Quick and dirty test to see how tries compare to hash tables in this specific case

redis_bench_tries.py

from redis import Redis
import time

r = Redis()

keys = []
for a in xrange(ord('a'), ord('z')):
    for b in xrange(ord('a'), ord('z')):
        for c in xrange(ord('a'), ord('z')):
            keys.append('{0}{1}{2}'.format(a, b, c))

keyvals = []
for k in keys:
    keyvals.append(k)
    keyvals.append('xxx')

r.execute_command('hmset', 'hash', *keyvals)
r.execute_command('tmset', 'trie', *keyvals)

start = time.time()
[r.execute_command('hmget','hash', *keys) for i in range(0, 1000)]
end = time.time()
print '1000 HGET performed in {0} seconds'.format(end - start)

start = time.time()
[r.execute_command('tmget', 'trie', *keys) for i in range(0, 1000)]
end = time.time()
print '1000 TGET performed in {0} seconds'.format(end - start)