HTML PrimerHTML Primer-- ii --
-- 1 --
HTML stands for the Hypertext Markup Language. It was defined by Tim Berners-Lee in 1990 as the method of marking up pages of information to be looked at by a browser. The Standard Generalized Markup Language (SGML) had been used widely at CERN for documentation and HTML was a cut down version of the CERN documentation language. HTML is not a static language and has gone though a number of iterations over the last 10 years:
XHTML 1.0 (functionality defined in HTML 4.01) is the version of HTML widely supported by today's browsers. An aim from the early days of the Web was that style and content of web pages should be kept distinct. This helps to ensure that pages can be displayed on the widest set of devices and allows them to be used by people with accessibility problems (either physical or due to their location). In the early days, HTML was used for defining both the style and the content. Today, the aim is that styling should be provided through a style sheet and that HTML should be responsible for defining the content. HTML 4.01 includes a number of legacy styling attributes which are deprecated as such styling should be done using a style sheet. These attributes do not appear in XHTML 1.1. This Primer will stick to describing the basic features of XHTML1.1 (functionality defined in HTML 4.01) but ignoring all the deprecated styling attributes of HTML 4.01 and XHTML 1.0.
There is currently a move towards application-specific markup by using XML (Extensible Markup Language) as the language for defining markup languages. XML derives from SGML. This means that a reformulation of HTML as an application of XML makes a great deal of sense. The facilities available in XML and the tools surrounding XML become available. For this to happen, HTML must be written more strictly than is required in the HTML specification. For example, each start tag must have an end tag. In consequence, this Primer is described as though all that a browser supports is XHTML 1.0. The result is a succinct Primer.
Frames in HTML 4.01 and XHTML 1.0 are not deprecated but, in many applications and environments can cause problems. In consequence there are actually six different flavours of XHTML 4.01/XHTML 1.0:
-- 2 --
Information in HTML is marked up by adding to the information markup that describes its structure. Marked up information consists of a set of elements that have the general form:
<xyz>Some content</xyz>
The terminology used is that this defines an xyz element. The start tag for the element is <xyz> and the end tag is </xyz>. The end tag looks very similar to the start tag except for the addition of the / before the name of the element. Spaces are not allowed after the < in the start tag and the </ in the end tag. Spaces are allowed before the closing >. The content of the element is any text to be marked-up and can contain other elements. So, for example, this is also allowed:
<xyz>Some <abc>more</abc>content<def>added</def></xyz>
The tags must be correctly nested. This would be illegal:
<xyz>Some <abc>more content<def>added</abc></def></xyz>
Either the def element must be completely inside the abc element or completely outside it.
Elements may have attributes and these form part of the start tag:
<xyz attr1="sometext" attr2='somemoretext'>Some content</xyz>
Attributes consist of the attribute name followed by the equal sign and a value which is a string of text inside one of two types of quotation marks (double quotes: ASCII decimal 34; single quotation marks: ASCII decimal 39). Two types are allowed in case the attribute's value contains quotation marks. The start and end quotation marks must either both be single or both be double quotes.
A shorthand is provided for element's that contain no content. For example:
<xyz attr1="sometext" attr2='somemoretext'></xyz>
may also be written:
<xyz attr1="sometext" attr2='somemoretext' />
A strict XHTML 1.0 document would have the following structure:
<?xml version="1.0" encoding="iso-8859-1" ?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html> <head> . . . </head> <body> . . . </body> </html>
-- 3 --
The document starts by indicating that it is XML and follows with a document type declaration that says what type of document this is.
The XML declaration is optional. IE6 runs the browser in a non-standard mode called Quirks mode if the XML declaration is present. In consequence, until IE7 replaces IE6 completely, users are recommended not to include the XML declaration. Then, all the main browers will produce similar results for core markup and styling. As we are going to be very strict in what we write, we can tell the browser this and it may help it to run more efficiently. After this comes the complete document which is an html element. It consists of two elements, the head element and the body element. The first contains elements that define meta information about the document while the second gives the content of the document. Most of this Primer will concentrate on defining what can be in the body of the document. Treat the first two lines as text that you have to start the document with. About the simplest document that we can have is:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html> <head> <title>My First Document</title> </head> <body> <p>My first document</p> </body> </html>
The head consists of a single title element and the body consists of a single p for paragraph element. Your browser is likely to display the title as part of its decoration around the document. The document displayed will be the paragraph My first document displayed something like:
My first document
We have added a background to whatever our browser has produced for this page just so that the results of the markup are differentiated from the text in the Primer. We will start by looking at the elements that can be used in the body of the document and come back to discuss the head later. From now on you can assume that the document has the structure given above and all we are discussing are the elements that make up the contents of the body element.
The HTML document will normally be stored in a file. It is usual for that file to have an extension .htm or .html.
If we were going to be less strict and define the document as an HTML 4.01 document that allows styling attributes, the first two lines would be:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"> <html lang="en">
The lang attribute can be added to indicate that the text is English. It is posible that the browser is presenting the page as audio output in which case this is needed for pronunciation.
h1 h2 h3 h4 h5 h6 p
We have already introduced the paragraph element above. HTML allows you to define 6 levels of headings with h1 being the most important and h6 the least important. So for example you could have:
<h1>1. My Book</h1> <h2>1.1 The First Section in My Book</h2> <h3>1.1.1 The First Sub-Section in my book</h3> <h4>1.1.1.1 The First sub-sub-section</h4> <p>This is the very first paragraph in my major work which tells in great detail all the things that I have ever done. Sit back and enjoy it!<p> <h4>1.1.1.2 The second sub-sub-section</h4> <p>Which does not add a great deal to the first section which sets the scene quite well.</p> <h3>1.1.2 The Second Sub-Section in my book</h3> <p>And so on. . .</p>
-- 4 --
A browser might display this as follows:
This is the very first paragraph in my major work which tells in great detail all the things that I have ever done. Sit back and enjoy it!
Which does not add a great deal to the first section which sets the scene quite well.
And so on. . .
How the browser chooses to display it will depend on the browser. They are expected to make the h1 elements more prominent than the h2 elements and so on. Users normally do not use a higher number heading if they have not first used all the lower numbers. It would be regarded as bad practice to use h1, h2 and then h4 without using h3 but it is not illegal. We have made the h1 and h2 elements more exotic by centring them and making them a different colour. Most browsers are not that adventurous and just change the font size, style and boldness to differentiate the headings.
<h1 id= class= style= title= lang= . . . onclick= onmousedown= . . . >
There are some attributes that apply to nearly all elements. They fall into three classes. The first set identifies and styles the element more precisely, a second set concerns the language the element is written in and a third set is concerned with interacting with the element. We will only look at the first set at the moment.
The first set are:
An example of their use is:
<h1 title="My First Title" id="b123" class="header first topic"> . . . </h1>
Unless stated otherwise, it can be assumed that any element in the body of the document may have these attributes and these will be the only attributes unless others are mentioned. Appendix B has the element names in bold that can have all the global attributes.
em strong
Within a heading or paragraph, it is likely that some phrase or word may need to be emphasised to add strength to some point. HTML provides two elements that can occur within the text of other elements to achieve that goal. The element em emphasises the enclosing text and the element strong emphasises it even more. It is up to the browser to decide how it will represent the two levels of emphasis. The text enclosed by the em element might be in italic and the text within the strong element might be bold but that is up to the browser. All that is hoped for is that the latter is emphasised greater than the former. An example might be:
<p>This is a book that you really <em>must</em> read. Phone me tomorrow <strong>in the afternoon</strong> and tell me what you think.</p>
-- 5 --
This might be rendered as:
This is a book that you really must read. Phone me tomorrow in the afternoon and tell me what you think.
It might also be rendered as:
This is a book that you really must read. Phone me tomorrow in the afternoon and tell me what you think.
These are not the only elements that you can use to delineate some parts of an element like paragraph and headers. We have described these two now because they are frequently used and it allows us now to discuss a major breakdown of HTML elements into different types.
