Webmaster''''s Guide to the Wireless Internet part 12 potx
Bạn đang xem bản rút gọn của tài liệu. Xem và tải ngay bản đầy đủ của tài liệu tại đây (179.89 KB, 10 trang )
82 Chapter 3 • A New Markup: WML
Figure 3.3 shows that the SDK has picked up that the closing element for
<card> did not match the closing element for <p>, which causes an error.
DTDs are essentially a set of rules that say what elements can exist and in
what form they can appear.Thus, it determines whether the element contains
content or not, what attributes it has, and where it can appear in the document.
In the following example, taken from the WML DTD, you can see how the head
element should appear (we added line numbers for reference):
1 <!ELEMENT head ( access | meta )+>
2 <!ATTLIST head
3 %coreattrs;
4 >
Getting to know how to interpret DTDs can be worthwhile because they
often provide a definitive guide to what can and cannot be used in the language
they define.Although appearing complex at first, the DTD is written according
to a convention known as Backus Naur Form (BNF). BNF is a notation that
describes the syntax a language must use. Knowing how BNF works will allow
you to “read” the specification with ease. Our preceding example reads as follows:
■
In Line 1, !ELEMENT specifies the name of the element (<head>) and
what other elements it may contain. For the <head> element, this can
be <access> or <meta>.
www.syngress.com
Figure 3.3 The Error in Detail
159_wg_wi_03 10/22/01 4:33 PM Page 82
A New Markup: WML • Chapter 3 83
■
In line 2, !ATTLIST specifies which attributes the <head> element can
contain. In this case, the <head> element can contain only those defined
as core attributes elsewhere in the DTD; the attribute %coreattrs; is a refer-
ence to another part of the DTD where the core attributes are specified
in a similar fashion to the head.
The various WML elements are covered in the later section “WML Elements.”
NOTE
You can find out more about DTDs and XML at www.xml.com.
Using WML Syntax
As mentioned earlier,WML is XML-based, which means that every WML file is
an XML file and must contain a reference to the DTD.Therefore, the following
must be included at the beginning of every WML file:
<?xml version="1.0"?>
<!DOCTYPE wml PUBLIC "//WAPFORUM//DTD_WML_1.1//EN"
/>This code tells the device interpreting the file that this file is an XML docu-
ment (in this case, conforming to XML version 1.1) and that the DTD is located
at www.wapforum.org/DTD/wml_1.1.xml.
WARNING
Unlike HTML, the XML declaration must be on the first line. Even a single
space before the XML declaration will cause an error—the browser
cannot interpret white space until it knows the file is an XML file and to
which DTD it should be referring. In HTML, white space at the beginning
of a file is highly prevalent because pages are frequently constructed
dynamically from a database, which results in files with expanses of
white space where the server-side code was stripped out (this can of
course be avoided by sensibly constructing the server-side code).
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 83
84 Chapter 3 • A New Markup: WML
Following Syntax Rules
WML can consist of a mixture of entities, elements, attributes, comments, and
variables. Many of the syntax rules in WML are directly inherited from XML:
■
All WML documents must contain the root element WML that contains
any other WML elements.
■
All elements must be lowercase (with the exception of keywords such as
DOCTYPE and ENTITY, which are in uppercase).
■
All element attribute values must be quoted.
■
All elements must have corresponding endings.
■
Elements which are empty elements (that is, have no content) must use
/> to signify closure of the element.
■
All white space is treated as significant.
■
Elements must not overlap, though they may be nested (as with HTML).
■
Element names are case-sensitive: name and Name are not the same.
Replacing Entities
Either numeric or named, entities are specific characters within the document
character set that require escaping in WML or that may not be supported for
entry within editors (due to their using a different character set, for example).All
WML entities have the following form:
&entityreference;
Thus, the commonly used ampersand (&) becomes:
&
In WML, certain characters are used to denote special situations, such as the
start and end of an element (< and >). Simply placing the character (<) in the
content would thus generate an error. Entities are used to reference content by
placing the entity on the content as a placeholder.Thus, to show a code snippet,
relevant entities would be replaced as follows (Figure 3.4 shows the result when
viewed on a device or emulator):
<p>The <p> element</p>
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 84
A New Markup: WML • Chapter 3 85
As you can see in Figure 3.4, the < and > entities as output on a device
show how you can insert content into WML that would not otherwise be possible.
