bethaus-app/app.py

import os
import sqlite3

# Use eventlet only in production; keep dev on threading to avoid monkey_patch issues with reloader
FLASK_ENV = os.environ.get('FLASK_ENV', 'production')
if FLASK_ENV == 'production':
    import eventlet
    eventlet.monkey_patch()

from flask import Flask, render_template, send_file, url_for, jsonify, request, session, send_from_directory, make_response, abort
from PIL import Image, ImageOps, ExifTags
import io
from functools import wraps
import mimetypes
from datetime import datetime, date, timedelta
import diskcache
import threading
from flask_socketio import SocketIO, emit
import geoip2.database
from functools import lru_cache
from urllib.parse import urlparse, unquote
from werkzeug.middleware.proxy_fix import ProxyFix
from pathlib import Path
import re
import qrcode
import base64
import json
import io
import search
import auth
import analytics as a
import folder_secret_config_editor as fsce
import helperfunctions as hf
import search_db_analyzer as sdb
import fnmatch
import openpyxl
from collections import OrderedDict

app_config = auth.return_app_config()
BASE_DIR = os.path.realpath(app_config['BASE_DIR'])

cache_audio = diskcache.Cache('./filecache_audio', size_limit= app_config['filecache_size_limit_audio'] * 1024**3)
cache_image = diskcache.Cache('./filecache_image', size_limit= app_config['filecache_size_limit_image'] * 1024**3)
cache_video = diskcache.Cache('./filecache_video', size_limit= app_config['filecache_size_limit_video'] * 1024**3)
cache_other = diskcache.Cache('./filecache_other', size_limit= app_config['filecache_size_limit_other'] * 1024**3)

_logged_request_ids = OrderedDict()
_logged_request_ids_lock = threading.Lock()
_LOGGED_REQUEST_IDS_MAX = 2048


def _is_duplicate_request(req_id: str) -> bool:
    if not req_id:
        return False
    with _logged_request_ids_lock:
        return req_id in _logged_request_ids


def _mark_request_logged(req_id: str):
    if not req_id:
        return
    with _logged_request_ids_lock:
        _logged_request_ids[req_id] = None
        _logged_request_ids.move_to_end(req_id)
        if len(_logged_request_ids) > _LOGGED_REQUEST_IDS_MAX:
            _logged_request_ids.popitem(last=False)

app = Flask(__name__)
app.wsgi_app = ProxyFix(app.wsgi_app, x_for=1, x_proto=1)

app.config['SECRET_KEY'] = app_config['SECRET_KEY']
app.config['PERMANENT_SESSION_LIFETIME'] = timedelta(days=90)
if FLASK_ENV == 'production':
    app.config['SESSION_COOKIE_SAMESITE'] = 'None'
    app.config['SESSION_COOKIE_SECURE'] = True

app.add_url_rule('/dashboard', view_func=auth.require_admin(a.dashboard))
app.add_url_rule('/file_access', view_func=auth.require_admin(a.file_access))
app.add_url_rule('/connections', view_func=auth.require_admin(a.connections))
app.add_url_rule('/mylinks', view_func=auth.require_secret(auth.mylinks))
app.add_url_rule('/songs_dashboard', view_func=auth.require_admin(a.songs_dashboard))

app.add_url_rule('/remove_secret', view_func=auth.remove_secret, methods=['POST'])
app.add_url_rule('/remove_token', view_func=auth.remove_token, methods=['POST'])
app.add_url_rule('/searchcommand', view_func=search.searchcommand, methods=['POST'])

app.add_url_rule('/admin/folder_secret_config_editor', view_func=auth.require_admin(fsce.folder_secret_config_editor), methods=['GET', 'POST'])
app.add_url_rule('/admin/folder_secret_config_editor/data', view_func=auth.require_admin(auth.load_folder_config))
app.add_url_rule('/admin/folder_secret_config_editor/action', view_func=auth.require_admin(fsce.folder_secret_config_action), methods=['POST'])

@app.route('/admin/generate_qr/<secret>')
@auth.require_admin
def generate_qr_code(secret):
    scheme = request.scheme
    host = request.host
    url = f"{scheme}://{host}?secret={secret}"
    qr = qrcode.QRCode(version=1, box_size=10, border=4)
    qr.add_data(url)
    qr.make(fit=True)
    img = qr.make_image(fill_color="black", back_color="white")
    buffer = io.BytesIO()
    img.save(buffer, format="PNG")
    buffer.seek(0)
    img_base64 = base64.b64encode(buffer.getvalue()).decode('ascii')
    return jsonify({'qr_code': img_base64})

app.add_url_rule('/admin/search_db_analyzer', view_func=auth.require_admin(sdb.search_db_analyzer))
app.add_url_rule('/admin/search_db_analyzer/query', view_func=auth.require_admin(sdb.search_db_query), methods=['POST'])
app.add_url_rule('/admin/search_db_analyzer/folders', view_func=auth.require_admin(sdb.search_db_folders))

