Upgrade StorageWhich export format via API returns the most complete dataset?
Hi
This is not about backup of Zotero or its groups! I know there is a different approach suggested. This is about the data in Zotero Groups.
We want to download the data in the Zotero Groups on a regular basis to have them available in a worst case scenario (Zotero online version as well as sync gone).
Therefore we are using the API and at the moment the css format (https://www.zotero.org/support/dev/web_api/v3/basics#item_export_formats)
I would like to confirm if this is the most complete format we can use for this, or if any of the other formats contains more complete datasets?
Thanks
Rainer
This is not about backup of Zotero or its groups! I know there is a different approach suggested. This is about the data in Zotero Groups.
We want to download the data in the Zotero Groups on a regular basis to have them available in a worst case scenario (Zotero online version as well as sync gone).
Therefore we are using the API and at the moment the css format (https://www.zotero.org/support/dev/web_api/v3/basics#item_export_formats)
I would like to confirm if this is the most complete format we can use for this, or if any of the other formats contains more complete datasets?
Thanks
Rainer
The only truly complete data format would be the format that Zotero itself uses for syncing — the API's default JSON format — but you can't actually import that directly into the Zotero client (at least without some scripting). You'd also need to download all attachment files via the API. But this is, of course, what Zotero itself does, which is why you should just make a backup of the Zotero data directory on a computer where Zotero is set up to sync and download files automatically.
So: I want to use the export API, and have as complete data as possible.
I also mentioned that after a re-import of the data the zoteroID of the work would be different - but we can live with that.
I also do not care so much about the attached files, as they are (in all but one group) ephemeral and can always be re-downloaded. And they do not have notes.
I am looking for a format which has
1. as much properties of the data as possible
2. is readable (css, bibtex, json, etc) which are most of the formats as I would guess
3. If 1 and 2 is the case, we can read it in in a script and convert it to a format which I can import into Zotero again or a different reference manager if really needed. The groups can be re-created, although with different ids for all things.
Which format of the ones listed in the link supplied above (https://www.zotero.org/support/dev/web_api/v3/basics#item_export_formats) are the most suitable for that. Bonus: if they can directly be imported, that is a plus.
If you must use export for whatever reason, Zotero RDF is most complete, but it's very much not loss-less (it's also comparatively heavy and harder to parse)
1. I get native Zotero API JSON, when I do not specify an export format in the API call - correct? The call "https://api.zotero.org/groups/2352922/items?limit=100&start=0" will result in Zotero API JSON.
2. In the data field, the fields required but e=the itemType are in contrast to the css, - where all are present
3. So the two options are:
- API JSON: most complete, takes time to download, can not be imported to Zotero directly (but as Zotero is using it as well, there should be an API approach to upload it again?
- Zotero RDF: the most complete and can be imported in Zotero again, but difficult to parse for anything else.
Correct?
2. In the data field of API JSON, the fields required by the respective `itemType` are in the data field, in contrast to the `csv`, - where all are present