`[cco-core]` Implement OAI-ORE Data Models via Pydantic

[clnsmth]

Description
We need to establish the foundational data structures for the cco-core library. According to ADR-001, we are using Pydantic to enforce the OAI-ORE (Object Reuse and Exchange) specification. These models will serve as the contract between the ingestion logic, the API, and the final JSON-LD output.

Context
The CCO Generator acts as a middleware. It needs internal Python objects that represent the complex citation graph before we serialize them to JSON-LD for export or storage.

Requirements

• Create a Pydantic model structure that reflects the OAI-ORE standard.
• The models must support serialization to JSON-LD.
• The models must support valid citation attributes (e.g., DOI, Title, Author, etc.).

Technical Scope (Suggested)
Based on the architecture document, we need three primary classes:

1. AggregatedResource: Represents the individual Linked Digital Object (LDO) (e.g., a specific dataset or software DOI).
2. Aggregation: The central collection object that groups the resources.
3. ResourceMap: The top-level wrapper document that describes the Aggregation.

Acceptance Criteria

• Pydantic models are defined for the three classes listed above.
• A test script exists that instantiates these objects and successfully serializes them into valid JSON-LD.
• Validation rules are in place (e.g., ensuring an Aggregation has at least one AggregatedResource).

References

• ADR-001 Section 3: Data Model
• RDA Recommendations: https://doi.org/10.15497/RDA00130
• OAI-ORE Primer: openarchives.org/ore/1.0/primer