# to read from the file
open (FILE,"</myfile");
flock FILE,1;
while (<FILE>){
($tmp_state,$tmp_job) = split /\|/,$_;
push @job, $tmp_job;
}
# sort @job or @state how you would like
# Print html header
foreach $tmp_job(@job){
# Print html with @job for a value and 
}
# Print html footer
[download]

PS A good way to learn how to write a program is to write all the steps you need to do in comments. Then you can write the code that will do what you wrote in your comment.

#open file
#lock file
#read user input
#write user input to file
#close file

#open file
open (FILE,"</myfile");
#lock file
flock FILE,2;
#read user input
$input = <STDIN>;
#write user input to file
print FILE "$input";
#close file
close FILE;
[download]

it is very unlikly "|" will be found in a text file

Relying on unlikelihoods is a dangerous programming practice, especially when you have a concern for security. You'd be much better off escaping the string or validating the input to guarantee your delimiter can't be found within your data. However, this can be a significant pain, so a good compromise is to use a multi-character delimiter instead of a single character, say '|||'. This is a minimal change which does a lot more to protect your data.

I wish I had saved one of todays votes.. (oh well, tomorrow is another day).
In addition to c-era's reply, I would like to add that you could also write a Perl program to read your customers' email messages and put the data into the appropriate text file(s), thus saving you considerable effort each time new data arrives.
This is assuming, of course, that there is a somewhat consistent template or format in which you receive your data.

In addition to c-era's reply, I would like to add that you could also write a Perl program to read your customers' email messages and put the data into the appropriate text file(s), thus saving you considerable effort each time new data arrives.
This is assuming, of course, that there is a somewhat consistent template or format in which you receive your data.

Yes, this is also what I would like to do. Could I just read the e-mail then start processing from there with out putting into a textfile first? Another thing that I didn't make clear above was that the two main files (sorted by state & job) are already built, and the new information will be added as a row in a pre-existing table. Perhaps I could post urls of sample pages, and a sample e-mail (which comes from a database, so they're all the same).

andy j.

For extracting the data from the HMTL files, you should be looking at HTML::Parser or one of it's subclasses (probably HTML::TokeParser).

If you want most of the flexibility of a database but without the iverheads, why not store the data in a set of CSV files and use DBD::CSV to access it.