Generate WebP in liip:imagine:cache:resolve CLI command and async resolve cache messages

[peter-gribanov]

Q	A
Branch?	2.0
Bug fix?	yes
New feature?	yes
BC breaks?	no
Deprecations?	no
Tests pass?	yes
Fixed tickets	#1319
License	MIT
Doc PR

The liip:imagine:cache:resolve CLI command warms up the cache for the specified image sources with all or specified filters applied, and prints the list of cache files. I forgot that this command should also create a image in WebP format if WebP generation is enabled.

• Add new mthod FilterService::warmsUpCache();
• The mthod FilterService::bustCache() now return result of busting;
• The liip:imagine:cache:resolve command now warms up image in WebP format too;
• The liip:imagine:cache:remove command now remove image in WebP format too;
• The ResolveCacheProcessor wat async warms up cache now warmed up image in WebP format too;
• Add unit tests for FilterPathContainer;
• Add unit tests for FilterService.

[coveralls]

Coverage increased (+2.3%) to 86.295% when pulling 4c3e68f on peter-gribanov:cache_resolve_command into 7282ab2 on liip:2.x.

[peter-gribanov]

Build failed because for PHP 8 used PHPUnit 9 what is not compatible with current tests. See fix #1381

[peter-gribanov]

@lsmith77 it would be nice to merge this PR. It fixes some bugs not implemented in #1307, adds unit tests and adds a cache warming function. If this PR is too large, then i can split it into several PRs.

[dbu]

sorry for the late reply. this looks good to me, and a nice refactoring to move the warmup into the filter service.

i have some nitpicks, can you please have a look?

[dbu]

if #1360 is merged first, we need to adjust the logic in the symfony messenger processor to follow the changes done here.

[peter-gribanov]

@dbu renamed the method to warmUpCache.

[peter-gribanov]

The order of the merger is not critical. I can make changes to my PR after #1360 merge.

[dbu]

thanks a lot! i really like that we move from talking about "resolving" to "warmup" 👍

#1360 won the race ;-) can you please rebase this branch on 2.x and apply your refactoring to WarmupCacheHandler?

i am worried a bit by how much mocking we are doing in one of the unit tests. but if you are happy with it, we can leave it like that, just want to ask the question.

[dbu]

there is a LOT of mocking going on in this test. i wonder if we could not move towards a more functional test that uses the real managers instead of mocking. if tests fail, it will be a bit harder to know where the mistake was - but if they succeed, they create more trust that things actually work as expected.

[peter-gribanov]

I agree that it would be great to supplement the unit tests with functional tests, but unfortunately i don't have time to do this 😕

[peter-gribanov]

#1360 won the race ;-) can you please rebase this branch on 2.x and apply your refactoring to WarmupCacheHandler?

done

[dbu]

thanks a lot!

i will look into creating a new minor release next week.