X-Git-Url: https://git.verplant.org/?a=blobdiff_plain;f=contrib%2FextractDS.px;h=80d873b9a7480be87ea1b74ae1bbc9d9d2d7c4ac;hb=d5272dbabf4c7a47f34066c05f479523d7b8beae;hp=3bf4b9e01c402758976095c7b470ecbb58e3c493;hpb=ef14b1348206799487e8c444f16a7f7b5b89be90;p=collectd.git

diff --git a/contrib/extractDS.px b/contrib/extractDS.px
index 3bf4b9e0..80d873b9 100755
--- a/contrib/extractDS.px
+++ b/contrib/extractDS.px
@@ -1,25 +1,43 @@
 #!/usr/bin/perl
 
+# collectd - contrib/rrd_filter.px
+# Copyright (C) 2007-2008  Florian octo Forster
+#
+# 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; only version 2 of the License is applicable.
+#
+# 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.
+#
+# You should have received a copy of the GNU General Public License along
+# with this program; if not, write to the Free Software Foundation, Inc.,
+# 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
+#
+# Authors:
+#   Florian octo Forster <octo at verplant.org>
+
 use strict;
 use warnings;
 
 =head1 NAME
 
-extractDS.px - Extract a single data-source from an RRD-file
+rrd_filter.px - Perform same advanced non-standard operations on an RRD file.
 
 =head1 SYNOPSYS
 
-  extractDS.px -i input.rrd -s source_ds -o output.rrd -d destination_ds
+  rrd_filter.px -i input.rrd -o output.rrd [options]
 
 =head1 DEPENDENCIES
 
-extractDS.px requires Perl and the included L<Getopt::Long> module, as well as
-the L<XML::Simple> module.
+extractDS.px requires the RRDTool binary, Perl and the included L<Getopt::Long>
+module.
 
 =cut
 
 use Getopt::Long ('GetOptions');
-use XML::Simple (qw(xml_in xml_out));
 use Data::Dumper ();
 
 our $InFile;
@@ -27,25 +45,95 @@ our $InDS = [];
 our $OutFile;
 our $OutDS = [];
 
+our $NewRRAs = [];
+
+our $Step = 0;
+
+=head1 OPTIONS
+
+The following options can be passed on the command line:
+
+=over 4
+
+=item B<--infile> I<file>
+
+=item B<-i> I<file>
+
+Reads from I<file>. If I<file> ends in C<.rrd>, then C<rrdtool dump> is invoked
+to create an XML dump of the RRD file. Otherwise the XML dump is expected
+directly. The special filename C<-> can be used to read from STDIN.
+
+=item B<--outfile> I<file>
+
+=item B<-o> I<file>
+
+Writes output to I<file>. If I<file> ends in C<.rrd>, then C<rrdtool restore>
+is invoked to create a binary RRD file. Otherwise an XML output is written. The
+special filename C<-> can be used to write to STDOUT.
+
+=item B<--map> I<in_ds>:I<out_ds>
+
+=item B<-m> I<in_ds>:I<out_ds>
+
+Writes the datasource I<in_ds> to the output and renames it to I<out_ds>. This
+is useful to extract one DS from an RRD file.
+
+=item B<--step> I<seconds>
+
+=item B<-s> I<seconds>
+
+Changes the step of the output RRD file to be I<seconds>. The new stepsize must
+be a multiple of the old stepsize of the other way around. When increasing the
+stepsize the number of PDPs in each RRA must be dividable by the factor by
+which the stepsize is increased. The length of CDPs and the absolute length of
+RRAs (and thus the data itself) is not altered.
+
+Examples:
+
+  step =  10, rra_steps = 12   =>   step = 60, rra_steps =  2
+  step = 300, rra_steps =  1   =>   step = 10, rra_steps = 30
+
+=item B<--rra> B<RRA>:I<CF>:I<XFF>:I<steps>:I<rows>
+
+=item B<-a> B<RRA>:I<CF>:I<XFF>:I<steps>:I<rows>
+
+Inserts a new RRA in the generated RRD file. This is done B<after> the step has
+been adjusted, take that into account when specifying I<steps> and I<rows>. For
+an explanation of the format please see L<rrdcreate(1)>.
+
+=back
+
+=cut
+
 GetOptions ("infile|i=s" => \$InFile,
-	"inds|s=s" => sub { push (@$InDS, $_[1]); },
 	"outfile|o=s" => \$OutFile,
-	"outds|d=s" => sub { push (@$OutDS, $_[1]); })
-	or exit (1);
+	'map|m=s' => sub
+	{
+		my ($in_ds, $out_ds) = split (':', $_[1]);
+		if (!defined ($in_ds) || !defined ($out_ds))
+		{
+			print STDERR "Argument for `map' incorrect! The format is `--map in_ds:out_ds'\n";
+			exit (1);
+		}
+		push (@$InDS, $in_ds);
+		push (@$OutDS, $out_ds);
+	},
+	'step|s=i' => \$Step,
+	'rra|a=s' => sub
+	{
+		my ($rra, $cf, $xff, $steps, $rows) = split (':', $_[1]);
+		if (($rra ne 'RRA') || !defined ($rows))
+		{
+			print STDERR "Please use the standard RRDTool syntax when adding RRAs. I. e. RRA:<cf><xff>:<steps>:<rows>.\n";
+			exit (1);
+		}
+		push (@$NewRRAs, {cf => $cf, xff => $xff, steps => $steps, rows => $rows});
+	}
+) or exit (1);
 
