URLs   Uniform Resource Locators

Absolute URLs

http://	cs.wellesley.edu/	~cs110/lectures/L02/	URLs.html
protocol	server.domain	path	filename

let's break it down in its components:

• http:// This is the protocol, which is almost always HTTP, the HyperText Transfer Protocol. You'll also occasionally see FTP, the File Transfer Protocol; HTTPS, the secure version of HTTP; and others.
• cs.wellesley.edu/  This is the server: a machine connected to the internet that hands out files to anyone who asks. cs is the name of the machine; wellesley.edu is the domain. The name of the server combined with the domain must be unique in the world. In fact, the domain must be unique in the world, and the server name must be unique within the domain.
• ~cs110/lectures/L02/  This is the path: a sequence of directories (AKA folders), separated by slashes, that indicates where on the server the file is located. In this path, there's a tilde (~), which means to start at the public_html directory under the home directory of the cs110 account. We'll always be using that kind of URL in this class, so keep this special rule in mind.
• URLs.html  This is the file that is being requested.

Because URLs can specify files on many different kinds of computers, you should stick to a lowest common denominator when naming your directories and files, to avoid confusing machines and browsers. The lowest common denominator means:

• use letters and digits, but no special characters like /, *, #, &, and so forth
• no spaces (this is the most common error for Mac and PC users). For example, if you want to have a web page named cats and dogs.html you're much safer with cats-and-dogs.html or cats_and_dogs.html

Furthermore, be careful with upper and lower case. To your eyes, cats.html and Cats.html may seem like the same filename, but to a computer, they're as different as cats.html and dogs.html. The same rule is true of directory names: they are case-sensitive. Many people like to stick to all lower-case, just so they don't have to remember. .

Certain parts of the end of an absolute URL can be omitted. For example, the filename can be omitted and (usually) defaults to a file named index.html in the specified directory (for an example, see our course homepage). Similarly, the path can be omitted, in which case the page is in the top directory of the web server. However, the protocol and server must be specified. You can't omit those. If a URL has a protocol, it's an absolute URL, otherwise it'll be treated as a relative URL.

Tree Structured Directories

As you know, most operating systems (including Mac, Windows, Unix, iOS, Android, etc.), organize the contents of a computer's hard drive into folders, also called directories. Folders and files form a tree structure, because of the ability of directories to contain other directories.

Here's an example:

For a website about Joss Whedon, we might have a folder about his Buffy the Vampire Slayer TV show, another about the TV show Angel, and a document about his comics. These might all be in a top-level folder called Joss. (The top-level folder is also called the root folder.) The folders about the TV shows might each have a sub-folder called cast. And so on.

While it's easy to conceptualize a tree of folders and files, also called a directory tree, there are many different ways to draw it. Here are several ways of depicting the Joss directory tree described above that you will see in this course:

For example, we can see that buffy.html and willow.html are both files in the same directory, the one called cast, which is, in turn, one of three things in the directory called Buffy, which is in the directory called Joss. Joss contains two directories (Buffy and Angel) and a file called comics.html and so on. Take a minute to be sure you understand the relationships among the directories and files. Notice that there are two files named angel.html; this is okay, because they are in different directories.

Relative URLs

We can use these relationships to form a shorthand way of specifying a URL. This is called a relative URL, because we specify the location of a file relative to a known file. Here are the basic rules:

• Relative URLs are only for files on the same server. Therefore, we can leave off the protocol and server name.
• If a file is in the same directory as the HTML file containing the link, we just give its name. If we want to put a link in the file buffy.html linking it to willow.html, we can say:
  <a href="willow.html">her best friend Willow</a>
• If a file is in a sibling subdirectory, we can just give the path starting from the sibling. So, if we are in the file called mythology.html and we want to give a link to the file angel.html, we can say:
  <a href="cast/angel.html">Angel, the vampire with a soul</a>
• If the file is in a directory that contains a subdirectory that contains the HTML file with the link, we have to go up the tree. To go up, we write ../. That's dot, dot, slash. For example, if we are in the file host.html and we want to make a link to season3.html, we can say:
  <a href="../season3.html">the Host returned in season 3</a>
  That last step, with the .. is tricky: to go up, we write ..; while to go down we have to indicate which subdirectory we mean. In either case, we separate these up/down steps with a slash. Here's another example: suppose we want to put a link from buffy.html to comics.html. In this case, we can write:
  <a href="../../comics.html">another layer in the comics</a>

Fragments

The model above establishes a one-to-one correspondence between a URL and a file (web page) somewhere in the vast Internet. Sometimes, however, you want to address (point to) a particular location within a page. Such locations are called fragments.

While fragments can be used in many ways, one common use is to have a table of contents at the top of a page, allowing readers to skip down to the particular section they are interested in. An example of such a table of contents is at the left side of these notes.

Fragments are addressed by extending the URL by adding a # character and the id attribute of the fragment (the HTML element that refers to the fragment). Thus, we have:

http://	cs.wellesley.edu/	~cs110/lectures/L03-html/	URLs.html#	relative
protocol	server.domain	path	filename	fragment id

One key step is marking the destination of the URL, by naming the fragment, just as we name a file or folder. This is done with the id attribute, which can be applied to every HTML tag. We typically add it to the header tag (such as h2) of the section we want to skip to. For example:

Given the code above, we can then point to that section from any web page anywhere, using an absolute URL, like the one in the table above. We can also use fragments with relative URLs. For example, to point to that section from another file in the same directory, we can just give the filename and the fragment name: URLs.html#relative. In fact, if we want to point to it from within the same file, we can omit the filename and the relative URL becomes just #relative. You can see examples of that kind of URL in the table of contents on the sidebar.