Just noticed that Bastien Dumont faced a same issue writing a filter for 
ODT output: https://github.com/jgm/pandoc/issues/7262 . 

On Monday, April 18, 2022 at 2:15:06 PM UTC+1 Julien Dutant wrote:

> And related question, shouldn't the WriterOptions Lua constructor include 
> a `standalone` field? 
>
> On Monday, April 18, 2022 at 2:10:12 PM UTC+1 Julien Dutant wrote:
>
>> Hi all,
>>
>> I'm writing a new Lua filter to handle theorems and the like in various 
>> format including JATS (https://github.com/jdutant/statement). Given the 
>> following markdown and a suitable bibliography:
>>
>> ```markdown
>> ::: theorem
>> (from @article) Some very interesting fact holds.
>> :::
>> ```
>>
>> The filter is trying to generate a JATS statement ( 
>> https://jats.nlm.nih.gov/archiving/tag-library/1.1/element/statement.html 
>> ):
>> ```xml
>> <statement>
>> <label>Theorem 1.1</label>
>> <title>from Doe, J (2003)</title>
>> <p>Some very interesting fact holds.</p>
>> </statement>
>> ```
>>
>> The problem I encounter is that the JATS writer converts pandoc.Plain to 
>> <p> blocks. So if I build inlines list:
>> inlines = { pandoc.RawInline('jats','<title>'), ... more inlines ..., 
>> pandoc.RawInline('jats','<title>')}
>>
>> and try to insert it in the document with:
>> blocks:insert( pandoc.Plain(inlines) )
>>
>> I get the unintended output:
>> ```xml
>> <p><title>from Doe, J (2003)</title><p>
>> ```
>>
>> Now I could of course stringify the title inlines first, add the title 
>> tags and insert the result in a RawBlock. But that's bad too, because the 
>> title inlines may contain things that aren't yet to be stringified like a 
>> citation. 
>>
>> Is it necessary for the JATS writer to turn Plain elements into <p> ones? 
>> The HTML one doesn't, after all.
>>
>> At the moment my best approach is to use `pandoc.write` (thanks so much 
>> Albert for giving us this!) to convert the inlines on the spot. There are 
>> some issues with this though.
>>
>> * If something in the inlines needs to be handled by another filter it 
>> won't be handled properly. For instance, a pandoc-crossref crossreference 
>> will be turned into plain text and unretrievable. AFAIK my filter can't 
>> tell which other filters are run and can't pass them as WriterOptions 
>> anyway.
>> * If I don't pass PANDOC_WRITER_OPTIONS, pandoc.write won't know which 
>> citation_mode to use, and may lack other relevant settings to write the 
>> inlines.
>> * If I pass all of PANDOC_WRITER_OPTIONS, pandoc.write uses standalone 
>> mode if the document is in standalone mode, so I get a whole preamble 
>> within my label. Yet there is no standalone setting in WriterOptions, and I 
>> can unset it (I've tried to remove `template` but it didn't work).
>> * So my best guess so far is to create a new WriterOptions by copying 
>> only those fields of  PANDOC_WRITER_OPTIONS that might be useful to format 
>> the inlines.
>>
>> All in all, it's pretty heavy handed just to handle a label of inlines. 
>> Is there a better approach? Any chance that the JATS writer converts Plain 
>> blocks to plain blocks without <p> tags?
>>
>> Best,
>> J
>>
>>

-- 
You received this message because you are subscribed to the Google Groups "pandoc-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pandoc-discuss+unsubscribe-/JYPxA39Uh5TLH3MbocFF+G/Ez6ZCGd0@public.gmane.org
To view this discussion on the web visit https://groups.google.com/d/msgid/pandoc-discuss/633d80cc-6595-4998-8f30-f15765ef1265n%40googlegroups.com.