Contract PDFs in Java

Why contracts become PDFs

A contract has to be fixed at signature. Both parties need to be certain they are looking at identical words, which a rendered page cannot guarantee. That is the difference between a document and a view of a database, and it is why legal and sales operations keep asking for this.

What a contract has to carry

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

• the parties, by legal entity name and registered number
• the effective date and the term
• numbered clauses with stable numbering, since amendments reference them
• the commercial schedule, usually as an annexe rather than in the body
• signature blocks with name, title and date for each party

The Java implementation

java.net.http.HttpClient ships with the runtime, so this adds no dependency. The render happens in a controller returning StreamingResponseBody, and you BodyHandlers.ofInputStream into a StreamingResponseBody.

@GetMapping(value = "/contract/{id}.pdf", produces = MediaType.APPLICATION_PDF_VALUE)
public ResponseEntity<StreamingResponseBody> contractPdf(@PathVariable String id) throws Exception {
    String html = renderContract(id);

    String payload = mapper.writeValueAsString(Map.of(
        "html", html,
        "options", Map.of("format", "A4", "printBackground", true)
    ));

    HttpRequest request = HttpRequest.newBuilder()
        .uri(URI.create("https://api.pdfpipe.xyz/v1/pdf"))
        .header("Authorization", "Bearer " + System.getenv("PDFPIPE_KEY"))
        .header("Content-Type", "application/json")
        .POST(HttpRequest.BodyPublishers.ofString(payload))
        .build();

    // ofInputStream, not ofByteArray: a long contract never lands on the heap.
    var response = client.send(request, HttpResponse.BodyHandlers.ofInputStream());

    StreamingResponseBody body = out -> response.body().transferTo(out);
    return ResponseEntity.ok(body);
}

The CSS that makes a contract page correctly

The layout problem specific to this document is that numbered clauses with cross-references, where a page break in the wrong place separates a clause from the term it defines. These rules handle it.

/* A clause must not be separated from the term it defines, and a
   signature block alone on a final page invites a dispute. */
.clause     { break-inside: avoid; }
.signatures { break-inside: avoid; break-before: avoid; }

p { orphans: 3; widows: 3; }

@page { size: A4; margin: 25mm 20mm; }

What goes wrong in Java

BodyHandlers.ofByteArray materialises the whole document on the heap. On a servlet container sized for JSON payloads, a few concurrent report renders is all it takes to meet the GC.

What people try first

Most Java projects reach for iText, whose AGPL terms mean commercial use needs a paid licence, or PDFBox, which draws primitives rather than laying out HTML. 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 contract lives afterwards

Rendering is the short part. A contract is retained for the term plus the limitation period, which typically means well over a decade. Immutability matters more here than anywhere else on this list: both parties must be able to prove the words have not moved.

Getting the document right

• Check the parties and the effective date, which is what every later dispute turns on.
• Generation is triggered by agreement being reached, often with a signature workflow following, 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 Java documents, and the same contract in other stacks.