# Messages/News routes
MESSAGES_FILENAME = 'messages.json'
CALENDAR_DB = 'calendar.db'

def load_messages():
    """Load messages from JSON file."""
    try:
        with open(MESSAGES_FILENAME, 'r', encoding='utf-8') as f:
            return json.load(f)
    except (FileNotFoundError, json.JSONDecodeError):
        return []

def save_messages(messages):
    """Save messages to JSON file."""
    with open(MESSAGES_FILENAME, 'w', encoding='utf-8') as f:
        json.dump(messages, f, ensure_ascii=False, indent=2)


def init_calendar_db():
    """Ensure calendar table exists."""
    conn = sqlite3.connect(CALENDAR_DB)
    conn.execute(
        """
        CREATE TABLE IF NOT EXISTS calendar_entries (
            id INTEGER PRIMARY KEY AUTOINCREMENT,
            date TEXT NOT NULL,
            time TEXT,
            title TEXT NOT NULL,
            location TEXT,
            details TEXT,
            created_at TEXT NOT NULL
        )
        """
    )
    # Ensure details column exists for legacy DBs
    cols = [r[1] for r in conn.execute("PRAGMA table_info(calendar_entries)").fetchall()]
    if 'details' not in cols:
        conn.execute("ALTER TABLE calendar_entries ADD COLUMN details TEXT")
        conn.commit()
    conn.commit()
    conn.close()


def serialize_calendar_row(row):
    return {
        'id': row['id'],
        'date': row['date'],
        'time': row['time'] or '',
        'title': row['title'],
        'location': row['location'] or '',
        'details': row['details'] or ''
    }


init_calendar_db()

@app.route('/api/messages', methods=['GET'])
@auth.require_secret
def get_messages():
    """Get all messages."""
    messages = load_messages()
    # Sort by datetime descending (newest first)
    messages.sort(key=lambda x: x.get('datetime', ''), reverse=True)
    return jsonify(messages)

@app.route('/api/messages', methods=['POST'])
@auth.require_admin
def create_message():
    """Create a new message."""
    data = request.get_json()
    messages = load_messages()
    
    # Generate new ID
    new_id = max([m.get('id', 0) for m in messages], default=0) + 1
    
    new_message = {
        'id': new_id,
        'title': data.get('title', ''),
        'content': data.get('content', ''),
        'datetime': data.get('datetime', datetime.now().isoformat())
    }
    
    messages.append(new_message)
    save_messages(messages)
    
    return jsonify(new_message), 201

@app.route('/api/messages/<int:message_id>', methods=['PUT'])
@auth.require_admin
def update_message(message_id):
    """Update an existing message."""
    data = request.get_json()
    messages = load_messages()
    
    for message in messages:
        if message['id'] == message_id:
            message['title'] = data.get('title', message['title'])
            message['content'] = data.get('content', message['content'])
            message['datetime'] = data.get('datetime', message['datetime'])
            save_messages(messages)
            return jsonify(message)
    
    return jsonify({'error': 'Message not found'}), 404

@app.route('/api/messages/<int:message_id>', methods=['DELETE'])
@auth.require_admin
def delete_message(message_id):
    """Delete a message."""
    messages = load_messages()
    messages = [m for m in messages if m['id'] != message_id]
    save_messages(messages)
    
    return jsonify({'success': True})


@app.route('/api/calendar/<int:entry_id>', methods=['PUT'])
@auth.require_admin
def update_calendar_entry(entry_id):
    """Update a calendar entry."""
    data = request.get_json() or {}
    entry_date = data.get('date')
    title = data.get('title')
    if not entry_date or not title:
        return jsonify({'error': 'date and title are required'}), 400

    time_value = data.get('time') or ''
    location = data.get('location') or ''
    details = data.get('details') or ''

    conn = sqlite3.connect(CALENDAR_DB)
    conn.row_factory = sqlite3.Row
    cur = conn.execute(
        """
        UPDATE calendar_entries
        SET date = ?, time = ?, title = ?, location = ?, details = ?
        WHERE id = ?
        """,
        (entry_date, time_value, title, location, details, entry_id)
    )
    if cur.rowcount == 0:
        conn.close()
        return jsonify({'error': 'not found'}), 404

    row = conn.execute(
        "SELECT id, date, time, title, location, details FROM calendar_entries WHERE id = ?",
        (entry_id,)
    ).fetchone()
    conn.commit()
    conn.close()
    return jsonify(serialize_calendar_row(row))


@app.route('/api/calendar/<int:entry_id>', methods=['DELETE'])
@auth.require_admin
def delete_calendar_entry(entry_id):
    """Delete a calendar entry."""
    conn = sqlite3.connect(CALENDAR_DB)
    cur = conn.execute("DELETE FROM calendar_entries WHERE id = ?", (entry_id,))
    conn.commit()
    conn.close()
    if cur.rowcount == 0:
        return jsonify({'error': 'not found'}), 404
    return jsonify({'success': True})


@app.route('/api/calendar/export', methods=['GET'])
@auth.require_admin
def export_calendar():
    """Export all calendar entries to an Excel file."""
    conn = sqlite3.connect(CALENDAR_DB)
    conn.row_factory = sqlite3.Row
    rows = conn.execute(
        """
        SELECT date, time, title, location, details
        FROM calendar_entries
        ORDER BY date ASC, time ASC
        """
    ).fetchall()
    conn.close()

    wb = openpyxl.Workbook()
    ws = wb.active
    ws.title = "Kalender"
    headers = ["date", "time", "title", "location", "details"]
    ws.append(headers)

    def fmt_date(date_str):
        try:
            return datetime.strptime(date_str, '%Y-%m-%d').strftime('%d.%m.%Y')
        except Exception:
            return date_str

    for row in rows:
        ws.append([
            fmt_date(row['date']),
            row['time'] or '',
            row['title'],
            row['location'] or '',
            row['details'] or ''
        ])

    output = io.BytesIO()
    wb.save(output)
    output.seek(0)

    response = make_response(output.getvalue())
    response.headers['Content-Disposition'] = 'attachment; filename=kalender.xlsx'
    response.headers['Content-Type'] = 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet'
    return response


