summaryrefslogtreecommitdiff
path: root/rt/etc/upgrade/upgrade-mysql-schema.pl
blob: bc59c97a1c077ad6f51eb6fe90b94650c2184263 (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
#!/usr/bin/perl

use strict;
use warnings;

use DBI;
use DBD::mysql 4.002;

unless (@ARGV) {
    print STDERR "usage: $0 db_name[:server_name] db_user db_password\n";
    exit 1;
}

# pretty correct support of charsets has been introduced in mysql 4.1
# as RT doesn't use it may result in issues:
# 1) data corruptions when default charset of mysql server has data restrictions like utf8
# 2) wrong ordering (collations)

# we have to define correct types for all columns. RT uses UTF-8, ascii and binary.
# * ascii is subset of many mysql's charsets except may be one or two rare where some ascii
#   characters replaced with local
# * for many charsets mysql allows us to store any octets sequences even when those are
#   invalid for this particula set, for example we can store UTF-8 data in latin1
#   column and fetch it as UTF-8, however sorting will be wrong

# here is tricky algorithm to change column to desired charset:
# * text to binary convertion is pretty straight forward except that text types
#   have length definitions in terms of characters and in some cases we must
#   use longer binary types to satisfy space requirements
# * binary to text is much easier as we know that there is ascii or UTF-8 then
#   we just make convertion, also 32 chars are long enough to store 32 bytes, so
#   length changes is not required
# * text to text convertion is trickier. no matter what is the current character set
#   of the column we know that there is either ascii or UTF-8, so we can not use
#   direct convertion, instead we do text to binary plus binary to text convertion
#   instead
# * as well we add charset definition for all tables and for the DB as well,
#   so all new columns by default will be in UTF-8 charset

my @tables = qw(
    ACL
    Attachments
    Attributes
    CustomFields
    CustomFieldValues
    GroupMembers
    Groups
    Links
    ObjectCustomFields
    ObjectCustomFieldValues
    Principals
    Queues
    ScripActions
    ScripConditions
    Scrips
    sessions
    Templates
    Tickets
    Transactions
    Users
);

my %charset = (
    ACL                      => {
        RightName     => 'ascii',
        ObjectType    => 'ascii',
        PrincipalType => 'ascii',
    },
    Attachments              => {
        MessageId  => 'ascii',
        Subject  => 'utf8',
        Filename  => 'utf8',
        ContentType  => 'ascii',
        ContentEncoding  => 'ascii',
        Content  => 'binary',
        Headers  => 'utf8',
    },
    Attributes               => {
        Name  => 'utf8',
        Description  => 'utf8',
        Content  => 'binary',
        ContentType  => 'ascii',
        ObjectType  => 'ascii',
    },
    CustomFields             => {
        Name  => 'utf8',
        Type  => 'ascii',
        Pattern  => 'utf8',
        Description  => 'utf8',
        LookupType => 'ascii',
    },
    CustomFieldValues        => {
        Name  => 'utf8',
        Description  => 'utf8',
    },
    Groups                   => {
        Name  => 'utf8',
        Description  => 'utf8',
        Domain  => 'ascii',
        Type  => 'ascii',
    },
    Links                    => {
        Base  => 'ascii',
        Target  => 'ascii',
        Type  => 'ascii',
    },
    ObjectCustomFieldValues  => {
        ObjectType  => 'ascii',
        Content  => 'utf8',
        LargeContent  => 'binary',
        ContentType  => 'ascii',
        ContentEncoding  => 'ascii',
    },
    Principals               => {
        PrincipalType  => 'ascii',
    },
    Queues                   => {
        Name  => 'utf8',
        Description  => 'utf8',
        CorrespondAddress  => 'ascii',
        CommentAddress  => 'ascii',
    },
    ScripActions             => {
        Name  => 'utf8',
        Description  => 'utf8',
        ExecModule  => 'ascii',
        Argument  => 'binary',
    },
    ScripConditions          => {
        Name  => 'utf8',
        Description  => 'utf8',
        ExecModule  => 'ascii',
        Argument  => 'binary',
        ApplicableTransTypes  => 'ascii',
    },
    Scrips                   => {
        Description  => 'utf8',
        ConditionRules  => 'utf8',
        ActionRules  => 'utf8',
        CustomIsApplicableCode  => 'utf8',
        CustomPrepareCode  => 'utf8',
        CustomCommitCode  => 'utf8',
        Stage  => 'ascii',
    },
    sessions                 => {
        id         => 'binary', # ascii?
        a_session  => 'binary',
    },
    Templates                => {
        Name  => 'utf8',
        Description  => 'utf8',
        Type  => 'ascii',
        Language  => 'ascii',
        Content  => 'utf8',
    },
    Tickets                  => {
        Type  => 'ascii',
        Subject  => 'utf8',
        Status  => 'ascii',
    },
    Transactions             => {
        ObjectType  => 'ascii',
        Type  => 'ascii',
        Field  => 'ascii',
        OldValue  => 'utf8',
        NewValue  => 'utf8',
        ReferenceType  => 'ascii',
        Data  => 'utf8',
    },
    Users                    => {
        Name  => 'utf8',
        Password  => 'binary',
        Comments  => 'utf8',
        Signature  => 'utf8',
        EmailAddress  => 'ascii',
        FreeformContactInfo  => 'utf8',
        Organization  => 'utf8',
        RealName  => 'utf8',
        NickName  => 'utf8',
        Lang  => 'ascii',
        EmailEncoding  => 'ascii',
        WebEncoding  => 'ascii',
        ExternalContactInfoId  => 'utf8',
        ContactInfoSystem  => 'utf8',
        ExternalAuthId  => 'utf8',
        AuthSystem  => 'utf8',
        Gecos  => 'utf8',
        HomePhone  => 'utf8',
        WorkPhone  => 'utf8',
        MobilePhone  => 'utf8',
        PagerPhone  => 'utf8',
        Address1  => 'utf8',
        Address2  => 'utf8',
        City  => 'utf8',
        State  => 'utf8',
        Zip  => 'utf8',
        Country  => 'utf8',
        Timezone  => 'ascii',
        PGPKey  => 'binary',
    },
);

