Client-side programming with
JavaScript
scripts vs. programs
JavaScript vs. JScript vs. VBScript
common tasks for client-side scripts
JavaScript
data types & expressions
control statements
functions & libraries
strings & arrays
Date, document, navigator, user-
defined classes
HTML is good for developing static pages
can specify text/image layout, presentation, links, …
Web page looks the same each time it is accessed
Client-side programming
programs are written in a separate programming (or scripting) language
e.g., JavaScript, JScript, VBScript
programs are embedded in the HTML of a Web page, with (HTML) tags
to identify the program component
e.g., <script type="text/javascript"> … </script>
the browser executes the program as it loads the page, integrating
the dynamic output of the program with the static content of HTML
could also allow the user (client) to input information and process it,
might be used to validate input before it’s submitted to a remote
server
A scripting language is a simple, interpreted programming language
scripts are embedded as plain text, interpreted by application
simpler execution model: don't need compiler or development environment
saves bandwidth: source code is downloaded, not compiled executable
platform-independence: code interpreted by any script-enabledbrowser
but: slower than compiled code, not as powerful/full-featured
JavaScript: the first Web scripting language, developed by Netscape in
1995 syntactic similarities to Java/C++, but simpler, more flexible in
some respects, limited in others (loose typing, dynamic variables, simple
objects)
JScript: Microsoft version of JavaScript, introduced in 1996
• same core language, but some browser-specific differences
• fortunately, IE, Netscape, Firefox, etc. can (mostly) handle both
VBScript: client-side scripting version of Microsoft Visual Basic
adding dynamic features to Web pages
validation of form data (probably the most commonly used
application)
image rollovers
time-sensitive or random page elements
handling cookies
defining programs with Web interfaces
utilize buttons, text boxes, clickable images, prompts, etc
limitations of client-side scripting
since script code is embedded in the page, it is viewable to
the world
for security reasons, scripts are limited in what they can do
e.g., can't access the client's hard drive
since they are designed to run on any machine platform,
scripts do not contain platform specific commands
JavaScript code can be embedded in a Web page using <script> tags
the output of JavaScript code is displayed as if directly entered in HTML
<html>
<!–- CS443 js01.html 16.08.06 -->
document.write displays text in the
page.
<head>
<title>JavaScript Page</title> - text to be displayed can include HTML
</head> tags the tags are interpreted by the
browser when the text is displayed as in
<body>
<script type="text/javascript"> C++/Java, statements end with ; but a
// silly code to demonstrate output line break might also be interpreted as
document.write("<p>Hello
the end of a statement (depends upon
world!</p>"); browser)
document.write(" <p>How are <br/> "
+" <i>you</i>?</p> ");
</script> JavaScript comments similar to
C++/Java
<p>Here is some static text as
well.</p> // starts a single line
comment
</body> /*…*/ enclose multi-line
</html>
comments
JavaScript has only three primitive data types
String : "foo" 'how do you do?' "I said 'hi'." ""
Number: 12 3.14159 1.5E6
Boolean : false true *Find info on Null, Undefined
<html> assignments are as in C++/Java
<!–- CS443 js02.html 16.08.06 -->
message = "howdy";
<head> pi = 3.14159;
<title>Data Types and
Variables</title> variable names are sequences of letters,
</head> digits, and underscores that start with a
<body>
letter or an underscore variables names
<script type="text/javascript">
var x, y; are case-sensitive
x= 1024;
y=x; x = "foobar"; you don't have to declare variables, will
document.write("<p>x = " + y + be created the first time used, but it’s
"</p>"); better if you use var statements
document.write("<p>x = " + x +
"</p>");
</script> var message, pi=3.14159;
</body>
</html> variables are loosely typed, can be
assigned different types of values
(Danger!)
