You may assume that you will only see the following HTML tags:

singletons:

   <br>
   <li>
   <!-->
   <hr>

opener/closer pairs:

   <html> ... </html>
   <head> ... </head>
   <title> ... </title>
   <body> ... </body>
   <center> ... </center>
   <h1> ... </h1>
   <h2> ... </h2>
   <h3> ... </h3>
   <h4> ... </h4>
   <b> ... </b>
   <i> ... </i>
   <pre> ... </pre>
   <code> ... </code>
   <sup> ... </sup>
   <sub> ... </sub>
   <p> ... </p>
   <a ... "filename"> ... </a>
   <ul> ... </ul>
   <ol> ... </ol>
   <dl> ... </dl>
   <dd> ... </dd>
   <dt> ... </dt>
   <table> ... </table>
   <tr> ... </tr>
   <td> ... </td>
   <th> ... </th>

Those of you familiar with HTML may think of the <p> tag as being a singleton. Although it is often used this way, it does have a closer form as well, and all examples provided to the class will have the opener and closer.
You should assume that each tag may have additional arguments that come after the name of the tag (given above) but before the closing >. These can be ignored for every tag other than the anchor (<a href = "filename">) tag, from which you need to extract the filename.
You may assume that all tags are well-formed. For example, there will never be a tag with < and no >; anchor tags will always have a filename in double quotes; etc. The only error that may occur is that tags are not properly nested.
For simplicity, you may use fixed-sized arrays to store things, just make sure to pick a size that seems like it will be large enough for common cases. For example, if you are using an array implementation of a List, Stack, or Queue, you may use a non-resizing implementation if you like -- just be sure to use a size that's big enough (you'll have to decide what "big enough" is for each one you use). You may also assume that all filenames and tags are a fixed size in order to avoid dynamically allocating character arrays.