Newlines and whitespaces were not handled as before. This caused
issues for some users because our indent-recuction was broken.
The cause seems to be an upstream issue in phpstan parser which
is not resolved yet. But this work around post processing the tokens
helps us to make it work as before.
The phpstan parser is not consuming the full description when parsing
docblocks with a more complex description. For them it's mostlikely not an
issue as phpstan doesn't use the descriptions. But it will also parse
the descriptions into unexpected tags. This could be an advantage but is
not according to the phpdoc spec.
Our own tokenizer is already tokenizing the docblocks into the correct parts.
So all we needed to do is assume all remaining tokens in the phpstan ast belong
to the description. From there our own code is able to handle this as before in
v5.3.
fixes#365
Throw on invalid type definitions and unexpected type definitions.
Not all types resolved by phpstan's parser are valid for docblocks,
they might in a more complex type system but I do not see how these
types would ever apply to param tags.
... or rather add the missing `@coversNothing` as this is an integration test.
This prevents the test from being marked as "risky" due to the missing tag on PHPUnit 9.x:
```
There was 1 risky test:
1) phpDocumentor\Reflection\DocblockSeeTagResolvingTest::testResolvesSeeFQSENOfInlineTags
This test does not have a @covers annotation but is expected to have one
```
... to prevent it from being considered "risky" and being listed as "risky" below each test run.
```
There was 1 risky test:
1) phpDocumentor\Reflection\ModifyBackTraceSafeTest::testBackTraceModificationDoesNotImpactFunctionArguments
This test did not perform any assertions
/home/runner/work/ReflectionDocBlock/ReflectionDocBlock/tests/integration/ModifyBackTraceSafeTest.php:15
```
The See, Covers and Use tags can reference also methods, properties and
constants. Which means that the FqsenResolver cannot handle those properly.
This patch fixes that issue.
In #89 the regex to parse tags was to strict. The first character after a tag
had to be a single string. Annotations typically start with a parentheses. Only character not
allowed after a tag is a [.
Viktor Szépe
Jaapio
jrfnl
Michael Knappe
Chuck Burgess
TomasVotruba
Gabriel Caruso
Roberto Espinoza
Mike van Riel