to benchmark the pipeline, I tried sending 100 incrby commands back to back
Code
Pure redis:
```py
t1 = time.time()
iter_count = 100
for i in range(iter_count):
redis.incrby("marine", 10)
val = redis.get("marine")
assert val == str(iter_count*10)
t2 = time.time()
print(t2-t1)
```
Pipeline:
```py
import time
t1 = time.time()
iter_count = 100
with redis.multi() as pipeline:
for i in range(iter_count):
pipeline.incrby("marine", 10)
val = redis.get("marine")
assert val == str(iter_count*10)
t2 = time.time()
print(t2-t1)
```
Adding pipeline and transaction feature as described in #45.
Majority of the update is complete. Next steps:
with redis.pipeline() as pipe: ...
)Auto pipelining for asyncio