[Bug]: CMap parsing error

[15369411343]

Attach (recommended) or Link to PDF file

test-031.pdf

(test.z01.zip test.z02.zip test.zip)

Web browser and its version

Microsoft Edge 134.0.3124.93

Operating system and its version

Windows10

PDF.js version

4.10.38

Is the bug present in the latest PDF.js version?

Yes

Is a browser extension

No

Steps to reproduce the problem

1. Unzip file.
2. Open the test.pdf using PDF.js.
3. Navigate to page 31.

What is the expected behavior?

Text and watermark can be displayed normally.

What went wrong?

There are a lot of text display abnormalities starting from page 31, and there are watermark display abnormalities starting from page 55, but these abnormalities do not exist when the file is opened directly in the Edge browser.

Link to a viewer

No response

Additional context

No response

[Snuffleupagus]

Attach (recommended) or Link to PDF file

test.z01.zip test.z02.zip test.zip

Sorry, but why are you attaching ZIP files where we're asking for a PDF? Please just attach the PDF directly like we ask!
It seems like the first two archives don't contain any PDF file, and while the last one does it however fails to extract on Windows 11.

PDF.js version

4.10.38

That version is no longer supported, please find the latest releases at https://mozilla.github.io/pdf.js/getting_started/#download

Is the bug present in the latest PDF.js version?

Yes

Sorry, but did you actually try the latest version?

[15369411343]

Sorry, because my file is quite large, I can only upload it after compressing it in multiple volumes. test.z01.zip and test.z02.zip are sub-volumes of test.zip. Please first decompress test.z01.zip and test.z02.zip separately before decompressing test.zip.

I did try the latest versions via these two links:
Modern browsers: https://mozilla.github.io/pdf.js/web/viewer.html
Older browsers: https://mozilla.github.io/pdf.js/legacy/web/viewer.html

[Snuffleupagus]

The CMap errors are unrelated, and the bug is rather in the PDF document itself since it contains a non-standard and non-embedded font (SimSun). Since that font uses an /Identity-H encoding and has no /ToUnicode data, it's not possible to create a "correct" glyph-mapping.

[15369411343]

Thank you very much for your reply!

I found that the web version of Foxit Reader can render this file properly. It seems that they used the "queryLocalFonts" method to obtain all the local font data in advance. Maybe this is a good idea to solve this kind of font conversion problem.

By the way, MuPDF can also display this document. However, where there are garbled characters in PDF.js, MuPDF does show the content but the character spacing is extremely close.

[15369411343]

I solved the problem locally by modifying the code in pdf.worker.mjs to set SimSun as the standard font. SimSun is indeed pre-installed as a standard font in the Chinese version of Windows.

[wanlingyun]

I solved the problem locally by modifying the code in pdf.worker.mjs to set SimSun as the standard font. SimSun is indeed pre-installed as a standard font in the Chinese version of Windows.

I have encountered the same problem. Can you tell me how to set it up

[15369411343]

> I solved the problem locally by modifying the code in pdf.worker.mjs to set SimSun as the standard font. SimSun is indeed pre-installed as a standard font in the Chinese version of Windows. I have encountered the same problem. Can you tell me how to set it up

Please contact me by email. I have left my email address in my personal profile