Add comments and POD.

Add comments NewsStats.pm
Add POD to install.pl
Add POD to feedlog.pl
Add POD and comments to gatherstats.pl
Add POD and comments to groupstats.pl, beautify code

Signed-off-by: Thomas Hochstein <thh@inter.net>
This commit is contained in:
Thomas Hochstein 2010-09-17 21:53:59 +02:00
parent 610b5ef492
commit 3430c89868
5 changed files with 811 additions and 44 deletions

View file

@ -58,8 +58,8 @@ our $MyVersion = "$MySelf $::VERSION (NewsStats.pm $VERSION)";
################################################################################ ################################################################################
sub ReadOptions { sub ReadOptions {
################################################################################ ################################################################################
### read commandline options and act on standard options ### read commandline options and act on standard options -h and -V
### IN : $Params: containing list of commandline paramaters (without -h and -V) ### IN : $Params: list of legal commandline paramaters (without -h and -V)
### OUT: a hash containing the commandline options ### OUT: a hash containing the commandline options
$Getopt::Std::STANDARD_HELP_VERSION = 1; $Getopt::Std::STANDARD_HELP_VERSION = 1;
@ -116,7 +116,9 @@ sub OverrideConfig {
### $OverrideR: reference to a hash containing overrides ### $OverrideR: reference to a hash containing overrides
my ($ConfigR,$OverrideR) = @_; my ($ConfigR,$OverrideR) = @_;
my %Override = %$OverrideR; my %Override = %$OverrideR;
# Config hash empty?
warn "$MySelf W: Empty configuration hash passed to OverrideConfig().\n" if ( keys %$ConfigR < 1); warn "$MySelf W: Empty configuration hash passed to OverrideConfig().\n" if ( keys %$ConfigR < 1);
# return if no overrides
return if (keys %Override < 1 or keys %$ConfigR < 1); return if (keys %Override < 1 or keys %$ConfigR < 1);
foreach my $Key (keys %Override) { foreach my $Key (keys %Override) {
$$ConfigR{$Key} = $Override{$Key}; $$ConfigR{$Key} = $Override{$Key};
@ -129,7 +131,7 @@ sub InitDB {
################################################################################ ################################################################################
### initialise database connection ### initialise database connection
### IN : $ConfigR: reference to configuration hash ### IN : $ConfigR: reference to configuration hash
### $Die : if TRUE, die if connection failed ### $Die : if TRUE, die if connection fails
### OUT: DBHandle ### OUT: DBHandle
my ($ConfigR,$Die) = @_; my ($ConfigR,$Die) = @_;
my %Conf = %$ConfigR; my %Conf = %$ConfigR;
@ -147,8 +149,10 @@ sub InitDB {
################################################################################ ################################################################################
sub ListNewsgroups { sub ListNewsgroups {
################################################################################ ################################################################################
### count each newsgroup and each hierarchy level, but only once ### explode a (scalar) list of newsgroup names to a list of newsgroup and
### IN : $Newsgroups: a list of newsgroups (content of Newsgroups:) ### hierarchy names where every newsgroup and hierarchy appears only once:
### de.alt.test,de.alt.admin -> de.ALL, de.alt.ALL, de.alt.test, de.alt.admin
### IN : $Newsgroups: a list of newsgroups (content of Newsgroups: header)
### OUT: %Newsgroups: hash containing all newsgroup and hierarchy names as keys ### OUT: %Newsgroups: hash containing all newsgroup and hierarchy names as keys
my ($Newsgroups) = @_; my ($Newsgroups) = @_;
my %Newsgroups; my %Newsgroups;
@ -171,7 +175,8 @@ sub ListNewsgroups {
################################################################################ ################################################################################
sub ParseHierarchies { sub ParseHierarchies {
################################################################################ ################################################################################
### get all hierarchies a newsgroup belongs to ### return a list of all hierarchy levels a newsgroup belongs to
### (for de.alt.test.moderated that would be de/de.alt/de.alt.test)
### IN : $Newsgroup : a newsgroup name ### IN : $Newsgroup : a newsgroup name
### OUT: @Hierarchies: array containing all hierarchies the newsgroup belongs to ### OUT: @Hierarchies: array containing all hierarchies the newsgroup belongs to
my ($Newsgroup) = @_; my ($Newsgroup) = @_;
@ -194,9 +199,11 @@ sub ParseHierarchies {
################################################################################ ################################################################################
sub GetTimePeriod { sub GetTimePeriod {
################################################################################ ################################################################################
