Migrating from other libraries

Map openpyxl, XlsxWriter and polars.write_excel calls onto Jetxl

The mapping is mostly mechanical. Jetxl has no reader, so if your script also opens or edits workbooks, keep your existing library for that half and use Jetxl only for writing.

From openpyxl

openpyxl builds a workbook cell by cell. Jetxl takes the whole frame plus a description of how it should look.

Before

from openpyxl import Workbook
from openpyxl.styles import Font

wb = Workbook()
ws = wb.active
ws.title = 'Employees'
ws.append(['Name', 'Salary'])

for row in rows:
    ws.append(row)

ws.freeze_panes = 'A2'
for cell in ws[1]:
    cell.font = Font(bold=True)

wb.save('out.xlsx')

openpyxl	Jetxl
ws.title	sheet_name
ws.freeze_panes	freeze_rows, freeze_cols
ws.auto_filter.ref	auto_filter
cell.font = Font(bold=True)	styled_headers, or cell_styles
cell.number_format	column_formats
ws.column_dimensions[...].width	column_widths
ws.merge_cells(...)	merge_cells
wb.create_sheet()	write_sheets_arrow
wb.save(path)	the filename argument
Reading a workbook	Not available

From XlsxWriter

XlsxWriter is closer in spirit, since it’s also write-only. The main difference is that it defines format objects up front and applies them per write, while Jetxl takes plain dictionaries.

Before

import xlsxwriter

wb = xlsxwriter.Workbook('out.xlsx')
ws = wb.add_worksheet('Sales')

money = wb.add_format({'num_format': '$#,##0.00'})
bold = wb.add_format({'bold': True})

ws.write_row(0, 0, ['Product', 'Revenue'], bold)
ws.set_column(1, 1, 14, money)
ws.freeze_panes(1, 0)

wb.close()

XlsxWriter	Jetxl
wb.add_format({...})	A dictionary in cell_styles or column_formats
ws.set_column(...)	column_widths
ws.freeze_panes(...)	freeze_rows, freeze_cols
ws.add_table(...)	tables
wb.add_chart(...)	charts
ws.insert_image(...)	images
ws.data_validation(...)	data_validations
ws.conditional_format(...)	conditional_formats

From polars.write_excel

The closest starting point, since you already have a DataFrame.

# Before
df.write_excel('out.xlsx', table_style='TableStyleMedium9', autofit=True)

# After
jet.write_sheet_arrow(
    df.to_arrow(),
    'out.xlsx',
    auto_width=True,
    tables=[{'name': 'Data', 'start_row': 1, 'start_col': 0,
             'style': 'TableStyleMedium9'}],
)

The published benchmark puts polars.write_excel at 26.6s against 0.66s for Jetxl on a million rows, with 3.13 GB peak memory against 958 MB. See Performance.

What to check after switching

Indexing

openpyxl and XlsxWriter are each internally consistent about row bases. Jetxl isn’t: rows are 1-based for cell styles, header content and table ranges, and 0-based for chart and image positions. See Conventions.

Percentages

If you were writing percentages as whole numbers with a plain format, the built-in percentage formats multiply them by 100. Store decimals instead.

Colors

XlsxWriter accepts named colors such as red. Jetxl doesn’t, and drops them silently rather than raising. Convert to hex.

Header styling

styled_headers gives bold only, with no fill. If your openpyxl or XlsxWriter output had shaded headers, reproduce that with cell_styles on the header row.

Reading

Any part of your code that opens an existing workbook has to stay on openpyxl. Jetxl only writes.