A brief analysis of the principle of chain writing method of SQL operations in ThinkPHP framework

Introduction

If you have any After several interviews, it is not difficult to find that although domestic TPs have always been criticized. But this does not affect its popularity in the development of the majority of enterprises. It has a strong community and a practical and detailed Chinese manual. One thing I believe everyone is familiar with is his chain writing method. The chain writing method simplifies the SQL workload to a certain extent. OK, how is it implemented? Let's start with object-oriented and analyze the implementation principle of chain writing.

The following statement

$User->limit(10)->where('status=1')->select();

Code

We know that the object-oriented method can return a variety of data types. Of course, it can also return the object itself , so we can use this feature to achieve the printed result of

<?php
class Test{
    private $var = "";
    public function Func(){
        $this->var = "Var is change";
        return $this;
    }

}

    $obj = new Test();
    var_dump($obj);
    var_dump($obj->Func());

:

object(Test)[1]
  private 'var' => string '' (length=0)
object(Test)[1]
  private 'var' => string 'Var is change' (length=13)

It is not difficult to find: our private variable $var has changed. In other words, our $obj->Func(), after execution, returns an object with $var = "Var is change".

$User->limit(10)->where('status=1')->select();

Then this statement is not difficult to understand. After the method is executed, the object is passed to the next method, and so on.

Simple Select() implementation

<?php
    class UserModel{

        private $field     ="*"; 
        private $tableName ="";
        private $where     ="";
        private $order     ="";
        private $limit     ="";

        function field($field){
            $this->field = $field;
            return $this;
        }

        function table($tableName){
            $this->table = $tableName;
            return $this;
        }

        function order($order){
            $this->order = "ORDER BY ".$order;
            return $this;
        }

        function where($where){
            $this->where = "WHERE ".$where;
            return $this;
        }

        function limit($index, $limit = 0){
            $this->limit = "LIMIT ".$index;
            if($limit){
                $this->limit.= ",{$limit}";
            }
            return $this;
        }

        function select(){
            if(empty($this->tableName)){
                $this->tableName = str_replace("Model", "", __CLASS__);//如果表名不指定，则获取类名
            }
            $selectSql ="SELECT {$this->field} 
                         FROM `{$this->tableName}` 
                         {$this->where} 
                         {$this->order} 
                         {$this->limit}"; //构造SQL语句模版串
            echo $selectSql;
            //return mysql_query($selectSql);  执行拼接后的SQL语句
        }

    }

    $user = new UserModel();
    $user->where("`user` = 1")->order("`user` DESC")->limit(5)->select();

?>

Summary

The idea is probably to assign values ​​to each condition of the SQL statement through the chain operation method, and then process the SQL uniformly in the last step statement. This is just a simple implementation of the principle. Interested students can judge multiple types of method parameters and assign conditions more flexibly. For example, the where method can pass an array. Then you can also follow this idea and do things like INSERT(), UPDATE(), DELETE(), etc. This is just an introduction. If you want to learn more about chain writing, you can also look at the TP source code.