Download - LEXIS

Document technical information

Format pdf
Size 3.4 MB
First found Jun 9, 2017

Document content analysis

Language
English
Type
not defined
Concepts
no text concepts found

Persons

Mariah Carey
Mariah Carey

wikipedia, lookup

John Resig
John Resig

wikipedia, lookup

Organizations

Places

Transcript

jQuery Hotsh t
Ten practical projects that exercise your skill, build your
confidence, and help you master jQuery
Dan Wellman
BIRMINGHAM - MUMBAI
jQuery Hotsh t
Copyright © 2013 Packt Publishing
All rights reserved. No part of this book may be reproduced, stored in a retrieval system,
or transmitted in any form or by any means, without the prior written permission of the
publisher, except in the case of brief quotations embedded in critical articles or reviews.
Every effort has been made in the preparation of this book to ensure the accuracy of the
information presented. However, the information contained in this book is sold without
warranty, either express or implied. Neither the author, nor Packt Publishing, and its dealers
and distributors will be held liable for any damages caused or alleged to be caused directly or
indirectly by this book.
Packt Publishing has endeavored to provide trademark information about all of the
companies and products mentioned in this book by the appropriate use of capitals.
However, Packt Publishing cannot guarantee the accuracy of this information.
First published: March 2013
Production Reference: 1150313
Published by Packt Publishing Ltd.
Livery Place
35 Livery Street
Birmingham B3 2PB, UK.
ISBN 978-1-84951-910-6
www.packtpub.com
Cover Image by Girish Suryawanshi ([email protected])
Credits
Author
Dan Wellman
Reviewers
Kaiser Ahmed
Project Coordinator
Anurag Banerjee
Proofreader
Aaron Nash
Carlos Estebes
Olivier Pons
Hajan Selmani
Christopher Stephen Sidell
Acquisition Editor
Robin de Jongh
Lead Technical Editor
Sweny M. Sukumaran
Technical Editor
Veronica Fernandes
Indexer
Tejal R. Soni
Production Coordinator
Arvindkumar Gupta
Cover Work
Arvindkumar Gupta
Foreword
I am very honored to have the opportunity to write the foreword for a Dan Wellman book.
I've been a fan of Dan's since I first read his jQuery UI book. Then I got the opportunity to
meet him in Oxford, England in February of 2012. Needless to say when he asked me to write
the foreword for his latest book I didn't even think about how I wouldn't have the time to fit
it in until the very last minute, but I immediately said yes.
Unlike other traditional jQuery books that dwell on how a jQuery statement is structured and
functions, Dan assumes that you are already familiar with jQuery. What he is going to teach
you is how to use jQuery for fun. He will walk you through a series of fun projects. Most of
these projects will be very useful in your personal and professional websites. The projects
include building a jQuery Mobile application, interactive Google Maps, Chrome Extensions,
and Infinite Scrolling to name a few.
jQuery has been making web development easier since January 2006 when John Resig
debuted jQuery at a New York City Barcamp with two other projects that day. Seven years
later it's great to see that you can still learn to accomplish fun things with a library as simple
yet as vast as jQuery.
I look forward to reading Dan's next book as I am sure this won't be his last.
At least I hope not.
Ralph Whitbeck
Modern Web Advocate at appendTo
jQuery Board Member
About the Author
Dan Wellman is an author and front-end engineer who lives on the South Coast of the
UK and works in London. By day he works for Skype, writing application-grade JavaScript,
and by night he writes books and tutorials focused mainly on front-end development. He
is also a staff writer for the Tuts+ arm of the Envato network, and occasionally writes for
.Net magazine. He's a proud father of four amazing children, and the grateful husband of a
wonderful wife. This will be his seventh book.
I'd like to thank my family and friends for their continued support, you
guys rock. I'd also like to thank my tireless PA, Derek Spacagna, for his
persistent encouragement, and my friend Michael Chart, without whose
mathematical genius some of the examples would not have been possible.
About the Reviewers
Kaiser Ahmed is a professional web developer. He has gained his B.Sc. from Khulna
University of Engineering and Technology (KUET) and M.Sc. in Computer Science and
Engineering from United International University, Dhaka. He is also a co-founder of
CyberXpress.Net Inc (www.cyberxpress.net), based in Bangladesh.
He has been working as Sr. Software Developer at Krembo Interactive and D1SH.COM CORP.,
Canada, for 2 years.
He has a wide range of technical skills, Internet knowledge, and experience across a
spectrum of online development in the service of building and improving online properties
for multiple clients. He enjoys creating site architecture and infrastructure, backend
development using open source toolset (PHP, MySQL, Apache, Linux and others, that is
LAMP), and frontend development with CSS and HTML/XHTML.
I want to thank my loving wife, Maria Akther, for her great support.
Carlos Estebes is the founder of Ehxioz (http://ehxioz.com/), a Los Angeles-based
software development startup that specializes in developing modern web applications and
utilizing the latest web development technologies and methodologies. He has over 10 years
of web development experience and holds a B.Sc. in Computer Science from California State
University, Los Angeles.
He previously collaborated with Packt Publishing as a technical reviewer in the third edition
of Learning jQuery.
Olivier Pons is a developer who's been building websites since 1997. He's a teacher at the
University of Sciences (IUT) of Aix-en-Provence, France and Ecole d'Ingénieurs des Mines de
Gardanne, where he teaches PHP, jQuery, jQuery Mobile, C++, Java fundamentals, advanced
VIM techniques, and Eclipse environment. He has already done some technical reviews,
including one for Ext JS 4 First Look, Packt Publishing. In 2011, he left a full time job as a Delphi
and PHP developer to concentrate on his own company, HQF Development (http://hqf.fr).
He currently runs a number of websites, including http://www.livrepizzas.fr,
http://www.papdevis.fr, and http://olivierpons.fr, his own web development
blog. He works as a consultant, project manager, and sometimes as a developer.
Hajan Selmani is a Microsoft MVP in ASP.NET/IIS, a Microsoft Certified Trainer (MCT),
and a Microsoft Certified Professional (MCP) with deep knowledge and experience in web
development technologies. He works as a coordinator of CodeCademy at Seavus Education
and Development Center, software development consultant, technology advisor in a
few startups, and also focuses on research and innovation using cutting-edge tools and
technologies. He is a regular speaker at code camps, Microsoft conferences, and local user
group meetings and events where he actively promotes the latest modern web standards
and development practices. He is a board member of MKDOT.NET and leader of MK Web
User Group. He holds a Master's degree in Computer Science, Intelligent Systems.
He has also reviewed the books ASP.NET jQuery Cookbook, Packt Publishing and Entity
Framework 4.1: Expert's Cookbook, Packt Publishing.
Christopher Stephen Sidell is a college student attending UMBC. He has been doing
freelance web development work for people for the past few years, starting with HTML
and CSS in 2005, JavaScript in 2007, and later adding PHP into the mix in 2008. Since then
he has been working with others to create portfolio sites and self-made projects. He is selfemployed and a full-time student at UMBC.
I'd like to thank my parents for giving me a How-To guide for HTML when I
was in grade school. I would like to further thank my father who helped me
with any technology problems I may have had and taught me to fix them
myself. My mother has given me vital support in the decisions I've made in
my life.
www.PacktPub.com
Support files, eBooks, discount offers and more
You might want to visit www.PacktPub.com for support files and downloads related to
your book.
Did you know that Packt offers eBook versions of every book published, with PDF and ePub
files available? You can upgrade to the eBook version at www.PacktPub.com and as a print
book customer, you are entitled to a discount on the eBook copy. Get in touch with us at
[email protected] for more details.
At www.PacktPub.com, you can also read a collection of free technical articles, sign up
for a range of free newsletters and receive exclusive discounts and offers on Packt books
and eBooks.
http://PacktLib.PacktPub.com
Do you need instant solutions to your IT questions? PacktLib is Packt's online digital book
library. Here, you can access, read and search across Packt's entire library of books. Why Subscribe?
ff
Fully searchable across every book published by Packt
ff
Copy and paste, print and bookmark content
ff
On demand and accessible via web browser
Free Access for Packt account holders
If you have an account with Packt at www.PacktPub.com, you can use this to access
PacktLib today and view nine entirely free books. Simply use your login credentials for
immediate access.
I'd like to dedicate this book to my father. Cheers pop!
Table of Contents
Preface1
Project One: Sliding Puzzle
9
Mission Briefing
Laying down the underlying HTML
Creating a code wrapper and defining variables
Splitting an image into pieces
Shuffling the puzzle pieces
Making the puzzle pieces draggable
Starting and stopping the timer
Determining if the puzzle has been solved
Remembering best times and adding some final styling
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Project Two: A Fixed Position Sidebar with Animated Scrolling
Mission Briefing
Building a suitable demo page
Storing the initial position of the fixed element
Detecting when the page has scrolled
Handling browser window resizes
Automating scrolling
Restoring the browser's back button
Handling the hash fragment on page load
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
9
12
15
18
20
26
31
34
35
38
39
41
41
43
46
48
49
50
53
54
55
55
Table of Contents
Project Three: An Interactive Google Map
57
Project Four: A jQuery Mobile Single-page App
83
Mission Briefing
Creating the page and interface
Initializing the map
Showing the company HQ with a custom overlay
Capturing clicks on the map
Updating the UI with the start and end locations
Handling marker repositions
Factoring in weights
Displaying the projected distance and cost
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Mission Briefing
Building the welcome page
Adding a second page
Creating the script wrapper
Getting some bounties
Adding a JsRender template
Building the list view
Building an item view
Handling paging
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Project Five: jQuery File Uploader
Mission Briefing
Creating the page and plugin wrapper
Generating the underlying markup
Adding event handlers for receiving files to upload
Displaying the list of selected files
Removing files from the upload list
Adding a jQuery UI progress indicator
Uploading the selected files
Reporting success and tidying up
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
ii
57
60
61
64
67
70
72
74
76
81
81
83
86
89
92
96
98
100
108
115
116
116
117
117
119
123
126
129
135
137
139
142
145
145
Table of Contents
Project Six: Extending Chrome with jQuery
147
Project Seven: Build Your Own jQuery
173
Project Eight: Infinite Scrolling with jQuery
191
Project Nine: A jQuery Heat Map
209
Mission Briefing
Setting up the basic extension structure
Adding a manifest and installing the extension
Adding a sandboxed JsRender template
Posting a message to the sandbox
Adding a content script
Scraping the page for microdata
Adding a mechanism for saving the microdata
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Mission Briefing
Installing Git and Make
Installing Node.js
Installing Grunt.js
Configuring the environment
Building a custom jQuery
Running unit tests with QUnit
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Mission Briefing
Preparing the underlying page
Getting the initial feeds
Displaying the initial set of results
Handling scrolling to the bottom of the page
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
Mission Briefing
Determining and saving the environment
Capturing visitor clicks
Saving the click data
Adding the management console
Requesting click data
147
149
152
154
158
160
162
165
171
171
173
174
178
179
181
184
185
189
189
191
193
198
201
204
207
208
209
212
216
218
220
223
iii
Displaying a heat map
Allowing different layouts to be selected
Showing heat maps for each layout
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
226
229
233
237
237
Project Ten: A Sortable, Paged Table with Knockout.js
239
Index
269
Mission Briefing
Rendering the initial table
Sorting the table
Setting the page size
Adding Previous and Next Links
Adding numerical page links
Managing class names
Resetting the page
Filtering the table
Mission Accomplished
You Ready To Go Gung HO? A Hotshot Challenge
239
241
244
251
254
256
258
260
261
266
267
Preface
Welcome to jQuery Hotshot. This book has been written to provide as much exposure to
the different methods and utilities that make up jQuery as possible. You don't need to be
a jQuery hotshot to read and understand the projects this book contains, but by the time
you've finished the book, you should be a jQuery hotshot.
As well as learning how to use jQuery, we are also going to look at a wide range of related
technologies including using some of the more recent HTML5 and related APIs, such as
localStorage, how to use and create jQuery plugins, and how to use other jQuery libraries
such as jQuery UI, jQuery Mobile, and jQuery templates.
jQuery has been changing the way we write JavaScript for many years. It wasn't the first
JavaScript library to gain popularity and widespread usage among developers, but its powerful
selector engine, cross-browser compatibility, and easy-to-use syntax quickly propelled it to be
one of the most popular and widely-used JavaScript frameworks of all time.
As well as being easy-to-use and abstracting complex and powerful techniques into a simple
API, jQuery is also backed by an ever-growing community of developers, and is possibly the
only JavaScript library protected by a not-for-profit foundation to ensure that development
of the library remains active, and that it remains open source and free for everyone for as
long as it's available.
One of the best things is that anyone can get involved. You can write plugins for other
developers to use in order to complete common or not-so-common tasks. You can work
with the bug tracker to raise new issues, or work with the source to add features, or fix bugs
and give back in the form of pull requests through Git. In short, there is something to do for
everyone who wants to get involved, whatever their background or skillset.
Preface
Getting started with jQuery
Every project in this book is built around jQuery; it's the foundation for everything we do. To
download a copy of jQuery, we can visit the jQuery site at http://jquery.com/. There are
download buttons here to obtain production and development versions of the library, as well
as a wealth of other resources including full API documentation, tutorials, and much, much
more to help you familiarize yourself with using the library.
One of the core concepts of jQuery is based on selecting one or more elements from the
Document Object Model (DOM) of a web page, and then operating on those elements
somehow using the methods exposed by the library.
We'll look at a range of different ways of selecting elements from the page throughout the
projects in the book, as well as a wide selection of the different methods we can call on
elements, but let's look at a basic example now.
Let's say there is an element on a page that has an id attribute of myElement. We can select
this element using its id with the following code:
jQuery("#myElement");
As you can see, we use simple CSS selectors in order to select the elements from the page
that we wish to work with. These can range from simple id selectors as in this example,
class selectors, or much more complex attribute selectors.
As well as using jQuery to select elements, it is also common to use the $ alias. This would
be written using $ instead of jQuery, as follows:
$("#myElement");
Once the element has been selected in this way, we would say that the element is wrapped
with jQuery, or that it's a jQuery object containing the element. Using the jQuery (or $)
method with a selector always results in a collection of elements being returned.
If there are no elements that match the selector, the collection has a length of 0. When id
selectors are used, we would expect the collection to contain a single element. There is
no limit as to how many elements may be returned in the collection; it all depends on the
selector used.
We can now call jQuery methods that operate on the element or elements that have
been selected. One of the great features of most jQuery methods is that the same
method may be used to either get a value, or set a value, depending on the arguments
passed to the method.
2
Preface
So to continue our example where we have selected the element whose id attribute
is myElement, if we wanted to find out its width in pixels, we could use jQuery's
width() method:
$("#myElement").width();
This will return a number which specifies how many pixels wide the element is. However, if
we wish to set the width of our element, we could pass the number of pixels that we'd like
the element to have its width set to as an argument to the same method:
$("#myElement").width(500);
Of course, there is much more to using jQuery than these simple examples show, and we'll
explore much more in the projects contained in this book, but this simplicity is at the heart
of the library and is one of the things that have made it so popular.
What this book covers
Project 1, Sliding Puzzle, helps us build a sliding puzzle game. We'll use jQuery and jQuery UI
together to produce this fun application and also look at the localStorage API.
Project 2, A Fixed Position Sidebar with Animated Scrolling, helps us implement a popular
user interface feature – the fixed-position sidebar. We focus on working with the CSS of
elements, animation, and event handling.
Project 3, An Interactive Google Map, teaches us how to work with Google's extensive Maps
API in order to create an interactive map. We look at a range of DOM manipulation methods
and look at how to use jQuery alongside other frameworks.
Project 4, A jQuery Mobile Single-page App, takes a look at the excellent jQuery Mobile
framework in order to build a mobile application that combines jQuery with the Stack
Exchange API. We also look at jQuery's official template engine, JsRender.
Project 5, jQuery File Uploader, uses jQuery UI once again, this time implementing a
Progressbar widget as part of a dynamic front-end file uploader. We also cover writing jQuery
plugins by making our uploader a configurable jQuery plugin.
Project 6, Extending Chrome with jQuery, shows us how we can extend the popular Chrome
web browser with an extension built with jQuery, HTML, and CSS. Once again we make use
of JsRender.
Project 7, Build Your Own jQuery, takes a look at how we can build a custom version of
jQuery using a range of key web developer's tools including Node.js, Grunt.js, Git, and QUnit.
3
Preface
Project 8, Infinite Scrolling with jQuery, takes a look at another popular user-interface feature
– infinite scrolling. We focus on jQuery's AJAX capabilities, again use JsRender, and look at
the handy imagesLoaded plugin.
Project 9, A jQuery Heat Map, helps us build a jQuery-powered heat map. There are several
aspects to this project including the code that captures clicks when pages are visited, and the
admin console that aggregates and displays the information to the site administrator.
Project 10, A Sortable, Paged Table with Knockout.js, shows us how to build dynamic
applications that keep a user interface in sync with data using jQuery together with the
MVVM framework Knockout.js.
What you need for this book
Some of the projects covered in this book can be completed using nothing but a browser and
a simple text editor. Of course, a complete IDE is always going to make things easier, with
features such as code completion, code coloring, and collapsible blocks. So using an IDE over
a simple text editor is recommended.
Other projects rely on additional JavaScript frameworks or community-built plugins. Several
projects use third-party services hosted on the Internet in order to consume data. One
project requires the use of several additional and highly specialized applications.
Where additional software or scripts are required, or API access is needed, these
requirements are discussed in the relevant projects and information is included on
where to obtain the required code or applications, how to install them, and how to
use them sufficiently for the project to be completed.
Who this book is for
This book is aimed primarily at front-end developers that have some knowledge and
understanding of HTML, CSS, and JavaScript. Some jQuery experience is desired, but not
essential. All code, whether it be HTML, CSS, or JavaScript (including jQuery) is discussed in
full to explain how it is used to complete the project.
4
Preface
Conventions
In this book, you will find several headings appearing frequently.
To give clear instructions of how to complete a procedure or task, we use:
Mission Briefing
This section explains what you will build, with a screenshot of the completed project.
Why Is It Awesome?
This section explains why the project is cool, unique, exciting, and interesting. It describes
what advantage the project will give you.
Your Hotshot Objectives
This section explains the major tasks required to complete your project.
ff
Task 1
ff
Task 2
ff
Task 3
ff
Task 4, and so on
Mission Checklist
This section explains any prerequisites for the project, such as resources or libraries that
need to be downloaded, and so on.
Task 1
This section explains the task that you will perform.
Prepare for Lift Off
This section explains any preliminary work that you may need to do before beginning work
on the task.
Engage Thrusters
This section lists the steps required in order to complete the task.
5
Preface
Objective Complete - Mini Debriefing
This section explains how the steps performed in the previous section allow us to complete
the task. This section is mandatory.
Classified Intel
The extra information in this section is relevant to the task.
You will also find a number of styles of text that distinguish between different kinds of
information. Here are some examples of these styles, and an explanation of their meaning.
Code words in text are shown as follows: "First of all we define a new variable called
correctPieces and set its value to 0."
A block of code is set as follows:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<title></title>
<link rel="stylesheet" href="css/common.css" />
</head>
<body>
<script src="js/jquery-1.9.0.min.js"></script>
</body>
</html>
Two independent lines of code will appear as follows:
<div data-role="header">
<a href="bounty-hunter.html" data-icon="home"
A line of code that has overflown to the next due to space constraints would appear
as follows:
filter: "!)4k2jB7EKv1OvDDyMLKT2zyrACssKmSCX
eX5DeyrzmOdRu8sC5L8d7X3ZpseW5o_nLvVAFfUSf"
When we wish to draw your attention to a particular part of a code block, the relevant lines
or items are set in bold:
pieces.appendTo(imgContainer).draggable("destroy");
if (timer) {
clearInterval(timer);
6
Preface
timerDisplay.text("00:00:00");
}
timer = setInterval(updateTime, 1000);
currentTime.seconds = 0;
currentTime.minutes = 0;
currentTime.hours = 0;
pieces.draggable({
Any command-line input or output is written as follows:
cd C:\\msysgit\\msysgit\\share\\msysGit
New terms and important words are shown in bold. Words that you see on the screen, in
menus or dialog boxes for example, appear in the text like this: "Clicking on the Next button
moves you to the next screen".
Warnings or important notes appear in a box like this.
Tips and tricks appear like this.
Reader feedback
Feedback from our readers is always welcome. Let us know what you think about this book—
what you liked or may have disliked. Reader feedback is important for us to develop titles
that you really get the most out of.
To send us general feedback, simply send an e-mail to [email protected], and
mention the book title via the subject of your message.
If there is a topic that you have expertise in and you are interested in either writing or
contributing to a book, see our author guide on www.packtpub.com/authors.
Customer support
Now that you are the proud owner of a Packt book, we have a number of things to help you
to get the most from your purchase.
7
Preface
Downloading the example code
You can download the example code files for all Packt books you have purchased from your
account at http://www.packtpub.com. If you purchased this book elsewhere, you can
visit http://www.packtpub.com/support and register to have the files e-mailed directly
to you.
Errata
Although we have taken every care to ensure the accuracy of our content, mistakes do
happen. If you find a mistake in one of our books—maybe a mistake in the text or the
code—we would be grateful if you would report this to us. By doing so, you can save other
readers from frustration and help us improve subsequent versions of this book. If you find
any errata, please report them by visiting http://www.packtpub.com/submit-errata,
selecting your book, clicking on the errata submission form link, and entering the details
of your errata. Once your errata are verified, your submission will be accepted and the
errata will be uploaded on our website, or added to any list of existing errata, under the
Errata section of that title. Any existing errata can be viewed by selecting your title from
http://www.packtpub.com/support.
Piracy
Piracy of copyright material on the Internet is an ongoing problem across all media. At Packt,
we take the protection of our copyright and licenses very seriously. If you come across any
illegal copies of our works, in any form, on the Internet, please provide us with the location
address or website name immediately so that we can pursue a remedy.
Please contact us at [email protected] with a link to the suspected pirated material.
We appreciate your help in protecting our authors, and our ability to bring you valuable content.
Questions
You can contact us at [email protected] if you are having a problem with any
aspect of the book, and we will do our best to address it.
8
Project 1
Sliding Puzzle
In our first project we'll get to see a variety of techniques in action, in a fun and relaxed
setting. Consider this a gentle warm up for the rest of the book.
We'll see how to make elements draggable using jQuery UI and how we can configure the
behavior of the draggable elements. We'll also look at other subjects including sorting
algorithms, and client-side storage using the localStorage API.
Mission Briefing
In this project we'll be building a simple but fun puzzle game in which a picture is scrambled
and has to be unscrambled back to the original picture by sliding the different pieces around
the board – a modern web-based take on a classic game from yesteryear.
Typically there is one blank space on the board and pieces can only be moved into this blank
space so we will need to build a system that keeps track of where the blank space is, and
only allows pieces directly adjacent to it to be dragged.
Sliding Puzzle
To give the player an incentive, we can also look at keeping track of how long it takes the
player to solve the puzzle so that the player's best time can be recorded. The following is a
screenshot that shows the final result of this project:
Why Is It Awesome?
Games are fun and they can keep people coming back to your site, especially a younger
audience. Non-flash browser-based games are taking off in a big way, but getting into the
action at the top end of the scale can have a steep learning curve.
A simple drag-based game like this is the perfect way to ease yourself into the gaming market
without jumping straight in at the deep end, allowing you to hone your skills with some of
the simpler concepts of game development.
10
Project 1
This is also a great way to learn how to build a draggable interface in a precise and engaging
format that is well suited to its intended purpose and intuitive to use. We can also look at
some more advanced draggable concepts such as collision avoidance and precise positioning.
We will also be learning how to interact with the localStorage API in order to store and
retrieve data between sessions.
Your Hotshot Objectives
This project will be broken down into the following tasks, which we'll work through
sequentially in order to produce a working end result:
ff
Laying down the underlying HTML
ff
Creating a code wrapper and defining variables
ff
Splitting an image into pieces
ff
Shuffling the puzzle pieces
ff
Making the puzzle pieces draggable
ff
Starting and stopping the timer
ff
Determining if the puzzle has been solved
ff
Remembering best times and adding some final styling
Mission Checklist
As well as jQuery, we'll also be using jQuery UI in this project, so now is the time to grab these
libraries and put them in place. We can also take a moment to set up our project folder, which
is where we can store all of the files that we'll create over the course of the book.
Create a new folder somewhere called jquery-hotshots. Within this folder create three
new folders called js, css, and img. All the HTML pages we create will go into the root
jquery-hotshots folder, while the other files we use will be distributed amongst the
subfolders according to their type.
For the projects covered throughout the book we'll use a local copy of the latest version
of jQuery, which at the time of writing is the brand new 1.9.0. Download a copy of the
minified version from http://code.jquery.com/jquery-1.9.0.min.js and save
it in the js folder.
11
Sliding Puzzle
It's considered best practice to use Google's content delivery network (CDN)
to load jQuery and to link to the file without specifying a protocol. Using a
CDN means the file is more likely to be in the visitor's browser cache, making
the library much quicker to load.
It is also advisable to provide a fallback in the event that the CDN is not
accessible for some reason. We can very easily use the excellent yepnope
to load a local version of the script if the CDN version is not found. See the
yepnope site, http://yepnopejs.com/, for more information on this
and other resource-loading tips and tricks.
To download the jQuery UI components we'll require, visit the download builder at
http://jqueryui.com/. We'll be using various other components in later projects,
so for simplicity we can just download the complete library using the Stable button.
The current version at the time of writing is 1.10.0.
Once the build has been downloaded, you'll need to grab a copy of the jquery-uix.x.x.custom.min.js file (where x.x.x is the version number) from the js
directory inside the archive, and paste it into your js folder.
Recent versions of jQuery UI, as well as some of the more
popular pre-built themes generated with Themeroller, are
also available via Google's CDN.
Laying down the underlying HTML
First of all we need to build out the page that'll contain our sliding puzzle. The initial page
will be a shell with mostly just a few containers; the draggable elements that will make up
the individual pieces of the puzzle can all be created dynamically when required.
Prepare for Lift Off
We'll use a standard starting point for all of the different projects throughout this book, so
we'll look at this briefly now to save showing it in every project:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<title></title>
<link rel="stylesheet" href="css/common.css" />
</head>
12
Project 1
<body>
<script src="js/jquery-1.9.0.min.js"></script>
</body>
</html>
Downloading the example code
You can download the example code files for all Packt books you have
purchased from your account at http://www.packtpub.com.
If you purchased this book elsewhere, you can visit http://www.
packtpub.com/support and register to have the files e-mailed
directly to you.
Each project we cover will be contained in a page that starts out exactly like this. Save a copy
of the previous file now in your local project folder and call it template.html. At the start
of each project I'll say something like "save a copy of the template file as project-name.
html". This is the file I'll be referring to.
So in this case, save a copy of the previous HTML (or template.html if you wish) in the
main project directory (jquery-hotshots) and call it sliding-puzzle.html.
We'll also be using a common style sheet for basic styling that each project will utilize. It
contains things such as an HTML5 reset, clearfix, and other utilities, as well as some basic
typographical fixtures and theming for consistency between projects. While I won't go into
detail on that here, you can take a look at the common.css source file in the accompanying
download of this book for more information.
Each project will also need its own style sheet. These will be covered where applicable and
will be discussed on a per-project basis as and when required. We can create the custom
style sheet we'll be using in this project now.
Create a new file and call it sliding-puzzle.css, then save it in the css folder. We can
link to this file in the <head> of the page using the following code:
<link rel="stylesheet" href="css/sliding-puzzle.css" />
This should appear directly after the common.css style sheet reference.
We can also link to the script files that we'll be using in this project. First, the jQuery UI file
we downloaded and copied into the js folder can be linked to using the following code:
<script src="js/jquery-ui-1.10.0.custom.min.js"></script>
Remember to always add the script for jQuery UI on the next line after the script for
jQuery itself.
13
Sliding Puzzle
Lastly we can add the script file we'll use for this project. Create a new file and save it as
sliding-puzzle.js in the js folder. We can link to it by adding the following <script>
element directly after the jQuery UI reference:
<script src="js/sliding-puzzle.js"></script>
Engage Thrusters
Save a copy of the template file as sliding-puzzle.html in the root project folder
and then add the following mark-up to the <body> element (before the jQuery
<script> element):
<div id="puzzle" class="clearfix">
<figure>
<img src="img/space-girl-vera.jpg" />
</figure>
<div id="ui">
<p id="time">Current time: <span>00:00:00</span></p>
<button id="start">Start!</button>
</div>
</div>
Objective Complete - Mini Debriefing
This simple HTML is all that's required to start with. As this is a book about JavaScript, I won't
cover the HTML in much detail unless absolutely critical to the project at hand. In this case
most of the elements themselves aren't significant.
The main thing is that we have a series of containers with id attributes that make selecting
them fast and easy. The only really important element is the <img>, which displays the
original image that we'll be turning into the puzzle.
The awesome image used in this example was created by the extremely
talented Mr. Eamon O'Donoghue. More examples of his fine work can be
seen at http://eamonart.com/. The image used in this project can
be found at http://eamonart.com/IMAGES/PINUPSLINKS/
Space%20Girl%20Vera.html.
14
Project 1
Creating a code wrapper and defining
variables
All of our code will need to be contained within a wrapper function that is executed once the
page has finished loading.
Prepare for Lift Off
The steps that we'll complete in this part of the project are as follows:
ff
Add a wrapper function for our code that will execute as soon as the page has
finished loading
ff
Define the variables that we'll use throughout the script
Engage Thrusters
The first step is to create a wrapper function for our code that will be executed as soon as
the page has loaded. Add the following code to a new script file called sliding-puzzle.
js, which should be saved in the js directory we created earlier:
$(function () {
//all our code will be in here...
});
Most jQuery code that we see in the wild resides within some kind of wrapper like this. Using
$(function(){}); is a shortcut to jQuery's document.ready function, which is fired
once the DOM for the page has loaded.
Using $
We wouldn't normally use $ in the global scope like this if we were sharing
our code with other developers, as there may be other libraries on the page
also using it. Best practice is to alias the $ character within an automatically
invoked anonymous function, or an immediately invoked function expression
if you prefer. This can be done using the syntax (function($) { … }
(jQuery));.
Next we can set some variables near the top of the script file. This is so that we don't have
lots of values that we may want to change later distributed throughout the file. Organization
is one of the keys to writing maintainable code, and we should always strive
to make our code, as well as our intentions, as clear as possible.
15
Sliding Puzzle
Next add the following code inside the function we just defined, replacing the comment
shown in the previous code sample:
var numberOfPieces = 12,
aspect = "3:4",
aspectW = parseInt(aspect.split(":")[0]),
aspectH = parseInt(aspect.split(":")[1]),
container = $("#puzzle"),
imgContainer = container.find("figure"),
img = imgContainer.find("img"),
path = img.attr("src"),
piece = $("<div/>"),
pieceW = Math.floor(img.width() / aspectW),
pieceH = Math.floor(img.height() / aspectH),
idCounter = 0,
positions = [],
empty = {
top: 0,
left: 0,
bottom: pieceH,
right: pieceW
},
previous = {},
timer,
currentTime = {},
timerDisplay = container.find("#time").find("span");
These aren't all of the variables that we'll use, just the majority of them. The list also
includes any variables that we'll need to use inside callback functions so that we don't
run into scope issues.
Objective Complete - Mini Debriefing
The variables we defined first are a combination of simple (primitive) values and objects
or arrays that we'll use throughout the code, and cached jQuery elements. For best
performance when using jQuery, it's best to select elements from the page and store
them in variables instead of repeatedly selecting them from the page.
Although none of our variables are directly assigned to window, and are therefore not
actually global, because we are defining them right at the top of our outermost function,
they will be visible throughout our code and we can consider them as global. This gives us
the visibility of globals, without actually cluttering the global namespace.
16
Project 1
It is best practice to define variables at the top of the function they are
scoped to because of a phenomenon known as hoisting, in which variables
defined deep inside a function, inside a for loop for example, are "hoisted"
to the top of the function in some situations, potentially causing errors that
are hard to track down.
Defining variables at the top of the function where possible is a simple way to
avoid this occurring and is considered a good practice when writing jQuery, or
JavaScript in general.
Most of the variables are quite straightforward. We store the number of puzzle pieces we'd
like to use and the aspect ratio of the image being used. It's important that the number of
pieces can be equally divided by both the width and height components of the ratio.
We split the aspect ratio into its component parts using JavaScript's split() function and
specifying the colon as the character to split on. We also use the JavaScript parseInt()
function to ensure we end up with actual numbers and not strings in the aspectW and
aspectH variables.
The next three variables are all different elements selected from the page that we need to
manipulate. Following this is a new element that we create using jQuery.
Next we calculate the width and height each piece of the puzzle will need to be sized to,
based on the width and height of the original image and the aspect ratio, and we initialize
a counter variable that we'll use to add a unique, ordered id attribute to each puzzle piece.
We also add an empty array called positions, which we'll use to store the top and left
positions of each new piece.
We'll need a way of keeping track of the empty space as the pieces are moved around the
board, so we create an object called empty and give it top, left, bottom, and right
properties so that we'll know exactly where the blank is space at any given moment. We'll
also want to keep track of the previous location of any given piece so we create an empty
object called previous that we'll populate with properties when required.
The remaining three variables are all concerned with keeping track of the time it takes to
solve the puzzle. We defined, but didn't initialize the timer variable that we'll use to store a
reference to a JavaScript setInterval()-based timer later in the script. We also created an
empty object called currentTime, which again we'll populate when required, and cached a
reference to the element that we'll use to display the current time.
17
Sliding Puzzle
Splitting an image into pieces
Our next task is to divide the image into a specified number of squares to represent the
individual pieces of the puzzle. To do this we'll create a series of smaller elements which
each show a different part of the image and which can be manipulated individually.
Prepare for Lift Off
The single step required to complete this task is to create a specified number of puzzle pieces
and give each a unique background-position and position in order to recreate the image.
Engage Thrusters
We now want to generate the different pieces that make up the puzzle. We can do this with
the following code, which should be added directly after the variables we just defined in
sliding-puzzle.js:
for (var x = 0, y = aspectH; x < y; x++) {
for (var a = 0, b = aspectW; a < b; a++) {
var top = pieceH * x,
left = pieceW * a;
piece.clone()
.attr("id", idCounter++)
.css({
width: pieceW,
height: pieceH,
position: "absolute",
top: top,
left: left,
backgroundImage: ["url(", path, ")"].join(""),
backgroundPosition: [
"-", pieceW * a, "px ",
"-", pieceH * x, "px"
].join("")
}).appendTo(imgContainer);
positions.push({ top: top, left: left });
}
}
18
Project 1
Objective Complete - Mini Debriefing
We used a nested set of for loops to create the new puzzle pieces in a grid pattern. The first
loop will run for as many rows as required; with a 3:4 aspect-ratio image such as that used
in this example, we will need four rows of squares. The inner loop will for run for as many
columns as required, which in this case is three.
Within the inner loop we first create two new variables top and left. We need to use these
values in a couple of places so it makes sense to create them once and reuse them each time
they're required.
The top position is equal to the height of the piece multiplied by the current value of the
outer loop's counter variable (x), while the left position is equal to the width of the piece
multiplied by the current value of the inner loop's counter variable (a). These variables are
used to make the puzzle pieces line up in a grid.
We then copy our stored <div> element using jQuery's clone() method and use the
attr() method to set a unique id attribute using the idCounter variable that we
initialized in the first part of the project. Notice that we increment the variable at the same
time as setting it directly within the attr() method.
We could increment the variable either inside the method as we have done here, or outside
of the method; there's no real difference in performance or anything else. I just feel that it's
more succinct to update it in situ.
Next we use the css() method to set a style attribute on the new element. We set the
width and height of the puzzle piece and position it using our top and left variables, as
well as set its backgroundImage and backgroundPosition style properties.
Any style properties that are usually defined using hyphenated words, such
as background-image, should be camel-cased when used with jQuery's
css() method in conjunction with an object.
The backgroundImage property can be set using our path variable and the rest of the
string components of the style, but the backgroundPosition property will need to be
calculated individually for each puzzle piece.
The horizontal component of the backgroundPosition style property is equal to the
width of the piece multiplied by the value of the inner loop's counter variable (a), while the
vertical component is equal to the height of the piece multiplied by the value of the outer
loop's counter variable (x).
Once the new element has been created we can add its position to our positions array
using JavaScript's push() method, passing in an object containing the top and left
positional properties of the element for later use.
19
Sliding Puzzle
Classified Intel
Instead of using standard string concatenation to construct the backgroundImage and
backgroundPosition strings, we put the values into an array literal and then joined the
array using JavaScript's join() method. By specifying an empty string as the value to use to
join the string, we ensure that no additional characters are added to the string.
Joining an array of substrings to form a single string is much faster than building a string
using the + operator on substrings, and as we're working repetitively inside a loop, we
should optimize the code within the loop as much as possible.
Shuffling the puzzle pieces
In this step we need to randomly shuffle the pieces to make it a puzzle so that the visitor
can unscramble them. We can also remove the original image as it's no longer required, and
remove the first piece to create an empty space so that the other pieces can be moved around.
Prepare for Lift Off
The steps we'll cover in this task are:
ff
Removing the original image from the page
ff
Removing the first piece of the puzzle
ff
Removing the first item in the positions array
ff
Shuffling the pieces randomly
Engage Thrusters
Completing the first step requires just the following line of code, which should be added directly
after the closing curly-bracket of the outer for loop we added to sliding-puzzle.js in the
last task:
img.remove();
The second step is equally as simple; the following can be added directly after the previous
line of code:
container.find("#0").remove();
20
Project 1
We can also use a single line of code for the next step. Add the following directly after the
previous line of code:
positions.shift();
Shuffling the pieces will be slightly more complex; you'll remember from the first part of the
project when we added the underlying HTML that one of the elements was a start button.
We'll use this button to trigger the shuffle. Add the following code directly after the first two
lines we just added (make sure they are still within the outer function wrapper):
$("#start").on("click", function (e) {
var pieces = imgContainer.children();
function shuffle(array) {
var i = array.length;
if (i === 0) {
return false;
}
while (--i) {
var j = Math.floor(Math.random() * (i + 1)),
tempi = array[i],
tempj = array[j];
array[i] = tempj;
array[j] = tempi;
}
}
shuffle(pieces);
$.each(pieces, function (i) {
pieces.eq(i).css(positions[i]);
});
pieces.appendTo(imgContainer);
empty.top = 0;
empty.left = 0;
container.find("#ui").find("p").not("#time").remove();
});
21
Sliding Puzzle
Objective Complete - Mini Debriefing
jQuery's remove() method is used to remove the original image element from the page,
which we already selected when we declared our variables at the start of the script. We use
the same method to remove the first puzzle piece, which we should do before the pieces
are shuffled to avoid removing a key piece, such as a face. As with the image used in this
example, an image where the main item of interest is not in the top-left corner is beneficial.
As we've removed the first piece from the board, we should also remove the first item in the
positions array. We'll use this array when we come to check whether the puzzle has been
unscrambled and as there won't be a piece at the first position, we don't need to store its
position. We use JavaScript's unshift() method to do this, which simply removes the first
item in the array it is called on.
Adding an event handler to the button using on()
We added a click event handler for the button by selecting it and calling the jQuery on()
method. The on() method takes two arguments in this example (although it can take three
when event delegation is required).
The first argument is the event to listen for and the second is the handler function to be
executed each time the event is detected. We are listening for the click event in this case.
The all-encompassing on() method
jQuery's on() method, introduced in version 1.7, replaces the
bind(), live(), and delegate() methods, which are now
deprecated. Using on() is now the recommended way of attaching
event handlers in jQuery.
Within the handler function we first define a variable which stores the children of the
<figure> element. Although we need to select the pieces from the page again, we can still
use our cached imgContainer variable to avoid creating a new jQuery object.
Shuffling the pieces
Next we define a function called shuffle(), which accepts the array to shuffle as an
argument. This function performs a Fisher-Yates shuffle, which is an established pattern for
creating a random ordering of a given set of values.
Within the function, we first get the length of the array that was passed in, and return false
(exiting the function) if the array is empty. We then use a while loop to cycle through the
array. A while loop in JavaScript is similar to a for loop but executes while the condition
specified in brackets has a truthy value (or while it evaluates to true), instead of executing
a specified number of times. A pre-decrementing loop condition is used to avoid an
unnecessary iteration of the loop once the items have all been shuffled.
22
Project 1
In JavaScript, as well as the true or false Boolean values, other types of
variables can be said to be truthy or falsey. The following values are all
considered falsey:
ff
ff
ff
ff
ff
ff
The Boolean value false
The number 0
An empty string
null
undefined
NaN
All other values are considered truthy. This is so that non-Boolean values can be
used as conditionals. The similarities between the terms falsey and false may
lead to confusion; just remember that false is an actual value, and falsey is an
aspect of a value, which values other than false also have.
For more information on this subject, see http://james.padolsey.com/
javascript/truthy-falsey/.
Within the loop, which will be executed once for each item in the array except the first
item, we want to pick a random item from the array and swap its position in the array with
another item. To generate a random number to use as the index of the item to swap, we first
generate a random number using JavaScript's Math.random() function and multiply the
random number (which will be between 0 and 1) by the length of the array plus 1. This will
give us a random number, between 0 and the length of the array.
We then pull the item with the current index out of the array, along with the item at
the randomly generated index, and swap them. It may seem complex but this is almost
universally regarded as the most efficient way to randomly shuffle the items in the array. It
gives us the most random result for the least amount of processing.
Once we have defined the function, we then invoke it, passing in the pieces array as the
array to shuffle.
For more information on the JavaScript implementation of the Fisher-Yates
shuffle, see http://sedition.com/perl/javascript-fy.html.
23
Sliding Puzzle
Positioning the pieces
Once the array of elements has been shuffled, we iterate it using jQuery's each() method.
This method is passed the array to iterate over, which in this case is the pieces array we
have just shuffled. The second argument is an iterator function that will be called for each
item in the array.
Within this function we use our positions array to put the shuffled elements in the right
place on the page. If we didn't do this, the elements would be shuffled, but would still
appear in the same place on the page because of their absolute positioning. We can use
the positions array that we updated when creating the new elements to get the correct
top and left positions for each of the shuffled elements.
Once the collection of elements have been iterated and their positions set, we then append
them back to the page using jQuery's appendTo() method. Again we can specify our
imgContainer variable as the argument to appendTo() in order to avoid selecting the
container from the page once more.
Positioning the empty space
Lastly we should make sure that the empty space is definitely at 0 top and 0 left, that is the
top-left square of the board. If the button is clicked, some pieces are moved and then the
button is clicked again, we have to ensure that the empty space is in the right place. We do
this by setting both the top and left properties of the empty object to 0.
We can also remove any previous messages that may be displayed in the UI area (we'll cover
adding these messages towards the end of this project). We don't want to remove the timer
though, so we filter this element out of the selection using jQuery's not() method, which
accepts a selector for which matching elements are discarded and therefore not removed
from the page.
24
Project 1
At this point we should be able to run the page in a browser and shuffle the pieces by
clicking on the Start! button:
25
Sliding Puzzle
Making the puzzle pieces draggable
Now it's time to kickstart jQuery UI to make the individual pieces of the puzzle draggable.
jQuery UI is a suite of jQuery plugins used to build interactive and efficient user interfaces. It
is stable, mature, and is recognized as the official, although not the only UI library for jQuery.
Prepare for Lift Off
In this task we'll cover the following steps:
ff
Making the puzzle pieces draggable using jQuery UI's Draggable component
ff
Configuring the draggables so that only pieces directly next to the empty space
can be moved
ff
Configuring the draggables so that pieces can only be moved into the empty space
Engage Thrusters
First we'll make the pieces draggable and set some of the configuration options that the
component exposes. This code should be added to sliding-puzzle.js, directly after the
code added in the previous task:
pieces.draggable({
containment: "parent",
grid: [pieceW, pieceH],
start: function (e, ui) {
},
drag: function (e, ui) {
},
stop: function (e, ui) {
}
});
The next few steps in this task will see additional code added to the start, drag, and stop
callback functions in the previous code sample.
We also need to configure the draggability so that the pieces can only be moved into the
empty space, and not over each other, and so that only pieces directly adjacent to the empty
space can be moved at all.
26
Project 1
Next add the following code in to the start callback function that we just added:
var current = getPosition(ui.helper);
if (current.left === empty.left) {
ui.helper.draggable("option", "axis", "y");
} else if (current.top === empty.top) {
ui.helper.draggable("option", "axis", "x");
} else {
ui.helper.trigger("mouseup");
return false;
}
if (current.bottom < empty.top ||
current.top > empty.bottom ||
current.left > empty.right ||
current.right < empty.left) {
ui.helper.trigger("mouseup");
return false;
}
previous.top = current.top;
previous.left = current.left;
Next, add the following code to the drag callback function:
var current = getPosition(ui.helper);
ui.helper.draggable("option", "revert", false);
if (current.top === empty.top && current.left === empty.left) {
ui.helper.trigger("mouseup");
return false;
}
if (current.top > empty.bottom ||
current.bottom < empty.top ||
current.left > empty.right ||
current.right < empty.left) {
ui.helper.trigger("mouseup")
.css({
top: previous.top,
left: previous.left
});
return false;
}
27
Sliding Puzzle
Finally, we should add the following code to the stop callback function:
var current = getPosition(ui.helper);
if (current.top === empty.top && current.left === empty.left) {
empty.top = previous.top;
empty.left = previous.left;
empty.bottom = previous.top + pieceH;
empty.right = previous.left + pieceW;
}
In each of our callbacks we've used a helper function that returns the exact position of the
current draggable. We should also add this function after the draggable()method:
function getPosition(el) {
return {
top: parseInt(el.css("top")),
bottom: parseInt(el.css("top")) + pieceH,
left: parseInt(el.css("left")),
right: parseInt(el.css("left")) + pieceW
}
}
Objective Complete - Mini Debriefing
We wrote a lot of code in that last task, so let's break it down and see what we did. We
started by making the pieces draggable using the jQuery UI draggable component. We
did this by calling the draggable() method, passing in an object literal that sets various
options that the draggable component exposes.
First we set the containment option to parent, which stops any of the pieces being
dragged out of the <figure> element that they are within. We also set the grid option,
which allows us to specify a grid of points that the piece being dragged should snap to. We
set an array as the value of this option.
The first item in this array sets the horizontal points on the grid and the second item sets the
vertical points on the grid. Setting these options gives the movement of the pieces a more
realistic and tactile experience.
The next and final three options that we set are actually callback functions that are invoked
at different points in the life-cycle of a drag. We use the start, drag, and stop callbacks.
28
Project 1
When the drag begins
The start callback will be invoked once at the very start of the drag interaction following a
mousedown event on a draggable. The stop callback will be invoked once at the very end of
a drag interaction, once a mouseup event has registered. The drag callback will fire almost
continuously while a piece is being dragged as it is invoked for every pixel the dragged
element moves.
Let's look at the start callback first. Each callback is passed two arguments by jQuery UI when
it is invoked. The first of these is the event object, which we don't require in this project, while
the second is an object containing useful properties about the current draggable.
At the beginning of the function we first get the exact position of the piece that dragging has
started on. When we call our getPosition() function, we pass in the helper property of
the ui object, which is a jQuery-wrapped reference to the underlying DOM element that has
started to be dragged.
Once we have the element's position, we first check whether the element is in the same
row as the empty space by comparing the left property of the current object (the object
returned by getPosition()) with the left property of the empty object.
