There are ways and means of figuring out where cable breaks are - lots of cables are installed in situations where you can't inspect their length and it's pretty important to know where to look.

With multiple breaks it would probably still look like a single segment was missing, but at least you know the extent of the damage. Most of these cables aren't contiguous anyway, you have repeaters which you can also use to figure out break placements.

Gerv, check Wikipedia for "optical time domain reflectometer"; I think this will answer your question. If a cable is cut in several places then presumably one could use ODTR techniques from either end to identify the distance to the nearest cut from that end, fix the cuts, and then use ODTR to identify any additional cuts further on, until all cuts were identified and fixed. As AlexH notes, in practice there are repeaters on the cable, and presumably that's where you'd generate the signals used for ODTR.

s/ODTR/OTDR/

Frank: that's dead clever. So presumably, my supposed nefarious terrorists would need to attach a device to the cut cable end which caused the ODTR to return incorrect results. Clearly, they couldn't make the cut appear nearer to the shore than it was, but they could make it appear further away, thereby getting the ship to look in the wrong place.

As for the repeaters, I guess it depends whether they are active or passive. Early cables, I seem to remember reading, had passive repeaters. If new ones have active ones capable of being told to generate pulses independent of incoming light, then yes, finding a fault would be much easier.

Frank, i agree