Module xmp 

XMP document metadata (ISO 32000-1 §14.3.2 / Adobe XMP). The catalog’s /Metadata entry is a stream carrying an XMP packet — RDF/XML describing the document with Dublin Core (dc:), XMP Basic (xmp:), and PDF-schema (pdf:) properties. PDF 2.0 deprecates the /Info dictionary in favour of this, so XMP is increasingly the only place some metadata lives.

This reads the common properties with a bounded tag/attribute scrape, not a full XML parser. That is deliberate: an XML engine that resolves general entities is vulnerable to “billion laughs” entity-expansion bombs, and a DOM builder can blow the stack on deeply-nested input. Here no general entity is ever resolved (only the five predefined XML entities and numeric character references, each of which maps to exactly one character), every scan is linear over the byte string, and every field and array is length-capped. Like the other navigation/metadata readers it runs only when explicitly called — never during open or rendering.

The trade-off: a producer that binds the schema namespaces to non-standard prefixes (not dc/xmp/pdf) is not recognized. In practice these prefixes are universal.

Structs

XmpMetadata

Common document metadata read from the XMP packet. Every field is optional; producers populate an arbitrary subset, and a field may be present in XMP but not in /Info (or vice-versa).

Functions

metadata_bytes

Decode and return the raw bytes of the catalog’s /Metadata XMP stream, or None when the document carries none. Routes through the parser’s filter pipeline, so it respects ParseLimits.

parse_xmp

Parse the catalog’s XMP /Metadata packet into XmpMetadata. Returns None when the document carries no /Metadata, it cannot be decoded, or it holds none of the recognized properties.