The important point in terms of the HTML document is that the reader can see that the word must is emphasised and that the phrase in the afternoon is emphasised even more. In terms of content, that is all that the reader of the page needs to know and that is all that is conveyed by the elements.
Some guidance is given within HTML as to how the elements in the document should appear. The headers should appear on separate lines. If there are two paragraphs, one following another then the first paragraph should normally be before the second and the end of one and the start of the other should be obvious. For emphasis, it should be clear that the emphasised words are not separate paragraphs but emphasised words within the paragraph.
In HTML, this is achieved by telling the browser what type of element each one is. They fall into two major classes:
Unless told otherwise, the browser is expected to present the two types differently. Block-level elements are expected to start on a newline while inline elements do not. When we get round to discuss styling, things get a bit more complicated but while we are just looking at HTML, this is sufficient in describing how the browser is expected to present the elements to the reader. The complications start to appear when there are block-level elements enclosing block-level elements. In some sense we already have that as the body element is a block-level element and we have shown that other block-level elements can be contained within it.
There are some elements that are used to define how other elements should appear. These are not block-level elements themselves and so fall outside the layout model. There are also elements that can really appear anywhere (script elements are a good example) but apply globally. Their appearance at a particular point in the document is just to aid readability.
<br />
For block-level elements like paragraphs, the normal pattern is to ignore additional white space characters like extra spaces and fill each line up before carrying on with the paragraph on the next line. If there is a need to force a line break within a paragraph, inserting the inline element br will cause this to happen. For example:
<p>This is a paragraph with a great deal of text in it to make sure that we have more than one line in it and here<br /> is where we want to force a line break</p>
This will be presented as:
This is a paragraph with a great deal
of text in it to make sure that we have
more than one line in it and here
is where
we want to force a line break
Appendix B gives a Quick Guide to all the elements allowed in our strict version of HTML and for each it lists their attributes, whether they are inline, block or other and any special conditions that apply. Once familiar with the elements, this should be useful for jogging your memory on a particular element.
The rest of this Primer will consider the remaining elements in some detail.
-- 6 --
ul ol li
Two simple lists that are frequently used in documents are unordered lists, a set of bullet points that are equally important and ordered lists where the order of reading the list has some importance. HTML allows you to define both types of lists. The element ul element defines an unordered list and the element ol defines an ordered list. They both contain a set of list elements. Both use the li element to define the items in the list. For example:
<p>Here is an unordered list:</p> <ul> <li>A bullet point</li> <li>Another bullet point</li> <li>And a third</li> </ul> <p>Here is an ordered list:</p> <ol> <li>First point</li> <li>Second point</li> <li>Third point</li> </ol>
This would appear something like:
Here is an unordered list:
Here is an ordered list:
The only element allowed within a ul element or an ol element is a li element. Please note!
In this case, the browser has chosen a certain style of bullet symbol to indicate the individual items and has numbered the ordered list using the numbers 1, 2, 3 etc. Styling the document allows the user to specify different bullet symbols and different number representations (a, b, c or i, ii, iii even). That does not concern us here. We will just use the ones provided by the browser.
Well it all looks pretty straightforward so far. We have a set of list elements each containing text (some of which could be emphasised) and that is as far as it goes. However, it is a great deal more complicated than that because the element li is a block-level element and within an li element it is possible to have other block-level elements. A very obvious example is to put another list within a list, a markup that is often required. This works in much the same way for both ul and ol elements so we will only consider one of them. Let us start with our example above:
<p>Here is an unordered list:</p> <ul> <li>A bullet point</li> <li>Point about bags and boxes and rings</li> <li>And a third</li> </ul>
Anywhere within the second bullet (or the others) we can add another ordered or unordered list so here is what happens when we insert a list after the word bags:
-- 7 --
<p>Here is an unordered list:</p> <ul> <li>A bullet point</li> <li>Point about bags <ol> <li>First point</li> <li>Second point</li> </ol> and boxes and rings together</li> <li>And a third</li> </ul>
The result is as follows:
Here is an unordered list:
Let us go wild and put lists after the words bags, boxes and rings:
<p>Here is an unordered list:</p> <ul> <li>A bullet point</li> <li>Point about bags <ol> <li>First bag point</li> <li>Second bag point</li> </ol> and boxes <ol> <li>First box point</li> <li>Second box point</li> </ol> and rings <ul> <li>A ring point</li> <li>Another ring point</li> </ul> together </li> <li>And a third</li> </ul>
This is displayed by my browser as:
Here is an unordered list:
-- 8 --
We could go on playing this game with lists inside lists inside lists. Note that the browser has chosen a different bullet symbol for the inner list and has been kind enough to indent each sub-list to make it more readable. When we do our own styling, we can provide even more control but that is sufficient for now. However we need to show a common error that you must not do:
<p>The main points to be covered are:</p> <ul> <li>Lists</li> <ul> <li>Ordered Lists</li> <li>Unordered Lists</li> </ul> <li>Headers</li> </ul>
This looks alright and most likely the browser will accept it but it is illegal. Here the inner unordered list appears as an element inside the outer unordered list at the same level as the li elements and that is not allowed. It is only the li element that can have other block-level elements within it. What should be written in this case is:
<p>The main points to be covered are:</p> <ul> <li>Lists <ul> <li>Ordered Lists</li> <li>Unordered Lists</li> </ul></li> <li>Headers</li> </ul>
We have seen that we can add lists within lists but the li element can contain any block-level element so paragraphs and headers can also appear as part of the list element:
<p>The main points to be covered are:</p> <ul> <li> <h4>Lists</h4> <p>Ordered Lists</p> <p>Unordered Lists</p> </li> <li>Headers</li> </ul>
This will be displayed as:
The main points to be covered are:
Ordered Lists
Unordered Lists
dl dt dd
The third type of list available in HTML is a definition list which consists of a dl element containing pairs of dt and dd elements. The dt element defines the term to be defined and the dd element gives the definition. For example:
<p>The following definitions apply:</p> <dl> <dt>Unordered List</dt> <dd>List in any order that has a bullet in front of it</dd> <dt>Ordered List</dt> <dd>List in a precise order that are numbered</dd> </dl>
-- 9 --
The format for the definition list is again up to the browser. the browser might display it as:
The following definitions apply:
HTML does not put any constraints on the positions of dt and dd elements within the dl element so that the user can define terms without giving a definition and vice versa. This can sometimes be useful when composing a document but in its final form it is sensible to have pairs of elements with each term followed by its definition. A dt element may be followed by several dd elements when there are several definitions that apply.
The dt element acts rather like a header element and should just contain text possibly with some inline elements to give emphasis etc. On the other hand the dd element is very much like the li element and can contain any mixture of block-level and inline elements allowed in the body of the document.
<a href= > <h1 id= >
The success of the Web has come about because it allows documents on the web to link to other documents or even to parts of the current document. The a element, short for anchor, is the element that achieves that in HTML. It is mainly used to define the source of the link. An element within a document is defined as the destination of the link by the id attribute. For example:
<p>More information will be found on <a href="http://www.w3.org/fred/intro.htm#mydestination">the W3C site</a>. Please visit it.</p>
This will be displayed something like:
More information will be found on the W3C site. Please visit it.
Here the anchor element defines the source of the anchor by the href attribute. Such an anchor is displayed by most browsers with some indication that the inline text that is the content of the element is emphasised in some way. In the example above, the text is underlined, and in a different colour. Some browsers will show the source element differently depending on whether the link has been traversed before or not. Clicking on the link will cause the page with name intro.htm in the directory fred on the W3C site to be opened and positioned so the element with its id attribute set to mydestination is visible. If this is in the middle of the document, the document will be scrolled to a position where this element is visible, usually at the top of the screen. The browsers usually default to the current web page being replaced by the one referenced in the anchor element.
