refactor: Nonempty

[Cobord]

Replacements and SiblingSubgraph have nonempty vecs upon construction. Attempting to create them with empty vecs gives errors (InvalidSubgraph::EmptySubgraph) so following "illegal state unrepresentable".

[codecov]

Codecov Report

Attention: Patch coverage is 85.00000% with 3 lines in your changes are missing coverage. Please review.

Project coverage is 85.56%. Comparing base (a2e4d05) to head (e689e91).

❗ Current head e689e91 differs from pull request most recent head 5d408d6. Consider uploading reports for the commit 5d408d6 to get more accurate results

Files	Patch %	Lines
quantinuum-hugr/src/hugr/views/sibling_subgraph.rs	78.57%	0 Missing and 3 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #902      +/-   ##
==========================================
- Coverage   85.63%   85.56%   -0.08%     
==========================================
  Files          78       78              
  Lines       14444    14415      -29     
  Branches    14444    14415      -29     
==========================================
- Hits        12369    12334      -35     
- Misses       1436     1443       +7     
+ Partials      639      638       -1     

Flag	Coverage Δ
rust	85.56% <85.00%> (-0.08%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[aborgna-q]

I'd prefer to leave the NonEmpty out of the public interface here,

Suggested change

	pub fn nodes(&self) -> &NonEmpty<Node> {
	&self.nodes
	pub fn nodes<'a>(&'a self) -> impl Iterator<Item = &'a Node> + &'a {
	self.nodes.iter()

that breaks some code, you'll need some collect_vecs around (which implies allocating new arrays :/).

So something like

let portgraph: NodeFiltered<_, NodeFilter<Vec<Node>>, Vec<Node>> =
    NodeFiltered::new_node_filtered(
        other.portgraph(),
        |node, ctx| ctx.contains(&node.into()),
        subgraph.nodes().cloned().collect_vec(),
    );

in HugrMut::insert_subgraph.

[aborgna-q]

This requires every dependent to add a direct dependency on NonEmpty. Although it's nice to statically check this internally, i'd prefer the external interface be left more generic.

Adding a Replacement::try_new(removel: Vec<Node>, ...) could help with this.

[Cobord]

fair enough

[aborgna-q]

Hey, we'll close this for now as it requires exposing transitive dependencies in the public API.
This change would improve the static compiler checks, but it's not worth the added complexity.

Thanks nonetheless for suggesting the change :)