As I already mentioned in my previous post, to measure the distance between two points on a sphere one can use Haversine formula. For a detailed description I suggest visiting the wikipedia site. In our example we will use a simplified version of Haversine formula, which is caused by a really simple reason. Executing a web query with the actual Haversine formula might lead to a really slow response from the database and our simplified version will retrieve results which do not differentiate that much with the original version.

For retrieving the company branches in a certain radius we will first build a php class, which will be responsible for translating the zip code into geographical lengths and then use this values to retrieve the results and return them as json_encoded() data. Because our class will be relatively long, I will split it into parts and explain it line by line.

Since we are using an AJAX request to retrieve our data and the location of our php file is easy to guess just by looking at our jQuery code (will come after php class), we add a simple check of the apache environmental variable to check, if it actually is an XMLHttpRequest. A lot of you will probably say that this security measure is not enough, but for this example, it will at least stop all the robots trying access this file and since looking up addresses might consume a lot of resources of your server, this will save you some speed lost on servers with high traffic.

1
2

if ($_SERVER["HTTP_X_REQUESTED_WITH"] != 'XMLHttpRequest')
  die('why are you doing this??');

In line 3 we create our class. Since the subject of this example is finding company branches, lets call it Branch. In lines 5 to 13, we define some variables, which we will use in our class. I think they are pretty much self explanatory, and you will understand, what is their usage just by looking at the rest of the class.

3
4
5
6
7
8
9
10
11
12
13

class Branches {
 
  private static $_key = 'YOUR GOOGLE MAPS KEY';
  private static $_host = 'localhost';
  private static $_db_user = 'root';
  private static $_db_pass = 'pass';
  private static $_db_name = 'db';
 
  private $zip, $radius, $data_set;
 
  public $results = array();

In our constructor we connect to the database and set the radius and zip properties. You could notice that I convert the zip property to integer as well and I additionally add leading zero for strings, which length equals 4. Well, this class was written for German zip codes. You might want to change this parts so it suits your needs. After all it will be important in the part where we will convert the zip code to geographical length, so Google can find this zip code in their database.

14
15
16
17
18
19
20
21
22
23
24
25

public function __construct()
{
  $sql = mysql_connect(self::$_host, self::$_db_user, self::$_db_pass) or die(mysql_error());
  mysql_select_db(self::$_db_name, $sql) or die(mysql_error());
 
  $this->radius = (int)$_GET['radius'];
  $this->zip = (int)$_GET['zip'];
 
  // add leading zeros in case of 0... zip codes
  if (strlen($this->zip) == 4)
    $this->zip = sprintf("%05s", $this->zip);  
}

In line 26 we define a public method translateZipToLatLng() and in line 39 a private method getGeolocation(), which both will be responsible for translating the zip code into the geographical length. To get a detailed description, what this two methods actually do, refer to my other post: Geolocation class - retrieving longitude and latitude using google maps.

26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

public function translateZipToLatLng()
{
  $url = 'http://maps.google.com/maps/geo?q=' . urlencode($this->zip . ', Germany') . '&output=csv&key=' . self::$_key;
 
  $response = $this->getGeolocation($url);
 
  if (substr($response, 0, 3) === '200') {
    $geo = explode(',', $response);
    return $data = array($geo[2], $geo[3]);
  }
  return false;
}
 
private function getGeolocation($url)
{
  $init = curl_init();
  curl_setopt($init, CURLOPT_URL, $url);
  curl_setopt($init, CURLOPT_HEADER,0);
  curl_setopt($init, CURLOPT_USERAGENT, $_SERVER["HTTP_USER_AGENT"]);
  curl_setopt($init, CURLOPT_RETURNTRANSFER, 1);
  $response = curl_exec($init);
  curl_close($init);
 
  return $response;
}

In line 51 we define a private method getResults, which, as you can see, will be responsible for retrieving all the information from the “branches" table and additionally measure the approximate distance from the zip code provided by the user to our actual company branches. In line 56 we define private method populate results, which will be responsible for creating an array of the results retrieved from the database. As you can see, if our query returns zero rows we define an error, if our query returns at least one row, we create an array with all the information about the company. The last public method in line 75, findBranches() will be the one, which we will call after instantiating our class and translating zip code into geographical length.

