Crate pdfrum_doc 

pdfrum-doc

Everything the catalog hangs off the page tree (ISO 32000-1 §12): outline bookmarks, destinations, links, annotations, the AcroForm data model, page labels, viewer preferences, the tagged structure tree, XMP metadata, and a PDF/A conformance report. It reads the object graph; it never rasterizes and never runs JavaScript.

use pdfrum_common::{Diagnostics, Limits};
use pdfrum_doc::page_label;
use pdfrum_object::{Array, Dict, Name, NoResolve, Object, PdfString};

// `/PageLabels` maps a *starting* page index to a rule; page 1 onwards is
// upper-case roman with the prefix `A-` (ISO 32000-1 §12.4.2).
let rule = Dict::from_pairs([
    (Name::from("S"), Object::Name(Name::from("R"))),
    (Name::from("P"), Object::Str(PdfString::literal(b"A-"))),
]);
let labels = Dict::from_pairs([(
    Name::from("Nums"),
    Object::Array(Array::of([Object::Int(1), Object::Dict(rule)])),
)]);
let catalog = Dict::from_pairs([(Name::from("PageLabels"), Object::Dict(labels))]);

let (limits, mut diags) = (Limits::default(), Diagnostics::default());
let label = page_label(&catalog, 2_u32, 3, &NoResolve, &limits, &mut diags);
assert_eq!(label.as_deref(), Some("A-II"));

Appearance generation returns an AnnotOverlay; it does not mutate the store. Readers take Option<&AnnotOverlay>, so the same annotation list can be read with generated appearances or without, and a caller that never asks for one pays nothing. DocOptions gates that generation, and both of its switches are off by default: the oracle builds every annotation list through the form-fill environment, which disables the /NeedAppearances widget path outright. Turn them on for the documented viewer behaviour instead.

A page label is a lower-bound lookup plus arithmetic, never a direct hit, and the two “no label” answers differ — a page index outside the document has no label at all, while a page inside it with no matching tree entry gets its one-based index as a decimal.

module	ISO 32000-1	what it reads
nav	§12.3	outline, destinations, actions, links, name and number trees
annot	§12.5	annotation dictionaries, subtypes, flags
ap	§12.5.5	appearance streams and the generated overlay
form	§12.7	the AcroForm field tree, as data
structure	§14.7	the tagged structure tree
page_label	§12.4.2	the page numbers a reader sees
pdfa	ISO 19005	conformance level and failed clauses

Part of pdfrum. #![forbid(unsafe_code)].

MIT OR Apache-2.0

Re-exports

pub use annot::AnnotFlags;

pub use annot::Annotation;

pub use annot::Subtype;

pub use ap::AnnotOverlay;

pub use ap::Focus;

pub use ap::FocusBox;

pub use ap::GeneratedAp;

pub use color::Color;

pub use error::Error;

pub use nav::AActionType;

pub use nav::Action;

pub use nav::ActionKind;

pub use nav::Bookmark;

pub use nav::Dest;

pub use nav::FileSpec;

pub use nav::Link;

pub use nav::NameTree;

pub use nav::ZoomMode;

pub use page_label::page_label;

pub use pdfa::Clause as PdfaClause;

pub use pdfa::Level as PdfaLevel;

pub use pdfa::Report as PdfaReport;

pub use prefs::ViewerPrefs;

pub use structure::StructElement;

pub use structure::StructTree;

Modules

annot

Annotations (ISO 32000-1 §12.5): the subtype table, the flag word, and the record every reader in this crate works from.

annot_render

Painting a page’s annotation appearances onto the page.

ap

Appearance-stream generation: turning an annotation’s dictionary into the content stream a viewer draws.

color

The annotation colour value (/C, /IC, /MK /BC, /MK /BG) and the three different ways PDFium converts it to bytes.

error

What can go wrong at this level.

form

The AcroForm data model (ISO 32000-1 §12.7): fields, their inherited attributes, their values, and the edits that write a filled form back out.

geom

PDF rectangle arithmetic with the exact semantics the annotation and appearance code depends on.

nav

Navigation: outlines, name and number trees, destinations, actions, file specifications and links.

page_label

Page labels (ISO 32000-1 §12.4.2): the page numbers a reader sees, which need not be 1, 2, 3.

pdfa

PDF/A conformance checking (ISO 19005): what a document fails, and why.

prefs

Viewer preferences (ISO 32000-1 §12.2): how a document asks to be displayed and printed.

signature

Signature fields, read as the file wrote them.

structure

The logical structure tree (ISO 32000-1 §14.7), read per page and bottom-up.

vt

The variable-text layout engine: a field value in, positioned glyphs out.

Structs

DocOptions

How much of the appearance-generation surface a document walk turns on.

Functions

xmp

The catalog’s /Metadata stream, decoded.