### get time period using -m / -p ### get a time period to act on, in order of preference: by default the
### last month; or a month submitted by -m YYYY-MM; or a time period submitted
### by -p YYYY-MM:YYYY-MM
### IN : $Month,$Period: contents of -m and -p ### IN : $Month,$Period: contents of -m and -p
### OUT: $StartMonth, $EndMonth ### OUT: $StartMonth, $EndMonth (identical if period is just one month)
my ($Month,$Period) = @_; my ($Month,$Period) = @_;
# exit if -m is set and not like YYYY-MM # exit if -m is set and not like YYYY-MM
die "$MySelf: E: Wrong date format - use '$MySelf -m YYYY-MM'!\n" if not &CheckMonth($Month); die "$MySelf: E: Wrong date format - use '$MySelf -m YYYY-MM'!\n" if not &CheckMonth($Month);
@ -220,7 +227,7 @@ sub GetTimePeriod {
################################################################################ ################################################################################
sub LastMonth { sub LastMonth {
################################################################################ ################################################################################
### get last month from today in YYYY-MM format ### get last month from todays date in YYYY-MM format
### OUT: last month as YYYY-MM ### OUT: last month as YYYY-MM
# get today's date # get today's date
my (undef,undef,undef,undef,$Month,$Year,undef,undef,undef) = localtime(time); my (undef,undef,undef,undef,$Month,$Year,undef,undef,undef) = localtime(time);
@ -237,7 +244,7 @@ sub LastMonth {
################################################################################ ################################################################################
sub CheckMonth { sub CheckMonth {
################################################################################ ################################################################################
### check for valid month ### check if input is a valid month in YYYY-MM form
### IN : $Month: month ### IN : $Month: month
### OUT: TRUE / FALSE ### OUT: TRUE / FALSE
my ($Month) = @_; my ($Month) = @_;
@ -248,7 +255,7 @@ sub CheckMonth {
################################################################################ ################################################################################
sub SplitPeriod { sub SplitPeriod {
################################################################################ ################################################################################
### split a time period YYYY-MM:YYYY-MM into start and end month ### split a time period denoted by YYYY-MM:YYYY-MM into start and end month
### IN : $Period: time period ### IN : $Period: time period
### OUT: $StartMonth, Â$EndMonth ### OUT: $StartMonth, Â$EndMonth
my ($Period) = @_; my ($Period) = @_;
@ -265,7 +272,7 @@ sub SplitPeriod {
################################################################################ ################################################################################
sub ListMonth { sub ListMonth {
################################################################################ ################################################################################
### return a list of month (YYYY-MM) between start and end month ### return a list of months (YYYY-MM) between start and end month
### IN : $StartMonth, $EndMonth ### IN : $StartMonth, $EndMonth
### OUT: @Months: array containing all months from $StartMonth to $EndMonth ### OUT: @Months: array containing all months from $StartMonth to $EndMonth
my ($StartMonth, $EndMonth) = @_; my ($StartMonth, $EndMonth) = @_;
@ -293,12 +300,11 @@ sub ListMonth {
################################################################################ ################################################################################
sub OutputData { sub OutputData {
################################################################################ ################################################################################
### output information with formatting from DBHandle ### read database query results from DBHandle and print results with formatting
### IN : $Format : format specifier ### IN : $Format : format specifier
### $DBQuery: database query handle with executed query, ### $DBQuery: database query handle with executed query,
### containing $Month, $Key, $Value ### containing $Month, $Key, $Value
### $PadGroup: padding length for newsgroups field (optional) for 'pretty' ### $PadGroup: padding length for newsgroups field (optional) for 'pretty'
### OUT: $Output: formatted output
my ($Format, $DBQuery,$PadGroup) = @_; my ($Format, $DBQuery,$PadGroup) = @_;
while (my ($Month, $Key, $Value) = $DBQuery->fetchrow_array) { while (my ($Month, $Key, $Value) = $DBQuery->fetchrow_array) {
print &FormatOutput($Format, $Month, $Key, $Value, $PadGroup); print &FormatOutput($Format, $Month, $Key, $Value, $PadGroup);
@ -308,12 +314,12 @@ sub OutputData {
################################################################################ ################################################################################
sub FormatOutput { sub FormatOutput {
################################################################################ ################################################################################