51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79

private function getResults($latlng)
{
  return mysql_query("SELECT *, (POW((69.1*(longitude-{$latlng[1]})*COS({$latlng[0]}/57.3)), '2') + POW((69.1*(latitude-{$latlng[0]})), '2')) AS distance FROM branches HAVING distance < {$this->radius} ORDER BY distance ASC;");
}
 
private function populateResults()
{
  if (!mysql_num_rows($this->data_set)) {
    $this->results = array(array('error' => 'none'));
  } else {
    while ($row = mysql_fetch_assoc($this->data_set)) {
      $this->results[] = array(
        'name'      => utf8_encode($row['name']),
        'zip'       => $row['zip'],
        'city'      => utf8_encode($row['city']),
        'latitude'  => $row['latitude'],
        'longitude' => $row['longitude'],
        'distance'  => round(sqrt($row['distance']),1),
      );
    }
  }
  return $this->results;
}
 
public function findBranches($latlng)
{
  $this->data_set = $this->getResults($latlng);
  return $this->populateResults();
}

Our class for retrieving results is done. For the simplicity I did not assume the case, when the database query returns zero rows, in which radius should be automatically extended. You can easily do this, by writing one more method, which will be responsible for extending the radius and adding a simple if...else check in our findBranches() method, to make it loop until your query returns any results. As I said before though, in a database with many entries, executing a web query with a formula complicated like the one we are using, retrieving results might consume a lot of time, and creating a loop, in which you will repeat the same query might lead to a really long time a user will have to wait for a response. In my case, the database was a little bit over 1000 entries and extending the radius and repeating the query did not result in any remarkable performance lost though.

The last part is the actual usage of the class we just wrote. Basically in line 84 we instantiate our class, in line 85 we translate the zip code to geographical length and according to the value returned by this method we use json_encode() function to return the results as a json representation of the array we just created or an error.

84
85

$branches = new Branches();
echo ($latlng = $branches->translateZipToLatLng()) ? json_encode($branches->findBranches($latlng)) : json_encode(array('error' => 'zip_invalid'));

Since our class for retrieving the results in a certain radius is ready, we should move to our client side code.

To make this simple, I will just show you the whole jQuery code and discuss it line by line.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71

$(document).ready(function() {  
  var mapContainer = $('#map')[0];
  if (GBrowserIsCompatible()) {
    if (mapContainer) {
      var map = new GMap2(mapContainer);
      map.addControl(new GLargeMapControl);
      map.setCenter(new GLatLng(52.519015140666,13.419671058654785), 14);
 
      $('#submit').click(function() {
        var zip = $('#zip').val();
        var radius = $('#radius').val();
        if (zip == '') {
          alert("You need to provide a zip code");
          return;
        }
        var rand = new Date().getTime();
        $.getJSON('zip.php?rand=' + rand, { zip: zip, radius: radius }, function(data) {
          var markers = [];
          var descriptions = [];
          var bounds = new GLatLngBounds();
 
          map.clearOverlays();
          $('#error').html('');
          $('#locations').html('');
          $('#closest_location').html('');
          var flag = true;
 
