Repository navigation
[Bug]: $nrOfBytes must be greater than 0, -324 given #295
Description
Is there an existing issue for this?
- I have searched the existing issues and found no similar reports
Are you using the latest version of this package?
- The issue I'm reporting exists in the latest release
Can other PDF readers read the file?
- The PDF I'm trying to read opens correctly in at least one other PDF reader
When running this snippet
$pdfParser = new PdfParser();
$pdfParsed = $pdfParser->parseString($pdfContent);
$fileSpecs = $pdfParsed->getCatalog()->getFileSpecifications();
$fileSpecs = array_filter(
$fileSpecs,
static function ($fileSpec) {
return !is_null($fileSpec->getEmbeddedFile());
}
);
foreach ($fileSpecs as $fileSpec) {
$this->attachmentList[] = new Attachment(
$fileSpec->getEmbeddedFile()->getStream()->toString(),
$fileSpec->getFileSpecificationString(),
ltrim($fileSpec->getEmbeddedFile()->getSubType() ?? '', '/')
);
}The error occurs when calling "$fileSpec->getEmbeddedFile()->getStream()->toString(),"
I run into the following issue/exception (Please attach the pdf)
/home/user/vendor/prinsfrank/pdfparser/src/Stream/InMemoryStream.php:26
/home/user/vendor/prinsfrank/pdfparser/src/Document/Object/Item/CompressedObject/CompressedObjectContent/CompressedObjectContentParser.php:27
/home/user/vendor/prinsfrank/pdfparser/src/Document/Object/Item/UncompressedObject/UncompressedObject.php:89
/home/user/vendor/prinsfrank/pdfparser/src/Document/Object/Decorator/DecoratedObject.php:33
Do you allow attachment files to be used in tests to prevent regressions?
- Yes, I give permission to use this file as a test file to prevent future regressions (And am authorized to give this permission)
Activity
PrinsFrank commentedon Dec 23, 2025
Sorry for the late reply, I was on holiday for a bit. I have found the issue here, the parser currently doesn't handle nested object markers properly, where a PDF object contains a bunch of other PDF objects. I'll keep you updated on a fix!
PrinsFrank commentedon Dec 23, 2025
I wrote a naive implementation for backward marker search from the next byte offset for objects, but that doesn't work on incrementally updated PDFs. I'll look at a more thorough implementation over the next few days, I'll keep you updated about a proper fix!
Do you see this issue in multiple embedded PDFs or is this a bug that only occurs with some files?
horstoeko commentedon Dec 24, 2025
Hi @PrinsFrank
Thank you very much for your support. It's not that urgent – enjoy the Christmas holidays.
I will indeed need to check that. At the moment, I have only come across this one PDF. I have not yet looked at the other sample PDFs from the official ZUGFeRD/Factur-X documentation.
Kind regards and happy Christmas
niranjanbai commentedon Mar 7, 2026
I occasionally notice this issue occurring on our server as well.
$nrOfBytes must be greater than 0, 0 givenAny update on fixing this issue?
horstoeko commentedon May 8, 2026
Hi @PrinsFrank
I saw that you've released a new version – I'm glad to hear that, as I'd already thought the project had been abandoned... Unfortunately, the bug mentioned in that issue is still there. How can I help?
Kind regards
PrinsFrank commentedon May 9, 2026
@horstoeko sorry about the delay. I've written some context in #320. In short: I'm back! And I'll fix this soon! I just need to find a few hours to focus on this, I'll keep you updated!
horstoeko commentedon May 9, 2026
Hi @PrinsFrank
It’s really good to hear from you. I hope you’re well. I’m also delighted that you may have found a solution to my problem. I look forward to hearing from you or reading your reply.
PrinsFrank commentedon May 21, 2026
@niranjanbai I've fixed issues with support for PDF files within PDF files. If you do see this with other files, please open a new issue. ;)
PrinsFrank commentedon May 21, 2026
@horstoeko This has been fixed! I'll release an alpha release for v3 soon, (that also supports encrypted PDFs), can you try the main branch for now!