Use Related metadata for autocomplete API discovery

[dheeraj12347]

Summary

Fixes apache/cloudstack#10442

CloudMonkey already parses and stores the Related metadata exposed by CloudStack APIs in APIArg.Related, but findAutocompleteAPI() did not use this information during autocomplete API discovery.

This change updates autocomplete API selection to prefer authoritative Related metadata before falling back to the existing heuristic-based lookup logic. This improves reliability for entity-reference parameter completion while preserving backward compatibility with the current behavior.

Changes

• Use APIArg.Related when selecting the autocomplete lookup API.
• Prefer Related APIs whose noun matches the expected resource noun.
• Fall back to any valid Related list* API if no noun match is found.
• Preserve all existing heuristic-based lookup logic as a fallback.
• Add unit test coverage for Related-based autocomplete API discovery.

Validation

Verified runtime behavior for common entity-reference parameters:

• zoneid → listZones
• domainid → listDomains
• networkid → listNetworks

Added unit tests covering:

• Related noun match selection.
• Related API fallback selection.
• Empty Related metadata fallback to existing heuristics.
• Non-list Related metadata fallback to existing heuristics.
• Map-type argument handling.

Build and validation:

go fmt ./...
go test ./...
go build -o bin/cmk .

Test results:

ok github.com/apache/cloudstack-cloudmonkey/cli

[github-actions]

✅ Build complete for PR #214.

📦 Binary artifacts are available in the workflow run (expires on June 19, 2026).

Note: Download artifacts by clicking on the workflow run link above, then scroll to the "Artifacts" section.
Artifacts from PR builds are for testing only and may contain unreviewed, malicious code.

[dheeraj12347]

Hi @Pearl1594 and @soreana,

I've updated PR #214 based on my planned Phase 1 deliverable. The implementation now uses APIArg.Related metadata for autocomplete API discovery, with fallback to the existing heuristic-based logic to preserve current behavior.

Since opening the PR, I've also added unit tests covering:

• Related noun match selection
• Related API fallback selection
• Empty Related metadata fallback
• Non-list Related metadata fallback
• Map-type argument handling

The PR has been marked ready for review. Whenever you have time, I would appreciate your feedback on the approach and implementation.

Thank you!