You missed the last paragraph, I think: > For a production implementation it may be worth packing i into the flags field since it only needs a few bits, even w

You missed the last paragraph, I think: > For a production implementation it may be worth packing i into the flags field since it only needs a few bits, even with larger alphabets. Also, I bet, as in Deutsch-Schorr-Waite, the p field could be eliminated and instead one of the child pointers is temporarily reversed. With these changes, this technique would fit into the original struct trie without changes, eliminating the extra memory usage.