by Jim O'Donnell
Creating an accurate document head is the first step to writing good HTML. Fortunately, it's also the easiest. The head section of an HTML document precedes the main content of the document. Similar to the banner page of a magazine, the head provides information for both the viewer software and the end user.
This chapter answers the following questions:
HTML documents are platform-independent, meaning that they don't conform to any one system standard. If they are created properly, you can move home pages to any server platform, or you can access them with any compliant WWW viewer. One way to indicate this independence is the <HTML> tag. Because HTML documents are not compiled (or processed) for execution, some applications need a hint to know how to interpret the plain text in a home page. That's where the <HTML> tag comes into play.
Remember that the <HTML> and </HTML> tags should be the first and last elements of your HTML document. Although most viewers can handle a home page without the <HTML> tag, it is recommended that all your HTML documents use it (see fig. 5.1). The end of the HTML container is defined with the end tag </HTML>.
Figure 5.1 : Using the <HTML> and </HTML>
tags to open and close your HTML documents is a good idea.
| Note |
Technically, the <HTML> tag is not part of the head section, as it contains all the HTML portions of the current document, including the head section. But, for purposes of clarity, the tag is presented here, where users begin to write their HTML code. The tag's closing component, </HTML>, comes at the very end of the document, like the traditional "The End" at the end of a book or movie. Logically, the closing is unnecessary (after all, if the file contains no more text, the document is ended). But, as a matter of good usage, take the extra second or two to include the </HTML> line. |
Files without the <HTML> tag can be misinterpreted as text-only documents, and the markup tags as just more text on the page. This fact is particularly relevant as other applications increasingly access existing HTML documents without the presumption that the document is HTML and not a plain text file (mail and news readers, for instance).
The head section is like a quick reference for WWW viewers and other applications that access HTML files. The head supplies the document title and establishes relationships between HTML documents and file directories. The document head can signal the WWW viewer to use its search capabilities to index the current document.
HTML provides the HEAD element to define the head section in a document. The <HEAD> tag encloses or contains the head section (which is enclosed by the <HTML> tag). The closing </HEAD> tag sets the bounds for the head section. The only element in the head section displayed by the end user's viewer is the value of the TITLE element. Figure 5.2 shows a typical document head.
Writing proper document heads is not only good HTML, it also prepares
your documents to be used by additional applications (such as
WAIS searches) and other future, undefined uses.
| Tip |
It's not a head element, but it's a good idea to include a comment, enclosed in "<!-" "->" to explain, describing the document in the head section. |
Using the TITLE element is as simple as it sounds-the TITLE element
"names" your document. The title doesn't assign a file
name to a document; it defines a text string that is interpreted
as the HTML title of the document. The actual file name is incidental
(thankfully); most file systems either limit the number of characters
in a file name or limit the use of "special" characters
that are required by the system (such as the / character). In
HTML titles, any character can be displayed.
| Caution |
The HTML character set does reserve some characters for special uses, such as the "less than" and "greater than" angle brackets. However, you can display these characters in your software viewer by using their HTML "entity" equivalents. If you try to use the special characters as normal, the viewer software either ignores them or displays the rest of the document's body text in unexpected (and unwanted) ways. |
Many Windows-based viewers display the TITLE text in a title bar, or at the top of the document (see fig. 5.3).
Figure 5.3 : Windows viewers display the text in the viewer's interface.
HTML doesn't limit the length of the TITLE element. However, before you rush off to give your documents voluminous and wonderfully expository titles, consider the space where the title is displayed (the viewer's title bar or window label). A good rule of thumb for the length of a title is no more than a single phrase or no longer than 60 characters. See figure 5.4 for an incorrect use of TITLE.
| Note |
When a user adds your document to his or her viewer's "hot list" or bookmark list, the TITLE value is saved as the name of your document. Avoid nondescript TITLE values, such as "Page 1," for documents likely to be linked to. Or play with your audience's expectations by providing a tantalizing TITLE for the link. |
| Troubleshooting |
I put a TITLE statement in the head section, but some people complain that their viewers display something else. What's happening?