JAVASCRIPT OPERATORS & CONTROL STATEMENTS
<html>
standard C++/Java operators &
<!–- CS443 js03.html 08.10.10 --> control statements are
<head>
provided in JavaScript
<title>Folding Puzzle</title> • +, -, *, /, %, ++, --, …
</head> • ==, !=, <, >, <=, >=
• &&, ||, !,===,!==
<body>
<script type="text/javascript"> • if , if-else, switch
var distanceToSun = 93.3e6*5280*12;
var thickness = .002; • while, for, do-while, …
var foldCount = 0;
while (thickness < distanceToSun) { PUZZLE: Suppose you took a
thickness *= 2; piece of paper and folded it in
foldCount++;
} half, then in half again, and so
document.write("Number of folds = " + on.
foldCount);
</script>
</body> How many folds before the
</html> thickness of the paper reaches
from the earth to the sun?
*Lots of information is available
online
JAVASCRIPT MATH ROUTINES
<html>
<!–- CS443 js04.html 08.10.10 -->
the built-in Math object
<head> contains functions and
<title>Random Dice Rolls</title>
</head> constants
<body> Math.sqrt
<div style="text-align:center"> Math.pow
<script type="text/javascript"> Math.abs
var roll1 = Math.floor(Math.random()*6) + 1; Math.max
var roll2 = Math.floor(Math.random()*6) + 1;
document.write("<img
Math.min
src='http://www.csc.liv.ac.uk/"+ Math.floor
"~martin/teaching/CS443/Images/die" + Math.ceil
roll1 + ".gif‘ alt=‘dice showing ‘ + Math.round
roll1 />");
document.write(" ");
document.write("<img Math.PI
src='http://www.csc.liv.ac.uk/"+ Math.E
"~martin/teaching/CS443/Images/die" +
roll2 + ".gif‘ alt=‘dice showing ‘ +
roll2 />"); Math.randomfunction returns
</script> a real number in [0..1)
</div>
</body>
</html>
<html>
<!-- CS443 js05.html 08.10.10 --> crude user interaction can
<head>
<title>Interactive page</title>
take place using prompt
</head>
<body>
<script type="text/javascript">
1st argument: the prompt
var userName = prompt("What is your name?", message that appears in the
""); dialog box
var userAge = prompt("Your age?", "");
var userAge = parseFloat(userAge);
2nd argument: a default value
document.write("Hello " + userName + ".") that will appear in the box (in
if (userAge < 18) { case the user enters nothing)
document.write(" Do your parents know "
+ "you are online?");
the function returns the value
} entered by the user in the
else { dialog box (a string)
document.write(" Welcome friend!");
}
if value is a number, must
</script> use parseFloat (or
<p>The rest of the page...</p> parseInt) to convert
</body>
</html> forms will provide a better
interface for interaction (later)
function definitions are similar to C++/Java, except:
no return type for the function (since variables are loosely typed)
no variable typing for parameters (since variables are loosely typed)
by-value parameter passing only (parameter gets copy of argument)
function isPrime(n)
// Assumes: n > 0 Can limit variable scope
// Returns: true if n is prime, else false to the function.
