validation: info dict: missing object

[HuabeiYou]

Hi there,

I was trying to add watermarks on pdf files using api.AddTextWatermarksFile but got the error validation error: pdfcpu: validateString: missing object for some of them. It seems like the error is coming from pdfcpu/validate/info.go line 60 to 64: if xRefTable.ValidationModa == pdf.ValidationRelaxed { ... }.

I then tried the CLI mode.
validate -m relaxed F18E1000.pdf gives error:

validating(mode=relaxed) F18E1000.pdf ...
validation error: pdfcpu: validateString: missing object

However pdfcpu validate -m strict F18E1000.pdf worked

validating(mode=strict) F18E1000.pdf ...
validation ok

I attached the sample pdf below. Also, is there a way to change the validation mode and make api.AddTextWatermarksFile work?

Thanks!
F18E1000.pdf

[HuabeiYou]

Hi there,

I was trying to add watermarks on pdf files using api.AddTextWatermarksFile but got the error validation error: pdfcpu: validateString: missing object for some of them. It seems like the error is coming from pdfcpu/validate/info.go line 60 to 64: if xRefTable.ValidationModa == pdf.ValidationRelaxed { ... }.

I then tried the CLI mode.
validate -m relaxed F18E1000.pdf gives error:

validating(mode=relaxed) F18E1000.pdf ...
validation error: pdfcpu: validateString: missing object

However pdfcpu validate -m strict F18E1000.pdf worked

validating(mode=strict) F18E1000.pdf ...
validation ok

I attached the sample pdf below. Also, is there a way to change the validation mode and make api.AddTextWatermarksFile work?

Thanks!
F18E1000.pdf

Update: Found out I can change the validation mode by passing in the Configuration, but it didn't solve my problem, here is the error:

• Add text watermark using ValidationNone:

text: corrupt -  <<
	<BaseFont, Helvetica>
	<Encoding, WinAnsiEncoding>
	<Subtype, Type1>
	<Type, Font>
>>

• Add text watermark using ValidationStrict:

dict=type1FontDict required entry=FirstChar missing

[hhrutter]

OK, I shall check it out.
Thank you for reporting this 💚

PS: relaxed validation is the default for pdfcpu validate.

[hhrutter]

Looks like this file was created/modified on an Apple device (possibly Mac Preview) and then processed with "pdf-tools".
All info dict entries using indirect references point to empty objects:

175:   offset=  841786 generation=0 pdfcpu.Dict
<<
	<AAPL:Keywords, (180 0 R)>
	<CreationDate, (178 0 R)>
	<Creator, (177 0 R)>
	<Keywords, (179 0 R)>
	<ModDate, (D:20180922122321Z)>
	<Producer, (3-Heights\(TM\) PDF Security Shell 4.8.25.2 \(http://www.pdf-tools.com\) / pdcat \(www.pdf-tools.com\))>
	<Title, (176 0 R)>
>>
  176: f   next=     177 generation=0
  177: f   next=     178 generation=0
  178: f   next=     179 generation=0
  179: f   next=     180 generation=0

We are working on a fix.

Thank you for using pdfcpu 💚

[hhrutter]

Fixed with latest commit.