my %max_type_length = (
    char       => int 1<<8,
    varchar    => int 1<<8,
    tinytext   => int 1<<8,
    mediumtext => int 1<<16,
    text       => int 1<<24,
    longtext   => int 1<<32,
);

my @sql_commands;

my ($db_datasource, $db_user, $db_pass) = (shift, shift, shift);
my $dbh = DBI->connect("dbi:mysql:$db_datasource", $db_user, $db_pass, { RaiseError => 1 });
my $db_name = $db_datasource;
$db_name =~ s/:.*$//;

my $version = ($dbh->selectrow_array("show variables like 'version'"))[1];
($version) = $version =~ /^(\d+\.\d+)/;

push @sql_commands, qq{ALTER DATABASE $db_name DEFAULT CHARACTER SET utf8};
convert_table($_) foreach @tables;

print join "\n", map(/;$/? $_ : "$_;", @sql_commands), "";
my $use_p = $db_pass ? " -p" : '';
print STDERR <<ENDREMINDER;
-- ** NOTICE: No database changes have been made. **
-- Please review the generated SQL, ensure you have a full backup of your database 
-- and apply it to your database using a command like:
-- mysql -u ${db_user}${use_p} $db_name < queries.sql";
ENDREMINDER
exit 0;

my %alter_aggregator;
sub convert_table {
    my $table = shift;
    @alter_aggregator{'char_to_binary','binary_to_char'} = (['DEFAULT CHARACTER SET utf8'],[]);

    my $sth = $dbh->column_info( undef, $db_name, $table, undef );
    $sth->execute;
    while ( my $info = $sth->fetchrow_hashref ) {
        convert_column(%$info);
    }
    for my $conversiontype (qw(char_to_binary binary_to_char)) {
        next unless @{$alter_aggregator{$conversiontype}};
        push @sql_commands, qq{ALTER TABLE $table\n   }.
            join(",\n   ",@{$alter_aggregator{$conversiontype}});
    }
}

sub convert_column {
    my %info = @_;
    my $table = $info{'TABLE_NAME'};
    my $column = $info{'COLUMN_NAME'};
    my $type = $info{'TYPE_NAME'};
    return unless $type =~ /(CHAR|TEXT|BLOB|BINARY)$/i;

    my $required_charset = $charset{$table}{$column};
    unless ( $required_charset ) {
        print STDERR join(".", @info{'TABLE_SCHEMA', 'TABLE_NAME', 'COLUMN_NAME'})
            ." has type $type however mapping is missing.\n";
        return;
    }

    my $collation = column_info($table, $column)->{'collation'};
    # mysql 4.1 returns literal NULL instead of undef
    my $current_charset = $collation && $collation ne 'NULL'? (split /_/, $collation)[0]: 'binary';
    return if $required_charset eq $current_charset;

    if ( $required_charset eq 'binary' ) {
        char_to_binary(%info);
    }
    elsif ( $current_charset eq 'binary' ) {
        binary_to_char( $required_charset, %info);
    } else {
        char_to_char( $required_charset, %info);
    }
}

sub char_to_binary {
    my %info = @_;

    my $table = $info{'TABLE_NAME'};
    my $column = $info{'COLUMN_NAME'};
    my $new_type = calc_suitable_binary_type(%info);
    push @{$alter_aggregator{char_to_binary}},
        "MODIFY $column $new_type ".build_column_definition(%info);

