summaryrefslogtreecommitdiffstats
path: root/collectstats.pl
blob: d8987b2b99ce1d5ec06aef25c094d4111837151f (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
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
319
320
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
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
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
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
#!/usr/bin/perl -w
# -*- Mode: perl; indent-tabs-mode: nil -*-
#
# 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.
#
# The Original Code is the Bugzilla Bug Tracking System.
#
# The Initial Developer of the Original Code is Netscape Communications
# Corporation. Portions created by Netscape are
# Copyright (C) 1998 Netscape Communications Corporation. All
# Rights Reserved.
#
# Contributor(s): Terry Weissman <terry@mozilla.org>,
#                 Harrison Page <harrison@netscape.com>
#                 Gervase Markham <gerv@gerv.net>
#                 Richard Walters <rwalters@qualcomm.com>
#                 Jean-Sebastien Guay <jean_seb@hybride.com>
#                 Frédéric Buclin <LpSolit@gmail.com>

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

use strict;
use lib qw(. lib);

use List::Util qw(first);
use Cwd;

use Bugzilla;
use Bugzilla::Constants;
use Bugzilla::Error;
use Bugzilla::Util;
use Bugzilla::Search;
use Bugzilla::User;
use Bugzilla::Product;
use Bugzilla::Field;

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

# Tidy up after graphing module
my $cwd = Cwd::getcwd();
if (chdir("graphs")) {
    unlink <./*.gif>;
    unlink <./*.png>;
    # chdir("..") doesn't work if graphs is a symlink, see bug 429378
    chdir($cwd);
}

my $dbh = Bugzilla->switch_to_shadow_db();


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

my $datadir = bz_locations()->{'datadir'};

my @myproducts = map {$_->name} Bugzilla::Product->get_all;
unshift(@myproducts, "-All-");

# As we can now customize statuses and resolutions, looking at the current list
# of legal values only is not enough as some now removed statuses and resolutions
# may have existed in the past, or have been renamed. We want them all.
my $fields = {};
foreach my $field ('bug_status', 'resolution') {
    my $values = get_legal_field_values($field);
    my $old_values = $dbh->selectcol_arrayref(
                             "SELECT bugs_activity.added
                                FROM bugs_activity
                          INNER JOIN fielddefs
                                  ON fielddefs.id = bugs_activity.fieldid
                           LEFT JOIN $field
                                  ON $field.value = bugs_activity.added
                               WHERE fielddefs.name = ?
                                 AND $field.id IS NULL

                               UNION

                              SELECT bugs_activity.removed
                                FROM bugs_activity
                          INNER JOIN fielddefs
                                  ON fielddefs.id = bugs_activity.fieldid
                           LEFT JOIN $field
                                  ON $field.value = bugs_activity.removed
                               WHERE fielddefs.name = ?
                                 AND $field.id IS NULL",
                               undef, ($field, $field));

    push(@$values, @$old_values);
    $fields->{$field} = $values;
}

my @statuses = @{$fields->{'bug_status'}};
my @resolutions = @{$fields->{'resolution'}};
# Exclude "" from the resolution list.
@resolutions = grep {$_} @resolutions;

# --regenerate was taking an enormous amount of time to query everything
# per bug, per day. Instead, we now just get all the data out of the DB
# at once and stuff it into some data structures.
my (%bug_status, %bug_resolution, %removed);
if ($regenerate) {
    %bug_resolution = @{ $dbh->selectcol_arrayref(
        'SELECT bug_id, resolution FROM bugs', {Columns=>[1,2]}) };
    %bug_status = @{ $dbh->selectcol_arrayref(
        'SELECT bug_id, bug_status FROM bugs', {Columns=>[1,2]}) };

    my $removed_sth = $dbh->prepare(
        q{SELECT bugs_activity.bug_id, bugs_activity.removed,}
        . $dbh->sql_to_days('bugs_activity.bug_when')
        . q{FROM bugs_activity
           WHERE bugs_activity.fieldid = ?
        ORDER BY bugs_activity.bug_when});

    %removed = (bug_status => {}, resolution => {});
    foreach my $field (qw(bug_status resolution)) {
        my $field_id = Bugzilla::Field->check($field)->id;
        my $rows = $dbh->selectall_arrayref($removed_sth, undef, $field_id);
        my $hash = $removed{$field};
        foreach my $row (@$rows) {
            my ($bug_id, $removed, $when) = @$row;
            $hash->{$bug_id} ||= [];
            push(@{ $hash->{$bug_id} }, { when    => int($when),
                                          removed => $removed });
        }
    }
}

