55 lines
1.8 KiB
Python
55 lines
1.8 KiB
Python
import os
|
|
import tempfile
|
|
import pyzipper
|
|
|
|
from indexer.config import BACKUP_PASSWORD
|
|
|
|
|
|
def blob_path(backup_root: str, content_hash: bytes) -> str:
|
|
"""Vrátí cestu k ZIP souboru: BACKUP/ab/cd/abcdef...zip"""
|
|
hex_hash = content_hash.hex()
|
|
return os.path.join(backup_root, hex_hash[:2], hex_hash[2:4], hex_hash + ".zip")
|
|
|
|
|
|
def ensure_backed_up(files_with_hash: list, backup_root: str) -> int:
|
|
"""
|
|
Vytvoří AES-256 šifrovaný ZIP pro každý soubor v content-addressable storage.
|
|
files_with_hash: [(full_path, content_hash_bytes), ...]
|
|
Přeskočí soubory, jejichž zip už existuje (deduplikace).
|
|
Returns: počet nově zálohovaných souborů.
|
|
"""
|
|
password = BACKUP_PASSWORD.encode("utf-8")
|
|
backed_up = 0
|
|
for full_path, content_hash in files_with_hash:
|
|
target = blob_path(backup_root, content_hash)
|
|
if os.path.exists(target):
|
|
continue
|
|
|
|
target_dir = os.path.dirname(target)
|
|
os.makedirs(target_dir, exist_ok=True)
|
|
|
|
tmp_path = None
|
|
try:
|
|
# Atomický zápis: temp soubor + přejmenování
|
|
fd, tmp_path = tempfile.mkstemp(dir=target_dir, suffix=".tmp")
|
|
os.close(fd)
|
|
|
|
hex_hash = content_hash.hex()
|
|
with pyzipper.AESZipFile(
|
|
tmp_path, "w",
|
|
compression=pyzipper.ZIP_DEFLATED,
|
|
encryption=pyzipper.WZ_AES,
|
|
) as zf:
|
|
zf.setpassword(password)
|
|
zf.write(full_path, arcname=hex_hash + ".blob")
|
|
|
|
os.replace(tmp_path, target)
|
|
backed_up += 1
|
|
except (FileNotFoundError, PermissionError, OSError) as e:
|
|
print(f" WARN: backup failed for {full_path}: {e}")
|
|
if tmp_path and os.path.exists(tmp_path):
|
|
os.remove(tmp_path)
|
|
continue
|
|
|
|
return backed_up
|