bethaus-app/helperfunctions.py

from flask import session
import re
import os
import sqlite3
from datetime import datetime, timedelta
from typing import Optional
import auth

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

log_db = sqlite3.connect("access_log.db", check_same_thread=False)



# Precompiled regex to find date-like patterns: either dotted X.Y.Z or ISO dashed YYYY-MM-DD
_DATE_REGEX = re.compile(
    r"("  # start group
    r"\d{4}-\d{1,2}-\d{1,2}"          # ISO: YYYY-M-D or YYYY-MM-DD
    r"|"                              # or
    r"\d{1,4}\.\d{1,2}\.\d{1,4}"      # dotted: X.Y.Z, where each is 1–4 digits (year may be 1–4)
    r")"
)


def _try_parse(date_str: str, fmt: str) -> Optional[datetime]:
    """Try to parse date_str with fmt, return datetime or None."""
    try:
        return datetime.strptime(date_str, fmt)
    except ValueError:
        return None


def extract_date_from_string(text: str) -> Optional[str]:
    """
    Extract the first date-like substring from text and return it in ISO format (YYYY-MM-DD).
    Supports:
    - ISO-style dates with dashes (YYYY-M-D or YYYY-MM-DD)
    - Dotted dates (DD.MM.YYYY, YYYY.MM.DD, DD.MM.YY, YY.MM.DD)
    """
    match = _DATE_REGEX.search(text)
    if not match:
        return None

    date_str = match.group(1)

    # 1) ISO dashed format takes priority
    if '-' in date_str:
        dt = _try_parse(date_str, '%Y-%m-%d')
        return dt.strftime('%Y-%m-%d') if dt else None

    # 2) Dotted formats
    parts = date_str.split('.')
    candidates = []

    # Unambiguous: last part 4 digits → DD.MM.YYYY
    if len(parts) == 3 and len(parts[2]) == 4:
        candidates.append('%d.%m.%Y')
    # Unambiguous: first part 4 digits → YYYY.MM.DD
    if len(parts) == 3 and len(parts[0]) == 4:
        candidates.append('%Y.%m.%d')
    # Ambiguous two-digit groups: try DD.MM.YY, then YY.MM.DD
    if len(parts) == 3 and all(len(p) == 2 for p in parts):
        candidates.extend(['%d.%m.%y', '%y.%m.%d'])

    # Try each candidate
    for fmt in candidates:
        dt = _try_parse(date_str, fmt)
        if dt:
            return dt.strftime('%Y-%m-%d')

    # no valid parse
    return None


def extract_structure_from_string(input_string):
    # extract category and titel from filename
    filepathname_ext = os.path.splitext(input_string)[0] # remove file extension
    filename_ext = os.path.basename(filepathname_ext) # get only the filename
    left_side, right_side = filename_ext.split('-', 1) if '-' in filename_ext else (filename_ext, None)

    try:
        int(left_side.strip())
        # first part is only a number
        previous_right_side = right_side
        left_side, right_side = previous_right_side.split('-', 1) if '-' in previous_right_side else (previous_right_side, None)
    except:
        # first part not a number
        pass
    
    # define your mapping: category → list of trigger-words
    CATEGORY_KEYWORDS = {
        'Predigt':              ['predig', 'thema'],
        'Vorwort':              ['wort', 'einladung', 'begrüßung', 'ansprache', 'einleitung', 'aufruf zum', 'zuruf zum'],
        'Kinderchor':           ['kinderchor'],
        'Jugendchor':           ['jugendchor'],
        'Orchester':            ['orchester', 'sinfonie', 'symphonie'],
        'Chor':                 ['chor'],
        'Gemeinsamer Gesang':   ['gemeinsam', 'gemeindelied', 'gemeinsamer gesang'],
        'Gruppenlied':          ['gruppenlied', 'jugend', 'lied', 'musikgruppe'],
        'Gedicht':              ['gedicht'],
        'Erzählung':            ['vortrag', 'erzä', 'program'],
        'Instrumental':         ['instrumental', 'musikstück', 'harfenstück'],
    }

    # walk the dict in your desired priority order
    category = None
    text = left_side.lower()
    for cat, keywords in CATEGORY_KEYWORDS.items():
        if any(kw in text for kw in keywords):
            category = cat
            break

    if right_side:
        titel, name = right_side.split('-', 1) if '-' in right_side else (right_side, None)
        if category == 'Predigt' or category == 'Vorwort' or category == 'Gedicht':
            if not name: # kein Titel, nur Name
                name = titel
                titel = None
    else:
        titel = None
        name = None

    return category, titel, name

def generate_top_list(category):
    
    now = datetime.now()

    # We'll compare the timestamp
    start_dt = now - timedelta(days=14)
    start_str = start_dt.isoformat()
    
    # Filter for mimes that start with the given type
    params_for_filter = (start_str,)

    # 1. Top files by access count
    query = f'''
        SELECT rel_path, COUNT(*) as access_count
        FROM file_access_log
        WHERE timestamp >= ? AND mime LIKE 'audio/%'
        GROUP BY rel_path
        ORDER BY access_count DESC
        LIMIT 1000
    '''
    with log_db:
        cursor = log_db.execute(query, params_for_filter)
        rows = cursor.fetchall()

    # Filter by allowed base folders
    allowed_basefolders = list(session['folders'].keys())
    rows = [
        (rel_path, access_count) for rel_path, access_count in rows
        if any(rel_path.startswith(folder) for folder in allowed_basefolders)
    ]

    # Convert rows to a list of dicts and add category
    records = [
        {
            'rel_path': rel_path,
            'access_count': access_count,
            'category': extract_structure_from_string(rel_path)[0]
        }
        for rel_path, access_count in rows
    ]
    # Filter by requested category and limit
    records = [r for r in records if r['category'] == category][:20]

    # Build file list and check existence
    filelist = []
    for record in records:
        rel_path = record['rel_path']
        if os.path.exists(os.path.join(BASE_DIR, rel_path)):  # ensure file exists on disk // slow operation. maybe improve later
            filelist.append({
                'name': os.path.basename(rel_path),
                'path': rel_path,
                'file_type': 'music'
            })

    return filelist