If each record has a consistent separator you can set $/ equal to it,
which changes the default value of $_ (normally a newline). I think
the value of $/ is normally thrown out in these cases so setting it
to the email might not work.
#!/bin/perl
$recordseparator = "########"; # or whatever your separator is
$/ = $recordseparator;
while (<STDIN>)
{
my @array = $_; # this includes entire record now, not just a line
foreach my $line (@array)
{
# process lines of record, match regexs, store variables, etc
}
# construct your mail message, send it off
} # on to the next loop iteration (next email)
Or you could make it two tasks - if they are fixed length you could
use unix "split" to splinter them into separate records, or you could
use a really simple perl command to just open a new file (and close the
previous one) whenever it reaches the new email address. Then write
a perl script to process one record and call it on every file.
lots of ways. Some perl guru out there could probably give you a
one-liner.
Curt
On Mon, Apr 03, 2000 at 03:42:12PM -0700, Bob Thompson wrote:
> umm, here's my problem.. I think I have a resonable facsimile of a
> script, but still have some questions.. I could send it along.. having
> only about 20 lines.
>
> thanks much for any replies
>
> bob
>
> ---------- Forwarded message ----------
> Date: Mon, 3 Apr 2000 14:55:23 -0700
> From: EFN Member Service <[EMAIL PROTECTED]>
> To: Bob Thompson <[EMAIL PROTECTED]>
> Subject: Re: Bash
>
> On Mon, Apr 03, 2000 at 02:39:56PM -0700, Bob Thompson wrote:
>
> >Patrick..
> > Having nothing at all, to do with 'bash'.. but your message
> >arrived at the right time. Do you do Perl? Can I ask you a few
> >questions about it, if so? I'm trying to write a little script to parse
> >a file into a bunch of mail-messages.. this file has 100+ reports, done
> >on data submitted by individuals to a survey, each preceeded by the e-mail
> >address of the person. The reports are a fixed length. I figured it was
> >a natural for Perl, but don't know that much about it. ..in particular
> >how can one read the file, line by line, extract the address and put it
> >into a To: line, provide the remainder of information, and have it loop
> >through the rest of the report, dumping into a new message as it goes,
> >finally sending it off..!
> >
> >bob
> >
>
> My knowledge of Perl is very limited (sorry :-( ), but what you describe
> sounds like what i've heard Perl being designed for. Perhaps if you were
> to ask on the Eugene Linux User's Group list, [EMAIL PROTECTED], you might
> be able to get some better info...
>
> --
> ----------------------------------------------------------------
> |efn Member Service| |Patrick R. Wade|
> |Oregon Public Networking / Eugene Free Net|http://www.efn.org/|
> |448 Charnelton St, Eugene, OR 97401 | (541) 484-9637|_________|
>