If the two properties are equal, we set the axis option of the draggable to y so that it
can only move horizontally. Configuration options can be set in any jQuery UI widget or
component using the option method.
If it isn't in the same row, we check whether it is in the same column instead by comparing
the top properties of the current and empty objects. If these two properties are equal, we
instead set the axis option to x so that the piece may only move vertically.
If neither of these conditions is true, the piece cannot be adjacent to the empty space, so we
manually trigger a mouseup event to stop the drag using jQuery's trigger() method, and
also return false from the function so that our stop handler is not triggered.
We need to make sure that only squares in the same row or column as the empty space are
draggable, but we also need to make sure that any pieces that are not directly adjacent to
the empty space cannot be dragged either.
To stop any pieces not adjacent to the empty space being dragged, we just check that:
ff
The bottom of the current piece is less than the top of the empty space
ff
The top of the current piece is greater than the bottom of the empty space
ff
The left of the current piece is greater than the right of the empty space
ff
The right of the current piece is less than the left of the empty space
29
Sliding Puzzle
If any of these conditions are true, we again stop the drag by triggering a mouseup event
manually, and stop any further event handlers on the draggable being called (but only for the
current drag interaction) by returning false.
If the callback function has not returned at this point, we know we are dealing with a
draggable that is adjacent to the empty space, thereby constituting a valid drag object. We
therefore store its current position at the start of the drag for later use by setting the top
and left properties of the previous object that we initialized at the start of the project.
The position of ui.helper
The ui object passed to our callback function actually contains an object
called position, which can be used to obtain the current draggable's
position. However, because we are using the grid option, the values
contained in this object may not be granular enough for our needs.
During the drag
Next we can walk through the drag callback, which will be called every time the position of
the current draggable changes. This will occur during a mousedown event.
First of all we need to know where the piece that's being dragged is, so we call our
getPosition() helper function again.
Then we want to check whether the piece being dragged is in the empty space. If it is, we
can stop the drag in the same way that we did before – by manually triggering a mouseup
event and returning false.
During the drag, only valid pieces will be draggable because we've already filtered out any
pieces that are not directly adjacent to the empty space. However, we still need to check that
the piece being dragged is not being dragged away from the empty space. We do this in the
same way that we filtered out pieces not adjacent to the empty space in the start callback.
When the drag ends
The stop callback is the simplest of the three callbacks. We get the position of the piece that
was dragged, and if it's definitely in the empty space, we move the empty space so that it
is in the position the dragged piece was in when the drag began. Remember we stored this
information in an object called previous.
30
Project 1
Starting and stopping the timer
At this point, our game is fully functional and the puzzle can be unscrambled; however to make
it more fun we should introduce an element of competitiveness by incorporating a timer.
Prepare for Lift Off
In this task we'll need to complete the following steps:
ff
Check the timer isn't already running when the Start button is clicked
ff
Start the timer from 0
ff
Increment the timer every second
ff
Update the display on the page so that the player can see how long the current
game has taken so far
Engage Thrusters
To check whether the timer is already running when the Start button is clicked we should
add the following code directly after where we appended the shuffled pieces to the page,
and directly before the call to draggable():
pieces.appendTo(imgContainer).draggable("destroy");
if (timer) {
clearInterval(timer);
timerDisplay.text("00:00:00");
}
timer = setInterval(updateTime, 1000);
currentTime.seconds = 0;
currentTime.minutes = 0;
currentTime.hours = 0;
pieces.draggable({
Next we can add the function that increments the timer and updates the display. This code
should come directly after where we update currentTime.hours in the previous code:
function updateTime() {
if (currentTime.hours === 23 && currentTime.minutes === 59 &&
currentTime.seconds === 59) {
31
Sliding Puzzle
clearInterval(timer);
} else if (currentTime.minutes === 59 && currentTime.seconds ===
59) {
currentTime.hours++;
currentTime.minutes = 0;
currentTime.seconds = 0;
} else if (currentTime.seconds === 59) {
currentTime.minutes++;
currentTime.seconds = 0;
} else {
currentTime.seconds++;
}
newHours = (currentTime.hours <= 9) ? "0" + currentTime.hours :
currentTime.hours;
newMins = (currentTime.minutes <= 9) ? "0" + currentTime.minutes :
currentTime.minutes;
newSecs = (currentTime.seconds <= 9) ? "0" + currentTime.seconds :
currentTime.seconds;
timerDisplay.text([
newHours, ":", newMins, ":", newSecs
].join(""));
}
Objective Complete - Mini Debriefing
The first thing we have to do in this task is check whether a timer is already running. The
timer will be stored in one of our "global" variables so we can check it easily. We use an if
statement to check whether timer contains a truthy value (see the previous information on
JavaScript's truthy and falsey values).
If it does, we know the timer is already running, so we cancel the timer using JavaScript's
clearInterval() function, passing in our timer variable as the timer to clear. We can
also reset the timer display if the timer is already running. We selected the timer display
element from the page and cached it when we initially declared our variables at the start of
this project.
Next we start the timer using JavaScript's setInterval() method and assign it to our
timer variable. When the timer begins this variable will contain the ID of the timer, not the
value of the timer, which is how clearInterval() knows which timer to clear.
32
Project 1
The setInterval() function accepts a function to execute after the specified interval as
the first argument, and the interval as the second argument. We specify 1000 milliseconds
as the interval, which is equal to 1 second, so the function passed as the first argument will
be called every second until the timer is cleared.
Once the timer has started we can also reset the values stored in the object we'll use to keep
track of the timer – the currentTime object. We set the seconds, minutes, and hours
properties of this object to 0. We need an object to keep track of the time because the
timer variable itself just contains the ID of the timer.
Next we added the updateTime() function that will be called by our interval every second.
All we do in this function is update the relevant properties of the currentTime object, and
update the display. We use an if conditional to check which parts of the timer to update.
We first check that the timer has not reached 24 hours. I would hope that no one would
actually spend that long playing the game, but if the browser is left open for some reason
for this length of time, we don't want the time display to say, for example, 24 hours and
1 minute, because at that point, we really should update the display to say 1 day, 0 hours,
and 1 minute. But we aren't bothering with days so instead we just stop the timer.
If the timer has not reached this length of time we then check whether the current minutes
equal 59 and the current seconds equal 59. If they do we need to increment currentTime.
hours by 1 and reset the currentTime.minutes and currentTime.seconds properties
back to 0.
If this check fails we then check whether the seconds equal 59. If they do, we increment the
currentTime.minutes property and then reset currentTime.seconds back to 0. If this
second test also fails we know that all we have to do is increment currentTime.seconds.
Next we need to check whether we need to pad any of the time components with a leading
0. We could use another if else conditional for this, but the JavaScript ternary construct is
neater and more compact so we use this instead.
First we test whether currentTime.hours is equal to or less than 9 and if so we add 0 to the
start of the value. We do the same for currentTime.minutes and currentTime.seconds.
Finally, we build the string which we will use to update the timer display. Instead of using
boring and slow string concatenation, we again use an awesome array comprising the
various parts of the display and then join the array.
The resulting string is set as the value of the <span> element contained in the timerDisplay
variable and the element on the page is updated using jQuery's text() method.
At this point we can now click on the button to shuffle the puzzle pieces, and watch as the
timer starts to increment.
33
Sliding Puzzle
Determining if the puzzle has been
solved
In this task we'll focus on determining whether the pieces have been put back into their
correct locations, unscrambling and therefore solving the puzzle.
Prepare for Lift Off
The following steps will be covered in this task:
ff
Checking the order of pieces to see if they match the starting order of the pieces
ff
Stopping the timer
ff
Displaying a congratulatory message
Engage Thrusters
First of all we need to decide when we should check whether the puzzle has been
completed. A good place to do the check would be on the stop event of the drag.
First add the following new variable directly after the existing current variable at the top of
the stop() callback:
var current = getPosition(ui.helper),
correctPieces = 0;
Don't forget to add a trailing comma after the first variable, as shown in the previous code
sample. Next add the following code directly after the if statement:
$.each(positions, function (i) {
var currentPiece = $("#" + (i + 1)),
currentPosition = getPosition(currentPiece);
if (positions[i].top === currentPosition.top && positions[i].left
=== currentPosition.left) {
correctPieces++;
}
});
if (correctPieces === positions.length) {
clearInterval(timer);
$("<p/>", {
text: "Congratulations, you solved the puzzle!"
}).appendTo("#ui");
}
34
Project 1
Objective Complete - Mini Debriefing
First of all we defined a new variable called correctPieces and set its value to 0. We then
used jQuery's each() method to iterate the positions array that we populated much
earlier in the code, when we initially shuffled the pieces.
What we need to do at this point is get each piece from the puzzle and check whether the
pieces are in the correct order. However, we can't just select the elements from the page
using jQuery's children() method, for example, or find(), because jQuery does not
return the elements in the order that they are found in the DOM, especially as we have
already dragged them all around their parent container.
What we have to do instead is select each element by its id attribute, and check to see what
top and left CSS properties it has in its style attribute. The length of the positions
array is the same as the number of pieces so we can iterate this array and use the index
argument that jQuery automatically passes to the iterator function.
Within the iterator we first select the current element. The id attributes for each piece will
start at 1 instead of 0 because we already removed the first piece from the puzzle, so we add
1 to the index value when selecting each piece. We also get the position of the current element
using our existing getPosition() function, passing in the element we just selected.
Next we compare the current piece's top and left properties with the equivalent item
from the positions array, and if both the top and left properties match, we increment
the correctPieces variable.
Once each piece from the page and each item in the positions array have been compared
and the each() method has finished iterating, we then check whether the value of the
correctPieces variable is equal to the length of the positions array. If it is, we know
that each piece is in the correct place.
We can stop the timer at this point in the same way that we did before – using the
clearInterval() function, and then create the congratulatory message and append it to
the element with an id of ui.
Remembering best times and adding
some final styling
The game is now pretty playable as it stands. We can shuffle the pieces, only allow them
to be dragged according to the rules, and the game will detect when the puzzle has been
solved. Using a simple timer we can tell the player how long it took for them to solve it, but
then what? What is the player supposed to do, just remember his/her best score?
35
Sliding Puzzle
Of course, we now need a way to save the player's best time. It would also be handy if
we could display an additional message if they beat their stored best time. We'll use the
JavaScript localStorage API to store the best time.
We can also add a little extra styling to finish the appearance of the game and lay out the
different elements a little better.
Prepare for Lift Off
The steps that we'll cover in this task are as follows:
ff
Checking whether a best time has been saved
ff
Checking whether the current best time is better than the saved best time
ff
Updating the saved best time when the current best time is better than it
ff
Displaying an additional message when the saved best time is beaten
ff
Tidying up the presentation of the game with CSS
Engage Thrusters
Everything we need to do in this task can be done in the if statement that is executed once
the pieces are back in the correct order. Directly after where we displayed the congratulatory
message in the last task add the following code:
var totalSeconds = (currentTime.hours * 60 * 60) +
(currentTime.minutes * 60) + currentTime.seconds;
if (localStorage.getItem("puzzleBestTime")) {
var bestTime = localStorage.getItem("puzzleBestTime");
if (totalSeconds < bestTime) {
localStorage.setItem("puzzleBestTime", totalSeconds);
$("<p/>", {
text: "You got a new best time!"
}).appendTo("#ui");
}
} else {
localStorage.setItem("puzzleBestTime", totalSeconds);
$("<p/>", {
text: "You got a new best time!"
}).appendTo("#ui");
}
36
Project 1
We already created the style sheet that we'll use for this – sliding-puzzle.css, so we
just need to add the following selectors and style rules to this file:
#puzzle {
width:730px; padding:5px; margin:auto;
border:1px solid #aaa; border-radius:5px;
background-color:#eee;
}
#puzzle figure {
width:510px; height:676px; border:1px solid #aaa;
position:relative; float:left; background-color:#fff;
}
#ui { padding:10px 0 0 10px; float:left; }
#ui button { margin-bottom: 2em; }
#ui p { font-size:1.7em; }
#start { width:204px; height:50px; font-size:1.75em; }
Objective Complete - Mini Debriefing
First of all we convert the current time into seconds so that we have only a single value to
work with and store. The seconds are calculated using the hours, minutes, and seconds
properties of the currentTime object used to update the visible timer on the page.
The hours property is multiplied by 60 to convert to minutes, and then by 60 again to
convert to seconds. The minutes property is multiplied by 60 a single time, then these two
values are added to the seconds remaining in the seconds property to give the final total,
which we store in the totalSeconds variable.
Next we check the localStorage to see if a key exists with the name puzzleBestTime. If it
does, we store the value held in localStorage in the bestTime variable. If the value of
our totalSeconds variable is less than the bestTime variable, we have a new high score,
which we save in localStorage with the puzzleBestTime name in order to overwrite the
old best time. We then display a second congratulatory message to say a new high score has
been achieved.
If localStorage doesn't contain a key with this name, this must be the first time the game
has been played in this browser, so we set the name of the key and store the value of
the currentTime variable as the new best time, and again display the second
congratulatory message.
There's nothing really crucial in the CSS that we added; it was just a little bit of light styling to
tidy up the various elements we've used and present the game in a cleaner style.
37
Sliding Puzzle
Classified Intel
The localStorage API is one of the more stable JavaScript APIs that fall within the general
umbrella term of HTML5, and enjoys wide support by all of the latest versions of all
common browsers.
Old browsers, which we may still need to support, such as IE7 or Firefox 2, do not support
localStorage. Luckily there are plenty of polyfills and workarounds that exist to add a basic
level of support in these legacy browsers.
See https://github.com/Modernizr/Modernizr/wiki/HTML5-Cross-BrowserPolyfills for a wide range of polyfills and patches that add support for modern APIs to
legacy browsers.
Mission Accomplished
We used a wide range of jQuery and plain-vanilla JavaScript over the course of this project to
create this simple game. We also looked at using jQuery UI's draggable component as well as
the localStorage API.
We covered a lot of code so let's briefly look back at what we did.
We first declared most of the variables that we used throughout the project right at the
start of our document.ready function. It's useful to do this so that variables can be used
throughout our code without making them global in scope. For performance reasons, it's
also best to cache jQuery objects so that they can be manipulated frequently without having
to keep selecting them from the page.
We then saw how we can easily split an image of a known aspect-ratio into a number
of equally-sized pieces laid out in a grid using nothing but some nested for loops and
some simple mathematics. We also saw that using an array of substrings to create a string
instead of using string concatenation is a very easy optimization that can help speed up our
applications when long strings need to be constructed.
We then saw how to shuffle the individual pieces into a random order using an accepted
algorithm for randomizing – the Fisher-Yates shuffle. We didn't actually use jQuery at all to
do this, but don't forget that the code to produce the shuffle was executed inside an event
handler added using jQuery's on() method.
Next we looked at how to make the pieces of the puzzle draggable using jQuery UI. We
looked at some of the configurable options exposed by the component, as well as how to
react to different events generated when the pieces were dragged. Specifically, we used the
start, drag, and stop callbacks to enforce the rules of the game concerning which pieces
could be moved, and how they could be moved during game play.
38
Project 1
After this we looked at using a standard JavaScript timer to keep track of how long it took to
solve the puzzle, and how to keep the visible timer on the page updated so that the player
could see the time that has elapsed since they started.
Detecting when the puzzle was solved was also a crucial ability of the code. Our main
obstacle here was the fact that the pieces weren't selected from the page in the visible order
we could see on the screen, but this was easily overcome by selecting the pieces using their
numbered id attributes and then manually checking their CSS position.
Lastly we looked at how to keep a record of the player's best time in solving the puzzle.
localStorage is the obvious choice here, and it was a small step to check whether a score was
already stored, and then compare the current time with the stored time to see if the record
had been beaten.
You Ready To Go Gung HO? A
Hotshot Challenge
There is still much more functionality we could add to our simple game. Why not update the
game so that it has different skill levels available for the player to choose from?
All we'd need to do to achieve this would be to provide some kind of interface to allow the
visitor to select the skill level, and then think of a way in which the game could be made
more difficult.
If we assume that the game in its current format is the easiest skill level, one very simple
way to make it harder is to increase the number of pieces that the original image is split into.
Have a go at doing this yourself. Those of you with a deep understanding of mathematics
may realize that our game has another flaw – some random combinations of the pieces
will simply not be solvable. Storing or computing all of the possible combinations that are
solvable is probably beyond practical, but there is another option.
Instead of randomly shuffling the array of pieces and then writing their positions to the
board, we could instead shuffle the pieces by programmatically moving them around the
board. A puzzle shuffled according to the rules of the game by which the player is bound
would result in a solvable puzzle every time.
39
Project 2
A Fixed Position
Sidebar with
Animated Scrolling
The position:fixed CSS style adds an interesting effect that allows a targeted element
to retain its position on the screen even when the page it is on is scrolled. However, its
effectiveness is limited by the fact that no matter how deep the element is nested within
other elements, it is always fixed relative to the document as a whole.
Mission Briefing
In this project we'll create a sidebar that emulates the position:fixed CSS style but
doesn't suffer from the same limitations as a pure CSS solution. We can also add an attractive
animation to the page so that when navigation items in the sidebar are clicked, different
parts of the page are scrolled into view.
A Fixed Position Sidebar with Animated Scrolling
The following is a screenshot that shows the final result of this project:
Why Is It Awesome?
Being able to fix an element in place on the page is an incredibly popular UI design pattern
used by many large and popular websites.
Keeping the visitor's main tools or calls-to-action within reach at all times improves the user
experience of the site and can help keep your visitors happy. Making things convenient is
important, so if a visitor has to scroll down a long page, then scroll all the way up just to click
something, they will soon lose interest in the page.
This same principle is also an emerging trend on mobile devices. Actual position:fixed
styling has pretty poor support on mobile devices in general, but the idea of keeping
important tools in hand, so to speak, without requiring excessive scrolling or a change of
screen is being picked up and implemented in some of today's most well-known apps.
42
Project 2
Your Hotshot Objectives
To complete this project we will need to work through the following tasks:
ff
Building a suitable demo page
ff
Storing the initial position of the fixed element
ff
Detecting when the page has scrolled
ff
Handling browser window resizes
ff
Automating scrolling
ff
Restoring the browser's back button
ff
Handling the hash fragment on page load
Building a suitable demo page
In this task we'll prepare the demo page and the other files we'll need ready for the script.
To make the benefits of this technique obvious, we'll need to use a number of extra
elements that strictly speaking aren't part of the required elements for the sidebar that we'll
be fixing in place.
The sidebar that we'll use as the focus of this example will need to sit within the structure of a
complete page, and to see the fixed position effect, the page will also need to be quite long.
We'll be using a range of HTML5 elements when building our demo page and you should be
aware that these are not supported in older versions of some browsers. If you find that you
do need to support legacy browsers, you'll need to use the html5shiv script available at
Google Code (http://code.google.com/p/html5shiv/).
Prepare for Lift Off
We should first save a new copy of the template file to the root folder of our project and call
the new file fixed-sidebar.html. We can also create a new style sheet called fixedsidebar.css, which we can save in the css folder, and a new JavaScript file called fixedsidebar.js, which should be saved to the js folder.
We can link to the new style sheet in the <head> part of the HTML page using the following
new <link> element, which should be added directly after the link to common.css:
<link rel="stylesheet" href="css/fixed-sidebar.css" />
Remember that the common.css style sheet is used to provide useful things such as a reset,
a simple typography framework, and some common layout styles in order to minimize the
CSS each project requires.
43
A Fixed Position Sidebar with Animated Scrolling
We can link to the new JavaScript file using the following new <script> element, which should
be directly after the jQuery <script> file in the <body> part of fixed-sidebar.html:
<script src="js/fixed-sidebar.js"></script>
The underlying page is now set up ready for us to add the elements that are required
for this project.
Engage Thrusters
We'll use a basic layout for our page consisting of the following elements, which should be
added to fixed-sidebar.html:
<header>
<h1>jQuery fixed sidebar example page</h1>
</header>
<div class="wrapper">
<article>
<h1>Example content area</h1>
<section id="part1">
</section>
<section id="part2">
</section>
<section id="part3">
</section>
<section id="part4">
</section>
<section id="part5">
</section>
</article>
<aside>
<h2>Important content to fix in place</h2>
<nav>
<h3>Some in page navigation</h3>
<ul>
<li><a href="#part1">Section 1</a></li>
<li><a href="#part2">Section 2</a></li>
<li><a href="#part3">Section 3</a></li>
<li><a href="#part4">Section 4</a></li>
<li><a href="#part5">Section 5</a></li>
</ul>
</nav>
</aside>
</div>
44
Project 2
These elements should be added to the <body> of the page, directly before the <script>
element that links to jQuery.
We'll also need some basic CSS for our example page in order to create the layout that this
example requires. In the fixed-sidebar.css style sheet that we created for this example,
add the following styles:
header, .wrapper { width:80%; max-width:1140px; margin:auto; }
header {
padding-bottom:2em; border-bottom:4px solid;
margin-bottom:3em;
}
header h1 { margin-top:.75em; }
article {
width:70%; padding-right:4%; border-right:4px solid;
margin-right:5%; float:left;
}
aside { width:20%; float:left; }
As before, none of this code is actually required, we're using it just to lay out the demo page
as we need to for the purposes of this example.
Objective Complete - Mini Debriefing
We've added a very simple layout to create our demo page. The HTML5 <article> is filled
with five different HTML5 <section> elements, each with their own id attributes. We'll
use these a little later in the project to allow animated navigation between them.
In the previous code sample each <section> element is empty. But if you're following
along and writing the example code as you go, you should fill each of them with a variety of
random elements in order to increase the length of the page.
None of the elements we're using in this example matter at all. The HTML5 <aside> is
the element that we'll be fixing in place, but the fact that it's an <aside> element is not
important – any element can be used with this technique.
Inside the <aside> element is an HTML5 <nav> element. As I mentioned, this will allow us
to add another cool feature later on, but again, is not essential for the basic technique to be
used. Any content can be used in the element that is to be fixed in place.
Notice also that in the CSS we don't use position:fixed anywhere at all. The reason for
this is simple. An element that has a fixed position is positioned relative to the document as
a whole, not to its parent container.
45
A Fixed Position Sidebar with Animated Scrolling
If no pixel coordinates are supplied, a fixed position element is rendered where the flow of
elements on the page dictates depending on its DOM position (although it is still technically
out of the normal flow of the page).
If we try to do this with our example layout, it ends up at the far-left of the outer .wrapper
element, because the float specified on the <article> element also removes the
<article> element from the normal document flow. This is not good.
If pixel coordinates are supplied, these are interpreted by the rendering engine to be relative to
the window, just like absolutely positioned elements. In some situations it may be acceptable
to specify pixel coordinates, but when using a liquid layout such as in this example, the
required coordinates to set the left and top style properties of the <aside> element will
vary depending on the resolution of the screen used to view the page, hence the conundrum
we face and hence the reason to use jQuery to achieve it instead of simple CSS.
Classified Intel
To save time when creating example layouts, like the one used in this project, we can use
services such as Placehold It (http://placehold.it/) for placeholder images of any
dimensions, and HTML Ipsum (http://html-ipsum.com) for a range of common HTML
elements pre-filled with Lorem Ipsum placeholder text.
Storing the initial position of the
fixed element
Before we can fix the element in place, we'll need to know where that place is. In this task
we'll obtain the current starting position of the <aside> element that we're going to be
fixing in place.
Engage Thrusters
In fixed-sidebar.js we should start with the following code:
$(function() {
});
We can cache a couple of jQuery-selected elements at the top of our function, and to store
the initial position of the fixed element, we can then add the following code within the
function we just added:
var win = $(window),
page = $("html,body"),
46
Project 2
wrapper = page.find("div.wrapper"),
article = page.find("article"),
fixedEl = page.find("aside"),
sections = page.find("section"),
initialPos = fixedEl.offset(),
width = fixedEl.width(),
percentWidth = 100 * width / wrapper.width();
Objective Complete - Mini Debriefing
We've used the same outer wrapper for our code that we used in the first project. As I
mentioned then, it's a very common way to execute code once the page has finished loading.
We'll probably use it in every project throughout the book.
We then cache references to the elements that we're going to be referring to so that we
don't have to keep selecting them from the DOM. We'll be querying these elements inside
event handlers a little later so it's much better for performance to select them from the
page once and then refer to the saved, or cached, version throughout our code, instead of
repeatedly selecting elements from the page.
We store references to the window object as we'll be attaching several event handlers to it.
We'll be scrolling the entire page a little later on and for full cross-browser compatibility we
should select and store a reference to both the <html> and <body> elements, as different
browsers use either the <html> or <body> element, so this covers all bases.
We'll need to select the element with the class name wrapper, the containing <article>,
all of the different <section> elements, and of course the <aside> element, which we'll
be working with frequently throughout the remaining code.
We also store the initial position of the fixed element so that we know the coordinates
on the page to fix the element to. We use jQuery's offset() method, which returns an
object containing top and left properties that show the current position relative to the
document, exactly what we require.
Depending on the styles applied to the surrounding elements, the width of the element
being fixed may change. To alleviate this we also store the initial width of the element using
jQuery's width() method, which returns an integer expressing the width in pixels.
Lastly, we can also compute and store the width as a percentage. We'll need to know this
later when we want to react to the browser window being resized. Working it out is easy by
multiplying 100 by the width of the fixed element, then dividing this figure by the width of
its container, which again we use jQuery's width() method to obtain. This also means that
the width of the fixed sidebar can easily be changed in just the CSS file, and the script will
continue to work.
47
A Fixed Position Sidebar with Animated Scrolling
Detecting when the page has
scrolled
Our next task is to detect when the page has been scrolled and fix the element in place when
that occurs. Detecting the scroll event is made easy for us by jQuery, as is setting the position
to fixed, because there are simple jQuery methods we can call to do these exact things.
Engage Thrusters
Add the following code to the script file directly after the variables we initialized in the
last task:
win.one("scroll", function () {
fixedEl.css({
width: width,
position: "fixed",
top: Math.round(initialPos.top),
left: Math.round(initialPos.left)
});
});
Objective Complete - Mini Debriefing
We can use jQuery's one() method to attach an event handler to the window object that we
stored in a variable. The one() method will automatically unbind the event handler as soon
as the event is detected for the first time, which is useful because we only need to set the
element to position:fixed once. In this example we are looking for the scroll event.
When the event is detected, the anonymous function we pass as the second argument
to one() will be executed. When this occurs we use jQuery's css() method to set some
style properties. We set the width of the element to counter situations where the width
of our target element increases because of float and/or margin on surrounding elements.
We set the position to fixed and also set the top and left style properties using the
initial position of the element that we stored in the initialPos variable at the start of
the project. We use JavaScript's Math.round() method in order to round the top and
left pixel positions to whole numbers, which helps to avoid any cross-browser issues with
subpixel rounding.
48
Project 2
Handling browser window resizes
At the moment, our <aside> element will be fixed in place as soon as the page scrolls,
which suits our needs while the browser remains the same size.
However, if the window is resized for some reason, the <aside> element will fall out of its
fixed position and could be lost outside of the boundaries of the viewport. In this task, we'll
fix that by adding an event handler that listens for the window's resize event.
Engage Thrusters
To maintain the fixed element's correct location relative to the rest of the page, we should
add the following code directly after the one() method that we added in the last task:
win.on("resize", function () {
if (fixedEl.css("position") === "fixed") {
var wrapperPos = wrapper.offset().left,
wrapperWidth = wrapper.width(),
fixedWidth = (wrapperWidth / 100) * percentWidth;
fixedEl.css({
width: fixedWidth,
left: wrapperPos + wrapperWidth - fixedWidth,
top: article.offset().top
});
}
});
Objective Complete - Mini Debriefing
This time we use jQuery's on() method to attach our event handler. We pass two
arguments to this method; the first is the event we are listening for, which in this task is
the window's resize event, and the second is the function that we wish to execute when
the event is detected.
We only want to reposition and resize the <aside> element if the page has already been
scrolled and the element has had its position set to fixed, so before we do anything else
we first check that this is the case.
If the element's position is set to fixed, we first determine the current left style property
of the wrapper element using the left property of the object returned by jQuery's offset()
method. We also get the wrapper element's width using jQuery's width() method.
49
A Fixed Position Sidebar with Animated Scrolling
Because our layout is liquid we also need to adjust the width of the fixed element. In the
CSS we originally set the width to 20%, so we can ensure that it stays at 20 percent of its
container by dividing the container's current width by 100 and then multiplying it by the
percentWidth variable we stored in the first task.
We then use jQuery's css() method to set the width of the fixed element and it's top
and left style properties to make sure that it stays in the correct location when the
window is resized.
Automating scrolling
At this point, we should be able to click on any of the links in the navigation menu we added
to the fixed element, and the page will jump to bring the corresponding section into view.
The fixed element will still be fixed into place.
The jump to the section is quite jarring however, so in this task we'll scroll each section into
place manually so that the jump to each section is not so sudden. We can also animate the
scroll for maximum aesthetic effect.
Engage Thrusters
For this task we should add another event handler, this time for click events on the links
in the navigation list, and then animate the page scroll to bring the selected <section>
into view.
First, we can add a general function for scrolling the page which accepts some arguments and
then performs the scroll animation using those arguments. We should define the function
using the following code directly after the one() method that we added in the last task:
function scrollPage(href, scrollAmount, updateHash) {
if (page.scrollTop() !== scrollAmount) {
page.animate({
scrollTop: scrollAmount
}, 500, function () {
if (updateHash) {
document.location.hash = href;
}
});
}
}
50
Project 2
Next, we can add a handler for click events on the navigation in our fixed element. This
should be added directly after the scrollPage() function that we just added:
page.on("click", "aside a", function (e) {
e.preventDefault();
var href = $(this).attr("href"),
target = parseInt(href.split("#part")[1]),
targetOffset = sections.eq(target - 1).offset().top;
scrollPage(href, targetOffset, true);
});
Objective Complete - Mini Debriefing
First we defined the scrollPage() function which accepts three arguments. The first is
href, the second is an integer that represents the figure that the scrollTop property of
the page will need to be animated to, and the third is a Boolean that will tell the function
whether or not to update the hash fragment in the location bar of the browser.
The first thing we do in this function is check whether the page actually needs to be scrolled.
To ensure it does, we just check that the current scroll of the page, obtained using jQuery's
scrollTop() method, is different from the amount that we wish to scroll to.
The jQuery animate() method also accepts three arguments. The first is an object where
each key is a property to animate, and each value is the value to animate it to. In this case
we want to animate the scrollTop property using the scrollAmount argument which is
passed to our function.
The second argument to the animate() method is the duration that the animation should
run for. It accepts an integer that represents the duration in milliseconds. We specify 500 so
the animation will take half a second to complete.
The third argument is a callback function that we would like executed as soon as the
animation ends. If the updateHash argument passed to our function is set to true, we can
update the location bar of the browser to show the id of the desired <section> element.
We can do this by updating the hash property of the document.location object with the
href argument passed to our scrollPage() function. This updates the location bar but
because it is just a hash fragment, it doesn't cause the page to reload.
51
A Fixed Position Sidebar with Animated Scrolling
After adding the scrollPage() function, we then added a handler for click events on
the navigation inside the fixed element. We use jQuery's on() method once again to attach
this event, but this time we pass three arguments to the method, which enables event
delegation. The handler is attached to the <body> of the page that we have already stored in
a variable.
The first argument is the event that we want to bind the handler to, which in this case is the
click event. The second argument is a selector; the on() method will filter all click events
so that only those originating from elements that match the selector will invoke the bound
handler function.
In this case we are only interested in clicks on the <a> elements in our fixed element –
<aside>. The third argument is the function to bind as the handler, which is automatically
passed the original event object by jQuery.
Within this function we first stop the browser navigating to the corresponding <section>
element, using the preventDefault() method of the event object that is passed to our
handler function.
Next, we set a variable that tells us which <section> the user would like to navigate to.
Inside our event handler function the $(this) object is scoped to the link that was clicked,
so we can easily get the required section id by getting the href attribute of the clicked link
using jQuery's attr() method. We store this in a variable called href.
We need to know where on the page the required <section> element is, which we obtain
by using JavaScript's split() method to split the string stored in the href variable that we
just set.
If we specify #part as the string to split on, the split() method will return an array
consisting of two items, where the second item is a string version of the section number
that was clicked. By wrapping this statement in JavaScript's parseInt(), we end up with an
integer. We store this integer in the target variable.
The last variable we set is the offset of the desired <section> element. To select the
correct <section> element, we can use the sections array that we stored at the start of
the project.
To pull the correct element from this array, we use jQuery's eq() method and pass it the
value that we just saved in the target variable minus 1. We need to subtract 1 because
arrays in JavaScript start at 0, but our <section> id attributes start at 1.
Once we have this information we can then call our scrollPage() function, passing in
the values we have just computed to animate the page scroll in order to bring the desired
<section> element into view.
52
Project 2
Restoring the browser's back button
At this point, we can click any of the links in the <aside> element and the page will be
smoothly scrolled to the desired location on the page. The address bar of the browser will
also be updated.
However, if the user tries to go back to a previous <section> using the back button of
his/her browser, nothing will happen. In this task we'll fix that so that the back button
works as expected, and can even use smooth scrolling when the back button is used to
go back to the previous <section>.
Engage Thrusters
We can enable the back button very easily by adding another event handler directly after the
one for click events that we just added:
win.on("hashchange", function () {
var href = document.location.hash,
target = parseInt(href.split("#part")[1]),
targetOffset = (!href) ? 0 : sections.eq(target - 1).offset().
top;
scrollPage(href, targetOffset, false);
});
Objective Complete - Mini Debriefing
We use jQuery's on() method to attach our event once again, and this time we don't need
to make use of event delegation, so we revert to the two-argument form of the method.
This time we are listening for the hashchange event, which as before is passed as the
first argument and occurs whenever the hash property of the document.location
object is changed.
In our handler function, which is passed as the second argument, we set the variables for the
different values that we need to pass to the scrollPage() function in order to perform the
scroll. We don't need to prevent the default behavior of the browser this time, and the href
variable is set using the document.location.hash property as it will be the back button
that will trigger the event, not one of the links in the <aside>.
Actually, this handler will also be triggered when one of the links is clicked, because the links
also update the hash, but the conditional check inside the scrollPage() function will
prevent unnecessary calls to jQuery's animate() method.
53
A Fixed Position Sidebar with Animated Scrolling
The target variable is computed in exactly the same way as it was in the last event handler,
but this time, the targetOffset variable needs to handle cases where there is no hash
fragment in the address bar of the browser. To handle this, we use a JavaScript ternary
construct that checks whether the target variable that we just defined has a falsey value,
which would indicate an empty string. If it does, we want to just animate the scroll back to
zero. If it doesn't, we determine the required scroll amount in the same way as we did before.
We should now be able to load the page, scroll to a part of the page by clicking on one of
the links in the <aside> element, and then scroll back to the top of the page using the
browser's back button.
Handling the hash fragment on page
load
At the moment the functionality of the browser's back button has been restored, and the
visitor can see the bookmarkable URL in the address bar.
If the page is requested with a hash fragment in it, the page will automatically jump to the
specified <section> when the page loads. In this part we'll add some code that checks the
hash property of the document.location object and if a hash is detected, it will scroll to
the corresponding part of the page smoothly.
Engage Thrusters
To enable this, we should add the following code directly after where we define our
starting variables near the top of the script file, and directly before where we listen
for the scroll event:
if (document.location.hash) {
var href = document.location.hash,
target = parseInt(href.split("#part")[1]),
targetOffset = sections.eq(target - 1).offset().top;
page.scrollTop(0);
document.location.hash = "";
scrollPage(href, targetOffset, true);
}
54
Project 2
Objective Complete - Mini Debriefing
In this bit of code, which will be executed as soon as the page has loaded, we first check
whether the document.location object contains a hash (or at least, contains a hash that
is not an empty string).
If it does, we obtain the hash, get the number of the <section>, and calculate the offset
from the top of the page in the same way that we have done in previous tasks. We then
set the scrollTop of the page to 0 to force the browser to the top of the page. We also
remove the hash at this point.
Finally we can call our scrollPage() function, passing in the new href fragment, the
amount of scroll required, and set the final argument to true so that the correct hash
fragment is added back to the browser's location bar. It should all happen so quickly that the
user does not notice that the page load has been intercepted and the behavior modified.
Mission Accomplished
In this project we looked at a very simple way of mimicking CSS's position:fixed styling
to fix an important element into place. The technique to only apply the fixed positioning
when the page starts to scroll is simple but effective, and is an excellent way to circumvent
the shortcomings of actual position:fixed when working with complex or liquid layouts.
We saw how to handle window resizes and added a smooth scrolling facility that scrolled the
page between different named sections of the page.
We also looked at how we can read and write to the document.location.hash property
of the window object, and how to manually scroll to the requested section when the page is
loaded. We also fixed the browser's back button to work with our smooth-scrolling animations.
You Ready To Go Gung HO?
A Hotshot Challenge
Very often, with the kind of in-page navigation we've used in this project, it is useful to
show an on-state on the navigation links when a section is scrolled to, either manually, or by
clicking on one of the links. Have a go at adding this simple but effective addition to the code
that we've looked at over the course of this project.
55
Project 3
An Interactive
Google Map
In this project we'll create a highly interactive Google map that works with the latest version
of Google's API to produce a map with custom overlays and markers, geocoded addresses,
and computed distances. We'll also look at how to keep our simple UI in sync with the
locations added to the map using a combination of Google and jQuery event handlers.
Mission Briefing
For the purposes of this project, we'll have a scenario where we need to build a map-based
application for a company that takes things from one place to another. They want a page
that their customers can visit to calculate the cost of, and maybe order, the transport of
something from one place to another by clicking on different areas of a localized map.
We'll see how to listen for clicks on the map so that markers can be added and the precise
locations of each marker can be recorded. We can then update the UI to show the actual
street addresses of the locations that were clicked and allow the visitor to generate a quote
based on the computed distance between the two addresses.
An Interactive Google Map
Why Is It Awesome?
Google Maps is a fantastic API to build on. Already highly interactive and packed with
features, we can build robust and highly functional applications on top of the solid
foundation it provides. Google provides the mapping data and interactivity with the map,
while jQuery is used to build the UI – a winning combination.
The page that we'll end up with will resemble the following screenshot:
Your Hotshot Objectives
This project will be broken down into the following tasks:
ff
Creating the page and interface
ff
Initializing the map
ff
Showing the company HQ with a custom overlay
ff
Capturing clicks on the map
ff
Updating the UI with the start and end locations
ff
Handling marker repositions
ff
Factoring in weights
ff
Displaying the projected distance and cost
58
Project 3
Mission Checklist
We'll need to link to a script file provided by Google in order to initialize the map and load in
the API. We can also create the new files that we'll be using in the project at this point.
Don't worry, we don't need an API key from Google or anything like that for this project to
work, we can just use the script by linking directly to it.
The Google Maps API is feature-rich and stable, and contains entry points for
all of the best known mapping features, including Street View, geolocation,
and the directions service. As well as the configuration options we used
here, there are many, many others. For further information, see the
documentation site at http://developers.google.com/maps/.
First we should save a new copy of the template file to our root project folder and call it
google-map.html. Also create a google-map.css file and a google-map.js file and
save them in the css and js folders respectively.
We can link to the style sheet for this example by adding the following <link> element to
the <head> of the page, directly after the <link> element for common.css:
<link rel="stylesheet" href="css/google-map.css" />
Don't forget, we're using common.css with each project so that
we can focus on the styles we actually need for the project, without
all of the boring reset, float-clears, and other common CSS styling
required for most web pages.
We can link to Google's script file, as well as the JavaScript file we just created, using the
following <script> elements, directly after the <script> element for jQuery:
<script src="http://maps.googleapis.com/maps/api/js?sensor=false">
</script>
<script src="js/google-map.js"></script>
We'll also be using a couple of images in this project, hq.png and start.png, which can
both be found in the accompanying code download for this book. You should copy them into
the img directory in your local jquery-hotshots project directory. Our page is now set up
ready for the first task.
59
An Interactive Google Map
Creating the page and interface
In our first task we can add the different containers for the map, and the initial UI elements
needed by the page. We can also add some basic styling to lay things out as we want.
Engage Thrusters
We should add the following elements to the <body> element in the google-map.html
page that we just set up:
<div id="map"></div>
<div id="ui">
<h1>I Am Mover</h1>
<p>Enter the weight of your goods below and click on two
different places on the map to see the distance between
them and the cost of moving your goods.</p>
<h3>Our charges</h3>
<dl class="clearfix">
<dt>Base rate (per mile)</dt>
<dd>&pound;3</dd>
<dt>Cost per kg per mile</dt>
<dd>&pound;0.25</dd>
</dl>
<input id="weight" placeholder="Weight (kg)" />
</div>
For some basic styling and to lay out the page ready for when we initialize the map, we can
add the following selectors and styles to the google-map.css file that we just created:
#map { width:100%; height:100%; }
#ui {
width:16%; height:99.8%; padding:0 2%;
border:1px solid #fff; position:absolute; top:0; right:0;
z-index:1; box-shadow:-3px 0 6px rgba(0,0,0,.5);
background-color:rgba(238,238,238,.9);
}
#ui h1 { margin-top:.5em; }
#ui input { width:100%; }
#ui dl {
width:100%; padding-bottom:.75em;
border-bottom:1px dashed #aaa; margin-bottom:2em;
}
#ui dt, #ui dd { margin-bottom:1em; float:left; }
#ui dt { width:50%; margin-right:1em; clear:both; }
#ui dd { font-weight:bold; }
60
Project 3
Objective Complete - Mini Debriefing
In this task we're just getting started by adding the underlying HTML elements that we'll
populate properly over the next few tasks. A slightly boring, but somewhat necessary, first
step in getting the example page up and running, and the project under way.
The first element we added is the container that the Google Maps API will render the map
tiles into. We give it an id of map so that it can be efficiently selected, but it is completely
empty to start with.
The next element is the container for the various UI elements the example requires. It too
has an id of ui for easy selecting from our script, as well as for adding the CSS styling with.
Styling with IDs
Avoiding the use of ID selectors to add CSS styling is well on its way to becoming
a general best practice, with tools such as CSSLint advising against its use.
While the arguments for doing this and sticking to classes, element, or attribute
selectors are compelling, we'll be working with them in some of the projects
throughout this book for simplicity.
CSSLint is an open source CSS code quality tool that performs static analysis of
source code and flag patterns that might be errors or otherwise cause problems
for the developer. See http://csslint.net/ for more information.
Within the interface container we have the name of the fictional company, some basic
instructions for using the page, a list of the different charges, and an <input> element to
allow weights to be entered.
Most of the CSS that we added in this task was purely decorative and specific to this
example. It could easily be wildly different if a different look and feel was required.
We've made the map container take up the full width and height of the page, and
styled the interface so that it appears to float over the right-hand side of the page.
Initializing the map
Getting a zoomable and panable interactive Google map up and running takes a ludicrously
small amount of code. In this task we'll add that code, as well as set up some of the variables
that we'll use later in the script.
61
An Interactive Google Map
Prepare for Lift Off
In this task we'll initialize the variables needed to configure the map and make a call to the
Google Maps API. We should start by adding the standard jQuery wrapper to the empty
google-map.js file that we created earlier:
$(function () {
//all other code in here...
});
Remember, the $(function () { … }); construct is a shortcut for jQuery's document.
ready event handler.
Engage Thrusters
Within the wrapper we just added, we should add the following code:
var api = google.maps,
mapCenter = new api.LatLng(50.91710, -1.40419),
mapOptions = {
zoom: 13,
center: mapCenter,
mapTypeId: api.MapTypeId.ROADMAP,
disableDefaultUI: true
},
map = new api.Map(document.getElementById("map"), mapOptions),
ui = $("#ui"),
clicks = 0,
positions = [];
Objective Complete - Mini Debriefing
In this task we start by creating some variables that we'll need to initialize the map. We'll be
addressing the google.maps namespace throughout our code so the first variable we set is
the contents of the top two namespaces for convenience.
Having a locally scoped copy that reaches right into the actual API that we want to use will
make our code marginally more efficient because it is easier for our code to resolve a single
variable. It's also much quicker to type in the first place.
All properties and methods used by the Google Maps API are namespaced. They all sit within
the maps namespace, which itself sits in the google namespace. Google has such a large
code-base for use in so many different applications that it makes sense to keep everything
isolated and organized using namespaces.
62
Project 3
For an excellent in-depth discussion on the intricacies of namespacing in
JavaScript, see the excellent article on the subject by JavaScript supremo
Addy Osmani (http://addyosmani.com/blog/essentialjs-namespacing/).
Next we store the latitude and longitude that we'd like to center the map on. This is done
using the Google Maps API's LatLng() method, which takes two arguments, the latitude
and longitude values, and returns a LatLng object for use with other API methods. Notice
how we call the LatLng constructor using our local api variable.
We can then create an object literal containing some of the configuration options that
our map will need. These options include the zoom level, the location the map should
be centered on, the type of map, and an option which disables the default map type
and zoom/pan controls. We can use the LatLng object contained in mapCenter for
the center configuration option.
Following this we create a new map instance using the map API's Map() constructor
function. This function accepts two arguments: the first is the DOM element that the map
should be rendered into and the second is the object literal containing the configuration
options that we wish to set.
The first argument takes an actual DOM element, not a jQuery-wrapped DOM element. So
although we could select the element from the page using jQuery and then extract the raw
DOM element, it is more efficient to use JavaScript's native getElementById() function
to retrieve the map container we added to the page in the previous task and pass it to the
Map() constructor.
Next, we cache a jQuery selector for the UI container so that we can access it from the
page repeatedly without having to actually select it from the DOM each time, and define a
variable called clicks, which we'll use to record how many times the map has been clicked.
We need to define it here in the top-level function scope so that we can reference it from
within a click handler later in the code.
Lastly, we add an empty array literal in the variable positions, which we'll populate later