{
if (n < 2) { if the first use of a variable
return false;
is preceded with var, then
}
else if (n == 2) {
that variable is local to the
return true; function
}
else {
for (var i = 2; i <= Math.sqrt(n); i++) {
if (n % i == 0) {
for modularity, should
return false; make all variables in
} a function local
}
return true;
}
}
<html>
<!–- CS443 js06.html 16.08.2006 --> Function
<head> definitions
<title>Prime Tester</title>
<script type="text/javascript"> (usually) go in
function isPrime(n) the
// Assumes: n > 0
// Returns: true if n is prime <head>
{ section
// CODE AS SHOWN ON PREVIOUS SLIDE
} <head> section is
</script> loaded first, so then
</head>
<body>
the function is
<script type="text/javascript"> defined before code in
testNum = parseFloat(prompt("Enter a positive integer", the
"7"));
if (isPrime(testNum)) { <body> is executed
document.write(testNum + " <b>is</b> a prime number."); (and, therefore, the
} function can
else {
document.write(testNum + " <b>is not</b> a prime be used later in the
number."); body of the HTML
} document)
</script>
</body>
</html>
<html>
<!–- CS443 js07.html 11.10.2011 --> recall the dynamic dice
<head>
<title> Random Dice Rolls Revisited</title> page could define a
<script type="text/javascript"> function for generating
function randomInt(low, high)
// Assumes: low <= high random numbers in a
// Returns: random integer in range [low..high] range, then use
{
return Math.floor(Math.random()*(high-low+1)) + low;
whenever needed
} easier to remember,
</script> promotes reuse
</head>
<body>
<div style="text-align: center">
<script type="text/javascript">
roll1 = randomInt(1, 6);
roll2 = randomInt(1, 6);
document.write("<img src='http://www.csc.liv.ac.uk/"+
"~martin/teaching/CS443/Images/die" +
roll1 + ".gif'/>");
document.write(" ");
document.write("<img src='http://www.csc.liv.ac.uk/"+
"~martin/teaching/CS443/Images/die" +
roll2 + ".gif'/>");
</script>
</div>
</body>
</html>
better still: if you define functions that may be useful
to many pages, store in a separate library file and
load the library when needed load a library using the
SRC attribute in the SCRIPT tag (put nothing between
the beginning and ending tags)
<script type="text/javascript"
src="random.js">
</script>
<html>
<!–- CS443 js08.html 11.10.2011 -->
<head>
<title> Random Dice Rolls Revisited</title>
<script type="text/javascript"
src="random.js">
</script>
</head>
<body>
<div style="text-align: center">
<script type="text/javascript">
roll1 = randomInt(1, 6);
roll2 = randomInt(1, 6);
document.write("<img src='http://www.csc.liv.ac.uk/"+
"~martin/teaching/CS443/Images/die" +
roll1 + ".gif'/>");
document.write(" ");
document.write("<img src='http://www.csc.liv.ac.uk/"+
"~martin/teaching/CS443/Images/die" +
roll2 + ".gif'/>");
</script>
</div>
</body>
</html>
an object defines a new type (formally, Abstract Data Type)
encapsulates data (properties) and operations on that data (methods)
a String object encapsulates a sequence of characters, enclosed in quotes
properties include
• length : stores the number of characters in the string
methods include
• charAt(index): returns the character stored at the given
index (as in C++/Java, indices start at 0)
• substring(start, end) : returns the part of the string
between the start (inclusive) and end (exclusive) indices
• toUpperCase() : returns copy of string with
uppercase letters
• toLowerCase()
lowercase : returns copy of string with
letters
to create a string, assign using new or (in this case) just make a direct
assignment (new is implicit)
word = new String("foo"); word = "foo";
properties/methods are called exactly as in C++/Java
• word.length word.charAt(0)
function strip(str)
// Assumes: str is a string suppose we want to
// Returns: str with all but letters removed test whether a word or
{
var copy = ""; phrase is a palindrome
for (var i = 0; i < str.length; i++) {
if ((str.charAt(i) >= "A" && str.charAt(i) <= "Z")
|| noon Radar
(str.charAt(i) >= "a" && str.charAt(i) <= "z")) Madam, I'm
{
Adam.
copy += str.charAt(i);
} A man, a plan, a
} canal: Panama!