          $.each(data, function(i, item) {
            if (item.error == 'none') {
              alert('No results found or ZIP incorrect');
              $('#error').append('No results found or ZIP incorrect');
              flag = false
              return false;
            }
            var description =
              '<div class="branch" style="float: left;">' +
              item.name + ' (~ ' + item.distance + ' km)<br />' +
              item.zip + ' ' + item.city + '<br />' +
              '<span class="branch_show" style="cursor: pointer; color: blue; text-decoration: underline;" id="' + i + '">Show</span></div>';
            var lat = item.latitude;
            var lng = item.longitude;
            var point = new GLatLng(lat, lng);
            var marker = new GMarker(point);
            markers[i] = marker;
            descriptions[i] = description;
            $('#locations').append(description);
            map.addOverlay(marker);
            bounds.extend(point);
            GEvent.addListener(marker, 'click', function() {
              marker.openInfoWindowHtml(description);
            });
          });
 
          $('#closest_location').append(descriptions[0]);
 
          map.setZoom(map.getBoundsZoomLevel(bounds));
          // just in case we did not find anything
          if (!flag)
            map.setCenter(new GLatLng(51.08282186160976,10.26123046875),6);
          else
            map.setCenter(bounds.getCenter());
 
          $('.branch_show').click(function() {
            var id = $(this).attr('id');
            markers[id].openInfoWindowHtml(descriptions[id]);
          });
        });
      });
    }
  }
});

As usually we want to execute the code on onLoad of our document and we do that in line 1. In line 2 we define our container for the map and in line 3 and 4 we check if browser is compatible and if container was found respectively.

In line 5 we instantiate the Gmap2 inside of our map container, we add some controls in line 6 and set the center in line 7. In line 9 we add a event listener which triggers after a user clicks submit button.

In lines 10 to 15 we add a simple check, if the field with the zip code is not empty and it if is we alert an according message and we break the execution of the code. Additionally we define a zip variable in line 10 and radius variable in line 11, which will hold according information to use in our json request.

In line 16 we define rand variable which we pass during our json request to prevent internet explorer to cache the resource. If we would not do that, and changes you would do to zip.php file would not be visible, because IE would use the cached version of this file.

In line 17 we make a request to the zip.php file, which we wrote earlier and we pass the zip code and the radius to our php file. After the calculation is done and you do not have any errors in your php file we retrieve the information, which we can finally process.

In line 18 and 19 we define two variables as arrays, which will be responsible for holding the information about markers and their corresponding descriptions and in line 20 we instantiate GLatLngBounds object, which we will later use to display the information on our map with matching zoom level.

In line 22 to 26 we remove all markers and we empty all containers holding the information about the places from a previous request. We do that to prevent displaying of the same markers after a user requests the zip.php file again. This way, if a user changes his or hers query, we will be sure that we are displaying only the information relevant to the current request and we wont have any trash from the previous requests. In line 26 we define a flag as boolean to use it later to display error information and redisplay the center of the map.

In line 28 we loop through the information we received in response. If an error was found, we break our loop and we display and alert according information to the user. We also set the flag to false to redisplay center of the map later in the code. This would be lines 29 to 34.

In lines 35 to 39 we put together the HTML code which we will use to display the information about the branches we are going to display on the map.

In lines 40 to 43 we define two variables lat and lng and use them to instantiate GLatLng object which we define as point variable. In line 44 we define marker variable as an instance of GMarker object and in lines 45 and 46 we add it to the arrays we defined earlier. In line 47 we append the description to the container we need to define in our HTML and in line 47 we add markers to the map. In line 48 we extend the bounds of our map by using a method extend() of GLatLngBounds object we defined earlier. This amazing functionality of google maps will let us display a square with all our markers and google map will take care of setting the correct zoom level.

In lines 49 to 51 we add a listener which will be triggered on a click event and will display the description in an info cloud. In line 52 we close the loop and in line 54 we append the first description to the closest_location container. To assure that the first description is really the closest place to the zip code provided by the user, you cannot forget about putting the “ORDER BY distance ASC” in the php code of zip.php file.

In line 56 we set the zoom of our map using our bound object I was mentioning before by utilizing getBoundsZoomLevel method of GMap. As you can see in the description in line 57, if our php code did not return any results, we override the zoom level we defined in line 56 be setting it again to the default value (line 59). On the other hand, if our php code returned any results we set the center of the map using our bounds object again (line 61).

