I’m a Physics researcher and I’ve created an Obsidian a custom plugin to reproduce a pure Obsidian Bibliography Manager. The plugin works that I have a command palette command (with possibly a hot-key), it opens a windows that asks for a DOI. After pressing submit, it creates a note with a basic template and the metadata containing all the most important fields of the reference like authors list, year, journal, bibtex code, and others.
Now, I’m interested in Anytype, is there a way to have such functionality also here?
I don’t fully understand what you want.
Are the metadata fields empty, after creating that note?
– If yes, then it’s easy. You can use Templates in Anytype.
It may be a good idea to create an own Type for it. but you could also simply make a Template from the Type “Page”.
I wouldn’t recommend to use the Type “Note” for that btw.
In fact, I completely avoid the usage of “Note”.
You might want to create feature requests on broader supports for general citation and apply metadata directly to the corresponding relations, including but not limited to BibTex and RIS. Or a function expansion of web-clipper, like Zotero | Connectors.
In the Obsidian case, it creates a new note with all the metadata corresponding to the reference. So the medadata are not empty. In the case of Anytype, it should create some kind of object with already all the fields filled.
Curious…did you mean to say “if an API is announced soon”, or did you mean it as in “API will be announced soon” - meaning you know it will be actually rolled out pretty soon?
At the last Town Hall, they presented the RayCast integration as an introduction to the API. It’s probably just the beginning (from memory, I’ll let you watch the video again), but it’s well under way.
Hello, as far as I understand, the APIs are now public. Is that correct? If yes, could this Bibliography manager be implemented? What I would require is just the DOI of the paper or book, and then it will create the object with all the properties already filled.
You just give the doi to a command line terminal, and it will create the authors, the journal, and the article objects. It will also generate a bibtex string code to copy and paste in latex projects.
It uses CrossRef API to extract the metadata. It should support all the DOI, as this is the standard.
I should probably still need some stability check to avoid duplicates of authors, journals or articles. But I have added some tests to keep track of it.
