Use unordered_set to fix quadratic running time in FASTA parsing

[jnalanko]

In FASTA parsing, sequence names were stored in a std::vector, and every new sequence name was searched versus every previous sequence name with std::find. This took O(n^2) time total, where n is the number of sequences, making the parsing extremely slow in my use case, where the index is a set of reads.

In this patch, I add an std::unordered_set of sequence names for O(1) sequence name lookup, reducing the running time to O(n).

[lrenders]

Thank you for the patch! I hadn't considered the use case where the index is a set of reads, so your patch is a welcome improvement.