Re: foreach type of deal?

[toolic]

One way to get rid of the g is to use the substitution operator (s///). This deletes the g if it is the 1st character on a line:

while (<>) {
    s/^g//;
}

• Show a few lines of your input (in "code" tags), if you need more help.
• perlintro

[kennethk]

Since 500k+ is a big number and the spec says "They all start out with a g", I would rather use substr since there is no need for a test.

while (<>) {
    my $word = substr($_,1); # Start with the second letter
}

As a note for the OP, while is a nicer option than Foreach Loops in this scenario since, in general, foreach constructs the entire list before incrementing over it, whereas while will just hit your file-handle once per loop -- normally, that means while has a much smaller memory footprint than foreach does for large sets. I don't recall if in the particular case of foreach(<$fh>) { there is an optimization in perl to avoid this pitfall.

[BillKSmith]

Do we ever have perfect data? I prefer to use a test to warn about invalid data. The following code (untested) makes use of runtime switches (ref: perldoc perlrun). Use -p for the I/O and -i for backup.

#!perl -p -ibak
if (!s/^g(Ord_\d{4}\.png)\s*$/$1/) {
    warn "Invalid line: $_\n";
    $_ = <>;
    redo;
}