Becky has asked for the wisdom of the Perl Monks concerning the following question:

I have lines in a file that look like this: 

ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 36.32      D    N 
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96      D    C

I need to replace the 2 lone D characters in each line with an X. I DON'T want to replace thd D in CD1. Whatever regex I do I seem to replace the D in CD1 or muck up the formatting of the line, which must stay the same. Can you help? Thanks!

Replies are listed 'Best First'.
Re: Replacing characters in a line
by pelagic (Priest) on Mar 04, 2005 at 11:49 UTC
    Just use "word-boundary anchors" in your regex:
    /\bD\b/

    [download]

    pelagic
[reply]
[d/l]
Re: Replacing characters in a line
by holli (Abbot) on Mar 04, 2005 at 11:50 UTC
    looks like a fixed length line, so iīd simply go for substr(): 
    $_="ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 36.32
+      D    N";
substr($_,21,1)="X";
print;

    [download]
    You can use a regex too.


    holli, /regexed monk/
[reply]
[d/l]
Re: Replacing characters in a line
by monkey_boy (Priest) on Mar 04, 2005 at 13:52 UTC
    These are PDB files
    They are fixed position fields, so use holli's method.


    Im so sick of my Signature...
[reply]
[d/l]
Re: Replacing characters in a line
by sh1tn (Priest) on Mar 04, 2005 at 15:52 UTC
    Depending on which 'D' exactly must be replaced: 
    $_ =<<SQ;
ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 36.32    
+  D    N
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  D    C
SQ

s/(.+ CD1 .+) | (\b D \b)/$1 or X/gex;

__END__

STDOUT:
ATOM   3316  N   TYR X  24      97.479  18.805   2.990  1.00 36.32    
+  X    N
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  D    C

    [download]


    Or in the other case: 
    $_ =<<SQ;
ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 36.32    
+  D    N
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  D    C
SQ

s/(?<!CD1 \s TYR \s)(\b D \b)/X/gx;

__END__

STDOUT:
ATOM   3316  N   TYR X  24      97.479  18.805   2.990  1.00 36.32    
+  X    N
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  X    C

    [download]


[reply]
[d/l]
[select]
Re: Replacing characters in a line
by perlsen (Chaplain) on Mar 04, 2005 at 12:39 UTC

    Hi, just try this,

    $string='ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 
+36.32      D    N 
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  D    C';

$string=~s#((\w+)   \d{4}  \w   TYR )D#$1X#gsi;

print $string;

output:
*******
ATOM   3316  N   TYR X  24      97.479  18.805   2.990  1.00 36.32    
+  D    N
ATOM   3322  CD1 TYR D  24      96.225  15.974   2.812  1.00 17.96    
+  D    C

    [download]
[reply]
[d/l]
      Why not just s/(?<=TYR )D/X/;?

      Update:
      Or, to keep it fixed-length: 
      s/^(.{21})D/$1X/;

      [download]

      Sidenote:
      Has anyone an idea, why this: 
      s/^(?<=.{21})D/X/;

      [download]
      does not work, while this 
      s/(?<=.{21})D/X/;

      [download]
      does (but donīt use this, itīs unsafe)?


      holli, /regexed monk/
[reply]
[d/l]
[select]
        Are we sure that D will always be preceeded by "TYR " ??
        Manav
[reply]
Re: Replacing characters in a line
by Taulmarill (Deacon) on Mar 04, 2005 at 14:36 UTC
    the following simple regex should work:
    s/ D / X /g;
[reply]
[d/l]
      *UPDATE*

      Ignore me. I didnt read the message properly

      but that would also replace the other lonley "D"'s
      e.g. 

      ATOM   3316  N   TYR D  24      97.479  18.805   2.990  1.00 36.32      D    N



      Im so sick of my Signature...
[reply]

Back to Seekers of Perl Wisdom