Given three binary strings, find the maximum possible length of a contiguous block of 1's formed by shifting and overlapping the strings.
This may be interpreted as finding the maximum window size $k$ such that for each column $(a, b, c)$ across the three strings, it holds that $a \lor b \lor c$ is true. If we instead require $a = b = c$, then we recover the original longest common substring (LCS) problem, for which there is a linear time algorithm using suffix arrays. Can a similar linear time result be achieved here?