import os
import sys
import pandas as pd
from datetime import date
from pathlib import Path
from openpyxl import load_workbook
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter

sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from common.mongo_writer import get_db

STUDIES = ["77242113UCO3001", "42847922MDD3003"]

BASE_DIR   = Path(os.path.dirname(os.path.abspath(__file__)))
OUTPUT_DIR = BASE_DIR / "output"

DATE_COLUMNS = {
    "Orig Exp Date", "Exp Date", "Rcv Date",
    "Date Asgn", "Disp Date", "Date Ret", "Destroyed", "Max Visit Date",
    "Visit Date", "Scheduled Date",
}

N_SHIP_COLS = 9  # počet shipment sloupců před detail sloupci


# ── Načítání dat z MongoDB ────────────────────────────────────────────────────

INVENTORY_COLS = [
    ("site",                     "Site"),
    ("medication_id",            "Med ID"),
    ("packaged_lot_no",          "Lot No."),
    ("original_expiration_date", "Orig Exp Date"),
    ("expiration_date",          "Exp Date"),
    ("received_date",            "Rcv Date"),
    ("receipt_user",             "Rcpt User"),
    ("subject_identifier",       "Subject ID"),
    ("quantity_assigned",        "Qty Asgn"),
    ("irt_transaction",          "IRT Tx"),
    ("date_assigned",            "Date Asgn"),
    ("assignment_user",          "Asgn User"),
    ("dispensation_status",      "Disp Status"),
    ("dispensing_date",          "Disp Date"),
    ("quantity_dispensed",       "Qty Disp"),
    ("dispensing_user",          "Disp User"),
    ("quantity_returned",        "Qty Ret"),
    ("date_returned",            "Date Ret"),
    ("return_user",              "Ret User"),
]


def load_inventory(study):
    db = get_db()
    inv = list(db.iwrs_inventory.find({"study": study}))
    destr = list(db.iwrs_destruction.find({"study": study}))
    # map medication_id -> first basket+date
    destr_map = {}
    for d in destr:
        mid = d.get("medication_id")
        if mid and mid not in destr_map:
            destr_map[mid] = (d.get("basket_id"), d.get("destruction_date"))

    records = []
    for doc in inv:
        row = {label: doc.get(key) for key, label in INVENTORY_COLS}
        b, dt = destr_map.get(doc.get("medication_id"), (None, None))
        row["Destroyed"]  = dt
        row["Basket No."] = b
        records.append(row)

    df = pd.DataFrame(records)
    if df.empty:
        print("  Inventory: 0 kitu")
        return df

    df = df.sort_values(["Site", "Rcv Date", "Med ID"], na_position="last").reset_index(drop=True)
    for col in DATE_COLUMNS:
        if col in df.columns:
            df[col] = pd.to_datetime(df[col], errors="coerce")
    print(f"  Inventory: {len(df)} kitu")
    return df


SHIP_COLS = [
    ("shipment_id",       "Shipment ID"),
    ("status",             "IRT Shipment Status"),
    ("type",               "Type"),
    ("ship_from",          "Shipment From"),
    ("ship_to_site",       "Ship To:"),
    ("request_date",       "Request Date"),
    ("received_date",      "Received Date"),
    ("received_by",        "Received by"),
    ("expected_arrival",   "Expected Arrival"),
]

ITEM_COLS = [
    ("investigator",            "Investigator"),
    ("medication_description",  "Medication Description"),
    ("medication_id",           "Medication ID"),
    ("packaged_lot_no",         "Packaged Lot number"),
    ("expiration_date",         "Expiration Date"),
    ("item_status",             "Status"),
]


def load_shipments(study):
    db = get_db()
    ships  = list(db.iwrs_shipments.find({"study": study}))
    items  = list(db.iwrs_shipment_items.find({"study": study}))

