db-query-builder.md 15.7 KB
Newer Older
Alexander Makarov committed
1 2 3
Query Builder and Query
=======================

4
> Note: This section is under development.
Qiang Xue committed
5

6 7 8 9
Yii provides a basic database access layer as described in the [Database basics](database-basics.md) section.
The database access layer provides a low-level way to interact with the database. While useful in some situations,
it can be tedious and error-prone to write raw SQLs. An alternative approach is to use the Query Builder.
The Query Builder provides an object-oriented vehicle for generating queries to be executed.
Alexander Makarov committed
10

11
A typical usage of the query builder looks like the following:
Alexander Makarov committed
12 13

```php
14
$rows = (new \yii\db\Query())
15
    ->select('id, name')
16
    ->from('user')
17 18
    ->limit(10)
    ->all();
19 20

// which is equivalent to the following code:
Alexander Makarov committed
21

22
$query = (new \yii\db\Query())
23
    ->select('id, name')
24
    ->from('user')
25
    ->limit(10);
Alexander Makarov committed
26

27
// Create a command. You can get the actual SQL using $command->sql
Alexander Makarov committed
28
$command = $query->createCommand();
29 30

// Execute the command:
Alexander Makarov committed
31 32 33
$rows = $command->queryAll();
```

34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
Query Methods
-------------

As you can see, [[yii\db\Query]] is the main player that you need to deal with. Behind the scene,
`Query` is actually only responsible for representing various query information. The actual query
building logic is done by [[yii\db\QueryBuilder]] when you call the `createCommand()` method,
and the query execution is done by [[yii\db\Command]].

For convenience, [[yii\db\Query]] provides a set of commonly used query methods that will build
the query, execute it, and return the result. For example,

- [[yii\db\Query::all()|all()]]: builds the query, executes it and returns all results as an array.
- [[yii\db\Query::one()|one()]]: returns the first row of the result.
- [[yii\db\Query::column()|column()]]: returns the first column of the result.
- [[yii\db\Query::scalar()|scalar()]]: returns the first column in the first row of the result.
- [[yii\db\Query::exists()|exists()]]: returns a value indicating whether the query results in anything.
- [[yii\db\Query::count()|count()]]: returns the result of a `COUNT` query. Other similar methods
51 52
  include `sum($q)`, `average($q)`, `max($q)`, `min($q)`, which support the so-called aggregational data query. `$q`
  parameter is mandatory for these methods and can be either the column name or expression.
53 54 55 56 57


Building Query
--------------

58 59 60 61
In the following, we will explain how to build various clauses in a SQL statement. For simplicity,
we use `$query` to represent a [[yii\db\Query]] object.


62
### `SELECT`
Alexander Makarov committed
63

64
In order to form a basic `SELECT` query, you need to specify what columns to select and from what table:
Alexander Makarov committed
65 66 67

```php
$query->select('id, name')
68
    ->from('user');
Alexander Makarov committed
69 70
```

71 72
Select options can be specified as a comma-separated string, as in the above, or as an array.
The array syntax is especially useful when forming the selection dynamically:
Alexander Makarov committed
73 74

```php
75
$query->select(['id', 'name'])
76
    ->from('user');
Alexander Makarov committed
77 78
```

79 80 81 82 83
> Info: You should always use the array format if your `SELECT` clause contains SQL expressions.
> This is because a SQL expression like `CONCAT(first_name, last_name) AS full_name` may contain commas.
> If you list it together with other columns in a string, the expression may be split into several parts
> by commas, which is not what you want to see.

84
When specifying columns, you may include the table prefixes or column aliases, e.g., `user.id`, `user.id AS user_id`.
85
If you are using array to specify the columns, you may also use the array keys to specify the column aliases,
86
e.g., `['user_id' => 'user.id', 'user_name' => 'user.name']`.
87 88 89 90

To select distinct rows, you may call `distinct()`, like the following:

```php
91
$query->select('user_id')->distinct()->from('post');
92 93
```

94
### `FROM`
95 96 97 98

To specify which table(s) to select data from, call `from()`:

```php
99
$query->select('*')->from('user');
100 101 102
```

You may specify multiple tables using a comma-separated string or an array.
103
Table names can contain schema prefixes (e.g. `'public.user'`) and/or table aliases (e.g. `'user u'`).
104 105 106 107
The method will automatically quote the table names unless it contains some parenthesis
(which means the table is given as a sub-query or DB expression). For example,

```php
108
$query->select('u.*, p.*')->from(['user u', 'post p']);
109 110 111 112 113 114
```

When the tables are specified as an array, you may also use the array keys as the table aliases
(if a table does not need alias, do not use a string key). For example,

```php
maks feltrin committed
115
$query->select('u.*, p.*')->from(['u' => 'user', 'p' => 'post']);
116 117 118 119
```