The href attribute must be a Uniform Resource Locator although this can be relative or absolute. Given that the source document is http://www.w3.org/fred/index.htm, some examples of URLs are:
-- 10 --
If the destination of a link is part of a document, it is defined by the id attribute. For example:
<p>More information will be found in <a href="http://www.w3.org/fred/intro.htm#sec3">Section 3</a>. Please visit it.</p> . . . <h1 id="sec3">3. Section Three of Document<h1>
The id attribute must be unique within the document.
< base href= />
The base element appears in the head of the document and there should be only one of them. It has a single attribute, href that defines the base to be used for all relative URLs in the document. In Section 4.2, the base for the examples was http://www.w3.org/fred/index.htm. If there was a base element in the document, the value specified by the href attribute would be used in place of this. The base element is mainly used when the same document can be retrieved from different places. The document may have arrived from one destination but the author would like all the references in the document to be made as though it came from a different origin.
pre address blockquote hr
HTML identifies three types of paragraph for special treatment. They are:
A fourth element , the hard rule could either be regarded as styling or the null element. It is also described in this Section.
<pre>
The ability to output preformatted text is needed in HTML because the usual behaviour is to throw unnecessary whitespace away. For example:
-- 11 --
<p>
Cascading
Style
Sheets from the
World
Wide
Web
Consortium
</p>
will be displayed as:
Cascading Style Sheets from the World Wide Web Consortium
This loses all the flavour of cascading that was in the original layout. Replacing p by pre as the element produces something like:
Cascading
Style
Sheets from the
World
Wide
Web
Consortium
The browser is expected to take account of all the whitespace and render it correctly. To help the layout of spacing to be just right, the browser is asked to use a fixed-pitch or monospaced font as has been done here. The usual method of presenting paragraphs is to fill up each line and then wrap on to the next. For preformatted text, if the line is long it should be left that way. The user will then have to scroll across to see it. One problem with preformatted text is the positioning of tabs. as no guidance is given by HTML, browsers are likely to make different decisions on what a tab means. In consequence, it is recommended not to use tabs in preformatted text.
Because preformatted text is trying to get the spacing just right, some of the inline elements are disallowed. In particular, sup, sub, object, img are not allowed to be used within a preformatted paragraph
<address>
The address element has similar characteristics as the p element, and has no additional attributes. In consequence, it could easily be achieved by having normal paragraphs with a class address. It is included as separate markup because of the importance of ownership and thus addresses. Browsers tend to render the element using italic script but do not preformat the text. For example:
<address> Bob Hopgood<br /> Oxford Brookes University<br /> bhopgood@brookes.ac.uk<br /> </address>
This would be rendered as:
Note that the element needs to force the line breaks by using the br element. An alternative approach using CSS is to request that the browser displays it as pre-formatted text.
<blockquote cite= >
Copyright is an important issue particularly on the Web where borrowing material is relatively easy. As the Web allows every resource to be uniquely identified, it is appropriate to define a block-level element that provides the additional attribute to point to the location from where the quotation was taken. For example:
-- 12 --
<p>The role of W3C is as follows:</p> <blockquote cite="http://www.w3.org"> <p>The World Wide Web Consortium (W3C) develops interoperable technologies (specifications, guidelines, software, and tools) to lead the Web to its full potential as a forum for information, commerce, communication, and collective understanding.</p> </blockquote>
This would be displayed as:
The role of W3C is as follows:
The World Wide Web Consortium (W3C) develops interoperable technologies (specifications, guidelines, software, and tools) to lead the Web to its full potential as a forum for information, commerce, communication, and collective understanding.
Note that the browser has indented the text and changed the colour of the text as well. Unfortunately, it has not provided any mechanism for seeing the citation apart from looking at the source code.
<hr />
The hard rule element is probably more a styling rule than an information content one although it could be used to delineate major sections in a document. It has no additional parameters and is used as follows:
<p>My first paragraph</p> <hr /> <p>My second paragraph</p>
This would be presented as:
My first paragraph
My second paragraph
The rule can be styled using CSS.
div
HTML has a generic block-level element called div. It can have any of the inline or block-level elements within it and it has no additional attributes. Its purpose is to group elements together so that they can act as a set. The element is quite powerful in terms of styling as it can have a class attribute defined for it so that a specific div element is styled differently. If the user wished to interact with a part of the document, enclosing that part with a div element and assigning it a unique id attribute ensures that the part of the document is located precisely. Until those are discussed in detail, the element does not appear to provide much additional functionality. However, by indicating the major sections of a page, it can be useful to users with accessibility problems to indicate the structure of the page and thus help navigation by a screen reader.
cite defn code samp var abbr acronym
HTML provides a number of inline elements that more precisely define the semantic meaning of the enclosed text. This may be useful for the browser in terms of presentation but could also be of use to a search engine or a spell checker. They indicate phrases in the document that are not normal textual information that would appear in a paragraph. Here is an example of most of them:
-- 13 --
<p>As <cite>Bob Hopgood</cite> points out, <dfn>HTML stands for Hypertext Markup Language</dfn>, which is a <abbr title="World Wide Web Consortium">W3C</abbr> Recommendation used by <acronym title="North Atlantic Treaty Organisation">NATO</acronym> to markup documents. Something like <code>class="classified"</code> is used to indicate that some text like <samp>In Confidence</samp> is added to all the classified paragraphs. If the <var>class</var> has a value of <var>unclassified</var>, no additional text is added.</p>
This would be displayed as:
As Bob Hopgood points out, HTML stands for Hypertext Markup Language,
which is a W3C Recommendation
used by NATO to
markup documents. Something like class="classified" is used to indicate
that some text like In Confidence is added to all the classified paragraphs.
If the class attribute has a value of unclassified no additional text is added.
The meaning of each is:
As can be seen in the example, most browsers make some attempt at rendering the inline elements differently. Hovering over the acronyms and abbreviations should cause the meaning to be displayed as a tooltip. The element abbr is rarely used.
HTML makes no attempt to display mathematics correctly. There is a separate markup language, MathML, for that purpose. However superscripts and subscripts appear sufficiently often that markup is provided for them:
<p> x<sub>i</sub> = y<sub>i</sub> + 2<sup>a + b<sup>2</sup></sup></p>
This would be displayed as:
xi = yi + 2a + b2
As is shown, inline elements may have inline elements inside them if it makes sense. Broadly speaking anything that is sensible is allowed.
<q cite= />
As well as the blockquote element, there is also an inline quotation element that allows quotations to occur within a paragraph. For example:
<p>David said, <q cite="David Duce">I saw Bob and he said <q cite="Bob Hopgood">SVG is nearly finished</q> so we had better start using it</q>.</p>
This would be displayed as:
David said, "I saw Bob and he said 'SVG is nearly finished' so we had better start using it".
The browser is supposed to add the quotation marks and be clever enough to use different quotation marks for the outer and inner quotes. These facilities are currently poorly supported in browsers. For example. IE6 produces no quotation marks and Opera produces the same quotation marks for both the inner and outer quotes so this is a risky element to use. Using appropriate styling, it is possible for the page author to solve the problem.
-- 14 --
span
HTML has a generic inline element called span. It can have any of the inline elements within it and it has no additional attributes. Its purpose is to bracket some inline text and elements so that they can act as a set. The element is quite powerful in terms of styling as it can have a class attribute defined for it so that a specific span element is styled differently. If the user wished to interact with a part of a block-level element, enclosing that part with a span element and assigning it a unique id attribute ensures that the part of the block-level element is located precisely. Until those are discussed in detail, the element does not appear to provide much additional functionality.