    # index items by shipment_id
    items_by_ship = {}
    for it in items:
        items_by_ship.setdefault(it.get("shipment_id"), []).append(it)

    records = []
    for s in ships:
        base = {label: s.get(key) for key, label in SHIP_COLS}
        for it in items_by_ship.get(s.get("shipment_id"), []):
            row = dict(base)
            for key, label in ITEM_COLS:
                row[label] = it.get(key)
            records.append(row)

    df = pd.DataFrame(records)
    if df.empty:
        print("  Shipments: 0 zásilek, 0 kitu")
        return df

    df = df.sort_values(["Ship To:", "Shipment ID", "Medication ID"], na_position="last").reset_index(drop=True)
    for col in ("Request Date", "Received Date", "Expiration Date", "Expected Arrival"):
        if col in df.columns:
            df[col] = pd.to_datetime(df[col], errors="coerce")
    n_ship = df["Shipment ID"].nunique()
    print(f"  Shipments: {n_ship} zásilek, {len(df)} kitu")
    return df


def load_visits(study):
    db = get_db()
    cur = db.iwrs_visits.find({
        "study": study,
        "visit_type": "Past",
        "irt_transaction_no": {"$ne": None},
    })
    rows = []
    for v in cur:
        rows.append({
            "Subject":      v.get("subject"),
            "Visit Date":   v.get("actual_date") or v.get("scheduled_date"),
            "Scheduled Date": v.get("scheduled_date"),
            "IRT Tx No":    v.get("irt_transaction_no"),
            "Visit":        v.get("irt_transaction_description"),
            "Medication":   v.get("medication_assignment"),
            "medication_id": v.get("medication_id"),
            "quantity_assigned": v.get("quantity_assigned"),
        })
    df = pd.DataFrame(rows)
    if df.empty:
        print("  Visits: 0 radku")
        return df

    # GROUP BY subject/actual/scheduled/irt_no/desc/medication
    grouped = (
        df.groupby(["Subject", "Visit Date", "Scheduled Date", "IRT Tx No", "Visit", "Medication"],
                   dropna=False, as_index=False)
        .agg(**{
            "Med IDs": ("medication_id", lambda s: ", ".join(sorted([str(x) for x in s if pd.notna(x)]))),
            "Qty":     ("quantity_assigned", "sum"),
        })
    )
    grouped = grouped.sort_values(["Subject", "Visit Date"]).reset_index(drop=True)
    for col in ("Visit Date", "Scheduled Date"):
        if col in grouped.columns:
            grouped[col] = pd.to_datetime(grouped[col], errors="coerce")
    if study == "77242113UCO3001":
        grouped["Visit"] = grouped["Visit"].replace("Subject Number Creation", "Screening")
    print(f"  Visits: {len(grouped)} řádků")
    return grouped


# ── Odvozené sheety ───────────────────────────────────────────────────────────

def build_site_summary(shipments_df):
    STATUS_COLS = ["Available", "Assigned", "Dispensed", "Returned by Subject"]
    pivot = shipments_df.groupby("Ship To:")["Status"].value_counts().unstack(fill_value=0)
    for s in STATUS_COLS:
        if s not in pivot.columns:
            pivot[s] = 0
    pivot = (
        pivot[STATUS_COLS]
        .reset_index()
        .rename(columns={"Ship To:": "Site", "Returned by Subject": "Returned"})
        .sort_values("Site")
        .reset_index(drop=True)
    )
    pivot["Total"] = pivot[["Available", "Assigned", "Dispensed", "Returned"]].sum(axis=1)
    print(f"  Site Summary: {len(pivot)} center")
    return pivot


def build_expired(df):
    today = date.today()
    mask = (
        df["Basket No."].isna() &
        df["Subject ID"].isna() &
        (df["Exp Date"] < pd.Timestamp(today))
    )
    filtered = df[mask].copy().reset_index(drop=True)
    sheet_name = f"Expired as of {today.strftime('%d-%b-%Y')}"
    print(f"  Expired: {len(filtered)}")
    return filtered, sheet_name


def build_assigned_not_dispensed(df):
    mask = df["Subject ID"].notna() & df["Disp Date"].isna()
    filtered = df[mask].copy().reset_index(drop=True)
    print(f"  Assigned not dispensed: {len(filtered)}")
    return filtered


