#!/usr/bin/perl -w
#	summarize_epics_wbpm.perl [filename]
#
# 	by: Jaideep Singh
#
#	date: 05/16/2006 
#
#	about:	reads a raw epics data text file and create a summary file
#		creates a file:
#				(0) first line is the prescalar string
#				(1) afterwards, each line is an epics variable
#				(2) if the variable is numeric, then the line will be:
#
#					variable name <tab> number of values <tab> minimum <tab> maximum <tab> mean <tab> standard deviation
#			        (3) the numeric variables come at the beginning of the file and are separated from the alphanumeric variables by a "-----" line
#				(4) if the variable is alphanumeric, then the line will be:
#
#					variable name <tab> number of values <tab> first value <tab> number of occurences <tab> second value <tab> number of occurences ...
#
#					Note that some EPICS-like stuff still leaks through...
#
#	input file is of the format: 
#		first line might be prescalar string
#		all subsequent lines:
#					timestamp<tab>seconds from 0000 Jan 1 2003<tab>variable name<tab>varicable value
#
#
#	EPICS info is injected into the data stream at certain times invertals.
#	More important stuff is logged more often.
#	Immediately preceding one log cycle, a timestamp is grabbed
#	Each variable is recorded in ASCII format on its own line: 
#
#	[variable name, upto 30 characters][at least one blank space][upto 20 characters for value] 
#	
#	for example:
#
#	IPM1H04B.XPOS                  -0.0408508
#	000000000111111111122222222223333333333444444444455
#	123456789012345678901234567890123456789012345678901
#
#	Variable names can have ".", ":","+","_" as well as alphanumeric characters
#	
#	Modified to cut on beam trips
#	for bpm variables, beam trips result in the string "0"  --->  these are not included in the statistics


use strict "subs";

#getting name of file to search + ARGUMENT ERROR HANDLING
$filename = $ARGV[0];
die "Uh, what file do you want me to look at dude?: Too few arguments\n" unless ($filename);
die "Scared and confused: Too many arguments\n" unless (($#ARGV+1) < 2);

#start flagging beam trips
open(FILE,$filename) or die "d'oh!: - $filename - not found\n";
#initializing variables
$trip_line_count = 0;
$trip_line_check_count = 0;
$trip_line_found_count = 0;

while(<FILE>)
{
 $trip_line_count++;
 chomp;
 if (/(ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+)/)
 {
#	print "Found prescalars: $1\n";
 }
 elsif ($_ ne "")
 {
	@line_parts = split(/\t/);
	if ($line_parts[3] =~ /([0123456789eE\.\+\-]+)/) { $poop = $1; if ($line_parts[3] eq $1) { $type = "num" } else { $type = "str" } } else {$poop = ""; $type = "str"}
	if ($type eq "num")
	{
		$trip_flag{$line_parts[1]} = "on" if (not $trip_flag{$line_parts[1]});  #beam is on by default
		if (($line_parts[2] eq "IPM1C08.XPOS") || ($line_parts[2] eq "IPM1C10.XPOS") || ($line_parts[2] eq "IPM1C10.YPOS") || ($line_parts[2] eq "IPM1C12.XPOS") || ($line_parts[2] eq "IPM1C12.YPOS") || ($line_parts[2] eq "IPM1C14.XPOS") || ($line_parts[2] eq "IPM1C14.YPOS") || ($line_parts[2] eq "IPM1C16.XPOS") || ($line_parts[2] eq "IPM1C16.YPOS") || ($line_parts[2] eq "IPM1C18.XPOS") || ($line_parts[2] eq "IPM1C18.YPOS") || ($line_parts[2] eq "IPM1C20.XPOS") || ($line_parts[2] eq "IPM1C20.YPOS") || ($line_parts[2] eq "IPM1H01.XPOS") || ($line_parts[2] eq "IPM1H01.YPOS") || ($line_parts[2] eq "IPM1H04A.XPOS") || ($line_parts[2] eq "IPM1H04A.YPOS") || ($line_parts[2] eq "IPM1H04B.XPOS") || ($line_parts[2] eq "IPM1H04B.YPOS"))
#		if ($line_parts[2] eq "hac_bcm_average")
		{
			$trip_line_check_count++;
			if ($line_parts[3] eq "0")
#			if ($line_parts[3] < 0.10)
			{
				$trip_flag{$line_parts[1]} = "off";  #no beam
				$trip_line_found_count++;
			}
		}
	}
 } 
};
close(FILE);
#really important!
$after_trip_cut_time = 30.1;
$before_trip_cut_time = 6.1;
#really important!
#initializations
$timecut = 0.0;
$beam_on_sum = 0.0;
$beam_on_cut_sum = 0.0;
$number_of_trips = 0.0;
#counting beam on time and number of trips , cutting out after_trip_time
foreach $timestamp (sort keys %trip_flag) 
{
	if (not $previous_timestamp)
	{ $previous_timestamp = $timestamp }