@app.route('/api/calendar/import', methods=['POST'])
@auth.require_admin
def import_calendar():
    """Import calendar entries from an uploaded Excel file, replacing existing entries."""
    if 'file' not in request.files or request.files['file'].filename == '':
        return jsonify({'error': 'No file uploaded'}), 400

    file = request.files['file']
    try:
        wb = openpyxl.load_workbook(file, data_only=True)
    except Exception as e:
        return jsonify({'error': f'Invalid Excel file: {e}'}), 400

    ws = wb.active
    headers = [str(cell.value).strip().lower() if cell.value else '' for cell in next(ws.iter_rows(min_row=1, max_row=1, values_only=False))]
    required = {'date', 'title'}
    optional = {'time', 'location', 'details'}
    all_allowed = required | optional

    header_map = {}
    for idx, h in enumerate(headers):
        if h in all_allowed and h not in header_map:
            header_map[h] = idx

    if not required.issubset(set(header_map.keys())):
        return jsonify({'error': 'Header must include at least date and title columns'}), 400

    entries = []
    for row in ws.iter_rows(min_row=2, values_only=True):
        entry = {}
        for key, col_idx in header_map.items():
            value = row[col_idx] if col_idx < len(row) else ''
            entry[key] = value if value is not None else ''
        # Normalize
        entry_date_raw = entry.get('date', '')
        entry_date = str(entry_date_raw).strip()
        if not entry_date:
            continue
        try:
            # Allow Excel date cells
            if isinstance(entry_date_raw, (datetime, date)):
                entry_date = entry_date_raw.strftime('%Y-%m-%d')
            elif '.' in entry_date:
                entry_date = datetime.strptime(entry_date, '%d.%m.%Y').strftime('%Y-%m-%d')
            else:
                # Expect yyyy-mm-dd
                datetime.strptime(entry_date, '%Y-%m-%d')
        except Exception:
            return jsonify({'error': f'Invalid date format: {entry_date}'}), 400

        entries.append({
            'date': entry_date,
            'time': str(entry.get('time', '') or '').strip(),
            'title': str(entry.get('title', '') or '').strip(),
            'location': str(entry.get('location', '') or '').strip(),
            'details': str(entry.get('details', '') or '').strip()
        })

    conn = sqlite3.connect(CALENDAR_DB)
    try:
        cur = conn.cursor()
        cur.execute("DELETE FROM calendar_entries")
        cur.executemany(
            """
            INSERT INTO calendar_entries (date, time, title, location, details, created_at)
            VALUES (?, ?, ?, ?, ?, ?)
            """,
            [
                (
                    e['date'],
                    e['time'],
                    e['title'],
                    e['location'],
                    e['details'],
                    datetime.utcnow().isoformat()
                ) for e in entries
            ]
        )
        conn.commit()
    finally:
        conn.close()

    return jsonify({'success': True, 'count': len(entries)})

@app.route('/api/calendar', methods=['GET'])
@auth.require_secret
def get_calendar_entries():
    """Return calendar entries between start and end date (inclusive)."""
    start = request.args.get('start') or date.today().isoformat()
    end = request.args.get('end') or (date.today() + timedelta(weeks=4)).isoformat()

    conn = sqlite3.connect(CALENDAR_DB)
    conn.row_factory = sqlite3.Row
    rows = conn.execute(
        """
        SELECT id, date, time, title, location, details
        FROM calendar_entries
        WHERE date BETWEEN ? AND ?
        ORDER BY date ASC, time ASC
        """,
        (start, end)
    ).fetchall()
    conn.close()

    return jsonify([serialize_calendar_row(r) for r in rows])


@app.route('/api/calendar', methods=['POST'])
@auth.require_admin
def create_calendar_entry():
    """Store a new calendar entry."""
    data = request.get_json() or {}
    entry_date = data.get('date')
    title = data.get('title')
    if not entry_date or not title:
        return jsonify({'error': 'date and title are required'}), 400

    time_value = data.get('time') or ''
    location = data.get('location') or ''
    details = data.get('details') or ''

    conn = sqlite3.connect(CALENDAR_DB)
    cur = conn.execute(
        """
        INSERT INTO calendar_entries (date, time, title, location, details, created_at)
        VALUES (?, ?, ?, ?, ?, ?)
        """,
        (entry_date, time_value, title, location, details, datetime.utcnow().isoformat())
    )
    conn.commit()
    new_id = cur.lastrowid
    conn.close()

    return jsonify({
        'id': new_id,
        'date': entry_date,
        'time': time_value,
        'title': title,
        'location': location,
        'details': details
    }), 201

