This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[DependencyInjection] Fix circular in DI with lazy + byContruct loop
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39120
| License | MIT
| Doc PR | -
This fix another issue lazy service.
It partially revert #38980 and #39021
Initially, we trusted lazy services to be lazy and not beeing called while building the services graph
=> bug #38970 when lazy deps is injected in a factory, it may be consumed directly to build the object before the graph is fully built
Fixed by #38980 => lazy service are considered as "normal service"
=> bug #39015 some loop are not resolvable with "normal service", but it shouldn't be an issue when servie proxifyied
Fixed by #39021 => lazy service are considered as "normal service" except when proxyfied
=> bug #39120 some loop are not resolvable with "normal service", but it shouldn't be an issue because the lazy service is injected in the constructor and user
Fixed by this PR => that revert to the initial state. lazy service are trusted.
But now, The IterratorArgument injected in a factory (single exception) is not more considered as lazy
Commits
-------
54af139a4e [DependencyInjection] Fix circular in DI with lazy + byContruct loop
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
Add Romanian missing translations
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | kinda
| New feature? | no
| Deprecations? | no
| Tickets | Fix#38758
| License | MIT
| Doc PR | n/a
Commits
-------
6aa31a13c7 Add Romanian missing translations
This PR was squashed before being merged into the 4.4 branch.
Discussion
----------
[DependencyInjection][Translator] Silent deprecation triggered by libxml_disable_entity_loader
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39040
| License | MIT
| Doc PR | -
The XML entity loader is disabled by default since libxml 2.9
But, since PHP 8.0, calling the method `libxml_disable_entity_loader` triggers a deprecation which has been solved in symfony by calling `libxml_disable_entity_loader` only for libxml < 2.9
The issue is, some dependencies, enable the entity loader and does not restore the initial state afterward, leading to exceptions triggered by Symfony iteself.
In previous versions symfony was resilient by disabling the flag before working, which is not the case anymore to avoid the deprecation.
This PR restore the resiliency of Symfony for PHP < 8.0, which is not yet deprecated.
But we have no way to check the status of the entity loader without triggering a deprecation with Symfony 8.
Commits
-------
114b7a543a [DependencyInjection][Translator] Silent deprecation triggered by libxml_disable_entity_loader
This PR was merged into the 4.4 branch.
Discussion
----------
[Form] ignore the pattern attribute for textareas
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39066
| License | MIT
| Doc PR |
Commits
-------
e7698e7434 ignore the pattern attribute for textareas
This PR was merged into the 5.1 branch.
Discussion
----------
[Serializer] Fix denormalizing scalar with UnwrappingDenormalizer
| Q | A
| ------------- | ---
| Branch? | 5.1
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#38983
| License | MIT
| Doc PR | none
In order to work the `UnwrappingDenormlizer` needs to be called first to unwrap the data so that it can be handled by other denormalizers.
To not introduce any BC break this PR does not move the logic into it's own class, instead it checks if a denormalizer supports the data and only denormalize scalar values if there is none.
Commits
-------
98cf389fb9 fix denormalizing scalar with UnwrappingDenormalizer
* 4.4:
[PhpUnitBridge] Fix qualification of deprecations triggered by the debug class loader
Improve return phpdoc for Normalizer
Use a partial buffer in SymfonyStyle
Fix console closing tag
Fix typo in comment
[VarDumper] fix casting resources turned into objects on PHP 8
This PR was merged into the 4.4 branch.
Discussion
----------
[PhpUnitBridge] Fix qualification of deprecations triggered by the debug class loader
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | https://github.com/symfony/symfony/issues/36036
| License | MIT
| Doc PR | -
I tested it successfully against the two examples stof gave in the issue. However, I don't see how to write a working real test for this.
The solution I found is to add the missing checked file in the original files stack so when the type of the deprecation is computed instead of having "self" -> "self", we have "vendor" -> "self".
Commits
-------
dff539434e [PhpUnitBridge] Fix qualification of deprecations triggered by the debug class loader
This PR was merged into the 4.4 branch.
Discussion
----------
[Console] Use a partial buffer in SymfonyStyle
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix#39156
| License | MIT
| Doc PR | -
Symfony style needs to buffer output in order to get the last 2 chars in order to prepend Block.
By using a `BufferedOutput` symfony bufferize everything while it not needed.
This PR adds a new `TrimmedBufferOutput` that keep only the N last chars.
Commits
-------
18fca2984d Use a partial buffer in SymfonyStyle
This PR was merged into the 4.4 branch.
