You may occasionally want to regenerate image thumbnails in WordPress (such as after changing image quality settings). Here is a handy function that regenerates all thumbnails for a given attachment post

function regenerate_attachment_images( $attachment_id )
{
	// We only want to look at image attachments
	if ( !wp_attachment_is_image($attachment_id) )
		return;

	$filepath = get_attached_file( $attachment_id, true );
	$metadata = wp_generate_attachment_metadata( $attachment_id, $filepath );

	// Was there an error?
	if ( is_wp_error( $metadata ) )
		$this->die_json_error_msg( $attachment_id, $metadata->get_error_message() );
	if ( empty( $metadata ) )
		$this->die_json_error_msg( $attachment_id, 'Unknown failure reason.' );

	// If this fails, then it just means that nothing was changed (old value == new value)
	wp_update_attachment_metadata( $attachment_id, $metadata );
}

The above was taken mostly from the Regenerate Thumbnails plugin with a few modifications of my own.

Example Usage

To regenerate all thumbnails for all images use the following (Warning: This could take a while if you have alot of images and/or image sizes):

$images = new WP_Query(array(
	'post_type' => 'attachment',
	'posts_per_page' => -1,
	'post_mime_type' => 'image',
	'suppress_filters' => false,

	'offset' => 0,
	'post_status' => 'inherit',
	'ignore_sticky_posts' => true,
	'no_found_rows' => true,
));

while ( $images->have_posts() ): 
	$images->next_post();
	regenerate_attachment_images( $images->post->ID );
endwhile;

Read More »

Sometimes in WordPress you want to include associated taxonomy terms with your get_posts() or WP_Query lookups. Doing so can have a noticeable impact on performance. Not to mention it’s much cleaner code-wise.

Here’s an example. I wanted to group image attachments into genre – action, adventure etc. and display that information on my sites frontend. Firstly I added my genre taxonomy to the attachment post type:

register_taxonomy('genre', 'attachment', array(
	'label' => 'Genres',
	'rewrite' => array( 'slug' => 'genre' ),
	'hierarchical' => true,
	'capabilities' => array(
		'assign_terms' => 'edit_posts',
		'edit_terms' => 'publish_posts'
	)
));

I now needed to display that information on the images associated post page (single.php) on the frontend.

The dumb way

On my first attempt I looped through the images, grabbing the associated genres and displaying them:

$images = get_posts(array(
	'post_parent' => get_the_ID(),
	'post_type' => 'attachment',
	'numberposts' => -1,
	'orderby'        => 'title',
	'order'           => 'ASC',
	'post_mime_type' => 'image',
));
foreach ( $images as $image )
	echo $image->post_title . ': ' . strip_tags(get_the_term_list($image->ID, 'genre', '', ', ', ''));

My image: Action, Adventure

This resulted in one unnecessary database call per image which could add up quickly. I needed a better way.

A smarter approach

WP_Query (which get_posts() uses to retrieve its results) supports a filter posts_clauses that lets you modify various parts of the SQL query it is about to perform. I used this to JOIN the taxonomy tables on and include the genre name(s) in the result array.

Firstly the filter (only works if you drop it in functions.php):

/**
 * Include 'size' name in image attachment lookups. This only applies if
 * INCLUDE_GENRES global variable flag is set - otherwise it will affect
 * the_loop
 *
 * @param array $pieces Includes where, groupby, join, orderby, distinct, fields, limits
 *
 * @return array $pieces
 */
add_filter( 'posts_clauses', function( $pieces )
{
	global $wpdb, $INCLUDE_SIZE;

	if ( empty($INCLUDE_GENRES) )
		return $pieces;

	$pieces['join'] .= " LEFT JOIN $wpdb->term_relationships iqctr ON iqctr.object_id=$wpdb->posts.ID
						 LEFT JOIN $wpdb->term_taxonomy iqctt ON iqctt.term_taxonomy_id=iqctr.term_taxonomy_id AND iqctt.taxonomy='genre'
						 LEFT JOIN $wpdb->terms iqct ON iqct.term_id=iqctt.term_id";
	$pieces['fields'] .= ",GROUP_CONCAT(iqct.name SEPARATOR ', ') AS genres";

	return $pieces;
}, 10, 1 );

You’ll notice the $INCLUDE_GENRES variable. This is required because without it the filter will apply to all the_loop and other queries. We only want it to apply for one specific query. Now how to use it:

$INCLUDE_GENRES = true;
$images = get_posts(array(
	'post_parent' => get_the_ID(),
	'post_type' => 'attachment',
	'numberposts' => -1,
	'orderby'        => 'title',
	'order'           => 'ASC',
	'post_mime_type' => 'image',
	'suppress_filters' => false,
));
$INCLUDE_GENRES = false;

foreach ( $images as $image )
	echo $image->post_title . ': ' . $image->genres;

My image: Action, Adventure

Perfect!

