pcvisit-filetracker/archive_logs.py

import shutil
import xml.etree.ElementTree as ET
from dataclasses import dataclass
from datetime import datetime, timedelta
from pathlib import Path


@dataclass(frozen=True)
class ArchiveLogsResult:
    checked_count: int
    archived_count: int
    skipped_count: int


def find_xml_files(directory: Path) -> list[Path]:
    archive_directory = directory / "Archiv"
    return sorted(
        path
        for path in directory.rglob("*")
        if path.is_file()
        and path.suffix.lower() == ".xml"
        and not path.name.lower().startswith("sessions")
        and archive_directory not in path.parents
    )


def get_archive_year(xml_path: Path, now: datetime | None = None) -> int | None:
    try:
        root = ET.parse(xml_path).getroot()
    except ET.ParseError:
        return None

    session = root.find("Session")
    if session is None:
        return None

    if session.get("Duration", "").strip():
        start_time = session.get("StartTime", "")
        try:
            return datetime.fromisoformat(start_time.replace("Z", "+00:00")).year
        except ValueError:
            return None

    created_at = datetime.fromtimestamp(xml_path.stat().st_birthtime)
    now = now or datetime.now()
    if now - created_at >= timedelta(days=30):
        return created_at.year
    return None


def remove_empty_directories(directory: Path) -> int:
    archive_directory = directory / "Archiv"
    removed_count = 0
    subdirectories = sorted(
        (path for path in directory.rglob("*") if path.is_dir()),
        key=lambda path: len(path.parts),
        reverse=True,
    )
    for subdirectory in subdirectories:
        if subdirectory == archive_directory or archive_directory in subdirectory.parents:
            continue
        try:
            subdirectory.rmdir()
        except OSError:
            continue
        removed_count += 1
    return removed_count


def run(directory: Path) -> ArchiveLogsResult:
    if not directory.is_dir():
        raise ValueError(f"Das Verzeichnis existiert nicht: {directory}")

    checked_count = 0
    archived_count = 0
    skipped_count = 0
    for xml_file in find_xml_files(directory):
        checked_count += 1
        year = get_archive_year(xml_file)
        if year is None:
            skipped_count += 1
            continue

        target_directory = directory / "Archiv" / str(year)
        target_directory.mkdir(parents=True, exist_ok=True)
        target = target_directory / xml_file.name
        if target.exists():
            skipped_count += 1
            continue
        shutil.move(str(xml_file), str(target))
        archived_count += 1

    remove_empty_directories(directory)
    return ArchiveLogsResult(
        checked_count=checked_count,
        archived_count=archived_count,
        skipped_count=skipped_count,
    )