Lebbeous, thank you very much for testing. I think we can find a way to work around such cases without too much trouble.
As for whether this type of data is wrong, I'd feel pretty confident arguing that it is. If a holding doesn't conform to the caption in a truly meaningful and sensible way, it really needs a new caption. Take as many as you need; they're free! ;)
We already know that isn't always going to happen, so I think we can get away with marking the missing subfields as unknown/missing (the asterisk in MFHD), then throwing away the extras and seeing what happens. I'll see what I can do. Thanks again.
Lebbeous, thank you very much for testing. I think we can find a way to work around such cases without too much trouble.
As for whether this type of data is wrong, I'd feel pretty confident arguing that it is. If a holding doesn't conform to the caption in a truly meaningful and sensible way, it really needs a new caption. Take as many as you need; they're free! ;)
We already know that isn't always going to happen, so I think we can get away with marking the missing subfields as unknown/missing (the asterisk in MFHD), then throwing away the extras and seeing what happens. I'll see what I can do. Thanks again.