5 # This script will gather statistical information from a database
6 # containing headers and other information from a INN feed.
8 # It is part of the NewsStats package.
10 # Copyright (c) 2010-2012 Thomas Hochstein <thh@inter.net>
12 # It can be redistributed and/or modified under the same terms under
13 # which Perl itself is published.
16 our $VERSION = "0.01";
18 push(@INC, dirname($0));
22 use NewsStats qw(:DEFAULT :TimePeriods ListNewsgroups ReadGroupList);
25 use Getopt::Long qw(GetOptions);
26 Getopt::Long::config ('bundling');
28 ################################# Definitions ##################################
30 # define types of information that can be gathered
31 # all / groups (/ clients / hosts)
33 @LegalStats{('all','groups')} = ();
35 ################################# Main program #################################
37 ### read commandline options
38 my ($OptCheckgroupsFile,$OptClientsDB,$OptDebug,$OptGroupsDB,$OptTLH,
39 $OptHostsDB,$OptMonth,$OptRawDB,$OptStatsType,$OptTest);
40 GetOptions ('c|checkgroups=s' => \$OptCheckgroupsFile,
41 'clientsdb=s' => \$OptClientsDB,
42 'd|debug!' => \$OptDebug,
43 'groupsdb=s' => \$OptGroupsDB,
44 'hierarchy=s' => \$OptTLH,
45 'hostsdb=s' => \$OptHostsDB,
46 'm|month=s' => \$OptMonth,
47 'rawdb=s' => \$OptRawDB,
48 's|stats=s' => \$OptStatsType,
49 't|test!' => \$OptTest,
50 'h|help' => \&ShowPOD,
51 'V|version' => \&ShowVersion) or exit 1;
53 ### read configuration
54 my %Conf = %{ReadConfig($HomePath.'/newsstats.conf')};
56 ### override configuration via commandline options
58 $ConfOverride{'DBTableRaw'} = $OptRawDB if $OptRawDB;
59 $ConfOverride{'DBTableGrps'} = $OptGroupsDB if $OptGroupsDB;
60 $ConfOverride{'DBTableClnts'} = $OptClientsDB if $OptClientsDB;
61 $ConfOverride{'DBTableHosts'} = $OptHostsDB if $OptHostsDB;
62 $ConfOverride{'TLH'} = $OptTLH if $OptTLH;
63 &OverrideConfig(\%Conf,\%ConfOverride);
65 ### get type of information to gather, defaulting to 'all'
66 $OptStatsType = 'all' if !$OptStatsType;
67 &Bleat(2, sprintf("Unknown type '%s'!", $OptStatsType))
68 if !exists($LegalStats{$OptStatsType});
70 ### get time period from --month
71 # get verbal description of time period, drop SQL code
72 my ($Period) = &GetTimePeriod($OptMonth);
73 &Bleat(2,"--month option has an invalid format - please use 'YYYY-MM' or ".
74 "'YYYY-MM:YYYY-MM'!") if (!$Period or $Period eq 'all time');
76 ### reformat $Conf{'TLH'}
79 # $Conf{'TLH'} is parsed as an array by Config::Auto;
80 # make a flat list again, separated by :
81 if (ref($Conf{'TLH'}) eq 'ARRAY') {
82 $TLH = join(':',@{$Conf{'TLH'}});
88 # add trailing dots if none are present yet
89 # (using negative look-behind assertions)
90 $TLH =~ s/(?<!\.):/.:/g;
91 $TLH =~ s/(?<!\.)$/./;
92 # check for illegal characters
93 &Bleat(2,'Config error - illegal characters in TLH definition!')