The last part of this code is just a listener which is triggered when a user clicks on the pseudo link show. We retrieve the id of this link and display the according info window on the map.

That would be it. The code is done. I know it is not perfect and accurate either. You can still use it in many different ways. It does not necessarily need to be a zip code search. Using it you can display distances between users and any other stuff connected with distances. In the second part I will show you how you can make this code a little bit more accurate using google maps build in api to show the directions, how to get from the zip code provided by the user to one of the company branches.

If you do not want to copy all the code, here is a zip package for you with working files.

One thing we need to make a decision about before we will start coding is when we should use the Haversine formula and when we should display the markers according to the borders of the map. According to this study, using the Haversine formula in a web query on a MySQL database with 800k entries might take up to 8 seconds. Therefore it will be crucial to split our jQuery code into requests concerning retrieving the data based on the Haversine formula and data within the borders of the map. In this example I won't give a complete walk through of how to measure the distance between two points (it will probably be a good subject for another example). In this example I will only show how to display data on an "moveend" event.

Let's take a look at the code. In line 1, we are executing the code after the document is loaded. I would consider it a standard procedure in writing unobtrusive JavaScript. In line 2, we define our map container, which corresponds with a HTML tag whose id equals map. In line 3, we are checking if the user’s browser is compatible with Google Map API and if it is, and our map container also exists, we can finally execute the rest of the code.

In line 5, we are instantiating our Google map and in lines 6 and 7 we are defining some controls for our map, the center (GLatLng) and zoom level (14). In line 9, we would have our code to measure the distance from given zip code and to display the markers on our map. The most interesting part of our code starts in line 14. We add a listener, which triggers on "moveend" event. In line 15, we use the getBounds() method, which returns GLatLngBounds object for us. In lines 16 to 19, we use getSouthWest() and getNorthEast() methods of our GLatLngBounds objects. This methods return GLatLng object, therefore we can directly use its methods to retrieve the geographical lengths of our south west and northeast corner. The last part is pretty simple, and I will leave it to you to adjust it to your needs. In line 20, we use the AJAX request, and we pass all geographical lengths to our php file. We assume that the response will be a valid json_encode() representation of the data retrieved from the database, and in lines 21 to 23, we add our markers to the map.

Here is the code you have been waiting for:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

$(document).ready(function() {  
  var mapContainer = $('#map')[0];
  if (GBrowserIsCompatible()) {
    if (mapContainer) {
      var map = new GMap2(mapContainer);
      map.addControl(new GLargeMapControl);
      map.setCenter(new GLatLng(52.519015140666,13.419671058654785), 14);
 
      $('#submit').click(function() {
 
        // here goes the code handling calculatig of distance
        // truncated because it is not a part of this post
 
        GEvent.addListener(map, "moveend", function() {
          var bounds = this.getBounds();
          var swLat = bounds.getSouthWest().lat();
          var swLng = bounds.getSouthWest().lng();
          var neLat = bounds.getNorthEast().lat();
          var neLng = bounds.getNorthEast().lng();
          $.getJSON('helper.php', { swLat: swLat, swLng: swLng, neLat: neLat, neLng: neLng }, function(data) {
            var point = new GLatLng(data.lat, data.lng);
            var marker = new GMarker(point);
            map.addOverlay(marker);
          });
        });
      });	
    }
  }
});

You might ask why we are placing this code inside of jQuery's click event? The answer is simple. This way we can prevent triggering the "moveend" event, before a user uses the form. Only after he or she uses the form, our listener will react to the user's interaction with the map.

Another question might be, why we are not using GBounds.extend() method to adjust the zoom level? An answer for this question is as simple as the last one. We already defined the zoom level by passing the geographical lengths of our rectangle to the php file and since we triggered "moveend" event, we always have an actual part of the map, which a user is seeing at the point of triggering it.