	$current_flag = $trip_flag{$timestamp};
	if (not $previous_flag)
	{ $previous_flag = $current_flag }

	if ($current_flag eq "off")
	{ 
		$timecut = $timestamp + $after_trip_cut_time;
		if ($previous_flag eq "on") { $number_of_trips++}
	}
	else
	{ $beam_on_sum += $timestamp-$previous_timestamp }

	if ($timestamp < $timecut)
	{ $trip_flag{$timestamp} = "off-cut" }

#	print "$timestamp\t\t$current_flag\t\t$trip_flag{$timestamp}\n";
	$last_timestamp = $timestamp;
	$previous_flag = $current_flag;
	$previous_timestamp = $timestamp;
}
#cutting out before_trip_time
$timecut = $last_timestamp + 1.0;
#print "\n ***** \n";
foreach $timestamp (reverse sort keys %trip_flag) 
{
	$current_flag = $trip_flag{$timestamp};
	if (($current_flag eq "off") || ($current_flag eq "off-cut"))
	{ $timecut = $timestamp - $before_trip_cut_time }
	
	if ($timestamp > $timecut)
	{ $trip_flag{$timestamp} = "off-cut2" }
	
#	print "$timestamp\t\t$current_flag\t\t$trip_flag{$timestamp}\n";
	$first_timestamp = $timestamp;
}
#counting beam on time after cuts
$previous_timestamp = 0;
foreach $timestamp (sort keys %trip_flag) 
{
	if (not $previous_timestamp)
	{ $previous_timestamp = $timestamp }
	
	$current_flag = $trip_flag{$timestamp};
	if ($current_flag eq "on")
	{ $beam_on_cut_sum += $timestamp-$previous_timestamp }
	
	$previous_timestamp = $timestamp;
}
$total_run_time = $last_timestamp-$first_timestamp;
$beam_off_time = $total_run_time - $beam_on_sum;
#print "trip line count=\t$trip_line_count\n";
#print "trip line check count=\t$trip_line_check_count\n";
#print "trip line found count=\t$trip_line_found_count\n";
#print "approx number of trips=\t$number_of_trips\n";
#print "approx total run time(sec)=\t$total_run_time\n";
#print "approx beam off time (sec)=\t$beam_off_time\n";
#print "approx run time after beam cuts (sec) =\t $beam_on_cut_sum\n";
#end flagging beam trips
#die and print "\ndone!\n";

open(FILE,$filename) or die "d'oh!: - $filename - not found\n";
#print "\nSearching raw data file:\t$filename\n";

#initializing variables
$line_count = 0;

LINENUM: while(<FILE>)
{
 $line_count++;
 chomp;
 if (/(ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+\,ps\d\=\d+)/)
 {
	$prescalars = $1;
 }
 elsif ($_ ne "")
 {
	@line_parts = split(/\t/);
	if ($line_parts[3] =~ /([0123456789eE\.\+\-]+)/) { $poop = $1; if ($line_parts[3] eq $1) { $type = "num" } else { $type = "str" } } else {$poop = ""; $type = "str"}
#	$poop is used as just a diagnostic variable
#	print "\*$line_parts[3]\*\t\*$poop\*\t$type\n";
	if ($type eq "num")
	{
#for diagnostics
#	   if ($line_parts[2] eq "hac_bcm_average")
#	   if (($line_parts[2] eq "MSEPRIR") || ($line_parts[2] eq "hac_bcm_average") || ($line_parts[2] eq "HALLA\:p") || ($line_parts[2] eq "halla_MeV")
#|| ($line_parts[2] eq "IPM1H04A.XPOS"))
#	   { print "$_\t$trip_flag{$line_parts[1]}\n" }
#for diagnostics
#
# just use count good counter
#
	   if (not $data_num{$line_parts[2]}{"count"})
	   { 
		$data_num{$line_parts[2]}{"count"} = 1.0;
	 	$data_num{$line_parts[2]}{"count0"} = 0.0;
	   	$data_num{$line_parts[2]}{"count1"} = 0.0;
	   }
	   else
	   { $data_num{$line_parts[2]}{"count"}++ }

