Open lukavdplas opened 1 year ago
Suggestion: make the download format a CSV with corpus name + document ID
This would mean that you are exporting the tag's linked documents rather than the tag itself (the CSV would probably not include name/description/etc), but I think that would work fine. Given the constraint of unique names, it could save some headaches when importing a tag from someone else. Also, the result would be more generic than a JSON with all tag data.
Note that uploading saved tag data would just be uploading a CSV file, so this requires some validation. Checking corpus names is easy enough, but we should probably verify that the document IDs are also real documents.
Possible follow-up to #997 : allow users to share tags.
Could be done by adding an export/import function. So a user could download a tag (e.g. as a JSON file), e-mail it to their friends and then those users can import the tags in their account.
It would also be possible to let users share tags directly through the interface, either by sharing them with specific users or by adding them to a sort of 'public' tag library. However, I'm not keen on introducing 'social' features to I-analyzer.