### format information for output ### format information for output according to format specifier
### IN : $Format : format specifier ### IN : $Format : format specifier
### $PadGroup: padding length for newsgroups field (optional) for 'pretty'
### $Month : month (as YYYY-MM) ### $Month : month (as YYYY-MM)
### $Key : newsgroup, client, ... ### $Key : newsgroup, client, ...
### $Value : number of postings with that attribute ### $Value : number of postings with that attribute
### $PadGroup: padding length for key field (optional) for 'pretty'
### OUT: $Output: formatted output ### OUT: $Output: formatted output
my ($Format, $Month, $Key, $Value, $PadGroup) = @_; my ($Format, $Month, $Key, $Value, $PadGroup) = @_;
@ -324,6 +330,7 @@ sub FormatOutput {
die "$MySelf: E: Unknown output type '$Format'!\n" if !exists($LegalOutput{$Format}); die "$MySelf: E: Unknown output type '$Format'!\n" if !exists($LegalOutput{$Format});
my ($Output); my ($Output);
# keep last month in mind
our ($LastIteration); our ($LastIteration);
if ($Format eq 'dump') { if ($Format eq 'dump') {
# output as dump (ng nnnnn) # output as dump (ng nnnnn)
@ -348,7 +355,9 @@ sub FormatOutput {
################################################################################ ################################################################################
sub SQLHierarchies { sub SQLHierarchies {
################################################################################ ################################################################################
### amend WHERE clause to include hierarchies ### add exclusion of hierarchy levels (de.alt.ALL) from SQL query by
### amending the WHERE clause if $ShowHierarchies is false (or don't, if it is
### true, accordingly)
### IN : $ShowHierarchies: boolean value ### IN : $ShowHierarchies: boolean value
### OUT: SQL code ### OUT: SQL code
my ($ShowHierarchies) = @_; my ($ShowHierarchies) = @_;
@ -358,7 +367,7 @@ sub SQLHierarchies {
################################################################################ ################################################################################
sub GetMaxLenght { sub GetMaxLenght {
################################################################################ ################################################################################
### get length of longest field in query ### get length of longest field in future query result
### IN : $DBHandle : database handel ### IN : $DBHandle : database handel
### $Table : table to query ### $Table : table to query
### $Field : field to check ### $Field : field to check
@ -375,7 +384,8 @@ sub GetMaxLenght {
################################################################################ ################################################################################
sub SQLGroupList { sub SQLGroupList {
################################################################################ ################################################################################
### create part of WHERE clause for list of newsgroups separated by : ### explode list of newsgroups separated by : (with wildcards) to a SQL WHERE
### clause
### IN : $Newsgroups: list of newsgroups (group.one.*:group.two:group.three.*) ### IN : $Newsgroups: list of newsgroups (group.one.*:group.two:group.three.*)
### OUT: SQL code, list of newsgroups ### OUT: SQL code, list of newsgroups
my ($Newsgroups) = @_; my ($Newsgroups) = @_;

View file

@ -85,3 +85,143 @@ $DBHandle->disconnect;
syslog(LOG_NOTICE, "$MySelf closing down.") if !$Options{'q'}; syslog(LOG_NOTICE, "$MySelf closing down.") if !$Options{'q'};
closelog(); closelog();
__END__
################################ Documentation #################################
=head1 NAME
feedlog - log data from an INN feed to a database
=head1 SYNOPSIS
B<feedlog> [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<syslog> via I<news> facility. If B<feedlog>
fails to initiate a database connection at startup, it will log to
I<syslog> with I<CRIT> priority and go in an endless loop, as
terminating would only result in a rapid respawn.
=head2 Configuration
F<feedlog.pl> will read its configuration from F<newsstats.conf> 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<yapfaq> 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<INN> F<errlog> 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<INN> F<newsfeeds> 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<feedlog.pl>
The script itself.
=item F<NewsStats.pm>
Library functions for the NewsStats package.
=item F<newsstats.conf>
Runtime configuration file for B<yapfaq>.