Read More »

Bootstrap 3’s NavBar component has a convoluted markup that makes it difficult to integrate into WordPress’s wp_nav_menu() function but with the help of a custom Walker and a filter it’s quite possible to get happening.

Target Markup

We’re aiming to replicate the NavBar markup from the BS3 documentation. Here it is in full. Specifically wp_nav_menu() will cover the highlighted section.

<nav class="navbar navbar-default" role="navigation">
	<!-- Brand and toggle get grouped for better mobile display -->
	<div class="navbar-header">
		<button type="button" class="navbar-toggle" data-toggle="collapse" data-target=".navbar-ex1-collapse">
			<span class="sr-only">Toggle navigation</span>
			<span class="icon-bar"></span>
			<span class="icon-bar"></span>
			<span class="icon-bar"></span>
		</button>
		<a class="navbar-brand" href="#">Brand</a>
	</div>

	<!-- Collect the nav links, forms, and other content for toggling -->
	<div class="collapse navbar-collapse navbar-ex1-collapse">
		<ul class="nav navbar-nav">
			<li class="active"><a href="#">Link</a></li>
			<li><a href="#">Link</a></li>
			<li class="dropdown">
				<a href="#" class="dropdown-toggle" data-toggle="dropdown">Dropdown <b class="caret"></b></a>
				<ul class="dropdown-menu">
					<li><a href="#">Action</a></li>
					<li><a href="#">Another action</a></li>
					<li><a href="#">Something else here</a></li>
					<li><a href="#">Separated link</a></li>
					<li><a href="#">One more separated link</a></li>
				</ul>
			</li>
		</ul>
	</div><!-- /.navbar-collapse -->
</nav>

Read on for full details!

Read More »

If you’ve installed the MediaServer or PhotoStation packages on your Synology NAS you’ve probably noticed @eaDir directories popping up everywhere. These are “hidden” folders equivalent to thumbs.db on Windows where the package stores thumbnail files associated with iTunes support. If you’re not using iTunes you don’t need these directories. You can remove them in two steps:

Disable the Service Creating Them

SSH in as root and run the following:

cd /usr/syno/etc.defaults/rc.d/
chmod 000 S66fileindexd.sh S66synoindexd.sh S77synomkthumbd.sh S88synomkflvd.sh S99iTunes.sh

Remove the existing directories

Again in SSH use the following to locate them (cd to your volume root first):

find . -type d -name "@eaDir"

and if you’re feeling adventurous you can automatically delete them like so:

find . -type d -name "@eaDir" -print0 | xargs -0 rm -rf

Read More »

If you need to do multiple CURL requests with PHP, curl_multi is a great way to do it. The problem with curl_multi is that it waits for all requests to complete before it begins processing any of the responses – for example if you’re making 100 requests and just one is slow, the other 99 will be held off for processing until the one has come back. This is wasteful so I’ve whipped up a script originally written by Josh Fraser with my own modifications and improvements that will begin processing a response immediately before dispatching the next request.
 

Usage:

The $callback (second) argument of rolling_url() can take any callback supported by PHP.

curl_multi_download(array(
	"http://translink.com.au/travel-information/services-and-timetables/trains/rss",
	"http://translink.com.au/travel-information/services-and-timetables/buses/rss",
	"http://arandombrokenurl.com/should/give/an/error",
	"http://translink.com.au/travel-information/services-and-timetables/ferries/rss",
), 'process_response');

function process_response( $info, $response ) 
{
	if ( $info['http_code'] != 200 )
	{
		echo "Error retrieving URL " . $info['url'] . "<br/>";
		return;
	};
	var_dump($info);
	var_dump($response);
}

Bonus Round

Here’s a curl_download() function, used to grab the response from a single URL (For extra points switch out curl_exec() with curl_exec_utf8() using the same method as above):

function curl_download( $url )
{
	$ch = curl_init();
	curl_setopt($ch, CURLOPT_URL, $url);
	curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
	curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, 10);
	$data = self::curl_exec($ch);
	curl_close($ch);

	return $data;
}

See inside for the full code.

Read More »

I just installed Ubuntu 13.04 in Virtualbox 4.2.16 and found much to my annoyance that the VM thought my mouse was a little higher than it actually was:

It turns out this is caused by having 3D acceleration turned on in VM Settings – Display window. After doing a bit of sleuthing I came across a forum post on virtualbox.org with a command that did the trick nicely.

Simply open a terminal and run

VBoxManage setextradata global GUI/Customizations noStatusBar

Restart your VM and voila. Perfect mouse working with 3D acceleration!

