[poppler] Syntax Error: Unknown character collection 'DYNA-HK1'

[suzuki toshiya]

Hi,

As soon as this bug is filed in bugzilla, I will go to there,
but please let me post something before it...

In the PDF, I found such resource declaration:

...
76 0 obj
<< /Type /Font /Subtype /CIDFontType0 /BaseFont /VCGDKL+PingFangSC-Semibold
/CIDSystemInfo << /Registry (DYNA) /Ordering (HK1) /Supplement 0 >> /W 77 0 R
/DW 1000 /FontDescriptor 78 0 R >>
endobj
...

an embedded font "PingFangSC-Semibold" requires CMap
named "DYNA-HK1". Nothing to say, it is non-standard
CMap which the PDF producers are expected to embed
in PDF (and I could not find it in the PDF).

I'm interested in how the sample PDF was generated.
According to the metadata,

...
87 0 obj
(Mac OS X 10.11.1 Quartz PDFContext)
endobj
88 0 obj
(Keynote)
endobj
89 0 obj
(D:20151205141530Z00'00')
endobj
1 0 obj
<< /Title 86 0 R /Producer 87 0 R /Creator 88 0 R /CreationDate 89 0 R /ModDate
89 0 R >>
endobj
...

It seems that the application is "Keynote" and generated
via builtin PDF generator of Mac OS X. I think such
fundamental error (missing CMap with non-standard name)
is caused by poor handmade implementation, but it does not
look like so (if the metadata was not forged). Such error
could be caused by the font itself?

Regards,
mpsuzuki

Albert Astals Cid wrote:
> El Saturday 05 December 2015, a les 22:45:07, Martin Hong va escriure:
>> Hello guys,
> 
> Hi
> 
>> I am new to Poppler, I am developing a tool to support converting pdf to
>> jpeg files online in China. But I meet a strange error and I can not
>> resolve it through searching google. The below lists some details:
>>
>> ――――――――
>> $ pdftoppm -v
>> pdftoppm version 0.37.0
>> Copyright 2005-2015 The Poppler Developers - http://poppler.freedesktop.org
>> Copyright 1996-2011 Glyph & Cog, LLC
>> ――――――――
>> $ pdftoppm -png ~/Desktop/test.pdf tmp
>> Syntax Error: Unknown character collection 'DYNA-HK1’
>> ――――――――
>>
>> the `test.pdf` is exported from Mac OS’ Keynote.app and contains some
>> Chinese characters. If I try to export a pdf which doesn’t contain any
>> Chinese characters, it will work. Additionally, I also tried to use
>> pdftotext to print text in `test.pdf`, it can print text normally.
>>
>> You can also download the test pdf from here:
>> http://7xj84e.com1.z0.glb.clouddn.com/test.pdf
>> <http://7xj84e.com1.z0.glb.clouddn.com/test.pdf> .
>>
>> Hope someone can help to resolve this problem and thanks very much in
>> advance!
> 
> Please report bugs in the bug tracker.
> 
> http://bugs.freedesktop.org/
> 
> Cheers,
>   Albert
> 
>> Best Regards,
>> Martin
> 
> _______________________________________________
> poppler mailing list
> poppler at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/poppler