Redis: pipeline.Exec returns redis: nil when some key does not exist

[ginuerzh]

For example, a pipeline with a HGET command:

pipeline := client.Pipeline()
pipeline.HGet("not_exist_key", "foo")
cmds, err := pipeline.Exec()
if err != nil {
    log.Fatal(err)
}

It may be reasonable to return an error when some commands execute failed, but it make me confused.
The commands execute normally, and I got the result in the cmds, but the returned err indicates me that the pipeline executes failed.
If I just check the err, then I will miss the normal result. So I must check the cmds regardless of the pipeline err.

[vmihailenco]

Just check for redis.Nil

if err != nil && err != redis.Nil {
    log.Fatal(err)
}