Read, validate and sign PDFs, in pure Go

Why PDF Go

Invoices, contracts and delivery notes arrive as PDFs. If you process them automatically, you need a library that swallows nothing. It has to spot damaged files, report every repair openly and keep earlier versions of a file readable. And it should drop into a Go service without C dependencies. That is exactly why we wrote PDF Go.

What’s in the library

• pdf.OpenFile(ctx, path, opts)

  Reading without guessing

  Exact values, cross-reference tables and streams, object streams and incremental revisions. Strict mode rejects malformed input. You switch on Repair explicitly, and every repair then appears in the diagnostics.

• document.ResolveRevision(…)

  Reading earlier versions

  Older revisions of a file can be resolved on purpose, including objects that a later change replaced or deleted. So you can see what the file contained before a change or a signature.

• pdfgo validate --json

  Checking the structure

  Headers from PDF 1.0 to 2.0, cross-references, trailers, stream lengths and linearised files. The report gives every finding a code, as text or JSON, and clear exit codes for your pipeline.

• pdfgo sign · pdfgo verify

  Signing and verifying

  Detached CMS signatures are written incrementally into named form fields, and multiple signatures are preserved. When verifying, PDF Go checks the ByteRange and the trust chain against your X.509 roots.

• pdfgo encrypt · pdfgo decrypt

  Encryption

  The Standard Security Handler R2 to R6 with RC4, AES-128 and AES-256. New files are encrypted with R6 and AES-256.

• pdfgo extract · pdfgo render

  Text and images

  Text comes out in content order via ToUnicode, without guesswork. Pages with vectors, images, TrueType and Type3 fonts are rendered to PNG.

• document.BeginEdit()

  Editing and writing

  Changes run in isolated transactions. Then you rewrite the file deterministically or append the change incrementally.

Quickstart

$ go get gitlab.schukai.com/oss/libraries/go/documents/pdf@v0.1.0-alpha.1
$ go install gitlab.schukai.com/oss/libraries/go/documents/pdf/cmd/pdfgo@v0.1.0-alpha.1

document, err := pdf.OpenFile(ctx, "contract.pdf", pdf.OpenOptions{
    Mode: pdf.Strict,
})
// …
fmt.Printf("PDF %s, %d pages\n", document.Version(), len(pages))

What you can rely on

• Only the Go standard library.No CGo and no third-party modules. go list -m all shows exactly one module.
• Written by us, with a clear provenance.Code, rule tables and test files are our own. Nothing is taken from other PDF implementations; the ISO standards serve as the specification.
• A report that says what was checked.The validation report separates passed checks (checks_passed) from complete profile coverage (complete). So you know exactly what a result is based on.
• Limits instead of surprises.Configurable limits, cancellation via context, safe concurrent access and no network access of its own.
• Tested the way a parser needs to be.Release checks include tests with the race detector and go vet, plus fuzzing with damaged input.

Licences

• Open source under AGPL-3.0.For open-source projects, you use PDF Go under the GNU Affero General Public License, version 3.
• A commercial licence for your software.Want to use PDF Go in a product or service without disclosing your code under the AGPL? We offer a commercial licence for that. Write to us and we’ll send you an offer.

FAQ