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($0));
22 use NewsStats qw(:DEFAULT :TimePeriods ListNewsgroups ParseHierarchies ReadGroupList);
25 use Getopt::Long qw(GetOptions);
26 Getopt::Long::config ('bundling');
28 ################################# Definitions ##################################
30 # define types of information that can be gathered
31 # all / groups (/ clients / hosts)
33 @LegalStats{('all','groups')} = ();
35 ################################# Main program #################################
37 ### read commandline options
38 my ($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 . ')';
106 my $DBHandle = InitDB(\%Conf,1);
108 ### get data for each month
109 &Bleat(1,'Test mode. Database is not updated.') if $OptTest;
110 foreach my $Month (&ListMonth($Period)) {
112 print "---------- $Month ----------\n" if $OptDebug;
114 if ($OptStatsType eq 'all' or $OptStatsType eq 'groups') {
115 # read list of newsgroups from --checkgroups
117 my %ValidGroups = %{ReadGroupList(sprintf('%s-%s',$OptCheckgroupsFile,$Month))}
118 if $OptCheckgroupsFile;
120 ### ----------------------------------------------
121 ### get groups data (number of postings per group)
122 # get groups data from raw table for given month
123 my $DBQuery = $DBHandle->prepare(sprintf("SELECT newsgroups FROM %s.%s ".
124 "WHERE day LIKE ? AND NOT disregard",
126 $Conf{'DBTableRaw'}));
127 $DBQuery->execute($Month.'-%')
128 or &Bleat(2,sprintf("Can't get groups data for %s from %s.%s: ".
129 "$DBI::errstr\n",$Month,
130 $Conf{'DBDatabase'},$Conf{'DBTableRaw'}));
132 # count postings per group
134 while (($_) = $DBQuery->fetchrow_array) {
135 # get list of newsgroups and hierarchies from Newsgroups:
136 my %Newsgroups = ListNewsgroups($_,$TLH,
137 $OptCheckgroupsFile ? \%ValidGroups : '');
138 # count each newsgroup and hierarchy once
139 foreach (sort keys %Newsgroups) {
144 # add valid but empty groups if --checkgroups is set
146 foreach (sort keys %ValidGroups) {
147 if (!defined($Postings{$_})) {
148 # add current newsgroup as empty group
150 warn (sprintf("ADDED: %s as empty group\n",$_));
151 # add empty hierarchies for current newsgroup as needed
152 foreach (ParseHierarchies($_)) {
153 my $Hierarchy = $_ . '.ALL';
154 if (!defined($Postings{$Hierarchy})) {
155 $Postings{$Hierarchy} = 0;
156 warn (sprintf("ADDED: %s as empty group\n",$Hierarchy));
163 # delete old data for that month
165 $DBQuery = $DBHandle->do(sprintf("DELETE FROM %s.%s WHERE month = ?",
166 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}),
168 or &Bleat(2,sprintf("Can't delete old groups data for %s from %s.%s: ".
169 "$DBI::errstr\n",$Month,
170 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}));
173 print "----- GroupStats -----\n" if $OptDebug;
174 foreach my $Newsgroup (sort keys %Postings) {
175 print "$Newsgroup => $Postings{$Newsgroup}\n" if $OptDebug;
178 $DBQuery = $DBHandle->prepare(sprintf("INSERT INTO %s.%s ".
179 "(month,newsgroup,postings) ".
182 $Conf{'DBTableGrps'}));
183 $DBQuery->execute($Month, $Newsgroup, $Postings{$Newsgroup})
184 or &Bleat(2,sprintf("Can't write groups data for %s/%s to %s.%s: ".
185 "$DBI::errstr\n",$Month,$Newsgroup,
186 $Conf{'DBDatabase'},$Conf{'DBTableGrps'}));
191 # other types of information go here - later on
196 $DBHandle->disconnect;
200 ################################ Documentation #################################
204 gatherstats - process statistical data from a raw source
208 B<gatherstats> [B<-Vhdt>] [B<-m> I<YYYY-MM> | I<YYYY-MM:YYYY-MM>] [B<-s> I<stats] [B<-c> I<filename template>]] [B<--hierarchy> I<TLH>] [B<--rawdb> I<database table>] [B<-groupsdb> I<database table>] [B<--clientsdb> I<database table>] [B<--hostsdb> I<database table>]
216 This script will extract and process statistical information from a
217 database table which is fed from F<feedlog.pl> for a given time period
218 and write its results to (an)other database table(s). Entries marked
219 with I<'disregard'> in the database will be ignored; currently, you
220 have to set this flag yourself, using your database management tools.
221 You can exclude erroneous entries that way (e.g. automatic reposts
222 (think of cancels flood and resurrectors); spam; ...).
