1: \begin{abstract}

2: We introduce optimal quadrature rules for spline spaces that are frequently used

3: in Galerkin discretizations to build mass and stiffness matrices.

4: Using the homotopy continuation concept \cite{Homotopy-2015} that transforms

5: optimal quadrature rules from source spaces to target spaces, we derive optimal rules for splines

6: generated above finite domains.

7: Starting with the classical Gaussian quadrature for polynomials, which is an optimal rule for a discontinuous

8: odd-degree space, we derive rules for target spaces of higher continuity.

9: We further show how the homotopy methodology handles cases where the source and target rules

10: require different numbers of optimal quadrature points.

11: We demonstrate it by deriving optimal

12: rules for various odd-degree spline spaces, particularly with non-uniform

13: knot sequences and non-uniform multiplicities.

14: We also discuss convergence of our rules to their asymptotic counterparts, that is,

15: the analogues of the midpoint rule of Hughes et al. \cite{Hughes-2010}, that are exact

16: and optimal for infinite domains.

17: For spaces of low continuities, we numerically show that the derived rules quickly converge to their asymptotic

18: counterparts as the weights and nodes of a few boundary elements differ from the asymptotic values.

19: \end{abstract}