Preserve global Z offsets and support config groups in get_ordered_datasets

[msayr]

Motivation

• Ensure dataset ordering handles nested config groups and avoids double-counting Z offsets when upstream prep already produced global Z coordinates.
• Prevent incorrect stacking where a later group's offsets were blindly shifted by a previous group's length, causing duplicated Z displacement.

Description

• Accept nested config_paths groups and collect datasets per group before computing offsets so groups can be treated atomically.
• Compute group_offsets, group_z_start, and group_z_end arrays and detect whether a group already carries global Z offsets; preserve them when they overlap the current global cursor, otherwise shift the group's Z offsets by previous_offset.
• Append dataset stores in group order, skip empty groups, and compute previous_offset as the maximal adjusted Z end across the group to correctly position subsequent groups.
• Convert offsets to mutable lists and add informational logging about preserved global offsets.

Testing

• Ran the project unit test suite with pytest and the tests completed successfully.

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 14538825c4

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Stop treating long local groups as global

When a later config group uses local Z offsets starting at 0 but is longer than the previous group, group_z_end.max() > previous_offset is true, so the group is preserved instead of shifted after the prior stack. For example, two XY configs with local ranges [0, 100) and [0, 200) leave the second group at Z 0 because 200 > 100, causing it to sort before/overlap the first group and corrupting the Z alignment plan that previously concatenated such configs.

Useful? React with 👍 / 👎.