# Grab the HOST_RULE environment variable
host_rule = os.getenv("HOST_RULE", "")
# Use a regex to extract domain names between backticks in patterns like Host(`something`)
pattern = r"Host\(`([^`]+)`\)"
allowed_domains = re.findall(pattern, host_rule)

socketio = SocketIO(
    app,
async_mode='eventlet' if FLASK_ENV == 'production' else 'threading',
    cors_allowed_origins=allowed_domains
)
background_thread_running = False

# Global variables to track the number of connected clients and the background thread
clients_connected = 0
background_thread = None
thread_lock = threading.Lock()

@lru_cache(maxsize=10)
def get_cached_image(size):
    dimensions = tuple(map(int, size.split('-')[1].split('x')))
    original_logo_path = os.path.join(app.root_path, 'custom_logo', 'logoB.png')

    with Image.open(original_logo_path) as img:
        img = img.convert("RGBA")

        orig_width, orig_height = img.size

        if dimensions[0] >= orig_width and dimensions[1] >= orig_height:
            resized_img = img
        else:
            resized_img = img.copy()
            resized_img.thumbnail(dimensions, Image.LANCZOS)

        img_byte_arr = io.BytesIO()
        resized_img.save(img_byte_arr, format='PNG')
        return img_byte_arr.getvalue()

def check_path(access_path: str) -> Path:
    """
    Take an absolute access_path, then ensure it lives inside BASE_DIR.
    Raises ValueError or PermissionError on failure.
    """
    p = Path(access_path)
    if not p.is_absolute():
        raise ValueError(f"Path {access_path} is not a valid absolute path")

    # Resolve symlinks & eliminate “..” components
    candidate = p.resolve()
    base = Path(BASE_DIR).resolve()

    try:
        # Will raise ValueError if candidate is not under base
        candidate.relative_to(base)
    except ValueError:
        raise PermissionError(f"Access to {access_path} is forbidden")

    return candidate

def list_directory_contents(directory, subpath):
    """
    List only the immediate contents of the given directory.
    Also, if a "Transkription" subfolder exists, check for matching .md files for music files.
    Skip folders that start with a dot.
    """
    directories = []
    files = []
    folder_config = auth.return_folder_config()
    transcription_dir = os.path.join(directory, "Transkription")
    transcription_exists = os.path.isdir(transcription_dir)

    direct_directories = []
    for item in folder_config:
        for folder in item['folders']:
            direct_directories.append(folder['foldername'])

    # Define allowed file extensions.
    music_exts = ('.mp3',)
    image_exts = ('.jpg', '.jpeg', '.png', '.gif', '.bmp')
    
    blocked_filenames = [
        'Thumbs.db',
        '*.mrk', '*.md',
        '*.arw', '*.cr2', '*.lrf',
        '*.mov', '*.mp4', '*.avi', '*.mkv'
    ]


    try:
        with os.scandir(directory) as it:
            # Sorting by name if required.
            for entry in sorted(it, key=lambda e: e.name):
                # Skip hidden files and directories.
                if entry.name.startswith('.'):
                    continue
                
                # Skip blocked_filenames using fnmatch for wildcards
                if any(fnmatch.fnmatch(entry.name, pattern) for pattern in blocked_filenames):
                    continue

                if entry.is_dir(follow_symlinks=False):
                    # skip system/hidden helper folders
                    if entry.name in ["Transkription", "@eaDir", ".app", "#recycle"]:
                        continue
                    rel_path = os.path.join(subpath, entry.name) if subpath else entry.name
                    
                    # check if path is inside a direct directory --> shareable
                    if subpath.split('/')[0] in direct_directories:
                        allow_share = True
                    else:
                        allow_share = False
                    
                    # build directory entry
                    directories.append({'name': entry.name, 'path': rel_path.replace(os.sep, '/'), 'share': allow_share})

                elif entry.is_file(follow_symlinks=False):
                    lower_name = entry.name.lower()

                    # implement file type filtering here !!!
                    #if lower_name.endswith(music_exts) or lower_name.endswith(image_exts):
                    rel_path = os.path.join(subpath, entry.name) if subpath else entry.name
                    if lower_name.endswith(music_exts):
                        file_type = 'music'
                    elif lower_name.endswith(image_exts):
                        file_type = 'image'
                    else:
                        file_type = 'other'
                    
                    # build file entry
                    file_entry = {'name': entry.name, 'path': rel_path.replace(os.sep, '/'), 'file_type': file_type}
                    
                    # Only check for transcription if it's a audio file.
                    if file_type == 'music' and transcription_exists:
                        base_name = os.path.splitext(entry.name)[0]
                        transcript_filename = base_name + '.md'
                        transcript_path = os.path.join(transcription_dir, transcript_filename)
                        if os.path.isfile(transcript_path):
                            file_entry['has_transcript'] = True
                            transcript_rel_path = os.path.join(subpath, "Transkription", transcript_filename) if subpath else os.path.join("Transkription", transcript_filename)
                            file_entry['transcript_url'] = url_for('get_transcript', subpath=transcript_rel_path.replace(os.sep, '/'))
                        else:
                            file_entry['has_transcript'] = False
                    else:
                        file_entry['has_transcript'] = False
                    files.append(file_entry)
    except PermissionError:
        pass

