Regular Expression - Perl: Adding Specificity - regex

Question

I wanted to write a specific regex to do the following in a specific file format.
It should be able to check with a regular expression whether the third field is just an O or has anything following the O.
Currently, I use the following syntax as shown below:
if ($line !~ /^ATOM\s+\d+\s+(O)/)
{
}
Could you guys help me out?
ATOM 284 OD1 ASN 1 34 -7.92000 -6.74600 -4.73800 O_2 1 2 -0.55000 0 0
ATOM 308 O LEU 1 35 -10.48500 -13.59200 -8.35100 O_2 1 2 -0.51000 0 0
I want to be able to print out the lines from a file that contain something after the O. (Such as the OD1 line). I should be able to remove the lines with just an O.

Answer 1

Just add \S (meaning "a non-space character"):
/^ATOM\s+\d+\s+O\S/
Incidentally, I get the impression that you don't actually know regular expressions? I recommend the perlretut ("Perl regular expressions tutorial") manpage.

Answer 2

You currently use !~ for does not match. If you want it to match you'll have to change it to =~. You also don't need the parenthesis are the O. () are used for capture groups. If you wanted to capture the group, you could do (O[A-Za-z0-9]).
if ($line =~ /^ATOM\s+\d+\s+O/)
# we don't care what's after the O, could be nothing or some characters
or
if ($line =~ /^ATOM\s+\d+\s+(O[a-zA-Z0-9]*)/)
# this will capture OD1 or just O in $1
or if you want to see whether there are characters after the 0, you can use
if ($line =~ /^ATOM\s+\d+\s+(O[a-zA-Z0-9]+)/)
# this would only capture OD1 in $1

Answer 3

You can use split to split out that one field:
my $field = ( split /\s+/, $line )[2];
This will make the regular expression easier you want easier to do. Plus, it makes what you're doing more obvious:
if ( $field =~ /^O/ ) {
here be dragons...
}
In fact, you might want to do that for all of your fields to make it easier to manipulate. Since I don't know what your fields mean, I'm just calling them $fld1, $fld2, etc.
my ( $fld1, $fld2, $fld3, $fld4, ... ) = split /\s+/, $line;
if ( $fld3 =~ /^O/ ) {
here be dragons...
}
Now, you can easily refer to your individual fields in your program.

Answer 4

if you're prefer not to use regex you can use split as #David-W said
my #fields = split /\s+/, $line;
##now $field[2]
if ($fields[2] ne 'o'){
##this line has o and other letters
}
but this will be much slower than regex especially for large data files
as for regex your data file starts with white space (apperantly)
so your regex should be as the following
if ($line !~ /^\s+ATOM\s+\d+\s+(O)\s+/){
##this line has o with other letters beside it
} else {
## this line only has o in field 3
}
adding ^\s+ at the beginning or remove ^ mark totally
$line !~ /ATOM\s+\d+\s+(O)\s+/
then add \s+ after o (at the end) to make sure it followed by space immediately
if you're not interested in capturing fields value you better off capture group (o)
if ($line !~ /ATOM\s+\d+\s+O\s+/) {
#...
} else {
#...
}