collectstats.pl 17.2 KB
Newer Older
1
#!/usr/bin/perl -w
2 3
# -*- Mode: perl; indent-tabs-mode: nil -*-
#
4 5 6 7 8 9 10 11 12 13
# The contents of this file are subject to the Mozilla Public
# License Version 1.1 (the "License"); you may not use this file
# except in compliance with the License. You may obtain a copy of
# the License at http://www.mozilla.org/MPL/
#
# Software distributed under the License is distributed on an "AS
# IS" basis, WITHOUT WARRANTY OF ANY KIND, either express or
# implied. See the License for the specific language governing
# rights and limitations under the License.
#
14
# The Original Code is the Bugzilla Bug Tracking System.
15
#
16
# The Initial Developer of the Original Code is Netscape Communications
17 18 19 20
# Corporation. Portions created by Netscape are
# Copyright (C) 1998 Netscape Communications Corporation. All
# Rights Reserved.
#
21 22
# Contributor(s): Terry Weissman <terry@mozilla.org>,
#                 Harrison Page <harrison@netscape.com>
23 24 25
#                 Gervase Markham <gerv@gerv.net>
#                 Richard Walters <rwalters@qualcomm.com>
#                 Jean-Sebastien Guay <jean_seb@hybride.com>
26 27

# Run me out of cron at midnight to collect Bugzilla statistics.
28 29 30
#
# To run new charts for a specific date, pass it in on the command line in
# ISO (2004-08-14) format.
31

32
use AnyDBM_File;
33
use strict;
34
use IO::Handle;
35
use vars @::legal_product;
36

37
use lib ".";
38
require "globals.pl";
39 40
use Bugzilla::Search;
use Bugzilla::User;
41

42
use Bugzilla;
43
use Bugzilla::Config qw(:DEFAULT $datadir);
44

45 46 47 48 49
# Turn off output buffering (probably needed when displaying output feedback
# in the regenerate mode.)
$| = 1;

# Tidy up after graphing module
50
if (chdir("graphs")) {
51
    unlink <./*.gif>;
52 53 54
    unlink <./*.png>;
    chdir("..");
}
55

56
GetVersionTable();
57

58
Bugzilla->switch_to_shadow_db();
59

60 61 62
# To recreate the daily statistics,  run "collectstats.pl --regenerate" .
my $regenerate = 0;
if ($#ARGV >= 0 && $ARGV[0] eq "--regenerate") {
63
    shift(@ARGV);
64 65 66
    $regenerate = 1;
}

67 68 69
my @myproducts;
push( @myproducts, "-All-", @::legal_product );

70
my $tstart = time;
71
foreach (@myproducts) {
72
    my $dir = "$datadir/mining";
73

74
    &check_data_dir ($dir);
75 76 77 78 79 80
    
    if ($regenerate) {
        &regenerate_stats($dir, $_);
    } else {
        &collect_stats($dir, $_);
    }
81
}
82
my $tend = time;
83 84
# Uncomment the following line for performance testing.
#print "Total time taken " . delta_time($tstart, $tend) . "\n";
85

86 87
&calculate_dupes();

88 89
CollectSeriesData();

90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105
{
    local $ENV{'GATEWAY_INTERFACE'} = 'cmdline';
    local $ENV{'REQUEST_METHOD'} = 'GET';
    local $ENV{'QUERY_STRING'} = 'ctype=rdf';

    my $perl = $^X;
    trick_taint($perl);

    # Generate a static RDF file containing the default view of the duplicates data.
    open(CGI, "$perl -T duplicates.cgi |")
        || die "can't fork duplicates.cgi: $!";
    open(RDF, ">$datadir/duplicates.tmp")
        || die "can't write to $datadir/duplicates.tmp: $!";
    my $headers_done = 0;
    while (<CGI>) {
        print RDF if $headers_done;
106
        $headers_done = 1 if $_ eq "\r\n";
107 108 109
    }
    close CGI;
    close RDF;
110
}
111 112 113
if (-s "$datadir/duplicates.tmp") {
    rename("$datadir/duplicates.rdf", "$datadir/duplicates-old.rdf");
    rename("$datadir/duplicates.tmp", "$datadir/duplicates.rdf");
114 115
}

