PHP类共享SQL查询数据的设计模式

Question

I'm a beginner to OOP. The following is the jist of my code, which I am trying to find a proper design pattern for:

class data {
public $location = array();
public $restaurant = array();
}

$data = new data;
$query = mysqli_query($mysqli, "SELECT * FROM restaurants"); 
//actually a big long query, simplified for illustrative purposes here
$i = 0;
while ($row = mysqli_fetch_array($query)) {
$data->location[] = $i.')'.$row['location']."<br>";
$data->restaurant[] = $row['restaurant']."<br>";
$i++;
}

I'd like to access the data class from another PHP page. (To print out information in HTML, hence the
tags). I prefer not to run the query twice. I understand that classes are created and destroyed in a single PHP page load. I'd appreciated design pattern guidance for managing HTTP application state and minimizing computer resources in such a situation.

Answer 1

SELECT data in database is rather inexpensive, in general speaking. You didn't need to worry about running the query twice. MySQL will do the caching part.

From your codes, you mixed up database data with HTML. I suggest to separate it.

// fetch data part
while ($row = mysqli_fetch_array($query)) {
  $data->location[] = $row['location'];
  $data->restaurant[] = $row['restaurant'];
}

// print HTML part
$i = 0;
foreach($data->location as $loc) {
  echo $i . ')' . $loc . '<br />';
}

Answer 2

If you store the data object in a $_SESSION variable, you will have access to it from other pages and upon refresh. As mentioned in other post and comments, you want to separate out the HTML from data processing.

class data {
  public $location = array();
  public $restaurant = array();
}

// start your session
session_start();

$data = new data;
$query = mysqli_query($mysqli, "SELECT * FROM restaurants"); 
//actually a big long query, simplified for illustrative purposes here
$i = 0;
while ($row = mysqli_fetch_array($query)) {
  $data->location[] = $i.')'.$row['location'];
  $data->restaurant[] = $row['restaurant'];
$i++;
}

// HTML (separate from data processing)
foreach ($data->location as $location) {
  echo $location . '<br />';
}

// save your session
$_SESSION['data'] = $data;

When you wish to reference the data object from another page

// start your session
session_start();

// get data object
$data = $_SESSION['data'];

// do something with data
foreach($data->location as $location) {
  echo $location . '<br />';
}

Answer 3

First you say this:

I'm a beginner to OOP.

Then you say this:

I prefer not to run the query twice. I understand that classes are created and destroyed in a single PHP page load.

You are overthinking this. PHP is a scripting language based on a user request to that script. Meaning, it will always reload—and rerun—the code on each load of the PHP page. So there is no way around that.

And when I say you are overthinking this, PHP is basically a part of a LAMP stack (Linux, Apache, MySQL & PHP) so the burden of query speed rests on the MySQL server which will cache the request anyway.

Meaning while you are thinking of PHP efficiency, the inherent architecture of PHP insists that queries be run on each load. And with that in mind the burden of managing the queries falls on MySQL and on the efficiency of the server & the design of the data structures in the database.

So if you are worried about you code eating up resources, think about improving MySQL efficiency in some way. But each layer of a LAMP stack has its purpose. And the PHP layer's purpose is to just reload & rerun scripts in each request.

Answer 4

You are probably are looking for the Repository pattern.

The general idea is to have a class that can retrieve data objects for you.

Example:

$db = new Db(); // your db instance; you can use PDO for this.

$repo = new RestaurantRepository($db); // create a repo instance

$restaurants = $repo->getRestaurants(); // retrieve and array of restaurants instances

Implentation:

class RestaurantRepository {
    public function  __construct($db) {
        $this->db = $db;
    }

    public function getRestaurants() {
        // do query and return an array of instances
    }
}

Code is untested and may have typos but it's a starter.

Answer 5

Saving the query results to a $_SESSION variable in the form of an array results in not having to re-run the query on another page. Additionally, it manages state correctly as I can unset($_SESSION['name']) if the query is re-run with different parameters.

I can also save the output of class data to a session variable. It seems to me that this design pattern makes more sense than running a new query for page refreshes.