Fix O(m^2) assembly in HypreSolver::factorize

[Chenyang-Joe]

Problem

HypreSolver::factorize builds the HYPRE matrix with a nested loop: for each nonzero
in a column it rebuilds and re-sets the entire column, so every column is handed to
HYPRE_IJMatrixSetValues nnz(col) times. Assembly is therefore O(m^2) in the number
of nonzeros and dominates solve setup on large matrices.

Fix

Build each column's index/value arrays once and set it in a single HYPRE_IJMatrixSetValues
call -> O(nnz) total. The assembled matrix (and the
solution) is unchanged.

Testing

• Builds clean.
• Full ctest suite passes 27/27, including hypre and hypre_initial_guess.

@maxpaik16 PTAL, thanks!

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.73%. Comparing base (a7727e3) to head (8582129).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #111      +/-   ##
==========================================
- Coverage   80.76%   80.73%   -0.03%     
==========================================
  Files          51       51              
  Lines        2121     2118       -3     
  Branches      281      280       -1     
==========================================
- Hits         1713     1710       -3     
  Misses        408      408              

Flag	Coverage Δ
polysolve	80.73% <100.00%> (-0.03%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.