Add semantic operators module

[D3f0]

This PR adds the semantic_operators.py module containing four key semantic operations:

Operations

• sem_map: Semantically transforms records from source to target schema using natural language instructions
• sem_filter: Filters records based on natural language predicates using LLM-based logical classification
• sem_agg: Aggregates across all records for summarization tasks
• normalize_input_data: Helper function for input normalization

The module provides an agentic analogue of LOTUS-style semantic operations and supports both Agentics Graph (AG) and pandas DataFrame inputs/outputs.

Testing

Includes comprehensive test suite with 20 test cases covering:

• Input normalization with DataFrame and AG types
• sem_map operation with BaseModel/string targets and various merge modes
• sem_filter operation with predicate evaluation and DataFrame support
• sem_agg operation with aggregation logic and transduction type setting
• Integration tests for end-to-end workflows
• Edge cases and error handling

All tests passing with 100% success rate.