on when we need to store the different areas of the map that have been clicked on. The
array needs to be in the scope of the top-level function so that we can access it from within
different event handlers later in the code.
63
An Interactive Google Map
Showing the company HQ with a
custom overlay
In this task we'll put the company HQ on the map, literally, by adding a custom marker
complete with an overlay that provides some basic information about the company, and
perhaps an image of the premises.
Prepare for Lift Off
In this task we'll cover the following subtasks:
ff
Adding a marker to the map
ff
Adding a hidden element containing information about the company
ff
Adding a custom overlay to display the company information when the new marker
is clicked
ff
Adding a click handler to show the overlay when the marker is clicked
Engage Thrusters
Adding a custom marker to the map can be achieved with the following simple code block,
which should be added directly after the variables we added in the previous task:
var homeMarker = new api.Marker({
position: mapCenter,
map: map,
icon: "img/hq.png"
});
To create an information overlay, or info window to use the correct Google terminology, for
our new marker, we should first add an HTML element that contains the content we wish to
display in the overlay. We can add the following new collection of elements to google-map.
html directly after the UI container:
<div id="hqInfo">
<img class="float-left" src="http://placehold.it/140x100"/>
<h1>I Am Mover</h1>
<p>This is where we are based.</p>
<p>Call: 0123456789</p>
<p>Email: [email protected]</p>
</div>
64
Project 3
We're using the placehold.it service again so that we don't have to
worry about sourcing or creating an actual image for this bit of example
content. It's a great service to use when mocking up prototypes quickly.
To tell the map about the new info window, we can use the following code, which should be
added directly after the homeMarker code back in google-map.js:
var infoWindow = new api.InfoWindow({
content: document.getElementById("hqInfo")
});
We also need some extra CSS to style the contents of the info window and to hide it until it is
required. Add the following code to the bottom of google-map.css:
body > #hqInfo { display:none; }
#hqInfo { width:370px; }
#hqInfo h1 { margin-bottom:.25em; line-height:.9em; }
#hqInfo p { margin-bottom:.25em; }
Finally, we can add a simple click handler that displays the info window using the following
code, which should be added after the infoWindow variable that we added a moment ago
in google-map.js:
api.event.addListener(homeMarker, "click", function(){
infoWindow.open(map, homeMarker);
});
Objective Complete - Mini Debriefing
First of all we defined a new marker, which is done using Google's Marker() constructor.
This function takes a single argument, which is an object literal that defines different
properties of the marker.
We set the position of the marker to be the center of the map for simplicity, although
when defining other markers you'll see that any LatLng object can be used. We should also
define the map that the marker belongs to, which we set to the map variable that contains
our map instance. To specify the image to use as the marker, we can supply a relative path in
string format to the icon option.
We then added a new container to the page which contains the information we want to
display in our custom info window. The content here is not important; it's the technique that
matters. We also added some additional styling for the contents of the info window.
65
An Interactive Google Map
In order to add the info window to our map instance, we used Google's InfoWindow()
constructor. This method also takes a single argument, which again is an object literal which
contains the options we wish to set. In this example we just set the content option to the
element containing the content we just added to the page.
This should be an actual DOM element, hence we use JavaScript's document.
getElementById() to get the element, instead of selecting it with jQuery.
Lastly we added an event handler to the map using Google's addListener() method. This
method takes the element to attach the event handler to, which in this case is the marker
we added, as the first argument, the event we wish to listen for as the second argument,
and the callback function to handle the event as the third argument. The signature of this
method is very similar to the event handling methods found in other common JavaScript
libraries, although it is slightly different to how events handlers are added in jQuery.
Within the anonymous function we pass as the third argument to the addListener()
method, all we do is call the open() method of our info window. The open() method takes
two arguments; the first is the map that the info window belongs to, and the second is the
location the info window is added to, which we set to our marker.
At this point we should be able to run the page in a browser, click on our custom marker,
and have the contents of our hidden <div> displayed in the info window, as shown in the
following screenshot:
66
Project 3
Capturing clicks on the map
In this task we need to add a click handler for our map so that visitors can set the start and
end of their transportation journey.
Engage Thrusters
First of all we need to add the function that will be executed when the map is clicked. Directly
after the listener that we added in the last task, add the following function expression:
var addMarker = function (e) {
if (clicks <= 1) {
positions.push(e.latLng);
var marker = new api.Marker({
map: map,
position: e.latLng,
flat: (clicks === 0) ? true : false,
animation: api.Animation.DROP,
title: (clicks === 0) ? "Start" : "End",
icon: (clicks === 0) ? "img/start.png" : "",
draggable: true,
id: (clicks === 0) ? "Start" : "End"
});
api.event.trigger(map, "locationAdd", e);
} else {
api.event.removeListener(mapClick);
return false;
}
}
Then, to attach a listener for clicks on the map which fires this function, we can add the
following code directly after it:
var mapClick = api.event.addListener(map, "click", addMarker);
67
An Interactive Google Map
Objective Complete - Mini Debriefing
First of all we added the function that will be executed every time the map is clicked. The
function will automatically be passed the event object by the addListener() method,
which will contain a latLng object for the coordinates on the map that were clicked.
The first thing we do in the function is store the latLng property of the event object in our
positions array. We'll need to know both of the locations that were clicked so it is useful
to add them both to the positions array, which is visible throughout our code.
Then we check whether the clicks variable that we defined earlier is less than or equal
to 1. Provided it is, we go ahead and create a new marker using Google's Marker()
constructor. We used the constructor earlier when we added a marker to show the
company's headquarters, but this time we set some different properties.
We set the map property to be our map instance, and this time set the position of the
marker to the latLng object contained in the event object, which will match the point on
the map that was clicked.
We'll use a green marker image for the first click, which will represent the start of the
journey. The image we'll use already has its own shadow, so when we add the first marker,
which we can determine using a JavaScript ternary that checks whether clicks is equal to
0, we set the flat property to true to disable the shadow that Google will otherwise add.
We can easily add a nice drop animation so that when the map is clicked, the new marker
drops into place. The animation features a bounce easing effect, which is also visually
pleasing. The animation is set using the animation property, which is set to DROP using the
Animation API.
We can also set a title for the marker, which is displayed when the cursor hovers over it,
using the title property. Again we use a simple JavaScript ternary to set either the Start
or End as the string depending on value of our clicks variable.
We use the icon property to specify the path to the image that we'll use for the start
marker. When clicks is not equal to 0 we just specify an empty string, which causes the
default red marker to be added.
68
Project 3
We also set the draggable property to true to make the markers draggable. This will let
users modify the start or end locations of the journey if they wish. We can add the code that
will handle this a little later on.
Next we can use Google's event API to trigger a custom event. We use the trigger()
method, specifying the map instance as the object that the event will originate from,
locationAdd as the name of our custom event, and pass the event object that we've
worked with in our addMarker() function (stored in e) as a parameter to any handlers that
may be listening for our custom event. We add a handler for this event in the next section.
Lastly we can set a unique id attribute on the marker so that we can differentiate each
marker. We'll need this when we want to update our UI following a marker drag, which
we'll look at a little later on.
This is everything we want to do at this point while the clicks variable is still less than or
equal to 1. The second branch of the outer conditional in our addMarker() function deals
with situations when clicks is greater than 1.
In this case, we know the map has already been clicked twice, so when this occurs we want
to stop listening for clicks on the map. We can unbind our handler using the event API's
removeListener() method. This method simply takes a reference to the eventListener
returned by the addListener() method.
When we bind the click event on the map to our addMarker function, we store what is
returned in the mapClick variable, which is what is passed to the removeListener()
method.
At this point we should be able to run the page in a browser and add new markers to the
map by clicking at different locations.
Classified Intel
We used a function expression in this task, by assigning the event handler to a variable,
instead of perhaps the more familiar function declaration. This is generally considered a
good practice, and while not essential in this situation, it is certainly a good habit to get into.
For a thorough understanding of why function expressions are generally better than function
declarations, see John Resig's blog post at http://ejohn.org/blog/javascript-asa-first-language/.
69
An Interactive Google Map
Updating the UI with the start and
end locations
Once the two markers have been added to the map, we want to display their locations in the
UI sidebar at the right of the page ready for when we compute the cost of the journey.
We'll want to show the full street address of each location that is clicked and also add a
button that triggers the computation of a quote based on the locations that the visitor has
chosen on the map.
Prepare for Lift Off
In the last task we used Google's trigger() method to trigger a custom event each time a
new marker was added to the map following a click. In this task we'll add a handler for that
custom event.
So far in this project, we've stuck almost entirely to Google's map API and haven't really used
jQuery at all other than to add the initial document.load wrapper for the rest of code. In
this part of the project we'll rectify that and fire up jQuery in order to update our UI.
Engage Thrusters
The handler for our custom locationAdd event should be as follows, which can be added
directly after the mapClick variable from the last task:
api.event.addListener(map, "locationAdd", function (e) {
var journeyEl = $("#journey"),
outer = (journeyEl.length) ? journeyEl : $("<div>", {
id: "journey"
});
new api.Geocoder().geocode({
"latLng": e.latLng },
function (results) {
$("<h3 />", {
text: (clicks === 0) ? "Start:" : "End:"
}).appendTo(outer);
$("<p />", {
text: results[0].formatted_address,
id: (clicks === 0) ? "StartPoint" : "EndPoint",
"data-latLng": e.latLng
70
Project 3
}).appendTo(outer);
if (!journeyEl.length) {
outer.appendTo(ui);
} else {
$("<button />", {
id: "getQuote",
text: "Get quote"
}).prop("disabled", true).appendTo(journeyEl);
}
clicks++;
});
});
As we'll be adding some new elements to the page, we'll also need to update our style sheet
for this project. Add the following new styles to the bottom of google-map.css:
#journey { margin-top:2em; }
#journey h3 { margin-bottom:.25em; }
Objective Complete - Mini Debriefing
We add the event handler for our custom locationAdd event in the same way that we
added our click events, using Google's addListener() method.
Within the event handler we first define some variables. The first is a cached jQuery object
that represents the element that displays the start and end points.
The next variable we set is then one of two things. If the jQuery object we set as the first
variable has length, we know the journey element exists on the page, so we just store a
reference to it. If it doesn't exist, we create a new element to use as the journey element and
set its id to journey.
When the map is clicked for the first time, the journey element won't exist and will be
created. The second time the map is clicked, the element will exist, so it will be selected
from the page instead of being created.
Next we use the geocode() method of Google's Geocoder() API, which allows us to
reverse-geocode a latLng object to get a street address. This method takes two arguments.
The first is a configuration object, which we can use to specify the latLng object that we
want to convert.
The second argument is a callback function that is executed once the geocoding is complete.
This function is automatically passed a results object that contains the address.
71
An Interactive Google Map
Within this callback function we can use jQuery to create new elements to display the
address and then append them to the journey element. The complete street address is
found in the formatted_address property of the results object, which we can set as
the text of one of the new elements. We can also set an id attribute on this element so that
we can easily select it programmatically when required, and store the latLng object of the
location using a custom data-latLng attribute.
The results object also contains a range of other useful properties about the address, so be
sure to check it out in the object explorer of your favorite browser-based developer toolkit.
If the journey element doesn't exist we can then append it to the UI in order to display
the address of the location. If it does exist, we know that it is the second click and can
then create a new <button> that can be used to generate a quote based on the distance
between the two locations.
We disable the <button> element using jQuery's prop() method to set the disabled
property. We can enable the button later when a weight is added to the <input> in the UI.
Once we have added the new elements showing the journey start and end points in the UI,
we can then increment the clicks variable so that we can keep track of how many markers
have been added.
Now when we run the page and click on the map twice to add both the markers, the address
of the points that we clicked should be displayed in the UI area at the right of the page. We
should also now see the red end marker and be limited to adding only two markers now that
we're incrementing the clicks variable.
Handling marker repositions
We've made our map markers draggable, so we need to handle address changes following a
marker drag. This task will show just how easily that can be done. This will take just two steps:
ff
Binding each marker to the dragend event
ff
Adding the handler function for the event
Engage Thrusters
First we need to bind each marker to the dragend event when the marker is created. To do
this, we should add the following highlighted line of code to the addMarker() function,
directly after the marker's constructor:
var marker = new api.Marker({
map: map,
position: e.latLng,
72
Project 3
flat: (clicks === 0) ? true : false,
animation: api.Animation.DROP,
title: (clicks === 0) ? "Start" : "End",
icon: (clicks === 0) ? "img/start.png" : "",
draggable: true,
id: (clicks === 0) ? "start" : "end"
});
api.event.addListener(marker, "dragend", markerDrag);
Next we should add the markerDrag() function itself. This can go directly after the
locationAdd handler that we added in the last task:
var markerDrag = function (e) {
var elId = ["#", this.get("id"), "Point"].join("");
new api.Geocoder().geocode({
"latLng": e.latLng
}, function (results) {
$(elId).text(results[0].formatted_address);
});
};
Objective Complete - Mini Debriefing
In this task we first updated the addMarker() function to bind each new marker to the
dragend event, which will be fired once the marker stops being dragged. We specify the
marker as the first argument to Google's addListener() method, which is the object to
bind to the event. The name of the event, dragend, is specified as the second argument,
and markerDrag as the name of the function that will handle the event.
Then we added markerDrag() as a function expression. Because it's an event handler it will
automatically be passed to the event object, which once again contains the latLng that we
need to pass to a Geocoder() to get the new address that the marker was dragged to.
Inside the handler we first set a new variable that will be used as the selector for the
element in the UI we want to update. Instead of concatenating a string together, we use the
array.join() technique for performance reasons. The first and last items in the array we
join are simply text.
The second item will be a string containing either start or end depending on which marker
was dragged. Inside our event handler this refers to the marker, so we can use it get the
custom id property that we added to each marker when it was created, allowing us to
update the right element in the UI.
73
An Interactive Google Map
Once we have constructed the selector we just get the street address using Google's
geocode() method exactly as we did before, which will give us the new address
of the marker after the drag.
Inside the callback function for geocode() we use the selector we just created to select the
<p> element in the UI and update its text content to the newly geocoded address.
Now when we view the page, we should be able to add the markers to the map as before,
then drag them around and see the new address in the UI area at the right of the page.
Factoring in weights
We now have two addresses – the start and end markers for the journey. All the visitor needs
to do now is enter a weight and we'll be able to calculate and display the distance and cost.
Engage Thrusters
All we need to do in this task is add a handler for the <input> element in the UI area so
that once a weight is entered the <button> becomes clickable. We can achieve this with
the following code, which can be added directly after the markerDrag() function from the
previous task:
$("#weight").on("keyup", function () {
if (timeout) {
clearTimeout(timeout);
}
var field = $(this),
enableButton = function () {
if (field.val()) {
$("#getQuote").removeProp("disabled");
} else {
$("#getQuote").prop("disabled", true);
}
},
timeout = setTimeout(enableButton, 250);
});
74
Project 3
Objective Complete - Mini Debriefing
We can add the event handler for the user-generated keyup DOM event using jQuery's
on() method. Using the on() method is now the standard way of attaching event handlers
in jQuery. Old methods such as live() or delegate() have now been deprecated and
should not be used.
Within the event handler we first check whether a timeout has been set, and if it has, we
clear it.
We then cache a selector for the <input> element so that we can see it inside our
enableButton() function. We add the enableButton() function, again as a
function expression.
All this function does is check whether the <input> element has a value, and if it does, we
set the disabled property to false using jQuery's prop() method. If it doesn't have a
value, we just disable it once more by setting the disabled property to true. Lastly we set
a timeout using the JavaScript setTimeout() function, passing it the enableButton()
function as the first argument. We set 250, or a quarter of a second, as the timeout length.
The timeout is stored in the timeout variable, ready for us to check the next time the
function is executed.
Classified Intel
The reason we use the timeout here is to rate-limit the number of times the
enableButton() function is executed. The function will be invoked after every
character is entered into the field.
A quarter of a second is a barely discernible delay, but if someone types a long number into
the field quickly, it can drastically reduce the number of times the function runs. Within
the function, we select an element from the page and create a jQuery object. That's not
too intense and in this example we probably don't even need to worry about it. But using a
timeout like this is a robust solution that can help out when doing more intense operations
inside a frequently fired event handler.
We could have just used jQuery's one() method to attach an event handler that simply
enables the <button> and then removes itself. However, this wouldn't allow us to disable
the <button> once more if the figure entered into the field is removed.
75
An Interactive Google Map
Displaying the projected distance
and cost
Our last task in this project is to get the distance between the two markers and calculate the
cost of the journey. Once calculated, we should probably display the results to the visitor too.
Engage Thrusters
First we should attach a click event handler for our <button>. Add the following code
directly after the handler for the keyup event that we added in the last task:
$("body").on("click", "#getQuote", function (e) {
e.preventDefault();
$(this).remove();
});
Next we can get the distance between the two points. Directly after the remove() method
we just added (but still inside the click handler function), add the following code:
new api.DistanceMatrixService().getDistanceMatrix({
origins: [$("#StartPoint").attr("data-latLng")],
destinations: [$("#EndPoint").attr("data-latLng")],
travelMode: google.maps.TravelMode.DRIVING,
unitSystem: google.maps.UnitSystem.IMPERIAL
}, function (response) {
});
Now we just need to compute and display the cost, which we can do by adding the following
code to the empty callback function we just added. First we can add the variables we'll need:
var list = $("<dl/>", {
"class": "clearfix",
id: "quote"
}),
format = function (number) {
var rounded = Math.round(number * 100) / 100,
fixed = rounded.toFixed(2);
return fixed;
},
term = $("<dt/>"),
desc = $("<dd/>"),
76
Project 3
distance = response.rows[0].elements[0].distance,
weight = $("#weight").val(),
distanceString = distance.text + "les",
distanceNum = parseFloat(distance.text.split(" ")[0]),
distanceCost = format(distanceNum * 3),
weightCost = format(distanceNum * 0.25 * distanceNum),
totalCost = format(+distanceCost + +weightCost);
Next we can generate the HTML structure that we'll use to display the computed figures:
$("<h3>", {
text: "Your quote",
id: "quoteHeading"
}).appendTo(ui);
term.clone().html("Distance:").appendTo(list);
desc.clone().text(distanceString).appendTo(list);
term.clone().text("Distance cost:").appendTo(list);
desc.clone().text("£" + distanceCost).appendTo(list);
term.clone().text("Weight cost:")
.appendTo(list);
desc.clone().text("£" + weightCost).appendTo(list);
term.clone().addClass("total").text("Total:").appendTo(list);
desc.clone().addClass("total")
.text("£" + totalCost)
.appendTo(list);
list.appendTo(ui);
Lastly, we should probably add some additional styling for the new elements that we just
created and added to the page. At the bottom of google-map.css, add the following
new styles:
#quoteHeading {
padding-top:1em; border-top:1px dashed #aaa;
margin-top:1em;
}
#quote dt { margin-right:0; }
#quote dd { width:50%; }
#quote .total {
padding-top:.5em; border-top:1px dashed #aaa;
margin-bottom:0; font-size:1.5em;
}
77
An Interactive Google Map
Objective Complete - Mini Debriefing
We started out by binding a click event handler to the body of the page using jQuery's on()
method. This time we use the 3-argument form of the method where the first argument is
still the name of the event, the second argument is a selector to filter the event by, and the
third argument is the function to trigger when the event occurs.
Events in JavaScript bubble up through their containers and when the event hits the body,
it will be filtered by the selector used as the second argument and the function will only be
executed if it was dispatched by an element that matches the selector. In this example, only
events dispatched by the <button> will trigger the function.
Using the on() method in this form gives us a means of employing powerful event
delegation that allows us to bind events for elements which may or may not exist at the time
of the binding.
Within the handler function, we first prevent the default behavior of the browser. There
shouldn't be any default behavior because we don't have a <form> on the page so there is
nothing for the <button> to submit. But if someone were to try and run this on an ASPX
page, which usually does have a <form> enclosing most, if not all, of the elements on the
page, it could behave in unexpected ways. Unless strictly necessary, preventDefault()
should always be used.
We then remove the <button> from the page. Note that even though the event handler
is bound to the <body>, the this object inside the handler function still points at the
<button> element that triggered the event.
We then used another of Google's APIs – the DistanceMatrixService(), which allows
us to compute the distance between two or more points on the map. Because we don't need
to reference the object returned by the DistanceMatrixService() constructor, we can
chain the getDistanceMatrix() method directly onto it.
This method takes two arguments with the first being a configuration object and the
second a callback function to execute when the method returns. The callback function will
automatically be passed an object containing the response.
We set several configuration options using the first argument. The origins and
destinations options both take arrays where each item in each array is a latLng object.
We can easily get the latLng objects for both of the markers using the custom datalatLng attribute that we set when we showed the addresses.
We also set the travelMode option to the distance it would be via road using the google.
maps.TravelMode.DRIVING constant, and set the unitSystem option to google.maps.
UnitSystem.IMPERIAL to give a distance in miles instead of kilometers, for no other
reason than because I'm a Brit, and I'm used to using miles.
78
Project 3
The callback function we supply is automatically passed a results object that contains, of
course, the results returned by the distance matrix. The first half of the callback function is
concerned with creating variables and computing values. The second part of the function
deals with displaying the information that has been computed.
We first create a new <dl> element and give it a class that is required for use with our
common.css style sheet, and an id attribute, mostly for decorative styling. Then we add
a simple function expression that receives a number as an argument, rounds it, and then
fixes it to two decimal places before returning it. We'll use this function to ensure that our
financial figures are in the required format.
We also create a new <dt> element and a new <dd> element that we can clone as many
times as required without having to repeatedly create new instances of jQuery, and then
store the value entered into the weight text field using jQuery's val() method.
Next we extract the distance property from the object passed to the callback function.
Its structure may look complex, as the object we are actually interested in for this example
is buried within a multidimensional array, but as the method's name suggests, it is able to
return a complex matrix of results for multiple origins and destinations.
Following this we concatenate a string that includes the text property of the distance
object that we just stored and the full word miles. The distance matrix returns imperial
results as mi instead of the full miles, so we add the string les to the end of the value.
We then get the numerical distance by splitting the string on the space between the number
of miles and the letters mi. JavaScript's split() function will return an array of two items
containing the part of the string up to, but not including, the split-character and the part
after the split-character. We are only interested in the first item in this array, and also use
JavaScript's parseFloat() function to ensure that this value is definitely a number and not
a string.
Now we have enough information to actually work out the cost of the journey. We've
specified the charge per mile to be £3 so we multiply the distance by 3 and pass the result to
our format() function so that the number is in the correct format.
We can also work out the charge per kilogram per mile in a very similar way, first multiplying
the weight by the cost per kilogram, then multiplying by the distance. Again we pass this
figure into our format() function. Then we can work out the total cost by adding these two
figures together. The figures that we've been working with somehow become strings. To fix
this, we can still use our format() function, but we prefix each of the values we want to
add with the + character, which will force them to be numbers and not strings.
Once we have created the figures we wish to display, we can then create the new elements
that we need to use to display them, starting with a nice heading to help clarify the new set
of information we're adding to the UI.
79
An Interactive Google Map
We can then create the clones of the <dt> and <dd> elements which hold each label and
figure. Once these have been created, we append them to the <dl> element we created,
before finally appending the new list as a whole to the UI, as shown in the following screenshot:
Classified Intel
The astute of you will notice that the number rounding solution we've used in this example
isn't that robust, and won't round all fractions as precisely (or correctly) as would be
required for a genuine system that deals with real currency.
JavaScript does not handle floating point arithmetic as gracefully as some other languages
do, and so creating the perfect rounding system that rounds correctly 100 percent of the
time is beyond the scope of this book.
For those who are interested, the stackoverflow site has some extremely illuminating
answers posted to questions around currency formatting in JavaScript. For example, see:
http://stackoverflow.com/questions/149055/how-can-i-format-numbersas-money-in-javascript.
80
Project 3
Mission Accomplished
We've covered a lot of both Google and jQuery functionality in this project. Specifically we
looked at the following subjects:
ff
Adding markers and overlays to the map using the Marker() and
InfoWindow() constructors.
ff
Reacting to map-driven events such as clicks on markers or marker drags. Event
handlers are attached using the addListener() method of the google.maps API.
We also saw how to fire custom events using the trigger() method.
ff
Using Google's services to manipulate the data generated by the map. The
services we used were the Geocoder() to reverse-geocode the latLng of
each point on the map that was clicked in order to obtain its address, and the
DistanceMatrixService() to determine the distance between the points.
ff
Taking advantage of jQuery's event capabilities to add both standard and delegated
events using the on() method to detect when different parts of our UI were
interacted with, such as the <button> being clicked or the <input> being typed into.
ff
Using jQuery's powerful DOM manipulation methods to update the UI with
addresses and the quote. We used a range of these methods including clone(),
html(), text(), and prop(), as well both selecting and creating new elements.
You Ready To Go Gung HO?
A Hotshot Challenge
In this example, visitors are only able to generate a single quote. Once the getQuote
<button> is clicked, the results are displayed and no further interaction is possible. Why don't
you change it so that a reset button is added to the UI when the quote is generated? The visitor
can then clear the quote and the markers from the map and start over from scratch.
81
Project 4
A jQuery Mobile
Single-page App
jQuery mobile is an exciting project that brings the power of jQuery to the world of
handheld and mobile experience. Like jQuery UI, it builds on and extends the jQuery core
with a series of UI widgets and helpers. In this case these are optimized for mobile display
and a touch interface.
We're also going to use JsRender, the official templating solution for jQuery and the
successor to the jQuery template plugin tmpl.
Mission Briefing
In this project we'll build a simple application that looks for questions on stack overflow
that have an un-awarded bounty on them. We'll call it Bounty Hunter. It will contain just
a few individual pages, but will be made to feel like a native application rather than as a
standard website.
Although sites and apps built with jQuery Mobile will work perfectly fine on a laptop or
desktop, jQuery Mobile subscribes to the mobile-first philosophy of building the smallest
layouts first.
This is the layout that we'll be focusing on throughout this project. If you don't have a
smartphone or other capable mobile device, the example app we'll be building will still work
in a normal desktop browser.
A jQuery Mobile Single-page App
The app we'll be building in this project will look as shown in the following screenshot:
Why Is It Awesome?
jQuery Mobile offers full support and, importantly, consistency across all of the major
modern smartphones and tablets. It also offers limited support to a much wider sphere of
common, but perhaps older, and definitely less capable mobile devices. It builds on the solid
foundation of jQuery itself and borrows a lot of best-practices from jQuery UI, certainly with
regard to how widgets are initialized and configured.
jQuery Mobile offers two ways in which widgets can be initialized; we can use the extensive
HTML5 data- attribute system, which will trigger the initialization of widgets automatically
without any additional configuration, or we can create widgets dynamically and invoke them
purely via script.
84
Project 4
Both techniques have their advantages and disadvantages and we'll get to look at both
techniques over the course of this project, so you'll be able to decide which way suits you best.
Your Hotshot Objectives
These are the tasks that this project will be broken down into:
ff
Building the welcome screen
ff
Adding a second page
ff
Creating the script wrapper
ff
Getting some bounties
ff
Adding a JsRender template
ff
Building the list view
ff
Building an item view
ff
Handling paging
Mission Checklist
The jQuery Mobile site provides a page template to use as a starting point when developing
with the framework. We can use that template as the basis for this project. To get set up
we should visit http://jquerymobile.com/demos/1.2.0/docs/about/gettingstarted.html.
Copy the template shown in the Create a basic page template section and save it as
bounty-hunter.html in our main working directory. This template contains everything we
need to get started.
We should also link to JsRender at this point; add the following code directly after the
<script> element that links to jQuery Mobile in the template we just saved:
<script src="http://borismoore.github.com/jsrender/jsrender.js">
</script>
At the time of writing, the current version of jQuery Mobile is not compatible
with jQuery 1.9. The template that we'll get from the jQuery Mobile site will
already link to a compatible version of jQuery, and 1.9 support will shortly be
available once jQuery Mobile hits the 1.3 milestone.
In order to test our mobile app, we should also use a web server for this project so that the
test page is viewed using a proper http:// URL and not a file:/// URL. You may already
have an open source web server, such as Apache, installed on your computer, and if so that
should be fine.
85
A jQuery Mobile Single-page App
If you don't have a web server already installed and configured, I'd recommend downloading
and installing Microsoft's Visual Web Developer Express (VWDE). This is a free version
of Microsoft's industry-standard IDE Visual Studio, and as well as including a built-in
development web server, it's also a very capable IDE with Intellisense support for JavaScript
and jQuery and a range of features for frontend developers.
For developers who prefer open source software, the Apache web server, along with PHP
and MySQL, can be installed on both Mac and Windows systems. To make installation and
configuration easier, a range of packages have been created which install the software
together and configure it automatically, such as XAMPP.
VWDE can be installed by visiting http://www.microsoft.com/
visualstudio/en-us/products/2010-editions/visualweb-developer-express.
The XAMPP downloads are available at http://www.
apachefriends.org/en/xampp.html.
Building the welcome page
Many apps have a welcome or home screen that the user can return to in order to select
common actions. In our first task of this project, we'll build the welcome screen, which will
consist of some simple page furniture such as a header, footer, a logo, and will also feature a
search box and button that will trigger a call to Stack Exchange's API.
Prepare for Lift Off
At this point we can create the additional resources that we'll be using in the project. We
should create a new style sheet called bounty-hunter.css in the css folder, and a new
script file called bounty-hunter.js in the js folder.
We should add a <link> element to the <head> of the page for the style sheet. The
following code should be added directly after the jQuery mobile style sheet (and before the
jQuery mobile <script> elements):
<link rel="stylesheet" href="css/bounty-hunter.css" />
We can add the <script> element in the usual place right before the closing </body> tag:
<script src="js/bounty-hunter.js"></script>
Since jQuery Mobile provides its own baseline styling that includes
a reset and typography defaults, we won't need to link to our
common.css file in this example.
86
Project 4
Engage Thrusters
The jQuery Mobile template that we downloaded contains the recommended basic structure
that most jQuery Mobile pages should be built from. We'll use the recommended structure,
but we'll be adding some extra attributes to the existing markup.
We should add an id attribute to the <div> element that has the data-role="page"
attribute in bounty-hunter.html; set the id attribute to welcome:
<div data-role="page" id="welcome">
Next we should change the original markup so that it appears as follows. First we can add a
header area:
<div data-role="header">
<h1>Bounty Hunter</h1>
</div>
Next we can add the main content area directly after the header area:
<div data-role="content">
<p>
Enter tag(s) to search for bounties on.
Separate tags with a semi-colon, or leave blank to get
all bounties.
</p>
<div class="filter-form">
<label for="tags" class="ui-hidden-accessible">
Search by tag(s):
</label>
<input id="tags" placeholder="Tag(s)" />
<button data-inline="true" data-icon="search">
Search
</button>
</div>
</div>
<img src="img/boba.png" alt="Bounty Hunter" />
Lastly we can add a footer area after the main content area:
<div data-role="footer" data-position="fixed"
data-id="footer">
<small>&copy; 2012 Some Company Inc.</small>
<a href="bounty-hunter-about.html" data-icon="info"
data-role="button" data-transition="slide">About</a>
</div>
87
A jQuery Mobile Single-page App
We can also add a few styles for our welcome screen. Add the following selectors and rules
to bounty-hunter.css:
.filter-form .ui-btn { margin:10px 0 0 0; float:right; }
.ui-footer small { display:block; margin:10px; float:left; }
.ui-footer .ui-btn { margin:2px 10px 0 0; float:right; }
Objective Complete - Mini Debriefing
First we updated the text inside the <h1> element within the container <div> that has the
data-role="header" attribute.
We then added some content to the content container, including a paragraph of introductory
text and a container <div>. Inside the container we added the <label>, <input>, and
<button> elements.
jQuery Mobile recommends using a <label> element with a valid for attribute for all
<input> elements for accessibility reasons, so we add one, but then hide it using the
ui-hidden-accessible class. This will allow assistive technologies to still see it,
without it cluttering up the page visually.
The <input> is just a simple text field with an id attribute for easy selection from scripts
and a placeholder attribute, which adds the specified text inside the <input> as
placeholder text. This is nice to use to give a visual cue now that the label is hidden, but
may not be supported in older browsers.
The <button> element has several custom jQuery Mobile data- attributes and will be
enhanced automatically by the framework when the page initially loads. jQuery Mobile
automatically enhances a range of different elements based on which element they are
and any data- attributes they have. Enhancements usually include wrapping the original
element in a container or adding other additional elements to sit alongside it.
The data-inline="true" attribute sets the container that is wrapped around the
<button> to inline-block so that it doesn't span the full width of the viewport. The
data-icon="search" attribute gives it a search icon.
We added some extra data- attributes to the container <div> element in the original
template with the data-role="footer" attribute. The data-position="fixed"
attribute works in conjunction with the data-id="footer" attribute to fix the element to
the bottom of the viewport, and to ensure that is not transitioned when we change pages.
Inside the footer container, we added a <small> element with some fake copyright info,
such as would usually be found in a web page's footer. We also added a new <a> element
that links to another page, which we'll add in the next task.
88
Project 4
This element is also given several custom data- attributes. The data-icon="info"
attribute gives the enhanced element an icon. The data-role="button" attribute
triggers enhancement by the framework and gives this simple link its button-like
appearance. The data-transition="slide" attribute uses the slide transition
when navigating to the new page.
Lastly we added some basic styling to the style sheet for this project. We floated the
search button to the right and changed the margin given to it by jQuery Mobile. The style
is added using the class we added to our container and a class added by the framework.
We need to use both classes to ensure that our selector is more specific than the one
used by the framework.
We also styled our footer elements to float them left and right and position them
as desired. Again we have to beat the specificity of the selector used by the default
jQuery Mobile theme.
At this point, we should be able to run the page in a browser and see the home page with
the header and footer at the top and bottom respectively, the super-simple search form, and
the big orange image that gives the application a basic identity.
Classified Intel
jQuery Mobile is built upon a custom data- attribute system in which we can give certain
attributes to elements and have the framework initialize widgets based on them. This
custom data- attribute framework isn't mandatory; we can manually initialize and configure
widgets if we want.
But using the attributes is convenient and allows us to focus on the custom script code to
add the behavior we want without worrying about the setup and initialization of the jQuery
Mobile widgets we wish to use.
Adding a second page
In this task we'll add the page that the About hyperlink we added to the footer container of
the welcome page links to. This allows us to experience jQuery Mobile transitions in action,
configured purely via the data- attributes system.
For more information, see the jQuery Mobile data-attributes
reference at http://jquerymobile.com/demos/1.2.0/
docs/api/data-attributes.html.
89
A jQuery Mobile Single-page App
Prepare for Lift Off
Save a new copy of the jQuery Mobile page template that we used in the last task but
this time call it bounty-hunter-about.html and save it in the main project directory
(alongside the bounty-hunter.html page).
We also still need to link to our bounty-hunter.css file, our bounty-hunter.js file,
and JsRender as we did before.
For more information on JsRender, see the documentation at
https://github.com/BorisMoore/jsrender.
Engage Thrusters
In our new bounty-hunter-about.html page, change the markup inside the <div> with
a data-role="page" to the following:
<div data-role="header">
<a href="bounty-hunter.html" data-icon="home"
data-shadow="false" data-iconpos="notext"
data-transition="slide" data-direction="reverse"
title="Home"></a>
<h1>About Bounty Hunter</h1>
</div>
<div data-role="content">
<p>
Bounty Hunter is an educational app built for the
jQuery Hotshots book by Dan Wellman
</p>
<a href="http://www.danwellman.co.uk">
danwellman.co.uk
</a>
</div>
<div data-role="footer" data-position="fixed"
data-id="footer">
<small>&copy; 2013 Some Company Inc.</small>
<a class="ui-disabled" href="#" data-icon="info"
data-role="button">About</a>
</div>
90
Project 4
Objective Complete - Mini Debriefing
This time, along with setting some different text in the <h1> inside the header container,
we also added a new link. This links back to the welcome screen of the app and uses several
custom data- attributes.
The data-icon, as before, sets the icon that should be used for the button. We can
disable the default shadow applied to the outer container element of the icon using datashadow="false", and set the data-iconpos="notext" attribute to make the button an
image-only button.
We also specified the data-transition="slide" attribute, so that the page
transitions nicely back to the welcome page, but this time we also set the datadirection="reverse" attribute so that the page appears to go backwards (that is, it
slides in the opposite direction) to the home page. Because we put this link before the <h1>
element it will be automatically floated to the left by the framework.
We added some basic content to the content container. This isn't important, and as you can
see, I've added a shameless plug for my personal website. This external link isn't completely
useless however, because it does show that when a link is prefixed with http://, jQuery
Mobile knows that it's an external link and doesn't hijack the click and try and transition it
into view.
You'll notice that the footer container has the same data- attributes as before, including
the same data-id="footer" attribute. This is what gives the footer container persistence.
When the page transitions into view, the footer will appear outside of the transitioned area
and remain fixed at the bottom of the page.
We've modified the <a> element in the footer container slightly. We've removed the
data-transition attribute and added the ui-disabled class instead. We've also
changed the href to a simple hash. As we're already on the About page, the About link
will do nothing, so we disable it to avoid the page being reloaded if it is clicked.
Classified Intel
jQuery Mobile adds its beautiful page-to-page transitions by hijacking any relative links.
When a relative link is clicked, jQuery mobile will fetch the page via AJAX, insert it into the
DOM of the current page, and transition it into view.
Generally when using a jQuery Mobile site, you will never move away from the page that
you started on because the framework will silently hijack same-domain links and dynamically
insert the content into the page. You might therefore think that each page need not link to
all of the CSS and script resources.
91
A jQuery Mobile Single-page App
This is not the case however – what if someone lands directly on one of the internal pages?
Or what if an external link is followed, but then the visitor returns using their browser's back
button? In both of these scenarios, they will be greeted with an unenhanced, dysfunctional
page that looks and feels nothing like the page they expected to see.
Now we should be able to reload the home page, then click on the About button in the
footer, and see the About page.
Creating the script wrapper
We won't be making use of jQuery's $(document).ready() { } function (or the
$(function() { }) shortcut) to execute our code when the page has loaded.
However, we still need to protect our top-level variables and functions from the global
scope so we still need a wrapper of some kind. In this task we'll create that wrapper, as
well as our top-level variables.
Engage Thrusters
In the empty bounty-hunter.js file, we can start by adding the following code:
(function() {
var tags = "",
getBounties = function(page, callback) {
$.ajax({
url: "https://api.stackexchange.com/2.0/questions/
featured",
dataType: "jsonp",
data: {
page: page,
pagesize: 10,
tagged: tags,
order: "desc",
sort: "activity",
site: "stackoverflow",
filter: "!)4k2jB7EKv1OvDDyMLKT2zyrACssKmSCX
eX5DeyrzmOdRu8sC5L8d7X3ZpseW5o_nLvVAFfUSf"
},
beforeSend: function () {
$.mobile.loadingMessageTextVisible = true;
$.mobile.showPageLoadingMsg("a", "Searching");
}
92
Project 4
}).done(function (data) {
callback(data);
});
};
}());
Objective Complete - Mini Debriefing
Our script wrapper consists of a self-executing anonymous function (or an immediately-invoked
function expression if you prefer). This outer function is wrapped in parentheses, and has an
extra pair of brackets at the end which cause the anonymous function to execute and return
immediately. This is an established JavaScript pattern often used in large-scale applications.
This creates a closure which encapsulates all of the code within it and shields it from the
global namespace, which makes the code more robust and less likely to break or fail when
used in conjunction with other libraries or plugins.
If you aren't sure of what a closure is or what it can do, there is
an excellent discussion about it on the Stack Overflow website
(http://stackoverflow.com/questions/111102/
how-do-javascript-closures-work).
It also allows us to run code almost as soon as the document has loaded. As the <script>
element it lives within is right at the bottom of the <body>, it will not be executed until the
rest of the page has been parsed by the browser.
Within the anonymous outer function we first define some variables. The first, called tags,
will be used in various functions over the course of this project, so it needs to be accessible
everywhere. Initially it can be set to an empty string.
The next variable is a function called getBounties(), which again we define in our toplevel scope so that it can be called from elsewhere in the code without issue. We'll use the
function to issue AJAX requests at different points in the app's lifecycle, and most of the
parameters of the request will not need to change.
We make an AJAX request to the Stack Exchange API using jQuery's ajax() method. This
method is jQuery's de facto method for making AJAX requests and is what the library's
helper methods, such as getJSON(), delegate to.
93
A jQuery Mobile Single-page App
The ajax() method accepts an object literal that can be used to configure any of
the standard AJAX options supported by jQuery in order to control how the request
is performed.
The url property sets the URL that the request is made to, which we set to the entry point
of the Stack Exchange API that we'd like to use. We set the dataType to JSONP so that we
can get the data from the Stack Exchange domain without triggering the browser's crossdomain security restrictions.
JSON (JavaScript Object Notation) is a data format, with an extremely similar syntax to
object literals in JavaScript, and it is used to exchange data across different platforms or
systems. JSONP (JSON with padding) is a technique that dynamically injects new scripts into
the page, which exposes JSON data to the JavaScript parser in the browser. It is necessary
because of the browser's same-origin security policy, which restricts the domains that data
can be loaded from to the current domain.
The Stack Exchange API can be configured, and the data we receive filtered in very specific
ways, using standard query string parameters to enable or disable particular functionality.
We can use jQuery's data AJAX property to add the query string parameters we wish to set.
For more information on the Stack Exchange API, see the documentation
at https://api.stackexchange.com/.
We use the page parameter to specify which page of the results we'd like to get,
which will be received by the function as a parameter. We set the number of questions
returned to 10 to page the amount of data shown at any one time. This is set using the
pagesize parameter.
The tagged parameter uses the value of the tags variable, which we can manipulate
when required later in the project. The Stack Exchange API won't complain if we send this
parameter without a value, so we can safely set it regardless of whether there are actually
any tags or not.
We specify that we'd like the results in descending order, and sort by activity, so questions
with the most recent activity will be listed first. The site is set to stackoverflow, so that
questions are not received from the entire Stack Exchange network of sites.
The last configuration property is a predefined filter that I have already created and saved
on Stack Exchange. There is a tool included for doing this when browsing any of the API
methods. The purpose of the filter is to control exactly which fields are returned in the
response, to ensure that we aren't receiving more data than we need.
94
Project 4
In this example we're just using the Stack Exchange API anonymously.
For full production-ready applications intended for public use, we
must always register the application with Stack Applications, and use
an API key with any requests we make.
Some fields that we want are not included in the default filter (which is used if no filter is
provided when making the request), and a lot of fields that we don't need are returned.
The filter we will use here gives us just the fields we need for this project, and does not
require authentication to use.
These are most of the AJAX options that we need to set for this request; those which are not
known at this point can be passed to the function when it is invoked. We'll see how to do
this in the next task.
We can make use of jQuery's beforeSend AJAX event to show the jQuery Mobile AJAX
spinner directly before the request is made. jQuery Mobile uses a spinner every time a
page is transitioned, but we can subvert it for our own requirements when making the
AJAX request.
The framework will automatically attach a mobile object to the instance of jQuery running
on the current page. This object contains various properties used to configure the jQuery
Mobile environment, and various methods to trigger different behavior with the framework.
We can use some of these now.
To ensure the message we wish to add is displayed, because by default the spinner uses
accessibly-hidden text, we set the loadingMessageTextVisible property of the mobile
object to true.
At page load, jQuery Mobile creates an object called mobile,
which contains a range of useful properties and methods.
To actually show the spinner, we can use the jQuery Mobile showPageLoadingMsg()
method. This method takes the theme swatch to use as the first argument, which in this
case we can set to the default theme a, and the text to display inside the spinner as the
second argument.
After the ajax() method, we chain the done() method. This is the new way of handling
successful AJAX requests as of jQuery 1.8 and replaces jQuery's success() method. We
pass an anonymous function to this method to be executed when the request object returns,
and this function receives the response as an argument. Within this function we simply
invoke the callback() function that will be passed into getBounties() as the second
argument, passing it the data from the response.
95
A jQuery Mobile Single-page App
Classified Intel
In this task we made use of the done() method to handle the successful response from
the Stack Exchange API instead of the more familiar success() method. This is now
the preferred way of handling successful responses (as of jQuery 1.8). The error() and
complete() callback methods of the jqXHR object returned by any of jQuery's AJAX
methods have been deprecated in favor of fail() and always().
As of jQuery 1.5, the AJAX suite of methods have returned the jqXHR object as a promise,
or deferred, object, so this change to the API brings the AJAX methods in line with other
implementations of the promise API within jQuery.
Getting some bounties
In this task we need to get some bounties from stack overflow. We'll want to initialize a
part of our script once the welcome page of our application has been initialized. Once this
happens we can attach a handler for the <button> on the page to trigger an AJAX request
using the getBounties() function that we add in the last part.
Engage Thrusters
Inside the outer function, but after the getBounties() function in bounty-hunter.js,
add the following code:
$(document).on("pageinit", "#welcome", function () {
$("#search").on("click", function () {
$(this).closest(".ui-btn")
.addClass("ui-disabled");
tags = $("tags").val();
getBounties(1, function(data) {
data.currentPage = 1;
localStorage.setItem("res", JSON.stringify(data));
$.mobile.changePage("bounty-hunter-list.html", {
transition: "slide"
});
});
});
});
96
Project 4
We can also add a handler for the pageshow event directly after the code we just added:
$(document).on("pageshow", "#welcome", function () {
$("#search").closest(".ui-btn")
.removeClass("ui-disabled");
});
Objective Complete - Mini Debriefing
We use the pageinit event to execute code when the page is initialized for the first time.
Due to the AJAX nature of how new pages are pulled into the DOM of the existing page and
displayed, this event is more reliable than document ready when using jQuery Mobile.
We use jQuery's on() method to bind an event handler for this event to the document
object, and set the first argument of the method to the pageinit event. Because our
script will be used on every page, but the code we've added here is only relevant on the
welcome page, we use the second argument of the method to ensure that the event handler
(which we add as the third argument) is only executed when the event originates from the
welcome page.
We then bind a handler for the click event to the search <button>, again using
jQuery's on() method. Within the handler we first add the ui-disabled class to the
outer <button> container to stop further requests being initiated. We then get any tags
that may have been entered in the text field using jQuery's val() method. This will return