def build_not_returned(df):
    no_ret = df[
        df["Date Ret"].isna() &
        df["Subject ID"].notna() &
        (df["Disp Status"].fillna("").str.upper() != "NOT DISPENSED")
    ].copy()
    max_asgn = df.groupby("Subject ID")["Date Asgn"].max().rename("Max Visit Date")
    no_ret = no_ret.join(max_asgn, on="Subject ID")
    filtered = no_ret[no_ret["Date Asgn"] < no_ret["Max Visit Date"]].copy()
    filtered = filtered.drop(columns=["Qty Ret", "Date Ret", "Ret User", "Destroyed", "Basket No."])
    filtered = filtered.reset_index(drop=True)
    print(f"  Not returned: {len(filtered)}")
    return filtered


def build_kits_for_destruction(df):
    mask = (
        df["Basket No."].isna() &
        (df["Date Ret"].notna() | (df["Disp Status"].fillna("").str.upper() == "NOT DISPENSED"))
    )
    filtered = (
        df[mask]
        .copy()
        .sort_values(["Site", "Date Ret"], ascending=[True, True])
        .drop(columns=["Destroyed", "Basket No."])
        .reset_index(drop=True)
    )
    print(f"  Kits for destruction: {len(filtered)}")
    return filtered


# ── Formátování ───────────────────────────────────────────────────────────────

STRIPE_GRAY  = PatternFill("solid", start_color="F2F2F2")
STRIPE_WHITE = PatternFill("solid", start_color="FFFFFF")

# pacienti — styly zachovány z create_subject_report.py
_PAT_HEADER_FILL  = PatternFill("solid", start_color="1F4E79")
_PAT_HEADER_FONT  = Font(name="Arial", bold=True, color="FFFFFF", size=10)
_PAT_NORMAL_FONT  = Font(name="Arial", size=10)
_PAT_BOLD_FONT    = Font(name="Arial", bold=True, size=10)
_PAT_STRIKE_FONT  = Font(name="Arial", size=10, strike=True, color="999999")
_PAT_ADOLESC_FONT = Font(name="Arial", bold=True, size=10)
_PAT_THIN         = Side(style="thin", color="CCCCCC")
_PAT_BORDER       = Border(left=_PAT_THIN, right=_PAT_THIN, top=_PAT_THIN, bottom=_PAT_THIN)
_PAT_EVEN_FILL    = PatternFill("solid", start_color="EBF3FB")
_PAT_ODD_FILL     = PatternFill("solid", start_color="FFFFFF")
_PAT_CENTER       = Alignment(horizontal="center", vertical="center")
_PAT_LEFT         = Alignment(horizontal="left",   vertical="center")


def _autofit(ws):
    for col_cells in ws.columns:
        max_len = 0
        col_letter = get_column_letter(col_cells[0].column)
        for cell in col_cells:
            if cell.value is None:
                continue
            # datum se zobrazí jako DD-MMM-YYYY = 11 znaků
            if hasattr(cell.value, "strftime") or cell.number_format == "DD-MMM-YYYY":
                length = 11
            else:
                length = len(str(cell.value))
            if length > max_len:
                max_len = length
        ws.column_dimensions[col_letter].width = min(max_len + 3, 50)


def format_sheet(ws, header_color, highlight_col=None, highlight_color=None):
    thin        = Side(style="thin", color="000000")
    border      = Border(left=thin, right=thin, top=thin, bottom=thin)
    header_fill = PatternFill("solid", start_color=header_color)
    header_font = Font(bold=True, color="FFFFFF", name="Arial", size=10)
    row_font    = Font(name="Arial", size=10)
    hi_fill     = PatternFill("solid", start_color=highlight_color) if highlight_color else None

    headers = [cell.value for cell in ws[1]]

    for cell in ws[1]:
        cell.fill      = header_fill
        cell.font      = header_font
        cell.alignment = Alignment(horizontal="center", vertical="center", wrap_text=False)
        cell.border    = border

    for row in ws.iter_rows(min_row=2, max_row=ws.max_row):
        stripe = STRIPE_GRAY if row[0].row % 2 == 0 else STRIPE_WHITE
        for cell in row:
            col_name = headers[cell.column - 1] if cell.column <= len(headers) else None
            cell.font      = row_font
            cell.border    = border
            cell.alignment = Alignment(horizontal="center")
            if col_name in DATE_COLUMNS:
                cell.number_format = "DD-MMM-YYYY"
            if hi_fill and col_name == highlight_col:
                cell.fill = hi_fill
            else:
                cell.fill = stripe