Inline images were added to HTML quite early on and provide a mechanism for adding pictorial information to Web pages. Two elements are provided:
HTML does not legislate which media types, sometimes called MIME type are supported by the browser. MIME stands for Multi purpose Internet Mail Extensions. In some cases, the browser can be enhanced by adding a plug-in provided by a different organisation to support another MIME type. For example, Scalable Vector Graphics (SVG) pictures are currently supported in IE6 using a plug-ins provided by Adobe or Corel. If the user has not down loaded the plug-in, the SVG pictures specified cannot be displayed. If the object element is used to specify the SVG file to be loaded, it could provide as an alternative either the picture in a different format that most browsers do support or it could display a message telling the user to download the plug-in. The browser has to work out what type of image is being provided and in the case of the img element it does that by data sniffing, that is MIME type detection. When the server sends the file it may send the MIME type with it. The file extension will normally give a good clue to the contents. If the browser is sufficiently intelligent, it can look at the contents of the file and determine it that way. So getting the type of file requires quite a bit of work on the part of the browser and server.
In the case of the object element the browser retrieves the value of the type attribute if provided and that gives the MIME type. This means that if it does not support the MIME type, it does not need to download the file. Some of the more usual images are (filename followed in brackets by the image type):
-- 15 --
<img src= alt= height= width= longdesc= usemap= />
The inline img element defines an image to be displayed inline. If there is insufficient space on the line for the image, it will be placed on the next line. Better control of images is available through styling. The main attributes are:
Here is an example that displays a JPEG image called bobbin.jpg that is 349 pixels wide and 91 pixels high:
<p>Hanging bobbins like this one: <img width="349" height="91" alt="a bobbin" src="bobbin.jpg" longdesc="bobbin.txt" /> were sold at hangings in England as souvenirs. The actual size is smaller: <img width="174" height="45" alt="a bobbin" src="bobbin.jpg" /> and I can make them look silly by changing the aspect ratio. Here the height has been halved: <img width="349" height="45" alt="a bobbin" src="bobbin.jpg" /> and here the width has been halved: <img width="174" height="91" alt="a bobbin" src="bobbin.jpg" > </p>
This would appear as:
Hanging bobbins like this
one: 
were
sold at hangings in England as souvenirs. The actual size is
smaller: and I can
make them look silly by changing the aspect ratio. Here the height has been
halved: and
here the width has been
halved:
-- 16 --
The file bobbin.txt might contain:
William Bradbury was murdered near Luton. William Worsley and Levi Welch were both charged with the crime but Levi Welch turned king's evidence admitting that he had robbed him but that Worsley had struck the blow that killed him. William Worsley was found guilty of highway robbery and murder. His death was the last public hanging to take place in Bedford on 31 March 1868. Bobbins were produced as souvenirs for the men in the crowd to give to their wives as a memento of the occasion. Being the last public hanging, this bobbin was produced in a reasonable quantity and so are not as rare as other hanging bobbins. At auction in good condition they are worth around 150 pounds sterling. In a shop you might pay a significant mark up on that price.
<object data= type= standby= height= width= >
The object element is designed to replace the img element and to provide facilities for inserting into a Web page files in a whole variety of formats.
The inline object element defines one of a set of objects to be displayed inline. The main attributes are:
Here is the previous example that displays an image called bobbin that is 349 pixels wide and 91 pixels high. The image is available in JPEG, GIF and PNG formats:
<object width="349" height="91" data="bobbin.jpg" type="image/jpg" standby="JPEG image downloading"> <object width="349" height="91" data="bobbin.png" type="image/png" standby="PNG image downloading" /> <img width="349" height="91" src="bobbin.gif" alt="bobbin photo" /> No image available </object>
In this case the browser will first attempt to download the JPG object. If this is not possible for any reason, it will try the PNG object and if that is also not possible, it will try to download the GIF image using the img element. If all three are not possible, it will display the text No image available. Adding an img element using a format that is almost certainly going to be available is useful for browsers that still do not recognise the object element. They will ignore those and just display the GIF image.
The support for the object element is improving but still not complete in some browsers. In consequence for most image formats (jpg, png, gif) it is often useful to continue with the img element.
< img usemap= /> <map id= > <area href= alt= shape= coords= />
It is possible to use an image as a mechanism for linking to other pages in the same way that anchors are used. To achieve this, it is necessary to define the parts of the image that will be activated when a click is made over them and also the action to take when this happens. For example:
<h3>Robert Cailliau and Murray Maloney</h3> <a href="#dummy"> <img alt="Robert and Murray" width="277" height="127" src="photorcmm.jpg" usemap="#maprcmm" /> </a> </p> <map id="maprcmm"> <area href="http://cern.web.cern.ch/CERN/Divisions/ETT/WPE/People/RobertCailliau/" shape="rect" coords="0, 0, 138, 127" /> <area href="http://www.muzmo.com/Murray/" shape="rect" coords="139, 0, 277, 127" /> </map>
-- 17 --
This would appear as:
The map element can be anywhere in the file. The img element uses the id attribute to indicate which map is being used. The img element's ismap attribute is set to the identifier of the map element. To ensure that the img element is recognised as an anchor, it is enclosed by an a element where the href attribute is not used and can be set to anything. It should be included as some browsers check that it is present.
Within the map element is a set of area elements which define parts of the image as clickable regions and the destination of the link when a user clicks on that part of the image. In the example, there are two areas. The coords attribute defines the active part of the image and the area shape is defined by the shape attribute. The example has two rectangular areas and the four coordinates define positions on the image in pixels. The four coordinates are minx miny maxx maxy where the origin is the top left corner of the image and y increases downwards. In the example, the two areas are the left and right sides of the photograph. Clicking on one or other will cause a link to the appropriate home page of Robert Cailliau or Murray Maloney. The possible shapes are:
There are a number of editors on the market that aid in the production of image maps. In most cases it is possible to achieve the same effect either by careful styling or by the use of Scalable Vector Graphics possibly with associated images. This is probably the least useful method of associating links with images.
HTML provides good support for tables. The table is divided into three main parts, a header, one or more body elements and a footer. There is also an optional caption that can be added to the table. Inside the three main parts, a table is divided into rows each having a number of columns. As with most word processing systems it is possible to give groups of columns special features and to merge two or more columns together or two or more rows together. The table is basically made up of cells and the content of a cell may be any set of block-level elements rather like the li element. So it is possible to put another table inside a single cell. In consequence, almost any table that is required can be achieved. CSS provides good control of the styling of individual cells and the table decoration. We shall not consider that further here.
<table summary= width= > <caption>. . .</caption> <colgroup align= span= width= /> <col align= span= width= /> <thead>.......</thead> <tfoot>.......</tfoot> <tbody>.......</tbody> <tbody>.......</tbody> </table>
The table head and foot are very similar to the table body elements. The main difference is that the browser is supposed to output the table head and table foot on every page of the output if it is a very long table. Most of the browsers do not appear to have implemented this. Multiple table bodies are allowed and this can be useful if different parts of the table are to be styled differently. Here is a very simple example:
-- 18 --
<table title="Membership of W3C July 2001" summary="5 column table with header row and first column the type of membership. Total in the final column and bottom row" width="90%"> <caption>W3C Membership</caption> <thead> <tr><th>Type</th><th>Americas</th><th>Europe</th><th>Pacific</th><th>Total</th></tr> </thead> <tbody> <tr><th>Full</th><td>62</td><td>29</td><td>17</td><td>108</td></tr> <tr><th>Affiliate</th><td>240</td><td>123</td><td>47</td><td>410</td></tr> <tr><th>Total</th><td>302</td><td>152</td><td>64</td><th>518</th></tr> </tbody> </table>
This would be displayed something like:
| Type | Americas | Europe | Pacific | Total |
|---|---|---|---|---|
| Full | 62 | 29 | 17 | 108 |
| Affiliate | 240 | 123 | 47 | 410 |
| Total | 302 | 152 | 64 | 518 |
The sections of the table are made up of table rows denoted by the tr element. Each row consists of a set of entries which can either be td elements (for data) or th elements (for headers). Note that the header elements can be anywhere. In this case the table has them in the top row and first column and the overall total. The browser is expected to render the two elements differently. the browser will most likely make the header data elements (th) bolder than the normal data elements (td).
