PDFjs.getMetadata() - XML parse bug

[dhollenbeck]

When getMetadata() is invoked on a specific PDF file, which I can not provide:

var loader = pdfjs.getDocument(source);
loader.promise.then(function getDocumentSuccess(pdf) {
  pdf.getMetadata().then(function () {}).catch(...);
});

I get the following error:

(node:7452) TypeError: Cannot read property '0' of undefined
    at SimpleDOMNode.get (...\node_modules\pdfjs-dist\build\pdf.js:14097:29)
    at Metadata._parse (...\node_modules\pdfjs-dist\build\pdf.js:13740:19)
    at new Metadata (...\node_modules\pdfjs-dist\build\pdf.js:13698:12)
    at ...\pdfjs-dist\build\pdf.js:9421:34
    at <anonymous>
    at process._tickCallback (internal/process/next_tick.js:188:7)

I believe the following line of code is the line with the bug:

pdf.js/src/display/xml_parser.js

Lines 280 to 282 in 2ce489c

	get firstChild() {
	return this.childNodes[0];
	}

Obviously, this.childNodes could be undefined as is the case with my pdf file. It is being invoked via an automatic property get function above:

pdf.js/src/display/metadata.js

Line 77 in 66ffdc4

rdf = rdf.firstChild;

The offending XML document has the structure of:

{ documentElement: SimpleDOMNode { nodeName: '#text', nodeValue: '' } }

[timvandermeij]

@dhollenbeck Would it be possible to provide a test case without sensitive data here as well like you did for the other issue? Either a file or the XML itself is fine. This would allow us to add a unit test for this so we can avoid regressions in the future, and allow us to verify the fix.

[dhollenbeck]

Absolutely, @timvandermeij
toissue.pdf