#!/usr/bin/env python3

import subprocess
import re
import time
import argparse

# ZFS-Pfad ermitteln
zfs = subprocess.check_output(["which", "zfs"]).decode().strip()

# Argumente verarbeiten
parser = argparse.ArgumentParser(description="ZFS Snapshot Übersicht")
parser.add_argument("--filter", help="Nur bestimmte Datasets anzeigen (Regex möglich, z.B. 'rpool/ROOT')", type=str)
args = parser.parse_args()

# Snapshots abrufen
_snapshots = subprocess.check_output([zfs, "list", "-t", "snapshot", "-Hpo", "name,creation"])

LABELS = ("frequent", "hourly", "daily", "weekly", "monthly", "yearly", "backup-zfs", "bashclub-zfs")
RE_LABELSEARCH = re.compile("|".join(LABELS))
_datasets = {}

# Snapshots parsen
for _datastore, _snapshot, _creation in re.findall(r"^([\w_./-]+)@([\w_.-]+)\t(\d+)", _snapshots.decode('utf-8'), re.M):
    if args.filter and not re.search(args.filter, _datastore):
        continue  # Dataset entspricht nicht dem Filter

    if _datastore not in _datasets:
        _datasets[_datastore] = {}
    _label = RE_LABELSEARCH.search(_snapshot)
    _label = _label.group(0) if _label else "other"

    if _label not in _datasets[_datastore]:
        _datasets[_datastore][_label] = []
    _datasets[_datastore][_label].append((_snapshot, int(_creation)))

# Ergebnisse anzeigen
for _datastore in _datasets:
    print(_datastore)
    print("-" * 40)
    for _label in _datasets[_datastore]:
        _data = _datasets[_datastore][_label]
        _first = time.strftime("%d.%m.%Y %H:%M:%S", time.localtime(_data[0][1]))
        _last = time.strftime("%d.%m.%Y %H:%M:%S", time.localtime(_data[-1][1]))
        _count = len(_data)
        print(f" {_label} {_count}")
        print(f"   {_first} {_data[0][0]}")
        if _count > 1:
            print(f"   {_last} {_data[-1][0]}")
    print("")