Read More »

Despite a global shift towards touch-oriented devices on the internet and the existence of a relatively old bug report, Google has decided not to add support for mobile/touch to their ChartRangeFilter for Google Charts (as of the time of writing). See for yourself:

Demo | Source – Doesn’t work on mobile. Frustrating!

This is really annoying, especially if you are dealing with a significant chunk of data that could take advantage of this functionality by allowing users to ‘zoom in’ on specific areas.

A Workaround (For Now)

Although we can’t get the ChartRangeFilter working on mobile, we can display something else instead for mobile devices such as the handy jQRangeSlider. To do so we’ll need to hook into the methods provided by google.visualization.ControlWrapper (which sits around the ChartRangeFilter).

In the source we have:

var controlWrapper = new google.visualization.ControlWrapper({
	controlType: 'ChartRangeFilter',
	...
});

We can update the selection in the ChartRangeFilter by calling the following on its ControlWrapper:

controlWrapper.setState({
	range: {
		start: new Date(2004, 01, 01),
		end: new Date(2006, 01, 01)
	}
});
controlWrapper.draw();

Combine this with a mobile UA check (terrible I know, but you can’t really use browser  feature detection for this) and a DateRangeSlider (I’m using dates on my X axis) and you get the following:

var is_mobile = /Android|webOS|iPhone|iPad|iPod|BlackBerry/i.test(navigator.userAgent);
// ChartRangeFilter doesn't work on mobile. Use a dateRangeSlider to manipulate it
if ( is_mobile )
{
	$('#filter').hide();
	$('#filter_mobile').show();

	$( "#filter_mobile" ).dateRangeSlider({
		bounds: {
			min: new Date(2001, 1, 1),
			max: new Date(2013, 3, 1)
		},
		defaultValues: {
			min: new Date(2001, 1, 1),
			max: new Date(2013, 3, 1)
		},
		step: {
			months: 1
		},
		arrows: true,
		wheelMode: null
	}).bind('valuesChanged', function(e, data) {
		controlWrapper.setState({range: { start: data.values.min, end: data.values.max }});
		controlWrapper.draw();
	});
}

Demo

ChartRangeFilter workaround for mobile devices

Try the demo below. View it on a mobile device (or use a mobile UA string). You get the standard Google Charts ChartRangeFilter for desktop and jQRangeSlider for mobile. Everything works nicely!

Demo | Source – Snazzy mobile-specific date range widget

Read More »

This morning I noticed one of my drives wasn’t mounted so I attempted to mount manually and got the following error message:

$ sudo mount /dev/sdi1 /mnt/my_drive/

mount: wrong fs type, bad option, bad superblock on /dev/sdi1,
missing codepage or helper program, or other error
In some cases useful info is found in syslog – try
dmesg | tail or so

Strange. Following the messages advice I checked out dmesg:

$ dmesg | tail

[ 213.962722] EXT4-fs (sdi1): no journal found

The Solution

If this happens to you, you can ignore the error and mount in readonly mode using the following command:

sudo mount -o loop,ro,noexec,noload /dev/sdi1 /mnt/your_broken_partition/

Thanks to Computer Forensics for their useful post on this issue.

ALWAYS BACKUP YOUR DATA

Read More »

As I’m sure is the case with just about every WordPress user, my site is constantly hit with failed login attempts from bots – usually originating from other countries. I had the idea last night to implement country-based login restrictions using CloudFlare’s IP Geolocation server variable.

My goal was to redirect redirect all users visiting the login page from countries other than Australia to the home page. This occurs before a login attempt can even be made.

Here’s how it’s done:

• Go into CloudFlare settings and make sure the IP Geolocation option is turned on
• Add the following to your active themes functions.php file

  // Limit login countries - requires cloudflare
  add_filter( 'authenticate', 'login_failed', 1, 3);
  function login_failed( $user, $username, $password )
  {
      if ( !isset( $_SERVER['HTTP_CF_IPCOUNTRY'] ) )
          return $user;
  
      if ( !in_array( $_SERVER['HTTP_CF_IPCOUNTRY'], array('AU') ) )
      {
          wp_redirect( home_url() );
          exit;
      }
  
      return $user;
  }

To test simply use a country other than your own. Once it’s confirmed working switch to your country and you’re good to go!

Read More »

Here is a super simple SQL query to determine the size of all tables in a given database:

SELECT table_name AS "Table",
	round(((data_length + index_length) / 1024 / 1024), 2) AS Size_in_MB
FROM information_schema.TABLES 
WHERE table_schema = 'my_db_name'
ORDER BY Size_in_MB DESC

This can be very useful when debugging for instance in figuring out which rogue plugin/module on your site is dumping hundreds of megs of data into your database.

Read More »