See attached Perl program. If you can run a Perl program, it does exactly
what you need, and a few more things too.
Craig
On Thu, Jan 31, 2019 at 12:03 PM gehtalles <toz...@t-online.de> wrote:
> I got a massive SDF file with some 300k structures I need to split into
> smaller files
>
> I tried
>
> babel infile.sdf -isdf -osdf -f 1 -l 1000 -O outfile.sdf
>
> but all meta data defined in
>
> > <data name>
> data content
>
> are discarded in the outfile.
>
> Is there an option I missed to include them in the new file?
>
>
>
> --
> Sent from: http://forums.openbabel.org/General-discussion-f3090658.html
>
>
> _______________________________________________
> OpenBabel-discuss mailing list
> OpenBabel-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
>
--
---------------------------------
Craig A. James
Chief Technology Officer
eMolecules, Inc.
---------------------------------
#!/usr/bin/perl
#======================================================================
# FILE: split_sdf.pl
# AUTHOR: Craig A. James
# DESCRIPTION:
# Splits an MDL SD File into manageable-sized chunks.
#
#======================================================================
# Copyright (c) 2006-2009, eMolecules Inc.
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation version 2 of the License.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#======================================================================
use strict;
$| = 1;
sub usage {
print "$_[0]\n" if defined($_[0]);
print <<EOF;
Usage: split_sdf.pl [N] [-skip N] [-do N] [file.sdf]
N is number of SD records per file
-skip N Skip N records before starting output
-do N Do N records and then quit
-random N/M Randomly only print approximately N/M of the total
-zap_extra_lf Zap extra linefeed between records
Files will be named with numeric suffix, e.g. "split_sdf.pl foo.sdf"
will generate files like these:
foo_1.sdf
foo_2.sdf
... etc.
EOF
print "Hit [Return] to exit...";
my $line = <STDIN>;
exit;
}
my $n = 50000;
if ($ARGV[0] =~ m/^[0-9]+$/) {
$n = $ARGV[0];
shift;
}
my $skip = 0;
if ($ARGV[0] eq "-skip") {
shift;
$skip = shift;
}
my $dorecords = 0;
if ($ARGV[0] eq "-do") {
shift;
$dorecords = shift;
}
my ($rand_n, $rand_m);
if ($ARGV[0] eq "-random") {
shift;
($rand_n, $rand_m) = split(/\//, $ARGV[0]);
shift;
$rand_n = int($rand_n);
$rand_m = int($rand_m);
$rand_n = 0 unless $rand_n > 0 && $rand_m > 0;
print "Random $rand_n/$rand_m\n";
}
my $zap_extra_lf = 0;
if ($ARGV[0] eq "-zap_extra_lf") {
shift;
$zap_extra_lf = 1;
}
my ($basename, $filename, $suffix);
if (defined($ARGV[0]) && $ARGV[0] ne "-") {
$filename = shift;
usage ("Can't read file '$filename'") if ! -r $filename;
$basename = $filename;
$basename =~ s/\.[a-zA-Z]*$//;
$suffix = $filename;
$suffix =~ s/^.*\.([a-zA-Z]+)$/$1/;
}
else { # no filename given?
$filename = "-"; # use standard input
$basename = "sdfile";
$suffix = "sdf";
}
my $filesuffix = 1;
my $count = 0;
my $total = 0;
open(SDFILE, "<$filename") || die("Failed to open '$filename' for reading!");
my $zf_filesuffix = sprintf("%04d", $filesuffix);
my $outfile = "${basename}_${zf_filesuffix}.$suffix";
open(OUTFILE, ">$outfile") || die("Couldn't open '$outfile' for writing.");
if ($dorecords > 0) {
print("Note: output limited to $dorecords records\n");
}
if ($skip > 0) {
print("Skipping $skip records...\n");
} else {
print "Writing file '$outfile' ... ";
}
my $skip_record;
my $r = int(rand($rand_m) + 1.0);
$skip_record = ($rand_n > 0) && ($r > $rand_n);
SDREC:
while (<SDFILE>) {
print OUTFILE $_ unless $skip > 0 || $skip_record;
if ($_ =~ m/^\$\$\$\$/) {
$r = int(rand($rand_m) + 1.0);
if (!$skip_record) {
if ($skip-- <= 0) {
$count++;
$total++;
}
elsif ($skip == 0) {
print "Writing file '$outfile' ... ";
}
if ($skip < 0 && ($count % $n) == 0) {
close OUTFILE;
print "$count records\n";
$filesuffix++;
$zf_filesuffix = sprintf("%04d", $filesuffix);
$outfile = "${basename}_${zf_filesuffix}.$suffix";
open(OUTFILE, ">$outfile") || die("Couldn't open '$outfile' for writing.");
print "Writing file '$outfile' ... ";
$count = 0;
}
if ($skip < 0 && $dorecords > 0) {
$dorecords--;
last SDREC if ($dorecords == 0);
}
if ($skip < 0 && $zap_extra_lf) {
# This is kind of hokey. Do a "read ahead" of five lines, and
# try to determine whether there's an extra line or not. If
# there is, discard it. Then print the lines that are left.
my @sdf_header;
foreach my $i (0..4) {
$sdf_header[$i] = <SDFILE>;
}
if ( $sdf_header[0] =~ m/^\n/
&& $sdf_header[4] =~ m/[ 0-9][ 0-9][0-9][ 0-9][ 0-9][0-9]/) {
shift @sdf_header;
}
print OUTFILE join("", @sdf_header);
}
}
$r = int(rand($rand_m) + 1.0);
$skip_record = ($rand_n > 0) && ($r > $rand_n);
}
}
print "$count records\n";
close OUTFILE;
print "Done: $total records total.\n";
_______________________________________________
OpenBabel-discuss mailing list
OpenBabel-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
