Commits
-------

4453e9fc fixed for the latest TwigFormExtension

Discussion
----------

fixed for the latest TwigFormExtension

Now TwigFormExtension requires CsrfProviderInterface as the first arg but there is no arg for this in TwigServiceProvider, so it occurs the error below:

> Catchable fatal error: Argument 1 passed to Symfony\Bridge\Twig\Extension\FormExtension::__construct() must implement interface Symfony\Component\Form\Extension\Csrf\CsrfProvider\CsrfProviderInterface, array given

then I set `$app['form.csrf_provider']` as the first arg.

Commits
-------

8d9bb8fb Added tests for Validator Provider
291bac8d Added Validator Component as submodule

Discussion
----------

Added tests for Validator Provider

I also had to add the Validator Component as a submodule

Commits
-------

90f24072 added mention of SymfonyBridgesServiceProvider to example of using path() Twig function

Discussion
----------

added mention of SymfonyBridgesServiceProvider to example of using path Twig function

I just spent an hour trying to figure out why using path() in Twig when using Silex wasn't working.  Turns out I hadn't included the SymfonyBridgesServiceProvider, which isn't mentioned in this piece of documentation.

I don't know if this makes the comment too verbose, but if it saves someone else wasting time it'll be worth it.

Commits
-------

80c96c85 Added link to protected closure section. See #235.

Discussion
----------

Added link to protected closure section. See #235.

Commits
-------

16599fbc fixed link to SwiftmailerServiceProvider doc

Discussion
----------

Fixed link to SwiftmailerServiceProvider doc

Commits
-------

7caa3122 [docs] remove duplicate "detect_unicode" line in pitfalls

Discussion
----------

[docs] remove duplicate "detect_unicode" line in pitfalls

Not sure what happened there, I suppose some merge failed.

Commits
-------

28399b6d add streaming to the changelog

Discussion
----------

add streaming to the changelog

Commits
-------

006d7ba4 add symfony/class-loader to composer.json

Discussion
----------

add symfony/class-loader to composer.json

Commits
-------

8dd779ec [stream] Fix rst formatting
73fe0b17 [stream] Add a nice API for streaming responses

Discussion
----------

[stream] Add a nice API for streaming responses

This PR adds a nice API for creating streaming responses with the new StreamedResponse class that was added to symfony [just recently](https://github.com/symfony/symfony/pull/2935).

Usage is documented in the usage doc, but for lazy people:

```php
<?php

$app->get('/images/{file}', function ($file) use ($app) {
    if (!file_exists(__DIR__.'/images/'.$file)) {
        return $app->abort(404, 'The image was not found.');
    }

    $stream = function () use ($file) {
        readfile($file);
    };

    return $app->stream($stream, 200, array('Content-Type' => 'image/png'));
});
```

Note: I had to point the autoloader to a copy of symfony/symfony because the subtree splits are not up-to-date. Thus the vendors are out of date too right now.

Commits
-------

58fc60f3 Adds doc about proxy headers trust

Discussion
----------

Adds doc about proxy headers trust

---------------------------------------------------------------------------

by igorw at 2011/12/21 07:07:57 -0800

No instantiation needed, just call it statically on the request class:

    Request::trustProxyData();

If there is something about this in the Symfony2 docs, it'd be nice to link to it.

---------------------------------------------------------------------------

by marcw at 2011/12/21 07:19:53 -0800

Well this is a configuration parameter in Symfony2...

Commits
-------

604b9177 [docs] set detect_unicode=off, so phar:// is no longer needed

Discussion
----------

[docs] set detect_unicode=off, so phar:// is no longer needed

detect_unicode was causing some issues with wrongly detecting the encoding. This is one of the reasons that phar:// was required, but I am not sure if this is the only case. I will leave the "phar-stub bug" section there for now.

Commits
-------

2434ad91 [doc] proofread documentation and fixed some typos.

Discussion
----------

[doc] proofread documentation and fixed some typos.

---------------------------------------------------------------------------

by igorw at 2011/11/10 05:30:58 -0800

Thumbs up!