the value of the text input, which we then store in our top-level tags variable.
Next we can call the getBounties() function that we added in the last task. As the request
is being initiated by the welcome page we need to get the first page of the results, so we
pass 1 to the function as the first argument.
We pass an anonymous function as the second argument to getBounties(). Remember,
the handler we added for the done() method will execute the function and automatically
pass the data from the response to it.
Within this function we first need to add a new property to our data object to store the
current page number. We can then store the data object so that we can use it in the next
page. We can do this using localStorage, but because only arrays and primitive types
can be stored in localStorage, we need to convert the object to a JSON string using the
browser's native JSON.stringify() method.
We then use the jQuery Mobile changePage() method to change the current page to the
page on which we'll display the response. This method is passed the URL of the page to
change to as the first argument, and a configuration object as the second argument.
97
A jQuery Mobile Single-page App
We use this configuration object to set the transition to use when showing the new page
with the transition option, which in this case we set to slide.
After the pageinit handler we also added an event handler for the pageshow event.
This event is dispatched every time a page is shown, unlike the pageinit event which is
dispatched only the first time a given page is initialized.
We bind the event to the document object again and filter the event by the #welcome
selector once again to ensure that the code only runs when the welcome page is shown.
Within the event handler we simply remove the ui-disabled class from the outer
<button> container. If we've returned to the welcome page, it's probably because we
want to perform a new search, maybe with a different tag.
Adding a JsRender template
At the end of the last task we used the changePage() method to call a new page, so now
we need to create that page. We can add our JsRender template to the new page ready for
when we build the list view in the next task.
Prepare for Lift Off
Create a new page using the starting template from jQuery Mobile once again. Call it
bounty-hunter-list.html and save it in the root of our project folder. Change the id
attribute of the data-role="page" wrapper to list.
The <h1> in the header <div> can be changed to something like Active Bounties, and
we can add the home icon again as we did on the About page. The footer can stay the same
as on the welcome page. The content <div> can be empty to begin with.
Engage Thrusters
Near the bottom of the new page we just created, inside the page container, add the
following JsRender template:
<script id="listTemplate" type="text/x-jquery-tmpl">
<ul data-role="listview">
{{for items}}
<li data-shadow="false" data-icon="arrow-r"
data-iconpos="right">
<a href="#" id="item-{{:#index}}">
<div class="bounty">
<span>+{{:bounty_amount}}</span>
98
Project 4
<span class="expires">Expires on:
<span class="value">
{{:bounty_closes_date}}
</span>
</span>
</div>
<h1 class="title">{{:title}}</h1>
<div class="meta">
<span>Answers:
<span class="value">
{{:answer_count}}
</span>
</span>
<span class="activity">
Last activity on:
<span class="value">
{{:last_activity_date}}
</span>
</span>
</div>
</a>
</li>
{{/for}}
</ul>
</script>
Objective Complete - Mini Debriefing
The <script> element that the template resides in is given a non-standard type attribute
to stop the browser parsing the script. It's also given an id attribute so that we can easily
select it from the page when we want to interpolate the template with data and render it
to the page.
Inside the <script> element, we first create the <ul> element that will be turned into
a Listview widget by jQuery Mobile. We give this element a data-role attribute of
listview. We then use JsRender's loop construct {{for}}, which accepts the object or
array to loop over. In this case we're interested in the items array that is part of the data
object that is saved in localStorage at the end of the last task, and which will be passed to
the template function that renders the template.
The code we add within the {{for}} loop will be repeated for each item in the items array,
which will consist of a series of questions from stack overflow. The object the template will
iterate will be passed into the loop when we call JsRender's template() method a little
later on.
99
A jQuery Mobile Single-page App
The first element we add is <li> as this should naturally be a child of the outer <ul> list.
We give the <li> element several data- attributes including data-shadow="false"
to disable shadows under each <li>, data-icon="arrow-r" to give each list item a
right-pointing arrow icon, and data-iconpos="right" to position the icon at the
right of the element.
Listitem icons
In order for the icons we've added to our list items to be displayed, each item
should contain a link. If no <a> elements are found within the item when the
widget is initialized, no icon will be added.
Inside the list item, we add an <a> element and a unique id for when we come to display
the item view later on. We can create a unique id using the loop index of the template,
which is available to us within the loop as #index.
Inside the <a> element we have several other elements. The first is a container for the
bounty offered on the current question. Inside this container we have another JsRender
token that will be replaced with the data from the object we are iterating. To access a
property of the object within our template, we use {{: followed by the property name and
ending with }}. The colon inside the opening double curly braces indicates that no HTML
encoding should be carried out. The Stack Exchange API will sanitize the data for us so we
can just use it as is.
We can also display some text and the date that the bounty expires using some nested
<span> elements, one of which has a class for some specific styling, and another property
from our data object.
We can output the title of the question using an <h1> element and another JsRender
template tag that pulls out the title property from the current item from inside the
data object.
Lastly we can display some meta-information about the question such as the number of
answers it has and the last time there was activity on the question. This information is added
in the same way as before, using a combination of <span> elements and JsRender template
tags to display various properties from our data object.
Building the list view
Our application should now be at the point where it has received some data that needs to
be formatted and displayed. We've also added a JsRender template ready to be used to build
the Listitem elements for a Listview widget.
100
Project 4
All we need to do now is render the template and display the results in our widget. We can
also add some additional controls to the widget to let the visitor navigate through the paged
results, although we won't make these functional just yet.
Engage Thrusters
First of all we can add some additional markup to the content container in our list page
(bounty-hunter-list.html):
<div class="ui-bar ui-bar-c">
<a href="#" data-role="button" data-icon="back"
data-inline="true" data-mini="true" class="ui-disabled">
Prev
</a>
<h2>Page
<span class="num"></span> of <span class="of"></span>
</h2>
<a href="#" data-role="button" data-icon="forward"
data-iconpos="right" data-inline="true"
data-mini="true" class="ui-disabled">
Next
</a>
</div>
<div id="results"></div>
<div class="ui-bar ui-bar-c footer-bar">
<a href="#" data-role="button" data-icon="back"
data-inline="true" data-mini="true" class="ui-disabled">
Prev
</a>
<h2>Page
<span class="num"></span> of <span class="of"></span>
</h2>
<a href="#" data-role="button" data-icon="forward"
data-iconpos="right" data-inline="true"
data-mini="true" class="ui-disabled">
Next
</a>
</div>
101
A jQuery Mobile Single-page App
Next we need to update our script in order to render the template and display the data.
In bounty-hunter.js, add the following code directly after the event handler for the
pageshow event:
$(document).on("pageinit", "#list", function () {
var data = JSON.parse(localStorage.getItem("res")),
total = parseInt(data.total, 10),
size = parseInt(data.page_size, 10),
totalPages = Math.ceil(total / size),
months = [
"Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul",
"Aug", "Sep", "Oct", "Nov", "Dec"
];
var createDate = function (date) {
var cDate = new Date(date * 1000),
fDate = [
cDate.getDate(), months[cDate.getMonth()],
cDate.getFullYear()
].join(" ");
return fDate;
}
$.views.helpers({ CreateDate: createDate });
$("#results").append($("#listTemplate")
.render(data))
.find("ul")
.listview();
var setClasses = function () {
if (data.currentPage > 1) {
$("a[data-icon='back']").removeClass("ui-disabled");
} else {
$("a[data-icon='back']").addClass("ui-disabled");
}
if (data.currentPage < totalPages) {
$("a[data-icon='forward']").removeClass("ui-disabled");
} else {
102
Project 4
$("a[data-icon='forward']").addClass("ui-disabled");
}
};
$("span.num").text(data.currentPage);
$("span.of").text(totalPages);
if (totalPages > 1) {
$("a[data-icon='forward']").removeClass("ui-disabled");
}
});
We also need to change our template slightly. There are two places in our template where
we show date properties; these both need to be changed so that they appear as follows:
{{:~CreateDate(bounty_closes_date)}}
And:
{{:~CreateDate(last_activity_date)}}
Lastly, we need to add some additional styling for our new elements, as well as the items
that will be added to the Listview widget. Add the following styles to the bottom of
bounty-hunter.css:
.ui-bar { margin:0 -15px 14px -15px; text-align:center; }
.ui-bar a:first-child { margin-left:-5px; float:left; }
.ui-bar a:last-child { margin-right:-5px; float:right; }
.ui-bar h2 { margin-top:10px; font-size:14px; }
.footer-bar { margin-top:14px; }
.bounty {
width:24%; border-radius:3px; margin-right:5%; float:left;
text-align:center; font-size:90%; line-height:1.5em;
font-weight:bold; color:#fff; background-color:#07d;
text-shadow:none;
}
.bounty span { display:block; }
.expires {
font-size:70%; font-weight:normal; line-height:1em;
}
.expires .value {
display:block; font-size:110%; font-weight:bold;
line-height:1.5em;
}
.title {
103
A jQuery Mobile Single-page App
width:70%; margin-top:-.25em; float:left;
white-space:normal; font-size:80%; line-height:1.25em;
color:#07d;
}
.meta { clear:both; }
.meta span {
width:24%; margin-right:5%; float:left; font-size:70%;
font-weight:normal; color:#999;
}
.meta .value {
width:70%; margin-right:0; float:none; font-size:90%;
font-weight:bold;
}
Objective Complete - Mini Debriefing
In the first step of this task we added some new HTML elements to the content container on
the page.
The first element we added will be used as a toolbar that sits above the Listview widget.
Within this toolbar are links for the visitor to navigate between different results pages. The
toolbar will pick up a lot of styling from jQuery Mobile because we have given it the ui-bar
and ui-theme class names.
The links are enhanced into Button widgets by jQuery Mobile because they have the
data-role="button" attribute. We also add icons to them using the data-icon
attribute, make them inline-block with the data-inline attribute, and make
them smaller than standard buttons using the data-mini attribute.
Lastly, we give the buttons the ui-disabled class name initially. We can enable each
button based on which page we are on and whether there are previous or next pages to
navigate to.
Along with the buttons, the toolbar also contains an <h2> element that will tell the visitor
which page they are on, and how many pages there are in total. The contents of the element
are split into spans with id attributes so that we can easily update them later on.
We also add an empty container <div> to the page with an id of results. This container will
be the element that we render our Listview widget into.
The second toolbar, which we added directly after the empty Listview container, is identical
to the first toolbar in every respect except that it has an additional class of footer-bar. We
use this to add a little CSS, which is only required by the bottom toolbar.
104
Project 4
In our script we first added a new event handler for the pageinit event of the list page.
This is bound in the same way as before using jQuery's on() method.
Within the event handler, we first set a series of variables. We store a reference to the data
that was stored in localStorage in an earlier task, and store the total number of results and
the page_size properties of the data object ready to use in various places in our code.
We also calculate the total number of pages based on the total and size variables that we
just saved, and create an array containing shortened month names, which we'll use when we
format the dates returned by Stack Exchange.
Next we need to add a new method that we can use as a helper function inside our
template. We call the method createDate and specify that the method may accept a single
argument, which will be a date string.
Inside the method, we first create a new date using the date string passed to the method.
This will be in UNIX epoch format, so we need to multiply it by 1000 for it to work with
JavaScript's Date() constructor.
The date string returned by the Date() constructor will be a full UTC date string, which is way
too long for displaying in our little bounty box, so next we define a new array where each item
in the array is part of the date string that we want to format the existing string into.
We can get the day of the month with the getDay() function. The getMonth() function
will return a zero-based number so we can use this to extract the correct shortened
month name from the array we created earlier. Lastly we get the four-digit year using the
getFullYear() function. Once the array has been filled, we join it straight away using a
space character as the join character and return the resulting string from the method.
Next we need to register our new method as a helper function so that the template we are
using can access it. This is done using the helpers() method of the views object, which
is created by JsRender and attached to jQuery. This method takes an object as its argument
where each key in the object is the name of a helper method, and each value is the actual
function we wish to use as the helper. In this example we map the CreateDate helper
method to the createDate function we just defined.
We then select the Listview widget using its id and append to it the rendered template.
The template is rendered using JsRender's render() method, which accepts the object
containing the data to be rendered as an argument.
Next we define another simple function that will add or remove the ui-disabled class name
for the buttons based on the currentPage property that we stored on the data object.
We can now update the headings to show the current page and the total number of pages.
We can do that using jQuery's text() method and displaying data.currentPage and
totalPages variables that we stored earlier.
105
A jQuery Mobile Single-page App
As this will only be the first time the list page is loaded, we know that only the Next button
needs to be enabled. We use an attribute selector to select just the two forward buttons
based on their data-icon attribute. We'll add the functionality that will make this button
work in the next and final task.
The last thing we do in our script is enable the forward buttons so that the next page can be
viewed, but only if there are more pages to display, which we can determine by checking the
totalPages variable again.
After adding the script we then updated our template to make use of the new date
formatting helper method we created. To use a helper method inside a template we just
need to use the ~ character followed by the registered name of the method. Any arguments
that need to be passed, such as the bounty_closes_date and last_activity_date
properties from each item in the array the template iterates, are passed using parentheses
as would be used when calling a normal JavaScript function.
Once the template has been rendered, we need to initialize the Listview. We do this by
first getting the new <ul> element inside the container and then using its widget method,
listview() in this case, to enhance it into a Listview widget.
Lastly we added some additional CSS styling to tweak the styling applied by the default
theme in jQuery Mobile. We need to make the toolbars full-width to match the Listview
widget, which we can do in the same way as the Listview widget itself does – simply by using
negative margin.
The Listview has negative margin for its top and bottom properties as well as its left and
right properties, so we need to counteract this by adding some positive margin to the
bottom of the top toolbar, and some positive top margin to the bottom toolbar.
We can also float the back and forward buttons left and right respectively and center the
heading text. We also shrink down the size of the heading text by a couple of pixels to help
ensure that it doesn't interfere with our buttons.
The styling for the elements inside the Listview is almost entirely for the visual appearance
of the elements we added using the template. The Listview itself will inherit a lot of styling
from the framework, so it's just the elements inside each Listitem we need to worry about.
106
Project 4
Once the bounties button has been clicked and the results have been returned, the list view
page should appear like the following screenshot:
Classified Intel
Like jQuery UI, jQuery Mobile widgets can be initialized completely from script, without
using any hardcoded data- attributes in the underlying markup. We could just as easily
have built the entire Listview widget from script instead of keeping the outer <ul>
element in the markup.
To initialize a widget we can just call its widget method. If we are creating a Listview, the
widget method is simply listview(). Other widgets can be initialized in the same way.
Also like jQuery UI widgets, jQuery Mobile widgets can accept configuration options and
event handlers, and have methods that can be called from script.
107
A jQuery Mobile Single-page App
Building an item view
The Listview provides links for each list item containing a question. In this task we can add
the page that is displayed when one of the questions is selected. This will be a much more
detailed view of a single question so we can make use of some of the other properties
returned to us by Stack Exchange. This time, instead of linking to an existing page, we'll
create a new one dynamically and inject it into the app.
Engage Thrusters
We'll use another template to render the item view because it's so convenient; add the
following code to bounty-hunter-list.html directly after the list template. We can
start by adding the outer <script> wrapper, the outer page container, and the header:
<script id="itemTemplate" type="text/x-jquery-tmpl">
<div data-role="page" id="{{:pageid}}" class="item-view">
<div data-role="header" data-position="fixed">
<a href="bounty-hunter-list.html"
data-shadow="false" data-icon="arrow-l"
data-transition="slide"
data-direction="reverse"
title="Back to list view">
Back
</a>
<h1>{{:title}}</h1>
</div>
</div>
</script>
Next we can add the content area wrapper and the content header. This should go into the
page container, directly after the header area:
<div data-role="content">
<header class="ui-helper-clearfix">
<div class="bounty">
<span>+{{:bounty_amount}}</span>
<span class="expires">
Expires on:
<span class="value">
{{:~CreateDate(bounty_closes_date)}}
108
Project 4
</span>
</span>
</div>
<div class="meta">
<span>Asked on:
<span class="value">
{{:~CreateDate(creation_date)}}
</span>
</span>
<span>Answers:
<span class="value">
{{:answer_count}}
</span>
</span>
<span class="activity">Last activity on:
<span class="value">
{{:~CreateDate(last_activity_date)}}
</span>
</span>
</div>
<h1 class="title">{{:title}}</h1>
<ul class="tags">
{{for tags}}
<li>{{:#data}}</li>
{{/for}}
</ul>
<div class="owner">
<a href="{{:owner.link}}"
title="{{:owner.display_name}}">
<img src="{{:owner.profile_image}}"
alt="{{:owner.display_name}}" />
<div>
<h3>{{:owner.display_name}}</h3>
<span>
{{:owner.accept_rate}}% accept rate
</span>
</div>
</a>
</div>
109
A jQuery Mobile Single-page App
<a data-role="button" data-icon="gear"
data-inline="true" href="{{:link}}"
title="Answer on Stack Overflow">
Answer on Stack Overflow
</a>
</header>
</div>
Next we can add the question and the list of answers. This should come directly after the
header element (but still inside the content <div>) we just added:
<div class="question">{{:body}}</div>
<h2>Answers</h2>
<ul class="answer">
{{for answers}}
<li data-shadow="false">
<h3>Answered by:
<span class="answer-name">
{{:owner.display_name}}
</span>, on
<span class="answer-date">
{{:~CreateDate(creation_date)}}
</span>
</h3>
<div>{{:body}}</div>
</li>
{{/for}}
</ul>
Lastly we can add the footer for the page. This should go directly after the content area, but
still inside the outer page container:
<div data-role="footer" data-position="fixed"
data-id="footer">
<small>&copy; 2012 Some Company Inc.</small>
<a href="bounty-hunter-about.html"
data-icon="info" data-role="button"
data-transition="slide">
About
</a>
</div>
110
Project 4
We also need to add some script to render the template and handle the page change
behavior. We can do this in the pageinit handler for the list page that we added in
the last task:
$("#results").on("click", "li", function () {
var index = $(this).find("a").attr("id").split("-")[1],
question = data.items[index];
question.pageid = "item-view-" + index;
$("body").append($("#itemTemplate").render(question));
var page = $("#item-view-" + index);
page.attr("data-external-page", true).on
("pageinit", $.mobile._bindPageRemove);
$.mobile.changePage(page, {
transition: "slide"
});
});
Lastly, we need some CSS for the new elements that are in the template we added. In
bounty-hunter.css, add the following code to the end of the file:
header {
padding:15px; border-bottom:1px solid #fff;
margin:-15px -15px 0 -15px;
box-shadow:0 1px 10px rgba(0,0,0,.3);
}
header:after {
content:""; display:block; clear:both; visibility:hidden;
}
header .bounty { margin-bottom:.75em; }
header .meta { width:70%; float:left; clear:none; }
header .meta span { width:100%; }
header .title {
width:auto; margin:0; float:none; clear:both;
font-size:125%;
}
.tags { padding:0; }
.tags li {
padding:.5%; border-right:1px solid #7f9fb6;
border-bottom:1px solid #3e6d8e; margin-right:1%;
111
A jQuery Mobile Single-page App
margin-bottom:1%; float:left; list-style-type:none;
font-size:90%; color:#4a6b82; background-color:#e0eaf1;
}
header a {
margin-left:0; float:left; clear:both;
text-decoration:none;
}
.owner {
padding:2.5%; margin:15px 0; float:left; clear:both;
font-size:70%; background-color:#e0eaf1;
}
.owner img { width:25%; margin-right:5%; float:left; }
.owner div { width:70%; float:left; }
.owner h3 { margin:-.25em 0 0; }
.owner span { font-size:90%; color:#508850; }
.question {
padding:15px; border-bottom:1px solid #000;
margin:-15px -15px 0 -15px;
}
.question img { max-width:100%; }
.answer { padding:0; list-style-type:none; }
.answer li { border-bottom:1px solid #000; font-size:80%; }
.answer h1, .answer h2, .answer h4 { font-size:100%; }
.item-view pre {
max-width:95%; padding:2.5%; border:1px solid #aaa;
background-color:#fff; white-space:pre-wrap;
}
Objective Complete - Mini Debriefing
We started out by adding a new template for the page to display a single question. This
template was considerably larger than the first one we added, for several reasons. Primarily,
because we're using this template to build an entire page, but also because we're showing
more stuff with this template. This is the detailed view of a question so we would naturally
want to show more than the summary shown in the list view.
The outer page container that we specify is given an id, which we'll add in our script, so that
we can easily select the correct page in order to show it. Aside from this we add some of
the same elements to our template as we have been adding in our actual pages, such as the
header, content, and footer containers.
112
Project 4
Within the content container is where most of the action is, although we're using the
template in exactly the same way as before – defining HTML elements and interpolating
them with properties from the object passed to the render() method.
The only new technique in this template is the creation of the tags list. We use the for
construct to iterate the list of tags, but this time the property we are iterating is a flat string
array instead of an object. As there isn't a key for us to use in the template tags to get the
value, we can instead use the special value #data, which will give us the current item in the
array being iterated.
All of the code we added to our script was contained within a click handling function that we
bind to the list of results display on the page, because we want to react to an individual list
item being clicked.
Within the handler function, we first set the variable which will contain the numerical part of
the id attribute of the list item that was clicked. We can easily get the numerical portion by
using JavaScript's split() function and specifying a hyphen as the character to split on.
When we render the template, we only want to show a single item, so we don't need to pass
in the entire object received from the AJAX request. Instead we use the index variable that
we just set to pull just the object representing the question we are interested in from the
items array within the data object.
Once we stored the object we'll be passing to our template to render, we need to add a new
property to it that is added as the id attribute for the page container in the template. This is
the pageid property that we set on the question object.
Next we render our template using JsRender's render() method once more. We pass
it the question object that we just prepared and this time the template is rendered to
the body of the page. Because it gets rendered outside of the page container it won't be
visible immediately.
Once the template has been rendered to the page, we select the outer page container and
store a reference to it in the page variable. When a new page is created dynamically and
appended to the page, jQuery Mobile will keep its markup in the page, even if we navigate
away from it.
To stop this from happening, we need to do two things; first of all we need to set the
data-external-page attribute of the page to true. Secondly we need to set a handler
for the pageinit event of the dynamic page. Once the new page has been initialized, we
mark it for deletion when the visitor navigates away from it using the internal jQuery Mobile
_bindPageRemove method.
Once this is done, we can transition to the new page using the changePage() method. We
pass the method the page element that we stored earlier, and use the configuration object
to set the transition.
113
A jQuery Mobile Single-page App
Because we've passed the changePage() method a DOM element and not specified a
URL, the address bar of the browser will not be updated and an entry will not be left in the
browser's history.
At this point we should be able to run the page on a smartphone or tablet, click on one of the
list items on the list view page and see the item view, as shown in the following screenshot:
114
Project 4
Handling paging
For our last task we'll look at wiring up the paging elements we added earlier. The Stack
Exchange API makes it easy to get results in a paged format, so we can leverage that to
our advantage.
The trade-off for requesting all available data from Stack Exchange, saving ourselves that
massive one-off hit, is that we make much smaller requests in response to the user initiating
a request for more data.
Engage Thrusters
Directly after the click handler we added for the <li> elements inside the Listview, add
the following code:
$("a[data-icon='forward'], a[data-icon='back']").on("click", function
() {
var button = $(this),
dir = button.attr("data-icon"),
page = parseInt($("span.num").eq(0).text(), 10);
if (dir === "forward") {
page++;
} else {
page--;
}
getBounties(page, function (newData) {
data = newData;
data.currentPage = page;
localStorage.setItem("res", JSON.stringify(newData));
$.mobile.hidePageLoadingMsg();
$("#results").empty()
.append($("#listTemplate")
.render(newData))
.find("ul")
.listview();
$("span.num").text(page);
setClasses();
});
});
115
A jQuery Mobile Single-page App
Objective Complete - Mini Debriefing
We attach a listener for all four buttons using the data-icon attribute once more to
select them from the page. Don't forget that this will only be done once, the first time
the list page loads.
We then store a reference to the button that was clicked, the value of the data-icon attribute
of the clicked button, and the current page. We then check the value of the dir attribute
and if it is equal to forward we increment the current page, otherwise we decrement it.
We can then call our getBounties() method once more, passing it the updated page
variable and the handler function to execute after the request.
Within this handler function, we first update the stored data by updating the data
variable with the new object returned by the most recent call to getBounties().
We add a currentpage property to the data object once more and update the
copy we have in localStorage.
We can then hide the spinner manually using the hidePageLoadingMsg() jQuery Mobile
method and then re-render the list template using the new data. Once this is done we can
update the display showing the current page, and call our setClasses() utility function to
enable or disable the forward and back buttons respectively.
Mission Accomplished
At this point we should have a fully working jQuery Mobile app than runs on both desktop
and mobile devices. It's a simple app, but we've gotten to explore a reasonable amount
of the framework. There is much more to learn, but seeing the bits that we've used in this
project should be more than enough to inspire you to delve deeper into the framework and
what it offers.
You Ready To Go Gung HO?
A Hotshot Challenge
One thing that we haven't looked at in this project so far is the theming capabilities of
jQuery Mobile. Like jQuery UI, jQuery Mobile benefits from the advanced theming
capabilities of Themeroller.
Your challenge for this project is to head on over to Themeroller at http://jquerymobile.
com/themeroller/ and build yourself a custom theme for the finished application.
116
Project 5
jQuery File Uploader
It's now possible to create a fully featured file upload widget with nothing but a few of the
latest HTML5 APIs and jQuery. We can easily add support for advanced features such as
multiple uploads, and a drag-and-drop interface, and with just a little help from jQuery UI we
can also add engaging UI features such as detailed file information and progress feedback.
Mission Briefing
In this project we'll build an advanced multifile upload widget using the HTML5 file API
to provide the core behavior, and using jQuery and jQuery UI to build an attractive and
engaging interface that visitors will find a pleasure to use.
We'll build the widget as a jQuery plugin because it's the kind of thing that we'd probably
like to be encapsulated so that we can drop it into numerous pages and have it work with
just a little configuration, instead of having to build a custom solution every time we require
this functionality.
Why Is It Awesome?
jQuery provides some great features that make writing reusable plugins a breeze.
Over the course of this project we'll see just how easy it is to package up specific
functionality and a mechanism for generating all of the necessary mark-up and
adding all of the required types of behaviors.
Handling file uploads on the client side gives us a lot of opportunity for adding experience
enhancing features, including information about each file selected for upload, and a rich
progress indicator that keeps the visitor informed about how long the upload is likely to take.
jQuery File Uploader
We can also allow the visitor to cancel the upload while it is in progress, or remove
previously selected files before the upload begins. These kinds of features just aren't
available using purely server-side techniques for handling file uploads.
At the end of this project we'll have produced the following widget:
Your Hotshot Objectives
To arrive at the completed project, we'll need to complete the following tasks:
ff
Creating the page and plugin wrapper
ff
Generating the underlying markup
ff
Adding event handlers for receiving files to upload
ff
Displaying the list of selected files
ff
Removing files from the upload list
ff
Adding a jQuery UI progress indicator
ff
Uploading the selected files
ff
Reporting success and tidying up
118
Project 5
Mission Checklist
Like in some of our previous projects, as well as using jQuery, we'll also be using jQuery UI
in this project. The copy of jQuery UI we downloaded at the start of the book should already
contain all of the widgets we require.
Like in the previous project, we'll also need to use a web server for this project, which means
running the page using a proper http:// URL and not a file:/// URL. See the previous
project for information on compatible web servers.
Creating the page and plugin
wrapper
In this task we'll create the page that links to the required resources, as well as add the
wrapper that our plugin will live within.
Prepare for Lift Off
At this point we should create the different files we'll need for this project. First, save a new
copy of our template file in the main project folder and call it uploader.html. We'll also
need a new style sheet, which should be saved in the css folder as uploader.css, and a
new JavaScript file, which should be saved in the js folder as uploader.js.
The new page should link to the jQuery UI style sheet so that we get the styling required by
the Progressbar widget, and also the style sheet for this project in the <head> of the page,
directly after the existing link to common.css:
<link rel="stylesheet" href="css/ui-lightness/jquery-ui-1.10.0.custom.
min.css" />
<link rel="stylesheet" href="css/uploader.css" />
We'll also need to link to jQuery UI and the JavaScript file for this example. We should add
both of these script files directly after the existing <script> element for jQuery:
<script src="js/jquery-ui-1.10.0.custom.min.js"></script>
<script src="js/uploader.js"></script>
119
jQuery File Uploader
Engage Thrusters
All our plugin will require is a container that the widget can render the required markup into.
In the <body> of the page, before the <script> elements linking to the different JavaScript
resources, add the following code:
<div id="uploader"></div>
As well as linking to the script file that contains the code for our plugin, we'll also need to
call the plugin in order to initialize it. Directly after the existing <script> elements, add the
following code:
<script>
$("#uploader").up();
</script>
The plugin's wrapper is a simple construct that we'll use to initialize the widget. In
uploader.js, add the following code:
;(function ($) {
var defaults = {
strings: {
title: "Up - A jQuery uploader",
dropText: "Drag files here",
altText: "Or select using the button",
buttons: {
choose: "Choose files",
upload: "Upload files"
},
tableHeadings: [
"Type", "Name", "Size", "Remove all x"
]
}
}
function Up(el, opts) {
this.config = $.extend(true, {}, defaults, opts);
this.el = el;
this.fileList = [];
this.allXHR = [];
}
$.fn.up = function(options) {
new Up(this, options);
120
Project 5
return this;
};
}(jQuery));
Objective Complete - Mini Debriefing
When building jQuery plugins, the best thing we can do is make our plugins easy to use.
Depending on what the plugin is for, it is best to have as few prerequisites as possible so if
a complex markup structure is required by the plugin, it is usually best to have the plugin
render the markup it needs, rather than making the users of the plugin try to add all of the
required elements.
In light of this, we'll write our plugin so that all it requires on the page is a simple container
that the plugin can render the markup into. We added this container to the page and gave it
an id attribute for easy selecting.
Developers using our plugin will need a way to invoke it. jQuery plugins extend the jQuery
object with additional methods and our plugin will add a new method to jQuery called up(),
which is called like any other jQuery method name – on a set of elements that have been
selected by jQuery.
The extra <script> element we add to the bottom of the <body> element calls our
plugin method in order to invoke the plugin, which is how someone using our plugin
would invoke it.
In our script file, we start out with a semicolon and an immediately-invoked anonymous
function. The semi-colon supports the modular nature of jQuery plugins and protects our
plugin from other plugins that don't stop execution correctly.
If another plugin in use on the page didn't end its final statement or expression with a
semi-colon, and we didn't start our plugin with a semi-colon, it could produce script
errors that prevented our plugin working.
We use an anonymous function as a wrapper for our plugin and invoke it immediately with
an extra set of parentheses after the function body. We can also ensure that our plugin
works with jQuery's noConflict() method by locally scoping the $ character within our
plugin and passing the jQuery object into the anonymous function as an argument.
Within the anonymous function we first define an object literal called defaults that will
be used as a configuration object for our plugin. This object contains another object called
strings, which is where we store all of the different bits of text that are displayed in
various elements.
121
jQuery File Uploader
The reason for using a configuration object for text strings is to make our plugin easy to
localize, making it easier for non-English speaking developers to use. Making plugins as
flexible as possible is a good way of making the plugin more appealing.
After the defaults object we define a constructor function that will generate instances of
our widget. The plugin is called Up and we capitalize the first letter of its name because this
is the general convention for functions that should be invoked using the new keyword.
The constructor function can accept two arguments; the first is a jQuery element or
collection of elements and the second a configuration object defined by the developer
using our plugin.
Within the constructor function we first attach some members to the instance. The first is
called config and will contain the object returned by jQuery's extend() method, which
is used to merge two objects together, and unlike most jQuery methods, it's called on the
jQuery object itself rather that a collection of HTML elements.
It takes four arguments; the first argument instructs the extend() method to deep-copy the
object being merged into the jQuery object, which we need to do because the defaults
object contains other objects.
The second argument is an empty object; any other objects will be merged together and
their own properties added to this object. This is the object that the method will return.
If we didn't pass in an empty object, the first object passed into the method would be
returned instead.
The next two arguments are the objects we are going to merge. These are the defaults
object that we defined a moment ago, and the opts object that may be passed to the
constructor when it is invoked.
This means that if a developer wishes to pass in a configuration object, they can overwrite
the values that we've defined in the defaults object. The properties not overwritten with
this configuration object will be set to the default values instead.
We also store a reference to the element, or collection of elements, as a member of the
instance as well so that we can easily operate on the elements in other parts of our code.
Lastly we add a couple of empty arrays that will be used to store the list of files to upload
and the XHR requests in progress. We'll see how these properties are used later in the
project, so don't worry too much about them now.
jQuery provides the fn object as a shortcut to its prototype, which is how we extend jQuery
with our plugin method. In this case the method is called up() and is the method we
invoked using the <script> element at the bottom of uploader.html. We specify that
the method may accept a single argument, which is an object containing the configuration
options that someone using our plugin may wish to provide.
122
Project 5
Within the method, we first create a new instance of the uploader using the new keyword in
conjunction with our constructor function. We pass the constructor function the element (or
collection of elements) that the method is called on, and the options object.
Lastly we returned this from the method. Inside a method added to jQuery's prototype,
the this object refers to the jQuery collection. It's very important to return the collection of
elements that the method was called on in order to preserve chaining.
Classified Intel
Chaining is an inherent feature of jQuery that developers using it have come to expect. It is
important that developers' expectations are met with regard to the style of programming
they use. People using our plugin will expect to be able to add additional jQuery methods
after calling our plugin's method.
Now that we are returning the collection of elements by returning the this object,
developers can do things like this:
$("#an-element").up().addClass("test");
So that's a simple example of what's possible, but it should illustrate why it is important to
always return this from a plugin.
Generating the underlying markup
In this task, we'll add an initialization method to our plugin which will generate the required
markup that the widget requires.
Engage Thrusters
First of all we should add the following code directly after the Up() constructor function in
uploader.js:
Up.prototype.init = function() {
var widget = this,
strings = widget.config.strings,
container = $("<article/>", {
"class": "up"
}),
heading = $("<header/>").appendTo(container),
title = $("<h1/>", {
text: strings.title
}).appendTo(heading),
drop = $("<div/>", {
123
jQuery File Uploader
"class": "up-drop-target",
html: $("<h2/>", {
text: strings.dropText
})
}).appendTo(container),
alt = $("<h3/>", {
text: strings.altText
}).appendTo(container),
upload = $("<input/>", {
type: "file"
}).prop("multiple", true).appendTo(container),
select = $("<a/>", {
href: "#",
"class": "button up-choose",
text: strings.buttons.choose
}).appendTo(container),
selected = $("<div/>", {
"class": "up-selected"
}).appendTo(container),
upload = $("<a/>", {
href: "#",
"class": "button up-upload",
text: strings.buttons.upload
}).appendTo(container);
widget.el.append(container);
}
We also need to call this new init() method. Change the method added to jQuery's fn
object so that it appears as follows:
$.fn.up = function(options) {
new Up(this, options).init();
return this;
};
We can also add the CSS for the markup generated by the plugin. In uploader.css, add the
following styling:
article.up { width:90%; padding:5%; }
article.up input { display:none; }
.up-drop-target {
height:10em; border:5px dashed #ccc; border-radius:5px;
margin-bottom:1em; text-align:center;
124
Project 5
}
.up-drop-target h2 {
margin-top:-.5em; position:relative; top:50%;
}
.up-selected { margin:1em 0; border-bottom:1px solid #ccc; }
Objective Complete - Mini Debriefing
We can add an init() method that takes care of creating and injecting the markup that the
widget is built from by adding it to our constructor's prototype. All objects created by the
constructor will inherit the method.
We first store the this object, which inside our init() method still refers to the jQuery
collection of elements, so that we can easily refer to it inside event handlers, which we'll add
in the next task.
We also locally scope the strings property to make resolution slightly faster, as we refer to
this property frequently in order to add the visible text strings to the widget's visible UI.
Next we create the new HTML elements and store each of them in variables. This means
that we can create the container and append all of the required elements to it while it's
still in memory, and then inject the entire widget into the page's DOM in one go, instead of
repeatedly modifying the DOM and adding the elements one by one.
The outer container for the widget is an <article> element which has a class name
for easy styling. The HTML5 specification describes an <article> as a self-contained
interactive widget, so I feel this is the perfect container for our widget. Although equally
as relevant, the <article> is not restricted to what we would traditionally describe as an
"article" – for example, a blog/news post or an editorial style piece.
We have a <header> element to contain the main heading for the widget, within which
we use a standard <h1>. We also use two <h2> elements inside the widget to show the
different sections (the drag-and-drop area, and the more traditional file <input>).
The <input> element has a type attribute of file and is also given the multiple
attribute, using jQuery's prop() method, so that multiple files can be uploaded in
supporting browsers. Current versions of IE (9 and below) do not support this attribute.
We also add an <a> element directly after the <input>, which we'll use to open the Open
dialog used to select the files to upload. The problem with the standard file type <input>,
is that there is no standard!
125
jQuery File Uploader
Almost every browser implements the file type <input> differently, with some browsers
showing an <input> as well as a <button> and some browsers just showing a <button>
and some text. It's also impossible to style the <input> or <button> generated by the
control as these are part of the shadow DOM.
For more information on what the shadow DOM is, see http://glazkov.
com/2011/01/14/what-the-heck-is-shadow-dom/.
To get around these cross-browser differences, we'll hide the <input> with CSS and use the
<a> element, styled to appear like an attractive button, to open the dialog.
We also add an empty <div> element that we'll use to list the selected files and display
some information about each one, followed by another <a> element that will be styled like a
button. This button will be used to initiate the upload.
We used the standard jQuery 1.4+ syntax for creating new HTML elements and supplied a
configuration object for most of the elements we created. Most elements are given a class
name and some also get text or HTML content. The class names we use are all scoped with a
sensible prefix so as to avoid potential conflicts with existing styles already used on the page.
The CSS we added is mostly just for presentation. The important aspects are that we hide
the standard file <input>, and give the drop target a fixed size so that files can easily be
dropped on it.
At this point we should be able to run the page in a browser (via a web server) and see the
basic elements and layout of the plugin. The page should appear as in the first screenshot of
this project.
Adding event handlers for receiving
files to upload
We can use the init() method that we added in the last task to attach the event handlers
that our widget will need to handle files being selected for upload. This may happen either
when files are dropped onto the drop target, or when they are selected using the button.
126
Project 5
Engage Thrusters
Directly after appending the new HTML elements to the container at the end of the init()
method in uploader.js (but still within the init() method), add the following code:
widget.el.on("click", "a.up-choose", function(e) {
e.preventDefault();
widget.el.find("input[type='file']").click();
});
widget.el.on("drop change dragover", "article.up", function(e) {
if (e.type === "dragover") {
e.preventDefault();
e.stopPropagation();
return false;
} else if (e.type === "drop") {
e.preventDefault();
e.stopPropagation();
widget.files = e.originalEvent.dataTransfer.files;
} else {
widget.files = widget.el
.find("input[type='file']")[0]
.files;
}
widget.handleFiles();
});
Objective Complete - Mini Debriefing
We first use jQuery's on() method, in event-delegation mode, to attach an event handler to
the outer container of the widget. We specify the click event as the first argument, and a
selector that matches our button with the class name up-choose as the second argument.
Within the handler function passed to on() as the third argument, we first prevent the
default behavior of the browser using JavaScript's preventDefault(), and then trigger a
click event for the hidden <input> element used to select the files to upload. This will
cause the File dialog to open in the browser and allow files to be selected.
We then attached another event handler. This time we are looking for either the drop,
dragover, or change events. The drop event will be triggered when files are dropped onto
the drop zone, the dragover event will be triggered while files are held over the drop zone,
and the change event will be triggered if files are removed.
127
jQuery File Uploader
All of these events will bubble up from either the drop zone (the <article> with the class
name up) or the hidden <input> and will pass through the outer container of the widget, to
which the event handler is bound.
Within this handler function we first check whether it's the dragover event; if it is,
we again prevent the default behavior of the browser using preventDefault() and
stopPropagation(). We also need to return false from this branch of the conditional.
The next branch of the if checks whether the event that triggered the handler was the drop
event. If it was we still need to use preventDefault() and stopPropagation(), but this
time we can also get the list of selected files using the event object that jQuery creates and
passes to the handler function, and store them in a property on the widget instance.
If neither of these conditions is true, we get the list of files from the <input>
element instead.
The property we need is part of the originalEvent object that jQuery packages into its
own event object. We can then get the files property from the dataTransfer object.
If the event was the change event instead, we can just get the files property of the
hidden <input>.
Whichever method is used, the collection of files selected for upload is stored on the
instance of the widget under the files property. This is just a temporary property, which
will get overwritten each time new files are selected, unlike the widget's filelist array,
which will store all the files for an upload.
Lastly we call the handleFiles() method. We'll add this method to the widget's
prototype in the next task so once this has been done, we'll be able to call the method
here without running into problems.
Combining the two events and detecting which event occurred in this way is much better
than attaching to separate event handlers. It means that we don't need two separate
handler functions that both do essentially the same thing, and regardless of whether files are
selected with the button and standard dialog, or by dropping files onto the drop target, we
can still get the list of files.
As this point we should be able to either drop files onto the drop zone, or click the button
and select files using the dialog. However, a script error will be thrown because we haven't
yet added the handleFiles() method of our plugin.
128
Project 5
Displaying the list of selected files
In this task we can populate the <div> we created in order to display the list of files that
have been selected for upload. We'll build a table where each row in the table lists a single
file with information such as the filename and type.
Engage Thrusters
Directly after the init() method in uploader.js, add the following code:
Up.prototype.handleFiles = function() {
var widget = this,
container = widget.el.find("div.up-selected"),
row = $("<tr/>"),
cell = $("<td/>"),
remove = $("<a/>", {
href: "#"
}),
table;
if (!container.find("table").length) {
table = $("<table/>");
var header = row.clone().appendTo(table),
strings = widget.config.strings.tableHeadings;
$.each(strings, function(i, string) {
var cs = string.toLowerCase().replace(/\s/g, "_"),
newCell = cell.clone()
.addClass("up-table-head "
+ cs)
.appendTo(header);
if (i === strings.length - 1) {
var clear = remove.clone()
.text(string)
.addClass("up-removeall");
newCell.html(clear).attr("colspan", 2);
} else {
newCell.text(string);
}
});
} else {
table = container.find("table");
129
jQuery File Uploader
}
$.each(widget.files, function(i, file) {
var fileRow = row.clone(),
filename = file.name.split("."),
ext = filename[filename.length - 1],
del = remove.clone()
.text("x")
.addClass("up-remove");
cell.clone()
.addClass("icon " + ext)
.appendTo(fileRow);
cell.clone()
.text(file.name).appendTo(fileRow);
cell.clone()
.text((Math.round(file.size / 1024)) + " kb")
.appendTo(fileRow);
cell.clone()
.html(del).appendTo(fileRow);
cell.clone()
.html("<div class='up-progress'/>")
.appendTo(fileRow);
fileRow.appendTo(table);
widget.fileList.push(file);
});
if (!container.find("table").length) {
table.appendTo(container);
}
}
We can also add some additional CSS for the new markup we've created. Add the following
code to the bottom of upload.css:
.up-selected table {
width:100%; border-spacing:0; margin-bottom:1em;
}
.up-selected td {
padding:1em 1% 1em 0; border-bottom:1px dashed #ccc;
font-size:1.2em;
}
.up-selected td.type { width:60px; }
130
Project 5
.up-selected td.name { width:45%; }
.up-selected td.size { width:25%; }
.up-selected td.remove_all_x { width:20%; }
.up-selected tr:last-child td { border-bottom:none; }
.up-selected a {
font-weight:bold; text-decoration:none;
}
.up-table-head { font-weight:bold; }
.up-remove-all { color:#ff0000; }
.up-remove {
display:block; width:17px; height:17px;
border-radius:500px; text-align:center;
color:#fff; background-color:#ff0000;
}
.icon {
background:url(../img/page_white.png) no-repeat 0 50%;
}
.doc, .docx {
background:url(../img/doc.png) no-repeat 0 50%;
}
.exe { background:url(../img/exe.png) no-repeat 0 50%; }
.html { background:url(../img/html.png) no-repeat 0 50%; }
.pdf { background:url(../img/pdf.png) no-repeat 0 50%; }
.png { background:url(../img/png.png) no-repeat 0 50%; }
.ppt, .pptx {
background:url(../img/pps.png) no-repeat 0 50%;
}
.txt { background:url(../img/txt.png) no-repeat 0 50%; }
.zip { background:url(../img/zip.png) no-repeat 0 50%; }
Objective Complete - Mini Debriefing
We started out by adding the handleFiles() method to the widget's prototype, making
the method call widget.handleFiles() that we added at the end of the last task work. It
was added in exactly the same way as the init() method earlier, and just like inside init(),
the this object points to the instance of the widget inside handleFiles(). This makes the
element on the page, the configuration options, and the selected file list easy to access.
Inside the method, we first created a series of variables. Like in the init() method we
create a local variable called widget that stores the this object. We won't be adding any
event handlers to this method so we don't absolutely have to do this, but we do access the
object several times so it makes sense to cache it in a variable.
131
jQuery File Uploader
We also cache the selected files container using widget.el – don't forget that el already
refers to the jQuery-wrapped instance of the outer widget container, so we can call jQuery
methods, such as find() directly on it without rewrapping it.
Next we create a series of new DOM elements ready for them to be cloned inside loops.
This is a much better way of creating elements, especially inside loops, and avoids having to
continually create new jQuery objects.
We also define a variable called table, but we don't initialize it straight away. Instead we
use the if conditional to check whether the container already contains a <table> element
by checking if jQuery's find("table") returns a collection that has length.
If length is equal to false, we know that no <table> elements were selected so we
initialize the table variable with a new <table> element created with jQuery. We then
create a header row for the <table> which we'll use to add headings for each of the
columns in the new table.
At this point the <table> element only exists in memory, so we can add the new row to it
without modifying the DOM of the page. We also cache a reference to the tableHeadings
property of the strings object used in our configuration object.
We then use jQuery's each() utility to create all of the <td> elements used as the table
headings. As well as being able to call each() on a collection of elements selected from the
page, we can also call each() on the jQuery object in order to iterate a pure JavaScript array
or object.
The each() method accepts the array or object to iterate. In this case it's an array, so the
iteration function called for each item in the array receives the index of the current item and
the value of the current item as arguments.
Inside the iterator, we first create a new string that we can use as a class name. The word
class is a reserved word in JavaScript, so we call our variable cs instead. To create the class
name, we just convert the current string to lowercase using JavaScript's toLowerCase()
function, and then remove any empty spaces using JavaScript's replace() function.
For a complete list of the reserved words in JavaScript, see the MDN
documentation at https://developer.mozilla.org/en-US/
docs/JavaScript/Reference/Reserved_Words.
The replace() function takes the regular expression to match as the first argument, and
the replacement string as the second argument. We could use the string " " as the first
argument instead, but then only the first blank space would be removed, whereas using a
regular expression with the g flag allows us to remove all spaces.
132
Project 5
We then create a new <td> element by cloning one of the elements we created and stored
in a variable at the start of the task. We give it a general class name for styling purposes, and
the unique class name that we just created so that each column can be styled independently
if required, and then append it straight to the header row that we created a moment ago.
We then check whether we're iterating the last item in the array by testing whether the
current index is equal to the length of the array minus 1. If it is the last item, we add a clear
all link by cloning the <a> element we created and cached at the start of the task.
We set the text of the new <td> element to the value of the current array item and add the
up-remove-all class for styling purposes so that we can filter events dispatched by it. We
can also use jQuery's attr() method to set a colspan attribute of 2 to this <td>. The new
<a> element is then added as the HTML content of the new <td> element.