	   if ($trip_flag{$line_parts[1]} ne "on")
	   { $data_num{$line_parts[2]}{"count0"}++ }
	   else
	   {
		$data_num{$line_parts[2]}{"count1"}++;
	   	if ($data_num{$line_parts[2]}{"count1"} == 1)
		{ 
			$data_num{$line_parts[2]}{"min"} = $line_parts[3];
			$data_num{$line_parts[2]}{"max"} = $line_parts[3];
			$data_num{$line_parts[2]}{"sum"} = $line_parts[3];
			$data_num{$line_parts[2]}{"sum_sq"} = $line_parts[3]*$line_parts[3];
			$data_num{$line_parts[2]}{"mean"} = $line_parts[3];
			$data_num{$line_parts[2]}{"std"} = 0.0;
		}
		else
		{ 
		 	$data_num{$line_parts[2]}{"min"} = $line_parts[3] if ($data_num{$line_parts[2]}{"min"} > $line_parts[3]);
		 	$data_num{$line_parts[2]}{"max"} = $line_parts[3] if ($data_num{$line_parts[2]}{"max"} < $line_parts[3]);
		 	$data_num{$line_parts[2]}{"sum"} = $data_num{$line_parts[2]}{"sum"} + $line_parts[3];
		 	$data_num{$line_parts[2]}{"sum_sq"} = $data_num{$line_parts[2]}{"sum_sq"} + $line_parts[3]*$line_parts[3];
		 	$data_num{$line_parts[2]}{"mean"} = $data_num{$line_parts[2]}{"sum"}/$data_num{$line_parts[2]}{"count1"};
		 	$data_num{$line_parts[2]}{"std"} = sqrt(abs($data_num{$line_parts[2]}{"sum_sq"}/$data_num{$line_parts[2]}{"count1"}-$data_num{$line_parts[2]}{"mean"}*$data_num{$line_parts[2]}{"mean"}));
		 	$data_num{$line_parts[2]}{"std"} = 0.0 if ($data_num{$line_parts[2]}{"max"} == $data_num{$line_parts[2]}{"min"});
		}
	   }
##################
#	   if (not $data_num{$line_parts[2]}{"count"})
#	   {
#		$data_num{$line_parts[2]}{"count"} = 1.0;
#		if ($trip_flag{$line_parts[1]} ne "on")
#		{ $data_num{$line_parts[2]}{"count0"} = 1.0 }
#		else
#		{ 
#			$data_num{$line_parts[2]}{"count0"} = 0.0;
#			$data_num{$line_parts[2]}{"min"} = $line_parts[3];
#			$data_num{$line_parts[2]}{"max"} = $line_parts[3];
#			$data_num{$line_parts[2]}{"sum"} = $line_parts[3];
#			$data_num{$line_parts[2]}{"sum_sq"} = $line_parts[3]*$line_parts[3];
#			$data_num{$line_parts[2]}{"mean"} = $line_parts[3];
#			$data_num{$line_parts[2]}{"std"} = 0.0;
#		}
#	   }
#	   else
#	   {	
#		$data_num{$line_parts[2]}{"count"}++;
#		if ($trip_flag{$line_parts[1]} ne "on")
#		{ $data_num{$line_parts[2]}{"count0"}++ }
#		else
#		{
#		 $data_num{$line_parts[2]}{"min"} = $line_parts[3] if ($data_num{$line_parts[2]}{"min"} > $line_parts[3]);
#		 $data_num{$line_parts[2]}{"max"} = $line_parts[3] if ($data_num{$line_parts[2]}{"max"} < $line_parts[3]);
#		 $data_num{$line_parts[2]}{"sum"} = $data_num{$line_parts[2]}{"sum"} + $line_parts[3];
#		 $data_num{$line_parts[2]}{"sum_sq"} = $data_num{$line_parts[2]}{"sum_sq"} + $line_parts[3]*$line_parts[3];
#		 $data_num{$line_parts[2]}{"mean"} = $data_num{$line_parts[2]}{"sum"}/($data_num{$line_parts[2]}{"count"} - $data_num{$line_parts[2]}{"count0"});
#		 $data_num{$line_parts[2]}{"std"} = sqrt(abs($data_num{$line_parts[2]}{"sum_sq"}/($data_num{$line_parts[2]}{"count"}-$data_num{$line_parts[2]}{"count0"})-$data_num{$line_parts[2]}{"mean"}*$data_num{$line_parts[2]}{"mean"}));
#		 $data_num{$line_parts[2]}{"std"} = 0.0 if ($data_num{$line_parts[2]}{"max"} == $data_num{$line_parts[2]}{"min"});
#		}
#	   }
##################