return copy;
}
function isPalindrome(str) must strip non-letters out of the
// Assumes: str is a string word or phrase
// Returns: true if str is a palindrome, else false
{
str = strim(str.toUpperCase()); make all chars uppercase in
order to be case-insensitive
for(var i = 0; i < Math.floor(str.length/2); i++) {
if (str.charAt(i) != str.charAt(str.length-i-1)) {
return false; finally, traverse and
} compare chars from each
} end
return true;
}
<html>
<!–- CS443 js09.html 11.10.2011 -->
<head>
<title>Palindrome Checker</title>
<script type="text/javascript">
function strip(str)
{
// CODE AS SHOWN ON PREVIOUS SLIDE
}
function isPalindrome(str)
{
// CODE AS SHOWN ON PREVIOUS SLIDE
}
</script>
</head>
<body>
<script type="text/javascript">
text = prompt("Enter a word or phrase", "Madam, I'm Adam");
if (isPalindrome(text)) {
document.write("'" + text + "' <b>is</b> a palindrome.");
}
else {
document.write("'" + text + "' <b>is not</b> a
palindrome.");
}
</script>
</body>
</html>
arrays store a sequence of items, accessible via an index
since JavaScript is loosely typed, elements do not have to be the same
type
to create an array, allocate space using new (or can assign directly)
items = new Array(10); // allocates space for 10 items
items = new Array(); // if no size given, will adjust dynamically
items = [0,0,0,0,0,0,0,0,0,0]; // can assign size & values []
to access an array element, use [] (as in C++/Java)
for (i = 0; i < 10; i++) {
items[i] = 0; // stores 0 at each index
}
the length property stores the number of items in the
array
for (i = 0; i < items.length; i++) {
document.write(items[i] + "<br>"); // displays elements
}
<html>
<!–- CS443 js10.html 11.10.2011 -->
<head>
<title>Dice Statistics</title> suppose we want to
<script type="text/javascript"
src="http://www.csc.liv.ac.uk/~martin/teaching/CS443/JS/rand
simulate dice rolls and
om.js"> verify even
</script>
</head> distribution
<body>
<script type="text/javascript">
numRolls = 60000;
diceSides = 6;
rolls = new Array(dieSides+1);
for (i = 1; i < rolls.length; i++) {
rolls[i] = 0;
}
for(i = 1; i <= numRolls; i++) {
rolls[randomInt(1, dieSides)]++;
}
for (i = 1; i < rolls.length; i++) {
document.write("Number of " + i + "'s = " +
rolls[i] + "<br />");
}
</script>
</body>
</html> view page
• Arrays have predefined methods that allow them to be
used as stacks, queues, or other common programming
data structures.
var stack = new Array();
stack.push("blue");
stack.push(12); // stack is now the array ["blue", 12]
stack.push("green"); // stack = ["blue", 12, "green"]
var item = stack.pop(); // item is now equal to "green"
var q = [1,2,3,4,5,6,7,8,9,10];
item = q.shift(); // item is now equal to 1, remaining
// elements of q move down one position
// in the array, e.g. q[0] equals 2
q.unshift(125); // q is now the array [125,2,3,4,5,6,7,8,9,10]
q.push(244); // q = [125,2,3,4,5,6,7,8,9,10,244]
String & Array are the most commonly used objects in JavaScript
other, special purpose objects also exist
the Date object can be used to access the date and time
to create a Date object, use new & supply year/month/day/… as
desired
today = new Date(); // sets to current date & time
newYear = new Date(2002,0,1); //sets to Jan 1, 2002 12:00AM
methods include:
newYear.getFullYear() can access individual components of a date
newYear.getMonth() number (0, 11)
newYear.getDay() number (1, 31)
newYear.getHours() number (0, 23)
newYear.getMinutes() number (0, 59)
newYear.getSeconds() number (0, 59)
newYear.getMilliseconds() number (0, 999)
DATE EXAMPLE
<html>
<!–- CS443 js11.html 16.08.2006 -->
<head>
<title>Time page</title>
</head>
<body> by default, a date will be
Time when page was loaded:
<script type="text/javascript">
displayed in full, e.g.,
now = new Date(); Sun Feb 03 22:55:20 GMT-0600
document.write("<p>" + now + "</p>"); (Central Standard Time) 2002
time = "AM";
