Open phyy-nx opened 1 year ago
@prjemian can you review? I don't know if this helps address all the concerns in #1124.
@gerrit-guenther: Can you comment?
I'm not clear how the one line of documentation added by commit satisfies the questions raised in #1124 or https://github.com/nexusformat/definitions/issues/886#issuecomment-804191753.
Ok, Ive re-read the original issue, and here are the specific questions, broken out and restated:
I've added new documentation to NXsubentry which I believe answers them.
3 and 4 are hopefully answered by 2. In addition I've noted that NXsubentries aren't designed to be contained within each other.
Ok, these are at least my interpretations of what is going on. Feedback appreciated :)
From code camp:
I have included this note in the PR: "An NXsubentry should not contain another NXsubentry". Is that correct? @yayahjb says we shouldn't need the restriction. @PeterC-DLS says it doesn't make sense for subentries to be nested. This would open the possibility of an application specifying a subentry. (Aaron: this might be useful for workflows?)
I based the assertion on my reading of the docs and @prjemian's comment showing the two use cases (with and without subentries) as meaning those are the only supported, but this might not be what was intended.
@mkoennecke what is the behavior of cnxvalidate if it finds a nested subentry?
Thoughts?
Closes #1124