I hope you will like this example. I did not want to provide the whole code for this example, because first of all I wanted to split it into two entries and second of all, I did not want to provide an example which you can copy and past, but rather give a solution for a certain problem and possibility start a discussion about doing it better. Have fun with this one.

To see, how our final solution will look like, go here

1. Geo Class - Geo.class.php

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

class Geo
{
 
    private $xml;
 
    protected function _init()
    {
        $this->xml = simplexml_load_file('geo.xml', 'SimpleXMLElement');
    }
 
    public function getDropDown($id, $option = array())
    {
        $this->_init();
 
        $select = '<select name="' . $id . '" id="' . $id . '">';
 
        if ($option['include_custom'] != '')
            $select.= '<option value="" selected="selected">' . $option['include_custom'] . '</option>';
 
        foreach ($this->xml->children() as $child) {
            $select.= '<option value="' . $child->name . '">' . $child->name . '</option>';
        }
 
        $select.= '</select>';
 
        return $select;
    }
 
    public function getLocation($identifier)
    {
        $this->_init();
 
        $coordinates = array();
 
        foreach ($this->xml->children() as $child)
            if ($child->name == $identifier) {
                $temp = explode(',', $child->Point->coordinates);
                $coordinates = array(
                                    'lat' => $temp[0],
                                    'lng' => $temp[1]
                );
            }
 
        return $coordinates;
    }
}

It wasn't really necessary to put this functions into a class, but it is just a habit We define a protected function _init(), which we call every time we want to use our KML file to retrieve information out of it. The function simplexml_load_file initializes an SimpleXMLElement object for us, which we can use later. The function getDropDown with the first mandatory argument and the second optional will be responsible for displaying the drop down field for us. I think I don't need to explain it in detail. The last function getLocation will retrieve the latitude and longitude from KML file. So let's move to index.php file, which will be the core of this example.

2. Core file - index.php

1
2

include('Geo.class.php');
$geo = new Geo();

Before we send any headers we include our Geo.class.php file and initialize our Geo class. After that goes the usual tag tags and the google maps and jQuery scripts. In the section of our document goes the actual jQuery part.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

$(document).ready(function() {  
  var mapContainer = $('#map')[0];
  var markerStart;
  var markerEnd;
  if(GBrowserIsCompatible()) {    
    if(mapContainer) {
      var map = new GMap2(mapContainer);
      map.addControl(new GLargeMapControl);  
      map.setCenter(new GLatLng(52.519015140666,13.419671058654785), 14);
 
      $('#start').change(function() {
        var location = $(this).val();
 
        if (location == '')
          return;
 
        $.getJSON('helper.php', { location: location }, function(data) {
          if (markerStart)
            map.removeOverlay(markerStart);
 
          pointA = new GLatLng(data.lat, data.lng);
          markerStart = new GMarker(pointA);
          map.addOverlay(markerStart);
        });
      });
 
      $('#end').change(function() {
        var location = $(this).val();
 
        if (location == '')
          return;
 
        $.getJSON('helper.php', { location: location }, function(data) {
          if (markerEnd)
            map.removeOverlay(markerEnd);
 
          var blueIcon = new GIcon(G_DEFAULT_ICON);
          blueIcon.image = "http://gmaps-samples.googlecode.com/svn/trunk/markers/blue/blank.png";
          var markerEndOptions = { icon: blueIcon };
 
          pointB = new GLatLng(data.lat, data.lng);
          markerEnd = new GMarker(pointB, markerEndOptions);
          map.addOverlay(markerEnd);
        });
      });
 
      $('#submit').click(function() {
        if (window.pointA === undefined || window.pointB === undefined) {
          alert("You need to choose start and end");
          return;
        }
 
        var directions = new GDirections(map);
        route = Array(pointA, pointB);
 
        GEvent.addListener(directions, 'load', function() {
          var dist = directions.getDistance().html;
          $('#distance').html(dist);
        });
 
        directions.loadFromWaypoints(route);
      });
    }
  }
});

Maybe it is long but pretty straight forward. As you can read in my other examples, we first initialize jQuery, so it starts after our document is loaded. After that we define our map container and our markers. After we check if our browser is compatible and our map container exists, we initialize the map with couple options (large controls, zoom level = 14 and the center of the map).

