Windows.Data.Pdf.PdfDocument Crash with file size greater than 2GB

Jérôme S 5

Same issue as here (no replies there, it's an old thread)
https://zcusa.951200.xyz/en-us/answers/questions/248422/windows-data-pdf-pdfdocument-crash

My notes:

Crash is not catchable which makes it really hard to find reason for issue or workaround.

I've tried the file in the official demo sample app and it crashes too. The file can be opened in other PDF software (including Edge - although slowly - which is funny because I think it uses the Windows.Data.Pdf library as well if I'm correct?)

After a bit more research, it seems the size is indeed what makes it crash. Another PDF lib I tried crashed similarly when the stream length exceeded Int32.MaxValue. The exception was catchable in this one so I could get a proper error message.

My laptop has 16GB of ram (nothing else running really) and 140GB free SSD space so I hope that's not the issue.

I can provide a storage link to a crashing file if necessary.

Same error on Windows 10 and Windows 11, both 64 bits, app is 64 bits.

Junjie Zhu - MSFT 19,131 Reputation points Microsoft Vendor

2024-11-13T02:39:50.7+00:00

Hello @Jérôme S ,

Welcome to Microsoft Q&A!

Could you please provide the code snippet and the pdf file to help us further check the problem for you? and be careful not to include any personal privacy.

Thank you.
Jérôme S 5 Reputation points

2024-11-13T07:58:32.9566667+00:00
Hi, sure, you can use the code from the sample:

https://github.com/microsoft/Windows-universal-samples/tree/main/Samples/PdfDocument/cs

Specifically that is the code line that will crash:

await PdfDocument.LoadFromFileAsync(file, PasswordBox.Password);

and this is a file that will crash (avoiding bots - remove REMOVE from the link below)

https://drive.google.com/file/d/1comy_coG5P6LREMOVElxLfsoe6F-QfEEaZpKxJ/view?usp=sharing
Junjie Zhu - MSFT 19,131 Reputation points Microsoft Vendor

2024-11-14T02:57:00.3433333+00:00

Hello @Jérôme S ,

Could you please share it through GitHub?
Jérôme S 5 Reputation points

2024-11-14T08:35:52.68+00:00

Not sure what you mean. Share a 2GB+ file through github?

if you can't access google drive I can use wetransfer or onedrive..
Junjie Zhu - MSFT 19,131 Reputation points Microsoft Vendor

2024-11-15T08:38:44.5033333+00:00

Due to policy reasons, I cannot download files on OneDrive or Google drive.
Jérôme S 5 Reputation points

2024-11-15T21:00:43.01+00:00

Would it work for you if I put it somewhere on an azure blob storage? Or maybe if you have policy restrictions you can provide a link where I can send it to.
Jérôme S 5 Reputation points

2024-11-18T09:21:42.5833333+00:00

I've put it here. Let me know if that's okay. Same thing remove REMOVE.

https://frenchfryfiles.blob.core.windows.net/cover/REMOVE6385f8b5-6597-4b0b-8c7e-cfe258d3a95a.pdf
EckiS 906 Reputation points

2024-11-18T11:39:31.0166667+00:00

the example project has
<Prefer32Bit>true</Prefer32Bit>
so even on Windows 64bit the process can not use more then 2GB of memory.
does it also crash when you set Prefer32Bit to false?
Jérôme S 5 Reputation points

2024-11-18T18:32:06.81+00:00

I'm not using AnyCPU but x64 when building, so as far as I know it should be able to use more than 2GB of memory.

I tried setting Prefer32Bits to false and it still crashes identically.
Junjie Zhu - MSFT 19,131 Reputation points Microsoft Vendor

2024-11-19T09:45:48.71+00:00

When I opened the link, the content is

This XML file does not appear to have any style information associated with it. The document tree is shown below. <Error> <Code>BlobNotFound</Code> <Message>The specified blob does not exist. RequestId:bd1deabd-201e-0004-2b67-3a3200000000 Time:2024-11-19T09:44:27.7151598Z</Message> </Error>

Share via

Windows.Data.Pdf.PdfDocument Crash with file size greater than 2GB

Your answer