    return directories, files


def generate_breadcrumbs(subpath=None):
    breadcrumbs = [{'name': 'Home', 'path': ''}]
    if subpath:
        parts = subpath.split('/')
        path_accum = ""
        for part in parts:
            path_accum = f"{path_accum}/{part}" if path_accum else part
            if 'toplist' in part:
                part = part.replace('toplist', 'oft angehört')
            breadcrumbs.append({'name': part, 'path': path_accum})
    return breadcrumbs


def human_readable_size(num_bytes):
    """
    Convert a byte count into a human-readable string (e.g., 1.2 MB).
    """
    if num_bytes is None:
        return "-"
    try:
        num = float(num_bytes)
    except (TypeError, ValueError):
        return "-"
    units = ['B', 'KB', 'MB', 'GB', 'TB']
    for unit in units:
        if num < 1024 or unit == units[-1]:
            return f"{num:.1f} {unit}" if num % 1 else f"{int(num)} {unit}"
        num /= 1024

@app.route('/icon/<string:size>.png')
@app.route('/icons/<string:size>.png')  # legacy path
def serve_resized_icon(size):
    cached_image_bytes = get_cached_image(size)
    response = send_file(
        io.BytesIO(cached_image_bytes),
        mimetype='image/png'
    )
    response.headers['Cache-Control'] = 'public, max-age=86400, immutable'
    return response

@app.route('/custom_logo/<string:filename>.png')
def custom_logo(filename):
    file_path = os.path.join('custom_logo', f"{filename}.png")
    if not os.path.exists(file_path):
        abort(404)
    with open(file_path, 'rb') as file:
        image_data = file.read()

    # Create a BytesIO object using the binary image data
    image_io = io.BytesIO(image_data)
    image_io.seek(0)  # Important: reset the stream position

    response = send_file(image_io, mimetype='image/png')
    response.headers['Cache-Control'] = 'public, max-age=86400'
    return response


@app.route('/sw.js')
def serve_sw():
    return send_from_directory(os.path.join(app.root_path, 'static'), 'sw.js', mimetype='application/javascript')

# API endpoint for AJAX: returns JSON for a given directory.
@app.route('/api/path/', defaults={'subpath': ''})
@app.route('/api/path/<path:subpath>')
@auth.require_secret
def api_browse(subpath):
    toplist_categories = app_config.get('toplist_categories', None)
    
    if subpath == '': # root directory
        foldernames = []
        for foldername, _ in session['folders'].items():
            foldernames.append({'name': foldername, 'path': foldername})

        return jsonify({
            'breadcrumbs': generate_breadcrumbs(),
            'directories': foldernames,
            'files': [],
            'folder_today': a.return_folder_today(),
            'folder_yesterday': a.return_folder_yesterday(),
            'toplist_enabled': bool(toplist_categories)
        })

    if subpath == 'heute' or subpath == 'gestern':
        foldernames = []
        if len(a.return_folder_today()) > 0:
            for item in a.return_folder_today():
                foldernames.append({'name': item['rel_path'], 'path': item['rel_path']})
        elif len(a.return_folder_yesterday()) > 0:
            for item in a.return_folder_yesterday():
                foldernames.append({'name': item['rel_path'], 'path': item['rel_path']})

        return jsonify({
            'breadcrumbs': generate_breadcrumbs(subpath),
            'directories': foldernames,
            'files': [],
            'folder_today': [],
            'folder_yesterday': []
        })
        
    if subpath.startswith('toplist'):
        foldernames = []
        files = []
        split_path = subpath.split('/')

        if len(split_path) == 1 and split_path[0] == 'toplist':
            foldernames = [
                {
                    'name': categorie,  'path': 'toplist/' + categorie
                } for categorie in toplist_categories
            ]
        elif len(split_path) > 1 and split_path[0] == 'toplist':
            files = hf.generate_top_list(split_path[1])

        return jsonify({
            'breadcrumbs': generate_breadcrumbs(subpath),
            'directories': foldernames,
            'files': files
        })
        

    root, *relative_parts = subpath.split('/')
    base_path = session.get('folders', {}).get(root)
    if not base_path:
        app.logger.warning(f"Requested root '{root}' not found in session folders")
        return jsonify({'error': 'Directory not found'}), 404
    directory = os.path.join(base_path, *relative_parts)
    
    playfile = None
    
    try:
        directory = check_path(directory)
    except (ValueError, PermissionError) as e:
        return jsonify({'error': str(e)}), 403

    # Check if the constructed directory exists.
    if not os.path.isdir(directory):
        # Assume the last segment is a filename; remove it.
        if relative_parts:
            playfile = relative_parts.pop()  # Get the filename.
            directory = os.path.join(base_path, *relative_parts)
            # Rebuild subpath to reflect the directory (without the file).
            subpath = '/'.join([root] + relative_parts)
        # If the parent directory still doesn't exist, return error.
        if not os.path.isdir(directory):
            return jsonify({'error': 'Directory not found'}), 404

    directories, files = list_directory_contents(directory, subpath)
    breadcrumbs = generate_breadcrumbs(subpath)
    
    response = {
        'breadcrumbs': breadcrumbs,
        'directories': directories,
        'files': files,
        'folder_today': a.return_folder_today(),
        'folder_yesterday': a.return_folder_yesterday()
    }
    
    # If a filename was selected include it.
    if playfile:
        response['playfile'] = os.path.join(subpath, playfile).replace(os.sep, '/')

    return jsonify(response)

