weasymerge 0.1.0

Mail merge and bulk document generation from CSV to PDF using WeasyPrint and Jinja2.

WeasyMerge is a tool for doing mail merges and other bulk document generation without having to write your own glue code. You get the power of WeasyPrint's best-in-class PDF generation with the ease of writing a Jinja2 template with some HTML/CSS and providing a CSV file of data to merge.

WeasyMerge can generate individual PDFs or batch rows together into one or more PDFs.

Installation

WeasyMerge is a command-line tool, so the easiest way to install it is with pipx or uv:

pipx install weasymerge

uv tool install weasymerge

You can also install it into an existing environment with pip:

pip install weasymerge

WeasyMerge relies on WeasyPrint, which needs some system libraries — see its installation guide. On Windows this is much smoother under WSL.

Usage

weasymerge --data guests.csv --template invite.html.j2 --output "pdfs/Invite {row_number} {row[Name]}.pdf"

pipx run and uvx run WeasyMerge on demand, fetching it if it isn't already available, so you can invoke it without installing it first:

pipx run weasymerge --data guests.csv --template invite.html.j2 --output "pdfs/Invite {row_number} {row[Name]}.pdf"

uvx weasymerge --data guests.csv --template invite.html.j2 --output "pdfs/Invite {row_number} {row[Name]}.pdf"

The first row of the CSV file is assumed to be the header row and the header is used as the key for the row dictionary.

HTML escaping is on by default in the Jinja2 template.

The row number for the first data row is 1.

WeasyMerge provides "safe" row data to the output filename template (ASCII letters, digits, space, underscore, hyphen, and dot). Non-ASCII letters are converted to their closest ASCII equivalent.

Batch mode (multiple rows per document)

Pass --rows-per-document=N to render N rows into a single document, or --rows-per-document=all to render every row into one document. Use this for label sheets, ID cards, or any layout where the template needs to see more than one row at a time.

weasymerge --data seeds.csv --template labels.html.j2 \
  --output "labels-{batch.from_row_number}-{batch.to_row_number}.pdf" \
  --rows-per-document=all

In batch mode the template receives a batch object instead of row and row_number:

• batch.rows — list of row dicts in this batch
• batch.numbered_rows — list of (row_number, row) pairs
• batch.from_row_number, batch.to_row_number — 1-indexed inclusive range
• batch.total_rows — total rows in the source data

The same fields are available in the --output template (e.g. {batch.from_row_number}, {batch.rows[0][Plant]}). Page breaks within a single document are the template's responsibility — use CSS page-break-after: always or break-after: page.

See examples/labels/ for a 70×37mm, 24-per-A4 label sheet.

Barcodes and QR codes

WeasyMerge comes with barcode and QR code Jinja2 filters, powered by pyStrich.

Each filter turns a value into a data: URL you can drop straight into an <img> tag in your template:

<img src="{{ row['TicketID'] | qrcode_svg_dataurl }}" alt="Ticket QR code">

The available filters are:

• qrcode_svg_dataurl
• datamatrix_svg_dataurl
• aztec_svg_dataurl
• pdf417_svg_dataurl
• code128_svg_dataurl
• ean13_svg_dataurl
• code39_svg_dataurl

See examples/tickets/ for A6 event tickets with a QR code and a Code128 barcode.