X-Git-Url: https://code.th-h.de/?p=usenet%2Fnewsstats.git;a=blobdiff_plain;f=groupstats.pl;h=fcacc30973f43a88916b197fad96e5c584cf857b;hp=b1e31e93c4cdc8e90bdbb7795ada58d6f121a5a1;hb=3430c898683771c0ecba9b53a55bb817da0e8fe8;hpb=610b5ef4928b81ab4eecbcdf2ac38197614f0953 diff --git a/groupstats.pl b/groupstats.pl index b1e31e9..fcacc30 100755 --- a/groupstats.pl +++ b/groupstats.pl @@ -23,10 +23,6 @@ use NewsStats qw(:DEFAULT :TimePeriods :Output :SQLHelper); use DBI; -################################# Definitions ################################## - -# ... - ################################# Main program ################################# ### read commandline options @@ -75,13 +71,9 @@ if ($Options{'l'}) { warn ("$MySelf: W: Output type forced to '-o pretty' due to usage of '-l'.\n"); }; -### get query type, default to 'postings' -#die "$MySelf: E: Unknown query type -q $Options{'q'}!\n" if ($Options{'q'} and !exists($LegalTypes{$Options{'q'}})); -#die "$MySelf: E: You must submit a threshold ('-t') for query type '-q $Options{'q'}'!\n" if ($Options{'q'} and !$Options{'t'}); - ### get time period my ($StartMonth,$EndMonth) = &GetTimePeriod($Options{'m'},$Options{'p'}); -# reset to one month for 'dump' type +# reset to one month for 'dump' output type if ($Options{'o'} eq 'dump' and $Options{'p'}) { $StartMonth = $EndMonth; warn ("$MySelf: W: You cannot combine time periods (-p) with '-o dump'. Month was set to $StartMonth.\n"); @@ -90,80 +82,110 @@ if ($Options{'o'} eq 'dump' and $Options{'p'}) { ### init database my $DBHandle = InitDB(\%Conf,1); -### get data +### create report # get list of newsgroups (-n) my ($QueryPart,@GroupList); my $Newsgroups = $Options{'n'}; if ($Newsgroups) { + # explode list of newsgroups for WHERE clause ($QueryPart,@GroupList) = &SQLGroupList($Newsgroups); } else { + # set to dummy value (always true) $QueryPart = 1; }; # manage thresholds if (defined($Options{'t'})) { if ($Options{'i'}) { + # -i: list groups below threshold $QueryPart .= ' AND postings < ?'; } else { + # default: list groups above threshold $QueryPart .= ' AND postings > ?'; }; + # push threshold to GroupList to match number of binding vars for DBQuery->execute push @GroupList,$Options{'t'}; } # construct WHERE clause +# $QueryPart is "list of newsgroup" (or 1), +# &SQLHierarchies() takes care of the exclusion of hierarchy levels (.ALL) +# according to setting of -s my $WhereClause = sprintf('month BETWEEN ? AND ? AND %s %s',$QueryPart,&SQLHierarchies($Options{'s'})); # get lenght of longest newsgroup delivered by query for formatting purposes +# FIXME my $MaxLength = &GetMaxLenght($DBHandle,$Conf{'DBTableGrps'},'newsgroup',$WhereClause,$StartMonth,$EndMonth,@GroupList); my ($OrderClause,$DBQuery); -# -b (best of) defined? +# -b (best of / top list) defined? if (!defined($Options{'b'}) and !defined($Options{'l'})) { + # default: neither -b nor -l + # set ordering (ORDER BY) to "newsgroups" or "postings", "ASC" or "DESC" + # according to -q and -d $OrderClause = 'newsgroup'; $OrderClause = 'postings' if $Options{'q'}; $OrderClause .= ' DESC' if $Options{'d'}; - # do query: get number of postings per group from groups table for given months and newsgroups + # prepare query: get number of postings per group from groups table for given months and newsgroups $DBQuery = $DBHandle->prepare(sprintf("SELECT month,newsgroup,postings FROM %s.%s WHERE %s ORDER BY month,%s",$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$WhereClause,$OrderClause)); } elsif ($Options{'b'}) { + # -b is set (then -l can't be!) # set sorting order (-i) if ($Options{'i'}) { $OrderClause = 'postings'; } else { $OrderClause = 'postings DESC'; }; - # push LIMIT to GroupList to match number of binding vars + # push LIMIT to GroupList to match number of binding vars for DBQuery->execute push @GroupList,$Options{'b'}; - # do query: get sum of postings per group from groups table for given months and newsgroups with LIMIT + # prepare query: get sum of postings per group from groups table for given months and newsgroups with LIMIT $DBQuery = $DBHandle->prepare(sprintf("SELECT newsgroup,SUM(postings) AS postings FROM %s.%s WHERE %s GROUP BY newsgroup ORDER BY %s,newsgroup LIMIT ?",$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$WhereClause,$OrderClause)); -} else { # -l +} else { + # -l must be set now, as all other cases have been taken care of # set sorting order (-i) if ($Options{'i'}) { $OrderClause = '<'; } else { $OrderClause = '>'; }; - # push level and $StartMonth,$EndMonth - again - to GroupList to match number of binding vars + # push level and $StartMonth,$EndMonth - again - to GroupList to match number of binding vars for DBQuery->execute + # FIXME -- together with the query (see below) push @GroupList,$Options{'l'}; push @GroupList,$StartMonth,$EndMonth; - # do query: get number of postings per group from groups table for given months and + # prepare query: get number of postings per group from groups table for given months and + # FIXME -- this query is ... in dire need of impromevent $DBQuery = $DBHandle->prepare(sprintf("SELECT month,newsgroup,postings FROM %s.