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

    &check_data_dir ($dir);

    if ($regenerate) {
        regenerate_stats($dir, $_, \%bug_resolution, \%bug_status, \%removed);
    } else {
        &collect_stats($dir, $_);
    }
}
my $tend = time;
# Uncomment the following line for performance testing.
#print "Total time taken " . delta_time($tstart, $tend) . "\n";

CollectSeriesData();

sub check_data_dir {
    my $dir = shift;

    if (! -d $dir) {
        mkdir $dir, 0755;
        chmod 0755, $dir;
    }
}

sub collect_stats {
    my $dir = shift;
    my $product = shift;
    my $when = localtime (time);
    my $dbh = Bugzilla->dbh;

    my $product_id;
    if ($product ne '-All-') {
        my $prod = Bugzilla::Product::check_product($product);
        $product_id = $prod->id;
    }

    # 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 $exists = -f $file;

    # if the file exists, get the old status and resolution list for that product.
    my @data;
    @data = get_old_data($file) if $exists;

    # If @data is not empty, then we have to recreate the data file.
    if (scalar(@data)) {
        open(DATA, '>', $file)
          || ThrowCodeError('chart_file_open_fail', {'filename' => $file});
    }
    else {
        open(DATA, '>>', $file)
          || ThrowCodeError('chart_file_open_fail', {'filename' => $file});
    }

    if (Bugzilla->params->{'utf8'}) {
        binmode DATA, ':utf8';
    }

    # Now collect current data.
    my @row = (today());
    my $status_sql = q{SELECT COUNT(*) FROM bugs WHERE bug_status = ?};
    my $reso_sql   = q{SELECT COUNT(*) FROM bugs WHERE resolution = ?};

    if ($product ne '-All-') {
        $status_sql .= q{ AND product_id = ?};
        $reso_sql   .= q{ AND product_id = ?};
    }

    my $sth_status = $dbh->prepare($status_sql);
    my $sth_reso   = $dbh->prepare($reso_sql);

    my @values ;
    foreach my $status (@statuses) {
        @values = ($status);
        push (@values, $product_id) if ($product ne '-All-');
        my $count = $dbh->selectrow_array($sth_status, undef, @values);
        push(@row, $count);
    }
    foreach my $resolution (@resolutions) {
        @values = ($resolution);
        push (@values, $product_id) if ($product ne '-All-');
        my $count = $dbh->selectrow_array($sth_reso, undef, @values);
        push(@row, $count);
    }

    if (!$exists || scalar(@data)) {
        my $fields = join('|', ('DATE', @statuses, @resolutions));
        print DATA <<FIN;
# Bugzilla Daily Bug Stats
#
# Do not edit me! This file is generated.
#
# fields: $fields
# Product: $product
# Created: $when
FIN
    }

    # Add existing data, if needed. Note that no count is not treated
    # the same way as a count with 0 bug.
    foreach my $data (@data) {
        print DATA join('|', map {defined $data->{$_} ? $data->{$_} : ''}
                                 ('DATE', @statuses, @resolutions)) . "\n";
    }
    print DATA (join '|', @row) . "\n";
    close DATA;
    chmod 0644, $file;
}

sub get_old_data {
    my $file = shift;

    open(DATA, '<', $file)
      || ThrowCodeError('chart_file_open_fail', {'filename' => $file});

    if (Bugzilla->params->{'utf8'}) {
        binmode DATA, ':utf8';
    }

    my @data;
    my @columns;
    my $recreate = 0;
    while (<DATA>) {
        chomp;
        next unless $_;
        if (/^# fields?:\s*(.+)\s*$/) {
            @columns = split(/\|/, $1);
            # Compare this list with @statuses and @resolutions.
            # If they are identical, then we can safely append new data
            # to the end of the file; else we have to recreate it.
            $recreate = 1;
            my @new_cols = ($columns[0], @statuses, @resolutions);
            if (scalar(@columns) == scalar(@new_cols)) {
                my $identical = 1;
                for (0 .. $#columns) {
                    $identical = 0 if ($columns[$_] ne $new_cols[$_]);
                }
                last if $identical;
            }
        }
        next unless $recreate;
        next if (/^#/); # Ignore comments.
        # If we have to recreate the file, we have to load all existing
        # data first.
        my @line = split /\|/;
        my %data;
        foreach my $column (@columns) {
            $data{$column} = shift @line;
        }
        push(@data, \%data);
    }
    close(DATA);
    return @data;
}

