# ============================================================
#  seaweed_backfill_v1.0.py
#  Verze:  1.0
#  Datum:  2026-06-15
#  Popis:  Jednorázový backfill — nahraje do SeaweedFS Filer
#          všechny dokumenty z VTMF.documents, které jsou na disku
#          (downloaded=True, file!=null) ale ještě nemají seaweed_path.
#          Placeholdery a záznamy bez souboru přeskočí.
#          Lze spustit opakovaně — HEAD check zajistí dedup,
#          přerušení kdykoli naváže příště.
# ============================================================

import hashlib
import mimetypes
import sys
import urllib.error
import urllib.request
from datetime import datetime
from pathlib import Path

from pymongo import MongoClient, ASCENDING

MONGO_URI  = "mongodb://192.168.1.76:27017"
MONGO_DB   = "VTMF"
MONGO_COLL = "documents"

SEAWEED_FILER  = "http://192.168.1.50:8888"
SEAWEED_PREFIX = "/vtmf-documents"


def log(msg):
    print(msg, flush=True)


def sw_path(sha256):
    return f"{SEAWEED_PREFIX}/{sha256[:2]}/{sha256[2:4]}/{sha256}"


def seaweed_store(data, mime="application/octet-stream"):
    """HEAD check + PUT. Vrací (path, url, uploaded)."""
    sha256 = hashlib.sha256(data).hexdigest()
    path = sw_path(sha256)
    url  = SEAWEED_FILER + path

    try:
        urllib.request.urlopen(
            urllib.request.Request(url, method="HEAD"), timeout=10)
        return path, url, False  # dedup hit
    except urllib.error.HTTPError as e:
        if e.code != 404:
            raise

    urllib.request.urlopen(
        urllib.request.Request(url, data=data, method="PUT",
                               headers={"Content-Type": mime}),
        timeout=120)
    return path, url, True


def main():
    client = MongoClient(MONGO_URI, serverSelectionTimeoutMS=5000)
    client.admin.command("ping")
    coll = client[MONGO_DB][MONGO_COLL]
    log(f"[ok] Mongo připojeno: {MONGO_URI} / {MONGO_DB}.{MONGO_COLL}")

    query = {
        "downloaded": True,
        "placeholder": {"$ne": True},
        "seaweed_path": None,
        "file": {"$ne": None},
    }
    todo = list(coll.find(query).sort([("vtmf", ASCENDING), ("version", ASCENDING)]))
    log(f"[i] Ke zpracování: {len(todo)} dokumentů\n")

    uploaded = dedup = skipped = failed = 0

    for n, doc in enumerate(todo, 1):
        key  = doc["_id"]
        path = doc.get("file")
        log(f"[{n}/{len(todo)}] {key}")

        if not path or not Path(path).exists():
            log(f"  [!] Soubor nenalezen na disku: {path} — přeskočeno.")
            skipped += 1
            continue

        try:
            data = Path(path).read_bytes()
            mime = mimetypes.guess_type(path)[0] or "application/octet-stream"
            sha256_hex = hashlib.sha256(data).hexdigest()

            sw_p, sw_url, was_new = seaweed_store(data, mime)

            coll.update_one({"_id": key}, {"$set": {
                "sha256":            sha256_hex,
                "seaweed_path":      sw_p,
                "seaweed_url":       sw_url,
                "seaweed_synced_at": datetime.now(),
            }})

            if was_new:
                uploaded += 1
                log(f"  [ok] Nahráno → {sw_p}")
            else:
                dedup += 1
                log(f"  [i]  Dedup hit → {sw_p}")

        except Exception as e:
            failed += 1
            log(f"  [!] Chyba: {e}")

    log(f"\n{'='*60}")
    log(f"  Hotovo: {uploaded} nahráno, {dedup} dedup, "
        f"{skipped} bez souboru, {failed} chyb.")
    log(f"{'='*60}")

    sys.exit(1 if failed else 0)


if __name__ == "__main__":
    main()