If it's not the last item in the array we simply set the text content of the new <td> element
to the value of the current array item.
This is all done in the first branch of the outer if statement which occurs when the table
does not exist. If the container does already contain a <table> element, we still initialize
the table variable by selecting the <table> from the page.
Don't forget that the handleFiles() method we're inside will be invoked once files
have been selected, so now we need to build a new row in the table for each of the files
that were selected.
Again we use jQuery's each() method, this time to iterate the collection of stored files in
the widget's files property. For each file that was selected (either by dropping onto the
drop zone, or using the button) we first create a new <tr> by cloning our row variable.
We then split the name property of the current file on the . character. We store the
extension of the file by getting the last item in the array created by the split() function.
At this point we also create a delete link, which can be used to remove an individual file from
the list of files to upload, by cloning the <a> element we created at the start of the task. It is
given the text x and the class name up-remove.
Next we create a series of new <td> elements by cloning the cached <td> in the cell
variable again. The first <td> is given a generic class name of icon, and the extension for
the current file so that we can add an icon for the different file types that can be uploaded,
and append it to the new row.
The second <td> element displays the name of the file. The third <td> element shows the
size of the file in kilobytes. If we knew that large files were likely to be uploaded, we could
instead convert to megabytes, but for the purpose of this project, kilobytes will be sufficient.
133
jQuery File Uploader
The fourth <td> element has the new delete link added to it using jQuery's html() method
and the last <td> element has an empty <div> element added to it which we'll use for the
jQuery UI Progressbar widget.
Once the new cells have been created and appended to the new row, the new row itself
is appended to the table. We can also add the current file to our fileList array ready
to be uploaded.
Lastly we need to check once more whether the selected files container already has a
<table> element inside it. It if doesn't, we append the new <table> we have just built to
the container. If it does contain the <table> already, the new rows will already have been
added to it.
The CSS we added in this part was purely presentational. One thing I've done is to add
some classes so that an icon can be shown for different file types that might be selected for
upload. I've only added a few as an example; the ones you actually require would depend
on the type of files that you expected your users to upload. A generic icon is also created for
types that do not match the selectors we've added.
The icons used in this example are part of the Farm Fresh icon pack. I've
renamed the files for conciseness and can be found in the code download
accompanying this book. The icons are available at Fat Cow web hosting at
(http://www.fatcow.com/free-icons).
At this point we should be able to run the page in a browser, select some files to upload, and
see the new <table> we have just created:
134
Project 5
Classified Intel
We manually created the elements needed to display the list of selected files in this example.
Another way to do it would be to use a templating engine, such as jsRender or Dust.js,
instead. This would have the benefit of being faster and more efficient than our manual
creation, would keep our plugin code simpler and more concise, and the file smaller.
Of course, it would add another dependency to our plugin because we'd have to include
the templating engine itself, as well as a precompiled template stored in a JavaScript file.
We aren't creating that many elements in this example, so it's probably not worth adding
another dependency. When many elements need to be created, the cost of adding a
dependency is outweighed by the efficiency it adds.
This is the kind of thing that needs to be considered on a case-by-case basis when writing a
jQuery plugin.
Removing files from the upload list
In this task we'll add the event handlers that will make the Remove and Remove all links in
the new file list work. We can attach the event handlers in the same place that we added
other event handlers earlier to keep things organized.
Engage Thrusters
In upload.js, within the widget's init() method and directly after the existing calls to
jQuery's on() method, add the following new code:
widget.el.on("click", "td a", function(e) {
var removeAll = function() {
widget.el.find("table").remove();
widget.el.find("input[type='file']").val("");
widget.fileList = [];
}
if (e.originalEvent.target.className == "up-remove-all") {
removeAll();
} else {
var link = $(this),
removed,
filename = link.closest("tr")
.children()
.eq(1)
135
jQuery File Uploader
.text();
link.closest("tr").remove();
$.each(widget.fileList, function(i, item) {
if (item.name === filename) {
removed = i;
}
});
widget.fileList.splice(removed, 1);
if (widget.el.find("tr").length === 1) {
removeAll();
}
}
});
Objective Complete - Mini Debriefing
We use jQuery's on() method to add a click event again. We attach it to the outer container
of the widget as we have our other events, and this time we filter the events based on the
selector td a as the event will only originate from <a> elements inside <td> elements.
Inside the event handler we first prevent the default behavior of the browser because we
don't want the link to be followed. We then define a simple helper function that removes the
<table> element from the widget, removes the value of the file <input>, and clears the
fileList array.
We need to clear the <input> because otherwise if we selected some files and then
removed them from the list of files, we wouldn't be able to then reselect the same group of
files. It's a bit of a fringe case, but this simple little trick allows it to work so we may as well
include it.
Next we check what the className property of the element that triggered the event is.
We can see this property using the target property of the originalEvent object that is
included in the jQuery event object which is passed to our handler function. We could also
use the srcElement property of the jQuery event object, but this does not work in current
versions of Firefox.
When the className property matches up-remove-all, we simply call our
removeAll() helper function to remove the <table> element and clear the
<input> and fileList array.
136
Project 5
If the className property doesn't match the Remove all link, we have to remove just the row
of the <table> element that contains the <a> that was clicked. We first cache a reference to
the <a> that triggered the event, which is set to this inside our handler function.
We also define a variable called removed, which we'll initialize with a value shortly. Lastly we
store the filename of the file that the row we are about to remove represents.
Once we've set our variables, the first thing we do is remove the row that we can get using
jQuery's closest() method, which finds the first parent element that matches the selector
passed to the method.
We then use jQuery's each() method to iterate the fileList array. For each item in the
array, we compare the item's name property with the filename variable we just initialized.
If the two match, we set the index number, which is passed automatically to the iterator
function by jQuery, to our removed variable.
Once the each() method has finished, we can use JavaScript's splice() function to
remove the file the current <tr> represented. The splice() function takes two arguments
(it can take more but we don't need them here), where the first argument is the index of the
item to begin removing at, and the second argument is the number of items to remove.
Lastly, we check whether the <table> element has more than one row left in it. If it only
has one row left, this will be the header row so we know that all files have been removed.
We can therefore call our removeAll() helper function to tidy up and reset everything.
Now when we've added files to the upload list, we should then be able to remove individual
files using the inline x buttons, or clear the list using the Remove all link.
Adding a jQuery UI progress
indicator
In this task we'll add the elements and initialization code required by the jQuery UI
Progressbar widget. The widget won't actually do anything yet because we won't be
actually uploading anything until the next task, but we need to wire everything up ready.
Engage Thrusters
We'll add an initProgress() method to our widget's prototype to select the <div>
elements that we added to the <table> element and convert them into Progressbar
widgets. We can also add the method that will be used to update the Progressbars.
137
jQuery File Uploader
Directly after the handleFiles() method, add the following code:
Up.prototype.initProgress = function() {
this.el.find("div.up-progress").each(function() {
var el = $(this);
if (!el.hasClass("ui-progressbar")) {
el.progressbar();
}
});
}
Next, we'll need to call this method after new rows have been added to <table>. Add the
following call to our new method right at the end of the handleFiles() method:
widget.initProgress();
Now we can add the code that updates the Progressbar. Add the following code directly after
the initProgress() method we just added:
Up.prototype.handleProgress = function(e, progress) {
var complete = Math.round((e.loaded / e.total) * 100);
progress.progressbar("value", complete);
}
We also need a tiny bit of CSS for the new progress bars. Add the following code to the end
of uploader.css:
.up-progress {
height:1em; width:100px; position:relative; top:4px;
}
Objective Complete - Mini Debriefing
This was a shorter task than some of those we've covered in this project so far, but no
less important. We added the new method in the same way that we've added most of the
functionality for our plugin.
Inside the method we first select all <div> elements with the class name up-progress.
Don't forget that we can access the widget's container element using this.el and as it's a
jQuery object already we can call jQuery methods, such as find() on it.
138
Project 5
We then iterate each element in the selection using jQuery's each() method. We're using
the standard each() method in this task, where the current element in the collection is set
to this inside the iterator function.
In the iterator function we first cache the current element. We then check whether it has
the jQuery UI class name ui-progressbar and if it doesn't, we convert the element into a
Progressbar using the jQuery UI method progressbar().
Doing it this way means that the progress bar will always be created, whether it is the
initial set of files being selected for uploading, or additional files that are being added
to an existing <table>.
We also added a call to the new initProgress() method at the end of the
handleFiles() method, which is called whenever new files are selected for upload.
Next we added the handleProgress() method, which we'll bind to an event in the next
task. This method will be passed two arguments, the first is the event object and the second
is an already-wrapped jQuery object representing an individual Progressbar.
Within the method, we first calculate how much of the file has been uploaded. We can
determine this by dividing the loaded property of the event object by the total property, and
then dividing by 100 to give us the percentage of the file that has been uploaded so far.
The loaded and total properties are special properties that are added to the event object
when the progress event is fired by the browser.
Once we have the percentage, we can call the value method of the Progressbar widget
in order to set the value to the percentage. This is a jQuery UI method and so is called in
a special way. Rather than calling value() directly, we instead call the progressbar()
method, and pass the name of the method to invoke, value, as the first argument. All
jQuery UI methods are called in this way.
Lastly, we added a little presentational CSS just to tweak the default styling provided by the
default jQuery UI theme in use. When we add files to be uploaded now, we should see an
empty Progressbar after each file in the <table>.
Uploading the selected files
We now have a list of files attached to the instance of our plugin ready to be uploaded. In
this task we'll do just that and upload the files asynchronously using jQuery. This behavior
will be tied to the Upload files button that we added to the markup generated by the plugin.
We can also use this task to update our Progressbars with the current progress of each file
being uploaded.
139
jQuery File Uploader
Engage Thrusters
As this is another event handler, we'll add it in the init() method along with all of the
other event handlers so that they're all kept in one place. Add the following code at the end
of the init() method, after the existing event handlers:
widget.el.on("click", "a.up-upload", function(e) {
e.preventDefault();
widget.uploadFiles();
});
Next add the new uploadFiles() method. This can go after the progress-related methods
that we added in the last task:
Up.prototype.uploadFiles = function() {
var widget = this,
a = widget.el.find("a.up-upload");
if (!a.hasClass("disabled")) {
a.addClass("disabled");
$.each(widget.fileList, function(i, file) {
var fd = new FormData(),
prog = widget.el
.find("div.up-progress")
.eq(i);
fd.append("file-" + i, file);
widget.allXHR.push($.ajax({
type: "POST",
url: "/upload.asmx/uploadFile",
data: fd,
contentType: false,
processData: false,
xhr: function() {
var xhr = jQuery.ajaxSettings.xhr();
if (xhr.upload) {
140
Project 5
xhr.upload.onprogress = function(e) {
widget.handleProgress(e, prog);
}
}
return xhr;
}
}));
});
}
}
Objective Complete - Mini Debriefing
Within our uploadFiles() method, we first store a reference to the widget as we have in
some of the other methods we've added. We also store a reference to the Upload files button.
The next thing to do is to check that the button doesn't have class name of disabled. If
it does have this class name, it means that an upload has already been initiated for the
selected files so we want to avoid making a duplicate request. If the button doesn't have the
disabled class, it means that this is the first time the button has been clicked. So to prevent
duplicate requests we then add the class disabled.
Next we iterate over the list of files that we've collected and which is stored in the
widget instance's fileList property. For each file in the array we first create a
new FormData object.
FormData is a part of the new XMLHttpRequest (XHR) level 2 specification which allows us
to dynamically create a <form> element and submit that form asynchronously using XHR.
Once we've created a new FormData object, we also store a reference to the Progressbar
widget associated with the current file. We then use the FormData's append() method to
append the current file to the new FormData object so that the file is encoded and sent to
the server.
Next we post the current FormData object to the server using jQuery's ajax() method. The
ajax() method will return the jqXHR object for the request. This is a special version of the
XHR object that jQuery enhances with additional methods and properties. We need to store
this jqXHR object so that we can use it a little later on.
We'll look at exactly how it's used in the next task, but for now just understand that the
jqXHR object returned by the ajax() method is pushed into the allXHR array that we
stored as a member on the widget instance right at the start of the project.
141
jQuery File Uploader
The ajax() method accepts a configuration object as an argument which allows us to
control how the request is made. We set the request to POST using the type option, and
specify the URL to post to using the url option. We add the FormData object as the payload
of the request using the data option and set the contentType and processData options
to false.
If we don't set the contentType option to false, jQuery will try to guess which content
type should be used for the request, which it may or may not do correctly, meaning some
uploads will work and some will fail, seemingly for no apparent reason. The content-type
of the request will be set to multipart/form-data by default as we are using FormData
which have files appended to them.
Setting the processData option to false will ensure that jQuery doesn't try to transform
the file into a URL-encoded query string.
We need to modify the underlying XHR object used to make the request, so that we can
attach our handler function to the progress event. The handler must be bound to the event
before the request is made and the only way to do that currently is using the xhr option.
The option takes a callback function which we can use to modify the original XHR object and
then return it for the request to be made. Inside the callback, we first store the original XHR
object, which we can get from jQuery's ajaxSettings object.
We then check whether the object has an upload property, and if it does we set an
anonymous function as the value of onprogress. Within this function, we simply call the
handleProgress() method of our widget that we added in the last task, passing it the
progress event object and the Progressbar widget that we stored at the start of this task.
Reporting success and tidying up
In this task we need to show when each file has finished uploading. We also need to
clear the <table> from the widget and re-enable the upload button once all of the
uploads have completed.
Engage Thrusters
We can show when each individual file has completed uploading using jQuery's done()
method, which we can chain after the ajax() method that we added in the last task:
.done(function() {
var parent = prog.parent(),
142
Project 5
prev = parent.prev();
prev.add(parent).empty();
prev.text("File uploaded!");
});
In order to tidy up following the uploads, we can make use of jQuery's when() method. We
should add the following code directly after the each() method in uploadFiles():
$.when.apply($, widget.allXHR).done(function() {
widget.el.find("table").remove();
widget.el.find("a.up-upload").removeClass("disabled");
});
Objective Complete - Mini Debriefing
Because jQuery's ajax() method returns a jqXHR object, and because this object is a
special object called a promise object, we can call certain jQuery methods on it. The done()
method is used to execute code when the request completes successfully.
You may be more used to using jQuery's success() method to handle
successful AJAX requests, or the error() or complete() methods.
These methods have been removed from the library in version 1.9, so we
should use their replacements done(), fail(), and always() instead.
Within this function all we need to do is remove the clear button and the Progressbar
widget for the file that has just finished uploading. We can find the elements that need
to be removed easily by navigating to them from the current Progressbar widget.
We stored a reference to each individual Progressbar in the last task and because the
done() method is chained to the ajax() method, we can still access this element using
the variable after the request has completed.
Notice that there appears to be an extra closing bracket at the end of the done() method.
The reason for this is because it's still within the push() method that we added in an earlier
task. It's critical that the done() method is added to the correct place – it must be chained
to the ajax() method inside the push() method.
Once these elements have been removed, we add a simple message that says the file has
finished uploading.
We also need to remove the <table> element from the page once all the requests have
been completed. It was for this reason that we stored all of the jqXHR objects generated
when uploading the files in the last task. We can use jQuery's when() method to do this.
143
jQuery File Uploader
The when() method can accept a series of promise objects and returns when they have all
been resolved. However, this method doesn't accept an array, which is why we call it using
JavaScript's apply() method instead of calling it normally.
We can again use the done() method to add a callback function to be invoked once the
when() method has returned. Within this callback, all we do is remove the <table>
element showing the files that were uploaded and re-enable the upload button by removing
the disabled class.
This is all we need to do to actually upload the file or files that were selected and receive
progress feedback on each file individually, as shown in the following screenshot:
Viewing the example file
To see this project in action you'll need to view the page we've
created using a web server (using http://localhost on your
own computer). It won't work if you open the file by double-clicking
on it in Explorer or Finder.
144
Project 5
Mission Accomplished
We've made it to the end of the project. At this point we should have an uploader plugin that
is easy to use and provides rich features in supporting browsers such as multiple files, file
information, an editable upload list, and upload progress reports.
Not all browsers are able to use the features that this widget is built
to exploit. The Opera browser for example, sees programmatically
triggering the file dialog box as a security risk and so does not allow it.
Also, legacy versions of Internet Explorer (anything prior to version 10)
will not be able to handle this code at all.
Supporting incompatible or legacy browsers is beyond the scope of this example, but
it would be relatively straight forward to add a fallback that made use of some other
technology, such as Flash, in order to provide support for some of the behavior our
plugin demonstrates.
Or there are a range of older jQuery plugins that make use of <iframe> elements to
simulate uploading files via AJAX. Instead of focusing on what isn't supported, I've chosen to
focus on what can be done in supporting browsers.
You Ready To Go Gung HO?
A Hotshot Challenge
By uploading the files individually, we were able to add an event handler to monitor the
progress of the files being uploaded. This also opens up of the possibility of aborting the
upload of individual files.
For this challenge why not see if you can add a mechanism for canceling the upload of a file.
We already have the remove button which is used to remove files before they are uploaded.
These could easily be updated so that they cancelled the upload if the upload is in progress.
A handler for the abort event can be added to the XHR object in the same way as the
progress event handler was attached, so it should be easily achievable.
145
Project 6
Extending Chrome
with jQuery
Building an extension for Chrome (or any other browser that can be extended with plugins
and extensions) is an easy way to create custom behavior or additional tools to enhance our
browsing experience.
Chrome allows us to leverage our web development skills to extend its browser interface,
using technologies we're already familiar with such as HTML, CSS, and JavaScript, and where
you can use JavaScript you can usually use jQuery.
Mission Briefing
In this project we'll build a Chrome extension that highlights elements on the page that
are marked up with Schema.org microdata. Microdata is a way of specifying descriptive
information about a variety of different entities, such as businesses, locations, or people
using standard HTML attributes, and is rumored to become an important factor in Google's
ranking algorithms.
Whenever we visit a page containing elements described as contact details, we can grab
them from the page and store them in our extension, allowing us to slowly build up a
directory of contact information for people that are using or making stuff we love.
Extending Chrome with jQuery
We can also use templating in this project to make creating a repeated group of elements
much more efficient, as well as easier to maintain. We used JsRender in the previous project,
so we can use it again, but this time we'll need to use it slightly differently than before. Once
completed, our extension will look something like that shown in the following screenshot:
Why Is It Awesome?
Microdata is used to describe the information contained within web pages in order to
promote better interoperability between autonomous systems, such as search engine
spiders and HTML documents.
When different elements on the page are described as being a company, a person, a product,
or a movie, it allows things such as search engines to better understand the information
contained on the page.
Microdata is rapidly becoming more commonplace on the Web and is set to take an ever
increasingly important role in the results generated by Google for search results, so there has
never been a better time to exploit it.
148
Project 6
Your Hotshot Objectives
The tasks this project is broken down into are as follows:
ff
Setting up the basic extension structure
ff
Adding a manifest and installing the extension
ff
Adding a sandboxed JsRender template
ff
Posting a message to the sandbox
ff
Adding a content script
ff
Scraping the page for microdata
ff
Adding a mechanism for saving the microdata
Setting up the basic extension
structure
In this task we'll create the underlying files required by the extension. All files used by an
extension need to reside in the same directory, so we'll set that up and make sure it contains
all the files we need.
Prepare for Lift Off
One thing I should point out, although hopefully you'll already have realized – we're going
to require the Chrome browser for the duration of this project. If you don't have it installed,
which as a web developer you really ought to, if only for testing purposes, get it and install
it immediately.
The latest version of Chrome can be downloaded from
https://www.google.com/intl/en/chrome/browser/.
We'll keep all of the files for this project in a single directory, so set one up in the project
folder now and call it chrome-extension. The extension will be built from the same base
code files as most of the other projects that we've created have been built from; the only
difference is that all files will need to be local to the extension.
We're going to need a copy of JsRender, so we should also download a copy of that and place
it in the chrome-extension directory. Last time we used JsRender we linked to the live
hosted version. This time we'll download it.
149
Extending Chrome with jQuery
The latest version of JsRender can be downloaded from
https://github.com/BorisMoore/jsrender/.
We can use the template file that we've started the other projects with, but we should make
sure the paths to jQuery, the JavaScript file, and the style sheets all point to files in the same
directory. All of the files used by a Chrome extension must be in the same folder, which is
why we download the scripts instead of linking to online versions.
We should place copies of jQuery, JsRender, and the common.css style sheet into the new
directory. We also need to create a new JavaScript file called popup.js and a new style
sheet called popup.css and save these files into the new directory also.
Lastly, we can create a new HTML page called popup.html. This file should also be saved in
the chrome-extension directory, and should contain the following code:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<title>jQuery-Powered Chrome Extension</title>
<link rel="stylesheet" href="common.css" />
<link rel="stylesheet" href="popup.css" />
</head>
<body>
<script src="jquery-1.8.0.min.js"></script>
<script src="jsrender.js"></script>
<script src="popup.js"></script>
</body>
</html>
Engage Thrusters
The HTML file we just created will be used as the extension's popup. This is the page that
is displayed as a popup when the extension's icon is clicked in the toolbar. In this project
we'll be creating a type of extension known as a browser action, which automatically adds a
button to Chrome's toolbar which is used to open the popup.
The popup will display a button used to trigger a scan of the current page for microdata and
display any previously saved contacts. Any previously stored contacts will be retrieved using
the localStorage API, and we can use a template to render them.
150
Project 6
First we can add the general markup to the page; in popup.html add the following code to
the <body> of the page:
<section role="main">
<header>
<h1>Web Contacts</h1>
</header>
<ul id="contacts"></ul>
</section>
<iframe id="poster" src="template.html"></iframe>
We can also add some basic styling for these elements. In popup.css, add the
following code:
body { width:32em; padding:0 2em; }
header { padding-top:2em; }
ul { padding:0 0 1em; font-size:1.5em; }
iframe { display:none; }
Objective Complete - Mini Debriefing
Chrome extensions are built using the same files that as web developers we're used to
working with – HTML, CSS, and JavaScript. The extension will add a button to the toolbar and
when this button is clicked, it will display a popup. The HTML page that we added in this task
is the basis of this popup.
We create the page like we would create any other standard HTML5 page. We linked to the
CSS and JavaScript files as usual, and then added a small <section> container which will be
used as a container for any previously saved contacts. Initially there won't be any, and when
there are, we'll render them using a template.
We've added a <header> containing a <h1> to give the saved contacts a title, and have
added an empty<ul> element that we'll populate shortly with a script.
Lastly we added an <iframe> to the page, which will be hidden from view. We'll use this to
communicate with another part of the extension a little later on. The src attribute of the
element is set to the page we want to send messages to.
The CSS we added was purely for presentation and just laid out the initial elements in a simple
layout. We're also linking to the common CSS file that each of the other projects have also
used, but don't forget, all files used by the extension must be in the extension's directory.
151
Extending Chrome with jQuery
Classified Intel
As we are creating a browser action we'll get a new button added to Chrome's toolbar which
will be visible as long as the unpacked extension is loaded. By default it will have the standard
extension icon – a puzzle piece, but we can replace this with an icon of our own creation.
We can also create other types of extension which don't add a button to the toolbar. We
could create a page action instead of a browser action, which adds an icon to the address bar
instead of the toolbar.
Whether this icon is visible on all pages or not will depend on how the extension behaves.
For example, if we wanted to run our extension every time a page was loaded in the browser,
but only display the icon if it found Schema.org microdata on the page, we could use a
page action.
A browser action, such as we'll create here, is accessible all of the time, regardless of the
page being viewed. We're using a browser action instead of a page action because the users
of our extension will probably want to be able to see contacts they have previously found
and saved, so a browser action is perfect for facilitating an always-available button to display
any data stored by the extension.
Adding a manifest and installing the
extension
In order to actually install our extension and see the fruits of our labor so far, we'll need to
create a manifest file. This special file, saved in JSON format, controls certain aspects of the
extension such as the pages it uses, and the content scripts it can run.
Prepare for Lift Off
In a new file add the following code:
{
"name": "Web Contacts",
"version": "1.0",
"manifest_version": 2,
"description": "Scrape web pages for Schema.org micro-data",
"browser_action": {
"default_popup": "popup.html"
}
}
Save this file in the chrome-extension directory that we created at the start of the task
within our main project directory as manifest.json.
152
Project 6
If the text editor you're using doesn't show .json in the Save as type:
(or equivalent) drop-down, select the All types (*) option and type the
full filename manifest.json in the File name: input field.
Engage Thrusters
In order to view the extension as it currently exists, it will need to be loaded in Chrome as an
extension. To do this, you should go to Settings | Tools | Extensions.
In recent versions of Chrome, the Settings menu is accessed by clicking
on the button that has three stripes as its icon, which is situated at the
top-right of the browser window.
When the extensions page loads, there should be a button to Load unpacked extension…. If
there isn't, tick the Developer mode checkbox and the button will appear.
Hit the button and select the chrome-extension folder as the extension directory. This
should install the extension and add the browser action button to the toolbar for us.
Objective Complete - Mini Debriefing
The simple manifest file is required by the extension before it can be loaded in the browser.
Only manifests of at least Version 2 are allowed by the current version of Chrome. The
manifest is required and the extension will not run without it. It is a simple text file, in JSON
format, that is used to tell the browser some of the basic information about the extension,
such as its name, who authored it, and the current version.
We can specify that our extension is a browser action, which adds a button for the extension
to Chrome's toolbar. We can also specify the page that will be displayed in the popup using
the manifest.
When the new button for our extension is clicked, the HTML page (popup.html) we added in
the previous task will be displayed in the extension popup, as shown in the following screenshot:
153
Extending Chrome with jQuery
Adding a sandboxed JsRender
template
In this task we can add the template that JsRender will use to display the saved contacts. At
this point we don't have any saved, but we can still wire it up ready, and then when we do
have some contacts, they'll be rendered into the popup without any fuss.
Prepare for Lift Off
Chrome uses a Content Security Policy (CSP) in order to prevent a large number of common
cross-site scripting (XSS) attacks, and because of this we are not allowed to execute any
scripts that use either eval() or new Function().
The JsRender templating library, like many other popular libraries and frameworks, uses new
Function() when compiling templates and therefore is not allowed to run directly inside
the extension. There are two ways we can overcome this problem:
ff
We could switch to a templating library that offers a pre-compilation of templates,
such as the popular Dust.js. We could then compile our template outside of the
extension in a browser and link to a JavaScript file containing the function that the
template is compiled into from within the extension. The function that would be
created with new Function() would then already have been created prior to the
extension even being installed, and the template could then be rendered inside the
extension, interpolated with any data made available within the extension.
ff
Or, Chrome's extension system allows us to use certain files inside a specified
sandbox. Unsafe string-to-function features such as eval() or new Function()
are allowed to run in the sandbox because the code is insulated from the extension's
data and API access within the browser.
We'll use the sandboxing feature in this example, so that we can continue using JsRender.
Engage Thrusters
First of all we have to set up the sandbox, which is done by specifying which pages
to sandbox using the manifest file that we created earlier. Add the following code to
manifest.json, directly before the final closing curly bracket:
"sandbox": {
"pages": ["template.html"]
}
154
Project 6
Don't forget to add a comma directly after the closing curly
bracket of the browser_action property.
We've specified template.html as the page to sandbox. Create a new file called
template.html and save it in the chrome-extension directory. It should contain the
following code:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<script id="contactTemplate" type="text/x-jsrender">
{{for contacts}}
<li>
<article>
<div class="details">
<h1>{{:name}}</h1>
{{if url}}
<span>website: {{url}}</span>
{{/if}}
{{if jobTitle}}
<h2>{{:jobTitle}}</h2>
{{/if}}
{{if companyName}}
<span class="company">
{{:companyName}}
</span>
{{/if}}
{{if address}}
<p>{{:address}}</p>
{{/if}}
{{if contactMethods}}
<dl>
{{for
~getMembers(contactMethods)}}
<dd>{{:key}}</dd>
<dt>{{:val}}</dt>
{{/for}}
</dl>
{{/if}}
</div>
</article>
155
Extending Chrome with jQuery
</li>
{{/for}}
</script>
<script src="jquery-1.9.0.min.js"></script>
<script src="jsrender.js"></script>
<script src="template.js"></script>
</head>
</html>
The template page also references the template.js script file. We should create this file in
the chrome-extension directory and add the following code to it:
(function () {
$.views.helpers({
getMembers: function (obj) {
var prop,
arr = [];
for (prop in obj) {
if (obj.hasOwnProperty(prop)) {
var newObj = {
key: prop,
val: obj[prop]
}
arr.push(newObj);
}
}
return arr;
}
});
} ());
Objective Complete - Mini Debriefing
We started out by adding a new HTML page to the extension. The page called template.
html is like a regular web page except that it doesn't have a <body>, just a <head> that
links to some JavaScript resources, and a <script> element containing the template we'll
be using.
156
Project 6
Usually in a Chrome extension, the CSP prevents us from running any inline
scripts – all scripts should reside within external files. Using a non-standard
type attribute on the <script> element allows us to circumvent this so
that we can store our template inside the page instead of having another
external file.
The bulk of the new page is the template itself. Schema.org microdata allows people to
add a lot of additional information to describe the elements on the page, so there are a
range of different bits of information that may be stored in the extension.
Our template therefore makes use of a lot of conditionals to display things if they are
present. The extension should always display the name, but aside from that it may
display an image, a job title and company, an address, or various contact methods,
or any combination thereof.
The most complex part of the template is the getMembers() helper function. We'll call this
helper function for each object in the contactMethods object using JsRender's {{for}}
tag, which calls the helper function using the tilde (~) character. Inside the loop we'll have
access to the values returned by the helper and can insert these into the relevant elements.
Next we added the template.js script file. At this point all we need to add to this script file
is the helper method used by the template to render any contact methods. These will be in
the format { email: [email protected] }.
The helper is registered using JsRender's helpers() method. This method is passed an
object where the name of the helper is specified as the key and the function that should be
invoked is the value.
The function receives an object. We first create an empty array and then iterate the
object using a standard for in loop. We first check that the property being iterated
belongs to the object and is not inherited from the prototype using JavaScript's
hasOwnProperty() function.
We then just create a new object and set the key as a property called key, and the value as a
property called val. These are the template variables that we use in the <dl> in our template.
This new object is then pushed into the array we created, and once the object passed
to the helper has been iterated, we return the array to the template for the {{for}}
loop to iterate.
157
Extending Chrome with jQuery
Posting a message to the sandbox
In this task we'll set up the communication between our popup and the sandboxed template
page to see how we can get the template to render when the popup is opened.
Engage Thrusters
First of all we can add the code that sends the message to the sandboxed page requesting
the template to render. In popup.js, add the following code:
var iframe = $("#poster"),
message = {
command: "issueTemplate",
context: JSON.parse(localStorage.getItem("webContacts"))
};
iframe.on("load", function () {
if (message.context) {
iframe[0].contentWindow.postMessage(message, "*");
} else {
$("<li>", {
text: "No contacts added yet"
}).appendTo($("#contacts"));
}
});
window.addEventListener("message", function (e) {
$("#contacts").append((e.data.markup));
});
Next we need to add the code that will respond to the initial message. Add the following
code to template.js directly after the helper method we added in the last task:
var template = $.templates($("#contactTemplate").html());
window.addEventListener("message", function (e) {
if (e.data.command === "issueTemplate") {
var message = {
markup: template.render(e.data.context)
};
e.source.postMessage(message, event.origin);
}
});
158
Project 6
Objective Complete - Mini Debriefing
First of all we set up the initial messaging in popup.js. We cached the <iframe> element
from the popup in a variable and then composed a message. The message is in the form of
an object literal with a command property and a context property.
The command property tells the code running in the <iframe> what to do, while the
context contains the data to be rendered into a template. The data that we'll render will be
stored in localStorage under the webContacts key, and the data will be in JSON format so
we need to convert it back to a JavaScript object using JSON.parse().
We then add a load handler to the <iframe> element using jQuery's on() method. The
code contained in the anonymous function passed to on() will be executed when the
contents of the <iframe> have loaded.
Once this occurs, we check whether the context property of the message object
has a truthy value. If it does, we post the message object to the <iframe> using the
postMessage() function of the iframe's contentWindow property.
The postMessage() function takes two arguments – the first is what to post, which in this
case is our message object, and the second argument specifies which files can receive the
message. We set this to a wildcard * so that any files can subscribe to our message.
If there aren't any stored contacts, the context property of our message object will have
the falsey value null. In this case we simply create a new <li> element with a text message
advising that there are no saved contacts and append this directly to the empty <ul>
hardcoded into popup.html.
Our script file popup.js will also need to receive messages. We use the standard JavaScript
addEventListener() function to attach a listener for message events to the window. By
default jQuery does not handle message events.
The messages received by popup.js will be the response from the sandboxed page
containing the HTML markup to render. The markup will be contained in a property called
markup, which will be contained in the data property of the event object. We simply select
the <ul> element in popup.html and append the markup we receive.
We also added some code to template.js, the script file referenced by the page inside
our <iframe>. We used the addEventListener() function here to subscribe to message
events again.
This time we first check whether the command property of the object sent as the message
is equal to issueTemplate. If it is, we then create and render the data into our JsRender
template and compose a new message object containing the rendered template markup.
159
Extending Chrome with jQuery
Once the message object has been created we post it back to popup.js. We can get
the window object to send the message using the source property of the event object,
and we can specify which files can receive the message using the origin property of
the event object.
These two properties are very similar except that source contains a window object and
origin contains a filename. The filename will be a special chrome extension name. At this
point we should be able to launch the popup and see the No contacts message as we don't
have any saved contacts yet.
Adding a content script
We're now at the stage where everything is in place to display stored contacts, so we
can focus on actually getting some contacts. In order to interact with pages that the user
navigates to in their browser, we need to add a content script.
A content script is just like a regular script, except that it interacts with the page being
displayed in the browser instead of with the files that make up the extension. We'll see
that we can post messages between these different areas (the page in the browser and the
extension) in a similar way that we posted a message to our sandbox.
Engage Thrusters
First we'll need to add some new files to the chrome-extension directory. We'll need a
JavaScript file called content.js and a style sheet called content.css. We need to tell
our extension to use these files, so we should also add a new section to the manifest file
(manifest.json) we created earlier in the project:
"content_scripts": [{
"matches": ["*://*/*"],
"css": ["content.css"],
"js": ["jquery-1.9.0.min.js", "content.js"]
}]
This new section should be added directly after the sandbox section that we added earlier
(as before, don't forget to add the trailing comma after the sandbox property).
160
Project 6
Next we can add the required behavior to content.js:
(function () {
var people = $("[itemtype*='schema.org/Person']"),
peopleData = [];
if (people.length) {
people.each(function (i) {
var person = microdata.eq(i),
data = {},
contactMethods = {};
person.addClass("app-person");
});
}
} ());
We can also add some basic styling that highlights any elements that contain microdata
attributes using the content.css style sheet. Update this file now so that it contains the
following code:
.app-person {
position:relative; box-shadow:0 0 3px rgba(0,0,0, .5);
background-color:#fff;
}
Objective Complete - Mini Debriefing
First of all we updated our manifest file to include a content script. As I mentioned, content
scripts are used to interact with the visible page being displayed in the browser instead of
any files used by the extension.
We can enable a content script using the content_script rule in the manifest. We need
to specify which pages the content script should be loaded into. We use a wildcard (*) for
the protocol, host, and path portions of URLs so that the script is loaded when any
page is visited.
When using Schema.org microdata to describe people, the different bits of information
that are present are placed within a container (usually a <div> element although any
element can be used) that has the special attribute itemtype.
161
Extending Chrome with jQuery
The value of this attribute is a URL that specifies what the data the elements it contains
describe. So to describe a person, this container would have the URL http://schema.org/
Person. This means that the elements the container has in it may have additional attributes
that describe a specific piece of data, such as a name, or a job title. These additional
attributes on the elements within the container will be itemprop.
In this case we're using a jQuery attribute-contains selector (*=) to attempt to select
elements containing this attribute from the page. If the array the attribute selector returns
has length (and is therefore not empty), we know that at least one of these elements exists
on the page and so can process the element further.
The collection of elements with this attribute are stored in a variable called people. We
also create an empty array in the variable peopleData ready to store all of the information
about all of the people found on the page.
We then use jQuery's each() method to iterate the elements selected from the page.
Instead of using $(this) inside our each() loop, we can use the collection of elements
that we've already selected from the page, in conjunction with jQuery's eq() method along
with the current index of the loop in order to reference each element, which we store in a
variable called person.
We also create an empty object and store it in a variable called data ready to store the
microdata for each person, and an empty object called contactMethods as any microdata
for a telephone number or e-mail address needs to be added to a subobject for our template
to consume.
All we do at this point is add a new class name to the container element. We can then use
the content.css style sheet to add some very basic styling to the element so that it is
brought to the user's attention.
Scraping the page for microdata
Now that we've got our content script in place, we can interact with any web page that the
user of the extension visits and check whether it has any microdata attributes.
At this point, any element containing microdata is highlighted to the user, so we need to add
the functionality that will allow the user to view the microdata and save it if he/she wishes,
which is what we'll be covering in this task.
162
Project 6
Engage Thrusters
Directly after where we add a class name to each element that has an itemtype attribute in
content.js, add the following code:
person.children().each(function (j) {
var child = person.children().eq(j),
iProp = child.attr("itemprop");
if (iProp) {
if (child.attr("itemscope") !== "") {
if (iProp === "email" || iProp === "telephone") {
contactMethods[iProp] = child.text();
} else {
data[iProp] = child.text();
}
} else {
var content = [];
child.children().each(function (x) {
content.push(child.children().eq(x).text());
});
data[iProp] = content.join(", ");
}
}
});
var hasProps = function (obj) {
var prop,
hasData = false;
for (prop in obj) {
if (obj.hasOwnProperty(prop)) {
hasData = true;
break;
}
}
return hasData;
163
Extending Chrome with jQuery
};
if (hasProps(contactMethods)) {
data.contactMethods = contactMethods;
}
peopleData.push(data);
Objective Complete - Mini Debriefing
In the last task we added a class name to each container of elements marked up with microdata.
In this task we're still in the context of the each() loop that processed each container.
So in the code we added in this task we first call each() again, this time on the direct children
of the container element; we can get these easily using jQuery's children() method.
Inside this each() loop we first get the current item from the existing cached person
variable using the loop counter passed to our iteration function (j) as an argument for
jQuery's eq() method. This avoids creating a brand new jQuery object inside our loop.
We also store the value of the current element's itemprop attribute in a variable called
iProp because we'll need to refer to it a number of times and using a nice short variable
means less typing for us.
At this point we don't know whether we're dealing with a regular element, or an element
containing microdata, so we use an if statement that checks whether the iProp variable
we just set has a truthy value. If the element doesn't have an itemprop attribute this
variable will hold an empty string, which is falsey and stops the code progressing further if
the element is just a regular element.
Inside this conditional we know we're dealing with an element containing microdata, but
there are different formats that the data may take. If the element contains an address
for example, it won't have any content directly, but will have child elements of its own
containing the data instead. In this case, the element will have an itemscope attribute.
First we want to process the elements that don't contain an itemscope attribute so the
first branch of our nested conditional checks the value returned by selecting the itemscope
attribute is not an empty string.
If you remember our template, we set up a helper function that displayed contact
information using an object. In order to create this new object instead of creating a new
property of the data object, we use another nested if statement to check whether the
iProp variable contains an e-mail or telephone number.
164
Project 6
If it does, we add the value of the iProp variable as the key of the contactMethods object,
and the text of the element as the value. If the iProp variable doesn't contain an e-mail
address or a telephone number, we set the iProp variable as a key of the data object, and
its value as the element's content.
The next branch of the second nested if statement is for elements that do have an
itemscope attribute. In this case, we first define an empty array and store it in a variable
called content. We then iterate over the child elements using jQuery's each() method and
push the text content of each element into the content array.
Once we have iterated over the child elements and populated the array, we can then add
the current iProp variable and the data in the content array to our data object. Any
element that has an itemscope attribute should still have an itemprop attribute, so this
should still work.
So at this point our data object should be an accurate representation of the microdata set
on elements inside our main container. Before we do anything with them however, we need
to check whether the contentMethods object has been populated, and if so add it to our
data object.
We can use the hasProps() function to check whether an object has its own properties.
The function will receive the object to test as an argument. Inside the function we first define
the hasData variable, which we set to false.
We then use a for in loop to iterate over each property of the object. For each property
we check whether the property actually exists on the object and was not inherited using
JavaScript's hasOwnProperty() function. If the property does belong to the object, we set
the hasData to true and then break out of the loop using break.
We then check whether the contactMethods object has any properties by passing it to
our hasProps() function, and if it does we add it to the data object. Finally, once all this
processing has been done, we add the data object to the peopleData array we defined at
the start of our code.
Adding a mechanism for saving the
microdata
At this point, if the page being displayed in Chrome contains any person microdata, we'll
have an array containing one or more objects that contain the microdata and the text it
describes. In this task we'll allow the user to store that data if he/she wishes.
Because our content script runs in the context of a web page and not our extension, we'll need
to use messaging once again to pass any gathered data back to the extension for storage.
165
Extending Chrome with jQuery
Prepare for Lift Off
In order to set up messaging between our content script and the extension, we'll need
to add a background page. A background page runs continuously while the extension is
installed and enabled and will allow us to set up handlers to listen and respond to messages
sent from the content script.
Background pages may be HTML or JavaScript. In this project we'll use the JavaScript version.
Create a new file now and save it in the chrome-extension directory as background.
js. We also need to register this file as a background script by adding a new background
section to the manifest.json file:
"background": {
"scripts": ["jquery-1.9.0.min.js", "background.js"]
}
This code should come directly after the array listing the content_scripts. Again, don't
forget the trailing comma after the array.
Engage Thrusters
First we'll add the required behavior to our background page. In background.js, add the
following code:
chrome.extension.onConnect.addListener(function (port) {
port.onMessage.addListener(function (msg) {
if (msg.command === "getData") {
var contacts = localStorage.getItem("webContacts")
|| '{ "message": "no contacts" }',
jsonContacts = JSON.parse(contacts);
port.postMessage(jsonContacts);
} else if (msg.command === "setData") {
localStorage.setItem("webContacts",
JSON.stringify({
contacts: msg.contacts
}));
port.postMessage({ message: "success" });
}
});
});
166
Project 6
Next, in content.js, directly after where we pushed a data object into the peopleData
array, add the following code:
$("<a/>", {
href: "#",
"class": "app-save",
text: "Save"
}).on("click", function (e) {
e.preventDefault();
var el = $(this),
port = chrome.extension.connect(),
contacts;
if (!el.hasClass("app-saved")) {
port.postMessage({ command: "getData" });
port.onMessage.addListener(function (msg) {
if (msg.message === "no contacts") {
contacts = [peopleData[i]];
port.postMessage({
command:"setData",
contacts:contacts
});
} else if (msg.contacts) {
contacts = msg.contacts;
contacts.push(peopleData[i]);
port.postMessage({
command: "setData",
contacts: contacts
});
} else if (msg.message === "success") {
el.addClass("app-saved")
.text("Contact information saved");
port.disconnect();
}
});
}
}).appendTo(person);
167
Extending Chrome with jQuery
Finally, we can add a little styling for the new save link that we just added. In content.css,
add the following code at the bottom of the file:
.app-save { position:absolute; top:5px; right:5px; }
.app-saved { opacity:.5; cursor:default; }
Objective Complete - Mini Debriefing
We added quite a lot of code in this task as there were a number of different files that we
updated in order to get the different parts of the extension communicating.
Adding the communication module
First of all we updated the behavior page that we added at the start of the task. We'll be
using localStorage to store saved contacts gathered by the extension, but the content script
that runs in the context of web pages viewed by the user of the extension only has access to
the localStorage area for any given page, but we need access to the localStorage area for the
extension itself.
To achieve this, our background.js file will act as a mediator that will access the
extension's localStorage area and pass data back and forth between the content script
and the extension.
First we added a listener to the onConnect event, which we can access via Chrome's
extension utility module. When a content script makes a connection to the extension, a
port will be opened automatically by the browser. An object representing this port will be
passed to our handler function automatically.
We can use the port to add an event handler for message events. As with our simple
<iframe> messaging from earlier in the project, this handler function will automatically be
passed the message that triggers the event.
Inside the message handler, we check whether the message's command property is equal to
getData. If it is, we first create a contacts object, which will consist of either the contacts
obtained from the localStorage getItem()method, or a very simple JSON object that simply
contains the message no contacts, which we can create manually.
Once we have either of these JSON objects, we can then parse it into a proper JavaScript
object using Chrome's native JSON parse() method. We can then pass this object back
to the port using the postMessage() method. A new port will be opened whenever a
new connection is made, so messages will automatically be passed back to the correct port
