The new posterior is the average of the old posterior and the new likelihood.

The posterior becomes the prior for the new data, such that $\pi(\theta|D, D_{new}) \propto L(D_{new}|\theta) \pi(\theta|D)$.

The posterior is updated using the ratio of the marginal likelihoods.

The posterior remains unchanged as the prior is already fixed.