#! PERL_COMMAND -w # $Cambridge: exim/src/src/exigrep.src,v 1.9 2007/03/13 16:37:57 ph10 Exp $ use strict; # Copyright (c) 2007 University of Cambridge. # See the file NOTICE for conditions of use and distribution. # Except when they appear in comments, the following placeholders in this # source are replaced when it is turned into a runnable script: # # PERL_COMMAND # ZCAT_COMMAND # COMPRESS_SUFFIX # PROCESSED_FLAG # This is a perl script which extracts from an Exim log all entries # for all messages that have an entry that matches a given pattern. # If *any* entry for a particular message matches the pattern, *all* # entries for that message are displayed. # We buffer up information on a per-message basis. It is done this way rather # than reading the input twice so that the input can be a pipe. # There must be one argument, which is the pattern. Subsequent arguments # are the files to scan; if none, the standard input is read. If any file # appears to be compressed, it is passed through zcat. We can't just do this # for all files, because zcat chokes on non-compressed files. # Performance optimized in 02/02/2007 by Jori Hamalainen # Typical run time acceleration: 4 times use Getopt::Std qw(getopts); use POSIX qw(mktime); # This subroutine converts a time/date string from an Exim log line into # the number of seconds since the epoch. It handles optional timezone # information. sub seconds { my($year,$month,$day,$hour,$min,$sec,$tzs,$tzh,$tzm) = $_[0] =~ /^(\d{4})-(\d\d)-(\d\d)\s(\d\d):(\d\d):(\d\d)(?>\s([+-])(\d\d)(\d\d))?/o; my $seconds = mktime $sec, $min, $hour, $day, $month - 1, $year - 1900; if (defined $tzs) { $seconds -= $tzh * 3600 + $tzm * 60 if $tzs eq "+"; $seconds += $tzh * 3600 + $tzm * 60 if $tzs eq "-"; } return $seconds; } # This subroutine processes a single line (in $_) from a log file. Program # defensively against short lines finding their way into the log. my (%saved, %id_list, $pattern, $queue_time, $insensitive, $invert); sub do_line { # Convert syslog lines to mainlog format, as in eximstats. if (!/^\d{4}-/o) { $_ =~ s/^.*? exim\b.*?: //o; } return unless my($date,$id) = /^(\d{4}-\d\d-\d\d \d\d:\d\d:\d\d (?:[+-]\d{4} )?)(?:\[\d+\] )?(\w{6}\-\w{6}\-\w{2})?/o; # Handle the case when the log line belongs to a specific message. We save # lines for specific messages until the message is complete. Then either print # or discard. if (defined $id) { $saved{$id} = '' unless defined($saved{$id}); # Save up the data for this message in case it becomes interesting later. $saved{$id} .= $_; # Are we interested in this id ? Short circuit if we already were interested. if ($invert) { $id_list{$id} = 1 if (!defined($id_list{$id})); $id_list{$id} = 0 if (($insensitive && /$pattern/io) || /$pattern/o); } else { $id_list{$id} = 1 if defined $id_list{$id} || ($insensitive && /$pattern/io) || /$pattern/o; } # See if this is a completion for some message. If it is interesting, # print it, but in any event, throw away what was saved. if (index($_, 'Completed') != -1 || index($_, 'SMTP data timeout') != -1 || (index($_, 'rejected') != -1 && /^(\d{4}-\d\d-\d\d \d\d:\d\d:\d\d (?:[+-]\d{4} )?)(?:\[\d+\] )?\w{6}\-\w{6}\-\w{2} rejected/o)) { if ($queue_time != -1 && $saved{$id} =~ /^(\d{4}-\d\d-\d\d \d\d:\d\d:\d\d ([+-]\d{4} )?)/o) { my $old_sec = &seconds($1); my $sec = &seconds($date); $id_list{$id} = 0 if $id_list{$id} && $sec - $old_sec <= $queue_time; } print "$saved{$id}\n" if ($id_list{$id}); delete $id_list{$id}; delete $saved{$id}; } } # Handle the case where the log line does not belong to a specific message. # Print it if it is interesting. elsif ( ($invert && (($insensitive && !/$pattern/io) || !/$pattern/o)) || (!$invert && (($insensitive && /$pattern/io) || /$pattern/o)) ) { print "$_\n"; } } # The main program. Extract the pattern and make sure any relevant characters # are quoted if the -l flag is given. The -t flag gives a time-on-queue value # which is an additional condition. getopts('Ilvt:',\my %args); $queue_time = $args{'t'}? $args{'t'} : -1; $insensitive = $args{'I'}? 0 : 1; $invert = $args{'v'}? 1 : 0; die "usage: exigrep [-I] [-l] [-t ] [-v] []...\n" if ($#ARGV < 0); $pattern = shift @ARGV; $pattern = quotemeta $pattern if $args{l}; # If file arguments are given, open each one and process according as it is # is compressed or not. if (@ARGV) { foreach (@ARGV) { my $filename = $_; if ($filename =~ /\.(?:COMPRESS_SUFFIX)$/o) { open(LOG, "ZCAT_COMMAND $filename |") || die "Unable to zcat $filename: $!\n"; } else { open(LOG, "<$filename") || die "Unable to open $filename: $!\n"; } do_line() while (); close(LOG); } } # If no files are named, process STDIN only else { do_line() while (); } # At the end of processing all the input, print any uncompleted messages. for (keys %id_list) { print "+++ $_ has not completed +++\n$saved{$_}\n"; } # End of exigrep