Discussion
----------
Improve return value for phpdoc of Normalizer
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | Phpdoc
| New feature? | no
| Deprecations? | no
| License | MIT
The inheritdoc return type is too large.
The return type type can be more precise for lot of Normalizer.
Commits
-------
f2713d6580 Improve return phpdoc for Normalizer
This PR was merged into the 4.4 branch.
Discussion
----------
[Console] Fix console closing tag
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
When using SymfonyStyle, in some cases, closing a tag, is called twice.
In the following code `<question>do you want <comment>something</>?</>` the first `</>` close both the `comment` and the `question` tags.
![Screenshot from 2020-11-25 01-21-06](https://user-images.githubusercontent.com/578547/100166475-191d9d80-2ebd-11eb-991a-6541210c479b.png)
The reason is, part of the content is sent in 2 Outputs (see #39160 for another issue), and both outputs share the same `$styleStack`.
This PR updates the `OutputFormatter::__clone` method to prevent sharing the same state.
Commits
-------
2834c279d7 Fix console closing tag
This PR was submitted for the 5.x branch but it was merged into the 4.4 branch instead.
Discussion
----------
Fix typo in comment
| Q | A
| ------------- | ---
| Branch? | 5.x <!-- see below -->
| Bug fix? | no
| New feature? | no <!-- please update src/**/CHANGELOG.md files -->
| Deprecations? | no <!-- please update UPGRADE-*.md and src/**/CHANGELOG.md files -->
| Tickets | - <!-- prefix each issue number with "Fix #", no need to create an issue if none exist, explain below instead -->
| License | MIT
| Doc PR | - <!-- required for new features -->
<!--
Replace this notice by a short README for your feature/bugfix. This will help people
understand your PR and can be used as a start for the documentation.
Additionally (see https://symfony.com/releases):
- Always add tests and ensure they pass.
- Never break backward compatibility (see https://symfony.com/bc).
- Bug fixes must be submitted against the lowest maintained branch where they apply
(lowest branches are regularly merged to upper ones so they get the fixes too.)
- Features and deprecations must be submitted against branch 5.x.
-->
fixed spelling below.
* possibe -> possible
Commits
-------
ca93ae5b1a Fix typo in comment
This PR was merged into the 5.1 branch.
Discussion
----------
[Cache] Fix CI because of Couchbase version
| Q | A
| ------------- | ---
| Branch? | 5.1
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | -
| License | MIT
| Doc PR | -
It looks likes the version 3.0 of the couchbase extension does not exposes the `\CouchbaseCluster` class anymore.
I didn't find documentaiton about the BC break, but their documentation changes
version 2.6 => https://docs.couchbase.com/php-sdk/2.6/managing-connections.html
version 3.0 => https://docs.couchbase.com/php-sdk/current/howtos/managing-connections.html
It wasn't reported before, because `shivammathur/setup-php` added the extension 3days ago https://github.com/shivammathur/setup-php/pull/337
Sounds like the adapter were never tested, because the extension where missing and phpunit skipped the tests.
Commits
-------
fcbf0bf76e Display debug info
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpClient] don't fallback to HTTP/1.1 when HTTP/2 streams break
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | yes
| New feature? | no
| Deprecations? | no
| Tickets | Fix https://github.com/composer/composer/issues/9481
| License | MIT
| Doc PR | -
With this change, I don't reproduce the failures that I describe in https://github.com/composer/composer/issues/9481 when running the script in #38690
Apparently curl has an issue when both h1.1 and h2 connections are open to the same host.
Instead of switching to HTTP/1.1 when retrying requests that failed because of an HTTP/2 stream error, I propose to close the http/2 connection when issuing a retry.
With this change, running the mirroring script of packagist works like a charm.
No need to investigate your mirrors @Seldaek, this was definitely a data corruption issue.
Commits
-------
0c92bc5a83 [HttpClient] don't fallback to HTTP/1.1 when HTTP/2 streams break
* 4.4:
do not depend on the actual time to fix a transient test
Run Redis Sentinel tests in GithubAction
Minor : Removed typo (extra "the" term)
Check if method inheritEnvironmentVariables exists
[PhpUnitBridge] Fix test fixture file name
This PR was merged into the 4.4 branch.
Discussion
----------
[HttpKernel] do not depend on the actual time to fix a transient test
| Q | A
| ------------- | ---
| Branch? | 4.4
| Bug fix? | no
| New feature? | no
| Deprecations? | no
| Tickets |
| License | MIT
| Doc PR |
Commits
-------
b7e5c7d67d do not depend on the actual time to fix a transient test