
<strong>{{ zip_name }}:</strong>
|
|
<ul>
|
|
{% for pdf_file in pdf_files %}
|
|
<li>{{ pdf_file }}</li>
|
|
{% endfor %}
|
|
</ul>
|
|
</div>
|
|
{% endfor %}
|
|
{% endif %}
|
|
{% if job.status == 'completed' %}
|
|
{% if job.zip_name %}
|
|
<p><a href="/download/{{ job.zip_name }}" class="download-btn">Download ZIP</a></p>
|
|
{% else %}
|
|
<p><em>No files were successfully processed to create a ZIP archive.</em></p>
|
|
{% endif %}
|
|
{% if job.skipped %}
|
|
<p><strong>Skipped Files:</strong></p>
|
|
<ul>
|
|
{% for skip_info in job.skipped %}
|
|
<li><strong>{{ skip_info[0] }}</strong>: {{ skip_info[1] }}</li>
|
|
{% endfor %}
|
|
</ul>
|
|
{% endif %}
|
|
{% endif %}
|
|
</div>
|
|
{% endfor %}
|
|
{% if not jobs %}
|
|
<p>No jobs yet. <a href="/">Upload some files</a> to get started.</p>
|
|
{% endif %}
|
|
</body>
|
|
</html>
|
|
''', jobs=job_queue)
|
|
|
|
@app.route('/download/<filename>')
|
|
@login_required
|
|
def download(filename):
|
|
# Security check: only allow downloading files that belong to completed jobs
|
|
for job_id, job in job_queue.items():
|
|
if job.get('status') == 'completed' and job.get('zip_name') == filename:
|
|
file_path = os.path.join(app.config['UPLOAD_FOLDER'], filename)
|
|
if os.path.exists(file_path):
|
|
return send_from_directory(app.config['UPLOAD_FOLDER'], filename, as_attachment=True)
|
|
return "File not found or access denied", 404
|
|
|
|
if __name__ == '__main__':
|
|
app.run(debug=True, port=3000) |