Bug 701474

Summary:	Crossreferences are gone during compression in version 9.27 and 9.26 while they are maintained in 9.20
Product:	Ghostscript	Reporter:	Bernhard Kleine <bernhard.kleine>
Component:	PDF Writer	Assignee:	Ken Sharp <ken.sharp>
Status:	RESOLVED DUPLICATE
Severity:	normal
Priority:	P4
Version:	9.27
Hardware:	PC
OS:	Windows 7
Customer:		Word Size:	---
Attachments:	The docbook testcase example of attaching PDF file This is the original pdf. this is the compressed file 9.20 with crossreferences intact. the cpmpressed file vers 9.26 without crossreferences

Description Bernhard Kleine 2019-09-01 10:53:02 UTC

A large pdf is normally compressed by the following batch file:

C:\"Program Files"\gs\gs9.20\bin\gswin64 -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dNOPAUSE -dBATCH -dDownsampleColorImages=true -dColorImageResolution=400 -dDownsampleGrayImages=true -dGrayImageResolution=400 -dDownsampleMonoImages=true -dMonoImageResolution=400 -dNOPAUSE -dBATCH -sOutputFile=FalterinLenzkirchV2_Compressed.pdf FalterInLenzkirchV2.pdf

In the version 9.20 the crossreferences were maintained. However, in the new versions (9.26/7) the crossreferences are lost while compression. Since the actual files I work with are very large, I can provide an example, if necessary.

Comment 1 Ken Sharp 2019-09-01 18:45:22 UTC

(In reply to Bernhard Kleine from comment #0)
> A large pdf is normally compressed by the following batch file:

Well actually, no. What you mean is a new PDF is produce which is smaller than the original. You should read this:

https://www.ghostscript.com/doc/9.27/VectorDevices.htm#Overview

to understand the process. Ghostscript's pdfwrite device does *not* compress PDF files.


> In the version 9.20 the crossreferences were maintained.

PDF files don't have an 'cross refrence' object. There are Link annotations and the Outlines tree. I have no idea whaich of these you are referring to.


> However, in the new
> versions (9.26/7) the crossreferences are lost while compression. Since the
> actual files I work with are very large, I can provide an example, if
> necessary.

Its completely impossible to investigate a problem without an example file, you should always provide teh simplest file and commadn line you can, which reprocues the problem.

Comment 2 Bernhard Kleine 2019-09-02 12:58:40 UTC

Created attachment 18068 [details]
The docbook testcase

Comment 3 Ken Sharp 2019-09-02 13:07:52 UTC

(In reply to Bernhard Kleine from comment #2)
> Created attachment 18068 [details]
> The docbook testcase

This is an XML file, Ghostscript doesn't handle XML.

Comment 4 Bernhard Kleine 2019-09-02 13:09:24 UTC

I have a minimal testcase which shows all the elements to show. However, a pdf is not allowed to upload. 

<?xml version="1.0" encoding="UTF-8"?>
<?xml-model href="http://docbook.org/xml/5.1/rng/docbookxi.rng" 
    schematypens="http://relaxng.org/ns/structure/1.0"?>
<?xml-model href="http://docbook.org/xml/5.1/sch/docbook.sch" 
    type="application/xml" 
    schematypens="http://purl.oclc.org/dsdl/schematron"?>
<book xmlns="http://docbook.org/ns/docbook" xml:lang="de" xmlns:xi="http://www.w3.org/2001/XInclude"
    xmlns:ns5="http://www.w3.org/2000/svg" xmlns:ns="http://docbook.org/ns/docbook"
    xmlns:xlink="http://www.w3.org/1999/xlink" version="5.1" 
     status="draft"
    >
    <chapter>
        <title>Chapter Test</title>
    <para>Here we test crossreferences<indexterm>
                <primary>test</primary>
            </indexterm></para>
        <para><indexterm>
                <primary>crossreferences</primary>
            </indexterm> in PDF<indexterm>
                <primary>PDF</primary>
            </indexterm></para>
        
    </chapter>
 
 
   
    <index>
        <title>Index</title>
    </index>
 
</book>

The batch file I have already provided.
I could send the pdfs if you let me know where to.

Comment 5 Ken Sharp 2019-09-02 13:16:48 UTC

Created attachment 18069 [details]
example of attaching PDF file

(In reply to Bernhard Kleine from comment #4)
> I have a minimal testcase which shows all the elements to show. However, a
> pdf is not allowed to upload. 

Its perfectly possible to add PDF files to our Bugzilla, a quick persue of the database would show *many* cases where PDF files have been attached. I'm not aware of this having changed, and I've attached a PDF file here as a test.

Possibly your end is blocking the attachment, I can't help you with that.

 
> The batch file I have already provided.
> I could send the pdfs if you let me know where to.

You need to attach the PDF file used as input to Ghostscript, I can't do anything with an XML file.

Comment 6 Ken Sharp 2019-09-02 14:26:07 UTC

Without the benefit of a file to look at, its impossible to be sure, but I strongly suspect this is a duplicate of bug #699896 which was itself a duplictae of bug #699830.

See especially comments 13 and 16 of bug #699830.

Comment 7 Bernhard Kleine 2019-09-02 14:38:43 UTC

Created attachment 18070 [details]
This is the original pdf.

Comment 8 Bernhard Kleine 2019-09-02 14:39:37 UTC

Created attachment 18071 [details]
this is the compressed file 9.20 with crossreferences intact.

Comment 9 Bernhard Kleine 2019-09-02 14:40:31 UTC

Created attachment 18072 [details]
the cpmpressed file vers 9.26 without crossreferences

Comment 10 Bernhard Kleine 2019-09-02 14:42:20 UTC

sorry for the noise. now the minimal pdfs are uploaded.

Comment 11 Ken Sharp 2019-09-02 15:03:26 UTC

See my comment #6.

Once again, Link annotations with no /F (Flag) value specified, which means they have a default value of 0, and are therefore 'non printing'.

In order to have the PDF interpreter process non-printing annotaions, you must set -dPrinted=false, by default the PDF interpreter is set to behave as a printer, so Printed is true.

A more lengthy explanation can be found in bug #699830. The behaviour change was due to *fixing* a bug, which caused the Printed flag to be incorrectly ignored in some cases.

*** This bug has been marked as a duplicate of bug 699830 ***