PdfElixide.Document.Metadata (pdf_elixide v0.9.0)

Document Info dictionary metadata — a PDF's classic /Info fields.

Every field is optional and defaults to nil; a document with no /Info dictionary yields a struct with all fields nil. Obtain it with PdfElixide.Document.metadata/1.

For richer, XML-based metadata (which many modern PDFs carry instead of, or in addition to, the Info dictionary) see PdfElixide.Document.xmp_metadata/1.

Text encoding

String fields are decoded as PDF text strings (ISO 32000-1 §7.9.2.2): a FE FF or FF FE byte-order mark selects UTF-16, PDF 2.0's EF BB BF selects UTF-8, and anything else is PDFDocEncoding — Latin-1 except over 0x80–0x9F, where the PDF glyphs (0x85 en dash, 0x90 right single quote, 0x92 trademark) replace the C1 controls.

A byte-order-mark-less string that happens to be valid UTF-8 is decoded as UTF-8. That is pdf_oxide's deliberate leniency towards the many producers that write raw UTF-8 rather than what the spec says, so a Latin-1 string whose bytes are also well-formed UTF-8 decodes as the latter. A value that decodes to whitespace only is nil, as an absent one is.

Fields

:title, :author, :subject — document title, author, and subject.
:keywords — the raw /Keywords string (PDF stores it as a single string; the split list, when available, is on the XMP :subjects field).
:creator — the application that created the original document.
:producer — the application that produced the PDF.
:creation_date, :mod_date — the raw PDF date strings (e.g. "D:20230101120000+00'00'"); not parsed into DateTime.
:trapped — the /Trapped value ("True", "False", or "Unknown").

Summary

Types

t()

Types

t()

@type t() :: %PdfElixide.Document.Metadata{
  author: String.t() | nil,
  creation_date: String.t() | nil,
  creator: String.t() | nil,
  keywords: String.t() | nil,
  mod_date: String.t() | nil,
  producer: String.t() | nil,
  subject: String.t() | nil,
  title: String.t() | nil,
  trapped: String.t() | nil
}