Files
janssen/format_accountability.py
T
2026-04-08 14:59:00 +02:00

119 lines
4.6 KiB
Python

import pandas as pd
from openpyxl import load_workbook
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter
INPUT_FILE = "accountability_combined.xlsx"
OUTPUT_FILE = "accountability_formatted.xlsx"
SHEET_NAME = "CountryMedicationOverview"
COLUMN_RENAMES = {
"Site": "Site",
"Medication ID": "Med ID",
"Packaged Lot number": "Lot No.",
"Original Expiration Date when Packaged Lot was Added": "Orig Exp Date",
"Expiration date": "Exp Date",
"Received Date": "Rcv Date",
"Shipment 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",
"DestroyedOn": "Destroyed",
"Basket number": "Basket No.",
}
DATE_COLUMNS = {
"Orig Exp Date", "Exp Date", "Rcv Date",
"Date Asgn", "Disp Date", "Date Ret", "Destroyed",
}
COLUMN_WIDTHS = {
"Site": 14,
"Med ID": 10,
"Lot No.": 12,
"Orig Exp Date": 16,
"Exp Date": 14,
"Rcv Date": 14,
"Rcpt User": 22,
"Subject ID": 14,
"Qty Asgn": 9,
"IRT Tx": 8,
"Date Asgn": 14,
"Asgn User": 20,
"Disp Status": 16,
"Disp Date": 14,
"Qty Disp": 9,
"Disp User": 20,
"Qty Ret": 10,
"Date Ret": 14,
"Ret User": 18,
"Destroyed": 14,
"Basket No.": 12,
}
# ── 1. Load with pandas and convert date columns ─────────────────────────────
df = pd.read_excel(INPUT_FILE)
df.rename(columns=COLUMN_RENAMES, inplace=True)
for col in DATE_COLUMNS:
if col in df.columns:
df[col] = pd.to_datetime(df[col], dayfirst=True, errors="coerce")
df.sort_values(["Site", "Rcv Date", "Med ID"], inplace=True, ignore_index=True)
df.to_excel(OUTPUT_FILE, index=False, sheet_name=SHEET_NAME)
# ── 2. Format with openpyxl ───────────────────────────────────────────────────
wb = load_workbook(OUTPUT_FILE)
ws = wb[SHEET_NAME]
header_fill = PatternFill("solid", start_color="1F4E79")
header_font = Font(bold=True, color="FFFFFF", name="Arial", size=10)
new_col_fill = PatternFill("solid", start_color="E2EFDA")
row_font = Font(name="Arial", size=10)
thin = Side(style="thin", color="000000")
border = Border(left=thin, right=thin, top=thin, bottom=thin)
headers = [cell.value for cell in ws[1]]
new_cols = {"Destroyed", "Basket No."}
# Header row
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
# Data rows
max_row = ws.max_row
for row in ws.iter_rows(min_row=2, max_row=max_row):
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 col_name in new_cols:
cell.fill = new_col_fill
# Column widths
for cell in ws[1]:
width = COLUMN_WIDTHS.get(cell.value, 14)
ws.column_dimensions[get_column_letter(cell.column)].width = width
ws.auto_filter.ref = ws.dimensions
ws.freeze_panes = "A2"
wb.save(OUTPUT_FILE)
print(f"Saved: {OUTPUT_FILE} ({max_row - 1} rows, sheet '{SHEET_NAME}')")