without additional configuration by us.
168
Project 6
If the command property of the msg object does not equal getData, it may equal setData
instead. If it does, we want to store one or more new contacts to localStorage. In this case
we'll pass the contacts to store as an object contained in the contacts property of the msg
object, so we can simply use the stringify() method on the object in this property as the
second argument to the setItem() method.
We can then pass back a short message confirming that saving the data has been a success
using the port object's postMessage() method once more.
Updating the content script
Secondly we updated the content.js file in order to harvest and store any contact
information found on the web page being viewed by the visitor.
We started out by adding a new <a> element that will be used as a button to save contact
information and which will be added to any element containing microdata. We added a
simple # href attribute to the new element, a class name for styling purposes, and the
text Save.
Most of the new functionality is contained in a click event handler attached directly to each
new <a> element when it is created using jQuery's on() method.
Within this event handler we first stop the default behavior of the browser using
preventDefault(), as we usually do when attaching event handlers to <a> elements.
We then cache a reference to the current <a> element by storing $(this) in a variable
called el. We also open up a new port to handle our communication needs using the
extension module's connect() method. A contacts variable is declared but not
defined straight away.
The rest of the code resides within an if conditional that checks the element does not
already have the class name app-saved, which will help prevent duplicate entries for the
same person on a single page being saved to localStorage.
In the conditional we first need to get any previously stored contacts, so we request the saved
contacts from the behavior page by posting a message to the port we opened a moment ago.
We send an object as the message which has a command property set to getData.
We then add a handler for the response to this message using the addListener() method
on the onMessage event. The rest of our code is within this handler, which consists of
another conditional that reacts differently depending on the response message.
The first branch of the conditional deals with when the message property of the response
msg contains the string no contacts. In this case we create a new array which contains
the contact information harvested from the person for whichever save link was clicked. We
already have this information in the peopleData array and as we're still inside the loop that
updates each person, we can use the i variable to store the correct person.
169
Extending Chrome with jQuery
We can then send this array to the behavior page for permanent storage in the extension's
localStorage area.
If the msg object doesn't have a message property it may have a contacts property
instead. This property will contain the array of previously stored contacts so we can save the
array to a variable, and add the new contact to this array before posting the updated array
back to the behavior page for permanent storage.
The final branch of the conditional deals with a successful save of the contact. In this
case the message property of the msg object will contain the success string. In this
case we add the class name app-saved to the <a> element and change the text to
Contact information saved. As the port is no longer required we can close it using the
disconnect() method of the port object.
Adding the simple styling
Lastly we added some very simple styling for the save links. It's important to show feedback
once the operation that the user initiates has completed.
In this example, we do that by changing the text of the link simply making it more opaque
using CSS to make it look as if it is no longer clickable, which is the case because of the if
statement we used in the script.
At this point we should now be able to browse to a page that contains microdata and save
contact information. When the browser action button is clicked, we'll see the popup, which
should display the saved contact, as shown in the screenshot at the start of the project.
Classified Intel
When testing content scripts, it's important to realize that whenever any of the content files
change, which in this case means either the JavaScript file or the style sheet, the extension
must be reloaded.
To reload the extension, there's a Reload (Ctrl + R) link below the extension listed on
Chrome's Extensions page. We'll need to click on this link to apply changes made to any of
the content files. Other parts of the extension, such as the popup files for example, do not
require that the extension be reloaded.
Another useful tool for extension programmers is the developer tool, which can be
opened specifically to monitor the code in the background page. This can be useful for
troubleshooting errors and script debugging when using a background page.
170
Project 6
Mission Accomplished
In this project we've covered most of the basics of building a Chrome extension.
We covered creating a browser action that triggers a popup when it is clicked in
order to display saved contacts.
We also saw how we can safely sandbox pages that need to run dangerous code such as
eval() or new Function in order to protect our extension from XSS attacks, and how we
can use the simple messaging API to send messages and receive responses to an <iframe>
element containing the sandboxed page.
We saw that as well as defining scripts that run in the context of the extension, we can also
add content scripts that run in the context of the web page being displayed in the browser.
We also learned how to use the manifest.json file to specify these different areas of
our extension.
We also saw that we can use a much more advanced messaging system that allows us to
open ports that allow for more than simple one-way messages. Communicating via ports
allows us to send as many messages back and forth from different areas of our extension
as we need to in order to complete a given task such as saving data to the extension's
localStorage area.
We also learned about the type of data that can be described using Schema.org microdata,
and the HTML attributes that can be added to elements in order to do the describing. As well
as being able to describe people, there are also Schema.org formats for describing places,
companies, movies, and much more.
We've learned a lot about creating extensions in Chrome, but we've also used a ton of
jQuery methods in order to simplify the scripts that we've written to power the extension.
You Ready To Go Gung HO?
A Hotshot Challenge
When our extension saves new contacts, the highlighted elements representing the
elements containing microdata are given a new CSS class name and have some very minimal
additional styling added to them.
This is ok, but a better way of confirming success would be to make use of Chrome's desktop
notification system to generate growl-style pop-up notifications that confirm success instead.
Check out the notification documentation at http://developer.chrome.com/
extensions/notifications.html and see if you can update the extension to
include this feature.
171
Project 7
Build Your Own jQuery
With the 1.8 release of jQuery, something new was introduced that the development
community as a whole had been wanting for a while – the ability to build custom versions
of jQuery that contained only the functionality that was required for a given task.
Mission Briefing
In this project we'll set up the environment we need in order to use jQuery's build tool. We'll
see what other software we need to use, how to run the build tool itself, and what we can
expect as output of the build tool.
Why Is It Awesome?
While it's fairly common for someone to say that they use jQuery in every site they build
(this is usually the case for me), I would expect it much rarer for someone to say that they
use the exact same jQuery methods in every project, or that they use a very large selection
of the available methods and functionality that it offers.
The need to reduce file size as aggressively as possible to cater for the mobile space, and the
rise of micro-frameworks such as Zepto for example, which delivers a lot of jQuery functionality
at a much-reduced size, have pushed jQuery to provide a way of slimming down.
As of jQuery 1.8, we can now use the official jQuery build tool to build our own custom
version of the library, allowing us to minimize the size of the library by choosing only the
functionality we require.
For more information on Zepto, see http://zeptojs.com/.
Build Your Own jQuery
Your Hotshot Objectives
To successfully conclude this project we'll need to complete the following tasks:
ff
Installing Git and Make
ff
Installing Node.js
ff
Installing Grunt.js
ff
Configuring the environment
ff
Building a custom jQuery
ff
Running unit tests with QUnit
Mission Checklist
We'll be using Node.js to run the build tool, so you should download a copy of this now. The
Node website (http://nodejs.org/download/) has an installer for both 64 and 32-bit
versions of Windows, as well as a Mac OS X installer. It also features binaries for Mac OS X,
Linux, and SunOS. Download and install the appropriate version for your operating system.
The official build tool for jQuery (although it can do much more besides build jQuery) is
Grunt.js, written by Ben Alman. We don't need to download this as it's installed via the
Node Package Manager (NPM). We'll look at this process in detail later in the project.
For more information on Grunt.js, visit the official site at
http://gruntjs.com.
First of all we need to set up a local working area. We can create a folder in our root project
folder called jquery-source. This is where we'll store the jQuery source when we clone
the jQuery Github repository, and also where Grunt will build the final version of jQuery.
Installing Git and Make
The first thing we need to install is Git, which we'll need in order to clone the jQuery source
from the Github repository to our own computer so that we can work with the source
files. We also need something called Make, but we only need to actually install this on Mac
platforms because it gets installed automatically on Windows when Git is installed.
As the file we'll create will be for our own use only and we don't want
to contribute to jQuery by pushing code back to the repository, we
don't need to worry about having an account set up on Github.
174
Project 7
Prepare for Lift Off
First we'll need to download the relevant installers for both Git and Make. Different
applications are required depending on whether you are developing on Mac or
Windows platforms.
Mac developers
Mac users can visit http://git-scm.com/download/mac for Git.
Next we can install Make. Mac developers can get this by installing XCode. This can be
downloaded from https://developer.apple.com/xcode/.
Windows developers
Windows users can install msysgit, which can be obtained by visiting https://code.
google.com/p/msysgit/downloads/detail?name=msysGit-fullinstall-1.8.0preview20121022.exe.
Engage Thrusters
Once the installers have downloaded, run them to install the applications. The defaults
selected by the installers should be fine for the purposes of this mission. First we should
install Git (or msysgit on Windows).
Mac developers
Mac developers simply need to run the installer for Git to install it to the system. Once this
is complete we can then install XCode. All we need to do is run the installer and Make, along
with some other tools, will be installed and ready.
Windows developers
Once the full installer for msysgit has finished, you should be left with a Command Line
Interface (CLI) window (entitled MINGW32) indicating that everything is ready for you to
hack. However, before we can hack, we need to compile Git.
To do this we need to run a file called initialize.sh. In the MINGW32 window, cd into
the msysgit directory. If you allowed this to install to the default location, you can use the
following command:
cd C:\\msysgit\\msysgit\\share\\msysGit
Once we are in the correct directory, we can then run initialize.sh in the CLI. Like the
installation, this process can take some time, so be patient and wait for the CLI to return a
flashing cursor at the $ character.
175
Build Your Own jQuery
An Internet connection is required to compile Git in this way.
Windows developers will need to ensure that the Git.exe and MINGW resources can be
reached via the system's PATH variable. This can be updated by going to Control Panel |
System | Advanced system settings | Environment variables.
In the bottom section of the dialog box, double-click on Path and add the following two
paths to the git.exe file in the bin folder, which is itself in a directory inside the msysgit
folder wherever you chose to install it:
ff
;C:\msysgit\msysgit\bin;
ff
C:\msysgit\msysgit\mingw\bin;
Update the path with caution!
You must ensure that the path to Git.exe is separated from the rest
of the Path variables with a semicolon. If the path does not end with a
semicolon before adding the path to Git.exe, make sure you add one.
Incorrectly updating your path variables can result in system instability
and/or loss of data. I have shown a semicolon at the start of the previous
code sample to illustrate this.
Once the path has been updated, we should then be able to use a regular command prompt
to run Git commands.
Post-installation tasks
In a terminal or Windows Command Prompt (I'll refer to both simply as the CLI from this
point on for conciseness) window, we should first cd into the jquery-source folder we
created at the start of the project. Depending on where your local development folder is, this
command will look something like the following:
cd c:\jquery-hotshots\jquery-source
To clone the jQuery repository, enter the following command in the CLI:
git clone git://github.com/jquery/jquery.git
Again, we should see some activity on the CLI before it returns to a flashing cursor to indicate
that the process is complete.
176
Project 7
Depending on the platform you are developing on, you should see something like the
following screenshot:
Objective Complete - Mini Debriefing
We installed Git and then used it to clone the jQuery Github repository in to this directory in
order to get a fresh version of the jQuery source. If you're used to SVN, cloning a repository
is conceptually the same as checking out a repository.
Again, the syntax of these commands is very similar on Mac and Windows systems, but
notice how we need to escape the backslashes in the path when using Windows. Once this
is complete, we should end up with a new directory inside our jquery-source directory
called jquery.
If we go into this directory, there are some more directories including:
ff
build: This directory is used by the build tool to build jQuery
ff
speed: This directory contains benchmarking tests
ff
src: This directory contains all of the individual source files that are compiled
together to make jQuery
ff
Test: This directory contains all of the unit tests for jQuery
It also has a range of various files, including:
ff
Licensing and documentation, including jQuery's authors and a guide to contributing
to the project
ff
Git-specific files such as .gitignore and .gitmodules
ff
Grunt-specific files such as Gruntfile.js
ff
JSHint for testing and code-quality purposes
177
Build Your Own jQuery
Make is not something we need to use directly, but Grunt will use it when we build the
jQuery source, so it needs to be present on our system.
Installing Node.js
Node.js is a platform for running server-side applications built with JavaScript. It is trivial
to create a web-server instance, for example, that receives and responds to HTTP requests
using callback functions.
Server-side JS isn't exactly the same as its more familiar client-side counterpart, but you'll
find a lot of similarities in the same comfortable syntax that you know and love. We won't
actually be writing any server-side JavaScript in this project – all we need Node for is to run
the Grunt.js build tool.
Prepare for Lift Off
To get the appropriate installer for your platform, visit the Node.js website at
http://nodejs.org and hit the download button. The correct installer for
your platform, if supported, should be auto-detected.
Engage Thrusters
Installing Node is a straightforward procedure on either the Windows or Mac platforms as
there are installers for both. This task will include running the installer, which is obviously
simple, and testing the installation using a CLI.
On Windows or Mac platforms, run the installer and it will guide you through the installation
process. I have found that the default options are fine in most cases. As before, we also need
to update the Path variable to include Node and Node's package manager NPM. The paths
to these directories will differ between platforms.
Mac
Mac developers should check that the $PATH variable contains a reference to usr/local/
bin. I found that this was already in my $PATH, but if you do find that it's not present, you
should add it.
For more information on updating your $PATH variable, see
http://www.tech-recipes.com/rx/2621/os_x_
change_path_environment_variable/.
178
Project 7
Windows
Windows developers will need to update the Path variable, in the same way as before,
with the following paths:
ff
C:\Program Files\nodejs\;
ff
C:\Users\Desktop\AppData\Roaming\npm;
Windows developers may find that the Path variable already contains
an entry for Node so may just need to add the path to NPM.
Objective Complete - Mini Debriefing
Once Node is installed, we will need to use a CLI to interact with it. To verify Node has
installed correctly, type the following command into the CLI:
node -v
The CLI should report the version in use, as follows:
We can test NPM in the same way by running the following command:
npm -v
Installing Grunt.js
In this task we need to install Grunt.js, which is extremely quick and easy to do, just like
the Node installation. We don't even need to download anything manually and as before
the same commands should work on either Mac or Windows systems, with only very
minor adjustments.
179
Build Your Own jQuery
Engage Thrusters
We need to use Node Package Manager NPM to install it, and can do this by running the
following command (note that Node itself must not be running):
npm install -g grunt-cli
Mac users may need to use superuser do at the start of the command:
sudo –s npm install –g grunt
Be prepared to wait a few minutes. Again, we should see a lot of activity as the resources
Grunt needs are downloaded and installed. The prompt will return to a flashing cursor once
the installation is complete. The CLI should appear like the following screenshot, depending
on which platform you are developing on:
Objective Complete - Mini Debriefing
If all goes according to plan (which, unless you have problems with your system, it generally
should), you should see a lot of activity in the CLI while Grunt and its dependencies are
downloaded and installed globally via NPM, and once that finishes Grunt will be installed
and ready to use.
An Internet connection is required for packages to be downloaded
and installed automatically using NPM.
To verify that Grunt has installed correctly, we can type the following command into the CLI:
grunt -version
180
Project 7
This will output the current version of Grunt and should work from any directory as Grunt
was installed globally.
Classified Intel
As well as building custom versions of jQuery, Grunt can also be used to create a few
different common projects. We start by choosing one of the following project types:
ff
gruntfile
ff
commonjs
ff
jquery
ff
node
We can run the built-in init task and specify one of these projects, and Grunt will go ahead
and set up a skeleton project containing resources commonly used with that project.
For example, running the jquery init task will set up a working directory for creating a
jQuery plugin. Inside the directory, Grunt will create folders for the source script files and the
unit tests, as well as create a range of files including a package.json file.
It is likely that at some point all new jQuery plugins will need to be structured in the way that
Grunt structures the working directory when creating this project type, so Grunt will become
an invaluable, time-saving tool for any jQuery plugin developer.
Configuring the environment
There're just a couple of things left we need to do before we're ready to build our own
version of jQuery. We can also test our installation and configuration by building the full
version of jQuery in order to test that everything is working as it should.
Prepare for Lift Off
We'll need to install some additional Grunt dependencies so that we can create the jQuery
Script files using the source files that we cloned from Github. There are also a range of NPM
modules that the project uses, which also need to be installed. Fortunately the NPM can
install everything for us automatically.
181
Build Your Own jQuery
Engage Thrusters
Before we can build the jQuery source, we need to install some additional Grunt
dependencies in the jquery source folder. We can do this using NPM and so can
enter the following command into the CLI:
npm install
Make sure you have already used the cd command to navigate to
the jquery directory before running the install command.
There should be a lot of activity in the CLI after running the install command, and by the
end of the process, the CLI should appear something like the following screenshot:
182
Project 7
To test that everything is working together as it should, we can build the full version of
jQuery. Simply run the grunt command in the CLI:
grunt
If you get any errors or warnings at this point, something has not
been installed or configured correctly. There could be any number of
reasons why the process has failed, so the best thing to do is to uninstall
everything we have installed, and start the process again, ensuring that
all steps are followed to the letter.
Again, we should see a lot of activity on the CLI to indicate that things are happening:
183
Build Your Own jQuery
Objective Complete - Mini Debriefing
Once the install process has completed, we should find that the Node dependencies have
been installed into a directory within the jquery directory called node_modules. Within
this folder are any additional files that Grunt needs for this specific project.
To test everything, we then ran the default build task for jQuery using the grunt command.
This task will do the following things:
ff
Read all of the jQuery source files
ff
Create a /dist directory for the output of the task
ff
Build the jquery.js distribution file
ff
Lint the distribution file with jshint
ff
Run the units tests
ff
Build a source map of the distribution file
ff
Build the jquery.min.js distribution file
The script files should be 230 KB for the full file and 81 KB for the .min file, although these
figures may differ as the jQuery version number increases.
Building a custom jQuery
In this task we'll build a custom version of jQuery, which will not consist of all of the different
modules that make up the "full" version of jQuery combined into the single file which we
would normally download from the jQuery site, just like the files we built at the end of the
last task, instead it will contain only the core modules.
Engage Thrusters
Now we can build a custom version of jQuery. To build a barebones version of jQuery,
omitting all of the non-core components, we can enter the following command into the CLI:
grunt custom:-ajax,-css,-deprecated,-dimensions,-effects,-offset
Objective Complete - Mini Debriefing
Once we have the source and have configured our local environment, we are able to build
a custom version of jQuery containing only the core components and omitting all of the
optional components.
184
Project 7
In this case we've excluded all of the optional components, but we could exclude any one
of them, or any combination of them to produce a script file that is only as large as it needs
to be.
If we check the /dist directory at this point, we should find now that the full-fat script file is
now 159 KB, while the .min version is just 57 KB, saving us approximately 30 percent of the
file size; not bad for a few minutes work!
Changes in the functionality or scope of your project may necessitate
that the source file be rebuilt and previously excluded modules
included. Once excluded, optional modules cannot be added to the
built files without rebuilding.
Classified Intel
As jQuery evolves, especially once the 2.0 milestone is passed, more and more of jQuery's
components will be exposed to the build tool as optional components, so that it will become
possible to exclude a wider portion of the library.
While at the time of writing the savings we make in file size may well be offset by the fact
that the majority of our visitors won't have our custom version of jQuery in their cache and
will therefore have to download it, there may come a time when we are able to shrink the
size of the file so much that it will still be more efficient to download our super-light script
file than it is to load the full source from the cache.
Running unit tests with QUnit
QUnit is the official test suite for jQuery and is included in the source that we cloned from
Git earlier in the project. If we take a look in the test folder inside the jquery folder, we
should find that there are a lot of unit tests written to test the different components that
make jQuery.
We can run these tests against the individual components of jQuery in order to look at the
environment that QUnit needs, and to see how easy testing JavaScript files using it can be.
For this task we'll need to install a web server and PHP.
For more information on QUnit, see the documentation at
http://qunitjs.com.
185
Build Your Own jQuery
Prepare for Lift Off
Mac developers should already have everything required in order to run QUnit through a
web server, because Mac computers come with Apache and PHP already installed. Windows
developers will probably have some setup to do however.
There are two options for the web server in this case, Apache or IIS. Both support PHP. Those
developers wishing to use Apache can install something like WAMP (Windows Apache Mysql
PHP) to have Apache installed and configured, and MySQL and PHP installed as modules.
To download and install WAMP, visit the Download section of the Wamp Server website
(http://www.wampserver.com/en/).
Pick the appropriate installer for your platform and run it. This should install and configure
everything necessary.
Those wishing to use IIS can install it via the Add/Remove Windows Components area of
the Programs and Features page in the control panel (the Windows installation CD will be
required in this case), or using the Web Platform Installer (WPI), which can be downloaded
from http://www.microsoft.com/web/downloads/platform.aspx.
Download and run the installer. Once it has launched, search for IIS and let the application
install it. Once this has installed, search for PHP and install that through the WPI also.
To run QUnit using a web server and PHP, you'll need to either copy the source files from the
jquery directory inside our project folder into the folder used by the web server to serve
files from, or configure the web server to serve files from the jquery directory.
On Apache, we can configure the default directory (the directory from which pages are
served when requested by the browser) by editing the httpd.conf file (there should be an
entry for this in the Start menu). Read down through the configuration file until you find the
line for the default directory and change it so that it points to the jquery directory in your
project folder.
On IIS, we can add a new website using the IIS Manager. Right-click on Sites in the
Connections pane at the left and choose Add Web Site…. Fill out the details in the dialog
that opens and we should be good to go.
Engage Thrusters
To run the tests, all we need to do is visit the /test directory using localhost:8080 (or
whichever hostname/port is configured) in a browser:
localhost:8080/test
186
Project 7
The tests should appear as shown in the following screenshot:
Objective Complete - Mini Debriefing
When the URL for the test suite is visited in a browser, QUnit will run all of the unit tests that
have been written for jQuery. There are currently over 6000 tests for the full build of jQuery,
and around 4000 tests for a custom build with all optional modules excluded.
You may find that some of the tests fail. Don't worry, this is normal and the reason for it is
because the default version of jQuery we get from Git will be the latest development version.
As I write this, the current version of jQuery is 1.8.3, but the version cloned from Git is
2.0.0pre.
To fix this, we could switch to the current stable branch and do a build from there instead. So
if I wanted to get version 1.8.3, I could use the following command in the CLI:
git checkout 1.8.3
187
Build Your Own jQuery
We could then build the source again, run QUnit, and all the tests should pass.
After checking out another version of the jQuery source, we'll need to run
npm install in the jquery directory to reinstall the node dependencies.
Classified Intel
Unit testing is a process not always followed religiously by front-end developers, but once
your applications cross a rough threshold in size and complexity, or when working in a team
environment, unit testing becomes essential for maintenance, so it is best to at least learn
the basics.
QUnit makes writing unit test for JavaScript easy. It features a simple API based around the
concept of making assertions that we can prove with simple functions. QUnit's API includes
methods we can use to make these assertions including:
ff
equal()
ff
notEqual()
ff
ok()
This makes it easy to check that a variable is equal to a particular value, or the return value
of a function does not equal a particular value, and so on.
A test in QUnit is built using the global test() method, which accepts two arguments; a
string describing the test and a function which is invoked to perform the test:
test("Test the return value of myCustomMethod()", function() {
//test code here
});
Within the function we can then use one or more of the assertions to check the result of
some operation performed by the method or function we're testing:
var value = myCustomMethod();
equal(value, true, "This method should return true");
The equal() method checks that the first and second arguments are equal and the final
argument is a string which describes what we expect to happen.
If you open up some of the script files in the jquery/test/
unit directory, it is easy to see how the tests can be structured.
188
Project 7
The documentation on the QUnit site is excellent. It not only describes the API clearly and
concisely, but it also has a lot of information about the concepts of unit testing, and so is a
great place to start for those new to unit testing.
On the site you can also find the source files required to run QUnit outside of Grunt and a
HTML template page you can use to run the test suite in a browser.
Mission Accomplished
In this mission we've learned not just how to build a custom version of jQuery by excluding
components that we don't require, and how to run jQuery's unit test suite, but also, perhaps
more importantly, we've learned how to set up a decent build environment for writing clean,
lint and error free, application-grade JavaScript.
You Ready To Go Gung HO?
A Hotshot Challenge
We've learned how to build our own jQuery, and excluded the maximum number of
components, so at the time of writing, there isn't much more we can do.
If you're reading this after version 1.9 of jQuery has been released, there may be more
components that you can exclude, or other techniques for building jQuery, so to really
cement your understanding of the build process, build a new custom build that also
excludes any new optional components.
If there aren't any new optional components, I'd recommend you to spend some time
writing QUnit tests for any custom scripts that you've written. The idea is to write a test
that replicates a bug. Then you can fix the bug and watch the test pass.
189
Project 8
Infinite Scrolling
with jQuery
Infinite scrolling is a technique employed by a lot of popular sites that minimizes the amount
of data loaded on a page initially, and then incrementally loads more data when the user
has scrolled to the bottom of the page. You can see this effect on the Facebook or Twitter
timelines among others.
Mission Briefing
In this project we'll build an infinite scrolling system using jQuery that mimics the effect seen
on sites such as those mentioned earlier. We'll make a request for some data and display
it on the page. Once the user has scrolled to the bottom of the page, we'll make another
request for the next page of data, and so on and so forth as the user continues scrolling.
Infinite Scrolling with jQuery
Once we've built our infinite scrolling system, we should end up with something similar to
the following screenshot:
Why Is It Awesome?
If you have a lot of data to display and it can easily be arranged in reverse-chronological
order, using the infinite scrolling technique is an easy way to maximize the user experience of
the page with progressive disclosure – showing incrementally more to the user.
A small subset of the data, which will make the page load quicker, as well as prevent your
visitor being overwhelmed with data, can be displayed at first, which gradually increases
following user interaction.
The data that we'll consume in this project will be a list of videos uploaded to the TEDTalks
channel on YouTube, in JSON format.
192
Project 8
Remember, JSON is a lightweight, text-based data format perfect for
transportation across networks. For more information on JSON, see
http://www.json.org/.
There are thousands of videos that can be found on this channel, so it makes a good test bed
for our project. Data that can be sorted into a time-based order makes a great basis for an
infinite scroller.
The TEDTalks channel can be viewed directly on the YouTube website at
http://www.youtube.com/user/tedtalksdirector.
Your Hotshot Objectives
This project will be broken down into the following tasks:
ff
Preparing the underlying page
ff
Getting the initial feeds
ff
Displaying the initial set of results
ff
Handling scrolling to the bottom of the page
Mission Checklist
We can link to the hosted version of JsRender like we did in some of our previous examples,
but in this project we're going to make use of a handy little jQuery plugin called imagesLoaded,
which allows us to fire a callback function when all the images with a selected container have
finished loading.
The imagesLoaded plugin can be downloaded from https://github.com/desandro/
imagesloaded and should be saved in our project's js directory.
Preparing the underlying page
In this task we'll set up the files that we're going to use throughout this project and prepare
the underlying page that our infinite scroller will reside on.
193
Infinite Scrolling with jQuery
Prepare for Lift Off
As usual we'll be using a custom style sheet for this project and a custom script file, so let's
add them first. Create a new JavaScript file called infinite-scroller.js and save it in
the js directory. Then create a new style sheet called infinite-scoller.css and save it
in the css directory. Lastly, save a copy of the template.html file in the root project folder
and call it infinite-scroller.html.
Engage Thrusters
The underlying markup used by the example page will be minimal – a lot of the elements
that we'll be working with will be generated dynamically by our template, which we can also
add in this task.
First of all we should add references to our new files to the HTML page. Start by adding a
<link> element for the style sheet to the <head> of infinite-scroller.html, directly
after the link to common.css:
<link rel="stylesheet" href="css/infinite-scroller.css" />
Next we can link to the two new JavaScript files. Add the following <script> elements
directly after jQuery:
<script src="http://borismoore.github.com/jsrender/jsrender.js">
</script>
<scriptsrc="js/jquery.imagesloaded.min.js"></script>
<scriptsrc="js/infinite-scroller.js"></script>
We also need to add a simple container to render our data into. Add the following code to
the <body> of the page:
<div id="videoList"></div>
Now we can add the templates we'll use. In this project we'll use two templates – one to
render the outer container and user data, which will be rendered once, and one to render
the list of videos, which we can use repeatedly as required.
As before, they will sit inside the <script> elements in the <body> of the page. Before the
existing <script> elements, add following new templates:
<script id="containerTemplate" type="text/x-jsrender">
<section>
<header class="clearfix">
<imgsrc="{{>avatar}}" alt="{{>name}}" />
<hgroup>
<h1>{{>name}}</h1>
194
Project 8
<h2>{{>summary.substring(19, 220)}}</h2>
</hgroup>
</header>
<ul id="videos"></ul>
</section>
</script>
Now for the video template:
<script id="videoTemplate" type="text/x-jsrender">
<li>
<article class="clearfix">
<header>
<a href="{{>content[5]}}" title="Watch video">
<imgsrc="{{>thumbnail.hqDefault}}"
alt="{{>title}}" />
</a>
<cite>
<a href="{{>content[5]}}"
title="Watch video">{{>title}}</a>
</cite>
</header>
<p>
{{>~Truncate(12, description)}}
<a class="button" href="{{>content[5]}}"
title="Watch video">Watch video</a>
</p>
<div class="meta">
<dl>
<dt>Duration:</dt>
<dd>{{>~FormatTime(duration)}}</dd>
<dt>Category:</dt>
<dd>{{>category}}</dd>
<dt>Comments:</dt>
<dd>{{>commentCount}}</dd>
<dt>Views:</dt>
<dd>{{>viewCount}}</dd>
<dt>Likes:</dt>
<dd>{{>likeCount}}</dd>
</dl>
</div>
</article>
</li>
</script>
195
Infinite Scrolling with jQuery
We can also add the styling for these elements now as well. In infinite-scroller.css,
add the following selectors and rules:
section { width:960px; padding-top:20px; margin:auto; }
section {
width:960px; padding:2em 2.5em 0;
border-left:1px solid #ccc; border-right:1px solid #ccc;
margin:auto; background-color:#eee;
}
section> header {
padding-bottom:2em; border-bottom:1px solid #ccc;
}
img, hgroup, hgroup h1, hgroup h2 { float:left; }
hgroup { width:80%; }
headerimg { margin-right:2em; }
hgroup h1 { font-size:1.5em; }
hgroup h1, hgroup h2 { width:80%; }
hgroup h2 {
font-weight:normal; margin-bottom:0; font-size:1.25em;
line-height:1.5em;
}
ul { padding:0; }
li {
padding:2em 0; border-top:1px solid #fff;
border-bottom:1px solid #ccc; margin-bottom:0;
list-style-type:none;
}
article header a {
display:block; width:27.5%; margin-right:2.5%; float:left; }
aimg { max-width:100%; }
article cite {
width:70%; margin-bottom:10px; float:left;
font-size:1.75em;
}
article cite a { width:auto; margin-bottom:.5em; }
article p {
width:45%; padding-right:2.5%;
border-right:1px solid #ccc; margin:0 2.5% 2em 0;
float:left; line-height:1.75em;
}
article .button { display:block; width:90px; margin-top:1em; }
article dl { width:19%; float:left; }
article dt, article dd {
width:50%; float:left; font-size:1.15em; text-align:right;
196
Project 8
}
article dt { margin:0 0 .5em; clear:both; font-weight:bold; }
li.loading{ height:100px; position:relative; }
li.loading span {
display:block; padding-top:3em; margin:-3em 0 0 -1em;
position:absolute; top:50%; left:50%; text-align:center;
background:url(../img/ajax-loader.gif) no-repeat 50% 0;
}
The ajax-loader.gif image used in this project can be found
in the accompanying code download for this book.
Objective Complete - Mini Debriefing
So practically the entire page is built from the templates we added to the <body> of the
page, except for an empty <div> that will give us a container to render the data into. The
template contains the markup used not just for the list of videos, but also information about
the user whose videos we're displaying.
In the first template, the outer container for the data is a <section> element. Within this
is a <header> which displays information about the user, including his/her profile picture,
name, and bio.
The actual bio returned by YouTube for a specified user can be quite long, so we'll use the
JavaScript substring() function to return a shortened version of this summary. This
function is passed two arguments; the first is the character to start copying from and the
second is the character to end on.
In the second template the actual list of videos will be displayed in the<ul> element added
in the first template, with each video occupying an <li>. Within each <li>, we have an
<article> element, which is an appropriate container for an independent unit of content.
Within the <article>, we have a <header> containing some of the key information
about the video such as its title and a thumbnail. Following the <header>we display a
short summary of the video in a <p>, element. We also use our shortening helper function
Truncate(), starting at character 12.
Lastly we display some meta-information about the video such as the number of plays, the
number of likes, and the video's duration, using a <dl>.
We use another helper function to display the duration in the video, FormatTime().
YouTube returns the length of the video in seconds, so we can convert this into a nicely
formatted time string instead.
197
Infinite Scrolling with jQuery
We used the > character to HTML-encode any data we insert into the page. It's always best
to do this as a security measure.
The CSS that was added was purely presentational; merely used to lay out the page in a list
format, and make it look slightly interesting and presentable. Feel free to change any aspect
of how the layout is styled, or the elements are themed.
Classified Intel
The SEO-conscious among you will realize that a page that is built almost entirely from AJAX
delivered content is unlikely to be well placed in search results. Traditionally, this would almost
certainly have been true, but now we can use the awesome pushState() method of the
HTML History API to deliver a dynamic website that is completely indexable by search engines.
A complete description of pushState()is beyond the scope of this book, but there are
plenty of great examples and tutorials out there. Considered by many to be the definitive
guide to the History API is the documentation on Mozilla's Developer Network, which
includes a section on pushState(). You can see the documentation at https://
developer.mozilla.org/en-US/docs/DOM/Manipulating_the_browser_history.
Getting the initial feeds
In this task we'll focus on getting the initial set of data in order to create the page when it
first loads. We need to write our code so that the function for getting the first page of data is
reusable for any page of data so that we can make use of it later on in the project.
Prepare for Lift Off
We can use the standard document ready shortcut provided by jQuery, just like we have
in a number of previous projects. We can get ready by adding the following code to the
infinite-scroller.js file that we create earlier:
$(function () {
//rest of our code will go here...
});
198
Project 8
Engage Thrusters
First of all we can add the code that retrieves the data from YouTube. Replace the comment
in the preceding snippet of code with the following:
var data = {},
startIndex = 1;
var getUser = function () {
return $.getJSON("http://gdata.youtube.com/
feeds/api/users/tedtalksdirector?callback=?", {
v: 2,
alt: "json"
}, function (user) {
data.userdata = user.entry;
});
};
var getData = function () {
return $.getJSON("https://gdata.youtube.com/
feeds/api/videos?callback=?", {
author: "tedtalksdirector",
v: 2,
alt: "jsonc",
"start-index": startIndex
}, function (videos) {
data.videodata = videos.data.items;
});
};
Next we need to process the response a little. We can use the following code to execute a
callback function once both of the AJAX requests have completed, which should be added
directly after the code we added a moment ago:
$.when(getUser(), getData()).done(function () {
startIndex+=25;
var ud = data.userdata,
clean = {};
clean.name = ud.yt$username.display;
clean.avatar = ud.media$thumbnail.url;
clean.summary = ud.summary.$t;
data.userdata = clean;
});
199
Infinite Scrolling with jQuery
Objective Complete - Mini Debriefing
We started out by defining a couple of variables. The first is an empty object that we'll
populate with the results of our AJAX requests. The second is an integer that represents
the index number of the first video we wish to get. YouTube videos are not zero-based like
regular JavaScript arrays, so we initially define the variable with a value of 1.
Next we added our two functions that we'll use to get our data. The first is the request to get
the profile data of the user whose feed we're going to be displaying. We're only going to be
using this function once, when the page initially loads, but you'll see why it's important that
we define the function as a variable in this way in just a moment.
The second function will be reused, so storing it in a variable is a great way to store it ready
to be invoked whenever we wish to get a new page of video data. It's important that these
functions both return the jqXHR objects returned by the getJSON() method.
Both of the requests use jQuery's getJSON() method to make the requests. In the user
request we only need to set the v and alt query parameters, which are set in the object
passed to getJSON() as the second argument. The user whose profile data we want
actually forms part of the URL we are making the request to.
The callback function for this request simply adds the contents of the user.entry object
received from the request to the userdata property of our data object.
The second request requires slightly more configuration. We still set the API version we want
to use with the v parameter, but this time we set the format of the response to be jsonc
instead of json. In the callback function for this request we store the array of videos in the
videodata property of our data object.
JSON-C stands for json-in-script and is a format that Google can respond with for some
requests. Data returned in JSON-C format is generally more lightweight and more efficient to
process than the same response in JSON due to how Google's API has been engineered.
The properties that we need to use are only returned when this format is used. The only
reason we don't use it when requesting the user data is because there is not a JSON-C
response for that particular query.
For more information on JSON-C responses from Google's APIs, see
the documentation at https://developers.google.com/
youtube/2.0/developers_guide_jsonc.
Next we make use of jQuery's when()method to initiate both of our requests, and then
use the done() method to execute a callback function once both of the jqXHR objects
have been resolved. This is why it was important that the single-use getUser() function is
structured in the same way as the reusable getData() function.
200
Project 8
Inside the callback function for done(), we first increment the startIndex variable by 25
so that when we make another request we get the next "page" of 25 videos. As we have the
first page of data now, when we use the getData() function later on, we'll automatically
get the "next" page of results.
The when() and done() methods are the preferred way of
handling asynchronous operations since jQuery 1.5.
At this point we just need to do a little processing of our userdata object. There's a whole
bunch of data we don't need to use, and some of the data we do need to use is buried within
nested objects, so we simply create a new object called clean and set just the data we need
on this object directly.
Once this has been done we can save our clean object back to our data object, overwriting
the original userdata object. This should make the object easier to process in our template.
Displaying the initial set of results
Now that we have data being returned by YouTube's API, we can render our template.
However, in order to render our template, we need to add helper functions used to format
some of the data. In this task we can add those helper functions and then render the template.
Engage Thrusters
The template helpers don't need to reside within the $.done()callback function. We can
add them directly before this code in infinite-scroller.js:
var truncate = function (start, summary) {
return summary.substring(start,200) + "...";
},
formatTime = function (time) {
var timeArr = [],
hours = Math.floor(time / 3600),
mins = Math.floor((time % 3600) / 60),
secs= Math.floor(time % 60);
if (hours> 0) {
timeArr.push(hours);
}
if (mins< 10) {
timeArr.push("0" + mins);
201
Infinite Scrolling with jQuery
} else {
timeArr.push(mins);
}
if (secs< 10) {
timeArr.push("0" + secs);
} else {
timeArr.push(secs);
}
return timeArr.join(":");
};
Next we just need to register the helper functions. Directly after the previous code add
the following:
$.views.helpers({
Truncate: truncate,
FormatTime: formatTime
});
Lastly we can render our template. We want a function that we can call from anywhere in
our code, ready for when we make further requests later on. Add the following code after
registering the helper functions:
var renderer = function (renderOuter) {
var vidList = $("#videoList");
if (renderOuter) {
vidList.append(
$("#containerTemplate").render(data.userdata));
}
vidList.find("#videos")
.append($("#videoTemplate").render(data.videodata));
}
Now we just need to call this function at the end of our $.done() callback function:
renderer(true);
202
Project 8
Objective Complete - Mini Debriefing
Our first helper function, truncate() is ultra-simple. We simply return a shortened version
of the string that the function receives as an argument. The substring() function takes
two arguments; the first is the position in the string to begin copying at, and the second
argument is the number of characters to copy, which we fix at 200.
To show that the string has been shortened, we also append an ellipsis to the end of the
returned string, which is why we use a helper function here, instead of using a substring in
the template directly as we did earlier.
The formatTime() helper function is a little more complex, but still relatively
straightforward. This function will receive a time in seconds, which we want to format into a
slightly nicer string that shows the hours, if there are any, the minutes, and seconds.
We first create an empty array to store the different components of the string. We then
create a number of variables to hold the hours, minutes, and seconds portions of the time
string we're going to create.
The hours are calculated by dividing the total number of seconds by 3600 (the number of
seconds in an hour). We use Math.floor()on it so that we only get a whole number result.
We need to calculate the minutes slightly differently because we need to take into account
the hours if there are any.
We use the modulus operator (%) here to remove any hours first and then divide the
remainder by 60, which will tell us either the total number of minutes or the remaining
minutes after the hours have been accounted for. To work out the number of seconds, we
just need to use the modulus operator again and the value 60.
We then use a series of conditionals to determine which of the variables to add to the array.
If there are any hours (which is unlikely given the nature of the videos) we push them into
the array.
If there are less than 10 minutes, we add 0 to the minutes figure and then push it into the
array. If there are more than 10 minutes we just push the mins variable into the array. The
same logic is applied to the secs variable before it is pushed into the array.
This function returns a nicely formatted time by joining the items in our array and using a
colon as the separator. The string will be in the format H:MM:SS or MM:SS depending on
the length of the video. Then we registered the helper functions with the template using
JsRender's helpers object, which is itself nested within the views object, which is added to
jQuery by the templating library. The helper functions we wish to add are set as the values in
an object literal, where the keys match the function calls in our template.
203
Infinite Scrolling with jQuery
Next we added a function that we can call in order to render our template. The renderer()
function takes a single argument, which is a Boolean specifying whether to render both the
container template and video template, or just the video template. Inside the function we
first cache a reference to the outer container of the video list.
If the renderOuter argument has a truthy value (that is if it specifically holds the value
true), we render the containerTemplate and append it to the empty <div> we added
to the <body> of the page. We then render the videoTemplate, appending the rendered
HTML to the <ul> that is added by the containerTemplate.
Lastly we called our renderer() function for the first time, passing true as the argument
to render both the containers and the initial list of videos.
Handling scrolling to the bottom of
the page
Now that we've got the first page of videos, we want to add a handler that monitors the
window for scroll events and detects when the page has been scrolled right to the bottom.
Engage Thrusters
First of all we need to add a couple of new variables. Change the very first set of variables
near the top of the file so that they appear as follows:
var data = {},
startIndex = 1,
listHeight = 0,
win = $(window),
winHeight = win.height();
Now we need to update our renderer() function so that when the templates have been
rendered we update the new listHeight variable. Add the following code after where we
render the videoTemplate:
vidList.imagesLoaded(function () {
listHeight = $("#videoList").height();
});
204
Project 8
Next we can add a handler for the scroll event. Directly after the when() method in
infinite-scroller.js, add the following code:
win.on("scroll", function () {
if (win.scrollTop() + winHeight >= listHeight) {
$("<li/>", {
"class": "loading",
html: "<span>Loading older videos...</span>"
}).appendTo("#videos");
$.when(getData()).done(function () {
startIndex += 25;
renderer();
$("li.loading").remove();
});
}
}).on("resize", function() {
winHeight = win.height();
});
We're using a spinner to show the user that more data is being retrieved. We need a few
extra styles to handle the position of the spinner, so we can also add the following code to
the bottom of our infinite-scroller.css style sheet:
li.loading{ height:100px; position:relative; }
li.loading span {
display:block; padding-top:38px; margin:-25px 0 0 -16px;
position:absolute; top:50%; left:50%; text-align:center;
background:url(../img/ajax-loader.gif) no-repeat 50% 0;
}
205
Infinite Scrolling with jQuery
Objective Complete - Mini Debriefing
We attach a handler to the window using our cached win object and the on() method. The
event type is specified as scroll. Inside the callback function we first check whether the
current scrollTop property of the window, plus the height of the viewport is greater
than or equal to the height of our videolist container. We need to do this to know when
the page has been scrolled to the bottom.
If the heights are equal, we create a temporary loader to provide visual feedback to the user
that something is happening. We append a new <li> element to the <ul> containing the
videos and give it a class name of loading so that we can easily target it with some CSS. We
set a <span> element as the content of the new list item.
We can get the current value of the scrollTop property using jQuery's scrollTop()
method. We're using a cached value for the window's height. Our scroll handler will be
fairly intensive as it will be invoked every time the user scrolls, so using a cached value for
the window height makes this process slightly more efficient.
It does mean however that if the window is resized, this value will no longer be accurate.
We fix this by adding a resize handler for the window which recalculates this value every
time the window is resized. This is done by chaining another call to the on()method after
the scroll handler, which looks for the resize event of the window object and updates the
winHeight variable accordingly.
We then use jQuery's when() method once again, which invokes our getData() function
to retrieve the next 25 videos. We also use the done() method again to execute a callback
function once the request has completed.
Within this callback function we increment our startIndex variable by 25 again, ready to
request the next set of videos. The getData() function will populate our data object with
the new video data so all we need to do is call our renderer() function to display the new
videos, and then remove the temporary loader.
206
Project 8
At this point we should have a fully functional infinite loader that loads more videos when
the user scrolls to the bottom of the page. We should be able to run the page and see
something like the following when we scroll to the bottom:
Mission Accomplished
A lot of the code we've written in this project has been concerned with getting the data
we want to display. Actually adding the infinite scroll feature itself requires only a minimal
amount of code – a single handler that watches for the scroll event and triggers a new
request for more data when the document is scrolled to the bottom.
As you can see this is a feature that would be easy to retrofit to existing functionality as
an additional layer. This technique is best suited to data that can easily be arranged in
reverse-chronological order, with new items appearing at the top and older items
appearing at the bottom.
207
Infinite Scrolling with jQuery
It's not necessarily a complete replacement for paged data, but certainly makes sense when
dealing with things such as news stories, blog posts, tweets, or status updates. It works very
well with social data.
You Ready To Go Gung HO?
A Hotshot Challenge
In this project we're simply providing links back to a full-screen video player for each video
on the YouTube site. So when the visitor clicks on a video thumbnail or title, they'll be sent
off to YouTube to actually watch the video.
While there's nothing inherently wrong in doing this, a much cooler thing to do would be
to open up a lightbox containing a video player embedded in an <iframe>. This would let
the visitor view the video without ever leaving your site. The response from YouTube for the
videos contains a link that can be used as the src attribute of an <iframe>, so why not see
if you can hook this up yourself?
You'll notice that if you scroll to the bottom of the page, but then carry on scrolling down
immediately, the same set of videos are requested multiple times. As an alternative task, see
if you can prevent this from happening by only requesting more data if there is not currently
a request in progress.
This should be very easy to set up and simply involves setting a flag when the request begins,
and removing the flag when it ends. We can then only make requests if the flag is not set.
208
Project 9
A jQuery Heat Map
A heat map can tell you a lot about how your website is used. It's a valuable tool in the world