The table has been defined as having a width of 90% of the width of the page. Any of the usual length values may be used but it is quite common to specify the width as a fraction of the current window size.
The summary element is aimed at users who for one reason or another will have difficulty seeing the table in full. They might be blind or have a very small display or have had the table converted to text to transmit it by voice. The aim is to give some information to make it easier to understand the table. Other facilities are provided to help the user understand the table structure. Normally the summary would not be this long.
The col and colgroup elements allow the author to define properties for a whole column or group of columns.
The colgroup element is provided to create structural divisions in the table. For example, the table above might have immediately after the caption:
<colgroup span="1" width="30%" /> <colgroup span="3" width="10%" /> <colgroup span="1" width="30%" />
This would change the table to look like:
| Type | Americas | Europe | Pacific | Total |
|---|---|---|---|---|
| Full | 62 | 29 | 17 | 108 |
| Affiliate | 240 | 123 | 47 | 410 |
| Total | 302 | 152 | 64 | 518 |
The first column, being the headings for each row is put into one group, the three columns giving the information is a second group while the third group is the one containing the totals. The attribute span defines the number of columns in the group in each case. In the example, the widths have been made difference to demonstrate this. Styling can be applied differently to each of the groups.
The col element shares attributes between several columns. For example:
<colgroup span="3" align="right"> <col width="30%"> <col span="2" width="10%"> </colgroup> <colgroup span="2" width="20%">
The colgroup element has defined that the first three columns form a group, not a very sensible grouping, and that these should be aligned to the right. The attribute span defines the number of columns in the group. Within that group of columns, the two col elements define the first column as having a width of 30% and the next two have a width of 10%. For the col element, the attribute span defines the number of columns that the element applies to. Finally, the second colgroup element specifies the group of two columns having a width of 20%.
-- 19 --
| Type | Americas | Europe | Pacific | Total |
|---|---|---|---|---|
| Full | 62 | 29 | 17 | 108 |
| Affiliate | 240 | 123 | 47 | 410 |
| Total | 302 | 152 | 64 | 518 |
The col element can be used by itself and appear without being surrounded by a colgroup element. The colgroup element does not add any additional functionality as it has the same attributes as the col element. It does make the description simpler and brings out any inherent structure to the table.
<tr> <th rowspan= colspan= align=> <td rowspan= colspan= align=> </tr>
Frequently in the organisation of more complex tables, it is useful to be able to have a cell in the table that spans more than one column or more than one row. The attribute rowspan defines the number of rows that the cell spans (default is 1) and colspan defines the number of columns that the cell spans. For example, to have a row in the head of the table to replace the caption, this could be done as follows:
<table title="Membership of W3C July 2001" summary="5 ..." width="90%"> <thead> <tr> <th colspan="5">W3C Membership</th> </tr> <tr> <th>Type</th><th>Americas</th><th>Europe</th><th>Pacific</th><th>Total</th> </tr> </thead>
The original table would then appear as:
| W3C Membership | ||||
|---|---|---|---|---|
| Type | Americas | Europe | Pacific | Total |
| Full | 62 | 29 | 17 | 108 |
| Affiliate | 240 | 123 | 47 | 410 |
| Total | 302 | 152 | 64 | 518 |
To make the first and last titles straddle the first two rows, the table could be changed to:
<table title="Membership of W3C July 2001" summary="5 ..." width="90%"> <thead> <tr> <th rowspan="2">Type</th> <th colspan="3">W3C Membership</th> <th rowspan="2">Total</th> </tr> <tr><th>Americas</th><th>Europe</th><th>Pacific</th></tr> </thead>
Clearly some care must be taken in designing this. The first row consists of three cells. The first cell straddles this row and the next one as does the last. The middle cell straddles three columns. When the next row is to be defined, the first and last cells have already been defined by the previous row. In the second row, only the remaining three cells need to be defined. the result is as follows:
| Type | W3C Membership | Total | ||
|---|---|---|---|---|
| Americas | Europe | Pacific | ||
| Full | 62 | 29 | 17 | 108 |
| Affiliate | 240 | 123 | 47 | 410 |
| Total | 302 | 152 | 64 | 518 |
-- 20 --
Forms in HTML are much like forms in real life. The user must fill in a set of fields on the form and post it somewhere. For the moment, we will concentrate on filling the form in rather than posting it. For commercial activities, it is usual for the form to be sent back to the server for processing. However, if the form is just designed to provide interaction between the user and the Web page, it may be sent to a script for execution locally.
Forms in real life come up in many shapes and sizes. Many forms provide text fields for you to fill in (your name and address for example. Some present you with a set of options and ask you to tick all that apply and some insist that you only tick one of the options. Most of the standard methods of form filling are provided in HTML. The form element encloses the complete form. The input element handles most of the fields on the form and the attributes of the element define what type of field it is. A separate textarea element is provided for those fields on forms where you are allowed to write essays rather than simple one line answers. The select element handles those scrolling lists where you have to select your country of origin etc. Within the list, the individual entries are handled by the option element.
So far the form is a set of input fields that are independent. By using other HTML elements such as headings and tables, the form can be made to look as though it has some structure. To aid accessibility, the form has its own internal structure. A set of form fields can be grouped together by the fieldset element and this section can be given its own title using the legend element. Also, individual fields can have a label associated with them using the label element.
To make even prettier buttons, HTML has added the button element which forms much the same as the input element when it is defining a button. However the button element can embed any block-level elements within it including images so it greatly enhances the look of the button but does not really add much content capability.
<form method= action= onsubmit= > <input name= type= value= size= maxlength= checked= src= accesskey= />
Here is a very simple example of a form:
<form method= "post" action = "mailto:bhopgood@brookes.ac.uk"> <p>Please type the mail to send to Bob Hopgood: <input name="MyMessage" type="text" /></p> </form>
This would appear something like:
The input element will be discussed in detail in the next section. The example here just expects the user to type some information into a box. The form element has two major attributes. The method attribute indicates the HTTP method to be used in sending the information and the action attribute gives details of where it will be sent. In the example, the HTTP POST method is specified and the input is to be sent to the mail address specified. The possible methods are:
In the example, if the text input is Hello world followed by a Return, this will cause the form to be submitted. The content of the form that is emailed is:
MyMessage=Hello+world
-- 21 --
If the form element had been:
<form method= "get" action = "http://www.w3.org/">
The URL requested would have been:
http://www.w3.org/?MyMessage=Hello+world
If the form has several input values, the individual inputs will be separated with &. For example:
http://www.w3.org/?MyMessage=Hello+world&MySecondMessage=From+me
It is possible to send the input results to a script associated with the web page by:
<form onsubmit="process()">
In this case the function process can access the result of the input and perform the script based on this input.
To a large extent, what happens as a result of the input is outside the control of HTML and we will not discuss it further in this Primer. The aim here is to show the types of input possible and how they can be used.
<form method= action= onsubmit= > <input name= type= value= size= maxlength= checked= src= accesskey= />
The overall form of the input element depends on the three attributes name, type and value. The other attributes are either required for a subset of the devices or the meaning varies dependent on the device:
The text device has already been shown in the previous section but here is a more complete example:
<input name ="MyMessage" type="text" value="Some text" size="40" maxlength="30" accesskey="Z" />
This would appear something like:
The device will appear with a window capable of holding the 40 characters that may be input. If more than 30 characters are input, the rest are thrown away. The initial text appears in the window. If the user does not want to use that, this should be deleted before typing is started. Hitting the character Z on the keyboard should move the focus ready to input text in the text field. This is still poorly supported in today's browsers. Also, some browsers manage to miscalcluate the width of the text field so that even when the size is the same as maxlength, the field length is not sufficient to see all the characters. On some browsers, the window can be made a scrolling one by: specifying the maxlength greater than the size. IE6 does not do this. To be sure of seeing all the characters typed:
<input name ="MyMessage" type="text" value="Some text:" size="40" maxlength="10" /></p>
This would appear something like:
-- 22 --
Attempting to type after the ten characters displayed will not cause any action.
