This PR was merged into the 2.3 branch.
Discussion
----------
[DependencyInjection] make paths relative to __DIR__ in the generated container
| Q | A
| ------------- | ---
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #6484, #3079, partially #9238, #10894, #10999
| License | MIT
| Doc PR | n/a
This is an alternative approach to #10999 for removing absolute paths from the generated container:
instead of trying to fix the container file after it has been dumped, telling to the PhpDumper where its output will be written allows it to replace parts of strings by an equivalent value based on `__DIR__`.
This should be safe, thus the PR is on 2.3.
Commits
-------
edd7057 [DependencyInjection] make paths relative to __DIR__ in the generated container
* 2.2:
No Entity Manager defined exception
fixed CS
[Acl] Fix for issue #9433
[Validator] fix docblock typos
[DependencyInjection] removed the unused Reference and Parameter classes use statements from the compiled container class
Fix mistake in translation's service definition.
if handler_id is identical to null fix
CS fix
Fixed ModelChoiceList tests in Propel1 bridge.
[AclProvider] Fix incorrect behaviour when partial results returned from cache
Check if the pipe array is empty before calling stream_select()
re-factor Propel1 ModelChoiceList
[Locale] fixed the failing test described in #9455
[Process] fix phpdoc and timeout of 0
bug #9445 [BrowserKit] fixed protocol-relative url redirection
Conflicts:
src/Symfony/Component/BrowserKit/Tests/ClientTest.php
src/Symfony/Component/Locale/Tests/Stub/StubIntlDateFormatterTest.php
* 2.2:
[HttpKernel] added a check for private event listeners/subscribers
[FrameworkBundle] fixed registration of the register listener pass
[Form] Fixed regression causing invalid "WHERE id IN ()" statements
[DependencyInjection] fixed a non-detected circular reference in PhpDumper (closes#8425)
[Form] Fixed regression in BooleanToStringTransformer from ed83752
[FrameworkBundle] removed obsolete code
[Process] Close unix pipes before calling `proc_close` to avoid a deadlock
Conflicts:
src/Symfony/Bundle/FrameworkBundle/FrameworkBundle.php
src/Symfony/Component/HttpKernel/DependencyInjection/RegisterListenersPass.php
src/Symfony/Component/Process/Process.php
This circular reference cannot be detected by the compiler pass as we
don't check for method arguments there.
The Container itself already detects such circular references at runtime.
So this fix is about circular references that are not detected at
compile time, and are not even detected at runtime because the code that
would cause the detection is never run (generated after a return
statement.)
This PR was squashed before being merged into the master branch (closes#7890).
Discussion
----------
ProxyManager Bridge
As of @beberlei's suggestion, I re-implemented #7527 as a new bridge to avoid possible hidden dependencies.
Everything is like #7527 except that the new namespace (and possibly package/subtree split) `Symfony\Bridge\ProxyManager` is introduced
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #6140 (supersedes) #5012#6102 (maybe) #7527 (supersedes)
| License | MIT (attached code) - BSD-3-Clause (transitive dependency)
| Doc PR | Please pester me to death so I do it
This PR introduces lazy services along the lines of zendframework/zf2#4146
It introduces an **OPTIONAL** dependency to [ProxyManager](https://github.com/Ocramius/ProxyManager) and transitively to [`"zendframework/zend-code": "2.*"`](https://github.com/zendframework/zf2/tree/master/library/Zend/Code).
## Lazy services: why? A comprehensive example
For those who don't know what this is about, here's an example.
Assuming you have a service class like following:
```php
class MySuperSlowClass
{
public function __construct()
{
// inject large object graph or do heavy computation
sleep(10);
}
public function doFoo()
{
echo 'Foo!';
}
}
```
The DIC will hang for 10 seconds when calling:
```php
$container->get('my_super_slow_class');
```
With this PR, this can be avoided, and the following call will return a proxy immediately.
```php
$container->getDefinitions('my_super_slow_class')->setLazy(true);
$service = $container->get('my_super_slow_class');
```
The 10 seconds wait time will be delayed until the object is actually used:
```php
$service->doFoo(); // wait 10 seconds, then 'Foo!'
```
A more extensive description of the functionality can be found [here](https://github.com/Ocramius/ProxyManager/blob/master/docs/lazy-loading-value-holder.md).
## When do we need it?
Lazy services can be used to optimize the dependency graph in cases like:
* Webservice endpoints
* Db connections
* Objects that cause I/O in general
* Large dependency graphs that are not always used
This could also help in reducing excessive service location usage as I've explained [here](http://ocramius.github.com/blog/zf2-and-symfony-service-proxies-with-doctrine-proxies/).