You may specify a sub-query using a `Query` object. In this case, the corresponding array key will be used
as the alias for the sub-query.
Qiang Xue committed
120

121
```php
122
$subQuery = (new Query())->select('id')->from('user')->where('status=1');
123 124
$query->select('*')->from(['u' => $subQuery]);
```
Qiang Xue committed
125

126

127
### `WHERE`
Alexander Makarov committed
128

129
Usually data is selected based upon certain criteria. Query Builder has some useful methods to specify these, the most powerful of which being `where`. It can be used in multiple ways.
Alexander Makarov committed
130

131
The simplest way to apply a condition is to use a string:
Alexander Makarov committed
132 133

```php
Alexander Makarov committed
134
$query->where('status=:status', [':status' => $status]);
Alexander Makarov committed
135 136
```

137
When using strings, make sure you're binding the query parameters, not creating a query by string concatenation. The above approach is safe to use, the following is not:
Alexander Makarov committed
138

139 140 141 142 143
```php
$query->where("status=$status"); // Dangerous!
```

Instead of binding the status value immediately, you can do so using `params` or `addParams`:
Alexander Makarov committed
144 145 146

```php
$query->where('status=:status');
Alexander Makarov committed
147
$query->addParams([':status' => $status]);
Alexander Makarov committed
148 149
```

150
Multiple conditions can simultaneously be set in `where` using the *hash format*:
Alexander Makarov committed
151 152

```php
Alexander Makarov committed
153
$query->where([
154 155 156
    'status' => 10,
    'type' => 2,
    'id' => [4, 8, 15, 16, 23, 42],
Alexander Makarov committed
157
]);
Alexander Makarov committed
158 159
```

160
That code will generate the following SQL:
Alexander Makarov committed
161 162 163 164 165

```sql
WHERE (`status` = 10) AND (`type` = 2) AND (`id` IN (4, 8, 15, 16, 23, 42))
```

166
NULL is a special value in databases, and is handled smartly by the Query Builder. This code:
Alexander Makarov committed
167 168

```php
Alexander Makarov committed
169
$query->where(['status' => null]);
Alexander Makarov committed
170 171
```

172
results in this WHERE clause:
Alexander Makarov committed
173 174 175 176 177

```sql
WHERE (`status` IS NULL)
```

178 179 180 181 182 183 184 185 186 187 188 189 190 191
You can also create sub-queries with `Query` objects like the following,

```php
$userQuery = (new Query)->select('id')->from('user');
$query->where(['id' => $userQuery]);
```

which will generate the following SQL:

```sql
WHERE `id` IN (SELECT `id` FROM `user`)
```


Alexander Makarov committed
192
Another way to use the method is the operand format which is `[operator, operand1, operand2, ...]`.
Alexander Makarov committed
193 194 195 196

Operator can be one of the following:

- `and`: the operands should be concatenated together using `AND`. For example,
Alexander Makarov committed
197
  `['and', 'id=1', 'id=2']` will generate `id=1 AND id=2`. If an operand is an array,
Alexander Makarov committed
198
  it will be converted into a string using the rules described here. For example,
Alexander Makarov committed
199
  `['and', 'type=1', ['or', 'id=1', 'id=2']]` will generate `type=1 AND (id=1 OR id=2)`.
Alexander Makarov committed
200
  The method will NOT do any quoting or escaping.
201

Alexander Makarov committed
202
- `or`: similar to the `and` operator except that the operands are concatenated using `OR`.
203

Alexander Makarov committed
204 205
- `between`: operand 1 should be the column name, and operand 2 and 3 should be the
   starting and ending values of the range that the column is in.
Alexander Makarov committed
206
   For example, `['between', 'id', 1, 10]` will generate `id BETWEEN 1 AND 10`.
207

Alexander Makarov committed
208 209
- `not between`: similar to `between` except the `BETWEEN` is replaced with `NOT BETWEEN`
  in the generated condition.
210

211 212 213 214
- `in`: operand 1 should be a column or DB expression. Operand 2 can be either an array or a `Query` object.
  It will generate an `IN` condition. If Operand 2 is an array, it will represent the range of the values
  that the column or DB expression should be; If Operand 2 is a `Query` object, a sub-query will be generated
  and used as the range of the column or DB expression. For example,
Alexander Makarov committed
215
  `['in', 'id', [1, 2, 3]]` will generate `id IN (1, 2, 3)`.
Alexander Makarov committed
216
  The method will properly quote the column name and escape values in the range.
217 218
  The `in` operator also supports composite columns. In this case, operand 1 should be an array of the columns,
  while operand 2 should be an array of arrays or a `Query` object representing the range of the columns.
219

