perf: optimize get_element from O(n) to O(1) lookup

[panbanda]

Summary

• Fix critical performance bug where get_element was doing O(n) linear search instead of O(1) indexed lookup
• The elements_by_id index was being populated but not used effectively

Problem

The get_element method retrieved an index from elements_by_id but then discarded it and did a linear search through the element vectors:

pub fn get_element(&self, path: &str) -> Option<&dyn Element> {
    let _idx = self.elements_by_id.get(path)?;  // Index retrieved but NEVER USED!
    
    // Then does O(n) linear search...
    let parts: Vec<&str> = path.split('.').collect();
    match parts.len() {
        1 => {
            if let Some(sys) = self.systems.iter().find(|s| s.get_full_path() == path) {
                return Some(sys as &dyn Element);
            }
            self.persons.iter().find(|p| p.get_full_path() == path)
                .map(|p| p as &dyn Element)
        }
        // ...more linear searches
    }
}

Solution

Changed elements_by_id to store (ElementType, usize) instead of just usize. Now get_element uses true O(1) lookup:

pub fn get_element(&self, path: &str) -> Option<&dyn Element> {
    let (element_type, idx) = self.elements_by_id.get(path)?;
    
    match element_type {
        ElementType::Person => self.persons.get(*idx).map(|p| p as &dyn Element),
        ElementType::System => self.systems.get(*idx).map(|s| s as &dyn Element),
        ElementType::Container => self.containers.get(*idx).map(|c| c as &dyn Element),
        ElementType::Component => self.components.get(*idx).map(|c| c as &dyn Element),
    }
}

Performance Impact

For a model with N elements:

• Before: O(N) per lookup (linear search through vectors)
• After: O(1) per lookup (hash map + direct array index)

This affects:

• get_element - directly optimized
• get_elements_by_type - calls get_element internally
• get_elements_by_tag - calls get_element internally
• get_children - calls get_element internally

Other Changes

• Uses enumerate() instead of manual index tracking in build_indexes (more idiomatic)
• Includes conditional asset test fix (same as PR feat: implement build command with real parser and exporter #8) since this branch is based on main

Test plan

• All 214 tests pass
• cargo clippy passes with no warnings
• cargo fmt --check passes
• CI pipeline green

Generated with Claude Code

[panbanda]

@panbanda Ready for review. This fixes a significant performance bug where get_element was O(n) instead of O(1).