You probably made a mistake in your document's head section, either leaving off an angle bracket or forgetting the closing tag </TITLE>. Although some viewers try to catch these errors and display what they think the author intended, others don't. Viewers can display all sorts of nasty text with a TITLE error. Go back and double-check your code, or use an HTML validation service, such as the WebTechs HTML Validation Service, at one of the following URLs: http://www.webtechs.com/html-val-svc/index.html |
Although viewers have a limited capability to display a document's TITLE value, by combining the TITLE text with a lead heading statement, you can effectively create a "1-2" punch with your introductory text. This approach can provide a way of including a longer title for your document, including the longer title within the document as the lead heading and a shorter version as the actual title (see fig. 5.5.)
Computer files are glorious things: small, lightweight, easily transportable. With a few keystrokes, you can relocate entire directories of files, or files with similar names or extensions. Reorganizing a hard disk of files or creating copies on a different system doesn't take a great deal of work (or knowledge). And making havoc of an orderly file system doesn't take any effort at all.
As the volume of HTML files under your management increases, you'll be thankful for two elements HTML uses for document heads. These tags serve to connect HTML documents to each other and to their authors.
HTML documents often rely on the physical locations of other HTML files. A document might include a pointer to another document, for instance (see fig. 5.6).
The HTML <BASE> tag acts somewhat like the DOS PATH statement; it provides an additional file directory location for the WWW viewer to refer to when looking up a document link. By specifying a value for <BASE> in your document head, you can shorten the URL statements by using relative URLs in your document's anchor and image links. <BASE> protects relative URL links in the document from "breaking" should the file be physically moved. Figure 5.7 demonstrates a proper BASE statement.
If no BASE value exists in a document, the WWW viewer assumes that relative URLs derive from the current directory of the HTML document.
Often one problem associated with managing a growing volume of
HTML files is determining which files belong together or who is
the proper author of a file. Losing track of files is very easy
when a single home page can use an unlimited number of file links
(and these files can be local or on a remote server). Using LINK,
you can easily solve these problems.
| Note |
Authorship on the Internet is a sticky issue. The WWW and other Internet applications make retrieving and reusing documents easy. One manner of protection is to include a LINK reference to the original author or to the original document (or documents) from which the file is derived. LINK, combined with a text statement in the document's body, provides as much copyright protection as the Internet currently allows. |
LINK statements define relationships between the current document and other documents, the author, or Web clients. They generally include a hypertext reference in the form of a URL and an attribute value that explains what the document's relationship with this URL is. Refer to appendix A, "HTML Tags," for more information about LINK attributes.
A document can include multiple LINK statements using as many attributes as necessary (see fig. 5.8). These attributes are shown in table 5.1.
Figure 5.8 : LINK options and attributes apply to the
entire HTML document.
| Attribute | Function |
| HREF | Points to a URL |
| REL | Defines the relationship between the current document and an HREF value |
| REV | Like REL, defines the relationship between the HREF value and the document (the opposite association) |
| NAME | Defines a link from an anchor or URL to this document |
| URN | Defines a Uniform Resource Number for the current document |
| TITLE | Functions the same as the <TITLE> tag in the head of the associated HREF |
| METHODS | Provides a list of functions supported by the current document; how it can be used by a viewer |
In this example, the LINK statements are performing three tasks. The first statement
<LINK HREF="hockey.html" REL="precedes">
tells the Web viewer that the current document (index.html) comes before the identified URL document (hockey.html).
The second statement
<LINK HREF="hockey.html" TITLE="Hockey Stuff">
identifies the title (Hockey Stuff) for the specified document (hockey.html).
The third statement
<LINK HREF="mailto:odonnj@rpi.edu" REV="made">
says that the author of this document (REV="made") is described at the following hypertext reference-in this case, an e-mail window that allows you to send a message to the author, odonnj@rpi.edu.
