Official Coding Period: Week 5 (June 30 - July 6)
This week, I fixed the input parameters in the SQL file to return a set of seq and node for the ordering function. I focused on solving all the compilation and style errors this time.
One major change I implemented was replace all custom tuple types II_r_t with int64_t arrays. I removed the minute changes in the documentation references for now since we are not working with them for the time being.
Started with the docqueries and added the example and test configuration to the documentation build. Then, I further simplified the result tuple logic and removed unused helper and type includes.
Sequence:
User->>SQL_Function: CALL pgr_sloanOrdering(edges_sql)
SQL_Function->>C_Layer: _pgr_sloanordering(PG_FUNCTION_ARGS)
C_Layer->>Ordering_Process: pgr_process_ordering(edges_sql, which, int64_t **, size_t *)
Ordering_Process->>Ordering_Driver: do_ordering(edges_sql, which, int64_t **, size_t *, ...)
Ordering_Driver-->>Ordering_Process: result_tuples (int64_t array)
Ordering_Process-->>C_Layer: result_tuples, result_count
C_Layer-->>SQL_Function: Construct tuples from int64_t array
SQL_Function-->>User: Return result set I also changed the type of hint from nullptr to "" in the C file. Other bug fixes include correcting function signature comments to accurately reflect the current parameter list.