So far, to input the text, it has been necessary to hit the RETURN key which may not be that intuitive. The submit type allows a separate device to initiate the submission. For example:
<form method= "post" action = "mailto:bhopgood@brookes.ac.uk"> <p>Please type the mail to send to Bob Hopgood: <input name="MyMessage" type="text" /> <input type = "submit" name="Mail" value = "Send Mail" /></p> </form>
This would appear something like:
Typing abc and hitting the submit button would input MyMessage=abc&Mail=Send+Mail. As the input includes the contents of the submit button, you can have sveral submit buttons and process the input text dependent on the button pressed.
Radio buttons can be defined by:
<p>Indicate your choice of music:</p> <p>Jazz: <input type ="radio" name="music" value="jazz" checked="checked" /></p> <p>Classical: <input type ="radio" name="music" value="classical"></p> <p>Folk: <input type ="radio" name="music" value="folk" /></p>
This would appear something like:
By giving the three buttons the same name they act as a single device with only one being checked at any one time. The initial setting of the device is with the value jazz checked.
Checkboxes can be defined by:
<p>Indicate all the choices of music you like:</p> <p>Jazz: <input type = "checkbox" name="music" value="jazz" /></p> <p>Classical: <input type="checkbox" name="music" value="classical" /></p> <p>Folk: <input type="checkbox" name="music" value="folk" /></p> <input type="submit" name="Choice" value="Choices Made" />
This would appear something like:
Indicate all the choices of music you like:
Jazz:
Classical:
Folk:
If jazz and folk have been set followed by hitting the submit button, the input would be: music=jazz&music=folk&Choice=Choices+Made
Buttons are aimed at providing a method for invoking local scripts. Whereas the submit button causes the form as a whole to be submitted, the push button causes a script to be launched:
<input type="button" value="Click Me" onclick="processclick()" />
This would appear something like:
-- 23 --
It is usual for the value of the click to be the name given to the button.
In the original example of text input, the situation could arise that the input provided by the user is wrong and rather than try and patch it up, the best solution is to start again. This could apply to the other input devices in the form as well. The reset button goes through the form and resets all the input devices to their initial values. For example:
<form method= "post" action = "mailto:bhopgood@brookes.ac.uk"> <p>Please type the mail to send to Bob Hopgood: <input name="MyMessage" type="text" value="Initial text" /></p> <p><input type = "submit" name="Mail" value = "Send Mail" /></p> <p><input type = "reset" name="Reset" value = "Reset Text" /></p> </form>
This would appear something like:
If the initial text is overwritten and the reset button hit, the initial text will reappear.
Sometimes the text to be input is information that you do not wish somebody else to see who is looking over your shoulder. In this case it would be preferable for the text not to be visible:
<p>Please type the mail to send to Bob Hopgood: <input name="MyPassword" type="password" value="mypass" /></p>
This would appear something like:
Please type your password:
As the user types, a different character is presented (often a *) so that the input cannot be seen. Even the initial value is masked in this way.
Frequently, the input required is the name of a local file name. In this case it would be useful if the user could browse his filestore and indicate the file to be sent rather than typing the filename. this is achieved by:
<p>Input name of picture to be processed: <input type="file" name="imagefile" value="C:\Files\test\photo.jpg" /></p>
This would appear something like:
Input name of picture to be processed:
The browser may or may not put the initial file value in the window (IE6 does not and Opera does). What it puts on the button that allows you to search the filestore is also up to the browser.
Sometimes there is a need to send some information to the server that the form designer does not input from the user to fill in. For example, a form may have been sent already from this user and the current form being input needs to be associated with that form. This can be achieved as follows:
<input type="hidden" name="from" value="frah" />
Nothing appears on the form but when the form is submitted, one of the input values will be from=frah.
Buttons can be defined by:
<p>Please type the mail to send to Bob Hopgood: <input name="MyMessage" type="text" /> then mail by pressing: <input type="image" name="Pretty" src="frah.gif" /></p>
This would appear something like:
-- 24 --
Please type the mail to send to Bob Hopgood: then mail by pressing:
What gets sent when the form is submitted depends on where the user has hit on the image. It would be something like: Pretty.x=73amp;Pretty.y=122. It is possible to associate a usemap attribute to the input element so that different responses could be made dependent on the input values provided.
<select name= size= multiple= > <option value= label= > <optgroup label= >
A frequent input requirement is to present a set of options to a user and give the user the opportunity to select one. The two elements select and option provide that facility. For example:
<p>Please make your choice of music:</p> <select name="music" size="6" multiple="multiple"> <option value = "jazz">Jazz</option> <option value = "classical">Classical</option> <option value = "folk">Folk</option> <option value = "pop">Pop</option> <option value = "world">World</option> <option value = "zydeco" selected="selected">Zydeco</option> <option value = "country">Country</option> <option value = "fado">Fado</option> </select>
This would appear something like this:
Please make your choice of music:
The select element is designed to display a menu of options to the user. The name attribute gives the name for the information to be returned and the size attribute defines the number of options that would be visible initially. The multiple attribute defines whether one or more choices are allowed to be made. If the attribute is set, holding the control key down and making a selection allows several selections to be made. The option elements appear inside the select element and each defines one of the entries on the menu. If the select attribute is set, that option will be set when the menu is displayed. In this case zydeco is set. If jazz was also selected, the value returned would be: music=jazz&music=zydeco.
For very long lists, the amount of scrolling that has to be done is excessive. The normal solution to this is to have a main menu that defines groups of menu items and moving right from a group entry gives you the sub-group. The optgroup element has been added to provide this additional single level of structure (the element cannot currently be nested). To define two groups would be provided by:
<p>Please make your choice of music:</p> <select name="music" size="4" multiple="multiple"> <optgroup label="Old"> <option value = "jazz">Jazz</option> <option value = "classical">Classical</option> <option value = "folk">Folk</option> <option value = "pop">Pop</option> </optgroup> <optgroup label="New"> <option value = "world">World</option> <option value = "zydeco" selected="selected">Zydeco</option> <option value = "country">Country</option> <option value = "fado">Fado</option> </optgroup> </select>
The initial menu is a 2-item menu displaying Old and New. By selecting one or other and moving right brings up the sub-menu with the options in the group. The element does not provide any richer input content but does improve the user interface. It is not implemented in IE6.
-- 25 --
If the length of the menu is sufficiently small that the size attribute can be the length of the options, the display is likely to be as follows (we have limited the options to 6):
Please make your choice of music:
<textarea name= rows= cols= readonly= />
If a body of text is to be input, the input element does not provide the real estate for a large volume of text providing just a single line buffer. To provide an area consisting of several lines requires the textarea element. For example:
<textarea name="para" rows="7" cols="60" ></textarea>
This would appear something like this:
A window containing 7 lines of text, each 60 characters wide is presented to the user and the user can scroll that window so that input is unlimited. To provide an initial value for the input buffer requires:
<textarea name="para" rows="7" cols="60" readonly="readonly" > This is the first row of the input buffer and this is the second line. </textarea>
This would appear something like this:
By setting the readonly attribute, the initial text cannot be overwritten. If it does not appear, it can be overwritten.
