The parsed OIDs are usually prefixes, not full hashes, so rename the
fields appropriately. Also note that values that are too long to be
valid OIDs are still accepted by the library.
I believe the structure is correct, but there are a lot of details to
fill in - translating the C string parsing logic into Go is not always
straightforward.