Of the attributes listed in table 5.1, HREF, NAME, REL, and REV are most often used. As HTML documents begin to be used by more applications, these values and attributes will become important to assist programs in using HTML documents.
HTML documents can be long and complex. Searching for specific information in these documents is a tedious job, especially when the terminology you're looking for varies. What you need is a simple method for doing a difficult job, and in HTML, where there's a need, there's often a solution (or two).
Consider the example of an HTML document that lists classical composers and their works with associated music data files and a brief synopsis for each work. Searching such a document for a specific musical composition, or an obscure composer, could take some time. What you want is an efficient way to retrieve this information (especially if you're providing this document for wide use and you want people to come back for more).
HTML provides the ISINDEX element for just such a need. ISINDEX
signals the WWW viewer to use its internal capabilities to generate
a simple search form, where the user enters one or more search
variables (separated by commas) in a blank line and clicks the
Search button. (The viewer can still view and read the document
normally if a user doesn't want to perform a keyword search.)
The viewer passes the search information to the document's server,
which performs the search.
| Note |
Having the ISINDEX element in a document doesn't guarantee that the document can be searched. WWW viewers don't have the means to perform their own searches, so they rely on the document's server to have a "search engine" program. For this reason, most Web sites prefer to run a script that appends the ISINDEX element automatically to its HTML documents if a search engine is available. |
ISINDEX requires no additional information or attributes-just add it to the head section of a document you want to make searchable.
Figure 5.9 shows how ISINDEX is included in the head of an HTML document used for searching a long HTML document of classical music. Figure 5.10 shows the resulting search form at the bottom of the Web page; entering text into the form and pressing Enter (or clicking the Submit button that some viewers provide) begins a search for the next occurrence of the text string.
Your Web server may not have a search engine to make ISINDEX a useful tool. An HTML document author, with sufficient time and desperation (and a Web site administrator who can't provide the necessary search program), can create a "rolodex" or "organizer" effect in a document. This effect is possible using anchors. Figure 5.11 shows an anchored index in a document.
This search function works using the HTML ANCHOR element; by defining each letter as an anchor link to a named anchor, the user can click that letter and jump immediately to the specified point in the document.
For example, in figure 5.11, the list of letters would begin like this in HTML:
<A HREF="#A">A</A> <A HREF="#B">B</A>
And so on. Clicking the highlighted B in the viewer window would jump to the line of HTML in the document that includes the following named anchor:
<A NAME="B">BACH</A>
The HTML specification includes a mechanism to include other meta-information,
information about the document, beyond the things such as title
and base that have defined head section elements. This mechanism
is the META element, which you can use to embed specialized information
into the document header. The <META> element has the three
attributes shown in table 5.2.
| Attribute | Function |
| HTTP-EQUIV | Binds the META element to an HTTP response header |
| NAME | Names a property such as author, publication date, or similar. If the NAME element is not specified, it is assumed to be the same as HTTP-EQUIV |
| CONTENT | Supplies a value for a named property |
Suppose, for example, that the document contains the following META elements:
<META HTTP-EQUIV="Expires" CONTENT="Thu, 01 Feb 1996 00:00:00 GMT"> <META HTTP-EQUIV="Reply-To" CONTENT="odonnj@rpi.edu (Jim O'Donnell)"> <META HTTP-EQUIV="Keywords" CONTENT="before, after">
Then if you view it and display the document information, you see the information shown in figure 5.12 (notice that the HTTP server converted the expired time from GMT to local time). The HTTP server on which I have my home page supports the Expires HTTP-EQUIV attribute of the META element but not the Keywords or Reply-To attributes. Unsupported attributes are ignored by the server.
When no equivalent HTTP response headers are available, you should use the NAME attribute instead of HTTP-EQUIV. Examples of this use of the META element are as follows:
<META NAME="Last Validated" CONTENT="Mon, 01 Jan 1996 09:23:12 GMT"> <META NAME="Web Page Type" CONTENT="Personal"> <META NAME="Special Features" CONTENT="None">