Refactor: Refactor Scheduler to Support Dynamic Workflow Scheduling and Pipeline Pooling

[weijinglin]

This PR refactors the Scheduler class to introduce a more flexible and extensible workflow scheduling mechanism. The main changes include:

• Introduced a pipeline pool using a dictionary to manage different workflow types (e.g., build_vector_index, graph_extract), each with its own GPipelineManager, flow, prepare, and post-processing functions.
• Added a schedule_flow method to dynamically select and execute workflows based on the flow name, supporting pipeline reuse and resource management.
• Refactored the build_vector_index and graph_extract flows to separate preparation, execution, and post-processing logic, improving modularity and maintainability.
• Updated related utility functions (graph_index_utils.py, vector_index_utils.py) to use the new schedule_flow interface.
• Improved error handling and logging for schema parsing and pipeline execution.

These changes lay the foundation for supporting more complex and agentic workflows in the future, while also improving the efficiency and scalability of the current pipeline execution framework.