@app.route("/media/<path:subpath>")
@auth.require_secret
def serve_file(subpath):
    # 1) Locate the real file on disk
    root, *relative_parts = subpath.split('/')

    dltoken = request.args.get('dltoken')
    if dltoken:
        as_attachment = True
        full_path = auth.decode_token(dltoken)['filename']
    else:
        as_attachment = False
        base_path = session['folders'].get(root)
        full_path = os.path.join(base_path or '', *relative_parts)

    try:
        full_path = check_path(full_path)
    except (ValueError, PermissionError) as e:
        return jsonify({'Unauthorized': str(e)}), 403

    if not os.path.isfile(full_path):
        app.logger.error(f"File not found: {full_path}")
        return "File not found", 404
    
    filesize = os.path.getsize(full_path)
    filename = os.path.basename(full_path)

    # 2) Prep request info
    mime, _ = mimetypes.guess_type(full_path)
    mime = mime or 'application/octet-stream'
    is_cache_request = request.headers.get('X-Cache-Request') == 'true'
    is_audio_get     = mime.startswith('audio/') and request.method == 'GET'
    ip_address = request.remote_addr
    user_agent = request.headers.get('User-Agent')
    range_header = request.headers.get('Range', '')
    req_id = request.args.get('req') or request.headers.get('X-Request-Id')

    def is_range_prefetch(header, ua):
        """
        Detect tiny range requests (common Apple prefetch) so we can skip logging duplicates.
        """
        if not header:
            return False
        try:
            if not header.lower().startswith('bytes='):
                return False
            range_spec = header.split('=', 1)[1]
            start_str, end_str = range_spec.split('-', 1)
            if not start_str.isdigit() or not end_str.isdigit():
                return False
            start = int(start_str)
            end = int(end_str)
            length = end - start + 1
            if length <= 1024 and start == 0:
                return True
        except Exception:
            return False
        return False

    # Logging: log every client GET (cached or not), but skip CDN prefetches (X-Cache-Request)
    # and HEAD probes to avoid double-counting. Also skip tiny range-prefetches (e.g., Apple).
    do_log = (
        not is_cache_request    # skip if upstream CDN asked us to cache
        and request.method != 'HEAD'
        and not is_range_prefetch(range_header, user_agent)
    )

    # 3) Pick cache
    if mime.startswith('audio/'):
        cache = cache_audio
    elif mime.startswith('image/'):
        cache = cache_image
    elif mime.startswith('video/'):
        cache = cache_video
    else:
        cache = cache_other

    # Plain HEAD requests (without X-Cache-Request) should not populate the cache.
    # They are just probes and would otherwise turn the first real GET into a “cached hit”.
    if request.method == 'HEAD' and not is_cache_request:
        response = make_response('', 200)
        response.headers['Content-Type'] = mime
        response.headers['Content-Length'] = str(filesize)
        response.headers['Accept-Ranges'] = 'bytes'
        response.headers['Cache-Control'] = 'public, max-age=86400'
        return response

    # 4) Image and thumbnail handling first
    if mime.startswith('image/'):
        small = request.args.get('thumbnail') == 'true'
        name, ext = os.path.splitext(subpath)
        orig_key = subpath
        small_key = f"{name}_small{ext}"
        cache_key = small_key if small else orig_key

        try:
            # Try to read the requested variant
            with cache.read(cache_key) as reader:
                file_path = reader.name
                cached_hit = True
        except KeyError:
            if small: # do not create when thumbnail requested
                response = make_response('', 204)
                return response
            
            cached_hit = False
            # On miss: generate both full-size and small thumb, then cache
            with Image.open(full_path) as orig:

                img = ImageOps.exif_transpose(orig)
                exif_bytes = None
                try:
                    exif = img.getexif()
                    if exif:
                        exif.pop(0x0112, None)  # Drop orientation tag since we already rotated the image
                        exif_bytes = exif.tobytes()
                except Exception:
                    exif_bytes = None
                variants = {
                    orig_key:  (1920, 1920),
                    small_key: ( 480,  480),
                }
                for key, size in variants.items():
                    thumb = img.copy()
                    thumb.thumbnail(size, Image.LANCZOS)
                    if thumb.mode in ("RGBA", "P"):
                        thumb = thumb.convert("RGB")
                    bio = io.BytesIO()
                    save_kwargs = {'format': 'JPEG', 'quality': 85}
                    if key == orig_key and exif_bytes:
                        save_kwargs['exif'] = exif_bytes
                    thumb.save(bio, **save_kwargs)
                    bio.seek(0)
                    cache.set(key, bio, read=True)
            # Read back the variant we need
            with cache.read(cache_key) as reader:
                file_path = reader.name

        # Serve the image variant
        response = send_file(
            file_path,
            mimetype=mime,
            conditional=True,
            as_attachment=(request.args.get('download') == 'true'),
            download_name=os.path.basename(orig_key)
        )
        response.headers['Content-Disposition'] = 'inline'
        response.headers['Cache-Control'] = 'public, max-age=86400'

        if do_log and not small:
            if not _is_duplicate_request(req_id):
                a.log_file_access(
                    cache_key,
                    os.path.getsize(file_path),
                    mime,
                    ip_address,
                    user_agent,
                    session['device_id'],
                    cached_hit,
                    request.method
                )
                _mark_request_logged(req_id)
        return response

