[OT] Efficient file structure for very large lookup tables?

[Dylan Knutson]

It's not a datastructure persay, but Google's LevelDB is very 
good, quite fast arbitrary key-value storage system (I use it in 
my project relating a 1 word key to a 200 word value) supporting 
batch transactional operations.

Think of it as a key-value storage version of SQLite.

https://code.google.com/p/leveldb/

And then the D bindings:
https://github.com/bheads/d-leveldb

And a D wrapper providing a nicer interface to LevelDB:
https://github.com/bheads/leveldb