Schema Parser for AutoEDA Agent

[madhurlak0810]

Implement a module to parse a schema file (JSON or YAML) for AutoEDA, extracting:

• Column names
• Declared data types
• Metadata: primary keys, time indices, target variables

Input:

• Schema file (JSON or YAML) describing dataset structure and types

Output:

• Python dictionary with:
  • columns: list of column names
  • data_types: dict mapping columns to types
  • metadata: dict with keys for primary keys, time indices, target variables

Requirements:

• Validate presence of required fields and handle missing/ambiguous cases with clear error messages.
• Keep implementation modular for integration with agentic loop and notebook generation using LangGraph

@gpsaggese @tkpratardan