    # 5) Non-image branch: check if cached; if not, fill cache synchronously and then serve from cache
    try:
        with cache.read(subpath) as reader:
            file_path = reader.name
            cached_hit = True
    except KeyError:
        cached_hit = False
        try:
            # Copy full file into diskcache (blocks the request; avoids serving partial content)
            with open(full_path, 'rb') as source:
                cache.set(subpath, source, read=True)
            with cache.read(subpath) as reader:
                file_path = reader.name
            app.logger.info(f"Cached {subpath}")
        except Exception as e:
            app.logger.error(f"Cache fill failed for {subpath}: {e}")
            abort(503, description="Service temporarily unavailable - cache population failed")

    # 6) Build response for non-image
    if as_attachment:
        download_name = filename
        mimetype = 'application/octet-stream'
    else:
        download_name = None
        mimetype = mime
        

    # Single send_file call with proper attachment handling
    response = send_file(
        file_path,
        mimetype=mimetype,
        conditional=True,
        as_attachment=as_attachment,
        download_name=filename if as_attachment else None
    )
    
    if as_attachment:
        response.headers['X-Content-Type-Options'] = 'nosniff'
        response.headers['Content-Disposition'] = 'attachment'
    else:
        response.headers['Content-Disposition'] = 'inline'

    response.headers['Cache-Control'] = 'public, max-age=86400'
    
    if request.method == 'HEAD':
        response.set_data(b'')

    # 7) Logging
    if do_log:
        if not _is_duplicate_request(req_id):
            a.log_file_access(
                subpath,
                filesize,
                mime,
                ip_address,
                user_agent,
                session['device_id'],
                cached_hit,
                request.method
            )
            _mark_request_logged(req_id)
    return response


@app.route("/media_exif/<path:subpath>")
@auth.require_secret
def media_exif(subpath):
    root, *relative_parts = subpath.split('/')
    base_path = session.get('folders', {}).get(root)
    if not base_path:
        return jsonify({'error': 'Directory not found'}), 404

    full_path = os.path.join(base_path, *relative_parts)
    try:
        full_path = check_path(full_path)
    except (ValueError, PermissionError) as e:
        return jsonify({'error': str(e)}), 403

    if not os.path.isfile(full_path):
        return jsonify({'error': 'File not found'}), 404

    mime, _ = mimetypes.guess_type(full_path)
    if not mime or not mime.startswith('image/'):
        return jsonify({'error': 'Not an image'}), 400

    try:
        with Image.open(full_path) as img:
            exif_data = img.getexif()
            if not exif_data:
                return jsonify({'exif': {}})

            readable = {}

            ignore_tags = {'Orientation', 'ExifOffset', 'MakerNote', 'PrintImageMatching'}

            def add_items(exif_dict):
                for tag_id, value in exif_dict.items():
                    tag = ExifTags.TAGS.get(tag_id, str(tag_id))
                    if tag in ignore_tags:
                        continue
                    try:
                        if isinstance(value, bytes):
                            value = value.decode('utf-8', errors='ignore').strip()
                    except Exception:
                        pass
                    readable[tag] = str(value)

            add_items(exif_data)

            # Include Exif and GPS sub-IFDs when available to surface exposure/ISO/lens details.
            try:
                if hasattr(ExifTags, 'IFD'):
                    for ifd in (ExifTags.IFD.Exif, ExifTags.IFD.GPSInfo):
                        try:
                            sub_ifd = exif_data.get_ifd(ifd)
                            add_items(sub_ifd)
                        except Exception:
                            continue
            except Exception:
                pass
    except Exception as e:
        app.logger.warning(f"EXIF read failed for {subpath}: {e}")
        return jsonify({'exif': {}})

    return jsonify({'exif': readable})



@app.route("/transcript/<path:subpath>")
@auth.require_secret
def get_transcript(subpath):
    
    root, *relative_parts = subpath.split('/')
    base_path = session['folders'][root]
    full_path = os.path.join(base_path, *relative_parts)
    
    if not os.path.isfile(full_path):
        return "Transcription not found", 404

    with open(full_path, 'r', encoding='utf-8') as f:
        content = f.read()
    return content, 200, {'Content-Type': 'text/markdown; charset=utf-8'}


@app.route("/create_token/<path:subpath>")
@auth.require_secret
def create_token(subpath):
    scheme = request.scheme  # current scheme (http or https)
    host = request.host
    if 'admin' not in session and not session.get('admin'):
        return "Unauthorized", 403

    folder_config = auth.return_folder_config()
    paths = {}
    for item in folder_config:
        for folder in item['folders']:
            paths[folder['foldername']] = folder['folderpath']