116 117
sub check_data_dir {
    my $dir = shift;
118

119
    if (! -d $dir) {
120 121
        mkdir $dir, 0755;
        chmod 0755, $dir;
122 123
    }
}
124

125 126 127 128
sub collect_stats {
    my $dir = shift;
    my $product = shift;
    my $when = localtime (time);
129 130 131
    my $product_id = get_product_id($product) unless $product eq '-All-';

    die "Unknown product $product" unless ($product_id or $product eq '-All-');
132

133 134 135 136 137
    # NB: Need to mangle the product for the filename, but use the real
    # product name in the query
    my $file_product = $product;
    $file_product =~ s/\//-/gs;
    my $file = join '/', $dir, $file_product;
138 139 140 141 142
    my $exists = -f $file;

    if (open DATA, ">>$file") {
        push my @row, &today;

143 144
        foreach my $status ('NEW', 'ASSIGNED', 'REOPENED', 'UNCONFIRMED', 'RESOLVED', 'VERIFIED', 'CLOSED') {
            if( $product eq "-All-" ) {
145
                SendSQL("SELECT COUNT(bug_status) FROM bugs WHERE bug_status='$status'");
146
            } else {
147
                SendSQL("SELECT COUNT(bug_status) FROM bugs WHERE bug_status='$status' AND product_id=$product_id");
148 149 150 151 152 153 154
            }

            push @row, FetchOneColumn();
        }

        foreach my $resolution ('FIXED', 'INVALID', 'WONTFIX', 'LATER', 'REMIND', 'DUPLICATE', 'WORKSFORME', 'MOVED') {
            if( $product eq "-All-" ) {
155
                SendSQL("SELECT COUNT(resolution) FROM bugs WHERE resolution='$resolution'");
156
            } else {
157
                SendSQL("SELECT COUNT(resolution) FROM bugs WHERE resolution='$resolution' AND product_id=$product_id");
158 159 160 161 162 163 164
            }

            push @row, FetchOneColumn();
        }

        if (! $exists) {
            print DATA <<FIN;
165
# Bugzilla Daily Bug Stats
166
#
167
# Do not edit me! This file is generated.
168
#
169
# fields: DATE|NEW|ASSIGNED|REOPENED|UNCONFIRMED|RESOLVED|VERIFIED|CLOSED|FIXED|INVALID|WONTFIX|LATER|REMIND|DUPLICATE|WORKSFORME|MOVED
170 171
# Product: $product
# Created: $when
172
FIN
173 174
        }

175 176
        print DATA (join '|', @row) . "\n";
        close DATA;
177
        chmod 0644, $file;
178 179 180 181 182
    } else {
        print "$0: $file, $!";
    }
}

183
sub calculate_dupes {
184 185
    my $dbh = Bugzilla->dbh;
    my $rows = $dbh->selectall_arrayref("SELECT dupe_of, dupe FROM duplicates");
186 187 188 189 190 191

    my %dupes;
    my %count;
    my $key;
    my $changed = 1;

192
    my $today = &today_dash;
193 194 195 196

    # Save % count here in a date-named file
    # so we can read it back in to do changed counters
    # First, delete it if it exists, so we don't add to the contents of an old file
197
    if (my @files = <$datadir/duplicates/dupes$today*>) {
198
        map { trick_taint($_) } @files;
199
        unlink @files;
200 201
    }
   
202
    dbmopen(%count, "$datadir/duplicates/dupes$today", 0644) || die "Can't open DBM dupes file: $!";
203 204 205

    # Create a hash with key "a bug number", value "bug which that bug is a
    # direct dupe of" - straight from the duplicates table.
206 207
    foreach my $row (@$rows) {
        my ($dupe_of, $dupe) = @$row;
208
        $dupes{$dupe} = $dupe_of;
209 210 211 212 213 214 215
    }

