Count syntax errors caused by a mutatan as a failure of a mutant

[sanmai]

Fixes #1569, related to #1555

1. Suppose Infections introduces a change to make a method return string instead of a float. Engineer adds strict return type to the method, killing the mutant: your ordinary defensive programming. We count this mutant as killed, giving a bump to the MSI.
2. Suppose Infections introduces a different change, making a method fail with a syntax error. Currently we count this error as something else, as escaped mutant in terms of MSI. How can that be right?

Let's consider what are the options we're giving. How can an engineer fix this problem? Their tests are all right, it's just Infection broke their program to produce an unrecoverable error. Why should they be penalized with a lowered MSI?

It appears to me this is very unjust. If Infection causes a type error, or any error in principle, that's not what you can fix writing a better test. That's not something that leads to better tests. That's not something that makes people change their tests for the better. I hope we can agree on this.

[maks-rafalko]

Let's target 0.25 so we can faster release it, as 0.25.1

[sanmai]

0.25 is three commits behind master, I think you should be able to forward it to the master and release:

$git checkout 0.25
$ git merge master
Updating 77b684ad..854bc7f4
Fast-forward
 .github/workflows/mt-annotations.yaml                                       |  2 +-
 src/Metrics/Calculator.php                                                  |  2 +-
 tests/phpunit/Logger/JsonLoggerTest.php                                     |  4 ++--
 tests/phpunit/Logger/PerMutatorLoggerTest.php                               |  4 ++--
 tests/phpunit/Metrics/CalculatorTest.php                                    |  4 ++--
 tests/phpunit/TestFramework/Coverage/JUnit/TestLocationBucketSorterTest.php | 25 ++++++++++++++++++++++++-
 6 files changed, 32 insertions(+), 9 deletions(-)

[maks-rafalko]

Released https://github.com/infection/infection/releases/tag/0.25.1