    _autofit(ws)
    ws.auto_filter.ref = ws.dimensions
    ws.freeze_panes    = "A2"


def format_shipment_sheet(ws, header_color_ship, header_color_detail, n_ship_cols):
    thin        = Side(style="thin", color="000000")
    border      = Border(left=thin, right=thin, top=thin, bottom=thin)
    hfont       = Font(bold=True, color="FFFFFF", name="Arial", size=10)
    dfont       = Font(name="Arial", size=10)
    fill_ship   = PatternFill("solid", start_color=header_color_ship)
    fill_detail = PatternFill("solid", start_color=header_color_detail)

    for cell in ws[1]:
        cell.fill      = fill_ship if cell.column <= n_ship_cols else fill_detail
        cell.font      = hfont
        cell.alignment = Alignment(horizontal="center", vertical="center", wrap_text=True)
        cell.border    = border
    ws.row_dimensions[1].height = 30

    for row in ws.iter_rows(min_row=2, max_row=ws.max_row):
        stripe = STRIPE_GRAY if row[0].row % 2 == 0 else STRIPE_WHITE
        for cell in row:
            cell.font      = dfont
            cell.border    = border
            cell.alignment = Alignment(horizontal="center", vertical="center")
            cell.fill      = stripe
            if cell.value.__class__.__name__ in ("datetime", "date", "Timestamp"):
                cell.number_format = "DD-MMM-YYYY"

    _autofit(ws)
    ws.auto_filter.ref = ws.dimensions
    ws.freeze_panes    = "A2"


# ── Pacienti ─────────────────────────────────────────────────────────────────

def load_patients(study):
    db = get_db()
    docs = list(db.iwrs_subject_summary.find({"study": study}))
    if not docs:
        raise RuntimeError(f"Žádná data v Mongo pro pacienty {study}")

    base_cols = [
        ("subject",                          "Subject"),
        ("investigator",                     "Investigator"),
        ("age",                              "Subject's age collection"),
        ("cohort_per_irt",                   "Cohort per IRT"),
        ("irt_subject_status",               "IRT Subject Status"),
        ("last_irt_transaction",             "Last Recorded IRT Transaction"),
        ("next_irt_transaction",             "Next Expected IRT Transaction"),
        ("next_irt_transaction_date_local",  "Next Expected IRT Transaction Date [Local]"),
    ]
    uco_extra = [
        ("rescreened_subject",               "Rescreened Subject"),
        ("adt_ir",                           "ADT-IR"),
        ("three_or_more_advanced_therapies", "3+ Adv. Therapies"),
        ("only_oral_5asa_compounds",         "Only 5-ASA"),
        ("ustekinumab",                      "Ustekinumab"),
        ("isolated_proctitis",               "Isolated Proctitis"),
    ]
    cols = list(base_cols)
    if study == "77242113UCO3001":
        cols += uco_extra

    rows = [{label: d.get(key) for key, label in cols} for d in docs]
    df = pd.DataFrame(rows).sort_values("Subject").reset_index(drop=True)

    if "Next Expected IRT Transaction Date [Local]" in df.columns:
        df["Next Expected IRT Transaction Date [Local]"] = pd.to_datetime(
            df["Next Expected IRT Transaction Date [Local]"], errors="coerce"
        )
    print(f"  Pacienti: {len(df)} subjektů")
    return df


def _simplify_cohort(val):
    if pd.isna(val):
        return ""
    val = str(val)
    if "dolescent" in val:
        return "Adolescent"
    if val.startswith("Adult"):
        return "Adult"
    return val


