Dear all,
I am just playing around with Tropy and its export functionality and while I want to offer a big shout out for having built such a beautiful piece of software, I would also lobby for options to export absolutely every bit of information contained in the Tropy database. Most of use have experienced lock-ins in the past that left us with no means to access months and years worth of work to the extent that I would say they are probably the single most frustrating aspect of academic life. Blurb aside, I cannot see how people commit to Tropy’s fantastic note-taking features (the manual suggest to use them for time-consuming transcription of sources) without an easy means to export those notes in bulk and in structured format.
Thank you very much,
Till
Thanks for your feedback!
We’re currently working on an export plugin system that will likely make it into the next release; this will include the missing data (specifically, notes in a plain text and an HTML rendition) in the JSON-LD export, too.
That said, I would like to emphasise that you need not worry about lock-in of your data with Tropy. Tropy itself is free and open source and the project files are SQLite databases. We will continue to improve the JSON-LD export and add a number of plugins to export directly to other software such as Omeka, but it is also possible to access your data directly with SQLite. The database schema is not trivial but it’s definitely possible to access your data with SQL completely without the use of Tropy itself.
Thank you very much for your impressively fast reply.
I am happy about the good news and would suggest a quick comment in the manual as to the future inclusion of notes in the JSON-LD export.
Concerning the open nature of SQLite and therefore the general prevention of lock-ins: this seems to be a rather generic feature and one that is completely inaccessible to the majority of users (I have spent a fair amount of time to figure out the database schema of a discontinued reference manager and then to build some export routines. I would never want to do that again).
All the best,
Till
@inukshuk, I am curious as to what the status of the export plugin that you mentioned in this thread. I looked through the recent releases and I didn’t see mention of it. I am currently trying to spec out some deliverables for a few use cases in a grant extension for which I have signed on to do some of the development. The objective on this grant extension is to prototype an application that provides a graphical user interface for constructing a hierarchy of containers with metadata sourced from some domain specific metadata gazetteers, being able to attach image assets to those containers, and then serialize the entire metadata hierarchy with the associated resources into some interchange container, e.g, a Bagit tar ball, that could be shared or ingested into some digital asset management system.
It seems to me that Tropy would be an ideal platform to extend in order to create such functionality. I am curious whether any of that seems like it is in line with the scope of continued development of Trophy and could possibly be incorporated into the code base, or if it diverges too far from the scope to which you all are constraining the project and would be better as a separate entity forked from the Trophy code base.
Anyhow, I’d like to hear your thoughts on the matter and possible pick your brain on what the best route moving forward might be.
We’ve meanwhile added notes and other missing data to the JSON-LD output (the default export). The idea is that JSON-LD export covers all the information you add to an item in Tropy; we’ll be using it to make copies or transfer items to other projects, so, naturally, all the information will stay covered.
We’ve also rolled out a plugin system for additional export formats (we’re currently building a UI for managing plugin configurations that will go into the next release in a couple of weeks). Basically, an export plugin can register itself with Tropy; when you export items, it receives the respective JSON-LD data. At the moment there is an export plugin that uploads exported items to an Omeka server; we are also working on a plugin that takes the exported JSON-LD and saves it in an archive together with all the associated photos. It should be relatively easy to create a similar plugin that saves everything in a BagIt archive.
The plugin system is still evolving; we plan to make a richer API available to plugins as the need arises (e.g., to duplicate behavior already implemented in Tropy, like saving a selection from a photo) and also to allow plugins to hook into other aspects of Tropy (next on the list will be import plugins). My hope is that you’d be able to cover a range of use cases via the plugin system and would not need to fork the code base. In any case, I’m happy to answer any questions you might have here on the forums or on GitHub, just let me know!
1 Like
Hi! Jumping in here as a researcher from the Ajami Lab who has recently been exchanging with you on Twitter about export capabilities. Following up on your request that we post on the Forum, we’d also like to see some way to export notes in a useful way for laypeople.
We currently use Tropy to annotate manuscripts written primarily in Arabic but with marginal notes and sections etc. in African languages. We’ve written up some technical notes here that lay out our process more or less using Tropy. In short, we’d like to be able to export our notes and selection notes for which we have adopted a standard format of 1) typesetting 2) transliteration 3) interpretation and 4) translation.
Right now, we can export things as a JSON-LD file and we have to do a lot of work to get the output into a CSV or text format that can be easily manipulated into a nice looking text file or CSV for typesetting or further analysis. We explained the whole process as currently stands in this Twitter thread (see the technical notes for proper instructions). What would be nice for export plug-in, we think, would be something that allowed people to export all the annotations (selections and notes) and metadata from a single document (or whatever Tropy calls a collection of images) into a CSV. So basically you could right-click as now and spit out a CSV that would spit out one’s selections as rows with columns for your selection name, selection note, coordinates etc.
Can post an example of the CSV that we eventually get from the JSON-LD file if that’s helpful too. Not on my computer at this time, but when my officemate is in, I can grab it from him! Let me know.
Thanks! We’re currently working on a CSV and notes export plugin. I think this might fit very well into our CSV export use case. I’ll look to set up a preliminary version to export one row per note (regardless at which level) and some of the relevant metadata and post here as soon as there’s something to try out.
1 Like
Hello! Following on this topic: would exporting a Tropy database to Omeka-S keeping the annotation of specific parts of the image (made through the selection tool)? Thanks a lot for such an incredible piece of software!
Yes, the plugin also uploads selections to Omeka-S; though I believe at the moment the plugin will not support this for all image types supported by Tropy (JPEGs should be fine; PDF or TIFF will probably not work yet).
1 Like