How the raw wfn files are organized

[MegaGatlingPea]

I have a question regarding the organization of raw wfn archives. While exploring the repository, I noticed that unlike other datasets (such as the Hamiltonian database) which are typically organized with splits like test_full, I couldn't find a similar structure for the wfn archives.

Current observation:

Other datasets (e.g., Hamiltonian database) have clear splits (test_full, etc.)
The wfn archives seem to have a different organization structure
Question:
Is the only way to access specific wfn archives by:

First locating the corresponding entries in the CSV files
Then downloading the individual archives based on the information found in the CSV?
If this is the intended workflow, could you confirm? If there's an alternative approach or if I'm missing something in the documentation, please let me know.

Thank you for your help!

[KuzmaKhrabrov]

Hello!
Thank you for the issue.

First locating the corresponding entries in the CSV files Then downloading the individual archives based on the information found in the CSV? If this is the intended workflow, could you confirm? If there's an alternative approach or if I'm missing something in the documentation, please let me know.

Unfortunately, at this this point this is the only way to get a wfn for an entry in the CSV / databases.
We plan to reorganize the wfns to make it easier for reconstructing small datasplits.

If you need some particular help, please, let me know