We add two event listeners, which are triggered on onchange event of our drop downs (with id start and end respectively, the html code will follow). If you use the $option parameter in getDropDown function for displaying custom blank option, we check if the user chose this option and return false if he did. If he chose a valid option, we use an AJAX request and retrieve the geolocation from our helper file.

After our call returns some data, we check, if we have already any markers saved in our map and if we do, we first remove them and then place a new marker according to the option chosen in the drop down.

We repeat the same for the drop down with id 'end'. Actually we could have put this two events into a separate function, but to make this example clear I purposely didn't do that.

The last event listener is triggered, when a user clicks on a 'calculate route' button. First we check, if both points were placed by the user, and if not we display an alert, that he or she has to provide valid starting and ending point. This check might not work in IE, therefore you might need to do some changes, so it also works in IE.

We initialize GDirections Class and define our route with starting and ending point. I also use the getDistance method to display the amount of kilometers between this two points. As the last thing we need to do is to display the directions on the map and voila, we are done

What we still need it the rest of index.php file and our helper, and KML file.

1
2
3
4
5
6
7

<div id="navigation">
        Start <?php echo $geo->getDropDown('start', array('include_custom' => 'please choose...')); ?>
        Stop <?php echo $geo->getDropDown('end', array('include_custom' => 'please choose...')); ?>
        <input type="button" name="submit" id="submit" value="calculate route" />
</div>
<div id="distance"></div>
<div id="map" style="width: 1000px; height: 600px;"></div>

Pretty straight forward. We display our navigation with an empty option 'please choose...', the 'calculate route' button and containers for distance and map.

3. Our helper - helper.php

1
2
3

include('Geo.class.php');
$geo = new Geo();
echo json_encode($geo->getLocation($_GET['location']));

Pretty simple, isn't it Just initialize the class and call the method to retrieve location with the location argument used in the _GET request.

4. KML file - geo.xml

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

<?xml version="1.0" encoding="UTF-8"?>
<kml xmlns="http://earth.google.com/kml/2.2">
  <Placemark>
    <name>Alexanderplatz</name>
    <description>Center of Berlin</description>
    <Point>
      <coordinates>52.52238370531909,13.409671783447266</coordinates>
    </Point>
  </Placemark>
  <Placemark>
    <name>Frankfurter Tor</name>
    <description>A nice place to live</description>
    <Point>
      <coordinates>52.51595970749194,13.4527587890625</coordinates>
    </Point>
  </Placemark>
  <Placemark>
    <name>Warschauerstr.</name>
    <description>Place to go and spend some time with friends</description>
    <Point>
      <coordinates>52.508568336719854,13.451085090637207</coordinates>
    </Point>
  </Placemark>
  <Placemark>
    <name>Revelerstr.</name>
    <description>Cassiopeia Club</description>
    <Point>
      <coordinates>52.50721006978328,13.457050323486328</coordinates>
    </Point>
  </Placemark>
  <Placemark>
    <name>Frankfurter Allee</name>
    <description>Shopping Center</description>
    <Point>
      <coordinates>52.513818166144006,13.474087715148926</coordinates>
    </Point>
  </Placemark>
</kml>

It is a file I prepared by myself. As far as I know, you can create them using the Google Earth application. Probably they can be way more complex than this one. But I guess it will be a nice subject for next example Good luck with this one and as usually, in case of any question, you know how to find me. Cheers!

load content box 1,3,5
load content box 2,4

OK, let's break this code into pieces. First of all we need our HTML code, where our function can load the content. As usually, don't forget to include the jQuery library, because without nothing will really happen.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