## Implementation quirks of this PR
There's a couple of quirks in the implementation:
* `Symfony\Component\DependencyInjection\CompilerBuilder#createService` is now public because of the limitations of PHP 5.3
* `Symfony\Component\DependencyInjection\Dumper\PhpDumper` now with extra mess!
* The proxies are dumped at the end of compiled containers, therefore the container class is not PSR compliant anymore
Commits
-------
78e3710 ProxyManager Bridge
A service can now be marked as synchronized; when set, all method calls
involving this service will be called each time this service is set.
When in a scope, methods are also called to restore the previous version of the
service.
This PR was squashed before being merged into the master branch (closes#6846).
Commits
-------
9024c07 [2.3] [DI] Improve the generated PHPDoc of the dumped PHP container
Discussion
----------
[2.3] [DI] Improve the generated PHPDoc of the dumped PHP container
By adding "@throws InactiveScopeException" where applicable
---------------------------------------------------------------------------
by vicb at 2013-01-23T11:51:11Z
thanks @stof
> Native types should be lowercase
Never say that to @fabpot :)
---------------------------------------------------------------------------
by stof at 2013-01-23T12:10:56Z
@vicb he has an exception for ``Boolean`` (and explained why) but the reason cannot be applied to ``object``
---------------------------------------------------------------------------
by vicb at 2013-01-23T13:01:55Z
@stof all "Object"s should be covered now. Yeah I know he has such a mannerism !
---------------------------------------------------------------------------
by stof at 2013-01-23T19:01:07Z
@vicb none of the dumping tests modified ? this looks weird
---------------------------------------------------------------------------
by stof at 2013-01-23T19:06:06Z
hmm, travis is indeed agreeing with me...
---------------------------------------------------------------------------
by vicb at 2013-01-23T20:07:16Z
I'll fix that tomorrow
---------------------------------------------------------------------------
by vicb at 2013-01-24T08:37:55Z
travis I hope you like it !
* 2.1:
[Console] Fix style escaping parsing
[Console] Make style formatter matching less greedy to avoid having to escape when not needed
[Bundle] [FrameworkBundle] fixed indentation in esi.xml services file.
[Component] [Security] fixed PSR-2 coding violation in ClassUtilsTest class.
[Form] Fixed EntityChoiceList when loading objects with negative integer IDs
[TwigBundle] There is no CSS visibility of display, should be visible instead
[Form] corrected source node for a Danish translation
[DependencyInjection] fixed a bug where the strict flag on references were lost (closes#6607)
[HttpFoundation] Check if required shell functions for `FileBinaryMimeTypeGuesser` are not disabled
[CssSelector] added css selector with empty string
[HttpFoundation] Docblock for Request::isXmlHttpRequest() now points to Wikipedia
[DependencyInjection] refactored code to avoid logic duplication
[Form] Deleted references in FormBuilder::getFormConfig() to improve performance
[HttpFoundation] Update docblock for non-working method
Conflicts:
src/Symfony/Bundle/TwigBundle/Resources/views/Exception/trace.html.twig
src/Symfony/Bundle/TwigBundle/Resources/views/Exception/traces.html.twig
* 2.0:
[Bundle] [FrameworkBundle] fixed indentation in esi.xml services file.
[TwigBundle] There is no CSS visibility of display, should be visible instead
[DependencyInjection] fixed a bug where the strict flag on references were lost (closes#6607)
[HttpFoundation] Check if required shell functions for `FileBinaryMimeTypeGuesser` are not disabled
[CssSelector] added css selector with empty string
[HttpFoundation] Docblock for Request::isXmlHttpRequest() now points to Wikipedia
[DependencyInjection] refactored code to avoid logic duplication
Conflicts:
src/Symfony/Bundle/FrameworkBundle/Resources/config/esi.xml
src/Symfony/Component/DependencyInjection/Dumper/PhpDumper.php
src/Symfony/Component/HttpFoundation/File/MimeType/FileBinaryMimeTypeGuesser.php
* 2.1:
Restrict Monolog version to be in version <1.3
[Console] Make getTerminalWith & getTerminalHeight public
[DependencyInjection] fixed PhpDumper optimizations when an inlined service depends on the current one indirectly
[DependencyInjection] fixed PhpDumper when an inlined service definition has some properties
[DependencyInjection] added some tests for PhpDumper when the container is compiled
[DependencyInjection] fixed CS
[Process] Do not reset stdout/stderr pipes on Interrupted system call
