This PR was merged into the 2.3 branch.
Discussion
----------
[HttpKernel] set s-maxage only if all responses are cacheable
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #18160
| License | MIT
| Doc PR |
Commits
-------
b7d9338 set s-maxage only if all responses are cacheable
This PR was merged into the 3.0 branch.
Discussion
----------
Fix leftover conflict marker in UPGRADE-3.0.md
| Q | A
| ------------- | ---
| Branch | 3.0
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? |
| Fixed tickets |
| License | MIT
| Doc PR |
Presumably after resolving a merge conflict, a parasite line `>>>>>>> 2.8` was introduced as part of commit 67df429d99:
* 67df429d99 (diff-0901ceb76939b17cd920bc69aa52d21fR406)
* 67df429d99/UPGRADE-3.0.md (L406)
The present PR just deletes this line 406.
Commits
-------
b47220a Fix leftover conflict marker in UPGRADE-3.0.md
This PR was merged into the 2.3 branch.
Discussion
----------
Improved the "branch" row of the PR table
| Q | A
| ------------- | ---
| Branch | 2.3+
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #18128
| License | MIT
| Doc PR | -
Commits
-------
6af23cc Improved the "branch" row of the PR table
This PR was merged into the 2.7 branch.
Discussion
----------
[Process] Wait a bit less on Windows
| Q | A
| ------------- | ---
| Branch | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | -
| License | MIT
| Doc PR | -
By using `stream_select` instead of `usleep` when the input is at wait, and by not blocking when the process has halted, we can enhance a bit the performance on Windows.
Commits
-------
380a54f [Process] Wait a bit less on Windows
This PR was squashed before being merged into the 2.7 branch (closes#18130).
Discussion
----------
[Debug] Replaced logic for detecting filesystem case sensitivity
| Q | A
| ------------- | ---
| Branch | 2.7
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | none
| License | MIT
| Doc PR | none
When I cloned the master branch onto a Virtualbox Vagrant OSX El Capitan host, Ubuntu Wily guest, the `Symfony\Component\Debug\Tests\DebugClassLoaderTest::testFileCaseMismatch` failed because 'Failed asserting that exception of type "\RuntimeException" is thrown'.
@WouterJ confirmed he got the same problem, and it's because Virtualbox shared folders aren't case sensitive, even when the guest is using a case sensitive filesystem. So I've replaced the logic that looked at the name of the operating system.
I ran the tests in the following environments:
* Virtualbox/Vagrant - OSX Host, Ubuntu guest
* Virtualbox/Vagrant - OSX Host, Windows guest
* OSX native
* Ubuntu native
NB - I _didn't_ run it on native Windows (because I don't have easy access to one).
Commits
-------
2e81b0a [Debug] Replaced logic for detecting filesystem case sensitivity
This PR was merged into the 2.8 branch.
Discussion
----------
Autowiring the concrete class too - consistent with behavior of other services
| Q | A
| ------------- | ---
| Branch | 2.8
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #18132
| License | MIT
| Doc PR | n/a
This follows #17261. Without this, if you use the concrete class `Container`, it doesn't autowire, so it creates a *new* one. That is certainly not what the end-user wants, and it's a serious WTF :). We can talk all day long about not injecting the container and type-hinting interfaces, but this is needed to be consistent with how all the other services in the container work and to avoid this odd behavior.
Thanks!
Commits
-------
2ea3f68 Autowiring the concrete class too - consistent with behavior of other services
This PR was merged into the 2.3 branch.
Discussion
----------
[Hack Day] Issue for newcomers (2). Replace perfom by perform
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | #18089
| License | MIT
| Doc PR | reference to the documentation PR, if any
Commits
-------
f33f192 replace perfom by perform
This PR was merged into the 2.3 branch.
Discussion
----------
Rename Choosen to Chosen in ModelChoiceListTest
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | yes
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | #18091
| License | MIT
| Doc PR | n/a
Commits
-------
5e4b515 Fix for Isssue #18091
This PR was merged into the 2.3 branch.
Discussion
----------
Fix typos #18090
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | #18090
| License | MIT
| Doc PR |
1. PHPs session design to PHP's session design
2. Symfony HttpKernel offers to Symfony's HttpKernel offers
3. in which case it it should to in which case it should
Commits
-------
6276452 Fix typos #18090 1. PHPs session design to PHP's session design 2. Symfony HttpKernel offers to Symfony's HttpKernel offers 3. in which case it it should to in which case it should
1. PHPs session design to PHP's session design
2. Symfony HttpKernel offers to Symfony's HttpKernel offers
3. in which case it it should to in which case it should
As of Symfony 2.8, constraint errors are now string UUIDs rather than integers. The corresponding docblock typehint in getErrorName() should reflect this change.