<style type="text/css">
.box1, .box2, .box3, .box4, .box5 {
	width:100px;
	height:100px;
	border:1px solid #000;
	float:left;
}
span {
	cursor:pointer;
	font-weight:bold;
}
#unload {
	color:red;
}
</style>
<span class="1,3,5">load content box 1,3,5</span><br /><span class="2,4">load content box 2,4</span>
<div class="box1"></div>
<div class="box2"></div>
<div class="box3"></div>
<div class="box4"></div>
<div class="box5"></div>

I think this code is self explanatory. Our span elements will register an on click event and display the content in according boxes. The only thing you need to know is that class attribute defines the numbers of boxes, in which our content will be displayed (our first link will display content in box1, box3 and box5, therefore the name of the class is '1,3,5'). It's also important that you keep the names of the classes with according numbers in the div containers, because this code won't work either. If you decide to change them, don't forget to change it in the jQuery function as well. Let's move to the jQuery part.

1
2
3
4
5

$(document).ready(function(){
  $('span').click(function() {
    var boxes = $(this).attr('class');
  });
});

We load our event listeners as soon as our DOM is loaded. In the first part of this code an event listener, which will be triggered whenever a user clicks on a span element. The first thing our function does is saving of the class value in variable 'boxes'.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

$(document).ready(function(){
  $('span').click(function() {
    var boxes = $(this).attr('class');
    if($(this).html() == 'unload boxes '+boxes) {
      var empty = boxes.split(',');
      for(var i=0;i<5;i++) {
        $('.box'+empty[i]).html('');
        $(this).attr('id','')
      }
      $(this).html('load content box '+boxes);
    } else {
      $.getJSON('json.php',{boxes:boxes},function(data){
        $.each(data, function(i,item){
          $(item.class).html(item.text);
        });
      });
      $(this).html('unload boxes '+boxes);
      $(this).attr('id','unload')
    }
  });
});

Full code might be a little bit tricky, therefore I'll try to explain it as good as possible. We add and if else statement. The first part of our statement will be responsible for unloading the boxes (simply removing the content out of them), the second part will do exactly the opposite. Let's start with explaining the second part first, cause that's the part, which will be executed first. We use an HTTP GET request to retrieve JSON data from an external file. In our case it will be json.php (content of this file will follow). As an addition we pass an argument boxes, which will tell our php file, for which boxes we want to load the content. After we receive the JSON data, we loop through it and paste it into according boxes. It is worth to mention in this moment that our script will receive a multi-dimensional array, in which we will save the classes of the boxes, in which we want to display our text and according texts. After our loop is done, we will replace the text of the link and add an id to give it a different color.

The first part of the script checks the text of the link, which was clicked by the user. If it contains the string, with which we replaced it, after loading of JSON data was done, we use again the class of our span attribute to remove the text from according boxes and we replace the text again.

The only thing left is the content of php file, which we used to create JSON data.

1
2
3
4
5
6
7
8
9

$array = array ();
$boxes = explode(',',$_GET['boxes']);
 
foreach($boxes as $box) {
  $push = array('class' => '.box'.$box, 'text' => 'test for box '.$box);
  array_push($array,$push);
}
 
echo json_encode($array);

In our php file we first initialize our array and we split the received data. For each value we received we extend our array adding a class name and corresponding text. In the last part we encode the array into JSON data and echo it. You can do all the magic you can imagine in this file. You can use IDs of the articles or users from your database and save it in the array. You can even extend this array and the jQuery function by extra fields.

One thing worth mentioning is that we use id attribute to change the color of our links. It is not really best practice, because multiple ids with the same name in one document wont validate with W3C rules. However, we do that on client side, therefore it shouldn't be an issue for a validation.

That's all. I hope you like this example. I know it's pretty custom and you will need to adjust it for your needs, but I hope it's clear enough for you to understand how JSON works and how to proceed retrived JSON data.