[Locale] Adjust `StubIntlDateFormatter` to have new methods added in PHP 5.5
use the right RequestMatcherInterface
[Locale] Fix failing `StubIntlDateFormatter` tests in PHP 5.5
[Locale] Fix failing `StubIntlDateFormatter` in PHP 5.5
[Form] Fix failing `MonthChoiceList` in PHP 5.5
Update .travis.yml
Conflicts:
src/Symfony/Bridge/Monolog/composer.json
src/Symfony/Component/DependencyInjection/Tests/Fixtures/php/services9.php
* 2.0:
Restrict Monolog version to be in version <1.3
[Console] Make getTerminalWith & getTerminalHeight public
[DependencyInjection] fixed PhpDumper optimizations when an inlined service depends on the current one indirectly
[DependencyInjection] fixed PhpDumper when an inlined service definition has some properties
[DependencyInjection] added some tests for PhpDumper when the container is compiled
[DependencyInjection] fixed CS
[Locale] Adjust `StubIntlDateFormatter` to have new methods added in PHP 5.5
[Locale] Fix failing `StubIntlDateFormatter` tests in PHP 5.5
[Locale] Fix failing `StubIntlDateFormatter` in PHP 5.5
[Form] Fix failing `MonthChoiceList` in PHP 5.5
Update .travis.yml
Conflicts:
src/Symfony/Bridge/Monolog/composer.json
src/Symfony/Component/DependencyInjection/Tests/Fixtures/php/services9.php
src/Symfony/Component/Form/Extension/Core/ChoiceList/MonthChoiceList.php
tests/Symfony/Tests/Component/DependencyInjection/Fixtures/yaml/services9.yml
Commits
-------
e271b17 Remove the string optimization since it causes no real performance gain but increases generation time of the dumped PHP Container
Discussion
----------
PhpDumper and large strings
When the PhpDumper is dealing with longer strings, the regular expression performed to optimize this can be quite a performance hog.
In our case sometimes the dumper takes more then 30 seconds if leaving this enabled. Disabling it will bring it back to sub-second speed.
This patch makes the optimization optional by passing in an additional container option.
---------------------------------------------------------------------------
by fabpot at 2012-08-25T16:57:29Z
I don't like adding yet another option for something that should "just works". It would be better to find a better way to optimise the strings for all cases.
---------------------------------------------------------------------------
by mazen at 2012-08-25T17:22:07Z
I never really tested how much of a runtime difference it incurs when either using the "optimized" version or the non optimized version, so:
Having an example at hand which generates stable results yields (in a non-debug environment with a booted container using either of the optimization methods):
Without optimized strings:
```
Time taken for tests: 14.865 seconds
Complete requests: 1000
Failed requests: 0
Write errors: 0
Total transferred: 217000 bytes
HTML transferred: 8000 bytes
Requests per second: 67.27 [#/sec] (mean)
Time per request: 14.865 [ms] (mean)
Time per request: 14.865 [ms] (mean, across all concurrent requests)
Transfer rate: 14.26 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 0 0 0.0 0 0
Processing: 14 15 19.7 14 632
Waiting: 14 15 19.7 14 632
Total: 14 15 19.7 14 632
Percentage of the requests served within a certain time (ms)
50% 14
66% 14
75% 14
80% 14
90% 14
95% 14
98% 15
99% 23
100% 632 (longest request)
```
With Optimized Strings enabled
```
Time taken for tests: 14.077 seconds
Complete requests: 1000
Failed requests: 0
Write errors: 0
Total transferred: 217000 bytes
HTML transferred: 8000 bytes
Requests per second: 71.04 [#/sec] (mean)
Time per request: 14.077 [ms] (mean)
Time per request: 14.077 [ms] (mean, across all concurrent requests)
Transfer rate: 15.05 [Kbytes/sec] received
Connection Times (ms)
min mean[+/-sd] median max
Connect: 0 0 0.0 0 0
Processing: 14 14 1.3 14 48
Waiting: 14 14 1.3 14 48
Total: 14 14 1.3 14 48
Percentage of the requests served within a certain time (ms)
50% 14
66% 14
75% 14
80% 14
90% 14
95% 14
98% 14
99% 15
100% 48 (longest request)
```
So the response times differ by around 0.8ms
Building the non-optimized container takes around 800ms
Building the optimized container takes 43 seconds
From my Point of View it would just be viable to remove the optimization (since it already incurred some issues fixed in 808088a3ca).
I do not see a way how to improve the regexps (but by all means I'm no regular expression guru)
---------------------------------------------------------------------------
by fabpot at 2012-08-30T07:12:55Z
I'm also for removing these optimizations. What others think?
---------------------------------------------------------------------------
by Baachi at 2012-08-30T07:54:53Z
I'm +1 for removing this feature.
The performance boost is to small.