janssen/Covance/materialize_lab_reports_v1.0.py

# =============================================================================
# Název:   materialize_lab_reports_v1.0.py
# Verze:   1.0
# Datum:   2026-06-16
# Popis:   Materializuje Lab Reports z MongoDB (covance.labreports) do Dropboxu.
#          Cíl: adresář = IDENTICKÁ KOPIE Monga (plné zrcadlo) — doplní chybějící
#          PDF a SMAŽE přebytečné. Žádná jiná ambice.
#
#          Struktura: <base studie>/<centrum siteNum>/<fileName>.pdf
#          (podsložka per centrum, do ní všechna PDF toho centra).
#
#          Klíč dokumentu = record_id (reportId). Název souboru = doc.fileName.
#          Kdyz vic ruznych reportu (ruzne reportId) ma v jednom centru stejny
#          fileName, prida se ke VSEM z te skupiny " [<reportId[:8]>]" -> vzdy
#          unikatni a stabilni (idempotentni zrcadlo).
#
#          Efektivita: cte z Monga jen metadata; PDF binarku tahá jen pro
#          soubory, ktere chybi (existujici se shodnou velikosti se preskoci,
#          obsah reportu je v case nemenny). --verify navic kontroluje sha256.
#
#          Pojistka: pokud Mongo vrati pro studii 0 dokumentu, studie se
#          PRESKOCI (nesmaze cely adresar kvuli vypadku DB) — obejde --allow-empty.
#
#          Prepinace: --dry-run (jen nahled, nic nezapise/nesmaze),
#                     --verify (porovnat i sha256), --study <kod> (jen jedna),
#                     --allow-empty (povolit zrcadleni i kdyz Mongo vrati 0).
# =============================================================================
from pymongo import MongoClient
import argparse
import hashlib
import os

# ---------------------------------------------------------------------------
# Cilove cesty v Dropboxu — klic = studyCode (jak je v Mongo).
# ---------------------------------------------------------------------------
STUDY_PATHS = {
    "77242113UCO3001": r"u:\Dropbox\!77242113UCO3001\#020 Reporty\LabCorp",
    "42847922MDD3003": r"u:\Dropbox\!!42847922MDD3003\#000 Reports\LabCorp",
}

MONGO_URI  = "mongodb://192.168.1.76:27017"
DB_NAME    = "covance"
COLLECTION = "labreports"

parser = argparse.ArgumentParser(description="Materializace Lab Reports z Mongo do Dropboxu (zrcadlo).")
parser.add_argument("--dry-run", action="store_true", help="jen nahled, nic nezapise/nesmaze")
parser.add_argument("--verify", action="store_true", help="u existujicich porovnat i sha256 (jinak jen velikost)")
parser.add_argument("--study", default=None, help="zpracovat jen jednu studii (studyCode)")
parser.add_argument("--allow-empty", action="store_true", help="povolit zrcadleni i kdyz Mongo vrati 0 dokumentu")
ARGS = parser.parse_args()


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


def safe(s: str) -> str:
    import re
    return re.sub(r'[\\/:*?"<>|]', "", s or "").strip()


def sha256_file(path: str) -> str:
    h = hashlib.sha256()
    with open(path, "rb") as f:
        for chunk in iter(lambda: f.read(1 << 20), b""):
            h.update(chunk)
    return h.hexdigest()


def plan_study(col, study_code: str, base: str):
    """Spocita ocekavane soubory (cesta -> doc-meta) + kontrola kolizi nazvu."""
    docs = list(col.find(
        {"studyCode": study_code},
        {"record_id": 1, "site": 1, "fileName": 1, "pdfSize": 1, "pdfSha256": 1},
    ))
    # skupiny (centrum, fileName) kvuli kolizim
    groups = {}
    for d in docs:
        key = (safe(d.get("site") or "_"), d.get("fileName") or (d["record_id"] + ".pdf"))
        groups.setdefault(key, []).append(d)

    expected = {}  # abs_path -> doc
    for (center, fname), members in groups.items():
        collide = len(members) > 1
        for d in members:
            name = fname
            if collide:
                stem, ext = (name[:-4], name[-4:]) if name.lower().endswith(".pdf") else (name, "")
                name = f"{stem} [{d['record_id'][:8]}]{ext}"
            path = os.path.join(base, center, name)
            expected[os.path.normpath(path)] = d
    return docs, expected


