PDF not rendering and title displayed as \376\377\000A\000f\000s\000c\000h\000r\000i\000f\000t\000 \000b\000r\000i\000e\000f\000 \000a\000a\000n\000 \000k\000o\000n\000i\000n\000g\000 \000B\000e\000a\000t\000r\000i\000x\000 \000o\000v\000e\000r\000 \000o\

[user1113]

Hi,

This pdf does not render in PDF.js, and the title is shown as \376\377\000A\000f\000s\000c\000h\000r\000i\000f\000t\000 \000b\000r\000i\000e\000f\000 \000a\000a\000n\000 \000k\000o\000n\000i\000n\000g\000 \000B\000e\000a\000t\000r\000i\000x\000 \000o\000v\000e\000r\000 \000o\000n\000t\000s\000l\000a\000g\000 \000k\000a\000b\000i\000n\000e\000t instead of the actual title Afschrift brief aan koning Beatrix over ontslag kabinet.

Adobe gives the following additional information about it:

• Title
  • Afschrift brief aan koning Beatrix over ontslag kabinet
• Author
  • rvd527
• Application
  • PDFCreator Version 0.9.9
• PDF-version
  • 1.4
• Included fonts
  • Calibri
  • Verdana regular+bold+italic
  • (All are TrueType)
• It meets the PDF/A spec

(Of course, you can also download the document itself at the link given.)

Screenshots

My information:
Firefox add-on version 0.2.537
User-Agent: Mozilla/5.0 (Windows NT 6.0; rv:11.0) Gecko/20100101 Firefox/11.0

[gigaherz]

The title looks like it may be utf-16 encoded as ascii but with the non-ascii characters encoded in octal...
@saebekassebil may be interested in that since he wrote the metadata decoding stuff.

There doesn't seem to be any messages in the console so I don't know about the display issue.

[gigaherz]

For reference: the string begins with 0xFE,0xFF, which is the Byte order Mark for UTF-16BE (character 0xFEFF), this makes it clear how to decode the string to unicode or whatever the internal representation of javascript is.

[saebekassebil]

The title in the embedded Metadata XMP document is:
"\\376\\377\\000A\\000f\\000s\\000c\\000h\\000r\\000i\\000f\\000t\\000 \\000b\\000r\\000i\\000e\\000f\\000 \\000a\\000a\\000n\\000 \\000k\\000 o\\000n\\000i\\000n\\000g\\000 \\000B\\000e\\000a\\000t\\000r\\000i\\000x\\000 \\000o\\000v\\000e\\000r\\000 \\000o\\000n\\000t\\000s\\000l\\000a\\000g \\000 \\000k\\000a\\000b\\000i\\000n\\000e\\000t"

So actually the whole title is there, but padded with NUL characters. This is probably because your PDF Generator is using a wide-character set (UTF-16) set instead of only UTF-8. I think yury is fixing this with the PR right above.

[gigaherz]

Yes he is, and yes it is exactly that, as I said, it's UTF-16BE, because the octals for \376\377 translate to 0xFE,0xFF, which is the big-endian ordering for 0xFEFF, the Byte-order-mark (a Zero-width-no-break-space code which apps loading UTF-16 use to deduce the endianess).

[yurydelendik]

inside pdf format structures the title is set and is read by pdf.js properly. the xmp/xml data is set to exactly the same vales as the fields inside pdf format, which is wrong -- the xml encoding shall be used.

the information above is taken from adobe reader pdf metadata pdf; could you also check xmp tab if one is present?

[user1113]

could you also check xmp tab if one is present?

I don't see any XMP tab.

[brendandahl]

The issue should be resolved with the above pull requests. If I overlooked anything please leave a comment and we'll re-open the issue.