Follow-up 2:
The 24 remaining complaints are caused by nesting
\startparagraph
This
\startformula
1+1=2
\stopformula
is true.
\stopparagraph
Not a big deal either.
Fingers crossed, I have my first compliant PDF. The structure I can see in
Acrobat looks good, too.
Matthias
> On Aug 13, 2025, at 2:03 PM, Matthias Weber <matthiaswebe...@gmail.com> wrote:
>
> Follow-up: The 66 complaints come from \startsection .. \stopsection pairs
> (and \startsubsection .. \stopsubsection).
> Adding just one of them increases the number of complaints by two.
>
> Instead of your
>
> \setupheads
> [section,subsection]
> [color=darkred]
>
> \setuphead
> [section]
> [style=\bfb]
>
> \setuphead
> [subsection]
> [style=\bfa]
>
>
>
> I use
>
> \setuphead[subsection][number=yes,
> page=yes,
> before={\noindentation\hairline\blank},
>
> after={\nowhitespace\hairline\blank[line]\indentation}]
>
>
> \setuphead[section][number=yes,
> page=yes,
> before={\noindentation\hairline\blank},
>
> after={\nowhitespace\hairline\blank[line]\indentation}]
>
>
> which is causing trouble somehow. No big deal.
>
> Matthias
>
>
>> On Aug 13, 2025, at 1:40 PM, Matthias Weber <matthiaswebe...@gmail.com>
>> wrote:
>>
>> Thank you so much Hans & Mikhael!
>>
>> I tested one of my lecture notes and get two complaints from VeraPDF:
>> (90 failed checks and 577444 passed checks…)
>>
>>
>> Specification: ISO 32005:2023, Clause: Table 5. P-Div, Test number: 1
>> <P> shall not contain <Div> Failed
>> 24 occurrences Hide
>> SEDiv
>> parentStandardType != 'P'
>> root/document[0]/StructTreeRoot[0](231 0 obj PDStructTreeRoot)/K[0](239 0
>> obj SEDocument document)/K[0](240 0 obj SEPart documentpart)/K[21](605 0 obj
>> SESect section)/K[1](611 0 obj SEDiv sectioncontent)/K[3](1770 0 obj SESect
>> section)/K[1](1775 0 obj SEDiv sectioncontent)/K[4](1876 0 obj SEP
>> paragraph)/K[2](1885 0 obj SEDiv formula)
>>
>>
>> and
>>
>> Specification: ISO 14289-2:2024, Clause: 8.2.2, Test number: 1
>> Content that is not considered real shall be an artifact Failed
>> 66 occurrences Hide
>> SESimpleContentItem
>> isTaggedContent == true || parentsTags.contains('Artifact') == true
>> root/document[0]/pages[1](2 0 obj PDPage)/contentStream[0](407 0 obj
>> PDSemanticContentStream)/content[3]/contentItem[0]
>>
>>
>> I am absolutely clueless about these two. Is there a way to find out what
>> the problem is and how I can fix it?
>>
>> Thanks again,
>>
>> Matthias
>>
>>
>>
>>
>>
>>> On Aug 13, 2025, at 11:12 AM, Hans Hagen via ntg-context
>>> <ntg-context@ntg.nl> wrote:
>>>
>>> Hi,
>>>
>>> Due to some pressure on this list we have been looking into reasonable
>>> mapping from context structure onto what pdf sees as structure as of 2025.
>>> This is not without risk as the pdf standard, tools, interpretations,
>>> whatever are in constant flux. We don't fall into the target audience so
>>> our decisions are guess work at best.
>>>
>>> That said, we now do the following. If you enable tagging
>>>
>>> \setuptagging[state=start]
>>>
>>> you get normal tagging, without so called rolemapping. This can work out ok
>>> with (future) more clever parsers of the content stream (think llm). It
>>> doesn't pollute the pdf too much but of course bloats it.
>>>
>>> Then, if you want to go fuzzy, you can say
>>>
>>> \setupbackend
>>> [format=pdf/ua-2]
>>>
>>> \setuptagging
>>> [state=start]
>>>
>>> This gives some basic mapping. Then there is a basic mapping that or with
>>> Mikael figured out and tested on lecture notes and a math book. Of course
>>> runtime increases (so the book now needs 14 sec instead of 8 but so be it
>>> we might be able to improve that a little bit but all that math ... it's
>>> not like one tags all day).
>>>
>>> \setuptagging
>>> [state=start,
>>> preset=basic,
>>> % preset={basic,\jobname},
>>> comment={just a basic test document}]
>>>
>>> You can kick in your own mappings, e.g. which could already be done with
>>> commands but we now prefer using an "imp" file as hard pdf mapping doesn't
>>> belong in a document setup. You, as user / author knows best what mapping
>>> is nneeded and can work. There is tagging-basic.tex document that explains
>>> this.
>>>
>>> We advise to delay tagging (if needed at all, which in my case is never) to
>>> when a document is finished (maybe by wrapping in a mode tagging) so that
>>> no runtime is wasted during writing / designing.
>>>
>>> You can test / check some with
>>>
>>> mtxrun --script pdf --structure [--detail --save] yourdocument
>>> mtxrun --script pdf --validate [--interaction] yourdocument
>>>
>>> where the second one calls for verapdf, or simply test with
>>>
>>> mtxrun --script pdf --check yourdocument
>>>
>>> this is only for testing purposes; we might add some to it depending on our
>>> own or user demand. The possibly generated file serves no other purpose
>>> (and obeys crap-in-crap-out).
>>>
>>> Of course, different versions of verapdf can result in different results
>>>
>>> By default we ignore interaction because that is a mess in tagged pdf but
>>> you can use option=interaction to get something if needed. We think that
>>> when a document validates with exception of link/references no one is
>>> harmed.
>>>
>>> After the context meeting we will freeze the basic mapping but one can of
>>> course always modulate on it,
>>>
>>> Hans & Mikael
>>>
>>> (who will now pick up on more rewarding and fun things that we'd put in the
>>> hold)
>>>
>>> -----------------------------------------------------------------
>>> Hans Hagen | PRAGMA ADE
>>> Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
>>> tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl
>>> -----------------------------------------------------------------
>>>
>>> ___________________________________________________________________________________
>>> If your question is of interest to others as well, please add an entry to
>>> the Wiki!
>>>
>>> maillist : ntg-context@ntg.nl /
>>> https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
>>> webpage : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
>>> archive : https://github.com/contextgarden/context
>>> wiki : https://wiki.contextgarden.net
>>> ___________________________________________________________________________________
>>
>
___________________________________________________________________________________
If your question is of interest to others as well, please add an entry to the
Wiki!
maillist : ntg-context@ntg.nl /
https://mailman.ntg.nl/mailman3/lists/ntg-context.ntg.nl
webpage : https://www.pragma-ade.nl / https://context.aanhet.net (mirror)
archive : https://github.com/contextgarden/context
wiki : https://wiki.contextgarden.net
___________________________________________________________________________________
