Programming in Perl

Homework 3

Due Tuesday, March 4, 2003 11:59:59pm EST

Check the FAQ frequently

Write a perl program that searches an RCS user's public_html directory for a list of HTML files. You will do this by first prompting the user to enter the username to search for. Once you have this name, you should attempt to open the user's public_html directory. Inside this directory, obtain a list of each of the files that end in .html or .htm.

For each HTML file in the directory:

Print the name of the file to standard output
Print a list of all HTML hyperlinks. For each one, print both the text that the user would click to activate the link, as well as the destination of that link (see sample output for an example). The list should be formatted so that it is obvious which text goes with which address. Also, you should strip newlines from the addresses or text and replace them with space characters, so that the output is on one line per address/text pair.
Create a new file in the current working directory (ie, the directory from which your perlscript was run). This file must have the same name as the current HTML file, and should have the same contents (including formatting). The difference is that your new file will have all instances of "Rensselaer" and "Rensselaer University" replaced with "RPI". Note that you must not replace any portion of the phrase "Rensselaer Polytechnic Institute".

As you are doing this assignment, be sure to keep several things in mind:

On RCS, a user's home directory is found by placing a tilde (~) in front of the user's id. So my home directory is at ~lallip/
Perl cannot natively open a directory whose name contains a special Unix wildcard character, like a tilde. (Recall the lecture from 2/12/03)
A user id might not necessarily exist in the RCS system
A user might not necessarily have a public_html direcotry, or the directory might not be readable
It's possible the current working directory is not writable to the user of your program
If you don't know anything about HTML, you should review the HTML Basics slide on the main course webpage.
All HTML tags are case insensitive
An HTML hyperlink might have quotes surrounding the address, but it is not required*
An HTML hyperlink tag might have additional attributes beyond the href= attribute. Your program must print out only the actual address of the link
Whitespace is optional between href and the = sign, and between the = sign and the link address.
HTML tags can span multiple lines
Your program must be able to handle both absolute links (which start with "http://" or "ftp://" or some other protocol name), as well as relative links (which are simply a filename (and possibly a directory name))
A single HTML document may contain multiple links to the same address (with the same or different clickable text)
A single HTML document may contain multiple links with the same clickable text (with same or different addresses)

Grading Criteria

Prompt for and read in RCS Id	5
Open user's public_html directory	7
Get a list of all HTML files	15
Print each file's name	5
Print a list of all HTML hyperlink addresses	15
Print the corresponding list of HTML hyperlink clickable texts	10
Create local copy of each HTML file	8
Make all correct substitutions in local copy	15
Compilation	5
Error Checking & Reporting	5
Code Style	5
Output Style	5

As always, 5 Above & Beyond points are available for relevant contributions beyond the requirements.

Submission Instructions

You may again use either the submission CGI program at http://cgi2.cs.rpi.edu/~lallip/cgi-bin/perl/spring03/submit.cgi or the RCS program ~lallip/public/submit.pl Please do not use the opposite method for any re-submissions without emailing me first.

Here are some files for you to use: First Sample HTML file, Second Sample HTML file, Sample Standard Output, Local copy of Sample 1, Local copy of Sample 2. Note that your local copies should have the same file name as the original HTML files.

The program must work on rcs-sun1.rpi.edu. It is due at 11:59:59pm EDT on Tuesday, March 4, 2003.

* That is, quotes are not required for the vast majority of browsers to recognize a link. I am not concerned with what the official HTML or XHTML standards say about the matter.