}

sub binary_to_char {
    my ($charset, %info) = @_;

    my $table = $info{'TABLE_NAME'};
    my $column = $info{'COLUMN_NAME'};
    my $new_type = lc $info{'TYPE_NAME'};
    if ( $new_type =~ /binary/ ) {
        $new_type =~ s/binary/char/;
        $new_type .= '('. $info{'COLUMN_SIZE'} .')';
    } else {
        $new_type =~ s/blob/text/;
    }

    push @{$alter_aggregator{binary_to_char}},
        "MODIFY $column ". uc($new_type) ." CHARACTER SET ". $charset
        ." ". build_column_definition(%info);
}

sub char_to_char {
    my ($charset, %info) = @_;

    my $table = $info{'TABLE_NAME'};
    my $column = $info{'COLUMN_NAME'};
    my $new_type = $info{'mysql_type_name'};

    char_to_binary(%info);
    push @{$alter_aggregator{binary_to_char}},
        "MODIFY $column ". uc($new_type)." CHARACTER SET ". $charset
        ." ". build_column_definition(%info);
}

sub calc_suitable_binary_type {
    my %info = @_;
    my $type = lc $info{'TYPE_NAME'};
    return 'LONGBLOB' if $type eq 'longtext';

    my $current_max_byte_length = column_byte_length(@info{qw(TABLE_NAME COLUMN_NAME)}) || 0;
    if ( $max_type_length{ $type } > $current_max_byte_length ) {
        if ( $type eq 'varchar' || $type eq 'char' ) {
            my $new_type = $type;
            $new_type =~ s/char/binary/;
            $new_type .= $info{'COLUMN_SIZE'} >= $current_max_byte_length
                ? '('. $info{'COLUMN_SIZE'} .')'
                : '('. $current_max_byte_length .')';
            return uc $new_type;
        } else {
            my $new_type = $type;
            $new_type =~ s/text/blob/;
            return uc $new_type;
        }
    } else {
        my $new_type;
        foreach ( sort { $max_type_length{$a} <=> $max_type_length{$b} } keys %max_type_length ) {
            next if $max_type_length{ $_ } <= $current_max_byte_length;
            
            $new_type = $_; last;
        }
        $new_type =~ s/text/blob/;
        return uc $new_type;
    }
}

sub build_column_definition {
    my %info = @_;

    my $res = '';
    $res .= 'NOT ' unless $info{'NULLABLE'};
    $res .= 'NULL';
    my $default = column_info(@info{qw(TABLE_NAME COLUMN_NAME)})->{default};
    if ( defined $default ) {
        $res .= ' DEFAULT '. $dbh->quote($default);
    } elsif ( $info{'NULLABLE'} ) {
        $res .= ' DEFAULT NULL';
    }
    $res .= ' AUTO_INCREMENT' if $info{'mysql_is_auto_increment'};
    return $res;
}

sub column_byte_length {
    my ($table, $column) = @_;
    if ( $version >= 5.0 ) {
        my ($char, $octet) = @{ $dbh->selectrow_arrayref(
            "SELECT CHARACTER_MAXIMUM_LENGTH, CHARACTER_OCTET_LENGTH FROM information_schema.COLUMNS WHERE"
            ."     TABLE_SCHEMA = ". $dbh->quote($db_name)
            ." AND TABLE_NAME   = ". $dbh->quote($table)
            ." AND COLUMN_NAME  = ". $dbh->quote($column)
        ) };
        return $octet if $octet == $char;
    }
    return $dbh->selectrow_arrayref("SELECT MAX(LENGTH(". $dbh->quote_identifier($column) .")) FROM $table")->[0];
}

sub column_info {
    my ($table, $column) = @_;
    # XXX: DBD::mysql doesn't provide this info, may be will do in 4.0007 if I'll write a patch
    local $dbh->{FetchHashKeyName} = 'NAME_lc';
    return $dbh->selectrow_hashref("SHOW FULL COLUMNS FROM $table LIKE " . $dbh->quote($column));
}