Redis 基本操作

参考：An introduction to Redis data types and abstractions
中文翻译：看云

Redis命令查询：Command Reference - Redis

Strings

# set and get
set     key     value                   # set value
get     key# multiple operation
mset    key1  value1  [key2 value2...]  # multiple set
mget    key1  [key2...]                 # multiple get# string operation
getrange    key     start   end         # get sub string
setrange    key     offset  value       # rewrite from specific offset
strlen      key                         # get string length
append      key     value               # append value to existing value(returns added value length)# bit operation
setbit      key     offset  value       # set bit value by offset
getbit      key     offset              # get bit value by offset# expires
setex       key     expire              # create a key with expire
set         key     ex      expire      # creat a key with expire
psetex      key     expire              # expires after milie-seconds# if exists
setex       key     value               # set if exists
setnx       key     value               # set if not exits(returns 1 on success, 0 on failed)
set         key     value     nx        # set if not exits(return OK on success, (nil) on failed)# change self
incr        key                         # increment on value
incrby      key     increment           # increment on value by increment
incrbyfloat key     increment           # increment on value by float increment
decr        key                         # same
decrby      key     decrement           # same
getset                                  # get set(can be used to count without omit)

Lists

#other
llen    key                             # list length# push
lpush   key     ele1    [ele2 ...]      # push to the left(one by one)
rpush   key     ele1    [ele2 ...]      # push to the right(one by one)
lpushx  key     value                   # push if exists
rpushx  key     value                   # push if exists# pop(get and remove)
rpop    mylist                          # pop one from the right, return null if empty
lpop    mylist                          # same# get
lindex  index                           # get value by index
lrange  key     start     end           # get element[a, b], [-2,-1] # remove
lrem    key     value     count         # default 0 remove all, <0 remove from tail# Block pop (如果有多个client在wait, 先wait的client先被服务.)
brpop   key1   [key2 ...]    timeout    # block until new element(0 wait forever)
blpop   key1   [key2 ....]   timeout    # block until new element(0 wait forever)# pop and push
rpoplpush   source  desti               # rpop from source, lpsuh to desti
brpoplpush  source  desti               # same but blocking# slice
ltrim   key     start   end             # only keep elements in [start, end]

List的常用案例：

• 记忆用户最近的更新。
• 进程间通信。使用生产者消费者模型。

rpoplpush/brpoplpush用法:

• Reliable queue (消费者取得数据后可能因为某种原因未完成)
• Circular list(sour和desti相同, 逐个遍历列表而非使用lrange一次取出所有元素, 在有多个client同时遍历，或者在遍历过程中加入新的元素时，都能依次遍历，此外，由于原子性操作，不会在丢失过程中丢失数据)

Hashes

# set
hset    key   filed    value                        # hash set
hmset   key   filed1   value1 [field2 value2 ...]   # hash multiple set
hsetnx  key   field     value                       # set field if the filed not exists# delete
hdel    key     filed1     [field2 ...]     # delete multiple fields from hash# other
hlen    key                                 # get the # of fields
hscan   key     cursor [pattern] [count]    # similar to scan, but in hashes
hstrlen key     filed                       # get length of value of field
hexists key     filed                       # check existance# get
hget    key     filed                       # get one filed
hmget   key     filed1   [filed2 ...]       # get multiple field
hgetall key                                 # get all fields with values
hkeys   key                                 # get all fields
hvals   key                                 # get all values# change value
hincrby         key     filed   increment   # hash increment
hincrebyfloat   key     field   increment   # hash increment by float

Note:

• 尽管hash非常适合用于表示对象，但我们也可以将hash用在别的地方(因为hash里面的内容是没有限制的)
• 小哈希(例如，一些拥有小值的元素)在内存中以特殊的方式编码存储，以让内存存储变得更有效

Redis Sets

# get size （or cardinality 基数、集合的势）
scard   key# add
sadd    key     value1   [value2  ...]  # add multple values to set# get
smembers    key                         # get all elements of set
srandmember key                         # randomly get one member# check existance
sismenber   key     value               # check existence# pop
spop        key                         # **randomly** pop one element# UNION
sunionstore res     key1   [key2 ...]   # get union of sets and store result

Redis Sorted sets

# add
zadd  key  score1  mem1  [score2  mem2]         # add multiple elements# get by rank
zrange      key   start   end  [withscores]     # get a range of elements
zrerange    key   start   end  [withscores]     # reverse order# operate on ranges
zrangebysore        key    min    max           # get all between min and max
zremrangebyscore    key    min   max            # remove all between min and max# get rank
zrank       key     value                       # get rank
zrerank     key     value                       # reverse order# lexicographical scores(字典序操作)
(可以用于自动完成)
zrangebylex
zrevrangebylex
zremrangebylex
zlexcount

Note: 重复调用zadd可以更新元素的score(可用于游戏排行榜)

Bitmaps

# operations
setbit  key     offset    val                   # set bit
getbit  key     offset                          # get bit
bitop   [and/or/xor/not]  dest  key1 [key2...]  # bit operate and store to dest
bitcount
bitpos  key    [start  end]                     # find the first bit of 0 or 1

可用于记录网站用户每日访问情况。

HyperLogLogs

一种类似set的数据结构，用于估计set的大小(更快地)。

对Key自身的操作(无关乎数据和数据类类型)

type        key         # get key type
del         key         # delete key
exists      key         # check existance
expire      key         # set an expire for a key
pexpire     key         # set an expire in millis for a key
persist     key         # remove expire for a key
ttl         key         # time to live
pttl        key         # time to live in milliseconds

遍历keys

keys                    # 一次返回所有的key（当keys特别多时应采用scan）
scan                    # 逐个遍历key (可以模糊匹配)