verapdf does not report problems for invalid hex strings in content stream

[shibaev]

I created a test PDF/A-1b document InvalidHexStrings.pdf, containing different invalid hex strings in a page content stream:

BT
<41 4> Tj
<42 4> '
30 60 <434420 4> "
T*
0 Tc
0 Tw
[<45 4> -1000 <46 5>] TJ
T*
T*
<4x1> Tj
<42> Tj
<43x> '
<44> '
30 60 <x45> "
<46> '
[<4x7> -1000 <48>] TJ
[<49> -1000 <4xA>] TJ
<4B> '
ET

/Artifact <</Name <45 4>>> DP

/Artifact <</Name <46x>>> DP

/Artifact <</Name <47 4>>> BDC
EMC

/Artifact <</Name <<x48>>> BDC
EMC

I expected to get 6.1.6-1 Hexadecimal strings shall contain an even number of non-white-space characters and
6.1.6-2 All non-white-space characters in hexadecimal strings shall be in the range 0 to 9, A to F or a to f. However, veraPDF 1.30.1 does not report any issues for this document. Adobe Acrobat Preflight reports both issues.

[shibaev]

The issue probably relates to non-hex characters like <43x> because verapdf properly reports 6.1.6-1 for InvalidHexStrings-content stream,odd length-1b.pdf. This document only uses odd-length strings.

[shibaev]

I found the root cause - the typo here /Artifact <</Name <<x48>>> BDC. Should be /Artifact <</Name <x48>>> BDC, in this case verapdf works properly.

Probably, you can ignore this issue because it's unlikely to found a real document with such an incorrect object.

[bdoubrov]

@shibaev thanks for bringing this up anyway. I see we do generate a log message Error while parsing content stream. But it only has a FINE level. We'll upgrade it to WARNING.