Alexander Makarov committed
220
- `not in`: similar to the `in` operator except that `IN` is replaced with `NOT IN` in the generated condition.
221

Alexander Makarov committed
222 223
- `like`: operand 1 should be a column or DB expression, and operand 2 be a string or an array representing
  the values that the column or DB expression should be like.
224
  For example, `['like', 'name', 'tester']` will generate `name LIKE '%tester%'`.
Alexander Makarov committed
225
  When the value range is given as an array, multiple `LIKE` predicates will be generated and concatenated
226
  using `AND`. For example, `['like', 'name', ['test', 'sample']]` will generate
Alexander Makarov committed
227
  `name LIKE '%test%' AND name LIKE '%sample%'`.
228 229 230 231 232 233
  You may also provide an optional third operand to specify how to escape special characters in the values.
  The operand should be an array of mappings from the special characters to their
  escaped counterparts. If this operand is not provided, a default escape mapping will be used.
  You may use `false` or an empty array to indicate the values are already escaped and no escape
  should be applied. Note that when using an escape mapping (or the third operand is not provided),
  the values will be automatically enclosed within a pair of percentage characters.
234 235 236 237

  > Note: When using PostgreSQL you may also use [`ilike`](http://www.postgresql.org/docs/8.3/static/functions-matching.html#FUNCTIONS-LIKE)
  > instead of `like` for case-insensitive matching.

Alexander Makarov committed
238 239
- `or like`: similar to the `like` operator except that `OR` is used to concatenate the `LIKE`
  predicates when operand 2 is an array.
240

Alexander Makarov committed
241 242
- `not like`: similar to the `like` operator except that `LIKE` is replaced with `NOT LIKE`
  in the generated condition.
243

Alexander Makarov committed
244 245
- `or not like`: similar to the `not like` operator except that `OR` is used to concatenate
  the `NOT LIKE` predicates.
246

247
- `exists`: requires one operand which must be an instance of [[yii\db\Query]] representing the sub-query.
248
  It will build a `EXISTS (sub-query)` expression.
249

250
- `not exists`: similar to the `exists` operator and builds a `NOT EXISTS (sub-query)` expression.
Alexander Makarov committed
251

252 253 254 255 256 257 258 259 260 261
Additionally you can specify anything as operator:

```php
$userQuery = (new Query)->select('id')->from('user');
$query->where(['>=', 'id', 10]);
```

It will result in:

```sql
262
SELECT id FROM user WHERE id >= 10;
263 264
```

Qiang Xue committed
265
If you are building parts of condition dynamically it's very convenient to use `andWhere()` and `orWhere()`:
Alexander Makarov committed
266 267 268 269 270

```php
$status = 10;
$search = 'yii';

Alexander Makarov committed
271
$query->where(['status' => $status]);
Alexander Makarov committed
272
if (!empty($search)) {
273
    $query->andWhere(['like', 'title', $search]);
Alexander Makarov committed
274 275 276 277 278 279 280 281 282
}
```

In case `$search` isn't empty the following SQL will be generated:

```sql
WHERE (`status` = 10) AND (`title` LIKE '%yii%')
```

Qiang Xue committed
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
#### Building Filter Conditions

When building filter conditions based on user inputs, you usually want to specially handle "empty inputs"
by ignoring them in the filters. For example, you have an HTML form that takes username and email inputs.
If the user only enters something in the username input, you may want to build a query that only tries to
match the entered username. You may use the `filterWhere()` method achieve this goal:

```php
// $username and $email are from user inputs
$query->filterWhere([
    'username' => $username,
    'email' => $email,
]);
```

The `filterWhere()` method is very similar to `where()`. The main difference is that `filterWhere()`
will remove empty values from the provided condition. So if `$email` is "empty", the resulting query
will be `...WHERE username=:username`; and if both `$username` and `$email` are "empty", the query
will have no `WHERE` part.

A value is *empty* if it is null, an empty string, a string consisting of whitespaces, or an empty array.

You may also use `andFilterWhere()` and `orFilterWhere()` to append more filter conditions.


308
### `ORDER BY`
Alexander Makarov committed
309

310
For ordering results `orderBy` and `addOrderBy` could be used:
Alexander Makarov committed
311 312

```php
Alexander Makarov committed
313
$query->orderBy([
314 315
    'id' => SORT_ASC,
    'name' => SORT_DESC,
Alexander Makarov committed
316
]);
Alexander Makarov committed
317 318 319 320
```

Here we are ordering by `id` ascending and then by `name` descending.