<label for= > <fieldset > <legend >
Quite a few of the input methods provide the mechanism for inputting the value but do not add any information concerning that mechanism. In consequence, normal HTML markup has been added to indicate what the input value represents. The label element specifies labels for input devices that do not have a label associated with the device. Earlier, for example, a set of radio buttons was defined as:
<p>Indicate your choice of music:</p> <p>Jazz: <input type ="radio" name="music" value="jazz" checked="checked" /></p> <p>Classical: <input type ="radio" name="music" value="classical"></p> <p>Folk: <input type ="radio" name="music" value="folk" /></p>
-- 26 --
Using a label, this could be written as:
<p>Indicate your choice of music:</p> <label><input type ="radio" name="music" value="jazz" checked="checked" />Jazz</label> <label><input type ="radio" name="music" value="classical" />Classical</label> <label><input type ="radio" name="music" value="folk" />Folk</label>
Indicate your choice of music:
The transmitted form is the same in both cases. It is possible to separate the labels from the input devices while keeping the close association between the label and the device. This is frequently the case when styling is being used to improve the layout. Here is a simple example:
<p>Indicate your choice of music:</p> <label for="jz">Jazz</label> <label for="cl">Classical</label> <label for="fl">Folk</label> <label for="fl">World</label> <input id="jz" type ="radio" name="music" value="jazz" checked="checked" /> <input id="cl" type ="radio" name="music" value="classical" /> <input id="fl" type ="radio" name="music" value="folk" />
Indicate your choice of music:
The user now has the choice of clicking on the radio button or the label itself. This is a slightly contrived example so the benefits are not that obvious in this rather simple example. Note that we have put two labels to point to the third button thus lumping world music in with folk.
For long forms with many inputs, some structure can be added to parts of the form using the fieldset and legend elements. The first groups a set of inputs and the second gives them a title. For example:
<fieldset> <legend>Personal Details</legend> <p><label><input name ="NAME" type="text" size="40" maxlength="40" />Name:</label></p> <p><label><input name ="EMAIL" type="text" size="40" maxlength="40" />Email:</label></p> </fieldset> <p> </p> <fieldset> <legend>Musical Interests</legend> <p>Indicate your choice of music:</p> <label for="jz">Jazz</label> <label for="cl">Classical</label> <label for="fl">Folk</label> <label for="fl">World</label> <input id="jz" type ="radio" name="music" value="jazz" checked="checked" /> <input id="cl" type ="radio" name="music" value="classical" /> <input id="fl" type ="radio" name="music" value="folk" /> <p>Indicate why you like this music</p> <p><textarea name="para" rows="7" cols="60" >abc</textarea></p> </fieldset>
This would be displayed something like:
A box has beeen added around the two parts of the form and the legend displayed at the top of each part.
-- 27 --
<button name= value= type= >
The buttons that can be produced using the input element are not that glamorous and, in consequence, a button element was defined to replace the input element of types button, submit and reset. In no type is defined, a submit is produced. The functionality is identical to the input element except that it can have any content that you like in defining the information to be displayed on the button. For example:
<button name="form" value="submit" type="submit" > <h2>MUSIC FORM</h2> <h3>Submit to Server Now</h3> </button>
This might appear as:
As you now have complete control over the content of the button, it can be styled in much the same way as a Web page. Here, some styling has been added to the two HTML elements just to give it a bit more of an effect. The user could add images, background colours, and whatever else is required to produce the desired effect. Most browsers have also added some goodies to the button such as giving it a shadow to give the impression that the button has been physically pressed.
That completes the description of the forms section of the Primer. It does not give the complete picture. Some of the facilities associated with scripting have been dealt with in a rather simplistic way. This is more appropriate for a detailed discussion concerning scripting. Similarly, there are a number of facilities concerning focus to enable a user with accessibility problems to easily move through a form filling in the sections. This is particularly appropriate for users with mobile phones or using a text-to-speech browser. These will be dealt with in the discussion of accessibility.
In an HTML document, the head element contains a set of elements that are not concerned with the content of the web page but give additional meta information related to styling, what the page contains, how it is related to other pages etc. The set of elements allowed in the head of the document are:
-- 28 --
<title>
The title element must be present in a document for it to be legal HTML. It serves to identify the document and it is normal for the browser to add the title to its own decoration at the top of the browser window. When a user stores a page in a bookmark (or favourites) list, the bookmark will be identified with the title of the document. In consequence, the title chosen should place the document in its context.
<meta name= content= http-equiv= scheme=/>
The meta element was a major way that a search engine locates pages although the usage is declining. Browsers are not required to support the meta element (other than to parse it) so there are few hard and fast rules about what is allowed and what it does. In many cases it may be browser or search engine specific. The two attributes name and content define a property and its value. As HTML does not define any values that the attribute name can take, it is largely up to the users of the page to decide what to do with the information. Here is an example of a typical web site aiming to get itself recognised by the search engines (http://www.equineworld,co.uk):
<html> <head> <title>Horses and ponies, UK</title> <meta name="description" content="Horses and Ponies: about horses, ponies, horse riding and equestrianism in the UK. Horses for sale, equestrian news, equine care, horse chat and more." /> <meta name="keywords" content="horses, horse, equestrian, equine, ponies, pony, horses for sale, uk" /> <meta name="robots" content="index, follow" /> <!--Horses and Ponies: about horses, ponies, horse riding and equestrianism in the UK. Horses for sale, equestrian news, equine care, horse chat and more.--> <meta name="owner" content="Equine World UK Ltd"/> <meta name="subject" content="horses and ponies, uk"/> <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"/> <meta http-equiv="PICS-Label" content='(PICS-1.1 "http://www.icra.org/ratingsv02.html" l gen true for "http://www.equine-world.co.uk" r (ca 1 lz 1 nz 1 oz 1 vz 1) "http://www.rsac.org/ratingsv01.html" l gen true for "http://www.equine-world.co.uk" r (n 0 s 0 v 0 l 0))' />
There are two main types of search engines. One set crawl the web looking at pages while the other type tries to get some human cataloguing of the information. Many of the ones used today are a mixture of both approaches. The web crawl involves looking at pages and following links to other pages. The better ones may look at all the text on a page. The worst ones start at the top and may only look at the head of the document. The meta element is one of the elements that some search engines look for. But do not assume that by using lots of meta elements you are sure to be recognised. Search engine spamming is a problem. if you add words to get your pages listed in a search, the search engines will detect that so make sure any words you use in the meta information are appropriate. For a commercial web site, it is essential that for a relevant query, the site gets ranked in the top 10. If you land up on the second page, you are close to death! So target keywords important for your site must be in the title of the page and in the meta information. Search engines have realised that meta information can be rigged and so the use of this information by search engines is getting less and less important. For example, neither Google or Alta Vista use meta element information to rank web pages any more. On the other hand, Inktomi does. Most read and use the information in the meta elements in some way. The two most important properties defined by the name attribute are description and keywords. If the search engine believes the page is honest, it will use the description in place of one that it might try to derive from its own activities. In the example above, the description aims to give a good overview of the whole site. It is clear that the site is about both horses and ponies, that people in the UK are most likely to be interested, that you can buy a horse there and so on. The keywords property tries to guess the kinds of searches that people who need to find this site might type.
The robots property is primarily aimed at stopping robots visiting pages of no value If the content is set to noindex robots will not visit the page. In the example, the author has requested a robot to index the page and follow any links on the page. You may wonder why the author has put a comment containing all the keywords after the meta information itself. As some search engines do not look at meta information and believe information near the top of the page is most important, adding the comment effectively acts as a meta element for those search engines.
Some of the other values of the name attribute are:
-- 29 --
It is difficult to be precise as anybody is allowed to dream up values and there usefulness is dependent on industry using them as de facto standards. A set of standard terms would have made the meta element more useful.
A good site for further information about search engines and what they do is http://www.searchenginewatch.com/.
The attribute http-equiv effectively defines an HTTP header which controls the action of the browser receiving the page and is used to refine the information provided by the actual headers in the HTTP. Some servers translate the meta information into actual HTTP headers automatically. In the example above, the meta elements define the content type of the page and its PICS label (this will be covered in the discussion of metadata). Other possibilities might be:
<meta http-equiv="expires" content=" Wed 28 Feb 2002 23:59:59 GMT" /> <meta http-equiv="Pragma" content="no-cache" /> <meta http-equiv="Refresh" content="0;URL=http://www.newurl.com" /> <meta http-equiv="Set-Cookie" content="cookievalue=xxx;expires=Wednesday, 28-Feb-02 18:12:20 GMT; path=/" />
The first states when the page is no longer valid. The second stops Netscape caching the information. The third states when the page should be refreshed and the fourth defines a cookie. Fuller details of what is possible is defined in HTTP.