# This regenerates all statistics from the database.
sub regenerate_stats {
    my ($dir, $product, $bug_resolution, $bug_status, $removed) = @_;

    my $dbh = Bugzilla->dbh;
    my $when = localtime(time());
    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 $and_product = "";
    my $from_product = "";

    my @values = ();
    if ($product ne '-All-') {
        $and_product = q{ AND products.name = ?};
        $from_product = q{ INNER JOIN products 
                          ON bugs.product_id = products.id};
        push (@values, $product);
    }

    # 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.
    my $query = q{SELECT } .
                $dbh->sql_to_days('creation_ts') . q{ AS start_day, } .
                $dbh->sql_to_days('current_date') . q{ AS end_day, } . 
                $dbh->sql_to_days("'1970-01-01'") . 
                 qq{ FROM bugs $from_product 
                   WHERE } . $dbh->sql_to_days('creation_ts') . 
                         qq{ IS NOT NULL $and_product 
                ORDER BY start_day } . $dbh->sql_limit(1);
    my ($start, $end, $base) = $dbh->selectrow_array($query, undef, @values);

    if (!defined $start) {
        return;
    }

    if (open DATA, ">$file") {
        my $fields = join('|', ('DATE', @statuses, @resolutions));
        print DATA <<FIN;
# Bugzilla Daily Bug Stats
#
# Do not edit me! This file is generated.
#
# fields: $fields
# Product: $product
# Created: $when
FIN
        # For each day, generate a line of statistics.
        my $total_days = $end - $start;
        my @bugs;
        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.
            $query = qq{SELECT bug_id 
                          FROM bugs $from_product 
                         WHERE bugs.creation_ts < } . 
                         $dbh->sql_from_days($day - 1) . 
                         q{ AND bugs.creation_ts >= } . 
                         $dbh->sql_from_days($day - 2) . 
                        $and_product . q{ ORDER BY bug_id};

            my $bug_ids = $dbh->selectcol_arrayref($query, undef, @values);
            push(@bugs, @$bug_ids);

            my %bugcount;
            foreach (@statuses) { $bugcount{$_} = 0; }
            foreach (@resolutions) { $bugcount{$_} = 0; }
            # Get information on bug states and resolutions.
            for my $bug (@bugs) {
                my $status = _get_value(
                    $removed->{'bug_status'}->{$bug},
                    $bug_status,  $day, $bug);

                if (defined $bugcount{$status}) {
                    $bugcount{$status}++;
                }

                my $resolution = _get_value(
                    $removed->{'resolution'}->{$bug},
                    $bug_resolution, $day, $bug);

                if (defined $bugcount{$resolution}) {
                    $bugcount{$resolution}++;
                }
            }

            # Generate a line of output containing the date and counts
            # of bugs in each state.
            my $date = sqlday($day, $base);
            print DATA "$date";
            foreach (@statuses) { print DATA "|$bugcount{$_}"; }
            foreach (@resolutions) { 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;
    }
}

# A helper for --regenerate.
# For each bug that exists on a day, we determine its status/resolution
# at the beginning of the day.  If there were no status/resolution
# changes on or after that day, the status was the same as it
# is today (the "current" value).  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.
sub _get_value {
    my ($removed, $current, $day, $bug) = @_;

    # Get the first change that's on or after this day.
    my $item = first { $_->{when} >= $day } @{ $removed || [] };

    # If there's no change on or after this day, then we just return the
    # current value.
    return $item ? $item->{removed} : $current->{$bug};
}

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

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

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);
}

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));
    my $today = $ARGV[0] || today_dash();

    # 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.
    my $dbh = Bugzilla->switch_to_main_db();
    my $shadow_dbh = Bugzilla->switch_to_shadow_db();
    
    my $serieses = $dbh->selectall_hashref("SELECT series_id, query, creator " .
                      "FROM series " .
                      "WHERE frequency != 0 AND " . 
                      "MOD(($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 " .
                            "(series_id, series_date, series_value) " .
                            "VALUES (?, " . $dbh->quote($today) . ", ?)");

    # 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 
                                   WHERE series_id = ? AND series_date = " .
                                   $dbh->quote($today));
                                     
    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.
        my $user = new Bugzilla::User($serieses->{$series_id}->{'creator'});
        my $cgi = new Bugzilla::CGI($serieses->{$series_id}->{'query'});
        my $data;

        # Do not die if Search->new() detects invalid data, such as an obsolete
        # login name or a renamed product or component, etc.
        eval {
            my $search = new Bugzilla::Search('params' => $cgi,
                                              'fields' => ["bug_id"],
                                              'user'   => $user);
            my $sql = $search->getSQL();
            $data = $shadow_dbh->selectall_arrayref($sql);
        };

        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;

            $deletesth->execute($series_id);
            $sth->execute($series_id, $count);
        }
    }
}