Re: How best to grab SQL table names?

[drieux]

On Sunday, Nov 9, 2003, at 16:32 US/Pacific, Dan Anderson wrote:
[..]
while ($_ = <FILE>) {
  if ($_ =~ m/CREATE.*TABLE/) {
    $_ = $';
    while (not ($_ =~ m/)//)) {  # match last parenthesis
       $_ .= <FILE>;
    }
    my $table_name = $`;
    $table_name = tr/ \t\n\r//d;
    return $table_name;
  }
}
My Programming Perl book says not to use $' $` and $&.  Is there a
better way to do this?
well a bit of it has to do with how the Perl Regular Expression
Engine works, the other part has to do with making your code a
bit more readable, and maintainable. The use of the $' and $`
as the "perldoc perlvar" pages will tell you imposes a considerable
performance penalty.
right, a bit of clean up of the code:

        while (<FILE>) { # it will set the line to $_ by default
                # cut off everything BEFORE our 'pattern'
                if ( s/.*CREATE.*TABLE//)
                {
                        chomp;  # remove any 'newline tokens'
                        my $line = $_ ; # save off what we have
                        # read until we have come to the closing paren
                        while ( $line !~ m/\)/ ) {
                                $_ = <FILE>;
                                chomp;
                                $line .= $_;    
                        }
                        # we will either have seen that last paren, or it did not exist
                        ( my $table_name = $line ) =~ s/.*\)//;
                $table_name = tr/ \t\n\r//d;
                        return $table_name;
                
                } # found opening line
        }# end while loop
being a bit of a tight translation.

ciao
drieux
---

--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]