134 lines
4.6 KiB
Python
134 lines
4.6 KiB
Python
import os
|
|
import blake3
|
|
import sys
|
|
import json
|
|
from tqdm import tqdm
|
|
|
|
# ==============================
|
|
# ⚙️ NASTAVENÍ
|
|
# ==============================
|
|
LOCAL_ROOT = r"U:\#Syncthing"
|
|
CACHE_FILE = "local_hashes.jsonl"
|
|
ERROR_LOG = "errors_scan.txt"
|
|
|
|
|
|
def calculate_blake3_hex(file_path):
|
|
h = blake3.blake3()
|
|
# Zde ošetříme chyby při otevírání/čtení souboru
|
|
try:
|
|
with open(file_path, "rb") as f:
|
|
while True:
|
|
chunk = f.read(4 * 1024 * 1024)
|
|
if not chunk: break
|
|
h.update(chunk)
|
|
return h.hexdigest()
|
|
except Exception as e:
|
|
# Vyhodíme výjimku nahoru, abychom ji zachytili v hlavní smyčce i s názvem souboru
|
|
raise e
|
|
|
|
|
|
def count_files(path):
|
|
print("⏳ Sčítám soubory...", end="\r")
|
|
total = 0
|
|
for root, dirs, files in os.walk(path):
|
|
if ".stfolder" in dirs: dirs.remove(".stfolder")
|
|
if ".stversions" in dirs: dirs.remove(".stversions")
|
|
total += len(files)
|
|
print(f"✅ Celkem souborů: {total:,}")
|
|
return total
|
|
|
|
|
|
def load_processed_files(cache_file):
|
|
"""Načte cesty souborů, které už máme hotové"""
|
|
processed = set()
|
|
if os.path.exists(cache_file):
|
|
print("📥 Načítám již hotové záznamy pro navázání...")
|
|
try:
|
|
with open(cache_file, "r", encoding="utf-8") as f:
|
|
for line in f:
|
|
line = line.strip()
|
|
if not line: continue
|
|
try:
|
|
data = json.loads(line)
|
|
processed.add(data["p"]) # "p" je cesta
|
|
except:
|
|
continue
|
|
except Exception as e:
|
|
print(f"⚠️ Chyba při čtení cache (nevadí, jedeme dál): {e}")
|
|
|
|
print(f"⏭ Přeskočím {len(processed):,} již hotových souborů.")
|
|
return processed
|
|
|
|
|
|
def main():
|
|
print("🚀 KROK 1: Generování cache (RESUMABLE + SAFE MODE)")
|
|
print(f"📂 Zdroj: {LOCAL_ROOT}")
|
|
print(f"💾 Výstup: {CACHE_FILE}")
|
|
print("===================================")
|
|
|
|
# 1. Zjistit celkový počet
|
|
total_files = count_files(LOCAL_ROOT)
|
|
|
|
# 2. Zjistit, co už máme hotové
|
|
processed_paths = load_processed_files(CACHE_FILE)
|
|
|
|
errors = 0
|
|
newly_processed = 0
|
|
|
|
# Otevřeme v režimu 'a' (APPEND) - připisování na konec
|
|
with open(CACHE_FILE, "a", encoding="utf-8") as f_out:
|
|
|
|
with tqdm(total=total_files, unit="file", dynamic_ncols=True) as pbar:
|
|
|
|
# Pokud už máme nějaká data, aktualizujeme progress bar
|
|
if processed_paths:
|
|
pbar.update(len(processed_paths))
|
|
|
|
for root, dirs, files in os.walk(LOCAL_ROOT):
|
|
if ".stfolder" in dirs: dirs.remove(".stfolder")
|
|
if ".stversions" in dirs: dirs.remove(".stversions")
|
|
|
|
for filename in files:
|
|
file_path = os.path.join(root, filename)
|
|
|
|
# === SKIPPER: Pokud už to máme, jdeme dál ===
|
|
if file_path in processed_paths:
|
|
continue
|
|
# ============================================
|
|
|
|
# Pro dlouhé názvy ořízneme text v progress baru
|
|
short_name = (filename[:30] + '...') if len(filename) > 30 else filename
|
|
pbar.set_description(f"Hash: {short_name}")
|
|
|
|
# === SAFE BLOCK: Try-Except uvnitř smyčky ===
|
|
try:
|
|
hex_hash = calculate_blake3_hex(file_path)
|
|
|
|
if hex_hash:
|
|
record = {"h": hex_hash, "p": file_path}
|
|
f_out.write(json.dumps(record, ensure_ascii=False) + "\n")
|
|
# Flush, aby se to zapsalo na disk hned (pro jistotu)
|
|
if newly_processed % 100 == 0:
|
|
f_out.flush()
|
|
newly_processed += 1
|
|
|
|
except Exception as e:
|
|
errors += 1
|
|
error_msg = f"❌ CHYBA: {filename} -> {e}\n"
|
|
# Zapíšeme chybu do logu, aby to nerušilo output
|
|
with open(ERROR_LOG, "a", encoding="utf-8") as ferr:
|
|
ferr.write(error_msg)
|
|
|
|
# Posuneme progress bar
|
|
pbar.update(1)
|
|
|
|
print("\n\n✅ HOTOVO.")
|
|
print(f"📄 Nově přidáno: {newly_processed}")
|
|
print(f"⏭ Přeskočeno: {len(processed_paths)}")
|
|
|
|
if errors > 0:
|
|
print(f"⚠️ Počet chyb: {errors} (Detaily viz soubor '{ERROR_LOG}')")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main() |