    # Total up the number of bugs which are dupes of a given bug
    # count will then have key = "bug number", 
    # value = "number of immediate dupes of that bug".
    foreach $key (keys(%dupes)) 
    {
216
        my $dupe_of = $dupes{$key};
217

218 219 220
        if (!defined($count{$dupe_of})) {
            $count{$dupe_of} = 0;
        }
221

222
        $count{$dupe_of}++;
223 224 225 226 227 228
    }   

    # Now we collapse the dupe tree by iterating over %count until
    # there is no further change.
    while ($changed == 1)
    {
229 230 231 232 233 234 235 236 237 238 239 240 241 242
        $changed = 0;
        foreach $key (keys(%count)) {
            # if this bug is actually itself a dupe, and has a count...
            if (defined($dupes{$key}) && $count{$key} > 0) {
                # add that count onto the bug it is a dupe of,
                # and zero the count; the check is to avoid
                # loops
                if ($count{$dupes{$key}} != 0) {
                    $count{$dupes{$key}} += $count{$key};
                    $count{$key} = 0;
                    $changed = 1;
                }
            }
        }
243 244 245 246 247
    }

    # Remove the values for which the count is zero
    foreach $key (keys(%count))
    {
248 249 250
        if ($count{$key} == 0) {
            delete $count{$key};
        }
251 252 253 254 255
    }
   
    dbmclose(%count);
}

256 257 258 259 260
# This regenerates all statistics from the database.
sub regenerate_stats {
    my $dir = shift;
    my $product = shift;

261 262
    my $dbh = Bugzilla->dbh;
    my $when = localtime(time());
263 264 265 266 267 268 269 270 271 272 273 274 275 276
    my $tstart = time();

    # NB: Need to mangle the product for the filename, but use the real
    # product name in the query
    my $file_product = $product;
    $file_product =~ s/\//-/gs;
    my $file = join '/', $dir, $file_product;

    my @bugs;

    my $and_product = "";
    my $from_product = "";
                    
    if ($product ne '-All-') {
277 278 279
        $and_product = " AND products.name = " . SqlQuote($product);
        $from_product = "INNER JOIN products " .
                        "ON bugs.product_id = products.id";
280 281 282 283 284
    }          
              
    # Determine the start date from the date the first bug in the
    # database was created, and the end date from the current day.
    # If there were no bugs in the search, return early.
285 286 287 288
    SendSQL("SELECT " . $dbh->sql_to_days('creation_ts') . " AS start, " .
                        $dbh->sql_to_days('current_date') . " AS end, " .
                        $dbh->sql_to_days("'1970-01-01'") . 
            " FROM bugs $from_product WHERE " .
289
            $dbh->sql_to_days('creation_ts') . " IS NOT NULL " .
290
            $and_product .
291
            " ORDER BY start " . $dbh->sql_limit(1));
292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318
    
    my ($start, $end, $base) = FetchSQLData();
    if (!defined $start) {
        return;
    }
 
    if (open DATA, ">$file") {
        DATA->autoflush(1);
        print DATA <<FIN;
# Bugzilla Daily Bug Stats
#
# Do not edit me! This file is generated.
#
# fields: DATE|NEW|ASSIGNED|REOPENED|UNCONFIRMED|RESOLVED|VERIFIED|CLOSED|FIXED|INVALID|WONTFIX|LATER|REMIND|DUPLICATE|WORKSFORME|MOVED
# Product: $product
# Created: $when
FIN
        # For each day, generate a line of statistics.
        my $total_days = $end - $start;
        for (my $day = $start + 1; $day <= $end; $day++) {
            # Some output feedback
            my $percent_done = ($day - $start - 1) * 100 / $total_days;
            printf "\rRegenerating $product \[\%.1f\%\%]", $percent_done;