of analytics that can tell you which features of your website are the most used, and which
areas might need some improvement in order to truly engage visitors.
Mission Briefing
In this project we're going to build our own heat map to record which areas of any page are
clicked the most. We'll need to build a way of actually recording where every click occurs and
a way of transmitting that information somewhere so that it can be stored.
We'll actually be building two different bits of the overall heat map – the client-side part that
is executed in visitors' browsers to capture the clicks, and an admin console that displays the
heat map to the owner of the site.
We'll need to account for different resolutions and devices in order to capture the
maximum amount of information and ensure that our script is efficient enough to
run in the background unnoticed.
A jQuery Heat Map
Nothing visible will happen at the client side of course (all that part will do is record and
store the clicks) but at the end of the project we'll be able to display detailed information
in the admin console on the number and positions of all clicks on a page, as shown in the
following screenshot:
Why Is It Awesome?
All analytics are useful to the owner of a website and can give detailed information about the
people visiting the site, including things such as their computing environment, which pages
they enter the site on, which pages they leave from, and how many pages they visit.
A heat map can be just as informative, but from a developer's point of view instead of an
owner's. Which bits of your pages are clicked most frequently? A heat map can tell you.
The heat map that we'll build will be for responsive websites that are able to change their
layout to suit the screen width of the device being used to view the site. A single project
is not nearly enough to cover all aspects of responsive design, and because we're focusing
mainly on the script itself, we won't be covering it in much detail.
210
Project 9
If you've used responsive techniques already then you won't need the additional
information. If you haven't worked with responsive principles before, this should be a
gentle introduction to the subject that should act as a primer for the subject.
Your Hotshot Objectives
In this project we'll cover the following tasks:
ff
Determining and saving the environment
ff
Capturing visitor clicks
ff
Saving the click data
ff
Adding the management console
ff
Requesting click data
ff
Displaying a heat map
ff
Allowing different layouts to be selected
ff
Showing heat maps for each layout
Mission Checklist
This is the only project where we aren't going to build the HTML and CSS that we need
ourselves. We want our heat map to work with a variety of layouts and the best way to test
that is with a responsive layout. Were we to code this ourselves, it would take us most of this
project to code and discuss just the layout, before we even got to the heat map.
We'll use a prebuilt responsive template for this project then, so that we can jump straight
into the fun part without getting distracted. The template that we'll use is called Simplex,
but unfortunately it's no longer available online. You'll need to use the template files found
in the accompanying download for this book. Simply copy the simplex folder from the
downloaded archive in to the main jquery-hotshots project directory. All we need to
do is add a few script references to each of the HTML pages in the template. The files that
should be updated are:
ff
contact.html
ff
gallery.html
ff
index.html
ff
who-we-are.html
The new <script> elements can go at the bottom of the <body> in each of the four pages.
First we need jQuery obviously:
<script src="../js/jquery-1.9.0.min.js"></script>
211
A jQuery Heat Map
We'll also be making use of the imagesLoaded plugin that we used in the last project:
<script src="../js/jquery.imagesloaded.min.js"></script>
In this project we'll create two scripts, one to run on the client and one for the admin
console. Initially we'll be working with the client script, so we should add the following to
each page:
<script src="../js/heat-map-client.js"></script>
This file doesn't exist yet of course, so while we're getting set up we can go ahead and create
this now. It should be saved in the js directory along with our other scripts.
Determining and saving the
environment
In our first task we'll store some information about the current browsing environment,
such as the URL of the current page. We'll also parse any attached style sheets looking
for media queries.
Prepare for Lift Off
We'll start with our document ready shortcut, as we have in most of our other projects. In
the heat-map-client.js file, add the following code:
$(function () {
});
All the additional code we add to this file will go into this callback function.
Engage Thrusters
We'll first set a series of variables that will be used throughout the script. We'll also need to
parse any attached style sheets and look for media queries so that we can determine which
breakpoints are defined for the different layouts.
Media queries are a way to specify in CSS that a group of styles should only be
applied if certain conditions are met, such as the width of the screen. For more
information see http://en.wikipedia.org/wiki/Media_queries.
212
Project 9
Add the following code inside the callback function we just added:
var doc = $(document),
clickStats = {
url: document.location.href,
clicks: []
},
layouts = [];
$.ajaxSetup({
type: "POST",
contentType: "application/json",
dataType: "json"
});
$.each(doc[0].styleSheets, function (x, ss) {
$.each(ss.rules, function (y, rule) {
if (rule.media&&rule.media.length) {
var jq = $,
current = rule.media[0],
mq = {
min: (current.indexOf("min") !== -1) ?
jq.trim(current.split("min-width:")[1]
.split("px")[0]) : 0,
max: (current.indexOf("max") !== -1) ?
jq.trim(current.split("max-width:")[1]
.split("px")[0]) : "none"
};
layouts.push(mq);
}
});
});
layouts.sort(function (a, b) {
return a.min - b.min;
});
$.ajax({
url: "/heat-map.asmx/saveLayouts",
data: JSON.stringify({ url: url, layouts: layouts })
});
213
A jQuery Heat Map
Objective Complete - Mini Debriefing
We start out by defining a series of variables. We cache a reference to the document object
and wrap it with jQuery functionality. We then create an object called clickStats which
we'll use as a general storage container for the session.
Within the object we store the URL of the page and define an empty array called clicks
which will be used to store each click that occurs. Lastly we create another array, this time
outside of our clickStats object, which we'll use to store objects representing each layout
for the document.
We also set some defaults for any AJAX requests using jQuery's ajaxSetup() method,
which accepts an object containing the options to set. We'll be making a couple of requests,
so it makes sense to set defaults for any options that we set in both requests. In this example
we'll need to set the type to POST, the contentType to application/json, and the
dataType to json.
Our next block of code is concerned with parsing any style sheets attached to the document
via <link> elements and extracting any media queries defined in them.
We start by using jQuery's each() method to iterate the style sheet objects stored in the
StyleSheets collection of the document object. For each style sheet, there will be an object
in the collection that contains all of its selectors and rules, including any media queries.
The collection that we're iterating consists of objects, so the callback function we pass to
the each() method will receive the index of the current object, which we set to x, and the
current object itself, which we set to ss as arguments.
Within our callback function we again use jQuery's each() method. This time we're
iterating the rules collection of the ss object passed into the callback function. This
collection will contain a series of objects. The callback function we pass to this incantation
of the method will receive the index once again, this time set to y, and the current object,
this time set to rule.
The type of object will differ depending on what it is. It may be a CSSImportRule for @
import statements, a CSSFontFaceRule for @font-face rules, a CSSStyleRule for any
selectors defined by the style sheet, or a CSSMediaRule for any media queries.
We are only interested in the CSSMediaRule objects, so within the callback in the nested
each() we first check whether the rule object has a media property and if the media
property has a length.
Only CSSMediaRule objects will have a media property, but this property may be empty,
so we can check for the presence of this property and check that it has length using an if
conditional within the nested callback.
214
Project 9
If both of these conditions are true (or truthy) we know we've found a media query. We
first set a couple of new variables. The first variable is the first item of the media collection,
which will contain the text string defining the media query, and the second is an object called
mq that we'll use to store the breakpoints of the media query.
We set two properties of this object – the min and max values of the media query. We set
the min property by checking whether the text string contains the word min. If it does we
first split the string on the term min-width: and take the second item in the array that the
split() function will return, and then split this resulting string on the term px and take the
first item. We can chain split() like this because the function returns an array, which is
also what it is called on.
If the string does not contain the word min we set the value to 0. We do the same to extract
the max-width if there is one. If there isn't a max-width, we set it to the string none. Once
we've created our layout object we push it into the layouts array.
Lastly we sort our breakpoints array so that it is in ascending order. We can do this by passing
a sorting function to JavaScript's sort() method, which is called on an array. The function
we pass in will receive two items from the array we are sorting.
If the min property of first object is less than the min property of the second b object, the
function will return a negative number, which puts the smaller number before the larger
number in the array – exactly what we want.
So we'll end up with an array where each item is a specific breakpoint, which increases
throughout the array, making checking against it later on to detect which breakpoint is being
applied much easier.
Lastly we need to send this data to the server, potentially so that it can be saved. The only
options we need to set for this request are the URL to make the request to, and the data
option which we use to post the URL of the page and the array of media queries to the
server. The AJAX defaults we set earlier will also be used of course.
Classified Intel
If you're familiar with media queries already, feel free to skip to the start of the next task; if
not we'll just look at them briefly here so that we all know what our script is trying to do.
A media query is like an if conditional, but expressed in CSS. A media query in a CSS file will
look something like the following code snippet:
@media screen and (max-width:320px) {
css-selector { property: style; }
}
215
A jQuery Heat Map
The statement begins with @media to signify a media query. The query specifies a medium,
such as screen, and optionally additional conditions, such as max-width or min-width.
The styles contained within the query will only be applied if the query is satisfied.
Media queries are one of the staples of responsive web design, the other being relative
dimensions. Typically, a responsively built web page will have one or more media queries
that allow us to specify different layouts for a range of screen sizes.
Each media query we include will set a breakpoint between layouts. When a breakpoint is
exceeded, such as when the maximum width of the device is less than 320px in the previous
media query, the layout changes as directed by the media query.
Capturing visitor clicks
In this task we need to build the part that captures any clicks that occur on the page. While
the page is open we want to record information about the layout and the click itself.
Engage Thrusters
We can capture clicks and record the other information we want to store using the following
code, which should be added directly after the ajax() method that we added to heatmap-client.js in the last task:
$.imagesLoaded(function() {
doc.on("click.jqHeat", function (e) {
var x = e.pageX,
y = e.pageY,
docWidth = doc.outerWidth(),
docHeight = doc.outerHeight(),
layout,
click = {
url: url,
x: Math.ceil((x / docWidth) * 100),
y: Math.ceil((y / docHeight) * 100)
};
$.each(layouts, function (i, item) {
var min = item.min || 0,
max = item.max || docWidth,
216
Project 9
bp = i + 1;
if (docWidth>= min &&docWidth<= max) {
click.layout = bp;
} else if (docWidth> max) {
click.layout = bp + 1;
}
});
clickStats.clicks.push(click);
});
});
Objective Complete - Mini Debriefing
We can listen for clicks on the page by adding a handler using jQuery's on() method, and
we'll also want to ensure that any images in the page have loaded fully before we start
capturing clicks, because images will affect the height of the document, and that in turn will
affect our calculations. Therefore we need to attach our event handler inside the callback
function for the imagesLoaded() method.
We specify click as the event to listen for as the first argument, but also namespace the
event with jqHeat. We'll probably want to use this on a range of pages, each of which may
have its own event handling code that we don't want to interfere with this code.
Within the event handler we first need to set up some variables. The function receives the
event object as an argument and we use this to set our first two variables, which store the
x and y positions of the click. This figure will represent pixel points on the page.
We then store the width and height of the document. The reason we store this on every click
is because the width, and therefore the height of the document, may change while the page
is open.
People say that only developers resize their browsers while they are testing responsive
builds, but this is not always the case. Depending on the breakpoints defined by any media
queries in use, a change in device orientation could affect the width and height of the
document, which could happen at any time after the page has loaded.
We define the layout variable next but we don't assign a value to this for now. We also
create a new object to represent the click. Within this object we initially store the click
coordinates as percentages.
217
A jQuery Heat Map
Converting the pixel coordinates into percentage coordinates is a trivial operation that
simply involves dividing the pixel coordinate by the width (or height) of the document and
then multiplying the figure by 100. We use JavaScript's Math.ceil() function so that the
number is rounded up to the next integer.
Next we need to determine which layout we are in. We can iterate our layouts array using
jQuery's each() method again. The callback function receives the index of the current item
in the layouts array as the first argument and the second is the actual object.
Inside the callback function we first set our variables. This time the variables we need are the
minimum width of the layout, which we set to either the min property of the object, or zero
if there is no min defined. We also set the max variable to either the max property of the
current item, or the width of the document if there is no max property.
Our last variable simply adds 1 to the current index. The index will be zero-based but it
makes more sense for our layouts to be labeled 1 to the number of layouts instead of 0 to
the number of layouts.
We then use an if conditional to figure out which layout is currently being applied. We first
check whether the current document width is greater or equal to the minimum value for
the media query and less than or equal to the maximum. If it is, we know we're inside the
current layout and so save the converted layout index to our click object.
If we haven't matched any of our layouts, the browser must be larger than the highest
max-width value defined by a media query, so we set the layout to the converted layout
plus one again. Lastly we add the click object that we've created to the clicks array in
our clickStats object.
Saving the click data
Someone has visited a page on which our heat map client script is running, they've clicked
around, and our script so far has recorded each of those clicks. Now what? Now we need a
way of transmitting that information to the server for permanent storage and display in the
management console. This is what we'll look at in this task.
Engage Thrusters
We can ensure that any captured clicks are sent to the server for permanent storage using
the following code, which should be added after the imagesLoaded()callback function:
window.onbeforeunload = function () {
$.ajax({
async: false,
218
Project 9
type: "POST",
contentType: "application/json",
url: "/heat-map.asmx/saveClicks",
dataType: "json",
data: JSON.stringify({ clicks: clicks })
});
}
Objective Complete - Mini Debriefing
We attached a handler for the beforeunload event to the window object in order to post
the data to the server before leaving the page. Unfortunately, this event isn't handled 100
percent of the time – there may be occasions when it doesn't fire.
In order to minimize this as much as possible we've attached the event handler directly to
the native window object, not the jQuery wrapped one, which we can access via the first
item in the array that is the jQuery object.
Using any jQuery method, including on(), adds overhead in that jQuery method, and as the
underlying JavaScript functions are invoked. To reduce this overhead as much as possible we
avoid using jQuery here and revert back to the old-school way of attaching event handlers by
prefixing the event name with on and assigning a function as their value.
All we need to do inside this function is post the data to the server so that it can be inserted
into a database. We use jQuery's ajax() method to make the request, and set the async
option to false to make the request synchronous.
This is important and will ensure the request is made in Chrome. We're not interested in a
response from the server anyway – we just need to ensure the request is made before the
page unloads.
We also set the type to POST as we're sending data to the server, and set the contentType
to application/json, which will set the appropriate headers for the request to ensure
that the data is handled correctly on the server.
The url is clearly the URL for the web service that we're sending the data to, and we set the
dataType to json, which again can help to make consuming the data on the server easier.
Lastly we stringify the clicks array and wrap it in an object using the browser's native JSON
engine. We send the stringified data to the server using the data option.
At this point, when a page that this script is attached to is opened, the script will run quietly
in the background recording the coordinates of any points on the page that are clicked.
When the user leaves the page, the click data they've generated is dispatched to the server
for storage.
219
A jQuery Heat Map
Classified Intel
Browsers that do not have a JSON engine, such as Version 7 of Internet Explorer and below,
will not be able to run the code we added in this task, although there are polyfill scripts that
can be used in these situations.
For more information see the JSON repository at Github (https://github.com/
douglascrockford/JSON-js).
Adding the management console
I said at the start of this project that we wouldn't need to write any HTML or CSS. That
was a minor exaggeration; we'll have to build the management console page ourselves,
but don't worry, we won't have to write much – most of what we'll display on the page
will be created dynamically.
Prepare for Lift Off
Create a new HTML page based on our standard template file called console.html and
save it in the simplex directory that we've been working in for this project. Next create a
new script file called console.js and save it in the same folder. Lastly, create a new style
sheet called console.css and save it in the css folder inside the simplex directory.
We should link to the new style sheet from the <head> of our new HTML page:
<link rel="stylesheet" href="css/console.css" />
We should also link to jQuery and our new script file at the bottom of the <body>:
<script src="../js/jquery-1.9.0.min.js"></script>
<script src="console.js"></script>
Lastly, we should add the class name jqheat to the <body> element:
<body class="jqheat">
Engage Thrusters
The page will need to display an interface that can be used to select a page to view the click
stats for. Add the following code to the <body> of console.html:
<header>
<h1>jqHeat Management Console</h1>
<fieldset>
<legend>jqHeat page loader</legend>
<input placeholder="Enter URL" id="url" />
220
Project 9
<button id="load" type="button">Load page</button>
</fieldset>
</header>
<section role="main">
<iframe scrolling="no" id="page" />
</section>
We can also add some very basic CSS for these elements. Add the following code to
console.css:
.jqheat{ overflow-y:scroll; }
.jqheat header {
border-bottom:1px solid #707070; text-align:center;
}
.jqheat h1 { display:inline-block; width:100%; margin:1em 0; }
.jqheat fieldset {
display:inline-block; width:100%; margin-bottom:3em;
}
.jqheat legend { display:none; }
.jqheat input {
width:50%; height:34px; padding:0 5px;
border:1px solid #707070; border-radius:3px;
}
.jqheat input.empty{ border-color:#ff0000; }
.jqheat button { padding:9px5px; }
.jqheat section {
width:100%;margin:auto;
position:relative;
}
.jqheat iframe, .jqheat canvas {
Width:100%; height:100%; position:absolute; left:0; top:0;
}
.jqheat canvas { z-index:999; }
In this task we won't add any real functionality, but we can prepare our script file with the
usual document ready handler in preparation for the next task. In console.js, add the
following code:
$(function () {
});
221
A jQuery Heat Map
Objective Complete - Mini Debriefing
Our page starts out with a <header> element that contains a heading for the page in <h1>
and <fieldset>. Inside the <fieldset> is the mandatory <legend> and a very simple
UI for the page, which consists of a single <input> and a <button> element. Both the
<input> and the <button> elements have id attributes so that we can easily select them
from the page in our script.
The main content area of the page consists of a <section> element, which is given a role
attribute of main. It's standard practice to markup the main content area of the page using
this attribute, which helps clarify the intent of the area to assistive technologies.
Inside the <section> is an <iframe>. We'll use the <iframe> to display the page the
user wants to view click statistics for. For now it just has an id attribute, again so that we
can select it easily, and the non-standard scrolling attribute set to no. I'm not a huge
fan of using non-standard attributes, but in this case it is the simplest way to prevent the
<iframe> having a pointless scrollbar when the content document is loaded into it.
There is a high chance of the page causing a scrollbar, and rather than have the movement
that occurs when the scrollbar appears, we may as well set the <body> of our page to have a
vertical scrollbar permanently, which we do in the CSS. Other than this, the CSS is mostly just
positional stuff which we won't look at too deeply.
Classified Intel
We used the HTML5 placeholder attribute on our <input> element, which in supporting
browsers displays the text added as the value of the attribute inside the <input>, acting as
an inline label.
This is useful because it means we don't have to add a whole new element just to
display a <label>, but support, while good at the time of writing, is not 100 percent.
Luckily there are some excellent polyfills that handle providing a sensible fallback
in un-supporting browsers.
There are a whole range of placeholder polyfills (and many more
besides) recommended by the Modernizr team. You can see the list in
full by visiting https://github.com/Modernizr/Modernizr/
wiki/HTML5-Cross-Browser-Polyfills.
222
Project 9
Requesting click data
The console page starts out almost empty, containing mostly a form for loading the page
that we want to view click data for. In this task we'll look at how we can load that page and
request its data from the server.
Engage Thrusters
Start out by adding the following code to the empty function in console.js:
var doc = $(document),
input = doc.find("#url"),
button = doc.find("#load"),
iframe = doc.find("#page"),
canvas = document.createElement("canvas");
$.ajaxSetup({
type: "POST",
contentType: "application/json",
dataType: "json",
converters: {
"textjson": function (data) {
var parsed = JSON.parse(data);
return parsed.d || parsed;
}
}
});
We can then add a click handler for the <button> element:
doc.on("click", "#load", function (e) {
e.preventDefault();
var url = input.val(),
len;
if (url) {
input.removeClass("empty").data("url", url);
button.prop("disabled", true);
iframe.attr("src", url).load(function() {
$(this).trigger("iframeloaded");
});
} else {
input.addClass("empty");
button.prop("disabled", false);
}
});
223
A jQuery Heat Map
Lastly, we can add an event handler for the custom iframeloaded event:
doc.on("iframeloaded", function () {
var url = input.data("url");
$.ajax({
type: "POST",
contentType: "application/json",
url: "/heat-map.asmx/getClicks",
dataType: "json",
data: JSON.stringify({ url:url, layout: 4 }),
converters: {
"textjson": function (data) {
var parsed = JSON.parse(data);
returnparsed.d || parsed;
}
}
});
});
Objective Complete - Mini Debriefing
We began, as we usually do, by setting some variables. We stored a reference to the
document object wrapped in jQuery, which we can use to select any element on the page
using this reference as a starting point, without creating a new jQuery object each time we
select an element or bind an event handler.
We also stored a reference to the <input> element that will contain the URL of the
page, a reference to the <button> next to the <input>, and the <iframe> that we'll
render the requested page into. Lastly we set an undefined variable called canvas, which
we'll use to store a reference to a <canvas> element that we create using JavaScript's
createElement() function.
We could use jQuery to create this element of course, but we're only creating a single
element as opposed to a complex DOM structure, so we may as well use plain JavaScript and
get a performance boost at the same time.
As before we can use the ajaxSetup() method to set the type, contentType, and
dataType options for the requests that we'll be making to the server. We also used a
converter to transform the data that will be returned by the server.
The converters option takes an object where the data type the converter should be used
for is specified as a key, and the function to use as the converter is specified as a value.
224
Project 9
Some servers will return JSON data wrapped in an object and stored in the property d for
security, while other servers do not do this. Usually the text json data type would be
parsed using jQuery's parseJSON() method, but in this case our code would still need to
extract the actual data from the object before it could be used.
Instead our converter parses the JSON using the browser's native JSON parser and then
returns either the contents of d if it exists, or the parsed data. This means that the code
which processes the data is the same regardless of whether the data is wrapped in an object
or not.
While not critical in this particular example, converters can be extremely useful in situations
where code is being distributed and the platform on which it will run is not known in advance.
Next we added a click handler to the document using jQuery's on() method in
event-delegation mode. To add a delegated handler, we attach the handler to a parent
element, in this case document, and use the second argument to on() to provide a
selector that events should be filtered against.
Events bubble up the page, from the triggering element right up to the outer window object.
The handler is only executed when the triggering element matches the selector passed
as the second argument. The first argument is of course the type of event, and the third
argument is the handler function itself.
Within the function we first prevent the default browser action for the event and then store
the value of the <input> element in the variable url. We also set an undefined variable
called len. We don't need to use that yet, but we will later on.
Next we check whether the url variable we set has a truthy value, like a string with length
for example. If it does, we remove the class name empty if the <input> element has it,
then set the contents of the <input> as the element's data using jQuery's data() method.
Associating the URL to the element in this way is a great way to persist the data so that it
can be obtained from other functions in the code which do not have access to the event
handler's scope. We also disable the <button> to prevent duplicate requests. We can
enable it later on when the heat map has been painted to the screen.
We then added the URL that we obtained from the <input> element as the src property of
the <inframe>, which causes the <iframe> to load the page the URL is for. We add a handler
for the load event that will be fired by the <iframe> once the page has loaded. Inside this
handler we fire a custom iframeloaded event using jQuery's trigger() method.
If the url variable does not contain a truthy value, we add the empty class to the <input>
and enable the <button> once more.
225
A jQuery Heat Map
Lastly we added an event handler for the custom iframeloaded event. Custom events
will bubble up to the document just like regular events so we can attach the handler to our
cached <body> element and it will still be triggered at the appropriate time.
Within this handler we get the URL for the page that has been loaded by recalling the data
associated with the <input> element. We then make a request to the server using jQuery's
ajax() method.
We've already set some of the required AJAX options as defaults using ajaxSetup() once
again, so for this request we just set the url and the data options. This time the data we
send is a stringified object containing the URL of the page, and the layout to get click data
for. In response to this, we expect a JSON object containing a series of click objects, where
each object contains x and y coordinates that refer to a specific point on the page.
Note that at this point, we're hardcoding which layout to load into the request, which we
set to number 4. We'll come back to this in the next part and allow the user to select which
layout to view.
Displaying a heat map
We're all set to display a heat map. In this task we'll process the click data in order to generate
the heat map, and then display it using the <canvas> element on top of the <iframe>.
Engage Thrusters
First of all we can add a success handler for the AJAX request we made at the end of the last
task. We can do this by chaining the done() method to the ajax() method:
}).done(function (clicks) {
var loadedHeight = $("html", iframe[0].contentDocument)
.outerHeight();
doc.find("section").height(loadedHeight);
canvas.width = doc.width();
canvas.height = loadedHeight;
$(canvas).appendTo(doc.find("section"))
.trigger("canvasready", { clicks: clicks });
});
226
Project 9
We can then add a handler for the custom canvasready event. This should be added
directly after the iframeloaded event handler:
doc.on("canvasready", function (e, clickdata) {
var docWidth = canvas.width,
docHeight = canvas.height,
ctx = canvas.getContext("2d") || null;
if (ctx) {
ctx.fillStyle = "rgba(0,0,255,0.5)";
$.each(clickdata.clicks, function (i, click) {
var x = Math.ceil(click.x * docWidth / 100),
y = Math.ceil(click.y * docHeight / 100);
ctx.beginPath();
ctx.arc(x, y, 10, 0, (Math.PI/180)*360, true);
ctx.closePath();
ctx.fill();
});
}
button.prop("disabled", false);
});
Objective Complete - Mini Debriefing
Once the AJAX request has completed, we first store the height of the document that has
been loaded in the <iframe>. The jQuery method may be passed a second argument after
the selector, which sets the context that should be searched to match the selector. We can
set the context to be the contentDocument object of the first <iframe> on the page,
which we can access using frame[0].
Setting the height of the <section> element will automatically make the <iframe> and
the <canvas> element that we created earlier the width and height of the <section> so
that the page can be viewed full size.
227
A jQuery Heat Map
Next we set the width and height properties of the <canvas> element we created in the
last task. We haven't set the width or height attributes of the <canvas> element yet, so
by default it will only be 300 x 300 pixels in size, regardless of its visible size as set by the CSS.
We therefore set the attributes to the correct sizes.
We can then append the new <canvas> to the <section> element on the page, and then
fire the custom canvasready event. We're going to want to use the data passed by the
server in an event handler for this event, so we pass this to the handler function using the
second argument of the trigger() method.
We then added a handler for the canvasready event. This function receives the event
object and the click data as parameters. Within the function we first get the width and
height of the <canvas> element. We stored the click data as percentages and we'll need
to convert them back to pixel values.
In order to paint on the <canvas>, we'll need to get a context. We can get a 2D context for
the <canvas> and store it in a variable using the getContext() function of the canvas
object. If the <canvas> element isn't supported, the ctx variable will be set to null. We
can then only proceed to interact with the canvas if the context is not null.
If ctx is not null, we first clear the <canvas> using the clearRect() function of the
canvas API and then set the color that we'll be painting on the canvas. We can set it to the
RGBA (Red, Green, Blue, Alpha) string 0,0,255,.05 which is a semi-transparent blue. This
only needs to be set once.
We then iterate over the click data returned by the server using jQuery's each() method.
The iterator function, which will be executed for the number of items in the clicks array, is
passed the index of the current item in the array and the click object.
We first store the x and y positions of each pixel that was clicked. These figures are currently
percentages, so we need to convert them back to pixel values. This is just the opposite
calculation that we performed in the client part of the heat map. We just multiply the
percentage by the width or height of the <canvas> and then divide that figure by 100.
We can then paint a dot on the <canvas> where the click occurred. We do this by starting a
new path using the beginPath() method of the canvas object. The dot is drawn using the
arc() method, which is passed a number of arguments. The first two are the coordinates of
the center of the arc, which we set to the x and y values we've just calculated.
The third argument is the radius of the circle. If we set the dot to be a single pixel the data
would be quite difficult to interpret, so using a large dot instead of a single pixel improves
the appearance of the heat map dramatically.
The third and fourth arguments are the angles to start and stop the arc at, and are in
radians not degrees. We can paint a complete circle by starting at zero radians and
ending on about 6.5 radians.
228
Project 9
Once the arc has been defined, we can close the path using the closePath() method and
fill the arc with color using the fill() method. At this point we should be able to run the
console in a browser, enter the URL of one of the template pages, and see the page with dots
on it corresponding to the clicks.
Allowing different layouts to be
selected
In this task of the project, we need to allow the user to select each layout supported by the
page. We can do this by using a <select> box which we populate at page load with each of
the different layouts.
Engage Thrusters
First of all we can add the <select> element to the page. This can go in between the search
field and the button at the top of the page in console.html:
<select id="layouts"></select>
Next we need to make a request at page load to populate the <select>element with
an <option> for each of the different layouts. We can do this in the click handler for the
<button> that we added in console.js earlier.
It will need to go into the first branch of the conditional that checks a URL has been entered
into the <input>, directly before where we set the src of the <iframe>:
$.ajax({
url: "/heat-map.asmx/getLayouts",
data: JSON.stringify({ url: url })
}).done(function (layouts) {
var option = $("<option/>"),
max;
len = layouts.length;
function optText(type, i, min, max) {
var s,
t1 = "layout ";
switch (type) {
case "normal":
229
A jQuery Heat Map
s = [t1, i + 1, " (", min, "px - ", max, "px)"];
break;
case "lastNoMax":
s = [t1, len + 1, " (", min, "px)"];
break;
case "lastWithMax":
s = [t1, len + 1, " (", max, "px+)"];
break;
}
return s.join("");
}
$.each(layouts, function (i, layout) {
var lMin = layout.min,
lMax = layout.max,
text = optText("normal", i, lMin, lMax);
if (i === len - 1) {
if (lMax === "none") {
text = optText("lastNoMax", null, lMin, null);
} else {
max = lMax;
}
}
option.clone()
.text(text)
.val(i + 1)
.appendTo("#layouts");
});
if (max) {
var fText = optText("lastWithMax", null, null, max);
option.clone()
.text(fText)
.val(len + 1)
.prop("selected",true)
.appendTo("#layouts");
}
});
230
Project 9
We can also add a little CSS for our new <select> element. We can just drop this into the
bottom of console.css:
.jqheat select {
width:175px; height:36px; padding:5px;
margin:0 .25em 0 .5em; border:1px solid #707070;
border-radius:3px;
}
Objective Complete - Mini Debriefing
First of all, we make the request to the server to get the layout information. The url is set
to a web service that returns the layouts and the data is the URL of the page we'd like the
layouts for.
We set a success handler using the done() method as is the recommended technique for
adding success handlers to promise objects for when they become resolved. Within the
handler we first set some variables.
We create an <option> element as we'll need one of these for each layout and so can clone
it as many times as we need by using the clone() method. We also update the len variable
that we created earlier but left undefined, to the number of layouts, which is the length of
the array the function will receive, and an undefined variable called max.
Next we define a function called optText() that we can use to generate the text for each
<option> element we create. This function will accept the type of string to create, the
index, and the min and max values.
Inside this function we set a couple of variables. The first, called s, is undefined at this point.
The second variable t1 is used to store some simple text that is used in each variant of
the string.
Then we use a switch conditional to determine which string to build based on the type,
which will be passed into the function as the first parameter and will be set to normal,
lastNoMax, or lastWithMax, and which should account for the different types of media
query that may be found.
In the normal case, we specify the min and max values. When there is no max value, we
build the string using the min value and when there is a max value, we build the string using
the max value.
Each string is constructed using an array, then at the end of the function we return a string
by joining whichever array was created.
231
A jQuery Heat Map
We then use jQuery's each() method to iterate over the layouts object returned by
the server. As always, the iterator function is passed the index of the current item and the
current item itself as parameters.
Within the iterator function, we set our variables, which in this case are the min and max
property values from the current layout object, and the normal variant of the text string,
which we're definitely going to use at least once. We call our optText() function and
store the result for later use.
We then check whether we're on the last iteration, which we'll know when the index is equal
to the length of the layouts array, which we stored earlier, minus 1. If we are on the last
iteration we then check whether the max value is equal to the string none. If it is, we call our
optText() function again and set the text to the lastNoMax type, which generates the
required text string for us. If it isn't we set the max variable, which we initially declared as
undefined to the max value of the current object. Lastly, we create the <option> element
required for each object in the layouts array. The <option> is given the text that we've
set, and the value of the index plus 1. Once created, the <option> is appended to the
<select> element.
Lastly we check whether the max variable has a truthy value. If it does, we call our
optText() function once more, this time using the lastWithMax type, and create another
<option> element, which we set as the selected item. This is required because we have one
more layout than we have objects in the layouts array.
When we run the page in a browser now, we should find that as soon as we enter a URL into
the <input> and hit the load page, the <select> element becomes populated with an
<option> for each of the layouts.
Classified Intel
The middle case (lastNoMax) in the switch statement in our optText() function
actually won't be used in this example because of how the media queries in the template
we're using are structured. In this example the media query for the last break point is
769px to 1024px. Sometimes, media queries may be structured so that the final breakpoint
contains just a min-width.
I've included this case of the switch to make the code support this other type of
media query format as it is quite common and you're likely to run into it when using
media queries yourself.
232
Project 9
Showing heat maps for each layout
Now that we have each of the layouts in the <select> element, we can wire it up so that
when the selected layout is changed, the page is updated to show the heat map for that layout.
Engage Thrusters
In this task we'll need to modify some of the code written in a previous task. We need to
change the click handler for the <button> so that the layout isn't hardcoded into the request.
First of all we need to pass the len variable to the handler for the iframeloaded event. We
can do this by adding a second argument to the trigger() method:
$(this).trigger("iframeloaded", { len: len });
Now we need to update the callback function so that this object is received by the function:
doc.on("iframeloaded", function (e, maxLayouts) {
Now, we can change the bit where we hardcoded layout 4 into the data passed to the server
when making the request for click data:
data: JSON.stringify({ url: url, layout: maxLayouts.len + 1 }),
Now we're ready to update the heat map when <select> is changed. Add the following
code directly after the canvasready handler in console.js:
doc.on("change", "#layouts", function () {
var url = input.data("url"),
el = $(this),
layout = el.val();
$.ajax({
url: "/heat-map.asmx/getClicks",
data: JSON.stringify({ url: url, layout: layout })
}).done(function (clicks) {
doc.find("canvas").remove();
var width,
loadedHeight,
opt = el.find("option").eq(layout - 1),
text = opt.text(),
min = text.split("(")[1].split("px")[0],
section = doc.find("section"),
233
A jQuery Heat Map
newCanvas = document.createElement("canvas");
if (parseInt(layout, 10) === el.children().length) {
width = doc.width();
} else if (parseInt(min, 10) > 0) {
width = min;
} else {
width = text.split("- ")[1].split("px")[0];
}
section.width(width);
newCanvas.width = width;
loadedHeight = $("html",
iframe[0].contentDocument).outerHeight();
section.height(loadedHeight);
newCanvas.height = loadedHeight;
canvas = newCanvas;
$(newCanvas).appendTo(section).trigger("canvasready", {
clicks: clicks });
});
});
Objective Complete - Mini Debriefing
We begin by delegating our handler to the document as we have with most of our other
event handlers. This time we're listening for the change event triggered by the element with
an id of layouts, which is the <select> element we added in the last task.
We then continue following the previous form by setting a few variables. We get the URL
saved as the data of the <input> element. We also cache the <select> element and the
value of the selected <option>.
Next we need to make an AJAX request to get the heat map for the selected layout. We set
the url to a web service that will return this information, and send the url we want the
heat map for, and the layout, as part of the request. Don't forget that this request will also
use the defaults we set using ajaxSetup().
234
Project 9
We use the done() method once again to add a success handler for the request. When the
response is received we first remove the existing <canvas> element from the page, and
then set some more variables.
The first two variables are undefined to begin with; we'll populate these in just a moment.
We store the <option> that was selected so that we can get its text, which is stored in the
next variable. We get the minimum width of the breakpoint by splitting the text that we've
just stored, and then cache a reference to the <section> on the page. Lastly we create a
new <canvas> element to display the new heat map.
The conditional if statement that follows the variable deals with setting the first of our
undefined variables – width. The first branch tests whether the layout requested is the last
layout, and if so, sets the new <canvas> to the width of the screen.
If the last layout was not requested, the next branch of the conditional checks whether
the minimum width of the layout is greater than 0. If it is, the width variable is set to the
minimum breakpoint.
The final branch is used when the minimum width of the breakpoint is 0. If the minimum
width is 0, we use the maximum width of the breakpoint instead, which we can obtain by
splitting the text of the <option> once more.
We then set the width of the <section> element and the new <canvas> element using
the width that we've just computed.
Following this we can define our second undefined variable – loadedHeight. This is
calculated in the same way as it was before, by reaching into the document loaded into the
<iframe> and getting the height of its document object using jQuery's outerHeight()
method, which includes any padding the element may have. Once we have this value, we can
then set the height of the <section> element and the new <canvas> element.
We're about to trigger our canvasready event once more, as this will consume the
click data and generate the heat map. Before we do that however, we just need to save
the newly created <canvas> element back to the canvas variable that we set right
at the top of console.js.
235
A jQuery Heat Map
At this point we should be able to load the default heat map for a URL, then use the
<select> element to view the heat map for another layout:
236
Project 9
Classified Intel
I've used an MS SQL database to store the data, and a c# web service containing various web
methods required for this project. Included in the code download accompanying this book is
a backup of the database and a copy of the web service file for you to use if you wish.
MS SQL express is a free version of SQL server which the database can be restored to, and
Visual Studio 2012 for web, which is also free, will happily run the web service through its
built-in development server.
If you don't have these products installed, and you have access to a Windows machine,
I'd strongly recommend you install them so that you can see the code used in this project
in action. The open source alternatives PHP and MySQL could also be used just as easily,
although you'll need to write this code yourself.
Mission Accomplished
In this project we built a simple heat map generator that captured click data on web pages
built using responsive techniques. We built two parts to the heat map generator – some
code that runs in the browsers of the visitors of the website to capture every click on the
screen, and a part that is used in conjunction with a simple management console in which
the URL of the page the heat map is for, and the layout to display can be selected.
While we have to allow a margin of error to account for the pixel-to-percentage conversion
and back again, different screen resolutions, and the range between different breakpoints,
this easy-to-implement heat map can still give us valuable information about how our site is
used, which features are popular, and which features are wasted screen space.
You Ready To Go Gung HO?
A Hotshot Challenge
One thing we haven't dealt with is color. Our heat map is built by dots of a uniform blue. As
they're semi-opaque, they get darker as more dots appear in a condensed area, but with
enough data, we should aim to change the color, going through red, yellow, and right up to
white for the most-clicked areas. See if you can add this functionality yourself to really top
the project off.
237
Project 10
A Sortable, Paged
Table with Knockout.js
Knockout.js is a fantastic JavaScript Model-View-ViewModel (MVVM) framework that can
help you save time when writing complex, interactive user interfaces. It works very well with
jQuery and even has built-in basic templating support for building repeated elements that
show different data.
Mission Briefing
In this project we'll build a paged table from data using jQuery and Knockout.js. Client-side
paging itself is a great feature, but we'll also allow the table to be sorted by providing clickable
table headings, and add some additional features such as filtering the data based on a
particular property.
By the end of this mission we'll have built something that looks like the following screenshot:
A Sortable, Paged Table with Knockout.js
Why Is It Awesome?
Building complex UIs that respond rapidly to user interaction is hard. It takes time, and
the more complex or interactive an application is, the longer it takes and the more code it
requires. And the more code an application requires, the harder it is to keep it organized
and maintainable.
While jQuery is good at helping us to write concise code, it was never designed with building
large-scale, dynamic, and interactive applications in mind. It's powerful, and great at what it
does and what it was designed to do; it just wasn't designed to build entire applications with.
Something else is needed when building large-scale applications, something that provides a
framework within which we can keep code organized and maintainable. Knockout.js is one
such framework designed to do just that.
Knockout.js is known as an MVVM framework, which is based on three core components –
the Model, the View, and the ViewModel. This is similar to the better-known MVC pattern.
The intent of these and other similar patterns is to provide a clear separation of the visual
part of an application and the code required to manage the data.
The Model can be thought of as the data for the application. Really, the actual data is a result
of the Model, but when working on the client side, we can forget about how the data is
accessed by the server-side code because usually we just make an AJAX request and the data
is delivered to us.
The View is the visual representation of that data, the actual HTML and CSS used to present
the Model to the user. When using Knockout.js, this part of the application can also include
bindings, which map elements on the page to specific bits of data.
The ViewModel sits between the Model and the View, and is quite literally a model of the
View – a simplified representation of the state of the View. It manages the user interaction,
makes and handles requests for data, and then feeds the data back to the user interface.
Your Hotshot Objectives
The tasks needed to complete this mission are listed below:
ff
Rendering the initial table
ff
Sorting the table
ff
Setting the page size
ff
Adding Previous and Next links
ff
Adding numerical page links
ff
Managing class names
240
Project 10
ff
Resetting the page
ff
Filtering the table
Mission Checklist
We'll be using Knockout.js in this project, so you'll need to grab a copy of it now. The
latest version, 2.2.1 at the time this book went to print, can be downloaded from:
http://knockoutjs.com/downloads/index.html. This should be saved in the js
directory within the main jquery-hotshots project folder as knockout-2.2.1.js.
We also need some data for this project. We'll need to use a reasonably large data set
consisting of data that can be ordered in a variety of ways. We'll be using the periodic table
of elements, in JSON format, as our data source.
I've provided a file as part of this example called table-data.js, which contains an object
with a property called elements. The value of this property is an array of objects, where
each object represents an element. The objects are in the following format:
{
name: "Hydrogen",
number: 1,
symbol: "H",
weight: 1.00794,
discovered: 1766,
state: "Gas"
}
Rendering the initial table
In our first task of the project, we'll build a super-simple ViewModel, add a basic View,
and render the Model into a bare <table> without any enhancements or additional
functionality. This will allow us to familiarize ourselves with some of the basic principles of
Knockout, without dropping us in at the deep end.
Prepare for Lift Off
At this point we create the files we'll be using in this project. Save a copy of the template file as
sortable-table.html in the root project directory.
We'll also need a style sheet called sortable-table.css, which we should save in the
css folder, and a JavaScript file called sortable-table.js, which should of course be
saved in the js directory.
241
A Sortable, Paged Table with Knockout.js
The HTML file should link to each of these resources, as well as the knockout-2.2.1.js
file. The style sheet should be linked to directly after common.css, which we've used in most
of the projects in the book so far, while knockout.js, table-data.js, and the custom
script file for this project (sortable-table.js) should come after the link to jQuery, in
that order.
Engage Thrusters
First of all we can build the ViewModel. In sortable-table.js, add the following code:
$(function () {
var vm = {
elements: ko.observableArray(data.elements)
}
ko.applyBindings(vm);
});
Next, we can add the View, which is built from some simple HTML. Add the following markup
to the <body> of sortable-table.html, before the <script> elements:
<table>
<thead>
<tr>
<th>Name</th>
<th>Atomic Number</th>
<th>Symbol</th>
<th>Atomic Weight</th>
<th>Discovered</th>
</tr>
</thead>
<tbody data-bind="foreach: elements">
<tr>
<td data-bind="text: name"></td>
<td data-bind="text: number"></td>
<td data-bind="text: symbol"></td>
<td data-bind="text: weight"></td>
<td data-bind="text: discovered"></td>
</tr>
</tbody>
</table>
242
Project 10
Lastly, we can add some basic styling to our <table> and its contents by adding the
following code to sortable-table.css:
table {
width:650px; margin:auto; border-collapse:collapse;
}
tbody { border-bottom:2px solid #000; }
tbodytr:nth-child(odd) td { background-color:#e6e6e6; }
th, td {
padding:10px 50px 10px 0; border:none; cursor:default;
}
th {
border-bottom:2px solid #000;cursor:pointer;
position:relative;
}
td:first-child, th:first-child { padding-left:10px; }
td:last-child { padding-right:10px; }
Objective Complete - Mini Debriefing
In our script, we first added the usual callback function to be executed when the document
loads. Within this we created the ViewModel using an object literal stored in the variable vm.
The only property this object has is elements, the value of which is set using a Knockout
method. Knockout adds a global ko object that we can use to call methods. One of these
methods is observableArray(). This method accepts an array as an argument, and the
array passed in to the method becomes observable. This is the data for our application.
In Knockout, primitives such as strings or numbers can be observable and this allows them to
notify subscribers when their values change. Observable arrays are similar, except that they
are used with arrays. Whenever a value is added or removed from an observable array, it will
notify any subscribers.
After defining our ViewModel, we need to apply any bindings that may be present in the
View. We'll look at the bindings in just a moment; for now just be aware that until we call
Knockout's applyBindings() method, any bindings we add to our View won't be applied.
The HTML we added is almost unremarkable, it's just a simple <table>, with a column for
each property of an element. If you take a look inside the table-data.js file, you'll see
the properties of each element in the array match the <th> elements.
The first interesting thing is the data-bind attribute we added to the <tbody> element.
This is the mechanism Knockout uses to implement declarative binding. This is how we
connect elements in the View to properties of the ViewModel.
243
A Sortable, Paged Table with Knockout.js
The value of the data-bind attribute consists of two parts – the binding and the
ViewModel property to connect to. The first part is the binding, which we set to foreach.
This is one of Knockout's flow-control bindings, and acts in a similar way to a standard for
loop in regular JavaScript.
The second part of the binding is the ViewModel property to bind to. The one property our
ViewModel has at present is elements, which contains an observable array. The foreach
binding maps to an array and then renders any child elements it has for each item in the array.
The child elements that this element has are a <tr> and a series of <td> elements, so we'll
get a row in the table for each item in the elements array. In order to populate the <td>
elements with content we'll use another Knockout binding – the text binding.
The text binding binds to a single observable property, so we have a <td> that is bound to
each property in the objects within the elements array. The text of each <td> will be set to
the value of each property in the current array item.
The CSS we added at the end of the task was purely for presentational purposes and contains
nothing relevant to Knockout or jQuery. At this point we should be able to run the page in a
browser and see the data from table-data.js displayed in a neat <table>.
Classified Intel
The binding between elements of the View and properties of the ViewModel are at the heart
of Knockout. The ViewModel is a simplified version of the state of the UI. Because of the
bindings, any time the underlying ViewModel changes, the view will be updated to reflect
those changes.
So if we programmatically add a new element object to the observable array, the <table>