=back
=head1 BUGS
Please report any bugs or feature requests to the author or use the
bug tracker at L<http://bugs.th-h.de/>!
=head1 SEE ALSO
=over 2
=item -
doc/README
=item -
doc/INSTALL
=back
This script is part of the B<NewsStats> package.
=head1 AUTHOR
Thomas Hochstein <thh@inter.net>
=head1 COPYRIGHT AND LICENSE
Copyright (c) 2010 Thomas Hochstein <thh@inter.net>
This program is free software; you may redistribute it and/or modify it
under the same terms as Perl itself.
=cut

View file

@ -47,11 +47,11 @@ $ConfOverride{'DBTableHosts'} = $Options{'s'} if $Options{'s'};
$ConfOverride{'TLH'} = $Options{'n'} if $Options{'n'}; $ConfOverride{'TLH'} = $Options{'n'} if $Options{'n'};
&OverrideConfig(\%Conf,\%ConfOverride); &OverrideConfig(\%Conf,\%ConfOverride);
### get type of information to gather, default to 'all' ### get type of information to gather, defaulting to 'all'
$Options{'t'} = 'all' if !$Options{'t'}; $Options{'t'} = 'all' if !$Options{'t'};
die "$MySelf: E: Unknown type '-t $Options{'t'}'!\n" if !exists($LegalTypes{$Options{'t'}}); die "$MySelf: E: Unknown type '-t $Options{'t'}'!\n" if !exists($LegalTypes{$Options{'t'}});
### get time period ### get time period (-m or -p)
my ($StartMonth,$EndMonth) = &GetTimePeriod($Options{'m'},$Options{'p'}); my ($StartMonth,$EndMonth) = &GetTimePeriod($Options{'m'},$Options{'p'});
### init database ### init database
@ -94,9 +94,219 @@ foreach my $Month (&ListMonth($StartMonth,$EndMonth)) {
$DBQuery->finish; $DBQuery->finish;
}; };
}; };
} else {
# other types of information go here - later on
}; };
}; };
### close handles ### close handles
$DBHandle->disconnect; $DBHandle->disconnect;
__END__
################################ Documentation #################################
=head1 NAME
gatherstats - process statistical data from a raw source
=head1 SYNOPSIS
B<gatherstats> [B<-Vhdo>] [B<-m> I<YYYY-MM>] [B<-p> I<YYYY-MM:YYYY-MM>] [B<-t> I<type>] [B<-n> I<TLH>] [B<-r> I<database table>] [B<-g> I<database table>] [B<-c> I<database table>] [B<-s> I<database table>]
=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 will extract and process statistical information from a
database table which is fed from F<feedlog.pl> for a given time period
and write its results to (an)other database table(s).
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.
By default B<gatherstats> will process all types of information; you
can change that using the B<-t> switch and assigning the type of
information to process. Currently only processing of the number of
postings per group per month is implemented anyway, so that doesn't
matter yet.
Possible information types include:
=over 3
=item B<groups> (postings per group per month)
B<gatherstats> will examine Newsgroups: headers. Crosspostings will be
counted for each single group they appear in. Groups not in I<TLH>
will be ignored.
B<gatherstats> will also add up the number of postings for each
hierarchy level, but only count each posting once. A posting to
de.alt.test will be counted for de.alt.test, de.alt.ALL and de.ALL,
respectively. A crossposting to de.alt.test and de.alt.admin, on the
other hand, will be counted for de.alt.test and de.alt.admin each, but
only once for de.alt.ALL and de.ALL.
Data is written to I<DBTableGrps> (see doc/INSTALL).
=back
=head2 Configuration
F<gatherstats.pl> will read its configuration from F<newsstats.conf>
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<-n>, B<-r>, B<-g>,
B<-c> and B<-s> switches, respectively.
=head1 OPTIONS
=over 3
=item B<-V> (version)
Print out version and copyright information on B<yapfaq> and exit.
=item B<-h> (help)
Print this man page and exit.
=item B<-d> (debug)
Output debugging information to STDOUT while processing (number of
postings per group).
=item B<-o> (output only)
Do not write results to database. You should use B<-d> in conjunction
with B<-o> ... everything else seems a bit pointless.
=item B<-m> I<YYYY-MM> (month)
Set processing period to a month in YYYY-MM format. Ignored if B<-p>
is set.