The scheme attribute has been added as a qualifier on what the meta information means. For example:
<meta scheme="ISBN" name="identifier" content="0-201-17805-2" />
The meaning of content, is as defined by ISBN classifications scheme. By knowing this, the meta information can be correctly interpreted as the book Raggett on HTML 4.
<script type= src= > <noscript >
To a large extent, scripting is outside the scope of this Primer. Suffice is to say that scripts can be placed in either the head or the body of the HTML document. Several scripts can be provided at different points in the page but effectively all the scripts are put together as one script associated with the page. Putting the script in the head of the page is more elegant but may not be the best place. Remember that search engines do not normally read the whole page. In consequence if you have a long script associated with your page, it is possible that all the search engine reads is a load of non-English rubbish as far as it is concerned. An example use is:
<script type="text/vbscript" src="http://www.w3.org/script/script.js" ></script>
Here the script is defined external to the document. The type indicates the MIME type of the script
<script type="text/javascript"> . . . </script>
In this case the script is enclosed between the start and end tags. The body element has an attribute that can cause the script to be executed when the page is loaded or unloaded:
<body onload="processscript()"> <body onunload="processscript()">
Most elements have the ability to define a script invocation on being clicked or moved over. For example:
<h1 onclick="ps()">ABC</h1> <p onmousedown="ps()">ABC</p>
The complete set of events are:
-- 30 --
<style type= >
The style element defines a stylesheet to be used by the document in much the same was as the script element defines a script. An example is:
<style type="text/css" media="print"> . . . </style>
The type attribute defines the type of stylsheet following and the media attribute indicates to which media it applies.
<link href= type= media= hreflang= charset=/>
The link element has been available in HTML almost since the beginning. the intention was that the element would give information concerning how this page relates to other pages. As two pages can each link to the other, it is not always obvious which is the parent and which is the child. The link element was designed for that purpose and the hope was that search engines would use the link information to improve search relevance. To a large extent it is not used in that role. However there are some specific uses of the link element:
<link href="../primerstyle.css" rel="stylesheet" type="text/css" media="print"/> <link rel="alternate" hreflang="pt" type="text/html" /> <link rel="alternate" hreflang="ar" charset="ISO-8859-6" type="text/html" />
The first example shows a link to an external stylesheet to be used when the document is printed. The second gives an alternative to the current page in Portuguese. The third links to an alternative page in Arabic. In this case, the character set needs to be defined also. Search engines may offer the linked pages when the user as requested pages in a specific language.
<base href= />
The base element provides the base for relative URLs. For example:
<base href="http://www.w3.org/nuts/index.htm" /> . . . < a href="../brazil/origin.htm" />
This would take the relative URL and, as the base element
is declared, the relative URL will be resolved to:
http://www.w3.org/brazil/origin.htm
by effectively treating the current page as though it were the one defined
by the base element.
-- i --
There are some useful Web sites and books relevant to HTML:
This Appendix lists all the elements available with the following information:
The content allowed within an element has the following abbreviations:
Two elements are allowed anywhere script, noscript
| Element Name | Type | Attributes | Allowed in content |
|---|---|---|---|
| a | inline | charset type name href hreflang rel rev accesskey shape coords tabindex onfocus onblur | [inline] but not a element |
| abbr | inline | [inline] | |
| acronym | inline | [inline] | |
| address | block-level | [inline] | |
| area | shape coords href alt | [empty] alt is required | |
| base | head | href id | [empty] only one allowed |
| blockquote | block-level | cite | [block-level] |
| body | block-level | onload onunload | [block-level] |
| br | inline | id class style title | [empty] |
| button | inline | name value type disabled tabindex accesskey onfocus onblur | [block-level] [inline] |
-- ii --
| Element Name | Type | Attributes | Allowed in content |
|---|---|---|---|
| caption | [block-level] [inline] | ||
| cite | inline | [inline] | |
| code | inline | [inline] | |
| col | align span width | [empty] | |
| colgroup | align span width | col | |
| dd | block-level | [block-level] [inline] | |
| dfn | inline | [inline] | |
| div | block-level | [block-level] [inline] | |
| dl | block-level | dt dd | |
| dt | block-level | [inline] | |
| em | inline | [inline] | |
| fieldset | [block-level] [inline] form legend | ||
| form | block-level | action method enctype onsubmit onreset accept accept-charset | [block-level] form not allowed inside |
| h1 to h6 | block-level | [inline] | |
| head | head | lang xml:lang dir xmlns | script style meta link object title base (Only one title and base) |
| hr | block-level | [empty] | |
| html | root | lang xml:lang dir xmlns | head body |
| img | inline | src alt longdesc height width usemap | [empty] |
| input | inline | name type value size maxlength checked disabled readonly src alt usemap tabindex acceskey onfocus onblur onselect onchange accept accesskey | [empty] |
| label | for accesskey onfocus onblur | [inline] label | |
| legend | accesskey | [inline] | |
| li | block-level | [block-level] [inline] | |
| link | head | charset href hreflang type rel rev media lang xml:lang dir xmlns | [empty] |
| map | id | [block-level] id attribute is required | |
| meta | head | http-equiv name content scheme id lang xml:lang dir xmlns | [empty] |
| noscript | block-level inline | [block-level] | |
| object | inline | data type standby height width usemap | [block-level] [inline] param |
| ol | block-level | li | |
| optgroup | disabled label | option | |
| option | selected disabled label value | text only |
-- iii --
| Element Name | Type | Attributes | Allowed in content |
|---|---|---|---|
| p | block-level | [inline] | |
| pre | block-level | xml:space | [inline] (Not allowed: img object sub sup) |
| q | inline | [inline] | |
| samp | inline | [inline] | |
| script | head block-level inline | charset type src defer xml:space | only text |
| select | name size multiple disabled tabindex onfocus onblur onchange | optgroup option | |
| span | inline | [inline] | |
| strong | inline | [inline] | |
| style | head | type media title xml:space | text only (See CSS for text syntax) |
| sub | inline | [inline] (Not allowed in pre) | |
| sup | inline | [inline] (Not allowed in pre) | |
| table | block-level | summary width border frame rules cellspacing cellpadding | caption col colgroup thead tfoot tbody tr (caption followed by col and colgroup elements in any order, then thead, tfoot and tbody or tr elements if tbody omitted) |
| tbody | tr | ||
| td | align rowspan colspan valign | [block-level] [inline] | |
| textarea | name rows cols disabled readonly tabindex accesskey onfocus onblur onselect onchange | text only | |
| tfoot | align valign | tr | |
| th | align rowspan colspan valign | [block-level] [inline] | |
| thead | align valign | tr | |
| title | head | lang xml:lang dir xmlns | text only (only one allowed) |
| tr | align valign | td th | |
| ul | block-level | li | |
| var | inline | [inline] |
-- iv --
This Appendix lists the deprecated attributes and the elements to which they apply. These will still appear in HTML documents but their function can be achieved by styling. Future versions of HTML are likely not to support them.
| Element Name | Deprecated Attributes |
|---|---|
| address | align |
| blockquote | align |
| body | background text link vlink alink bgcolor |
| br | clear |
| caption | align |
| col | align |
| colgroup | align |
| dd | align |
| div | align |
| dl | align |
| dt | align |
| form | align |
| h1 to h6 | align |
| hr | align noshade size width |
| html | version |
| img | hspace vspace border align |
| input | align |
| legend | align |
| li | type value compact |
| noscript | align |
| object | hspace vspace border align |
| ol | type start compact |
| p | align |
| pre | width |
| script | align language |
| table | align bgcolor |
| tbody | |
| td | nowrap width height bgcolor |
| tfoot | |
| th | nowrap width height bgcolor |
| thead | |
| tr | bgcolor |
| ul | type compact |