Commits
-------

3871f24c [DoctrineServiceProvider] Fix default options was not used with tests

Discussion
----------

[DoctrineServiceProvider] Fix default options was not used

---------------------------------------------------------------------------

by fabpot at 2011/10/25 23:13:23 -0700

The code looks correct to as we are modifying the options by reference.

---------------------------------------------------------------------------

by jeromemacias at 2011/10/26 05:54:18 -0700

On the line 50 :

```php
$tmp = $app['dbs.options'];
```

This is a copy of options array, the reference used after does not has any effect on initial options array.

---------------------------------------------------------------------------

by DrBenton at 2011/10/26 06:02:46 -0700

@jeromemacias But in the ``` foreach ($tmp as $name => &$options) ``` the ```$options``` var is referenced (with the use of ```&```), not copied.
Thus, if I'm not wrong, this loop *does* have an effect on ```$app['dbs.options']```.

---------------------------------------------------------------------------

by jeromemacias at 2011/10/26 12:36:40 -0700

@DrBenton The $options var is a reference to $tmp but $tmp is a copy of $apps['dbs.options'],  not a reference.

---------------------------------------------------------------------------

by jeromemacias at 2011/10/26 12:58:38 -0700

I used this config (json) :

```json
{
  "db.options":
  {
    "dbname": "mydbname",
    "user": "mydbuser",
    "password": "mydbpass"
  }
}
```

With the current code :

```php
            $tmp = $app['dbs.options'];
            foreach ($tmp as $name => &$options) {
                $options = array_replace($app['db.default_options'], $options);

                if (!isset($app['dbs.default'])) {
                    $app['dbs.default'] = $name;
                }
            }

            var_export($app['dbs.options']); die;
```

The result is :

```php
array ( 'default' => array ( 'dbname' => 'mydbname', 'user' => 'mydbuser', 'password' => 'mydbpass', ), )
```

With the new code :

```php
            $tmp = $app['dbs.options'];
            foreach ($tmp as $name => &$options) {
                $options = array_replace($app['db.default_options'], $options);

                if (!isset($app['dbs.default'])) {
                    $app['dbs.default'] = $name;
                }
            }
            $app['dbs.options'] = $tmp;

            var_export($app['dbs.options']); die;
```

The result is :

```php
array ( 'default' => array ( 'driver' => 'pdo_mysql', 'dbname' => 'mydbname', 'host' => 'localhost', 'user' => 'mydbuser', 'password' => 'mydbpass', ), )
```

---------------------------------------------------------------------------

by igorw at 2011/10/26 13:59:51 -0700

Issue confirmed, please reopen and merge.

EDIT: To be clear: the defaults are not being applied to dbs.options.

---------------------------------------------------------------------------

by DrBenton at 2011/10/26 14:05:23 -0700

Sorry, you're right @jeromemacias. Since the values of ```$app['dbs.options']``` seem to be simple scalars, iterating though ```$tmp``` values, even by reference, does not affect ```$app['dbs.options']``` values. I thought ```$app['dbs.options']``` values where Objects (in that case, the code would have been ok).

My mistake !

---------------------------------------------------------------------------

by fabpot at 2011/10/26 14:13:41 -0700

We need some unit tests for this.

---------------------------------------------------------------------------

by jeromemacias at 2011/10/30 09:28:53 -0700

@fabpot Just added test

---------------------------------------------------------------------------

by fabpot at 2011/11/07 01:08:44 -0800

Can you squash your commits? Thanks.

---------------------------------------------------------------------------

by jeromemacias at 2011/11/07 15:18:40 -0800

@fabpot Done.

Commits
-------

554425d7 maintain line numbers when minifying php for the phar

Discussion
----------

maintain line numbers when minifying php for the phar

Currently the minification of a phar makes debugging impossible. This PR adds newlines to keep the line numbers intact.

Commits
-------

79e2684f [Silex] use @inheritdoc instead of a new phpdoc block.
46dc1dff [Silex] added missing PHPDoc blocks.

Discussion
----------

[Silex] added missing PHPDoc blocks.

Commits
-------

b711b707 [MonologServiceProvider] improve logging, log reponse status, improve test granularity

Discussion
----------

[MonologServiceProvider] improve logging, log response status, improve test granularity