    # use folder config file and ignore validity to get full path    
    root, *relative_parts = subpath.split('/')
    base_path = paths[root]
    full_path = os.path.join(base_path, *relative_parts)
    foldername = relative_parts[-1]
    validity_date = (datetime.now() + timedelta(days=90)).strftime('%d.%m.%Y')
    data = {
        "validity": validity_date,
        "folders": [
            {
                "foldername": foldername,
                "folderpath": full_path
            }
        ]
    }
    
    token = auth.generate_token(data)

    url = f"{scheme}://{host}?token={token}"
    qr = qrcode.QRCode(version=1, box_size=10, border=4)
    qr.add_data(url)
    qr.make(fit=True)
    img = qr.make_image(fill_color="black", back_color="white")
    buffer = io.BytesIO()
    img.save(buffer, format="PNG")
    buffer.seek(0)
    img_base64 = base64.b64encode(buffer.getvalue()).decode('ascii')
    token_item = auth.decode_token(token)

    return render_template('view_token.html',
                            token_qr_code=img_base64,
                            token_folder=token_item.get('folders'),
                            token_url=url,
                            token_valid_to=token_item.get('validity', 'Unbekannt'),
                            title_short=auth.return_app_config().get('TITLE_SHORT', ''),
                            title_long=auth.return_app_config().get('TITLE_LONG', '')
                           )


@app.route("/create_dltoken/<path:subpath>")
@auth.require_secret
def create_dltoken(subpath):
    scheme = request.scheme  # current scheme (http or https)
    host = request.host
    # 1) Locate the real file on disk
    root, *relative_parts = subpath.split('/')
    base_path = session['folders'].get(root)
    full_path = os.path.join(base_path or '', *relative_parts)
    
    try:
        full_path = check_path(full_path)
    except (ValueError, PermissionError) as e:
        return jsonify({'Unauthorized': str(e)}), 403

    if not os.path.isfile(full_path):
        app.logger.error(f"File not found: {full_path}")
        return "File not found", 404

    validity_date = datetime.now().strftime('%d.%m.%Y')
    data = {
        "validity": validity_date,
        "filename": str(full_path)
    }
    
    token = auth.generate_token(data)
    url = f"{scheme}://{host}/media/{subpath}?dltoken={token}"
    return url


def query_recent_connections():
    global clients_connected, background_thread_running
    background_thread_running = True
    last_connections = None
    try:
        while clients_connected > 0:
            rows = a.return_file_access()
            connections = [
                {
                    'timestamp': datetime.fromisoformat(row[0]).strftime('%d.%m.%Y %H:%M:%S'),
                    'full_path': row[1],
                    'filesize': row[2],
                    'filesize_human': human_readable_size(row[2]),
                    'mime_typ': row[3],
                    'location': row[4],
                    'user_agent': row[5],
                    'cached': row[7]
                }
                for row in rows
            ]

            if connections != last_connections:
                socketio.emit('recent_connections', connections)
                last_connections = connections.copy()

            socketio.sleep(1)
    finally:
        background_thread_running = False


@socketio.on('connect')
def handle_connect(auth=None):
    global clients_connected, background_thread_running
    clients_connected += 1
    print("Client connected. Total clients:", clients_connected)
    with thread_lock:
        if not background_thread_running:
            socketio.start_background_task(query_recent_connections)
            print("Started background query task.")

@socketio.on('disconnect')
def handle_disconnect():
    global clients_connected
    clients_connected -= 1
    print("Client disconnected. Total clients:", clients_connected)

@socketio.on('request_initial_data')
def handle_request_initial_data():
    rows = a.return_file_access()
    connections = [
        {
            'timestamp': datetime.fromisoformat(row[0]).strftime('%d.%m.%Y %H:%M:%S'),
            'full_path': row[1],
            'filesize' : row[2],
            'filesize_human': human_readable_size(row[2]),
            'mime_typ' : row[3],
            'location': row[4],
            'user_agent': row[5],
            'cached':  row[7]
        }
        for row in rows
    ]
    emit('recent_connections', connections)

@socketio.on('request_map_data')
def handle_request_map_data():
    """Send initial map data from in-memory log to avoid hitting the DB."""
    rows = a.return_file_access()
    connections = []
    for row in rows:
        # lat/lon are appended at indexes 10/11 in analytics.log_file_access
        lat = row[10] if len(row) > 10 else None
        lon = row[11] if len(row) > 11 else None
        city = row[8] if len(row) > 8 else None
        country = row[9] if len(row) > 9 else None
        if lat and lon:
            connections.append({
                'timestamp': row[0],
                'city': city,
                'country': country,
                'lat': lat,
                'lon': lon
            })
    emit('map_initial_data', {'connections': connections})

# Catch-all route to serve the single-page application template.
@app.route('/', defaults={'path': ''})
@app.route('/<path:path>')
@auth.require_secret
def index(path):
    app_config = auth.return_app_config()
    
    # Generate dynamic Open Graph meta tags for Telegram preview
    og_title = app_config.get('TITLE_LONG', 'Default Title')
    og_description = "... uns aber, die wir gerettet werden, ist es eine Gotteskraft."
    
    # If secret or token is provided, show folder names in description
    folder_names = list(session['folders'].keys())
    if folder_names:
        folder_list = ", ".join(folder_names)
        og_description = folder_list

    return render_template("app.html",
                           search_folders = folder_names,
                            title_short=app_config.get('TITLE_SHORT', 'Default Title'),
                            title_long=app_config.get('TITLE_LONG' , 'Default Title'),
                            features=app_config.get('FEATURES', None),
                            og_title=og_title,
                            og_description=og_description,
                            admin_enabled=auth.is_admin()
                           )

if __name__ == '__main__':
    socketio.run(app, debug=True, host='0.0.0.0')