#!/usr/bin/perl -w
#	get_all_epics.perl [filename]
#
# 	by: Jaideep Singh
#
#	date: 12/17/2005 
#
#	about:	reads a raw data file and looks for all possible EPICS variables
#		creates a file:
#				(1) the first line is the prescalar declaration for that run:
#
#					ps1=2,ps2=3,ps3=9999,ps4=9999,ps5=65535,ps6=65535,ps7=65535,ps8=65534
#
#				(2) contains 4 columns which contain the timestamp, epics variable name, and epics value:
#
#					timestamp<tab>seconds from Jan 1 2003<tab>EPICS variable name<tab>data value
#					2003-08-16 15:41:56     19669316        IPM1H04B.XPOS   -0.0408508
#					
#					Note that some EPICS-like stuff still leaks through...
#
#	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

use strict "subs";

sub jtime
        { ($sec, $min, $hr, $day, $mon) =  @_;
#          1041397200 in unix time, the number of seconds from midnight Jan 1 1970 to Jan 1 2003
#         (31,28,31,30,31,30,31,31,30,31,30,31);
          @mon2day = (0,31,59,90,120,151,181,212,243,273,304,334);
          $totsec = $sec + 60*($min + 60*($hr + 24*($day -1 + $mon2day[$mon-1])));
          return $totsec;
        };

#month look up hash
%mon2num = (
           "January"   => "01",
           "February"  => "02",
           "March"     => "03",
           "April"     => "04",
           "May"       => "05",
           "June"      => "06",
           "July"      => "07",
           "August"    => "08",
           "September" => "09",
           "October"   => "10",
           "November"  => "11",
           "December"  => "12",
           "Jan" => "01",
           "Feb" => "02",
           "Mar" => "03",
           "Apr" => "04",
           "May" => "05",
           "Jun" => "06",
           "Jul" => "07",
           "Aug" => "08",
           "Sep" => "09",
           "Oct" => "10",
           "Nov" => "11",
           "Dec" => "12"
           );

#getting name of file to search + ARGUMENT ERROR HANDLING
$filename = $ARGV[0];
print "Uh, what file do you want me to look at dude?: Too few arguments\n" and die unless ($filename);
print "Scared and confused: Too many arguments\n" and die unless (($#ARGV+1) < 2);
open(DIE_FILE,">>dead_file_list.txt");
open(FILE,$filename) or print DIE_FILE "$filename\n" and die "d'oh!: - $filename - not found\n";
#print "\nSearching raw data file:\t$filename\n";

#initializing variables
$count = 0;
$line_count = 0;

#searching line by line, grabbing the timestamp before and after searching
#print "\nstart\t";
#print scalar localtime;
#print "\n";
while(<FILE>)
{
 $line_count++;
 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;
 }
#                                            1                     2      3      4      5              6
#					    Mon			  Day    Hour   Min    Sec            Year
 if (/[MTWFS][ouehra][neduit]\s+([JFMASOND][aepuco][nbrylgptvc])\s+(\d+)\s+(\d+)\:(\d+)\:(\d+)\s+E[DS]T\s+(2003)/) 
 {
	chomp;
        $current_year = $6;
       	$current_month = $mon2num{$1};
        $current_day = $2;
        $current_day = "0".$2 unless (length($current_day) == 2);
        $current_hour = $3;
        $current_hour = "0".$2 unless (length($current_hour) == 2);
      	$current_minute = $4;
     	$current_minute = "0".$2 unless (length($current_minute) == 2);
     	$current_second = $5;
      	$current_second = "0".$2 unless (length($current_second) == 2);
  	$current_time_stamp = $current_year."-".$current_month."-".$current_day." ".$current_hour."\:".$current_minute."\:".$current_second;
	$current_number_of_seconds =  jtime($current_second,$current_minute,$current_hour,$current_day,$current_month);
#	print "\n $_ -----> $current_time_stamp \n" and die;
 } 
 if (/^([\w\-.:+]{5,30})(\s{1,26})([\w+\-.]{1}[\w+\-.\s:]{0,19})$/)
 {
  if (length($_) < 52)
  { 
   $count++; 
   $data{$1.$current_time_stamp}[0] = $current_time_stamp;
   $data{$1.$current_time_stamp}[1] = $current_number_of_seconds;
   $data{$1.$current_time_stamp}[2] = $1;
   $data{$1.$current_time_stamp}[3] = $3;
   chomp $data{$1.$current_time_stamp}[3];
  }
 }
};
close(FILE);
#print "end\t";
#print scalar localtime;
#print "\n\nFound $count total instances of possible EPICS variables\n\n";

#setting up output file
@path_name_parts = split(/\//,$filename);
@file_name_parts = split(/\./,$path_name_parts[$#path_name_parts]);
$file_out = $file_name_parts[0]."_".$file_name_parts[$#file_name_parts]."_epics_data.txt";
#print "OUTPUT filename : $file_out\n\n";
open(OUTPUT,">$file_out");
print OUTPUT "$prescalars\n";
foreach $variable_name (sort keys %data) 
{ 
	print OUTPUT "$data{$variable_name}[0]\t$data{$variable_name}[1]\t$data{$variable_name}[2]\t$data{$variable_name}[3]\n";
}
close(OUTPUT);
close(DIE_FILE);