def existing_pdfs(base: str):
    out = set()
    for root, _dirs, files in os.walk(base):
        for fn in files:
            if fn.lower().endswith(".pdf"):
                out.add(os.path.normpath(os.path.join(root, fn)))
    return out


def needs_write(path: str, doc, verify: bool) -> bool:
    if not os.path.exists(path):
        return True
    if os.path.getsize(path) != doc.get("pdfSize", -1):
        return True
    if verify and sha256_file(path) != doc.get("pdfSha256"):
        return True
    return False


def process_study(col, study_code: str, base: str):
    log(f"\n=== {study_code} -> {base} ===")
    if not os.path.isdir(base):
        log(f"  CHYBA: cilovy adresar neexistuje: {base} — PRESKAKUJI.")
        return
    docs, expected = plan_study(col, study_code, base)
    log(f"  Mongo: {len(docs)} dokumentu, ocekavanych souboru: {len(expected)}.")

    if len(docs) == 0 and not ARGS.allow_empty:
        log(f"  POJISTKA: Mongo vratilo 0 dokumentu -> studie PRESKOCENA (nemazu nic). "
            f"Pouzij --allow-empty kdyz je to zamer.")
        return

    have = existing_pdfs(base)
    to_write = [(p, d) for p, d in expected.items() if needs_write(p, d, ARGS.verify)]
    to_delete = sorted(have - set(expected.keys()))

    skip = len(expected) - len(to_write)
    log(f"  PLAN: +{len(to_write)} zapsat, -{len(to_delete)} smazat, ={skip} beze zmeny.")

    if ARGS.dry_run:
        for p, _ in to_write[:10]:
            log(f"    [DRY +] {os.path.relpath(p, base)}")
        if len(to_write) > 10:
            log(f"    [DRY +] ... a dalsich {len(to_write)-10}")
        for p in to_delete[:10]:
            log(f"    [DRY -] {os.path.relpath(p, base)}")
        if len(to_delete) > 10:
            log(f"    [DRY -] ... a dalsich {len(to_delete)-10}")
        return

    # ZAPIS chybejicich (PDF binarka se tahá az ted, jen pro tyto)
    written = 0
    for p, d in to_write:
        os.makedirs(os.path.dirname(p), exist_ok=True)
        full = col.find_one({"record_id": d["record_id"]}, {"pdf": 1})
        data = bytes(full["pdf"])
        with open(p, "wb") as f:
            f.write(data)
        written += 1
        if written % 50 == 0:
            log(f"    ... zapsano {written}/{len(to_write)}")

    # SMAZ prebytecne
    deleted = 0
    for p in to_delete:
        try:
            os.remove(p)
            deleted += 1
        except OSError as e:
            log(f"    CHYBA mazani {p}: {e!r}")

    # odstran prazdne podslozky center pod base
    removed_dirs = 0
    for root, dirs, files in os.walk(base, topdown=False):
        if root == base:
            continue
        if not os.listdir(root):
            try:
                os.rmdir(root)
                removed_dirs += 1
            except OSError:
                pass

    log(f"  HOTOVO: zapsano {written}, smazano {deleted}, smazano prazdnych slozek {removed_dirs}.")


def main():
    client = MongoClient(MONGO_URI, serverSelectionTimeoutMS=5000)
    client.admin.command("ping")
    col = client[DB_NAME][COLLECTION]

    items = STUDY_PATHS.items()
    if ARGS.study:
        items = [(k, v) for k, v in STUDY_PATHS.items() if k == ARGS.study]
        if not items:
            log(f"Neznamy studyCode: {ARGS.study}. Dostupne: {list(STUDY_PATHS)}")
            return

    log(f"Materializace {'(DRY-RUN) ' if ARGS.dry_run else ''}-> Dropbox, {len(list(items))} studii"
        f"{', --verify' if ARGS.verify else ''}.")
    for study_code, base in items:
        process_study(col, study_code, base)
    client.close()
    log("\nKONEC.")


if __name__ == "__main__":
    main()