xinvoice/ts/formats/pdf/extractors/associated.extractor.ts

79 lines
2.8 KiB
TypeScript
Raw Normal View History

import { PDFDocument, PDFDict, PDFName, PDFRawStream, PDFArray, PDFString } from '../../../plugins.js';
import { BaseXMLExtractor } from './base.extractor.js';
/**
* Associated files extractor for PDF/A-3 documents
* Extracts XML from associated files (AF entry in the catalog)
* Particularly useful for ZUGFeRD v1 and some Factur-X documents
*/
export class AssociatedFilesExtractor extends BaseXMLExtractor {
/**
* Extract XML from a PDF buffer using associated files
* @param pdfBuffer PDF buffer
* @returns XML content or null if not found
*/
public async extractXml(pdfBuffer: Uint8Array | Buffer): Promise<string | null> {
try {
const pdfDoc = await PDFDocument.load(pdfBuffer);
// Try to find associated files via the AF entry in the catalog
const afArray = pdfDoc.catalog.lookup(PDFName.of('AF'));
if (!(afArray instanceof PDFArray)) {
console.warn('No AF (Associated Files) entry found in PDF catalog');
return null;
}
// Process each associated file
for (let i = 0; i < afArray.size(); i++) {
const fileSpec = afArray.lookup(i);
if (!(fileSpec instanceof PDFDict)) {
continue;
}
// Get the file name
const fileNameObj = fileSpec.lookup(PDFName.of('F')) || fileSpec.lookup(PDFName.of('UF'));
if (!(fileNameObj instanceof PDFString)) {
continue;
}
const fileName = fileNameObj.decodeText();
// Check if it's a known invoice XML file name
const isKnownFileName = this.knownFileNames.some(
knownName => fileName.toLowerCase() === knownName.toLowerCase()
);
// Check if it's any XML file or has invoice-related keywords
const isXmlFile = fileName.toLowerCase().endsWith('.xml') ||
fileName.toLowerCase().includes('zugferd') ||
fileName.toLowerCase().includes('factur-x') ||
fileName.toLowerCase().includes('xrechnung') ||
fileName.toLowerCase().includes('invoice');
if (isKnownFileName || isXmlFile) {
// Get the embedded file dictionary
const efDict = fileSpec.lookup(PDFName.of('EF'));
if (!(efDict instanceof PDFDict)) {
continue;
}
// Get the file stream
const fileStream = efDict.lookup(PDFName.of('F'));
if (fileStream instanceof PDFRawStream) {
const xmlContent = await this.extractXmlFromStream(fileStream, fileName);
if (xmlContent) {
return xmlContent;
}
}
}
}
console.warn('No valid XML found in associated files');
return null;
} catch (error) {
console.error('Error in associated files extraction:', error);
return null;
}
}
}