%s WHERE newsgroup IN (SELECT newsgroup FROM %s.%s WHERE %s GROUP BY newsgroup HAVING MAX(postings) %s ?) AND %s ORDER BY newsgroup,month",$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$WhereClause,$OrderClause,$WhereClause)); }; # execute query -$DBQuery->execute($StartMonth,$EndMonth,@GroupList) or die sprintf("$MySelf: E: Can't get groups data for %s to %s from %s.%s: %s\n",$StartMonth,$EndMonth,$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$DBI::errstr); +$DBQuery->execute($StartMonth,$EndMonth,@GroupList) + or die sprintf("$MySelf: E: Can't get groups data for %s to %s from %s.%s: %s\n",$StartMonth,$EndMonth,$Conf{'DBDatabase'},$Conf{'DBTableGrps'},$DBI::errstr); -# output result +# output results +# print caption (-c) with time period if -m or -p is set +# FIXME - month or period should handled differently printf ("----- Report from %s to %s\n",$StartMonth,$EndMonth) if $Options{'c'} and ($Options{'m'} or $Options{'p'}); +# print caption (-c) with newsgroup list if -n is set printf ("----- Newsgroups: %s\n",join(',',split(/:/,$Newsgroups))) if $Options{'c'} and $Options{'n'}; +# print caption (-c) with threshold if -t is set, taking -i in account printf ("----- Threshold: %s %u\n",$Options{'i'} ? '<' : '>',$Options{'t'}) if $Options{'c'} and $Options{'t'}; if (!defined($Options{'b'}) and !defined($Options{'l'})) { - &OutputData($Options{'o'},$DBQuery,$MaxLength); + # default: neither -b nor -l + &OutputData($Options{'o'},$DBQuery,$MaxLength); } elsif ($Options{'b'}) { - while (my ($Newsgroup,$Postings) = $DBQuery->fetchrow_array) { + # -b is set (then -l can't be!) + # we have to read in the query results ourselves, as they do not have standard layout + while (my ($Newsgroup,$Postings) = $DBQuery->fetchrow_array) { + # we just assign "top x" or "bottom x" instead of a month for the caption + # FIXME print &FormatOutput($Options{'o'}, ($Options{'i'} ? 'Bottom ' : 'Top ').$Options{'b'}, $Newsgroup, $Postings, $MaxLength); }; -} else { # -l - while (my ($Month,$Newsgroup,$Postings) = $DBQuery->fetchrow_array) { +} else { + # -l must be set now, as all other cases have been taken care of + # we have to read in the query results ourselves, as they do not have standard layout + while (my ($Month,$Newsgroup,$Postings) = $DBQuery->fetchrow_array) { + # we just switch $Newsgroups and $Month for output generation + # FIXME print &FormatOutput($Options{'o'}, $Newsgroup, $Month, $Postings, 7); }; }; @@ -171,3 +193,283 @@ if (!defined($Options{'b'}) and !defined($Options{'l'})) { ### close handles $DBHandle->disconnect; +__END__ + +################################ Documentation ################################# + +=head1 NAME + +groupstats - create reports on newsgroup usage + +=head1 SYNOPSIS + +B [B<-Vhiscqd>] [B<-m> I] [B<-p> I] [B<-n> I] [B<-t> I] [B<-l> I] [B<-b> I] [B<-o> I] [B<-g> I] + +=head1 REQUIREMENTS + +See doc/README: Perl 5.8.x itself and the following modules from CPAN: + +=over 2 + +=item - + +Config::Auto + +=item - + +DBI + +=back + +=head1 DESCRIPTION + +This script create reports on newsgroup usage (number of postings per +group per month) taken from result tables created by +F. + +The time period to act on defaults to last month; you can assign +another month via the B<-m> switch or a time period via the B<-p> +switch; the latter takes preference. + +B will process all newsgroups by default; you can limit +that to only some newsgroups by supplying a list of those groups via +B<-n> (see below). You can include hierarchy levels in the output by +adding the B<-s> switch (see below). + +Furthermore you can set a threshold via B<-t> so that only newsgroups +with more postings per month will be included in the report. You can +invert that by the B<-i> switch so only newsgroups with less than +I postings per month will be included. + +You can sort the output by number of postings per month instead of the +default (alphabetical list of newsgroups) by using B<-q>; you can +reverse the sorting order (from highest to lowest or in reversed +alphabetical order) by using B<-d>. + +Furthermore, you can create a list of newsgroups that had consistently +more (or less) than x postings per month during the whole report +period by using B<-l> (together with B as needed). + +Last but not