Is someone using logseq with jabref?

I recently found logseq and it seems to fit my needs quite nicely. However, there is the only drawback that it seems to not work that great if you already have a lot of information stored in JabRef.

A usecase for me would be literature search, where I find it easier to collect thoughts in logseq than it is in JabRef. I found the logseq-citation-plugin which allows me to load a bib file and link to files, but unfortunately, the new format for the file entry does not really work with that plugin - but maybe that could be fixed on either side: I already created a bugreport for the plugin, but I wonder if it is possible to configure Jabref to write simply the filepath and not the description and filetype?

Any ways, I wonder if someone here is using JabRef in combination with logseq and might want to give me some more ideas what workflows work and how it can be configured?

3 Likes

I’m using logseq now for a month and I still like it. I also tried obsidian, but personally I like logseq a bit more - I think everything works a bit more fluent for me there.

The bibtex integration works, but there are a few things that do not work that well:

  • If you link the library to the folder the plugin requires, the plugin can no longer automatically re-index the library, when the bibtex file changes.
  • The plugin has a weird search field and you can, for example, not search for the citation key
  • I have not figured out how to link the files - it is possible, I have manually tried it - but from the plugin it does not work, because jabref stores more information than just the file path

Some thoughts, or ideas that could solve some of these problems:

  • an export option in jabref would be nice, to automatically store the library on-save somewhere else too
  • this export option should also allow to overwrite or normalize certain fields, such as the filename field (an option to export there the absolute path without the metadata would be nice)

Would be nice to hear some thoughts of other logseq+jabref users!

3 Likes

I tried, but at that time relative paths did not work, so you’d have to copy all files into the assets/ folder. Similar to this issue: Relative path for Zotero data directory - Feature Requests - Logseq

Yes, I have the same issue. Relative paths do not work and it looks like all trailing path components like .. are stripped from the path.

I found out you can circumvent this, by using something like this:

![does not work file path](file://../../papers/files/MyPaper.pdf)  # gives an error :(

![hacky file path](file://assets/../../papers/files/MyPaper.pdf)  # weird that this works...

I have the following structure on my disk:

papers/
 \--- references.bib  # JabRef library
 \--- files/
        \--- (all PDF files)
logseq/
  \--- assets/
  \--- ...  # All other logseq stuff

But unfortunately, I have not figured out how to automatically link the file path from the citation plugin, as the Jabref metadata in the files field interferes with the path as noted above…

2 Likes

That path issue is a shame. Otherwise Logseq is an amazing program.
Are you also using this plugin? GitHub - sawhney17/logseq-citation-manager
I think this is what I tried, but there is also direct Zotero integration, but it only works with the Zotero web server. Comprehensive Zotero Plugin - Plugins - Logseq
The plugin above was designed to work with Zotero exported .bib files, are they not the same as Jabref bib files?

Yes, that’s the one I’m using too.
I opened a feature request on github as well: parse JabRef file entries · Issue #44 · sawhney17/logseq-citation-manager · GitHub

I guess the bibtex formats are not different between zotero and jabref, but the difference is in the files field.

2 Likes

FIY, some work is being carried out related to relative/absolute path in JabRef. See https://github.com/JabRef/jabref/pull/9433

I’ve been trying to integrate the JabRef database with the plugin - did you get PDF files to work in the end? Being able to open files directly from logseq would be amazing.

I have a similar directory setup, but right now, it is difficult to implement - the v3 plugin requires the database to be inside logseq/assets/.

I simply symlinked the .bib file into that folder, required by the plugin. That works very well.
I read that a relative path should work as well - but it did not for me.
There is this PR, which I believe should fix that: Fixes #57; clarify behavior in plugin settings. by jmuchovej · Pull Request #58 · sawhney17/logseq-citation-manager · GitHub

No, it all depends on the correct handling of the information in the file field. If that is working, you should be able to open the PDFs without problems.

Hello, I have been using JabRef for 2 months and just started Logseq several days ago. I would like also to use the PDF annotation feature of logseq to keep track on my literature notes (later used as citations in a permanent note)…

I was trying to set up logseq-citation-manager but it looks really hard. I am encountering the same problems you mentioned in this thread.

If you can share the way you are currently managing this situation it will be very appreciate.

Thank you all

My solution is to not annotate in logseq right now…
Adding links works reasonable well though!

As long as the path issue is not solved, there seems to be no workaround (see parse JabRef file entries · Issue #44 · sawhney17/logseq-citation-manager · GitHub - if anyone knows how and where to implement it, please do :slight_smile: ).

One question would be, if jabref can be configured to write the plain path instead of the metadata enriched one. Is this possible? I could not find any hint that it is though…

JabRef has a menu item save selected as plain bibtex (e.g without any metadata like groups etc). I guess it would be possible to add special handling for the file field.

I would suggest you open an issue at GitHub - JabRef/jabref: Graphical Java application for managing BibTeX and biblatex (.bib) databases