            # Get a list of bugs that were created the previous day, and
            # add those bugs to the list of bugs for this product.
            SendSQL("SELECT bug_id FROM bugs $from_product " .
319 320
                    " WHERE bugs.creation_ts < " . $dbh->sql_from_days($day - 1) .
                    " AND bugs.creation_ts >= " . $dbh->sql_from_days($day - 2) .
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
                    $and_product .
                    " ORDER BY bug_id");
            
            my @row;        
            while (@row = FetchSQLData()) {
                push @bugs, $row[0];
            }

            # For each bug that existed on that day, determine its status
            # at the beginning of the day.  If there were no status
            # changes on or after that day, the status was the same as it
            # is today, which can be found in the bugs table.  Otherwise,
            # the status was equal to the first "previous value" entry in
            # the bugs_activity table for that bug made on or after that
            # day.
            my %bugcount;
            my @logstates = qw(NEW ASSIGNED REOPENED UNCONFIRMED RESOLVED 
                               VERIFIED CLOSED);
            my @logresolutions = qw(FIXED INVALID WONTFIX LATER REMIND 
                                    DUPLICATE WORKSFORME MOVED);
            foreach (@logstates) {
                $bugcount{$_} = 0;
            }
            
            foreach (@logresolutions) {
                $bugcount{$_} = 0;
            }
            
            for my $bug (@bugs) {
                # First, get information on various bug states.
                SendSQL("SELECT bugs_activity.removed " .
352 353 354 355 356
                        "  FROM bugs_activity " .
                    "INNER JOIN fielddefs " .
                        "    ON bugs_activity.fieldid = fielddefs.fieldid " .
                        " WHERE fielddefs.name = 'bug_status' " .
                        "   AND bugs_activity.bug_id = $bug " .
357 358
                        "   AND bugs_activity.bug_when >= " . $dbh->sql_from_days($day) .
                     " ORDER BY bugs_activity.bug_when " .
359
                        $dbh->sql_limit(1));
360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
                
                my $status;
                if (@row = FetchSQLData()) {
                    $status = $row[0];
                } else {
                    SendSQL("SELECT bug_status FROM bugs WHERE bug_id = $bug");
                    $status = FetchOneColumn();
                }
                
                if (defined $bugcount{$status}) {
                    $bugcount{$status}++;
                }

                # Next, get information on various bug resolutions.
                SendSQL("SELECT bugs_activity.removed " .
375 376 377 378 379
                        "  FROM bugs_activity " .
                    "INNER JOIN fielddefs " .
                        "    ON bugs_activity.fieldid = fielddefs.fieldid " .
                        " WHERE fielddefs.name = 'resolution' " .
                        "   AND bugs_activity.bug_id = $bug " .
380 381
                        "   AND bugs_activity.bug_when >= " . $dbh->sql_from_days($day) .
                     " ORDER BY bugs_activity.bug_when " . 
382
                        $dbh->sql_limit(1));
383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420
                        
                if (@row = FetchSQLData()) {
                    $status = $row[0];
                } else {
                    SendSQL("SELECT resolution FROM bugs WHERE bug_id = $bug");
                    $status = FetchOneColumn();
                }
                
                if (defined $bugcount{$status}) {
                    $bugcount{$status}++;
                }
            }

            # Generate a line of output containing the date and counts
            # of bugs in each state.
            my $date = sqlday($day, $base);
            print DATA "$date";
            foreach (@logstates) {
                print DATA "|$bugcount{$_}";
            }
            
            foreach (@logresolutions) {
                print DATA "|$bugcount{$_}";
            }
            
            print DATA "\n";
        }
        
        # Finish up output feedback for this product.
        my $tend = time;
        print "\rRegenerating $product \[100.0\%] - " .
            delta_time($tstart, $tend) . "\n";
            
        close DATA;
        chmod 0640, $file;
    }
}

421 422 423 424
sub today {
    my ($dom, $mon, $year) = (localtime(time))[3, 4, 5];
    return sprintf "%04d%02d%02d", 1900 + $year, ++$mon, $dom;
}
425

