Facturx.Extract (Facturx v0.7.0)

Copy Markdown View Source

Extract the embedded CII XML from a Factur-X / ZUGFeRD PDF (pure Elixir).

Locates the embedded file referenced by the PDF's /Filespec (/AFRelationship /Data), decodes its FlateDecode stream, and reports the detected profile.

Reads both cross-reference forms: the classic table, and the PDF 1.5+ stream with its objects compressed into object streams. The index is built by scanning for N G obj and then expanding every object stream — never from the cross-reference table, which an incremental update leaves pointing at the objects the update replaced.

Unlike Facturx.Embed, this module does not require its input to be PDF/A, and does not check: reading an attachment cannot damage the document, so refusing a file it can in fact read would only be in the caller's way. The library will therefore read from PDFs it would decline to write to.

Memory note (BEAM refc binaries)

A PDF is a large binary (> 64 bytes → refc binary). Slicing it yields sub-binaries that retain the whole PDF. Everything we return — the XML and the filename — is :binary.copy/1-ed so the multi-MB PDF can be collected.

Summary

Types

result()

Result of a successful extraction.

Functions

extract(pdf)

Extract the embedded CII XML and metadata from pdf.

Types

result()

@type result() :: %{
  xml: binary(),
  profile: Facturx.profile() | nil,
  filename: String.t()
}

Result of a successful extraction.

Functions

extract(pdf)

@spec extract(binary()) :: {:ok, result()} | {:error, term()}

Extract the embedded CII XML and metadata from pdf.