=item B<-p> I<YYYY-MM:YYYY-MM> (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<-t> I<type> (type)
Set processing type to one of I<all> and I<groups>. Defaults to all
(and is currently rather pointless as only I<groups> has been
implemented).
=item B<-n> I<TLH> (newsgroup hierarchy)
Override I<TLH> from F<newsstats.conf>.
=item B<-r> I<table> (raw data table)
Override I<DBTableRaw> from F<newsstats.conf>.
=item B<-g> I<table> (postings per group table)
Override I<DBTableGrps> from F<newsstats.conf>.
=item B<-c> I<table> (client data table)
Override I<DBTableClnts> from F<newsstats.conf>.
=item B<-s> I<table> (server/host data table)
Override I<DBTableHosts> from F<newsstats.conf>.
=back
=head1 INSTALLATION
See doc/INSTALL.
=head1 EXAMPLES
Process all types of information for lasth month:
gatherstats
Do a dry run, showing results of processing:
gatherstats -do
Process all types of information for January of 2010:
gatherstats -m 2010-01
Process only number of postings for the year of 2010:
gatherstats -p 2010-01:2010-12 -t groups
=head1 FILES
=over 4
=item F<gatherstats.pl>
The script itself.
=item F<NewsStats.pm>
Library functions for the NewsStats package.
=item F<newsstats.conf>
Runtime configuration file for B<yapfaq>.
=back
=head1 BUGS
Please report any bugs or feature requests to the author or use the
bug tracker at L<http://bugs.th-h.de/>!
=head1 SEE ALSO
=over 2
=item -
doc/README
=item -
doc/INSTALL
=back
This script is part of the B<NewsStats> package.
=head1 AUTHOR
Thomas Hochstein <thh@inter.net>
=head1 COPYRIGHT AND LICENSE
Copyright (c) 2010 Thomas Hochstein <thh@inter.net>
This program is free software; you may redistribute it and/or modify it
under the same terms as Perl itself.
=cut

View file

@ -23,10 +23,6 @@ use NewsStats qw(:DEFAULT :TimePeriods :Output :SQLHelper);
use DBI; use DBI;
################################# Definitions ##################################
# ...
################################# Main program ################################# ################################# Main program #################################
### read commandline options ### 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"); 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 !$Options{'t'});
### get time period ### get time period
my ($StartMonth,$EndMonth) = &GetTimePeriod($Options{'m'},$Options{'p'}); my ($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 $Options{'p'}) { if ($Options{'o'} eq 'dump' and $Options{'p'}) {
$StartMonth = $EndMonth; $StartMonth = $EndMonth;
warn ("$MySelf: W: You cannot combine time periods (-p) with '-o dump'. Month was set to $StartMonth.\n"); warn ("$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 ### init database
my $DBHandle = InitDB(\%Conf,1); my $DBHandle = InitDB(\%Conf,1);
### get data ### create report
# get list of newsgroups (-n) # get list of newsgroups (-n)
my ($QueryPart,@GroupList); my ($QueryPart,@GroupList);
my $Newsgroups = $Options{'n'}; my $Newsgroups = $Options{'n'};
if ($Newsgroups) { if ($Newsgroups) {
# explode list of newsgroups for WHERE clause
($QueryPart,@GroupList) = &SQLGroupList($Newsgroups); ($QueryPart,@GroupList) = &SQLGroupList($Newsgroups);
} else { } else {
# set to dummy value (always true)
$QueryPart = 1; $QueryPart = 1;
}; };
# manage thresholds # manage thresholds
if (defined($Options{'t'})) { if (defined($Options{'t'})) {
if ($Options{'i'}) { if ($Options{'i'}) {
# -i: list groups below threshold
$QueryPart .= ' AND postings < ?'; $QueryPart .= ' AND postings < ?';
} else { } else {
# default: list groups above threshold
$QueryPart .= ' AND postings > ?'; $QueryPart .= ' AND postings > ?';
}; };
# push threshold to GroupList to match number of binding vars for DBQuery->execute
push @GroupList,$Options{'t'}; push @GroupList,$Options{'t'};
} }
# construct WHERE clause # 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'})); my $WhereClause = sprintf('month BETWEEN ? AND ? AND %s %s',$QueryPart,&SQLHierarchies($Options{'s'}));
# get lenght of longest newsgroup delivered by query for formatting purposes # get lenght of longest newsgroup delivered by query for formatting purposes
# FIXME
my $MaxLength = &GetMaxLenght($DBHandle,$Conf{'DBTableGrps'},'newsgroup',$WhereClause,$StartMonth,$EndMonth,@GroupList); my $MaxLength = &GetMaxLenght($DBHandle,$Conf{'DBTableGrps'},'newsgroup',$WhereClause,$StartMonth,$EndMonth,@GroupList);
my ($OrderClause,$DBQuery); my ($OrderClause,$DBQuery);
