You didn't specify the operating system, so I'll assume you're running a decent UN*X server. Let's handle this in chronological order of service:

• To watch the FTP directory, use http://www.tripwire.com or a similar trigger mechanism. Otherwise, you'll need to set a cron job to read the directory once in a while which I consider polling and wasteful.
• To parse the pipe-delimited text, take the easy route and use DBD::CSV which has a variable that you can set to the pipe. That way, you open up the world of SQL to yourself. If you imagine that that is overkill, a simple line-reading-spilt-on-pipes 3-liner would do fine.
• To output the results of the parsing in the HTML table, use CGI. This will give you helpful functions to clean up your code that you may be writing in raw HTML and make sure that the end result is valid HTML.

Good luck!

AgentM Systems nor Nasca Enterprises nor Bone::Easy nor Macperl is responsible for the comments made by AgentM. Remember, you can build any logical system with NOR.