def _fmt_date(val):
    if pd.isna(val):
        return ""
    if hasattr(val, "strftime"):
        return val.strftime("%Y-%m-%d")
    return str(val)[:10]


def _write_prehled(wb, df_raw, study):
    ws = wb.create_sheet("Přehled", 0)
    ws.sheet_view.showGridLines = False

    is_uco = (study == "77242113UCO3001")

    if is_uco:
        display_headers = ["Subject", "Investigator", "Věk", "Cohort",
                           "Rescreened", "ADT-IR", "≥3 Adv.Th.", "5-ASA only",
                           "Uste.", "Isol.Proct.",
                           "Status", "Last IRT", "Next Visit", "Next Date"]
        col_widths       = [14, 22, 6, 12, 11, 8, 11, 10, 8, 12, 14, 12, 12, 13]
        status_col       = 11
        flag_cols        = set(range(5, 11))  # 1-indexed sloupce s Yes/No hodnotami
    else:
        display_headers = ["Subject", "Investigator", "Věk", "Cohort", "Status", "Last IRT", "Next Visit", "Next Date"]
        col_widths       = [14, 22, 6, 12, 14, 12, 12, 13]
        status_col       = 5
        flag_cols        = set()

    last_col = get_column_letter(len(display_headers))
    ws.merge_cells(f"A1:{last_col}1")
    title = ws["A1"]
    title.value     = f"Subject Summary — {study}   ({date.today().strftime('%d-%b-%Y')})"
    title.font      = Font(name="Arial", bold=True, size=12, color="1F4E79")
    title.alignment = Alignment(horizontal="left", vertical="center")
    ws.row_dimensions[1].height = 22

    for c, (h, w) in enumerate(zip(display_headers, col_widths), 1):
        cell = ws.cell(row=2, column=c, value=h)
        cell.font = _PAT_HEADER_FONT
        cell.fill = _PAT_HEADER_FILL
        cell.alignment = _PAT_CENTER
        cell.border = _PAT_BORDER
        ws.column_dimensions[get_column_letter(c)].width = w
    ws.row_dimensions[2].height = 18

    base = {
        "Subject":      df_raw["Subject"].fillna(""),
        "Investigator": df_raw["Investigator"].fillna(""),
        "Věk":          df_raw["Subject's age collection"].apply(lambda v: "" if pd.isna(v) else int(v)),
        "Cohort":       df_raw["Cohort per IRT"].apply(_simplify_cohort),
    }
    if is_uco:
        base.update({
            "Rescreened":  df_raw["Rescreened Subject"].fillna(""),
            "ADT-IR":      df_raw["ADT-IR"].fillna(""),
            "≥3 Adv.Th.": df_raw["3+ Adv. Therapies"].fillna(""),
            "5-ASA only":  df_raw["Only 5-ASA"].fillna(""),
            "Uste.":       df_raw["Ustekinumab"].fillna(""),
            "Isol.Proct.": df_raw["Isolated Proctitis"].fillna(""),
        })
    base.update({
        "Status":     df_raw["IRT Subject Status"].fillna(""),
        "Last IRT":   df_raw["Last Recorded IRT Transaction"].fillna("—"),
        "Next Visit": df_raw["Next Expected IRT Transaction"].fillna("—"),
        "Next Date":  df_raw["Next Expected IRT Transaction Date [Local]"].apply(_fmt_date),
    })
    display = pd.DataFrame(base).sort_values("Subject").reset_index(drop=True)

    for r_idx, row in display.iterrows():
        excel_row     = r_idx + 3
        status        = str(row["Status"])
        is_failed     = "Screen Failed" in status or "Discontinued" in status
        is_randomized = "Randomized" in status
        is_adolescent = row["Cohort"] == "Adolescent"
        fill = _PAT_EVEN_FILL if r_idx % 2 == 0 else _PAT_ODD_FILL