# -b (best of) defined? # -b (best of / top list) defined?
if (!defined($Options{'b'}) and !defined($Options{'l'})) { 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 = 'newsgroup';
$OrderClause = 'postings' if $Options{'q'}; $OrderClause = 'postings' if $Options{'q'};
$OrderClause .= ' DESC' if $Options{'d'}; $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)); $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'}) { } elsif ($Options{'b'}) {
# -b is set (then -l can't be!)
# set sorting order (-i) # set sorting order (-i)
if ($Options{'i'}) { if ($Options{'i'}) {
$OrderClause = 'postings'; $OrderClause = 'postings';
} else { } else {
$OrderClause = 'postings DESC'; $OrderClause = 'postings DESC';
}; };
# push LIMIT to GroupList to match number of binding vars # push LIMIT to GroupList to match number of binding vars for DBQuery->execute
push @GroupList,$Options{'b'}; 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)); $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,$OrderClause));
} else { # -l } else {
# -l must be set now, as all other cases have been taken care of
# set sorting order (-i) # set sorting order (-i)
if ($Options{'i'}) { if ($Options{'i'}) {
$OrderClause = '<'; $OrderClause = '<';
} else { } else {
$OrderClause = '>'; $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,$Options{'l'}; push @GroupList,$Options{'l'};
push @GroupList,$StartMonth,$EndMonth; 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)); $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 # 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 $Options{'p'}); printf ("----- Report from %s to %s\n",$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 $Options{'n'}; printf ("----- Newsgroups: %s\n",join(',',split(/:/,$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 $Options{'t'}; printf ("----- Threshold: %s %u\n",$Options{'i'} ? '<' : '>',$Options{'t'}) if $Options{'c'} and $Options{'t'};
if (!defined($Options{'b'}) and !defined($Options{'l'})) { if (!defined($Options{'b'}) and !defined($Options{'l'})) {
# default: neither -b nor -l
&OutputData($Options{'o'},$DBQuery,$MaxLength); &OutputData($Options{'o'},$DBQuery,$MaxLength);
} elsif ($Options{'b'}) { } elsif ($Options{'b'}) {
# -b is set (then -l can't be!)
# we have to read in the query results ourselves, as they do not have standard layout
while (my ($Newsgroup,$Postings) = $DBQuery->fetchrow_array) { while (my ($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, $MaxLength); print &FormatOutput($Options{'o'}, ($Options{'i'} ? 'Bottom ' : 'Top ').$Options{'b'}, $Newsgroup, $Postings, $MaxLength);
}; };
} else { # -l } else {
# -l must be set now, as all other cases have been taken care of
# we have to read in the query results ourselves, as they do not have standard layout
while (my ($Month,$Newsgroup,$Postings) = $DBQuery->fetchrow_array) { while (my ($Month,$Newsgroup,$Postings) = $DBQuery->fetchrow_array) {
# we just switch $Newsgroups and $Month for output generation
# FIXME
print &FormatOutput($Options{'o'}, $Newsgroup, $Month, $Postings, 7); print &FormatOutput($Options{'o'}, $Newsgroup, $Month, $Postings, 7);
}; };
}; };
@ -171,3 +193,283 @@ if (!defined($Options{'b'}) and !defined($Options{'l'})) {
### close handles ### close handles
$DBHandle->disconnect; $DBHandle->disconnect;
__END__
################################ Documentation #################################
=head1 NAME
groupstats - create reports on newsgroup usage
=head1 SYNOPSIS
B<groupstats> [B<-Vhiscqd>] [B<-m> I<YYYY-MM>] [B<-p> I<YYYY-MM:YYYY-MM>] [B<-n> I<newsgroup(s)>] [B<-t> I<threshold>] [B<-l> I<level>] [B<-b> I<number>] [B<-o> I<output type>] [B<-g> I<database table>]
=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<gatherstats.pl>.
