X-Git-Url: https://code.th-h.de/?p=usenet%2Fnewsstats.git;a=blobdiff_plain;f=feedlog.pl;h=61311ac35d83c063c899c514b900e676d67ae5fc;hp=a68b833ec88fc33d6953ecbc89a43142596cd1b8;hb=d3b6810d3d09c7a1dffaef75731624fd9712d914;hpb=741336c210429f37bebfd9882b9461e824320cd0 diff --git a/feedlog.pl b/feedlog.pl index a68b833..61311ac 100755 --- a/feedlog.pl +++ b/feedlog.pl @@ -1,87 +1,227 @@ -#! /usr/bin/perl -W -# -# feedlog.pl -# -# This script will log headers and other data to a database -# for further analysis by parsing a feed from INN. -# -# It is part of the NewsStats package. -# -# Copyright (c) 2010 Thomas Hochstein -# -# It can be redistributed and/or modified under the same terms under -# which Perl itself is published. - -BEGIN { - our $VERSION = "0.01"; - use File::Basename; - push(@INC, dirname($0)); -} -use strict; - -use NewsStats; - -use Sys::Syslog qw(:standard :macros); - -use Date::Format; -use DBI; - -################################# Main program ################################# - -### read commandline options -my %Options = &ReadOptions('qd'); - -### read configuration -my %Conf = %{ReadConfig('newsstats.conf')}; - -### init syslog -openlog($MySelf, 'nofatal,pid', LOG_NEWS); -syslog(LOG_NOTICE, "$MyVersion starting up.") if !$Options{'q'}; - -### init database -my $DBHandle = InitDB(\%Conf,0); -if (!$DBHandle) { - syslog(LOG_CRIT, 'Database connection failed: %s', $DBI::errstr); - while (1) {}; # go into endless loop to suppress further errors and respawning -}; -my $DBQuery = $DBHandle->prepare(sprintf("INSERT INTO %s.%s (day,date,mid,timestamp,token,size,peer,path,newsgroups,headers) VALUES (?,?,?,?,?,?,?,?,?,?)",$Conf{'DBDatabase'},$Conf{'DBTableRaw'})); - -### main loop -while (<>) { - chomp; - # catch empty lines trailing or leading - if ($_ eq '') { - next; - } - # first line contains: mid, timestamp, token, size, peer, Path, Newsgroups - my ($Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups) = split; - # remaining lines contain headers - my $Headers = ""; - while (<>) { - chomp; - # empty line terminates this article - if ($_ eq '') { - last; - } - # collect headers - $Headers .= $_."\n" ; - } - - # parse timestamp to day (YYYY-MM-DD) and to MySQL timestamp - my $Day = time2str("%Y-%m-%d", $Timestamp); - my $Date = time2str("%Y-%m-%d %H:%M:%S", $Timestamp); - - # write to database - if (!$DBQuery->execute($Day, $Date, $Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups, $Headers)) { - syslog(LOG_ERR, 'Database error: %s', $DBI::errstr); - }; - $DBQuery->finish; - - warn sprintf("-----\nDay: %s\nDate: %s\nMID: %s\nTS: %s\nToken: %s\nSize: %s\nPeer: %s\nPath: %s\nNewsgroups: %s\nHeaders: %s\n",$Day, $Date, $Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups, $Headers) if !$Options{'d'}; -} - -### close handles -$DBHandle->disconnect; -syslog(LOG_NOTICE, "$MySelf closing down.") if !$Options{'q'}; -closelog(); - +#! /usr/bin/perl -W +# +# feedlog.pl +# +# This script will log headers and other data to a database +# for further analysis by parsing a feed from INN. +# +# It is part of the NewsStats package. +# +# Copyright (c) 2010 Thomas Hochstein +# +# It can be redistributed and/or modified under the same terms under +# which Perl itself is published. + +BEGIN { + our $VERSION = "0.01"; + use File::Basename; + push(@INC, dirname($0)); +} +use strict; + +use NewsStats; + +use Sys::Syslog qw(:standard :macros); + +use Date::Format; +use DBI; + +################################# Main program ################################# + +### read commandline options +my %Options = &ReadOptions('qd'); + +### read configuration +my %Conf = %{ReadConfig('newsstats.conf')}; + +### init syslog +openlog($MySelf, 'nofatal,pid', LOG_NEWS); +syslog(LOG_NOTICE, "$MyVersion starting up.") if !$Options{'q'}; + +### init database +my $DBHandle = InitDB(\%Conf,0); +if (!