Memolanes archive export v2

[CaviarChen]

This feature will read all snapshots stored in your Fog Machine, then convert and optimize them one by one. After that, they will be merged into an archive file that can be read by MemoLanes (mldx).
Specifically:
1. All snapshots will be processed in the order of snapshot time and converted into journeys in MemoLanes.
2. Each journey's date is based on the snapshot time and your current time zone.
3. We subtract the previous snapshot from each snapshot, so each journey will only contain the diff.
4. Area that exists in a snapshot but not in the final snapshot will be removed. If you've used the eraser, ensure the final snapshot doesn't contain areas you don't want.
5. Snapshots that are too small will be ignored.

with the current optimization, it rought takes ~10s (instead of 60s) with my full data in debug (2s in release). We might need more memory on the server tho.