Add file de-duplication

[fcasson]

Some workflows upload files as input which were outputs of other workflows. This enables the parent / child detection feature to work

However, storage of identical objects is duplicated. This could be improved (a bit like git forks on github and gitlab but a lot simpler since there is only a single server).

Some attempt was made, but not merged I think.

https://git.iter.org/projects/IMEX/repos/simdb/commits/19f391fdfeaea65356b1029f5b576a90925a5e06#[src/simdb/remote/api.py]

Similar approach could be done for IMAS data but might need to based on IMAS checksums, not file checksums.

[olivhoenen]

While interesting, the attempt above relies solely on the application, making it potentially less robust. Standard deduplication at the FS level exist, but may require to use specific filesystems like XFS or btrfs.

We are currently working on enabling background tasks on the server (see #72 and #83), this could have the potential to enable file deduplication on the server without increasing the complexity and robustness of the application.