Redis is an in-memory data store and uses key-value structures. It is used as a caching layer between the app and the database. That is because redis is fast as it is stored on memory. The downside is that the data is volatile because the data would be lost if the server crashes. Now it can be used as the main database because of data persistance, JSON support, search, and object mapping. Having redis with modules, it is called Redis stack.
Redis Data Types
We still store in key value pairs. It could be a string, a set (an unordered collection of strings), hashes (like objects), lists (string based), sorted sets (sets where the values are ordered by some column).
Redis Keys
The keys of redis can be called anything. A common convention is to use books:1 and books:2 for example.
Making a Redis database
We can install Redis in our computer.
Basic commands
Strings Redis stores numbers as strings. If we want to create a new key value set we run the commands
SET name JohnTo get a value
GET name
To delete data, where name2 is the key of the key value pair.
DEL name2
To define multiple key value pairs
MSET name2 Yoshi color green rating 10
To get multiple values
MGET name name2
To get the first characters of a string
GETRANGE name 0 3
-1 is the last character.
GETRANGE name -3 -1
To set a range of characters in an existing key
SETRANGE name 2 abc
To increase a number by one
INCR rating
Even if rating is stored as a string, Redis knows that it is a number. To decrease
DECR rating
To increase more
INCRBY rating 5
Command options
Options are extra flags that we can add to some of the commands. For example, if we run
SET name yoshi EX 7
The value will be yoshi and after 7 seconds it will be deleted.
SET name Yoshi NX
NX is a flag to set something if it doesn’t exist.
SET name Yoshi XX
XX is a flag that indicates to only overrides if it does exist.
SET name Peach GET
Will return the previous value and set the new value as Peach.
Sets (Data type)
Sets are unordered collection of strings which all have to be unique.
SADD names mario peach
It will add them to the end if it does exist. To delete one element of the set
SREM names yoshi
If we want to return two sets in something like a join, we run
SUNION names morenames
This does not create a new set.
SISMEMBER names yoshi
Check whether a value exist in the set. Returns 1 if found.
Lists (Data type)
Is a collection of different values. They do not need to be unique. They have an unique order based on how they were added. The way values are queried are very different than sets. In a list we can think of a queue. Accessing a value in the middle of a list is slow. If we are working in two ends of the list, that operation is very fast. Lists can be good for when data progresses over time. Or when we want the last 10 updates or so. To add a value to the list from the left, do
LPUSH orders [value]
If we do it again, the last value added will be the first on on the list. To push from the right
RPUSH orders [name] [name]
These will be added at the end of the list. To take things off the list
LPOP orders 1
This removes 1 element from the left on the list. To get a specific item
LINDEX orders 1
To get the position
LPOS orders ryu