Enhance error handling and optimize random algorithm for course allocation

[vikrantwiz02]

This pull request significantly optimizes the course allocation logic in the academic information system by reducing the number of database queries and improving performance. The main changes involve batching queries and using bulk operations to replace inefficient per-object lookups and inserts, which should lead to faster allocations and less database load.

Performance and Query Optimization:

• In random_algo, replaced per-course and per-student queries with batched queries and precomputed dictionaries for course capacities, already allotted seats, student-curriculum mappings, and semester lookups. This reduces the number of queries from many per iteration to just a few per allocation run.
• In random_algo, switched from creating FinalRegistration objects one at a time to collecting them in a list and using bulk_create at the end, greatly improving insert performance.
• In allocate, replaced multiple queries per student with a single batched query per slot, and used dictionaries to map students to curricula and semesters. Final registrations are now inserted in bulk for each slot.

Code Simplification and Consistency:

• Simplified the way student and course relationships are handled by using ID-based lookups and bulk operations, making the code easier to maintain and less error-prone. [1] [2] [3]

Error Handling Improvement:

• Improved exception logging in start_allocation_api by logging the exception with a traceback and returning a more informative error message including the exception type.