Closing Elements
Elements are similar to HTML elements (or tags) in that they specify the markup
and structure of a WML file. However, due to the use of XML,WML elements
do not seek to provide formatting information.This is left to the end device to
work out.
Elements can be found in one of two forms:
<element>Content</element>
or:
<element/>
Although the first example, with its opening and ending tag, will be familiar
to anyone who has used HTML, notice the backslash on the second example.
WML is far less forgiving than HTML, and if an element is not a container for
content, you must use the backslash as shown to explicitly close it. If the element
does not contain any content but has a corresponding closing element, you must
still include the closing element, for example:
<element> </element>
www.syngress.com
Figure 3.4 WML Entities
159_wg_wi_03 10/22/01 4:33 PM Page 85
86 Chapter 3 • A New Markup: WML
Characterizing the Element with Attributes
Although an element can provide information, such as identifying enclosed con-
tent as a paragraph (as in the <p> element), further information about the ele-
ment’s characteristics is provided by adding attributes to the element, as necessary.
Thus, if you want to right-align a paragraph, the attribute would read as follows:
<p align="right">Right-aligned text</p>
All attributes must be enclosed in quotation marks (either single or double)—
this is strictly required by WML. Failure to enclose all attributes in quotes gen-
erate an error, and the content will fail to load in the device.
NOTE
No attribute may be named more than once in any element that
contains it.
You can find more information about which attributes are applicable to
which elements in the later “WML Elements” section.
Case Sensitivity
In addition to enclosing all element attributes in quotes,WML—unlike
HTML—is case-sensitive.All elements and attributes must be in lowercase.
Handling White Space
White space within text is handled by WML in the same manner as XML default
white space handling.White space before or after elements is ignored completely,
and in fact, sequences of white space (such as tabs and spaces) will be collapsed
into a single space, the treatment of which will depend on the locale.Thus, you
cannot space words within text by using multiple spaces or tabs.The handling of
white space within elements and attributes is dependent on the rules defined in
the XML specification referenced within the XML declaration at the beginning
of the WML deck.
Collapsing Carriage Returns
The numerous examples in this chapter make use of carriage returns and tabs to
format the content, making it easier to follow.You can continue this practice
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 86
A New Markup: WML • Chapter 3 87
through to actual WML files to make code easier to read when editing. In WML,
the user-agent (or the WAP Gateway) will take carriage returns, tabs, new lines,
and multiple spaces, and collapse them to a single space as part of the process of
minimizing the file size.
Commenting
WML comments are much like their HTML counterparts, though they must not
extend across more than one line.Thus, the following is valid:
<! A comment >
This, however, is invalid:
<! A Comment
and another >
Using Variables
A variable is a temporary parameter, commonly stored on the user-agent into
which a value can be inserted for future access during the session.A variable
within WML has a very exact syntax.
Where white space can be expected to signify the end of a variable, the
following is acceptable:
$identifier
If this is not the case, the variable must be enclosed in parentheses:
$(identifier)
Variables are covered in more detail in Chapter 4.
Formatting Text
Now for the bad news:Whereas HTML contains much to help you format your
content, such as font tags and Cascading Style Sheets, the WML specification
leaves that pretty much up to the receiving device.This behavior was intentional,
given the huge range of possible wireless devices that are bound to be created,
and means that if you rely on some kind of formatting to get across a key part of
your application functionality, you could be heading for problems.This is com-
pounded by the haphazard implementation of what little formatting control is
allowable in WML by the currently available devices.
