Bucket import issues

[khalidrafi6]

• I am on the latest ActivityWatch version.

• I have searched the issues of this repo and believe that this is not a duplicate.

• OS name and version: Arch Linux

• ActivityWatch version: 0.13.2 (rust)

Describe the bug

I am trying to import json buckets from Android to get all my data in one place. But there are two issues:

1. Buckets with the same name fail to import. So, I have to delete all the buckets one by one first. It would be nice to have the buckets updated with new data instead.

2. If I delete buckets and then import, the id column in the buckets table of sqlite db gets changed. So, I have to keep updating my sqlite query function to inspect my data.

To Reproduce

Try importing duplicate buckets.

It would be helpful for dealing with Android data if these small bugs were fixed.

[github-actions]

Hi there!
As you're new to this repo, please make sure you've used an appropriate issue template and searched for duplicates (it helps us focus on actual development!). We'd also like to suggest that you read our contribution guidelines and our code of conduct.
Thanks a bunch for opening your first issue! 🙏

[TimeToBuildBob]

Submitted a fix for the first issue (duplicate bucket import failing) in ActivityWatch/aw-server-rust#575.

The PR changes the import behaviour: when a bucket already exists, events are now merged into it instead of returning a 500 error. This lets you re-import Android exports without manually deleting buckets first.

The second issue (IDs changing on re-import) is a SQLite auto-increment behaviour — since event IDs are stripped during export/import the SQLite rowids are always re-assigned. That's separate from this fix and would need a different approach (e.g. preserving stable bucket-scoped IDs).

[TimeToBuildBob]

This is being addressed by ActivityWatch/aw-server-rust#575 which changes the import behavior to merge events into existing buckets instead of failing on duplicate bucket names. This should fix issue 1 directly.

For issue 2 (the ID column changing), that's a consequence of the delete-then-reimport workflow. With the merge approach in #575, you won't need to delete buckets first, so the IDs should remain stable.

[TimeToBuildBob]

Quick correction / status update: the linked fix in ActivityWatch/aw-server-rust#575 improves the duplicate-bucket failure mode, but it does not fully solve repeated re-imports for exports where events have no explicit id field. In that common case, re-importing the same file can still create duplicate rows because SQLite assigns fresh IDs.

So the current status is:

• ✅ duplicate bucket name no longer has to hard-fail
• ⚠️ repeated import of the same ID-less events is still not safely idempotent
• ⚠️ stable event IDs across delete/reimport still need a separate solution

I posted the scope correction in the PR thread as well so future review has the accurate framing.