94 if ($TLH !~ /^[a-zA-Z0-9:+.-]+$/);
98 # reformat $TLH from a:b to (a)|(b),
99 # e.g. replace ':' by ')|('
101 $TLH = '(' . $TLH . ')';
105 # read list of newsgroups from --checkgroups
107 my %ValidGroups = %{ReadGroupList($OptCheckgroupsFile)} if $OptCheckgroupsFile;
110 my $DBHandle = InitDB(\%Conf,1);
112 ### get data for each month
113 &Bleat(1,'Test mode. Database is not updated.') if $OptTest;
114 foreach my $Month (&ListMonth($Period)) {
116 print "---------- $Month ----------\n" if $OptDebug;
118 if ($OptStatsType eq 'all' or $OptStatsType eq 'groups') {
119 ### ----------------------------------------------
120 ### get groups data (number of postings per group)
121 # get groups data from raw table for given month
122 my $DBQuery = $DBHandle->prepare(sprintf("SELECT newsgroups FROM %s.%s ".
123 "WHERE day LIKE ? AND NOT disregard",
125 $Conf{'DBTableRaw'}));
126 $DBQuery->execute($Month.'-%')
127 or &Bleat(2,sprintf("Can't get groups data for %s from %s.%s: ".
128 "$DBI::errstr\n",$Month,
129 $Conf{'DBDatabase'},$Conf{'DBTableRaw'}));
131 # count postings per group
133 while (($_) = $DBQuery->fetchrow_array) {
134 # get list oft newsgroups and hierarchies from Newsgroups:
135 my %Newsgroups = ListNewsgroups($_,$TLH,
136 $OptCheckgroupsFile ? \%ValidGroups : '');
137 # count each newsgroup and hierarchy once
138 foreach (sort keys %Newsgroups) {
143 # add valid but empty groups if --checkgroups is set
145 foreach (sort keys %ValidGroups) {
146 if (!defined($Postings{$_})) {
148 warn (sprintf("ADDED: %s as empty group\n",$_));
153 # delete old data for that month
155 $DBQuery = $DBHandle->do(sprintf("DELETE FROM %s.%s WHERE month = ?",
156 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}),
158 or &Bleat(2,sprintf("Can't delete old groups data for %s from %s.%s: ".
159 "$DBI::errstr\n",$Month,
160 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}));
163 print "----- GroupStats -----\n" if $OptDebug;
164 foreach my $Newsgroup (sort keys %Postings) {
165 print "$Newsgroup => $Postings{$Newsgroup}\n" if $OptDebug;
168 $DBQuery = $DBHandle->prepare(sprintf("INSERT INTO %s.%s ".
169 "(month,newsgroup,postings) ".
172 $Conf{'DBTableGrps'}));
173 $DBQuery->execute($Month, $Newsgroup, $Postings{$Newsgroup})
174 or &Bleat(2,sprintf("Can't write groups data for %s/%s to %s.%s: ".
175 "$DBI::errstr\n",$Month,$Newsgroup,
176 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}));
181 # other types of information go here - later on
186 $DBHandle->disconnect;
190 ################################ Documentation #################################
194 gatherstats - process statistical data from a raw source
198 B<gatherstats> [B<-Vhdt>] [B<-m> I<YYYY-MM> | I<YYYY-MM:YYYY-MM>] [B<-s> I<stats] [B<-c> I<checkgroups file>]] [B<--hierarchy> I<TLH>] [B<--rawdb> I<database table>] [B<-groupsdb> I<database table>] [B<--clientsdb> I<database table>] [B<--hostsdb> I<database table>]
206 This script will extract and process statistical information from a
207 database table which is fed from F<feedlog.pl> for a given time period
208 and write its results to (an)other database table(s). Entries marked
209 with I<'disregard'> in the database will be ignored; currently, you
210 have to set this flag yourself, using your database management tools.
211 You can exclude erroneous entries that way (e.g. automatic reposts
212 (think of cancels flood and resurrectors); spam; ...).