        for c_idx, val in enumerate(row, 1):
            cell = ws.cell(row=excel_row, column=c_idx, value=val if val != "" else None)
            cell.fill      = fill
            cell.border    = _PAT_BORDER
            cell.alignment = _PAT_CENTER if (c_idx == 3 or c_idx in flag_cols) else _PAT_LEFT
            if is_failed:
                cell.font = _PAT_STRIKE_FONT
            elif c_idx == status_col and is_randomized:
                cell.font = _PAT_BOLD_FONT
            elif c_idx == 4 and is_adolescent:
                cell.font = _PAT_ADOLESC_FONT
            else:
                cell.font = _PAT_NORMAL_FONT
        ws.row_dimensions[excel_row].height = 16

    ws.freeze_panes = "A3"
    ws.auto_filter.ref = f"A2:{last_col}{len(display) + 2}"


def _write_next_visits(wb, df_raw, study, visits_df=None):
    ws = wb.create_sheet("Next Visits", 1)
    ws.sheet_view.showGridLines = False

    ws.merge_cells("A1:D1")
    title = ws["A1"]
    title.value     = f"Next Expected Visits — {study}   ({date.today().strftime('%d-%b-%Y')})"
    title.font      = Font(name="Arial", bold=True, size=12, color="1F4E79")
    title.alignment = Alignment(horizontal="left", vertical="center")
    ws.row_dimensions[1].height = 22

    nv_headers = ["Subject", "Investigator", "Next Visit", "Datum"]
    nv_widths   = [14, 22, 26, 13]
    for c, (h, w) in enumerate(zip(nv_headers, nv_widths), 1):
        cell = ws.cell(row=2, column=c, value=h)
        cell.font = _PAT_HEADER_FONT
        cell.fill = _PAT_HEADER_FILL
        cell.alignment = _PAT_CENTER
        cell.border = _PAT_BORDER
        ws.column_dimensions[get_column_letter(c)].width = w
    ws.row_dimensions[2].height = 18

    df = pd.DataFrame({
        "Subject":      df_raw["Subject"].fillna(""),
        "Investigator": df_raw["Investigator"].fillna(""),
        "Next Visit":   df_raw["Next Expected IRT Transaction"].fillna(""),
        "Datum":        df_raw["Next Expected IRT Transaction Date [Local]"],
        "Status":       df_raw["IRT Subject Status"].fillna(""),
    })

    # I-0: datum = screening date + 42 dní
    if visits_df is not None and not visits_df.empty:
        screen = (
            visits_df[visits_df["Visit"].str.contains("Screen", case=False, na=False)]
            .groupby("Subject")["Visit Date"].min()
            .rename("Screening Date")
        )
        df = df.join(screen, on="Subject")
        mask_i0 = df["Next Visit"].str.contains("I-0", na=False)
        df.loc[mask_i0, "Datum"] = df.loc[mask_i0, "Screening Date"] + pd.Timedelta(days=42)
        df = df.drop(columns=["Screening Date"])

    df = df[df["Datum"].notna()]
    df = df[~df["Status"].str.contains("Screen Failed|Discontinued", na=False)]
    df = df.sort_values("Datum").reset_index(drop=True)

    for r_idx, row in df.iterrows():
        excel_row = r_idx + 3
        fill      = _PAT_EVEN_FILL if r_idx % 2 == 0 else _PAT_ODD_FILL
        datum_val = row["Datum"]
        datum_str = datum_val.strftime("%Y-%m-%d") if hasattr(datum_val, "strftime") else str(datum_val)[:10]
        for c_idx, val in enumerate([row["Subject"], row["Investigator"], row["Next Visit"], datum_str], 1):
            cell = ws.cell(row=excel_row, column=c_idx, value=val if val != "" else None)
            cell.fill      = fill
            cell.border    = _PAT_BORDER
            cell.font      = _PAT_NORMAL_FONT
            cell.alignment = _PAT_LEFT
        ws.row_dimensions[excel_row].height = 16

    ws.freeze_panes = "A3"
    ws.auto_filter.ref = f"A2:D{len(df) + 2}"


