Shulou(Shulou.com)06/02 Report--

This article mainly introduces "how to achieve the like effect of PHP+Redis". In the daily operation, I believe that many people have doubts about how to achieve the like effect of PHP+Redis. The editor consulted all kinds of materials and sorted out simple and easy-to-use operation methods. I hope it will be helpful to answer the doubts of "how to achieve the like effect of PHP+Redis". Next, please follow the editor to study!

Preface

Giving likes is actually a very interesting function. There are two basic design ideas, one of which is naturally using mysql (hundreds of lines of code have not been written yet, which is toxic).

The database is directly stored on the ground, and the other is to use the business characteristics of giving likes to throw it into redis (or memcache), and then brush it back to mysql offline.

The functions I am talking about here are based on my previous projects, so there are some places that can be ignored. I mainly record the ideas for the implementation of this function. When you understand it, it is basically the same in whatever language you want to write.

Write directly to Mysql

Writing directly to Mysql is the easiest thing to do.

Just make three tables.

Comment_info

Record the main content of the article, mainly like_count,hate_count,score these three fields are the main fields of our function.

Comment_like

Record the number of times the article has been liked, and how many people have praised this data can be found directly from the table.

User_like_comment

Record which articles have been praised by users, and when you open the list of articles, the data displayed as to whether they have been liked is in it.

Shortcoming

There is great pressure on reading and writing of database.

Popular articles will be liked by many users, even in a short period of time, so it is not ideal to operate the database directly in the long run.

Redis storage is then brushed back to the database in batches

The main feature of redis is that it is fast. After all, the main data is in memory.

In addition, the main reason why I chose redis over memcache is that redis supports more data types, such as hash, set, zset and so on.

These types will be used in detail below.

Advantages

High performance

Relieve the pressure of database reading and writing

In fact, I am more concerned with alleviating the writing pressure, the real reading pressure, which can be solved through mysql master and slave and even by adding redis to cache hot data.

Writing pressure is really not good for the previous scheme.

Shortcoming

Development complexity

This is much more complicated than writing mysql directly, and there are a lot of things to consider.

Data security cannot be guaranteed

Data will be lost when redis is hung up, and if the data in redis is not synchronized in time, it may be eliminated when redis memory is replaced.

But for us to like, it's not a problem to lose a little bit of data.

In fact, the second shortcoming above can be avoided, which involves some design patterns of redis. It doesn't matter if I don't understand it. I will try my best to write it in detail. Later I will give you how to solve this shortcoming.

Knowledge preparation before design function

1. The redis data type to be used (for a specific type description, please see the link at the bottom for details):

This type of zset is mainly used for sorting or increasing or decreasing numbers. It is used as digital records of like and hate, as well as heat records.

Set is an unordered collection, which is mainly used to record whether it needs to be updated today, and to record the id of articles that have been liked (including a bit of a nuisance) today, so as to facilitate updating this part of the data at night or when there is time.

Hash this is a hash, mainly used to store data and indexes. This is used to record which article users click on and what to judge next time (I have read some online introductions using set to record, that is also possible, but I think it is more space-saving, easy to manage, and the speed of hash is fast).

List is the leader of the queue, and whether our data can safely return to mysql depends on it.

two。 How to judge the heat:

As we all know, the higher the number of likes, the higher the popularity of the article, so how to judge? Just record the number of likes directly, but what about the latest articles? For example, there is an article published a year ago, which got 50 likes, and the latest article got 49 likes, but according to the above, the article a year ago is more popular than the latest, which is unreasonable. The articles are all timely. Everyone wants to see the latest and hottest.

So! We need to deal with this timeliness in another way. Most languages have timestamps. The newer the timestamps are, the larger the number is. The timestamps are initialized and assigned to the score of the article directly, so that the latest articles will be ahead of the previous articles. Then there is the impact of likes on score. Let's assume that getting 20 likes a day is the hottest day, 60 "60" 24 "86400 seconds a day, and then getting a like is 86400 / 20 = 4320 points. The specific figures depend on your own business needs. I'm just giving an example. Of course, clicking hate also subtracts the corresponding number.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.