-if (!$InFile || !$OutFile || !@$InDS || !@$OutDS)
+if (!$InFile || !$OutFile)
 {
-	print STDERR "Usage: $0 -i <infile> -s <inds> -o <outfile> -d <outds>\n";
-	exit (1);
-}
-if (!-f $InFile)
-{
-	print STDERR "Input file does not exist\n";
-	exit (1);
-}
-if (-f $OutFile)
-{
-	print STDERR "Output file does exist\n";
+	print STDERR "Usage: $0 -i <infile> -m <in_ds>:<out_ds> -s <step>\n";
 	exit (1);
 }
 if ((1 + @$InDS) != (1 + @$OutDS))
@@ -54,7 +142,7 @@ if ((1 + @$InDS) != (1 + @$OutDS))
 	exit (1);
 }
 
-extract_ds ($InFile, $OutFile);
+main ($InFile, $OutFile);
 exit (0);
 
 {
@@ -65,10 +153,17 @@ my $current_index;
 # state 2 == parse values
 my $state;
 my $out_cache;
-sub handle_line
+sub handle_line_dsmap
 {
-	my $fh = shift;
 	my $line = shift;
+	my $index = shift;
+	my $ret = '';
+
+	if ((@$InDS == 0) || (@$OutDS == 0))
+	{
+		post_line ($line, $index + 1);
+		return;
+	}
 
 	if (!defined ($state))
 	{
@@ -119,7 +214,10 @@ sub handle_line
 			for (my $new_index = 0; $new_index < @$InDS; $new_index++)
 			{
 				my $old_index = $ds_index->[$new_index];
-				print $fh $out_cache->[$old_index];
+				while ($out_cache->[$old_index] =~ m/^(.*)$/gm)
+				{
+					post_line ("$1\n", $index + 1);
+				}
 			}
 
 			# Clear the cache - it's used in state1, too.
@@ -128,14 +226,14 @@ sub handle_line
 				$out_cache->[$i] = '';
 			}
 
-			print $fh $line;
+			$ret .= $line;
 			$current_index = -1;
 			$state = 1;
 		}
 		elsif ($current_index == -1)
 		{
 			# Print all the lines before the first DS definition
-			print $fh $line;
+			$ret .= $line;
 		}
 		else
 		{
@@ -156,7 +254,10 @@ sub handle_line
 			for (my $new_index = 0; $new_index < @$InDS; $new_index++)
 			{
 				my $old_index = $ds_index->[$new_index];
-				print $fh $out_cache->[$old_index];
+				while ($out_cache->[$old_index] =~ m/^(.*)$/gm)
+				{
+					post_line ("$1\n", $index + 1);
+				}
 			}
 
 			# Clear the cache
@@ -165,19 +266,19 @@ sub handle_line
 				$out_cache->[$i] = '';
 			}
 
-			print $fh $line;
+			$ret .= $line;
 			$current_index = -1;
 		}
 		elsif ($line =~ m#<database>#)
 		{
-			print $fh $line;
+			$ret .= $line;
 			$state = 2;
 		}
 		elsif ($current_index == -1)
 		{
 			# Print all the lines before the first DS definition
 			# and after cdp_prep
-			print $fh $line;
+			$ret .= $line;
 		}
 		else
 		{
@@ -189,7 +290,7 @@ sub handle_line
 	{
 		if ($line =~ m#</database>#)
 		{
-			print $fh $line;
+			$ret .= $line;
 			$current_index = -1;
 			$state = 1;
 		}
@@ -197,13 +298,14 @@ sub handle_line
 		{
 			my @values = ();
 			my $i;
-			my $output = "\t\t";
+			
+			$ret .= "\t\t";
 
 			if ($line =~ m#(<!-- .*? -->)#)
 			{
-				$output .= "$1 ";
+				$ret .= "$1 ";
 			}
-			$output .= "<row> ";
+			$ret .= "<row> ";
 