hours = now.getHours();
if (hours > 12) {
hours -= 12; can pull out portions of the date
time = "PM" using the methods and display
}
else if (hours == 0) {
as desired
hours = 12;
}
here, determine if "AM" or "PM"
document.write("<p>" + hours + ":" + and adjust so hour between 1-12
now.getMinutes() + ":" + 10:55:20 PM
now.getSeconds() + " " +
time + "</p>");
</script>
</body>
</html>
ANOTHER EXAMPLE
<html>
<!–- CS443 js12.html 12.10.2012 -->
<head>
<title>Time page</title>
</head> you can add and subtract
<body> Dates:
<p>Elapsed time in this year:
<script type="text/javascript">
the result is a number of
now = new Date(); milliseconds
newYear = new Date(2012,0,1);
secs = Math.round((now-newYear)/1000); here, determine the
days = Math.floor(secs / 86400); number of seconds since
secs -= days*86400;
hours = Math.floor(secs / 3600); New Year's day (note:
secs -= hours*3600; January is month 0)
minutes = Math.floor(secs / 60);
secs -= minutes*60 divide into number of days,
document.write(days + " days, " + hours, minutes and seconds
hours + " hours, " +
minutes + " minutes, and " +
secs + " seconds.");
</script>
</p>
</body>
</html>
Internet Explorer, Firefox, Opera, etc. allow you to access
information about an HTML document using the document object
<html>
<!–- CS443 js13.html 2.10.2012 -->
<head>
document.write(…)
<title>Documentation page</title> method that displays text in
</head> the page
<body>
<table width="100%">
<tr>
document.URL
<td><i> property that gives the
<script type="text/javascript">
document.write(document.URL); location of the HTML
</script> document
</i></td>
<td style="text-align: right;"><i> document.lastModified
<script type="text/javascript">
document.write(document.lastModified); property that gives the date
</script>
</i></td>
& time the HTML
</tr> document was last
</table>
</body>
changed
</html>
<html>
<!–- CS443 js14.html 16.08.2006 -->
navigator.appName
<head>
property that gives <title>Dynamic Style Page</title>
the browser name
<script type="text/javascript">
navigator.appVer if (navigator.appName ==
sion property that "Netscape") {
gives the browser document.write('<link
version rel=stylesheet '+
'type="text/css"
<!-- MSIE.css --> href="Netscape.css">');
}
<!-- Netscape.css else {
a {text-
--> document.write('<link
decoration:none;
rel=stylesheet ' +
font- 'type="text/css"
a {font- size:larger;
family:Arial; href="MSIE.css">');
color:red;
color:white; }
font-
background- </script>
family:Arial} </head>
color:red} a:hover
{color:blue}
<body>
Here is some text with a
<a href="javascript:alert('GO
AWAY')">link</a>.
</body>
</html>
•In order to use an HTML validator, and not get error messages from
the JavaScript portions, you must “mark” the JavaScipt sections in a
particular manner. Otherwise the validator will try to interpret the
script as HTML code.
•To do this, you can use a markup like the following in your inline code
(this isn’t necessary for scripts stored in external files).
<script type=“text/javascript”>
// <![CDATA[
document.write(“<p>The quick brown fox jumped over the lazy
dogs.</p>”);
// **more code here, etc.
// ]]>
</script>
•Since the (new) XHTML standard is written as an XML application,
validators such as the one from the W3C are actually attempting to
check an XML document for the correct structure.
•The two tags <![CDATA[ and ]]> together form an XML directive,
meaning to interpret the data between them as literal (non-parsed)
“character data”. An XML validator will effectively ignore the data
between these two tags, meaning that any symbols that would result
in an invalid document structure are ignored and do not result in an
error message from the validator.
•Because we are using these tags inside of a JavaScript block, and
they are not JavaScript commands, we precede each of them with
a (JavaScript) comment marker, hence the two forward slashes
before each tag.