# ── Jeden report pro jednu studii ─────────────────────────────────────────────

def create_study_report(study):
    today = date.today()

    # číslování: najdi nejvyšší existující verzi pro dnešní datum
    existing = sorted(OUTPUT_DIR.glob(f"{today} {study} CZ IWRS overview v*.xlsx"))
    if existing:
        last = existing[-1].stem  # např. "2026-05-12 42847922MDD3003 CZ IWRS overview v3"
        last_ver = int(last.rsplit("v", 1)[-1])
        version = last_ver + 1
    else:
        version = 1

    output_file = OUTPUT_DIR / f"{today} {study} CZ IWRS overview v{version}.xlsx"

    print(f"\n[{study}] Nacitam z MongoDB...")
    df           = load_inventory(study)
    shipments_df = load_shipments(study)
    df_patients  = load_patients(study)
    visits_df    = load_visits(study)

    expired_df, expired_sheet = build_expired(df)
    assigned_df               = build_assigned_not_dispensed(df)
    not_returned_df           = build_not_returned(df)
    destruction_df            = build_kits_for_destruction(df)
    site_summary_df           = build_site_summary(shipments_df)

    with pd.ExcelWriter(output_file, engine="openpyxl") as writer:
        df.to_excel(              writer, index=False, sheet_name="CountryMedicationOverview")
        expired_df.to_excel(      writer, index=False, sheet_name=expired_sheet)
        assigned_df.to_excel(     writer, index=False, sheet_name="Assigned not dispensed")
        not_returned_df.to_excel( writer, index=False, sheet_name="Not returned")
        destruction_df.to_excel(  writer, index=False, sheet_name="Kits for destruction")
        shipments_df.to_excel(    writer, index=False, sheet_name="Shipments")
        site_summary_df.to_excel( writer, index=False, sheet_name="Site Summary")
        visits_df.to_excel(       writer, index=False, sheet_name="Patient Visits")

    wb = load_workbook(output_file)

    ws_main = wb["CountryMedicationOverview"]
    format_sheet(ws_main, header_color="1F4E79")
    green_fill   = PatternFill("solid", start_color="E2EFDA")
    headers_main = [c.value for c in ws_main[1]]
    for row in ws_main.iter_rows(min_row=2, max_row=ws_main.max_row):
        for cell in row:
            col_name = headers_main[cell.column - 1] if cell.column <= len(headers_main) else None
            if col_name in ("Destroyed", "Basket No."):
                cell.fill = green_fill

    format_sheet(wb[expired_sheet],            header_color="C00000", highlight_col="Exp Date",       highlight_color="FFE0E0")
    format_sheet(wb["Assigned not dispensed"], header_color="833C00", highlight_col="Subject ID",     highlight_color="FFF2CC")
    format_sheet(wb["Not returned"],           header_color="375623", highlight_col="Max Visit Date", highlight_color="E2EFDA")
    format_sheet(wb["Kits for destruction"],   header_color="595959")
    format_shipment_sheet(wb["Shipments"],     "1F4E79", "375623", N_SHIP_COLS)
    format_sheet(wb["Site Summary"],           header_color="1F4E79")
    format_sheet(wb["Patient Visits"],         header_color="1F4E79")

    # ── pacienti (Přehled + Next Visits) na začátek ──────────────────────────
    _write_prehled(wb, df_patients, study)
    _write_next_visits(wb, df_patients, study, visits_df)

    # ── pořadí listů: Patient Visits jako první ──────────────────────────────
    names = wb.sheetnames
    wb._sheets = [wb["Patient Visits"]] + [wb[s] for s in names if s != "Patient Visits"]

    wb.save(output_file)
    print(f"  Uloženo: {output_file.name}  ({len(df)} řádků)")


# ── Main ──────────────────────────────────────────────────────────────────────

def main():
    OUTPUT_DIR.mkdir(exist_ok=True)
    for study in STUDIES:
        try:
            create_study_report(study)
        except Exception as e:
            import traceback
            print(f"\n[{study}] CHYBA: {e}")
            traceback.print_exc()
    print("\nHotovo.")


main()