 			$i = 0;
 			while ($line =~ m#<v>\s*([^<\s]+)\s*</v>#g)
@@ -215,19 +317,249 @@ sub handle_line
 			for (my $new_index = 0; $new_index < @$InDS; $new_index++)
 			{
 				my $old_index = $ds_index->[$new_index];
-				$output .= '<v> ' . $values[$old_index] . ' </v> ';
+				$ret .= '<v> ' . $values[$old_index] . ' </v> ';
 			}
-			$output .= "</row>\n";
-			print $fh $output;
+			$ret .= "</row>\n";
 		}
 	}
 	else
 	{
 		die;
 	}
-}} # handle_line
 
-sub extract_ds
+	if ($ret)
+	{
+		post_line ($ret, $index + 1);
+	}
+}} # handle_line_dsmap
+
+#
+# The _step_ handler
+#
+{
+my $step_factor_up;
+my $step_factor_down;
+sub handle_line_step
+{
+	my $line = shift;
+	my $index = shift;
+
+	if (!$Step)
+	{
+		post_line ($line, $index + 1);
+		return;
+	}
+
+	$step_factor_up ||= 0;
+	$step_factor_down ||= 0;
+
+	if (($step_factor_up == 0) && ($step_factor_down == 0))
+	{
+		if ($line =~ m#<step>\s*(\d+)\s*</step>#i)
+		{
+			my $old_step = 0 + $1;
+			if ($Step < $old_step)
+			{
+				$step_factor_down = int ($old_step / $Step);
+				if (($step_factor_down * $Step) != $old_step)
+				{
+					print STDERR "The old step ($old_step seconds) "
+					. "is not a multiple of the new step "
+					. "($Step seconds).\n";
+					exit (1);
+				}
+				$line = "<step> $Step </step>\n";
+			}
+			elsif ($Step > $old_step)
+			{
+				$step_factor_up = int ($Step / $old_step);
+				if (($step_factor_up * $old_step) != $Step)
+				{
+					print STDERR "The new step ($Step seconds) "
+					. "is not a multiple of the old step "
+					. "($old_step seconds).\n";
+					exit (1);
+				}
+				$line = "<step> $Step </step>\n";
+			}
+			else
+			{
+				$Step = 0;
+			}
+		}
+	}
+	elsif ($line =~ m#<pdp_per_row>\s*(\d+)\s*</pdp_per_row>#i)
+	{
+		my $old_val = 0 + $1;
+		my $new_val;
+		if ($step_factor_up)
+		{
+			$new_val = int ($old_val / $step_factor_up);
+			if (($new_val * $step_factor_up) != $old_val)
+			{
+				print STDERR "Can't divide number of PDPs per row ($old_val) by step-factor ($step_factor_up).\n";
+				exit (1);
+			}
+		}
+		else
+		{
+			$new_val = $step_factor_down * $old_val;
+		}
+		$line = "<pdp_per_row> $new_val </pdp_per_row>\n";
+	}
+
+	post_line ($line, $index + 1);
+}} # handle_line_step
+
+#
+# The _add RRA_ handler
+#
+{
+my $add_rra_done;
+my $num_ds;
+sub handle_line_add_rra
+{
+  my $line = shift;
+  my $index = shift;
+
+  my $post = sub { for (@_) { post_line ($_, $index + 1); } };
+
+  $num_ds ||= 0;
+
+  if (!@$NewRRAs || $add_rra_done)
+  {
+    $post->($line);
+    return;
+  }
+
+  if ($line =~ m#<ds>#i)
+  {
+    $num_ds++;
+  }
+  elsif ($line =~ m#<rra>#i)
+  {
+    for (my $i = 0; $i < @$NewRRAs; $i++)
+    {
+      my $rra = $NewRRAs->[$i];
+      my $temp;
+      $post->("\t<rra>\n",
+      "\t\t<cf> $rra->{'cf'} </cf>\n",
+      "\t\t<pdp_per_row> $rra->{'steps'} </pdp_per_row>\n",
+      "\t\t<params>\n",
+      "\t\t\t<xff> $rra->{'xff'} </xff>\n",
+      "\t\t</params>\n",
+      "\t\t<cdp_prep>\n");
+
+      for (my $j = 0; $j < $num_ds; $j++)
+      {
+	$post->("\t\t\t<ds>\n",
+	"\t\t\t\t<primary_value> NaN </primary_value>\n",
+	"\t\t\t\t<secondary_value> NaN </secondary_value>\n",
+	"\t\t\t\t<value> NaN </value>\n",
+	"\t\t\t\t<unknown_datapoints> 0 </unknown_datapoints>\n",
+	"\t\t\t</ds>\n");
+      }
+
+      $post->("\t\t</cdp_prep>\n", "\t\t<database>\n");
+      $temp = "\t\t\t<row>" . join ('', map { "<v> NaN </v>" } (1 .. $num_ds)) . "</row>\n";
+      for (my $j = 0; $j < $rra->{'rows'}; $j++)
+      {
+	$post->($temp);
+      }
+      $post->("\t\t</database>\n");
+    }
+  }
+
+  $post->($line);
+}} # handle_line_add_rra
+
+#
+# The _output_ handler
+#
+{
+my $fh;
+sub set_output
+{
+	$fh = shift;
+}
+
+sub handle_line_output
+{
+	my $line = shift;
+	my $index = shift;
+
+	if (!defined ($fh))
+	{
+		post_line ($line, $index + 1);
+		return;
+	}
+	
+	print $fh $line;
+}} # handle_line_output
+
+#
+# Dispatching logic
+#
+{
+my @handlers = ();
+sub add_handler
+{
+	my $handler = shift;
+
+	die unless (ref ($handler) eq 'CODE');
+	push (@handlers, $handler);
+} # add_handler
+
+sub post_line
+{
+	my $line = shift;
+	my $index = shift;
+
+	if (0)
+	{
+		my $copy = $line;
+		chomp ($copy);
+		print "DEBUG: post_line ($copy, $index);\n";
+	}
+
+	if ($index > $#handlers)
+	{
+		return;
+	}
+	$handlers[$index]->($line, $index);
+}} # post_line
+
+sub handle_fh
+{
+	my $in_fh = shift;
+	my $out_fh = shift;
+
+	set_output ($out_fh);
+
+	if (@$InDS)
+	{
+	  add_handler (\&handle_line_dsmap);
+	}
+
+	if ($Step)
+	{
+	  add_handler (\&handle_line_step);
+	}
+
+	if (@$NewRRAs)
+	{
+	  add_handler (\&handle_line_add_rra);
+	}
+
+	add_handler (\&handle_line_output);
+
+	while (my $line = <$in_fh>)
+	{
+		post_line ($line, 0);
+	}
+} # handle_fh
+
+sub main
 {
 	my $in_file = shift;
 	my $out_file = shift;
@@ -235,17 +567,53 @@ sub extract_ds
 	my $in_fh;
 	my $out_fh;
 