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 87
88 Chapter 3 • A New Markup: WML
The formatting that is available addresses the bare minimum necessary and
includes the following:
■
Elements for bold and strong emphasis (<b> or <strong>)
■
Elements for italic and emphasized text (<I> or <em>)
■
Elements for underlined text (<u>)
■
Elements to manipulate font size (<small> and <big>)
Displaying Fonts
You have very little influence over the fonts used in a device.They will almost
certainly be nonproportional in nature, which rules out fun ASCII text of the
type sometimes used to create basic pictures because these require proportional
fonts.Take a look at the differences in the interpretation of markup containing
formatting instructions in the comparison between Nokia and Phone.com
devices in Figures 3.5 and 3.6. Both emulators are displaying the same WML
content, but the Nokia device fails to display the bold text as intended. In fact,
the actual Nokia 7110 does not display formatting such as emphasized, bold, or
underlining.
www.syngress.com
Figure 3.5 Use of <b> on a Phone.com Browser
159_wg_wi_03 10/22/01 4:33 PM Page 88
A New Markup: WML • Chapter 3 89
Both emulators will interpret the WML, which is well-formed and valid, but
the Nokia browser doesn’t have the capacity to render bold text, so the <b> ele-
ment is simply ignored, and the text is displayed in the normal default font.
NOTE
Figures 3.5 and 3.6 are for guidance only. Although sophisticated, emu-
lators do not necessarily exactly reflect what your WML will look like on
an actual live device. Always remember to check your WML on a real
device to prevent any unwelcome surprises. You can learn more about
emulators later in this chapter in the section called “WML Editors, WAP
SDKs, and Emulators.”
Reserved Characters
WML uses a number of special characters to denote characters that may not be
present within the current document encoding.The characters have the following
syntax:
& entity reference ;
www.syngress.com
Figure 3.6 Use of <b> on a Nokia Browser
159_wg_wi_03 10/22/01 4:33 PM Page 89
90 Chapter 3 • A New Markup: WML
The entity reference can be numerical, named, or hexadecimal, for example:
 
<
 
These will all cause the < character to be displayed on the device screen.
It is vitally important to include the ampersand and semicolon because
missing either of these will generate an error.You need to consider seven impor-
tant entities within WML, as shown in Table 3.1.
Table 3.1
Character Entity Reference Table
Entity Name Entity Entity Description
quot " Quotation mark
amp &#38; Ampersand
apos ' Apostrophe
lt &#60; Less-than
gt > Greater-than
nbsp   Nonbreaking space
shy ­ Soft hyphen
Displaying Tables
Tables in WML are fairly simple in comparison to HTML tables, reflecting the
limitations of the display generally available.You cannot nest tables, nor can you
set explicit widths, although you must specify the number of columns within a
row set.
WARNING
Only the more advanced devices, such as the Ericsson R320, are likely to
be capable of displaying data marked up using tables. You must carefully
inspect the target device to ensure that the device is capable of handling
the display of tables and that the contents of the table will display as
intended.
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 90
A New Markup: WML • Chapter 3 91
Meta Information
Meta information is information designed for use by the browser or gateway and
is not displayed to the user. Meta information is included within the head of the
document, the precise range of which is generally left to the device manufacturer
to specify.
Controlling Caching
Whenever a request is made for a WML page, the device will cache the received
file, allowing it to be called up from the cache instead of having to download the
whole file again.This can considerably speed up browsing .The current devices
do not have large caches; when dealing with dynamic data that may have the
same URL but different contents each time the URL is accessed, you may want
to disable caching.
You can control caching in a number of ways. Caching is a somewhat com-
plicated issue to address because many devices cache differently than others and
also differ in what instructions can be given to them to control caching.The fol-
lowing meta information is intended to cater to as many devices as possible:
<meta http-equiv="Cache-Control" forua="true" content="no-cache,
max-age=0,must-revalidate, proxy-revalidate, s-maxage=0"/>
NOTE
As you can see from the examples in this section, you can specify from a
number of headers to cater for as many eventualities as possible.
Whenever caching is a concern, make sure you do the appropriate back-
ground research and testing to target the device on which you will be
deploying your solution.
Additionally, making use of the Expires header is an ideal way to control
browser caching. It works very well on many different devices. If the document is
set to expire in the future, the browser will always re-request the content, thus
ensuring the document is never reloaded from the cache.You can include this in
server-side script, such as ASP:
<%
Response.Expires = -1
www.syngress.com
159_wg_wi_03 10/22/01 4:33 PM Page 91