426 427 428 429 430
sub today_dash {
    my ($dom, $mon, $year) = (localtime(time))[3, 4, 5];
    return sprintf "%04d-%02d-%02d", 1900 + $year, ++$mon, $dom;
}

431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
sub sqlday {
    my ($day, $base) = @_;
    $day = ($day - $base) * 86400;
    my ($dom, $mon, $year) = (gmtime($day))[3, 4, 5];
    return sprintf "%04d%02d%02d", 1900 + $year, ++$mon, $dom;
}

sub delta_time {
    my $tstart = shift;
    my $tend = shift;
    my $delta = $tend - $tstart;
    my $hours = int($delta/3600);
    my $minutes = int($delta/60) - ($hours * 60);
    my $seconds = $delta - ($minutes * 60) - ($hours * 3600);
    return sprintf("%02d:%02d:%02d" , $hours, $minutes, $seconds);
}
447 448 449 450 451 452 453 454 455 456

sub CollectSeriesData {
    # We need some way of randomising the distribution of series, such that
    # all of the series which are to be run every 7 days don't run on the same
    # day. This is because this might put the server under severe load if a
    # particular frequency, such as once a week, is very common. We achieve
    # this by only running queries when:
    # (days_since_epoch + series_id) % frequency = 0. So they'll run every
    # <frequency> days, but the start date depends on the series_id.
    my $days_since_epoch = int(time() / (60 * 60 * 24));
457
    my $today = $ARGV[0] || today_dash();
458

459 460
    # We save a copy of the main $dbh and then switch to the shadow and get
    # that one too. Remember, these may be the same.
461 462
    my $dbh = Bugzilla->switch_to_main_db();
    my $shadow_dbh = Bugzilla->switch_to_shadow_db();
463
    
464
    my $serieses = $dbh->selectall_hashref("SELECT series_id, query, creator " .
465 466 467 468 469 470 471
                      "FROM series " .
                      "WHERE frequency != 0 AND " . 
                      "($days_since_epoch + series_id) % frequency = 0",
                      "series_id");

    # We prepare the insertion into the data table, for efficiency.
    my $sth = $dbh->prepare("INSERT INTO series_data " .
472
                            "(series_id, series_date, series_value) " .
473 474
                            "VALUES (?, " . $dbh->quote($today) . ", ?)");

475 476 477
    # We delete from the table beforehand, to avoid SQL errors if people run
    # collectstats.pl twice on the same day.
    my $deletesth = $dbh->prepare("DELETE FROM series_data 
478
                                   WHERE series_id = ? AND series_date = " .
479 480
                                   $dbh->quote($today));
                                     
481 482 483
    foreach my $series_id (keys %$serieses) {
        # We set up the user for Search.pm's permission checking - each series
        # runs with the permissions of its creator.
484
        my $user = new Bugzilla::User($serieses->{$series_id}->{'creator'});
485 486 487

        my $cgi = new Bugzilla::CGI($serieses->{$series_id}->{'query'});
        my $search = new Bugzilla::Search('params' => $cgi,
488 489
                                          'fields' => ["bugs.bug_id"],
                                          'user'   => $user);
490 491
        my $sql = $search->getSQL();
        
492 493 494 495 496 497 498 499
        my $data;
        
        # We can't die if we get dodgy SQL back for whatever reason, so we
        # eval() this and, if it fails, just ignore it and carry on.
        # One day we might even log an error.
        eval { 
            $data = $shadow_dbh->selectall_arrayref($sql);
        };
500
        
501 502 503 504
        if (!$@) {
            # We need to count the returned rows. Without subselects, we can't
            # do this directly in the SQL for all queries. So we do it by hand.
            my $count = scalar(@$data) || 0;
505

506 507 508
            $deletesth->execute($series_id);
            $sth->execute($series_id, $count);
        }
509 510 511
    }
}