$DBHandle) { + syslog(LOG_CRIT, 'Database connection failed: %s', $DBI::errstr); + while (1) {}; # go into endless loop to suppress further errors and respawning +}; +my $DBQuery = $DBHandle->prepare(sprintf("INSERT INTO %s.%s (day,date,mid,timestamp,token,size,peer,path,newsgroups,headers) VALUES (?,?,?,?,?,?,?,?,?,?)",$Conf{'DBDatabase'},$Conf{'DBTableRaw'})); + +### main loop +while (<>) { + chomp; + # catch empty lines trailing or leading + if ($_ eq '') { + next; + } + # first line contains: mid, timestamp, token, size, peer, Path, Newsgroups + my ($Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups) = split; + # remaining lines contain headers + my $Headers = ""; + while (<>) { + chomp; + # empty line terminates this article + if ($_ eq '') { + last; + } + # collect headers + $Headers .= $_."\n" ; + } + + # parse timestamp to day (YYYY-MM-DD) and to MySQL timestamp + my $Day = time2str("%Y-%m-%d", $Timestamp); + my $Date = time2str("%Y-%m-%d %H:%M:%S", $Timestamp); + + # write to database + if (!$DBQuery->execute($Day, $Date, $Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups, $Headers)) { + syslog(LOG_ERR, 'Database error: %s', $DBI::errstr); + }; + $DBQuery->finish; + + warn sprintf("-----\nDay: %s\nDate: %s\nMID: %s\nTS: %s\nToken: %s\nSize: %s\nPeer: %s\nPath: %s\nNewsgroups: %s\nHeaders: %s\n",$Day, $Date, $Mid, $Timestamp, $Token, $Size, $Peer, $Path, $Newsgroups, $Headers) if $Options{'d'}; +} + +### close handles +$DBHandle->disconnect; +syslog(LOG_NOTICE, "$MySelf closing down.") if !$Options{'q'}; +closelog(); + +__END__ + +################################ Documentation ################################# + +=head1 NAME + +feedlog - log data from an INN feed to a database + +=head1 SYNOPSIS + +B [B<-Vhdq>] + +=head1 REQUIREMENTS + +See doc/README: Perl 5.8.x itself and the following modules from CPAN: + +=over 2 + +=item - + +Config::Auto + +=item - + +Date::Format + +=item - + +DBI + +=back + +=head1 DESCRIPTION + +This script will log overview data and complete headers to a database +table for further examination by parsing a feed from INN. It will +parse that information and write it to a mysql database table in real +time. + +All reporting is done to I via I facility. If B +fails to initiate a database connection at startup, it will log to +I with I priority and go in an endless loop, as +terminating would only result in a rapid respawn. + +=head2 Configuration + +F will read its configuration from F which +should be present in the same directory via Config::Auto. + +See doc/INSTALL for an overview of possible configuration options. + +=head1 OPTIONS + +=over 3 + +=item B<-V> (version) + +Print out version and copyright information on B and exit. + +=item B<-h> (help) + +Print this man page and exit. + +=item B<-d> (debug) + +Output debugging information to STDERR while parsing STDIN. You'll +find that information most probably in your B F file. + +=item B<-q> (quiet) + +Suppress logging to syslog. + +=back + +=head1 INSTALLATION + +See doc/INSTALL. + +=head1 EXAMPLES + +Set up a feed like that in your B F file: + + ## gather statistics for NewsStats + newsstats! + :!*,de.* + :Tc,WmtfbsPNH,Ac:/path/to/feedlog.pl + +See doc/INSTALL for further information. + +=head1 FILES + +=over 4 + +=item F + +The script itself. + +=item F + +Library functions for the NewsStats package. + +=item F + +Runtime configuration file for B. + +=back + +=head1 BUGS + +Please report any bugs or feature requests to the author or use the +bug tracker at L! + +=head1 SEE ALSO + +=over 2 + +=item - + +doc/README + +=item - + +doc/INSTALL + +=back + +This script is part of the B package. + +=head1 AUTHOR + +Thomas Hochstein + +=head1 COPYRIGHT AND LICENSE + +Copyright (c) 2010 Thomas Hochstein + +This program is free software; you may redistribute it and/or modify it +under the same terms as Perl itself. + +=cut