Skip to content

Commit

Permalink
readme.md: updated doc
Browse files Browse the repository at this point in the history
  • Loading branch information
dg committed Jun 29, 2017
1 parent 2c000fe commit 9358712
Showing 1 changed file with 28 additions and 33 deletions.
61 changes: 28 additions & 33 deletions readme.md
Original file line number Diff line number Diff line change
Expand Up @@ -13,49 +13,54 @@ Nette provides a powerful layer for accessing your database easily.
- easily fetches data
- uses efficient queries and does not transmit unnecessary data

The `Nette\Database\Connection` class is a wrapper around the PDO and represents a connection to the database.
The core functionality is provided by `Nette\Database\Context`. `Nette\Database\Table` layer provides an enhanced layer for table querying.
The [Nette Database Core](https://doc.nette.org/database-core) is a wrapper around the PDO and provides core functionality.

To create a new database connection just create a new instance of [api:Nette\Database\Connection] class:
The [Nette Database Explorer](https://doc.nette.org/database-explorer) layer helps you to fetch database data more easily and in a more optimized way.

```php
$connection = new Nette\Database\Connection($dsn, $user, $password);
```

All connections are created as "lazy" by default. This means the connection is established when it's needed, not when you create a `Connection` instance. You can disable this behavior by passing `'lazy' => FALSE` configuration.
Documentation
-------------

This is just a piece of documentation. [Please see our website](https://doc.nette.org/database).


Queries
--------
Database Core
-------------

The core functionality is provided by `Nette\Database\Connection`. Connection allows you to easily query your database by calling `query` method:
To create a new database connection just create a new instance of `Nette\Database\Connection` class:

```php
$database = new Nette\Database\Context($connection);
$database = new Nette\Database\Connection($dsn, $user, $password); // the same arguments as uses PDO
```

Connection allows you to easily query your database by calling `query` method:

$database->query('INSERT INTO users', array( // an array can be a parameter
```php
$database->query('INSERT INTO users', [ // an array can be a parameter
'name' => 'Jim',
'created' => new DateTime, // or a DateTime object
'avatar' => fopen('image.gif', 'r'), // or a file
), ...); // it is even possible to use multiple inserts
], ...); // it is even possible to use multiple inserts

$database->query('UPDATE users SET ? WHERE id=?', $data, $id);
$database->query('SELECT * FROM categories WHERE id=?', 123)->dump();
```

Table Selection
---------------
Database Explorer
-----------------

Nette Database Explorer layer helps you to fetch database data more easily and in a more optimized way. The primary attitude is to fetch data only from one table and fetch them at once. The data are fetched into `ActiveRow` instances. Data from other tables connected by relationships are delivered by another queries - this is maintained by Database Explorer layer itself.

`Nette\Database\Table` layer helps you to fetch database data more easily and in a more optimized way. **The primary attitude is to fetch data only from one table and fetch them at once.** The data are fetched into [ActiveRow | database-activerow] instances. Data from other tables connected by relationships are delivered by another queries - this is maintained by Database\Table layer itself.
Let's take a look at common use-case. You need to fetch books and their authors. It is common 1:N relationship. The often used implementation fetches data by one SQL query with table joins. The second possibility is to fetch data separately, run one query for getting books and then get an author for each book by another query (e.g. in your foreach cycle). This could be easily optimized to run only two queries, one for books, and another for the needed authors - and this is just the way how Nette Database Explorer does it.

Let's take a look at common use-case. You need to fetch books and their authors. It is common 1:N relationship. The often used implementation fetches data by one SQL query with table joins. The second possibility is to fetch data separately, run one query for getting books and then get an author for each book by another query (e.g. in your foreach cycle). This could be easily optimized to run only two queries, one for books, and another for the needed authors - and this is just the way how Nette\Database\Table does it.
Selecting data starts with the table, just call `$context->table()` on the `Nette\Database\Context` object. The easiest way to get it is [described here](https://doc.nette.org/database-core#toc-configuration), but if we use Nette Database Explorer alone, it can be [manually created](https://doc.nette.org/database-explorer#toc-manual-creating-nette-database-context).

Creating Selection is quite easy, just call `table()` method on your database context.

```php
$selection = $context->table('book'); // db table name is "book"
```

Selection implements traversable interface: you can just iterate over the instance to get all books. The rows are fetched as ActiveRow instances; you can read row data from their properties.
We can simply iterate over the selection and pass through all the books. The rows are fetched as ActiveRow instances; you can read row data from their properties.

```php
$books = $context->table('book');
Expand All @@ -65,7 +70,7 @@ foreach ($books as $book) {
}
```

Getting just one specific row is done by `get()` method. It is "filtering" method, which directly returns an ActiveRow instance.
Getting just one specific row is done by `get()` method, which directly returns an ActiveRow instance.

```php
$book = $context->table('book')->get(2); // returns book with id 2
Expand All @@ -76,18 +81,6 @@ echo $book->author_id;
Working with relationships
--------------------------

As we mentioned in the chapter intro, Database\Table layer maintains the table relations for you. There are two possibilities how and where you can work with relationships.

1. **Filtering rows fetched by Selection.** In the introduction we stated the basic principle to select data only from one database table at once. However, Selection instance can do a table join to filter selected row. For example you need select only that authors who has written more than 2 books.
2. **Getting related data for fetched ActiveRows.** We denied getting data from more than one table at once. Sadly, printing `author_id` is not good enough. We need to get full author database row, ideally fetched as ActiveRow. Getting this type of relationships is maintained by ActiveRow.


In provided examples we will work with this database schema below. There are common OneHasMany and ManyHasMany relationships. OneHasMany relationship is doubled, a book must have an author and could have a translator (`translator_id` could be a `NULL`).

![](https://files.nette.org/git/doc-2.1/db-schema-1-.png)

In example below we are getting related data for fetched books. In author property (of book ActiveRow instances) is available another ActiveRow instance, which represents author of the book. Getting book_tag instances is done by `related()` method, which returns collection of this instances. In the cycle we get the tag name from another ActiveRow instance available in book_tag instance.

```php
$books = $context->table('book');

Expand All @@ -111,11 +104,13 @@ SELECT * FROM `book_tag` WHERE (`book_tag`.`book_id` IN (1, 4, 2, 3))
SELECT * FROM `tag` WHERE (`tag`.`id` IN (21, 22, 23))
```

If you use cache (defaults on), no columns will be queried unnecessarily. After the first query, cache will store the used column names and Nette\Database will run queries only with the needed columns:
If you use caching (defaults on), no columns will be queried unnecessarily. After the first query, cache will store the used column names and Nette Database Explorer will run queries only with the needed columns:

```sql
SELECT `id`, `title`, `author_id` FROM `book`
SELECT `id`, `name` FROM `author` WHERE (`author`.`id` IN (11, 12))
SELECT `book_id`, `tag_id` FROM `book_tag` WHERE (`book_tag`.`book_id` IN (1, 4, 2, 3))
SELECT `id`, `name` FROM `tag` WHERE (`tag`.`id` IN (21, 22, 23))
```

[Continue…](https://doc.nette.org/database-explorer).

0 comments on commit 9358712

Please sign in to comment.