This PR was merged into the 2.3 branch.
Discussion
----------
[ci] remove token for composer now that rate limiting is off
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | -
| License | MIT
| Doc PR | -
As of https://github.com/composer/composer/issues/4884#issuecomment-195229989
Commits
-------
218ff14 [ci] remove token for composer now that rate limiting is off
* 3.0:
[ci] use hirak/prestissimo
[Filesystem] Fix transient tests
[WebProfiler] Sidebar button padding
Updated some missing READMEs
[HttpFoundation] Avoid warnings when checking malicious IPs
[HttpFoundation] Set the Content-Range header if the requested Range is unsatisfied
* 2.8:
[ci] use hirak/prestissimo
[Filesystem] Fix transient tests
[WebProfiler] Sidebar button padding
Updated some missing READMEs
[HttpFoundation] Avoid warnings when checking malicious IPs
[HttpFoundation] Set the Content-Range header if the requested Range is unsatisfied
Conflicts:
appveyor.yml
src/Symfony/Component/Intl/Tests/DateFormatter/AbstractIntlDateFormatterTest.php
This PR was submitted for the 2.3 branch but it was merged into the 3.1-dev branch instead (closes#17255).
Discussion
----------
[Console][2.3] ApplicationTester - test stdout and stderr
| Q | A
| ------------- | ---
| Bug fix? | no
| New feature? | yes
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| License | MIT
Currently it is not possible to test application output of both `stdout` _and_ `stderr` using the `ApplicationTester`. This makes it hard to check if an application writes to the correct output.
Commits
-------
6ff6a28 [Console][2.3] ApplicationTester - test stdout and stderr
* 2.7:
[ci] use hirak/prestissimo
[Filesystem] Fix transient tests
[HttpFoundation] Avoid warnings when checking malicious IPs
[HttpFoundation] Set the Content-Range header if the requested Range is unsatisfied
Conflicts:
src/Symfony/Component/Intl/Tests/DateFormatter/AbstractIntlDateFormatterTest.php
* 2.3:
[ci] use hirak/prestissimo
[Filesystem] Fix transient tests
[HttpFoundation] Avoid warnings when checking malicious IPs
[HttpFoundation] Set the Content-Range header if the requested Range is unsatisfied
Conflicts:
appveyor.yml
src/Symfony/Component/Filesystem/Tests/FilesystemTest.php
This PR was merged into the 2.3 branch.
Discussion
----------
[ci] use hirak/prestissimo
| Q | A
| ------------- | ---
| Branch | 2.3
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | -
| License | MIT
| Doc PR | -
Commits
-------
adf1ca8 [ci] use hirak/prestissimo
This PR was merged into the 3.1-dev branch.
Discussion
----------
[Cache] Hash using B64+MD5 in FilesystemAdapter
| Q | A
| ------------- | ---
| Branch | master
| Bug fix? | no
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | no
| Fixed tickets | -
| License | MIT
| Doc PR | -
Using MD5+B64 is good enough to avoid any hash collision, even on case-insensitive filesystems.
On Windows where the total path length is limited, this saves 44 chars worth.
Even if collisions are extremely unlikely, they are detected by adding then comparing the raw key to saved payloads. This also has the added benefit of easing debugging/grepping the cached items on the filesystem.
Commits
-------
e96bb10 [Cache] Hash using B64+MD5 in FilesystemAdapter
This PR was merged into the 3.1-dev branch.
Discussion
----------
Fixing autowiring collision failure
| Q | A
| ------------- | ---
| Branch | master
| Bug fix? | yes (bug introduced in master)
| New feature? | no
| BC breaks? | no
| Deprecations? | no
| Tests pass? | yes
| Fixed tickets | none
| License | MIT
| Doc PR | n/a
In #17877, I introduced a bug: https://github.com/symfony/symfony/pull/17877/files#diff-62df969ae028c559d33ffd256de1ac49L200.
Namely, if some class cannot be autowired because there is an *odd* number of matching services, then it *would* autowire the last service found, instead of throwing an exception. The tests only tested even numbers, which is how it was missed. This fixes that.
Thanks!
Commits
-------
2aea337 Fixing a bug where an odd number of type collisions would incorrectly autowire (instead of an error)