321
### `GROUP BY` and `HAVING`
Alexander Makarov committed
322 323 324 325 326 327 328 329 330 331

In order to add `GROUP BY` to generated SQL you can use the following:

```php
$query->groupBy('id, status');
```

If you want to add another field after using `groupBy`:

```php
Alexander Makarov committed
332
$query->addGroupBy(['created_at', 'updated_at']);
Alexander Makarov committed
333 334 335 336 337 338
```

To add a `HAVING` condition the corresponding `having` method and its `andHaving` and `orHaving` can be used. Parameters
for these are similar to the ones for `where` methods group:

```php
Alexander Makarov committed
339
$query->having(['status' => $status]);
Alexander Makarov committed
340 341
```

342
### `LIMIT` and `OFFSET`
Alexander Makarov committed
343 344 345 346 347 348 349 350 351 352 353 354 355

To limit result to 10 rows `limit` can be used:

```php
$query->limit(10);
```

To skip 100 fist rows use:

```php
$query->offset(100);
```

356 357 358 359 360 361 362 363 364 365 366
### `JOIN`

The `JOIN` clauses are generated in the Query Builder by using the applicable join method:

- `innerJoin()`
- `leftJoin()`
- `rightJoin()`

This left join selects data from two related tables in one query:

```php
367 368 369
$query->select(['user.name AS author', 'post.title as title'])
    ->from('user')
    ->leftJoin('post', 'post.user_id = user.id');
370 371 372 373 374 375 376 377
```

In the code, the `leftJoin()` method's first parameter
specifies the table to join to. The second parameter defines the join condition.

If your database application supports other join types, you can use those via the  generic `join` method:

```php
378
$query->join('FULL OUTER JOIN', 'post', 'post.user_id = user.id');
379 380 381 382 383 384 385 386 387 388 389 390 391 392
```

The first argument is the join type to perform. The second is the table to join to, and the third is the condition.

Like `FROM`, you may also join with sub-queries. To do so, specify the sub-query as an array
which must contain one element. The array value must be a `Query` object representing the sub-query,
while the array key is the alias for the sub-query. For example,

```php
$query->leftJoin(['u' => $subQuery], 'u.id=author_id');
```


### `UNION`
Alexander Makarov committed
393 394 395 396 397

`UNION` in SQL adds results of one query to results of another query. Columns returned by both queries should match.
In Yii in order to build it you can first form two query objects and then use `union` method:

```php
398
$query = new Query();
399
$query->select("id, category_id as type, name")->from('post')->limit(10);
Alexander Makarov committed
400

401
$anotherQuery = new Query();
402
$anotherQuery->select('id, type, name')->from('user')->limit(10);
Alexander Makarov committed
403 404 405 406

$query->union($anotherQuery);
```

407 408 409 410 411 412 413 414 415 416 417 418 419 420

Batch Query
-----------

When working with large amount of data, methods such as [[yii\db\Query::all()]] are not suitable
because they require loading all data into the memory. To keep the memory requirement low, Yii
provides the so-called batch query support. A batch query makes uses of data cursor and fetches
data in batches.

Batch query can be used like the following:

```php
use yii\db\Query;

421
$query = (new Query())
422
    ->from('user')
423
    ->orderBy('id');
424

Qiang Xue committed
425
foreach ($query->batch() as $users) {
426
    // $users is an array of 100 or fewer rows from the user table
427 428
}

Qiang Xue committed
429
// or if you want to iterate the row one by one
430
foreach ($query->each() as $user) {
431
    // $user represents one row of data from the user table
432 433
}
```
434

Qiang Xue committed
435 436 437 438 439 440 441 442 443
The method [[yii\db\Query::batch()]] and [[yii\db\Query::each()]] return an [[yii\db\BatchQueryResult]] object
which implements the `Iterator` interface and thus can be used in the `foreach` construct.
During the first iteration, a SQL query is made to the database. Data are since then fetched in batches
in the iterations. By default, the batch size is 100, meaning 100 rows of data are being fetched in each batch.
You can change the batch size by passing the first parameter to the `batch()` or `each()` method.

Compared to the [[yii\db\Query::all()]], the batch query only loads 100 rows of data at a time into the memory.
If you process the data and then discard it right away, the batch query can help keep the memory usage under a limit.

444 445 446 447 448 449
If you specify the query result to be indexed by some column via [[yii\db\Query::indexBy()]], the batch query
will still keep the proper index. For example,

```php
use yii\db\Query;

450
$query = (new Query())
451
    ->from('user')
452
    ->indexBy('username');
453

Qiang Xue committed
454
foreach ($query->batch() as $users) {
455
    // $users is indexed by the "username" column
456 457 458 459 460
}

foreach ($query->each() as $username => $user) {
}
```