in reply to Re^2: Extracting Data from a second line
in thread Extracting Data from a second line

If I'm working with a regular filehandle, I always use -s to get the length to read, but unfortunately that doesn't work here. -s DATA will give you the length of the entire file, Perl code included, not just the part after __DATA__.

Update: With that stated, I do realize it wouldn't hurt anything to supply a bigger number than required as the third argument to read.

Replies are listed 'Best First'.
Re^4: Extracting Data from a second line
by ikegami (Patriarch) on Feb 16, 2006 at 18:54 UTC

    -s FILE is often too big — not just when using DATA — since read is to be given a number of characters (as opposed to a number of bytes) when binmode isn't set to raw. 

    open(FILE, $0)
   or die("Unable to open $0: $!\n");

$read = read(FILE, $buf='', -s FILE);

print("Character mode:\n");
print("Bytes requested: ", -s FILE,      "\n");  # 580
print("Chars read:      ", $read,        "\n");  # 560
print("Chars read:      ", length($buf), "\n");  # 560

seek(FILE, 0, 0);
binmode(FILE);
print("\n");

$read = read(FILE, $buf='', -s FILE);

print("Bin mode:\n");
print("Bytes requested: ", -s FILE,      "\n");  # 580
print("Chars read:      ", $read,        "\n");  # 580
print("Chars read:      ", length($buf), "\n");  # 580

    [download]

    But like you said, it's ok if it's too big.

[reply]
[d/l]
[select]
[reply]
        The number returned by that program is still larger than it needs to be in Windows. There are 41 characters to read in Windows, but $size_of_DATA is 44.
[reply]
[d/l]

In Section Seekers of Perl Wisdom