214 The time period to act on defaults to last month; you can assign
215 another time period or a single month via the B<--month> option (see
218 By default B<gatherstats> will process all types of information; you
219 can change that using the B<--stats> option and assigning the type of
220 information to process. Currently that doesn't matter yet as only
221 processing of the number of postings per group per month is
224 Possible information types include:
228 =item B<groups> (postings per group per month)
230 B<gatherstats> will examine Newsgroups: headers. Crosspostings will be
231 counted for each single group they appear in. Groups not in I<TLH>
234 B<gatherstats> will also add up the number of postings for each
235 hierarchy level, but only count each posting once. A posting to
236 de.alt.test will be counted for de.alt.test, de.alt.ALL and de.ALL,
237 respectively. A crossposting to de.alt.test and de.alt.admin, on the
238 other hand, will be counted for de.alt.test and de.alt.admin each, but
239 only once for de.alt.ALL and de.ALL.
241 Data is written to I<DBTableGrps> (see L<doc/INSTALL>); you can
242 override that default through the B<--groupsdb> option.
248 B<gatherstats> will read its configuration from F<newsstats.conf>
249 which should be present in the same directory via Config::Auto.
251 See L<doc/INSTALL> for an overview of possible configuration options.
253 You can override configuration options via the B<--hierarchy>,
254 B<--rawdb>, B<--groupsdb>, B<--clientsdb> and B<--hostsdb> options,
261 =item B<-V>, B<--version>
263 Print out version and copyright information and exit.
265 =item B<-h>, B<--help>
267 Print this man page and exit.
269 =item B<-d>, B<--debug>
271 Output debugging information to STDOUT while processing (number of
274 =item B<-t>, B<--test>
276 Do not write results to database. You should use B<--debug> in
277 conjunction with B<--test> ... everything else seems a bit pointless.
279 =item B<-m>, B<--month> I<YYYY-MM[:YYYY-MM]>
281 Set processing period to a single month in YYYY-MM format or to a time
282 period between two month in YYYY-MM:YYYY-MM format (two month, separated
286 =item B<-s>, B<--stats> I<type>
288 Set processing type to one of I<all> and I<groups>. Defaults to all
289 (and is currently rather pointless as only I<groups> has been
292 =item B<-c>, B<--checkgroups> I<filename>
294 Check each group against a list of valid newsgroups read from
295 I<filename>, one group on each line and ignoring everything after the
296 first whitespace (so you can use a file in checkgroups format or (part
297 of) your INN active file).
299 Newsgroups not found in I<filename> will be dropped (and logged to
300 STDERR), and newsgroups found in I<filename> but having no postings
301 will be added with a count of 0 (and logged to STDERR).
303 =item B<--hierarchy> I<TLH> (newsgroup hierarchy)
305 Override I<TLH> from F<newsstats.conf>.
307 =item B<--rawdb> I<table> (raw data table)
309 Override I<DBTableRaw> from F<newsstats.conf>.
311 =item B<--groupsdb> I<table> (postings per group table)
313 Override I<DBTableGrps> from F<newsstats.conf>.
315 =item B<--clientsdb> I<table> (client data table)
317 Override I<DBTableClnts> from F<newsstats.conf>.
319 =item B<--hostsdb> I<table> (host data table)
321 Override I<DBTableHosts> from F<newsstats.conf>.
331 Process all types of information for lasth month:
335 Do a dry run, showing results of processing:
337 gatherstats --debug --test
339 Process all types of information for January of 2010:
341 gatherstats --month 2010-01
343 Process only number of postings for the year of 2010,
344 checking against checkgroups-2010.txt:
346 gatherstats -m 2010-01:2010-12 -s groups -c checkgroups-2010.txt
352 =item F<gatherstats.pl>
356 =item F<NewsStats.pm>
358 Library functions for the NewsStats package.
360 =item F<newsstats.conf>
362 Runtime configuration file.
368 Please report any bugs or feature requests to the author or use the
369 bug tracker at L<http://bugs.th-h.de/>!
385 This script is part of the B<NewsStats> package.
389 Thomas Hochstein <thh@inter.net>
391 =head1 COPYRIGHT AND LICENSE
393 Copyright (c) 2010-2012 Thomas Hochstein <thh@inter.net>
395 This program is free software; you may redistribute it and/or modify it
396 under the same terms as Perl itself.