-	open ($in_fh,  '-|', 'rrdtool', 'dump', $in_file) or die ("open (rrdtool): $!");
-	open ($out_fh, '|-', 'rrdtool', 'restore', '-', $out_file) or die ("open (rrdtool): $!");
+	my $in_needs_close = 1;
+	my $out_needs_close = 1;
 
-	while (my $line = <$in_fh>)
+	if ($in_file =~ m/\.rrd$/i)
+	{
+		open ($in_fh,  '-|', 'rrdtool', 'dump', $in_file) or die ("open (rrdtool): $!");
+	}
+	elsif ($in_file eq '-')
 	{
-		handle_line ($out_fh, $line);
+		$in_fh = \*STDIN;
+		$in_needs_close = 0;
 	}
+	else
+	{
+		open ($in_fh, '<', $in_file) or die ("open ($in_file): $!");
+	}
+
+	if ($out_file =~ m/\.rrd$/i)
+	{
+		open ($out_fh, '|-', 'rrdtool', 'restore', '-', $out_file) or die ("open (rrdtool): $!");
+	}
+	elsif ($out_file eq '-')
+	{
+		$out_fh = \*STDOUT;
+		$out_needs_close = 0;
+	}
+	else
+	{
+		open ($out_fh, '>', $out_file) or die ("open ($out_file): $!");
+	}
+
+	handle_fh ($in_fh, $out_fh);
+
+	if ($in_needs_close)
+	{
+		close ($in_fh);
+	}
+	if ($out_needs_close)
+	{
+		close ($out_fh);
+	}
+} # main
+
+=head1 LICENSE
 
-	close ($in_fh);
-	close ($out_fh);
-} # extract_ds
+This script is licensed under the GNU general public license, versionE<nbsp>2
+(GPLv2).
 
 =head1 AUTHOR