Is there possibilities to import some file formats into fiduswriter, e.g. LaTeX, docx?
Importing to fiduswriter
Hey @opus,
The thing we currently have in place is a paste-analyzer. When you paste you paste content, it analyzes where it comes from and tries as much as possible to make it work well based on the source. So for example, it can detect that certain content comes from Word or from Google Docs and then treats that correspondingly.
This means it could be improved. It would be nice to pharse citation if it can find reference in bibliography file. Importing from docx and odt would certainly be very usable. Maybe we could make Template in FidusWriter, authors could use it for writing and submitting to journals and than journal editors could import it and export to JATS, LaTeX and HTML. I propose this workflow because there are always to be authors working offline.
Yes, so citation information is the one big thing we cannot get out of the pasted content from the word processors. I actually filed a ticket with LibreOffice about this exact thing a while back: https://bugs.documentfoundation.org/show_bug.cgi?id=116049 . Likely that won’t get fixed anytime soon or ever, so then the other option is for someone to either donate import code for ODT/DOCX or to pay for the programming of it. It’s certainly something we would be interested in, but it’s not something we can finance by ourselves at this stage.
Thank you, I understand now. Could this be resolved with pandoc maybe, odt2fidus file? I would like to hear more about fidus file.
I did check pandoc a few years back, and at least it didn’t look like it had sufficient information for citations and I thought we handled a lot of the imports better with our paste analysis. Likely pandoc has improved since then. Maybe the easiest way to get a number of reasonable importers is to just write one importer from pandoc’s main format?
The fidus file is basically a zip file with some json files inside. The format changes with every Fidus Writer version as we add new features (and in seldom cases remove some). We thought once that we would switcht o a standardized format as the main file, but that doesn’t really work because it restricts both what features we can add and remove to the editor, and it doesn’t allow us to do fixes to the way we store certain things in newer versions.
For example, if in version 2.2 we add support for 5 new languages in which we let users store their documents, we will then need to go to the committee that controls that file format to get those 5 new languages added. Or we added tracked changes in version 3.5 and if we did not have control over the file format, we would then need to wait for te standard we had chosen to also support that before we could release.
For that reason we keep Fidus Writer as the native fromat for Fidus Writer, and we make sure that users always can import their older fidus document, but it’s not a good idea to use it as an interchange format.
Yes Pandoc referencing is now much better. Ragarding Maybe the easiest way to get a number of reasonable importers is to just write one importer from pandoc’s main format?" could that be markdown or HTML?
Well, you tell me. Which of the formats carries most semantic information? I would have guessed it would be JATS, but that doesn’t contain all the information needed in relation to citations.
There is also the non-documented pandoc native format called “JSON AST”. I wonder if they have the same issue with that as we do with the fidus format - that they don’t really want others to convert to that format as they change it between each version.
Yes, there is custom pandoc writer that emits the article’s data as JSON–LD in Pandoc Scholar. Also is this project
Converter still aplicatable with new version of FidusWriter?
@opus Fidus Writer now has an export filter that is much better than that Java project from 2016. I believe the import filter through paste is also better, but I don’t remember all the details of that project. As far as I can recall, he used a pre-existing framework to parse and write docx, which meant that he could not add anything that this framework didn’t support. So that meant only support for a specific very limited form of citations which is what Word supports natively.
The project is written in Java, which is a different language than what is used in Fidus Writer and it has made it hard for us to maintain after the student who wrote that was done with his project. I don’t think it currently works, but if it can do anything that we currently cannot do, it should not be too difficult to update it.
Good to know. Thank you for awesome information!