Accessing elements on the page using
JavaScript functions
JavaScript and forms
Events, capturing user input
The Document Object Model, and
manipulating the webpage
In JavaScript, all numbers are floating point
Special predefined numbers:
Infinity, Number.POSITIVE_INFINITY -- the result of
dividing a positive number by zero
Number.NEGATIVE_INFINITY -- the result of dividing a
negative number by zero
NaN, Number.NaN (Not a Number) -- the result of dividing
0/0
• NaN is unequal to everything, even itself
• There is a global isNaN() function
Number.MAX_VALUE -- the largest representable number
Number.MIN_VALUE -- the smallest (closest to
zero) representable number
In JavaScript, string is a primitive type
Strings are surrounded by either single
quotes or double quotes
There is no “character” type
Special characters are:
\0 NUL \v vertical tab
\b backspace \' single quote
\f form feed \" double quote
\n newline \\ backslash
\r carriage return \xDD Unicode hex DD
\t horizontal tab \xDDDD Unicode hex DDDD
charAt(n)
Returns the nth character of a string
concat(string1, ..., stringN)
Concatenates the string arguments to the recipient string
indexOf(substring)
Returns the position of the first character of substring in
the recipient string, or -1 if not found
indexOf(substring, start)
Returns the position of the first character of substring in the
given string that begins at or after position start, or -1 if not
found
lastIndexOf(substring), lastIndexOf(substring,
start)
Like indexOf, but searching starts from the end of the recipient
string
match(regexp)
Returns an array containing the results, or null if no match
is found
On a successful match:
• If g (global) is set, the array contains the matched substrings
• If g is not set:
• Array location 0 contains the matched text
• Locations 1... contain text matched by parenthesized groups
• The array index property gives the first matched position
replace(regexp, replacement)
Returns a new string that has the matched substring
replaced with the replacement
search(regexp)
Returns the position of the first matched substring in
the given string, or -1 if not found.
The boolean values are true and false
When converted to a boolean, the
following values are also false:
0
"0" and '0'
The empty string, '' or ""
undefined
null
NaN
There are special values undefined and null
undefined is the only value of its “type”
This is the value of a variable that has been declared but
not defined, or an object property that does not exist
void is an operator that, applied to any value, returns
the value undefined
null is an “object” with no properties
null and undefined are == but not ===
As in C and Java, there are no
“true” multidimensional arrays
However, an array can contain arrays
The syntax for array reference is as in C
and Java
Example:
var a = [ ["red", 255], ["green", 128] ];
var b = a[1][0]; // b is now "green"
var c = a[1]; // c is now ["green", 128]
var d = c[1]; // d is now 128
The unary operator typeof returns one of
the following strings: "number", "string",
"boolean", "object", "undefined", and
"function"
typeof null is "object"
If myArray is an array, typeof myArray is "object"
JavaScript has “wrapper” objects for when a
primitive value must be treated as an object
var s = new String("Hello"); // s is now a String
var n = new Number(5); // n is now a Number
var b = new Boolean(true); // b is now a Boolean
Because JavaScript does automatic conversions as
needed, wrapper objects are hardly ever needed
JavaScript has no “casts,” but conversions can be
forced
var s = x + ""; // s is now a string
var n = x + 0; // n is now a number
var b = !!x; // b is now a boolean
Because JavaScript does automatic conversions as needed,
explicit conversions are hardly ever needed
Every variable is a property of an object
When JavaScript starts, it creates a global object
In client-side JavaScript, the window is the global
object
It can be referred to as window or as this
The “built-in” variables and methods are defined here
There can be more than one “global” object
For example, one frame can refer to another frame with
code such as parent.frames[1]
Local variables in a function are properties of a
special call object
In HTML the window is the global object
It is assumed that all variables are properties of this object, or
of some object descended from this object
The most important window property is document
HTML form elements can be referred to by
document.forms[formNumber].elements[elementNumber]
Every HTML form element has a name attribute
The name can be used in place of the array reference
Hence, if
• <form name="myForm">
<input type="button" name="myButton" ...>
• Then instead of document.forms[0].elements[0]
• you can say document.myForm.myButton
A variable is local to a function if
It is a formal parameter of the function
It is declared with var inside the function (e.g. var x = 5)
Otherwise, variables are global
Specifically, a variable is global if
It is declared outside any function (with or without var)
It is declared by assignment inside a function (e.g. x = 5)
When a function is a property of an object,
we call it a “method”
A method can be invoked by
either of call(object, arg1,
..., argN) or
apply(object, [arg1, ..., argN])
call and apply are defined for all functions
• call takes any number of arguments
• apply takes an array of arguments
Inside the funcBoth allow you to invoke a function as
if it were a method of some other object, object
tion, the keyword this refers to the object