will instantly be updated to show the new element. Similarly, if we remove an item from the
array in the ViewModel, the corresponding <tr> will be immediately removed.
Sorting the table
In this task we can change the <th> elements so that they are clickable. When one of them
is clicked, we can then sort the table rows by the column that was clicked.
244
Project 10
Engage Thrusters
First of all we can update the <tr>and the <th> elements that it contains in
sortable-table.html:
<tr data-bind="click: sort">
<th data-bind="css: nameOrder">Name</th>
<th data-bind="css: numberOrder">Atomic Number</th>
<th data-bind="css: symbolOrder">Symbol</th>
<th data-bind="css: weightOrder">Atomic Weight</th>
<th data-bind="css: discoveredOrder">Discovered</th>
</tr>
Next we can add some new observable properties to our ViewModel in
sortable-table.js:
nameOrder: ko.observable("ascending"),
numberOrder: ko.observable("ascending"),
symbolOrder: ko.observable("ascending"),
weightOrder: ko.observable("ascending"),
discoveredOrder: ko.observable("ascending"),
We also add a new method called sort:
sort: function (viewmodel, e) {
var orderProp = $(e.target).attr("data-bind")
.split(" ")[1],
orderVal = viewmodel[orderProp](),
comparatorProp = orderProp.split("O")[0];
viewmodel.elements.sort(function (a, b) {
var propA = a[comparatorProp],
propB = b[comparatorProp];
if (typeof (propA) !== typeof (propB)) {
propA = (typeof (propA) === "string") ? 0 :propA;
propB = (typeof (propB) === "string") ? 0 :propB;
}
if (orderVal === "ascending") {
245
A Sortable, Paged Table with Knockout.js
return (propA === propB) ? 0 : (propA<propB) ? -1 : 1;
} else {
return (propA === propB) ? 0 : (propA<propB) ? 1 : -1;
}
});
orderVal = (orderVal === "ascending") ? "descending" :
"ascending";
viewmodel[orderProp](orderVal);
for (prop in viewmodel) {
if (prop.indexOf("Order") !== -1 && prop !== orderProp) {
viewmodel[prop]("ascending");
}
}
}
Lastly, we can add some additional CSS to style our clickable <th> elements:
.ascending:hover:after {
content:""; display:block; border-width:7px;
border-style:solid; border-left-color:transparent;
border-right-color:transparent; border-top-color:#000;
border-bottom:none; position:absolute; margin-top:-3px;
right:15px; top:50%;
}
.descending:hover:after {
content:""; display:block; border-width:7px;
border-style:solid; border-left-color:transparent;
border-right-color:transparent; border-bottom-color:#000;
border-top:none; position:absolute; margin-top:-3px;
right:15px; top:50%;
}
Objective Complete - Mini Debriefing
We first updated our HTML with some more bindings. First we added the click binding
using the data-bind attribute on the parent <tr>. The click binding is used to add an
event handler to any HTML element.
246
Project 10
The handler function can be a ViewModel method or any regular JavaScript function.
In this example, we bound the handler to a function called sort, which will be a method
of our ViewModel.
Note also that we've added the binding to the parent <tr> and not the individual <th>
elements. We can exploit the fact that events bubble up to their parent elements to
implement a very simple and computationally cheap form of event delegation.
We also added the css binding to each of the <th> elements. The css binding is used to
add a class name to an element. So the class name that the element acquires depends on
the ViewModel property it is bound to. Each of our <th> elements is bound to a different
ViewModel property and will be used as part of our sorting.
Next we made some changes to our script file. First we added a series of new observable
properties. We added the following properties:
ff
nameOrder
ff
numberOrder
ff
symbolOrder
ff
weightOrder
ff
discoveredOrder
Each of these properties is observable, which is required to allow the <th> element's class
names to update automatically when any of the properties change. Each of the properties
are initially set to the string ascending so this is the class name that each <th>element will
be given.
Sorting the data
Next we added our sort method to the ViewModel. Because this method is part of an
event-handling binding (the click binding we added to the <tr>), the method will
automatically be passed two arguments – the ViewModel as the first, and an event object as
the second. We can make use of both of these within the function.
First we define some variables. We use jQuery to select whichever <th> element was
clicked. We can determine this using the target property of the event object, which we
wrap in jQuery so that we can call jQuery methods on the selected element.
We can then get the data-bind attribute of the element using jQuery's attr() method,
which we can then split on the space between the binding name and the property it is bound
to. So for example, if we were to click on the <th> containing Name in a browser our first
variable, orderProp, would be set to nameOrder.
247
A Sortable, Paged Table with Knockout.js
The next variable orderVal is set to the current value of the ViewModel property that the
orderProp variable points to. Knockout provides a simple way to get or set any ViewModel
property programmatically.
If we want to get the value of a property, we call the property as if it were a function,
as follows:
property();
If we want to set the property, we still call it like a function, but we pass in the value we
would like to set as an argument:
property(value);
So, continuing the example of when the <th> containing Name is clicked, the orderVal
variable would have the value ascending, because this is the default value of each of the
…Order properties. Notice how we get the correct value using the orderProp variable and
square-bracket notation.
Our last variable, comparatorProp, is convenient to store the property of the objects
within the elements array that we are going to sort by. Our ViewModel properties have
the string Order at the end of them, but the properties inside the objects in the elements
array do not. So to get the correct property we just have to split the string on the uppercase
O and take the first item from the array returned by split().
observableArray
Next we use the sort() method to perform the sort. It looks like we're using JavaScript's
regular sort() function for this, but actually, we aren't. Don't forget, the elements array
isn't just a regular array; it's an observableArray, so while we could get the underlying array
out of the element's viewModel property, and then call the regular JavaScript sort()
function on it, Knockout gives us a better way.
Knockout provides a range of standard array functions from JavaScript that we can call
on observable arrays. For the most part these work in very similar ways to their original
JavaScript counterparts, but it's almost always better to use the Knockout variants where
possible because they are better supported across browsers, especially legacy browsers, than
the original JavaScript versions. Some of the Knockout methods also give us little extra bits of
functionality or convenience too.
One example of this is with Knockout's sort() method. This isn't the reason we've
used the method here, but it is an example of how Knockout can improve the original
JavaScript function.
248
Project 10
JavaScript's built-in default sort() function does not sort numbers very well, because it
automatically converts numbers into strings and then performs the sort based on the string
instead of the number, leading to results that we do not expect.
Knockout's sort() method does not do this, and can sort arrays of strings or numbers
equally as successfully. We don't know at this point whether we'll be sorting strings,
numbers, or both because the objects in the elements array contains both strings and
numbers, sometimes in the same properties.
Like the JavaScript sort() function, a function passed to Knockout's sort() method will
automatically be passed two values, which are the current items to sort. Also like JavaScript's
sort() function, Knockout's sort() method should return 0 if the values being compared
are equal, a negative number if the first value is lower, or a positive number if the first value
is higher.
Within the function passed to sort(), we first get the values we're going to compare from
the objects. Both of the values passed to the function will be objects, but we only want to
compare a property from within each of these objects, so we store the properties we'll be
comparing in the propA and propB variables for convenience.
Comparing different value types
I mentioned a moment ago that sometimes we might be comparing values of different types.
This could occur if we are sorting by the date column, which may contain a number in the
form of a year, or it may be the string Antiquity, which some of the objects have.
So we check whether the two values being compared are of the same type using
JavaScript's typeof operator and a regular if statement. If they aren't the same type
we check whether each property is a string, and if so, convert its value to the number 0.
Within the if statement, we use JavaScript's ternary statement for conciseness.
Checking the order
We then check whether the orderProp variable we set a moment ago is set to ascending.
If it is, we perform a standard sort. We check whether the two values are equal and if so,
return 0. If the two values are not equal we can then check whether the first value is less
than the second and return -1 if it is, or 1 if it isn't. To keep this entire statement on a single
line we can use a compound ternary.
If the order is not ascending, it must be descending, so we can perform a descending
sort instead. The code for this is almost identical, except that we return 1 if the first value
is less than the second value, and -1 if not, the reverse of the statement in the first branch
of the conditional.
249
A Sortable, Paged Table with Knockout.js
We then need to update the value of the …Order property for the column we have just
sorted. This piece of code acts like a simple toggle switch – if the value is currently set
to ascending, we set it to descending. If it's set to descending, we simply set it to
ascending. The behavior that this allows is that when a <th> element is clicked for the first
time, it will perform the default ascending sort. If it is clicked a second time, it will perform a
descending sort.
Lastly we want to reset any other …Order properties of our ViewModel if they have been
changed. We use a simple JavaScript for in loop to iterate over the properties of our
ViewModel. For each property we check whether it contains the string Order, and that
it is not the property that we've just updated.
Provided both of these conditions pass, we reset the value of the current property to the
default value ascending.
Adding icons
The CSS we added is used to add a little sort icon to each <th> element when it is hovered.
We can exploit the CSS shape technique to create a down pointing arrow for ascending, and
an up pointing arrow for descending. We also use the :after CSS pseudo selector to avoid
hardcoding a non-semantic element, such as <span> or similar, to display the shape. Which
arrow is display is determined by the class name that we bound to the …Order properties of
our ViewModel.
If you've never used CSS shapes before, I'd strongly recommend that
you look into them because they are a fantastic way of creating icons
without requiring non-semantic placeholder elements, or HTTP-heavy
images. For more information, check out the CSS shapes guide at
http://css-tricks.com/examples/ShapesOfCSS/.
At this point we should be able to run the page in a browser and click on any of the headings
once to perform an ascending sort, or click twice to perform a descending sort:
250
Project 10
Setting the page size
So the sorting functionality we added is pretty awesome. But the <table> is still quite
large and unwieldy – too large in fact to be seen on the page in its entirety. So it's perfect
for paging.
One thing we need to do is determine how many items should constitute a single page of
data. We could hardcode a value into our script for the number of items to show per page,
but a better way is to add a facility to the UI so that the users can set the number of items to
display per page themselves. This is what we'll do in this task.
Engage Thrusters
We can start by adding some additional markup. Add the following elements directly after
the <tbody> element:
<tfoot>
<tr>
<tdcolspan="5">
<div id="paging" class="clearfix">
<label for="perPage">Items per page:</label>
<select id="perPage" data-bind="value: pageSize">
<option value="10">10</option>
<option value="30">30</option>
<option value="all">All</option>
</select>
</div>
</td>
</tr>
</tfoot>
We also need to make a minor change to the <tbody> element. It currently has a foreach
binding to the observed array of elements. We're going to add a new property to our
ViewModel in a moment and we need to update the binding in sortable-table.html so
that it is linked to this new property:
<tbody data-bind="foreach: elementsPaged">
Next we can add a few new ViewModel properties in sortable-table.js:
pageSize: ko.observable(10),
currentPage: ko.observable(0),
elementsPaged: ko.observableArray(),
251
A Sortable, Paged Table with Knockout.js
Lastly, we can add a special new variable known as a computed observable. This should
come after the vm variable:
vm.createPage = ko.computed(function () {
if (this.pageSize() === "all") {
this.elementsPaged(this.elements.slice(0));
} else {
var pagesize = parseInt(this.pageSize(), 10),
startIndex = pagesize * this.currentPage(),
endIndex = startIndex + pagesize;
this.elementsPaged(this.elements.slice(startIndex,
endIndex));
}
}, vm);
Objective Complete - Mini Debriefing
We started out in this task by adding a <tfoot> element containing a row and a single
cell. Inside the cell is a container for our paging elements. We then have a <label> and a
<select> element.
The <select> element contains a few options for showing different numbers of items,
including an option to see all the data. It also uses Knockout's value data-bind attribute
to link the value of the <select> element to a property on our ViewModel called
pageSize. This binding means that any time the <select> element's value changes, such
as when a user makes a selection, the ViewModel property will be updated automatically.
This binding goes both ways, so if we were to update the pageSize property
programmatically in our script, the element on the page would automatically be updated.
We then linked the <tbody>foreach binding to a new property on our VeiwModel
called elementsPaged. We'll use this new property to store a subset of the items in the
elements array. The actual items in this property will constitute a single page of data.
Next we added some new properties to the object literal stored in the vm variable, also
known as our ViewModel. These properties include currentPage, pageSize, and
elementsPaged properties that we just discussed.
The last thing we do is add a Knockout feature called a computed observable. This is an
extremely useful facility that lets us monitor one or more variables and execute code
whenever any of the observables change values.
252
Project 10
We use the ko.computed() method to set up the computed observable as a method of the
ViewModel, passing in a function as the first argument. The ViewModel is passed in as the
second argument. We're not within a method attached to our ViewModel now, so we need to
pass the ViewModel in to the computed() method in order to have this set to the ViewModel.
Within the function passed as the first argument we reference the three new ViewModel
properties we just added. Any ViewModel properties referenced within this function will be
monitored for changes and the function invoked whenever this occurs.
All the function does is check whether the pageSize() property is equal to the string all.
If it is, it simply adds all of the objects in the elements array to the elementsPaged array.
It does this by taking a slice of the elements array that starts at the very first item. When
slice() is used with a single argument it will slice to the end of the array, which is exactly
what we need to get the entire array.
If pageSize does not equal the string all, we first need to make sure it's an integer.
Because this ViewModel property is linked to the <select> element on the page,
sometimes the value might be a string of the number instead of an actual number. We can
ensure it's always a number by using the parseInt() JavaScript function on the property
and storing it in the variable pagesize for use throughout the rest of the function.
Next we need to determine what the start index passed as the first argument to slice()
should be. To work this out we just multiply the value of the pageSize property by the
currentPage property, which is initially set to 0.
We can then populate the elementsPaged array with a slice of the elements array starting
at the startIndex value we just determined and ending at the endIndex value, which will
be the startIndex plus the number of items per page.
When we run the page in our browser, the <select> box will initially have the value 10 set,
which will trigger our computed observable, selecting the first 10 items in the elements
array, and display them in the <table>.
We should find that we can use the <select> to change how many items are
displayed dynamically.
Classified Intel
In this task we used the slice() Knockout method. You may have thought that we were
using JavaScript's native Array.slice() method, but actually we were using the Knockout
version and there's an easy way to spot this.
Usually when we want to get the value inside an observable property, we invoke the
property like a function. So when we wanted to get the pageSize property of the
ViewModel, we used this.pageSize().
253
A Sortable, Paged Table with Knockout.js
When we called the slice() method however, we didn't invoke the elements property like
a function, so the actual array within the property wasn't returned. The slice() method
was called directly on the observable.
Knockout re-implements a range of native methods that can be called on arrays, including
push(), pop(), unshift(), shift(), reverse(), and sort(), which we used in the
last task.
It's recommended to use the Knockout versions of these methods rather than the native
JavaScript ones because they are supported across all browsers that Knockout supports,
and so that dependency tracking is maintained and the UI of your application is kept in sync.
Adding Previous and Next Links
At this point our page is now only displaying the first 10 items. We need to add an interface
that allows the user to navigate to other pages of data. In this task we can add Next and
Previous links so that the pages can be viewed in a linear sequence.
Engage Thrusters
We'll start out once again by adding the HTML component of this feature. Directly after the
<select> element within the <tfoot> element, add the following new markup:
<nav>
<a href="#" title="Previous page"
data-bind="click: goToPrevPage">&laquo;</a>
<a href="#" title="Next page"
data-bind="click: goToNextPage">&raquo;</a>
</nav>
Next we can add some new methods to our ViewModel. These can go directly after the sort
method that we added earlier in sortable-table.js:
totalPages: function () {
var totalPages = this.elements().length / this.pageSize() || 1;
return Math.ceil(totalPages);
},
goToNextPage: function () {
if (this.currentPage() < this.totalPages() - 1) {
this.currentPage(this.currentPage() + 1);
}
},
goToPrevPage: function () {
254
Project 10
if (this.currentPage() > 0) {
this.currentPage(this.currentPage() - 1);
}
}
Lastly, we can add a little CSS to tidy up the new elements we added in this part, as well as
the ones we added in the last part by adding the following code to sortable-table.css:
tfoot label, tfoot select, tfootnav {
margin-right:4px; float: left; line-height:24px;
}
tfoot select { margin-right:20px; }
tfootnav a {
display:inline-block; font-size:30px; line-height:20px;
text-decoration:none; color:#000;
}
Objective Complete - Mini Debriefing
We started out by adding a <nav> element containing two <a> elements to the page,
which make the Previous and Next links. We add data bindings to the links that connect
the Previous link to the goToPrevPage() method, and the Next link to the
goToNextPage() method.
We then added a small utility method, as well as these two new methods to our ViewModel.
Our methods don't have to accept parameters like the sort() method did, and we can
access our ViewModel within the methods using this.
The first method totalPages() simply returns the total number of pages by dividing the
total number of items in the elements array by the value held in the pageSize property.
Sometimes the currentPage property will equal the string all, which will return NaN
when used in a Math operation, so we can add the double-bar OR (||) to return 1 when this
is the case. We also use Math.ceil() to ensure we get a whole number, so when there are
11.8 pages of data (the default based on 10 items per page), the method will return 12. The
Ceil() function will always round up, because we can't have part of a page.
The createPage computed observable that we added in the last task actually does most of
the work for us. The next two methods simply update the currentPage property, which in
turn automatically triggers the createPage() computed observable.
In the goToNextPage() method we first check that we aren't already on the last page,
and as long as we aren't, we increase the currentPage property by one. We use the
totalPages() method when we check whether we're on the last page.
255
A Sortable, Paged Table with Knockout.js
The goToPrevPage() method is just as simple. This time we check that we aren't already
on the first page of data (if currentPage is equal to 0), and if we aren't, we decrease the
value of currentPage by 1.
The tiny bit of CSS we added simply tidies up the elements in the <tfoot> element, allowing
them to float alongside each other, and makes the new links a little bigger than they would
be by default.
Adding numerical page links
We can now add as many links as are required in order to allow the user to visit any of the
pages directly. These are the numerical page links that link to each single page directly.
Engage Thrusters
First of all we need to add a new observable property to our ViewModel, directly after the
existing observables in sortable-table.js:
pages: ko.observableArray(),
After this we can add a new method to our ViewModel. This can be added after the
goToPrevPage() method, within the vm object literal:
changePage: function (obj, e) {
var el = $(e.target),
newPage = parseInt(el.text(), 10) - 1;
vm.currentPage(newPage);
}
Don't forget to add a comma after the goToPrevPage() method! We can then add a new
computed observable, in the same way that we have previously. This can come directly after
the createPage computed observable that we added in the last task:
vm.createPages = ko.computed(function () {
var tmp = [];
for (var x = 0; x < this.totalPages(); x++) {
tmp.push({ num: x + 1 });
}
this.pages(tmp);
}, vm);
256
Project 10
Next, we need to add some new markup to the HTML page. This should be added between
the Previous and Next links that we added in the last task:
<ul id="pages" data-bind="foreach: pages">
<li>
<a href="#" data-bind="text: num,
click: $parent.changePage"></a>
</li>
</ul>
Lastly we can add a little CSS to position the new elements in sortable-table.css:
tfoot
tfoot
tfoot
tfoot
nav
nav
nav
nav
ul { margin:3px 0 0 10px; }
ul, tfootnav li { float:left; }
li { margin-right:10px; }
li a { font-size:20px; }
Objective Complete - Mini Debriefing
First of all we added a new pages observable array to our ViewModel. We didn't give it an
array to begin with; we'll add this dynamically at the appropriate time.
The computed observable we added as createPages is used to build an array where each
item in the array represents a page of data. We can get the total number of pages using our
totalPages() method as we did before.
Once this has been determined, which will be whenever the pageSize() observable
changes, we can then populate the observable array that we just added.
The objects added to the array are created using a simple for loop to create an object and
push it into an array. Once we've built an object for each page we can then set the array as
the value of the pages property.
Each object we create has just a single property, called num, the value of which is the current
value of the x counter variable used by the loop.
In the HTML page, we used the foreach data binding to iterate over the array we added
to the pages array. For each object in the array, we create an <li> element and an <a>
element. The <a> has two bindings specified using the data-bind attribute.
The first is the text binding, which sets the text of the element. In this case, we set the text
to be the value of the num property that each object has.
The second binding is a click binding, which calls a method called changePage. However,
within the foreach binding, the context is set to the current object in the pages array, so we
need to use the special $parent context property to access the method on the ViewModel.
257
A Sortable, Paged Table with Knockout.js
Lastly we added the changePage method used by the <a> elements. All we need to do in
this simple method is get the text of the element that was clicked, remove 1 from its value
because the actual page numbers are zero-based, and update the curentPage observable
property of our ViewModel. Inside this method for some reason the value of this is not set
to the element that was clicked, as we would expect from our encounters with the sort()
method that we added earlier.
Because the <a> elements that will trigger the changePage method are created within
a foreach binding, the first argument passed to changePage will be the object within
the pages array that the <a> element is associated with. Luckily we can still access the
ViewModel using the variable vm.
The CSS we added simply floats the list items alongside each other, spaces them out a little,
and sets the color and size of the text.
Classified Intel
As well as the $parent context property that allows us to access the parent object of the
ViewModel property being iterated in a foreach binding, we can also make use of $data,
which points to the array being iterated.
As well as this, there is also an $index property that allows us to access the current
iteration index, which we could have used in this example, instead of setting the num
property on each object.
Managing class names
In this task we can show feedback to the user to describe which page is currently being
viewed. We can also disable the Previous or Next links if we're on the first or last page of
data. We can do all this using a little bit more script and some simple CSS.
Engage Thrusters
First we need to add another method to our ViewModel, directly after the existing ones in
sortable-table.js:
manageClasses: function () {
var nav = $("#paging").find("nav"),
currentpage = this.currentPage();
nav.find("a.active")
.removeClass("active")
.end()
.find("a.disabled")
258
Project 10
.removeClass("disabled");
if (currentpage === 0) {
nav.children(":first-child").addClass("disabled");
} else if (currentpage === this.totalPages() - 1) {
nav.children(":last-child").addClass("disabled");
}
$("#pages").find("a")
.eq(currentpage)
.addClass("active");
}
We then need to call this method from several places in our existing code. First, we need to
call it at the end of the createPage() and createPages() computed observables, by
adding the following code to the end of the last line in each function (the line that begins
with this):
.manageClasses();
Then, in order to add the initial class names before the table is interacted with, we need to
call it after the applyBindings() method after the ViewModel:
vm.manageClasses();
Lastly, we can add the additional CSS that I mentioned in the task introduction:
tfoot nav a.disabled, tfoot nav a.disabled:hover {
opacity: .25; cursor: default; color:#aaa;
}
tfoot nav li a.active, tfoot a:hover { color:#aaa; }
Objective Complete - Mini Debriefing
The first thing we did in this task was to add a new method to our ViewModel – the
manageClasses() method. This method is responsible for adding or removing the
disabled class from the Previous and Next links, and for adding the active class to the
numbered link corresponding to the current page.
Inside the method, we first cache a selector for the containing <nav> element so that
we can access the elements we need to update as efficiently as possible. We also get the
curentPage ViewModel property as we'll be comparing its value a few times.
We then find the elements that have the disabled and active classes, and remove them.
Notice how we use jQuery's end() method after removing the active class to get back to
the original <nav> selection.
259
A Sortable, Paged Table with Knockout.js
All we need to do now is to put the classes back onto the appropriate elements. If the
currentPage is 0, we add the disabled class to the first link in the <nav> using
jQuery's :first-child selector in conjunction with the children() method.
Alternatively, if we're on the last page, we add the disabled class to the last child of the
<nav> instead, this time using the :last-child selector.
Selecting the element to apply the active class to is done easily using the jQuery eq()
method, which reduces a selection of elements down to a single element as the specified
index. We use the currentpage as the index of the element to retain in the selection.
The CSS was added purely to give the elements with the class names different styling so it
was easy to see when the classes are added and removed.
When we run the page in a browser now, we should find that the Previous link is disabled to
begin with, and the number 1 is active. If we visit any of the pages, that number will gain the
active class.
Resetting the page
Now that we've wired up our numeric paging links, a problem has become apparent.
Sometimes, when changing the number of items per page, an empty table is displayed.
We can fix this by adding another binding to the <select> element that resets the current
page whenever the <select>element's value changes.
Engage Thrusters
First of all we can add the new binding to the HTML. Change the <select> element so that
it appears as follows:
<select id="perPage" data-bind="value: pageSize, event: {
change: goToFirstPage
}">
Now we can add the goToFirstPage() method to the ViewModel:
goToFirstPage: function () {
this.currentPage(0);
}
260
Project 10
Objective Complete - Mini Debriefing
First of all we added the event binding as a second binding to the <select> element
responsible for setting the number of items per page. The format of this binding is slightly
different than other bindings we've used in this project.
After the name of the binding, event in this case, we specify the name of the event and the
handler to call when the event occurs within curly braces. The reason why this format is used
is because if we want, we are able to specify multiple events and handlers within the braces.
We then added the new event handler, goToFirstPage(), as a method of our ViewModel.
All we need to do within the handler is set the currentPage observable to 0, which
will automatically move us back to the first page of results. This will occur whenever the
<select> element's value changes.
Filtering the table
To finish off the project we can add filtering so that the different types of elements can be
shown. The data for the table contains a column we haven't used yet – the state of the
element (the actual physical element, not an HTML element!)
In this task, we can add a <select> element that allows us to filter elements by their state.
Engage Thrusters
First we need to add a new observable array to the ViewModel, which will be used to store
objects that represent the different states an element can be:
states: ko.observableArray(),
We can also add a simple non-observable property to the ViewModel:
originalElements: null,
Next we need to populate the new array. We can do this outside of our ViewModel, directly
after where we call vm.manageClasses():
var tmpArr = [],
refObj = {};
tmpArr.push({ state: "Filter by..." });
$.each(vm.elements(), function(i, item) {
var state = item.state;
261
A Sortable, Paged Table with Knockout.js
if (!refObj.hasOwnProperty(state)) {
var tmpObj = {state: state};
refObj[state] = state;
tmpArr.push(tmpObj);
}
});
vm.states(tmpArr);
Then we can add the new HTML that will create the <select> element used to filter the
<table> data:
<div class="filter clearfix">
<label for="states">Filter by:</label>
<select id="states" data-bind="foreach: states, event: {
change: filterStates
}">
<option data-bind="value: state, text: state">
</option>
</select>
</div>
Now we need to add a final method to our ViewModel that will actually filter the data when
a selection is made:
filterStates: function (obj, e) {
if (e.originalEvent.target.selectedIndex !== 0) {
var vm = this,
tmpArr = [],
state = e.originalEvent.target.value;
vm.originalElements = vm.elements();
$.each(vm.elements(), function (i, item) {
if (item.state === state) {
tmpArr.push(item);
}
});
vm.elements(tmpArr).currentPage(0);
var label = $("<span/>", {
262
Project 10
"class": "filter-label",
text: state
});
$("<a/>", {
text: "x",
href: "#",
title: "Remove this filter"
}).appendTo(label).on("click", function () {
$(this).parent().remove();
$("#states").show().prop("selectedIndex", 0);
vm.elements(vm.originalElements).currentPage(0);
});
label.insertBefore("#states").next().hide();
}
}
Lastly, we can add just a little CSS to sortable-table.css, just to tidy up the new
elements:
tfoot .filter { float:right; }
tfoot .filter label {
display:inline-block; height:0; line-height:0;
text-indent:-9999em; overflow:hidden;
}
tfoot .filter select { margin-right:0; float:right; }
tfoot .filter span {
display:block; padding:0 7px; border:1px solid #abadb3;
border-radius:3px; float:right; line-height:24px;
}
tfoot .filter span a {
display:inline-block; margin-left:4px; color:#ff0000;
text-decoration:none; font-weight:bold;
}
Objective Complete - Mini Debriefing
First we added a new observable array called states that will be used to contain the
different states of the elements that make up our data. These states are solid, liquid,
gas, or unknown.
We also added a simple property to the ViewModel called originalElements, which will
be used to store the complete collection of elements. This property is just a regular object
property because we don't need to observe its value.
263
A Sortable, Paged Table with Knockout.js
Populating the states array
Next we populate the states array with all of the unique states found in the data. We only
need to populate this array once, so it can appear outside of the ViewModel. We start out by
creating an empty array and an empty object.
We then add a single item to the array which will be used for the first <option> element
within the <select> element and will function as a label before the <select> box is
interacted with.
We can then use jQuery's each() method to iterate the elements array. For each item in
the array (which if you remember, will be an object representing a single element) we get its
state and check whether this is stored in the reference object. We can check this using the
hasOwnProperty() JavaScript function.
If the state doesn't already exist in the object, we add it. If it does already exist, we don't
need to do anything. If the object doesn't contain the state, we also push the state into the
empty array.
Once the each() loop has finished, we should then have an array that contains a single
instance of each state found in the data, so we can add this array as the value of the
states observable array.
Building the <select> box
The underlying markup for the filtering feature is quite straightforward. We added a
container <div> with a couple of class names, a <label> and a <select>. The <label>
class name is just added for accessibility, we won't display it because the first <option> of
the <select> element will function as a label.
The <select> element has several Knockout bindings. We use the foreach binding, which
is connected to the states array, so once this is populated, the <option> elements for the
<select> will be added automatically.
We also used the event binding once again to add a handler for the change event, which
will be fired whenever the <select> box is interacted with.
Inside the <select> element, we add a template for the <option> elements. Each option
will be given the text and value of the state property of the current object in the
states array.
264
Project 10
Filtering the data
We then added the method to our ViewModel responsible for filtering the data displayed in
the <table>. The first thing we do in the method is check that the first <option> has not
been selected, because this is just a label and doesn't correlate to a state.
We can determine this by looking at the selectedIndex property of the target element
(<select>), which is available in the originalEvent object. This itself is part of the event
object that is passed to our event handler automatically.
Because we're going to be changing the elements observable array (in order to trigger the
paging of the filtered elements) we want to store the original elements for later. We can
store them in the originalElements property of the ViewModel.
Next, we need to build a new array that contains only the elements that have the state that
was selected in the <select> element. To do this we can create an empty array and then
iterate over the elements array and check the state of each element. If it matches, we
push it into the new array.
We can get the state that was selected from the <select> element again using the event
object passed to our event handler. This time we use the value property of the target
element in the originalEvent object.
Once the new array has been populated, we update the elements array so that it contains
just the new array we have just created, and then set the currentPage to 0.
The filters that we've added are mutually exclusive, so only one filter can be applied at
any one time. Once a filter has been selected, we want to hide the <select> box so that
another filter cannot be selected.
We can also create a label that shows which filter is currently being applied. This label is
made from a <span> element which shows the text of the filter, and also contains an <a>
element that can be used to remove the filter and return the <table> back to its initial state
of showing all of the elements.
We can use jQuery's on() method to attach the handler for the <a> element as soon as
it is created and appended to the page. Within the handler, we simply set the elements
property of the ViewModel back to the array saved in the originalEvents property and
again reset the <table> back to the first page by setting the currentPage property to 0.
265
A Sortable, Paged Table with Knockout.js
We should now find that we can select one of the options in the <select> box, see just the
filtered data and the filter label, then click on the red cross sign in the filter label to go back
to the initial <table>. A filtered selection of the data and the filter label is shown in the
following screenshot:
Mission Accomplished
Our application runs mostly on Knockout functionality, which allows us to easily populate
dynamic elements with content, add event handlers, and generally manage the state of the
application. We use jQuery too, mostly in a DOM selection capacity, and also occasionally when
we wish to use a utility, such as the $.each() method that we leveraged several times.
It would have been equally as possible to build this application purely using jQuery and
without using Knockout at all; however, jQuery itself was never designed nor intended to be
the complete solution to building complex dynamic applications.
What we generally find when we try to build complex dynamic applications using jQuery
alone, is that our script very quickly becomes a bloated mess of event handlers that is
neither easy to read, or maintain, or update at a future point.
Using Knockout to handle maintaining the state of an application, and using jQuery to
fulfill the role it was intended for, gives us the ideal toolkit for building highly dynamic,
data-driven, complex applications using very little code.
266
Project 10
Throughout this example, I've tried to keep individual methods as simple as possible and
have them do one thing well and one thing only. Keeping individual units of functionality
isolated in this way helps to keep our code maintainable because it is easy to see what each
existing function does, and easy to add new features without breaking what already exists.
You Ready To Go Gung HO?
A Hotshot Challenge
Knockout makes it easy to build a <table> from an array of data, and because the data
is dynamic, it's easy to edit it or add new items to it, and have the data in our application
updated. Although the data is stored locally in a file in this example, it would be trivial to
store the data on the server and populate our elements array at page load using a simple
AJAX function.
This would be the first thing to do if you wanted to take this example further. Once this
has been done, why not see if you can make the table cells editable so that their values
can be changed, or add a feature that allows you to insert new rows into the <table>.
Once you've done this, you'll want to post the new data back to the server so that it
can be stored permanently.
267
Index
Symbols
$PATH variable 178
<a> element 100, 125, 133
<article> element 46, 125, 197
<aside> element 46, 49
<body> element 14, 121, 197
<button> element 88
<div> element 19, 88, 161, 197
<div> elements 137
<figure> element 22, 28
<form> element 141
<h1> element 151
<head> element 86
<header> element 125, 151
<iframe> element 151, 159
<img> element 14
<input> element 88, 125
<label> element 88
<link> element 59, 86
<nav> element 45
<script> element 14, 44, 86, 99, 121
<section> element 45, 197
<select> box
building 264
<select> element 252
<small> element 88
<span> element 33
<table> element 132, 137, 243
<td> element 133
<th> elements 243
<ul> element 151, 197
A
addEventListener() function 159
addListener() method 66, 68, 69, 73
addMarker() function 69, 72
AJAX 91
ajax() method 94, 141
ajaxSetup() method 214, 224
always() method 96
animate() method 51
animation property 68
append() method 141
appendTo() method 24
applyBindings() method 243, 259
apply() method 144
arc() method 228
aspectH variable 17
aspectW variable 17
attr() method 19, 52, 133, 247
B
back button
restoring, for browser 53
backgroundImage property 19
backgroundPosition property 19
beginPath() method 228
bestTime variable 37
bind() method 22
bounties
getting, for jQuery Mobile Single-page App
project 96, 97
browser
back button, restoring for 53
browser action 150
browser window
resizes, handling 49
C
callback() function 95
cd command 182
CDN 12
chaining 123
change event 127
changePage() method 97, 98, 114
children() method 35, 164
Chrome
about 147
extension structure, setting up 149-152
URL, for downloading 149
chrome-extension directory 149, 160
Chrome extension project
about 147
content script, adding 160, 161
extension, installing 152, 153
manifest, adding 152, 153
mechanism, adding for saving
microdata 165-168
message, posting to sandbox 158-160
objectives 149
page, scraping for microdata 162, 165
sandboxed JsRender template, adding 154-157
className property 136
class names
managing 258-260
clearInterval() function 32, 35
clearRect() function 228
click data
requesting 223-226
saving 218-220
click event 52
clicks
capturing, on map 67-69
clickStats object 214
clone() method 19, 231
closest() method 137
code wrapper
creating, for puzzle game 15-17
colspan attribute 133
command-line interface (CLI) 178
command property 159
communication module
about 168
adding 168
company HQ
displaying, with custom overlay 64-66
270
complete() method 96
computed() method 253
contactMethods object 165
containment option 28
content.css file 160
content delivery network. See CDN
content.js file 160
content script
about 160
updating 169
Content Security Policy (CSP) 154
contentWindow property 159
converters option 224
createElement() function 224
cross-site scripting (XSS) 154
CSS 147
CSSLint
about 61
URL 61
css() method 19, 50
currentTime object 33
currentTime variable 37
custom jQuery
building 184, 185
custom overlay
used, for displaying company HQ 64-66
D
data
sorting 247
data-bind attribute 243
Date() constructor 105
delegate() method 22
demo page
creating, for sidebar project 43-45
disabled property 72, 75
document.ready function 15
DOM 91
done() method 96, 144
dragend event 72
draggable() method 28
draggable property 69
dragover event 127, 128
drop event 127
Dust.js 135
E
each() method 24, 35, 132, 133, 162, 214
element
initial position, storing 46, 47
enableButton() function 75
eq() method 164
equal() method 188
error() method 96
event handlers
adding, for file upload 126-128
extend() method 122
F
Facebook 191
fail() method 96
Fat Cow
URL 134
fileList property 141
files
displaying 129-131
removing, from upload list 135-137
uploading 139-142
find() method 132
Fisher-Yates shuffle
about 22
URL, for info 23
for loop 22
format() function 79
formatTime() function 203
FormData object 141
full() method 229
function expression 69
G
games 10
geocode() method 71, 74
getBounties() function 93, 96, 97, 116
getContext() function 228
getData() function 200, 206
getDay() function 105
getDistanceMatrix() method 78
getElementById() function 63
getFullYear() function 105
getItem() method 168
getJSON() method 93, 200
getMonth() function 105
getPosition() function 29
getUser() function 200
Git
about 174
installing 175, 176
Github repository 174
Google Code
URL 43
google-map.css file 59, 60
google-map.js file 59
Google Maps 58
Google Maps API 59
google.maps namespace 62
goToNextPage() method 255
goToPrevPage() method 255
grid option 28
grunt command 183
Grunt.js
about 174
installing 180
URL 174
H
handleFiles() method 128, 131, 138, 139
handleProgress() method 139, 142
hasData variable 165
hashchange event 53
hash fragment
handling, on page load 54, 55
hasOwnProperty() function 165
hasProps() function 165
heat map
about 209
displaying 226-229
displaying, for layout 233-237
features 210
heat-map-client.js file 212
heat map project
checklist 211, 212
click data, requesting 223-226
click data, saving 218-220
environment, determining 212-216
environment, saving 212-216
layouts, selecting 229-232
management console, adding 220-222
271
objectives 211
visitor clicks, capturing 216-218
helpers() method 105
hidePageLoadingMsg() method 116
History API 198
hoisting 17
hours property 37
HTML 147
html5shiv script 43
HTML Ipsum
URL 46
html() method 134
HTML page
creating, for puzzle game 12-14
I
if statement 36
image
splitting, into pieces 18-20
imagesLoaded() method 217
imagesLoaded plugin
about 193
URL, for downloading 193
imgContainer variable 24
infinite scrolling system
about 191
checklist 193
features 192, 193
initial feeds, getting 198-201
initial set of results, displaying 201-204
objectives 193
underlying page, creating 193-198
InfoWindow() constructor 66
init() method 124, 125
initProgress() method 137-139
installation, Git 175, 176
installation, Grunt.js 180
installation, Make 175, 176
installation, Node.js 178, 179
install command 182
interactive Google map project
about 57
checklist 59
clicks, capturing on map 67-69
company HQ, displaying with custom
overlay 64-66
272
factoring, in weights 74, 75
features 58
interface, creating 60, 61
map, initializing 62
marker repositions, handling 72, 73
objectives 58
page, creating 60, 61
projected cost, displaying for journey 76-80
projected distance, displaying between markers
76-80
UI, updating with start and end locations 70-72
interface
creating, for interactive Google map
project 60, 61
itemprop attribute 164
itemscope attribute 165
item view
building, for jQuery Mobile Single-page App
project 108-114
J
JavaScript 147
JavaScript Object Notation. See JSON
join() method 20
jQuery
about 11
environment, configuring 181-184
jQuery 1.5 96
jquery directory 182
jQuery File Uploader project
about 117
checklist 119
event handlers, adding for receiving files to
upload 126-128
features 117, 118
files, removing from upload list 135-137
jQuery UI progress indicator, adding 137-139
list, displaying of selected files 129-135
objectives 118
page, creating 119-123
plugin wrapper, creating 119-123
success, reporting 142-144
underlying markup, generating 123-126
jquery-hotshots folder 11
jQuery Mobile 83
jQuery Mobile Single-page App project
about 83
bounties, getting for 96, 97
checklist 85, 86
features 84
item view, building 108-114
jsRender template, adding 98, 99, 100
list view, building 100-107
objectives 85
paging, handling 115, 116
script wrapper, creating 92-95
second page, adding 89, 91
selected files, uploading 139-142
welcome page, building 86-89
jQuery repository
cloning 176
jQuery UI
about 11, 26
URL 12
jQuery UI progress indicator
adding 137-139
jqXHR object 96, 141
JSON
about 94, 193
URL 193
JSON-C
about 200
URL, for documentation 200
JSON with padding (JSONP) 94
jsRender
about 83, 135
URL, for downloading 150
jsRender template
adding, for jQuery Mobile Single-page App
project 98-100
K
Knockout.js 239
L
LatLng() method 63
layout
heat maps, displaying for 233-237
list
displaying, of selected files 129-135
listHeight variable 204
list view
building, for jQuery Mobile Single-page App
project 100-107
listview() method 106
live() method 22
loadingMessageTextVisible property 95
localStorage 168
localStorage API 9
locationAdd event 70
Lorem Ipsum placeholder text 46
M
Make
about 174
installing 175, 176
manageClasses() method 259
management console
adding 220-222
map
clicks, capturing on 67-69
initializing 62
mapClick variable 69
Map() constructor 63
Marker() constructor 65
markerDrag() function 73
marker repositions
handling 72, 73
Math.random() function 23
Math.round() method 48
mechanism
adding, for saving microdata 165-168
media queries 212, 216
message
posting, to sandbox 158-160
microdata
about 147
features 148
page, scraping for 162-165
MINGW32 175
minutes property 37
Model-View-ViewModel (MVVM) 239
mousedown event 30
mouseup event 30
MS SQL database 237
MS SQL express 237
msysgit 175
273
N
noConflict() method 121
Node.js
about 178
installing 178, 179
URL 178
Node Package Manager (NPM) 174
not() method 24
O
observableArray() method 243, 248
offset() method 47, 49
one() method 48, 50, 75
on() method
about 22, 75, 78, 206, 217, 219
used, for adding event handler to button 22
open() method 66
optText() method 231, 232
orderProp variable 248, 249
originalEvent object 128, 136
P
page
creating, for interactive Google map
project 60, 61
scraping, for microdata 162-165
pageid property 113
pageinit event 97
page load
hash fragment, handling on 54, 55
page parameter 94
page scrolling
detecting 48
pageshow event 97
pagesize parameter 94
paged table project
about 239
checklist 241
class names, managing 258-260
features 240
initial table, rendering 241-244
numerical page links, adding 256-258
objectives 240
page, resetting 260, 261
274
page size, setting 251-253
Previous and Next Links, adding
for page 254, 255
table, filtering 261-263
table, sorting 244-247
paging
handling 115, 116
parseFloat() function 79
parseInt() function 17, 52
parseJSON() method 225
placeholder attribute 222
Placehold It
URL 46
plugin wrapper
creating 119-123
position*fixed CSS style 41
postMessage() function 159, 168
preventDefault() method 52, 128
projected cost
displaying, for journey 76-80
projected distance
displaying, between markers 76-80
promise object 143
prop() method 72, 125
push() method 19, 143
pushState() method 198
puzzle game
about 9
checklist 11
code wrapper, creating 15-17
congratulatory message, displaying 34, 35
empty space, positioning 24
event handler, adding to button 22
features 10, 11
final styling, adding 36-38
HTML page, creating for 12-14
image, splitting into pieces 18-20
objectives 11
order, verifying for puzzle pieces 34, 35
pieces, making draggable 26-30
pieces, positioning 24
puzzle pieces, shuffling 20-22
timer, starting 31-33
timer, stopping 31-35
variables, defining 15-17
puzzle pieces
making draggable 26-30
positioning 24
shuffling 20-23
Q
QUnit
about 185
unit tests, running with 185-188
R
ranking algorithms 147
removeAll() method 137
removeListener() method 69
remove() method 22, 76
renderer() function 204
render() method 105, 113
renderOuter argument 204
replace() function 132
reserved word 132
resizes
handling, for browser window 49
S
sandbox
message, posting to 158-160
sandboxed JsRender template
adding 154-157
script wrapper
creating, for jQuery Mobile Single-page App
project 92-95
scroll event 48
scrolling
automating 50, 52
handling, to bottom of page 204-207
scrollPage() function 51, 53, 55
scrollTop() method 51
scrollTop property 51, 206
second page
creating, for jQuery Mobile Single-page App
project 89, 91
seconds property 37
setClasses() utility 116
setInterval() method 32
setTimeout() function 75
shadow DOM 126
showPageLoadingMsg() method 95
sidebar project
about 41
back button, restoring for browser 53
browser window resizes, handling 49
demo page, creating 43-45
features 42
hash fragment, handling on page load 54, 55
initial position, storing of element 46, 47
objectives 43
page scrolling, detecting 48
scrolling, automating 50, 52
slice() method 254
sliding-puzzle.html file 13
sort method 247
sort() method 215
splice() function 137
split() function 17, 52, 113, 133
src attribute 151
Stack Exchange API
about 94
URL 94
states array
populating 264
stopPropagation() method 128
strings property 125
substring() function 197, 203
success() method 95
T
table
filtering 261-263
rendering 241-244
sorting 244-247
table-data.js file 243
tableHeadings property 132
tagged parameter 94
TEDTalks channel
about 192
URL, for viewing 193
template.html file 13
templating engine 135
test() method 188
text() method 33
Themeroller 12
title property 68
tmpl 83
275
totalPages() method 255, 257
totalSeconds variable 37
trigger() method 29, 70, 233
truncate() function 203
Truncate() function 197
Twitter 191
U
unit tests
running, with QUnit 185-188
unshift() method 22
updateTime() function 33
uploadFiles() method 140, 141
upload list
files, removing from 135-137
up() method 121
url property 94
userdata property 200
V
val() method 79
variables
defining, for puzzle game 15-17
videodata property 200
visitor clicks
capturing 216-218
Visual Web Developer Express (VWDE)
about 86
276
URL, for installing 86
W
WAMP 186
welcome page
creating, for jQuery Mobile Single-page App
project 86-89
when() method 144, 200, 205
while loop 22
width() method 47
Windows Apache Mysql PHP. See WAMP
winHeight variable 206
X
XAMPP 86
XMLHttpRequest (XHR) level 2 specification 141
Y
yepnope
URL 12
YouTube 192
Z
Zepto
about 173
URL 173
Thank you for buying
jQuery Hotshot
About Packt Publishing
Packt, pronounced 'packed', published its first book "Mastering phpMyAdmin for Effective MySQL
Management" in April 2004 and subsequently continued to specialize in publishing highly focused
books on specific technologies and solutions.
Our books and publications share the experiences of your fellow IT professionals in adapting and
customizing today's systems, applications, and frameworks. Our solution based books give you the
knowledge and power to customize the software and technologies you're using to get the job done.
Packt books are more specific and less general than the IT books you have seen in the past. Our
unique business model allows us to bring you more focused information, giving you more of what
you need to know, and less of what you don't.
Packt is a modern, yet unique publishing company, which focuses on producing quality,
cutting-edge books for communities of developers, administrators, and newbies alike.
For more information, please visit our website: www.packtpub.com.
About Packt Open Source
In 2010, Packt launched two new brands, Packt Open Source and Packt Enterprise, in order
to continue its focus on specialization. This book is part of the Packt Open Source brand, home
to books published on software built around Open Source licences, and offering information to
anybody from advanced developers to budding web designers. The Open Source brand also runs
Packt's Open Source Royalty Scheme, by which Packt gives a royalty to each Open Source project
about whose software a book is sold.
Writing for Packt
We welcome all inquiries from people who are interested in authoring. Book proposals should
be sent to [email protected] If your book idea is still at an early stage and you would like
to discuss it first before writing a formal book proposal, contact us; one of our commissioning
editors will get in touch with you.
We're not just looking for published authors; if you have strong technical skills but no writing
experience, our experienced editors can help you develop a writing career, or simply get some
additional reward for your expertise.
jQuery UI 1.8: The User
Interface Library for jQuery
ISBN: 978-1-84951-652-5
Paperback: 424 pages
Build highly interactive web applications with
ready-to-use widgets from the jQuery User
Interface Library
1. Packed with examples and clear explanations
of how to easily design elegant and powerful
front-end interfaces for your web applications
2. A section covering the widget factory including an
in-depth example on how to build a custom jQuery
UI widget
3. Updated code with significant changes and fixes to
the previous edition
jQuery Mobile Web
Development Essentials
ISBN: 978-1-84951-726-3
Paperback: 246 pages
Learn to use the touch-optimized, cross-device,
cross-platform, jQM web framework for smartphones
and tables
1. Create websites that work beautifully on a wide
range of mobile devices with jQuery mobile
2. Learn to prepare your jQuery mobile project by
learning through three sample applications
3. Packed with easy to follow examples and clear
explanations of how to easily build mobileoptimized websites
Please check www.PacktPub.com for information on our titles
Responsive Web Design with
HTML5 and CSS3
ISBN: 978-1-84969-318-9
Paperback: 324 pages
Learn responsive design using HTML5 and CSS3
to adapt websites to any browser or screen size
1. Everything needed to code websites in HTML5
and CSS3 that are responsive to every device
or screen size
2. Learn the main new features of HTML5 and
use CSS3’s stunning new capabilities including
animations, transitions and transformations
3. Real world examples show how to progressively
enhance a responsive design while providing fall
backs for older browsers
Git: Version Control for
Everyone
ISBN: 978-1-84951-752-2
Paperback: 180 pages
The non-coder's guide to everyday version control for
increased efficiency and productivity
1. A complete beginner's workflow for version control
of common documents and content
2. Examples used are from non-techie, day to day
computing activities we all engage in
3. Learn through multiple modes – readers learn
theory to understand the concept and reinforce it
by practical tutorials.
4. Ideal for users on Windows, Linux, and Mac OS X
Please check www.PacktPub.com for information on our titles

Similar documents

×

Report this document