Expand Salesforce metadata and frontend indexing

[krzysztofwojt]

Summary

• index every Salesforce *-meta.xml with stable fallback symbols
• add detailed symbols and selected references for Flow, object metadata, Custom Labels, LWC, Aura, and Visualforce
• add Salesforce-aware LWC imports, templates, targets, and properties plus dedicated salesforce_markup
• cover the real default worker path and the built bundle with a Salesforce DX fixture

Why

This builds on the Apex and metadata indexing merged in #9. The original adapter covered core Apex and selected XML metadata, but frontend bundles and much of Salesforce source-format metadata remained unknown or only generically searchable.

Validation

• npm run typecheck
• npm run lint
• npm run test:guards (13/13)
• targeted Salesforce, parser, worker, and embeddings tests (55/55)
• built dist/token-goat.mjs Salesforce smoke tests (2/2)
• synthetic scale test: 2,700 files, 5,500 symbols, 600 refs, 0 duplicate symbols/refs, 0 unknown *-meta.xml

Notes

• no database schema migration
• long-tail metadata gets a stable top-level symbol; deeper extraction is limited to high-value types

[krzysztofwojt]

Hi @Zelys-DFKH , please give me some time to manually review my changes here, I'll change the status to ready to review once I'm sure it's really ready ;)

[krzysztofwojt]

Okay, ready.