least you can create a "best of" list of the top x +newsgroups via B<-b> (or a "worst of" list by adding B). + +By default, B will dump a very simple alphabetical list of +newsgroups, one per line, followed by the number of postings in that +month. This output format of course cannot sensibly be combined with +time periods, so you can set the output format by using B<-o> (see +below). Captions can be added by setting the B<-c> switch. + +=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. + +You can override configuration options via the B<-g> switch. + +=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<-m> I (month) + +Set processing period to a month in YYYY-MM format. Ignored if B<-p> +is set. + +=item B<-p> I (period) + +Set processing period to a time period between two month, each in +YYYY-MM format, separated by a colon. Overrides B<-m>. + +=item B<-n> I (newsgroups) + +Limit processing to a certain set of newsgroups. I can +be a single newsgroup name (de.alt.test), a newsgroup hierarchy +(de.alt.*) or a list of either of these, separated by colons, for +example + + de.test:de.alt.test:de.newusers.* + +=item B<-t> I (threshold) + +Only include newsgroups with more than I postings per +month. Can be inverted by the B<-i> switch so that only newsgroups +with less than I postings will be included. + +This setting will be ignored if B<-l> or B<-b> is set. + +=item B<-l> I (level) + +Only include newsgroups with more than I postings per +month, every month during the whole reporting period. Can be inverted +by the B<-i> switch so that only newsgroups with less than I +postings every single month will be included. Output will be ordered +by newsgroup name, followed by month. + +This setting will be ignored if B<-b> is set. Overrides B<-t> and +can't be used together with B<-q> or B<-d>. + +=item B<-b> I (best of) + +Create a list of the I newsgroups with the most postings over the +whole reporting period. Can be inverted by the B<-i> switch so that a +list of the I newsgroups with the least postings over the whole +period is generated. Output will be ordered by sum of postings. + +Overrides B<-t> and B<-l> and can't be used together with B<-q> or +B<-d>. Output format is set to I (see below). + +=item B<-i> (invert) + +Used in conjunction with B<-t>, B<-l> or B<-b> to set a lower +threshold or level or generate a "bottom list" instead of a top list. + +=item B<-s> (sum per hierarchy level) + +Include "virtual" groups for every hierarchy level in output, for +example: + + de.alt.ALL 10 + de.alt.test 5 + de.alt.admin 7 + +See the B man page for details. + +=item B<-o> I (output format) + +Set output format. Default is I, consisting of an alphabetical +list of newsgroups, each on a new line, followed by the number of +postings in that month. This default format can't be used with time +periods of more than one month. + +I format is like I, but will print the month in front of +the newsgroup name. + +I format can only be use with a group list (see B<-n>) of +exactly one newsgroup and is like I, but will output months, +followed by the number of postings. + +If you don't need easily parsable output, you'll mostly use I +format, which will print a header for each new month and try to align +newsgroup names and posting counts. Usage of B<-b> will force this +format. + +=item B<-c> (captions) + +Add captions to output (reporting period, newsgroups list, threshold). + +=item B<-q> (quantity of postings) + +Sort by number of postings instead of by newsgroup names. + +Cannot be used with B<-l> or B<-b>. + +=item B<-d> (descending) + +Change sort order to descending. + +Cannot be used with B<-l> or B<-b>. + +=item B<-g> I (postings per group table) + +Override I from F. + +=back + +=head1 INSTALLATION + +See doc/INSTALL. + +=head1 EXAMPLES + +Show number of postings per group for lasth month in I format: + + groupstats + +Show that report for January of 2010 and de.alt.* plus de.test, +including display of hierarchy levels: + + groupstats -m 2010-01 -n de.alt.*:de.test -s + +Show that report for the year of 2010 in I format: + + groupstats -p 2010-01:2010-12 -o pretty + +Only show newsgroups with less than 30 postings last month, ordered +by number of postings, descending, in I format: + + groupstats -iqdt 30 -o pretty + +Show top 10 for the first half-year of of 2010 in I format: + + groupstats -p 2010-01:2010-06 -b 10 -o pretty + +Report all groups that had less than 30 postings every singele month +in the year of 2010 (I format is forced) + + groupstats -p 2010-01:2010-12 -il 30 + +=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 + +=item - + +gatherstats -h + +=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