Purchase order PDFs in Python

Why purchase orders become PDFs

A purchase order is the document a supplier's accounts team matches an invoice against, and it travels by email between two organisations that share no system. That is the difference between a document and a view of a database, and it is why procurement teams keep asking for this.

What a purchase order has to carry

Before any of the code below matters, the template has to produce a document that is actually a purchase order. These are the fields that make it one, and the ones a reader or an auditor will look for first.

• the PO number, which the supplier invoice must quote back
• the ordering entity and the ship-to address, which are often different
• the bill-to address and any cost centre or GL code
• each line with the agreed price, not the supplier list price
• the delivery date required, and the approval reference

The Python implementation

httpx is the client most projects here already have. The render happens in an async endpoint that returns a streaming response, and you stream the response with httpx.stream and yield chunks.

import os, httpx
from fastapi import FastAPI
from fastapi.responses import StreamingResponse

app = FastAPI()

@app.get("/purchase-order/{purchase_order_id}.pdf")
async def purchase_order_pdf(purchase_order_id: str):
    html = render_purchase_order(purchase_order_id)

    client = httpx.AsyncClient(timeout=60.0)
    request = client.build_request(
        "POST",
        "https://api.pdfpipe.xyz/v1/pdf",
        headers={"Authorization": f"Bearer {os.environ['PDFPIPE_KEY']}"},
        json={"html": html, "options": {"format": "A4", "printBackground": True}},
    )
    response = await client.send(request, stream=True)

    # stream=True keeps a long purchase order off the worker's heap.
    return StreamingResponse(
        response.aiter_bytes(),
        media_type="application/pdf",
        background=response.aclose,
    )

The CSS that makes a purchase order page correctly

The layout problem specific to this document is that delivery and billing addresses sit alongside a line item table, and both must stay on the first page where a supplier looks for them. These rules handle it.

/* Backgrounds are dropped unless printBackground is set on the
   request, which is the usual reason a document renders in plain
   black and white when it looked right in the browser. */
@page {
  size: A4;
  margin: 18mm 16mm;
}

table { break-inside: avoid; }
h2    { break-after: avoid; }

What goes wrong in Python

requests holds the entire body in memory before you touch it. httpx.stream does not, and on documents past a few megabytes that difference decides whether your worker survives a burst.

What people try first

Most Python projects reach for WeasyPrint, which is excellent at CSS Paged Media but needs Cairo, Pango and GDK-PixBuf present in the image, and diverges from a browser on modern layout. That works until it is running on more than one machine, at which point the browser becomes the thing you operate rather than the thing you use.

Where the purchase order lives afterwards

Rendering is the short part. A PO is retained alongside the invoice it authorised, because the pair is what an auditor matches. In practice it lives in the procurement system for as long as the supplier relationship does, which is longer than most teams plan for.

Getting the document right

• Check the PO number, because an invoice quoting the wrong one will not be paid.
• Generation is triggered by an approval completing in a procurement workflow, so size the timeout for that path rather than for a health check.
• These are produced one at a time by a person who is waiting, so latency is felt directly.
• Backgrounds are painted by default here, so a design that uses colour needs nothing set. Only an explicit print_background of false turns them off.

Frequently asked

Related

Other Python documents, and the same purchase order in other stacks.