224 The time period to act on defaults to last month; you can assign
225 another time period or a single month via the B<--month> option (see
228 By default B<gatherstats> will process all types of information; you
229 can change that using the B<--stats> option and assigning the type of
230 information to process. Currently that doesn't matter yet as only
231 processing of the number of postings per group per month is
234 Possible information types include:
238 =item B<groups> (postings per group per month)
240 B<gatherstats> will examine Newsgroups: headers. Crosspostings will be
241 counted for each single group they appear in. Groups not in I<TLH>
244 B<gatherstats> will also add up the number of postings for each
245 hierarchy level, but only count each posting once. A posting to
246 de.alt.test will be counted for de.alt.test, de.alt.ALL and de.ALL,
247 respectively. A crossposting to de.alt.test and de.alt.admin, on the
248 other hand, will be counted for de.alt.test and de.alt.admin each, but
249 only once for de.alt.ALL and de.ALL.
251 Data is written to I<DBTableGrps> (see L<doc/INSTALL>); you can
252 override that default through the B<--groupsdb> option.
258 B<gatherstats> will read its configuration from F<newsstats.conf>
259 which should be present in the same directory via Config::Auto.
261 See L<doc/INSTALL> for an overview of possible configuration options.
263 You can override configuration options via the B<--hierarchy>,
264 B<--rawdb>, B<--groupsdb>, B<--clientsdb> and B<--hostsdb> options,
271 =item B<-V>, B<--version>
273 Print out version and copyright information and exit.
275 =item B<-h>, B<--help>
277 Print this man page and exit.
279 =item B<-d>, B<--debug>
281 Output debugging information to STDOUT while processing (number of
284 =item B<-t>, B<--test>
286 Do not write results to database. You should use B<--debug> in
287 conjunction with B<--test> ... everything else seems a bit pointless.
289 =item B<-m>, B<--month> I<YYYY-MM[:YYYY-MM]>
291 Set processing period to a single month in YYYY-MM format or to a time
292 period between two month in YYYY-MM:YYYY-MM format (two month, separated
296 =item B<-s>, B<--stats> I<type>
298 Set processing type to one of I<all> and I<groups>. Defaults to all
299 (and is currently rather pointless as only I<groups> has been
302 =item B<-c>, B<--checkgroups> I<filename template>
304 Check each group against a list of valid newsgroups read from a file,
305 one group on each line and ignoring everything after the first
306 whitespace (so you can use a file in checkgroups format or (part of)
307 your INN active file).
309 The filename is taken from I<filename template>, amended by each B<--
310 month> B<gatherstats> is processing, so that
312 gatherstats -m 2010-01:2010-12 -c checkgroups
314 will check against F<checkgroups-2010-01> for January 2010, against
315 F<checkgroups-2010-02> for February 2010 and so on.
317 Newsgroups not found in the checkgroups file will be dropped (and
318 logged to STDERR), and newsgroups found there but having no postings
319 will be added with a count of 0 (and logged to STDERR).
321 =item B<--hierarchy> I<TLH> (newsgroup hierarchy)
323 Override I<TLH> from F<newsstats.conf>.
325 =item B<--rawdb> I<table> (raw data table)
327 Override I<DBTableRaw> from F<newsstats.conf>.
329 =item B<--groupsdb> I<table> (postings per group table)
331 Override I<DBTableGrps> from F<newsstats.conf>.
333 =item B<--clientsdb> I<table> (client data table)
335 Override I<DBTableClnts> from F<newsstats.conf>.
337 =item B<--hostsdb> I<table> (host data table)
339 Override I<DBTableHosts> from F<newsstats.conf>.
349 Process all types of information for lasth month:
353 Do a dry run, showing results of processing:
355 gatherstats --debug --test
357 Process all types of information for January of 2010:
359 gatherstats --month 2010-01
361 Process only number of postings for the year of 2010,
362 checking against checkgroups-*:
364 gatherstats -m 2010-01:2010-12 -s groups -c checkgroups
370 =item F<gatherstats.pl>
374 =item F<NewsStats.pm>
376 Library functions for the NewsStats package.
378 =item F<newsstats.conf>
380 Runtime configuration file.
386 Please report any bugs or feature requests to the author or use the
387 bug tracker at L<http://bugs.th-h.de/>!
403 This script is part of the B<NewsStats> package.
407 Thomas Hochstein <thh@inter.net>
409 =head1 COPYRIGHT AND LICENSE
411 Copyright (c) 2010-2012 Thomas Hochstein <thh@inter.net>
413 This program is free software; you may redistribute it and/or modify it
414 under the same terms as Perl itself.