Generic and container type mapping

[Shishir99-code]

Problem

list[Model], dict[str, Model], tuples and sets don't map to the right spec array/object schemas.

Acceptance criteria

• list[Model]/set[Model] → spec array with item schema; dict[str, Model] → object/additionalProperties
• Tuple types map to a reasonable array schema; nested containers resolve recursively
• Tests covering list-of-model, dict-of-model, and nested containers

Guardrails (must hold — the AI reviewer enforces these)

• Python 3.11+, from __future__ import annotations in every module, strict mypy, ruff clean (line length 100)
• Parser stays AST-only — never import or exec the analyzed project
• No new runtime dependency without strong justification in the PR (CI dep-count matters)
• .docguard.yaml config stays backwards-compatible
• No LLM in the check/report path; output stays deterministic (sorted, no clocks/network)
• Internal data = stdlib @dataclass with to_dict(); exit codes 0/1/2
• Add pytest coverage including a false-positive / missed-endpoint case where relevant

Likely files

• src/docguard/parsers/fastapi_parser.py
• src/docguard/core/comparator.py
• tests/test_type_fidelity.py

Out of scope

Numpy/pandas container types.

---

Part of epic: Epic: Deep type fidelity — eliminate the last false positives. Picked up autonomously by the DocGuard pipeline when moved to Ready.