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<groupstats> 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<threshold> 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<i> 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<i>).
By default, B<groupstats> 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<groupstats.pl> will read its configuration from F<newsstats.conf>
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<yapfaq> and exit.
=item B<-h> (help)
Print this man page and exit.
=item B<-m> I<YYYY-MM> (month)
Set processing period to a month in YYYY-MM format. Ignored if B<-p>
is set.
=item B<-p> I<YYYY-MM:YYYY-MM> (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<newsgroup(s)> (newsgroups)
Limit processing to a certain set of newsgroups. I<newsgroup(s)> 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> (threshold)
Only include newsgroups with more than I<threshold> postings per
month. Can be inverted by the B<-i> switch so that only newsgroups
with less than I<threshold> postings will be included.
This setting will be ignored if B<-l> or B<-b> is set.
=item B<-l> I<level> (level)
Only include newsgroups with more than I<level> 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<level>
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<n> (best of)
Create a list of the I<n> 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<n> 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<pretty> (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<gatherstats> man page for details.
=item B<-o> I<output type> (output format)
Set output format. Default is I<dump>, 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<list> format is like I<dump>, but will print the month in front of
the newsgroup name.
I<dumpgroup> format can only be use with a group list (see B<-n>) of
exactly one newsgroup and is like I<dump>, but will output months,
followed by the number of postings.
If you don't need easily parsable output, you'll mostly use I<pretty>
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<table> (postings per group table)
Override I<DBTableGrps> from F<newsstats.conf>.
=back
=head1 INSTALLATION
See doc/INSTALL.
=head1 EXAMPLES
Show number of postings per group for lasth month in I<dump> 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<pretty> 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<pretty> format:
groupstats -iqdt 30 -o pretty
Show top 10 for the first half-year of of 2010 in I<pretty> 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<pretty> format is forced)
groupstats -p 2010-01:2010-12 -il 30
=head1 FILES
=over 4
=item F<groupstats.pl>
The script itself.
=item F<NewsStats.pm>
Library functions for the NewsStats package.
=item F<newsstats.conf>
Runtime configuration file for B<yapfaq>.
=back
=head1 BUGS
Please report any bugs or feature requests to the author or use the
bug tracker at L<http://bugs.th-h.de/>!
=head1 SEE ALSO
=over 2
=item -
doc/README
=item -
doc/INSTALL
=item -
gatherstats -h
=back
This script is part of the B<NewsStats> package.
=head1 AUTHOR
Thomas Hochstein <thh@inter.net>
=head1 COPYRIGHT AND LICENSE
Copyright (c) 2010 Thomas Hochstein <thh@inter.net>
This program is free software; you may redistribute it and/or modify it
under the same terms as Perl itself.
=cut

View file

@ -150,3 +150,108 @@ Enjoy!
-thh <thh\@inter.net> -thh <thh\@inter.net>
TODO TODO
__END__
################################ Documentation #################################
=head1 NAME
install - installation script
=head1 SYNOPSIS
B<install> [B<-Vh>]
=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 will create database tables as necessary and configured.
=head2 Configuration
F<install.pl> will read its configuration from F<newsstats.conf> 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<yapfaq> and exit.
=item B<-h> (help)
Print this man page and exit.
=back
=head1 FILES
=over 4
=item F<install.pl>
The script itself.
=item F<NewsStats.pm>
Library functions for the NewsStats package.
=item F<newsstats.conf>
Runtime configuration file for B<yapfaq>.
=back
=head1 BUGS
Please report any bugs or feature requests to the author or use the
bug tracker at L<http://bugs.th-h.de/>!
=head1 SEE ALSO
=over 2
=item -
doc/README
=item -
doc/INSTALL
=back
This script is part of the B<NewsStats> package.
=head1 AUTHOR
Thomas Hochstein <thh@inter.net>
=head1 COPYRIGHT AND LICENSE
Copyright (c) 2010 Thomas Hochstein <thh@inter.net>
This program is free software; you may redistribute it and/or modify it
under the same terms as Perl itself.
=cut