$/ = ""; # read paragraphs
while (my $para = <>)
{
  my $lastHeading;

  my @lines = split(/\n/, $para);
  if (@lines == 1) # heading?
  {
     $lastHeading = $lines[0];
     next;
  }
  
  my $hitMe;
 
  if ($lastHeading eq 'MEANINGLESS TEXT')
  {
   $hitMe = $lines[4]; # fifth line
  }
  elsif ($lastHeading eq 'ALTERNATIVE TEXT')
  {
   $hitMe = $lines[5]; # sixth line
  }
  # now do something...
}
[download]

First, you could indicate what part of the file you are in with a flag, ie,

$meaningless=0;
$alternate=0;
while (<FILE>) {
  if (/MEANINGLESS/) {
    $meaningless=1;
    $alternate=0;
  } elsif (/ALTERNATE/) {
    $meaningless=0;
    $alternate=1;
  }
}
[download]

Next, I am not at all clear on why you need line numbers. A match is a match is a match. If it matches, you can print it, regardless of the line number. You can count line numbers if you combine with the above flag system:

$meaningless_count=0;
$alternate_count=0;
while (<FILE>) {
  #include flags from above
  if ($meaningless) {
    $meaningless_count++;
  } elsif($alternate) {
    $alternate_count++;
  }
}
[download]

Scott

Are you:

• Trying to compare two files, and if they have an identical line, print the lines from the second file surrounding that line?
• Searching a number of texts for a particular string from the command line, and if the string appears in one of the texts, print the first line of the text along with the lines surrounding the match?
• or perhaps something more complicated?

___
-DA
> perl -MPOSIX -le '$ENV{TZ}="EST";print ctime(1000000000)'
Sat Sep  8 20:46:40 2001
[download]

I'm searching a single text file. If I match on a phone number I want it to print certain lines within the block record that it find it in. The block records have different titles and data in them but always have the phone number. So if I find the phone number and devise a way of recognizing what record block it is then I can tell the script for this record block print these lines and for this record block print these lines only if you find the phone number and recognize the block record heading.

The Brassmon_k
The examples above helped me a little but am still not getting correct results. I can't seem to get the script to print off any other even if it matches them except for the line with the phone number in it.

If I can restate the problem to make sure I understand it:

You have a text file with records which start with a newline, a line of text, and another newline. You want to print selected lines from a record, if both the following are true: the record title matches a predefined list, and the phone number is in the record. The title of the record determines which lines of the record to print.

Your question for us is how to find out what record you're looking at.

It sounds like bikenomad's code should do almost exactly that, if you simply include print statements inside the if statement:

    if ($lastHeading eq 'MEANINGLESS TEXT') {
        print $lines[2]; # third line
        print $lines[5]; # sixth line
        print $lines[7]; # eighth line
    } elsif ($lastHeading eq 'ALTERNATIVE TEXT') {
        print $lines[3]; # fourth line
        print $lines[5]; # sixth line
        print $lines[7]; # eighth line
[download]

Does this help?

___
-DA
> perl -MPOSIX -le '$ENV{TZ}="EST";print ctime(1000000000)'
Sat Sep  8 20:46:40 2001
[download]

Ok, it is becoming a little more clear.

You need to keep every line in the block, probably by assigning each line as an element of an array. After you determine that a text block is the one you want, you can print each line you want from the array, or all of them. How you decide to print a given line could be done with a regex or grep or other possiblities. If you can't get it to work now, post some code, and we'll give you pointers.

Scott