@pdfa-lab/core / PDFALab

Class: PDFALab

Defined in: pdfa-lab.ts:99

Accessors

pdfDocument

Get Signature

get pdfDocument(): PDFDocument

Defined in: pdfa-lab.ts:112

Returns the internally used PDFDocument. Changing the structure of this document, and then calling other PDFALab methods, may result in undefined behaviour and is strongly discouraged.

Returns

PDFDocument

the internally used PDFDocument

Methods

collectFonts()

collectFonts(): Map<string, FontInfo>

Defined in: pdfa-lab.ts:352

Collects all fonts used in the PDF.

Returns

Map<string, FontInfo>

a Map with keys as PDFRef (reference of the font) and values as FontInfo

---

embedFonts()

embedFonts(fontkit, options?, references?): Promise<void>

Defined in: pdfa-lab.ts:232

Embed multiple fonts, but only if they are not already embedded.

If no references were passed, all currently missing fonts are embedded.

Parameters

Parameter	Type	Description
fontkit	FontkitAPI	a fontkit instance
options	FontEmbedOptions	control the font embedding
references?	PDFRef[]	can be determined by collectFonts()

Returns

Promise<void>

---

extractText()

extractText(): Promise<TextBlock[]>

Defined in: pdfa-lab.ts:369

Extract all tags from the PDF.

Returns

Promise<TextBlock[]>

an array of TextBlock objects.

---

extractXmp()

extractXmp(format?, baseIRI?, options?): string | null

Defined in: pdfa-lab.ts:389

Extract XMP meta information from the PDF.

Parameters

Parameter	Type	Default value	Description
format	string	'application/rdf+xml'	the desired serialisation format (default: application/rdf+xml)
baseIRI	string	DEFAULT_BASE_IRI	the base IRI (default: urn:xmp:doc)
options	RdfSerialisationOptions	{}	serialisation options for certain formats, see RdfSerialisationOptions

Returns

string | null

the serialised XMP or null if no meta information available

---

makePDFA()

makePDFA(options?): Promise<void>

Defined in: pdfa-lab.ts:213

Convert a document to PDF/A.

Parameters

Parameter	Type	Description
options	PDFAConversionOptions	-

Returns

Promise<void>

---

save()

save(): Promise<Uint8Array<ArrayBufferLike>>

Defined in: pdfa-lab.ts:128

Serialises the internally used PDF into a sequence of bytes. As a side effect, it also updates internal structures of the PDFDocument into a stable version.

The method is a thin wrapper around the save() method of PDFDocument. Fine-tuning the saving process can be achieved by getting the pdfDocument, and then calling the save() method with the desired options.

Returns

Promise<Uint8Array<ArrayBufferLike>>

---

setXmpMetadataPacket()

setXmpMetadataPacket(rdfxml): void

Defined in: pdfa-lab.ts:429

Parameters

Parameter	Type
rdfxml	string

Returns

void

---

from()

static from(input): Promise<PDFALab>

Defined in: pdfa-lab.ts:176

Creates a PDFALab instance from a variety of PDF-like inputs and normalizes them into a single canonical @cantoo/pdf-lib document representation.

This method performs a pre-processing step to ensure that the returned PDFDocument has a consistent internal object graph.

Why

PDF libraries such as pdf-lib (and forks like @cantoo/pdf-lib) rely heavily on runtime identity checks such as instanceof PDFDict, instanceof PDFArray, etc.

In bundled or duplicated dependency scenarios (e.g. monorepos, multiple builds, or linked packages), the same logical PDF object type may exist in multiple constructor instances. This causes instanceof checks to fail even though the objects are structurally identical, leading to weird runtime errors such as:

• "Expected instance of PDFDict but got instance of PDFDict"

To prevent this, the document is always serialised and reloaded. This forces all internal objects to be reconstructed using the active @cantoo/pdf-lib runtime, ensuring consistent prototype chains and reliable instanceof behaviour.

Behaviour

• If a raw PDF input is provided (base 64 encoded string, data URI, or binary), it is directly loaded.
• If a PDFDocument from another runtime instance is detected, it is first serialized via save() and then reloaded via PDFDocument.load().
• This process guarantees a canonical internal representation of the PDF.

This step is intentionally non-trivial and may be expensive, but is required for correctness in environments where multiple copies or builds of pdf-lib may exist.

Parameters

Parameter	Type	Description
input	string | ArrayBuffer | Uint8Array<ArrayBufferLike> | PDFDocument	A PDF source: raw bytes, base64 string, data URI, or a PDFDocument instance.

Returns

Promise<PDFALab>

A normalized PDFALab instance backed by a canonical PDF document.