	}
	else { if (not $data_str{$line_parts[2]}{$line_parts[3]}) {$data_str{$line_parts[2]}{$line_parts[3]} = 1.0} else {$data_str{$line_parts[2]}{$line_parts[3]}++} }
 } 
};
close(FILE);
#setting up output file
@path_name_parts = split(/\//,$filename);
@file_name_parts = split(/\./,$path_name_parts[$#path_name_parts]);
$file_out = $file_name_parts[0]."_summary_wbpm.txt";
#print "OUTPUT filename : $file_out\n\n";
open(OUTPUT,">$file_out");
print OUTPUT "$prescalars\n" if ($prescalars);
foreach $variable_name (sort keys %data_num) 
{ 
#	print "$variable_name\n";
	if ($data_num{$variable_name}{"count0"} == $data_num{$variable_name}{"count"})
	{
		$data_num{$variable_name}{"min"} = 0.0;
		$data_num{$variable_name}{"max"} = 0.0;
		$data_num{$variable_name}{"mean"} = 0.0;
		$data_num{$variable_name}{"std"} = 0.0;
	}
#$data_num{$variable_name}{"min"}
	$count = $data_num{$variable_name}{"count1"};
#	$count = $data_num{$variable_name}{"count"}."(".$data_num{$variable_name}{"count1"}." + ".$data_num{$variable_name}{"count0"}.")";
	  $min = sprintf( "%.9E" , $data_num{$variable_name}{"min"}   );
	  $max = sprintf( "%.9E" , $data_num{$variable_name}{"max"}   );
	 $mean = sprintf( "%.9E" , $data_num{$variable_name}{"mean"}  );
	  $std = sprintf( "%.9E" , $data_num{$variable_name}{"std"}   );
	print OUTPUT "$variable_name\t$count\t$min\t$max\t$mean\t$std\n"
}
print OUTPUT "-----\n";
# because %data_str is a hash of hashes, $data_str{$variable_name} is a reference to a hash
# therefore $temp is a reference to a hash
#       and %$temp is the hash
#       and $$temp{$variable_value} is the value corresponding to the $variable_value key
foreach $variable_name (sort keys %data_str) 
{
	$temp = $data_str{$variable_name};
	$num_values = scalar keys %$temp;
	print OUTPUT "$variable_name\t$num_values";
	foreach $variable_value (sort keys %$temp) { print OUTPUT "\t$variable_value\t$$temp{$variable_value}"} 
	print OUTPUT "\n";
}
print OUTPUT "-----\n";
print OUTPUT "trip line count=\t$trip_line_count\n";
print OUTPUT "trip line check count=\t$trip_line_check_count\n";
print OUTPUT "trip line found count=\t$trip_line_found_count\n";
print OUTPUT "approx number of trips=\t$number_of_trips\n";
print OUTPUT "approx total run time(sec)=\t$total_run_time\n";
print OUTPUT "approx beam off time (sec)=\t$beam_off_time\n";
print OUTPUT "approx run time after beam cuts (sec) =\t $beam_on_cut_sum\n";
close(OUTPUT);