while ( my $line1 = <$fh> )
{
    my $line2 = <$fh>;

    # etc
}
[download]

You'll need to check for eof on the second read though.

This sort of thing?

while ( my $line1 = <$fh> )
{
    my $line2 = <$fh>;

    # etc
}
[download]

Except that

while ( my $line1 = <$fh> )
[download]

is just the same as

while ( defined(my $line1 = <$fh>) )
[download]

so perhaps I would do:

while ( my $line1 = <$fh> )
{
    defined(my $line2 = <$fh>) or last;

    # etc
}
[download]

Why not just read both lines at a time? It even seems to handle the case of the file ending with an odd line correctly.

#! perl -slw
use strict;

while( my $line = <DATA> . <DATA> ) {
    print "'$line'";
}

__DATA__
1   2   3
    4   5   6
1   2   3
    4   5   6
1   2   3
    4   5   6
1   2   3
    4   5   6
1   2   3
    4   5   6
1   2   3
    4   5   6
1   2   3
    4   5   6
an  odd  record
[download]

for (;;) {
   my $line1 = <$fh>;
   last if not defined $line1;

   my $line2 = <$fh>;
   last if not defined $line2;

   ...
}
[download]

Please show us a little code. Read I know what I mean. Why don't you? then post about a dozen lines of code and data demonstrating your problem and tell us what you get and what you want. We don't want to see the database code, just a few line pertinent to showing the issue you have parsing the data and a few lines of sample data.

sounds like an XY problem to me. Why not something like this <pseudocode>?
Update After I finished writing this, I discovered FunkyMonk had already hit on essentially the same plan. ++, FunkyMonk

# Get the data into an array, or grab it a line at a time;

# Read a line/array element to a var, $input;    # ie, up to the first
+ return

$may_be_a_full_entry_or_may_not = $input;

Loop and read the next line to $input;
  use a regex to see if it starts with "day\t"
    if (it matches) {
      STORE $may_be_a_full_entry_or_may_not;
      s/"blank"//;    # or use a capturing regex 
                      # whatever that is: tab, space, multi-spaces???
                      # be careful to make only the first "blank" go a
+way!
      STORE $input;
    }
   else { 
     s/"blank"//;    # or use a capturing regex 
     $input = $may_be_a_full_entry_or_may_not; . $input; # concat the 
+two lines
     STORE $input to your db;
   }
[download]

repeating until done.

The details of the loop are left to you. Note, this is a) in-elegant and b) won't work, obviously, if the data entry folk spread their daily report over 3 lines.

Update2 Fixed formatting. Thank you, injunjoel!

The example code below allows for an arbitrary number of lines with an empty 'day' field, it reuses the last non-empty day field read.

#!/usr/bin/perl
use strict;use warnings;

my $day = '';
while (my $line = <DATA>) {
   chomp $line;
   my ($newday,$ptime,$stime,$ctime,$amount) = split /\t/,$line;
   $day = $newday if $newday..$newday;
   print "$day\t$ptime\t$stime\t$ctime\t$amount\n";
   }


__DATA__
01-May-2007     34      11am    567     665
        65      12am    657     4
        64      4pm     345     3
05-May-2007     87      10pm    987     56
06-May-2007     987     9am     8899    765
        567     unkn    4       3
[download]

01-May-2007     34      11am    567     665
01-May-2007     65      12am    657     4
01-May-2007     64      4pm     345     3
05-May-2007     87      10pm    987     56
06-May-2007     987     9am     8899    765
06-May-2007     567     unkn    4       3
[download]

Good solutions above, this is how you could do it in Perl 6:

my $fh = open("file");

for =$fh->$line1,$line2 {
  # do whatever with the two lines
}
[download]

I've been trying to do this in perl6, I can get your solution to work - perl6 may well have changed in the meantime.

In fact I get an error message about the =$

...
Preceding context expects a term, but found infix = instead.
Did you make a mistake in Pod syntax?
...
------> for =^$fh -> $line1, $line2 {
[download]

I had thought that something like this might work:

my $fh = open("file");
for $fh.lines,$fh.lines -> $line1,$line2 {
    say $line1;
    say $line2;
}
[download]

but this seems to produce a rather confusing result with all the lines stuck together is a Seq, and has some EOF related error, the 'singluar form':

for $fh.lines -> $line {
    say $line;
}
[download]

works fine.

What is true is already so. Owning up to it doesn't make it worse. - Eugene Gendlin