Hello there,

I've noticed that StandaloneLinePromotedPropertyFixer only applies if there are at least two properties in the constructor. I'm unsure if it's a conscious choice or a bug.

Should it be fixed ? Should it be an option to configure the behavior ?
I'm curious to hear your thoughts.

gnutix

I get this error when using the ChainMethodCallAnalyzer.

[Error]
Undefined constant "Symplify\CodingStandard\TokenAnalyzer\T_OPEN_CURLY_BRACKET"

Symplify\CodingStandard\TokenAnalyzer\NewlineAnalyzer->doesContentBeforeBracketRequireNewline()
in vendor/symplify/coding-standard/src/TokenAnalyzer/ChainMethodCallAnalyzer.php at line 36
Symplify\CodingStandard\TokenAnalyzer\ChainMethodCallAnalyzer->isPreceededByFuncCall()
in vendor/symplify/coding-standard/src/Fixer/Spacing/MethodChainingNewlineFixer.php at line 164
Symplify\CodingStandard\Fixer\Spacing\MethodChainingNewlineFixer->shouldBracketPrefix()
in vendor/symplify/coding-standard/src/Fixer/Spacing/MethodChainingNewlineFixer.php at line 108
Symplify\CodingStandard\Fixer\Spacing\MethodChainingNewlineFixer->shouldPrefixNewline()
in vendor/symplify/coding-standard/src/Fixer/Spacing/MethodChainingNewlineFixer.php at line 73
Symplify\CodingStandard\Fixer\Spacing\MethodChainingNewlineFixer->fix()
in .php-cs-fixer.php at line 73
NameWrapper->fix()
in vendor/friendsofphp/php-cs-fixer/src/Runner/Runner.php at line 173
PhpCsFixer\Runner\Runner->fixFile()
in vendor/friendsofphp/php-cs-fixer/src/Runner/Runner.php at line 114
PhpCsFixer\Runner\Runner->fix()
in vendor/friendsofphp/php-cs-fixer/src/Console/Command/FixCommand.php at line 300
PhpCsFixer\Console\Command\FixCommand->execute()
in vendor/symfony/console/Command/Command.php at line 326

Not sure what causes this and what the best approach to solve this is.

Hello, I'm not sure where to post this issue.

This repository has a dependency on symplify/rule-doc-generator-contracts, but that repository is a readonly subset of the archived Symplify monorepo.

There is a pending change on main branch allowing nette/utils 4.0.

Could you tag that repository to enable nette/utils upgrade on symplify/coding-standard ?

Thanks for your work!

Suppose you have code like this:

<?php

/**
 * @param string $one
 * @param string $one
 */
function someFunction($one, $two): void
{
}

The author's intention is to write a param for $one and $two, but it mistakenly duplicates $one.

I use SetList::CLEAN_CODE from ECS, but it seems to be replaced as follows:

 <?php
 
 /**
- * @param string $one
- * @param string $one
+ * @param string $two
+ * @param string $two
  */
 function someFunction($one, $two): void
 {
 }

ParamNameTypoMalformWorker::fixTypos() certainly seems to have that problem.

We don't want a quick fix as we can fix these manually, but will report them as issues. Thank you for your work!

<?php

declare(strict_types=1);

class TestClass
{

	/**
	 * This is just standard doc comment for method which declaration's length is actually <  LineLengthFixer max length
	 * but it computes this php doc to the method declaration itself
	 */
	private function testMethod($arg)
	{

	}
}
?>

when LineLengthFixer is called on this file, it breaks the line and outputs this

I don't know how to fix it, however

Running with php 7.4 and

 "require-dev": {
        "symplify/easy-coding-standard": "12.*"
}

I get

This doesnt happen until v11.4.*

Use return type when possible

/**
 * @return int
 */
public function getValue()
{
}

=>

public function getValue() : int
{
}

Use args type when possible

Same for args:

/**
 * @param int $amount
 */
public function count($amount)
{
}

=>

public function count(int $amount)
{
}

Unite return type format

// Class
public function getValue() : int

// Interface
public function getValue() : int;

Use strict_value when needed

When somewhere is some arg/return type, make this required.

<?php 

declare (strict_types = 1);

No

$array = [1 => 'one', 2 => 'two', 3 => 'three', 4 => 'four'];

Yes!

$array = [
    1 => 'one',
    2 => 'two',
    3 => 'three',
    4 => 'four'
];

After updating symplify/coding-standard version CS stopped working.

MethodCommentReturnTagSniff was renamed to MethodReturnTypeSniff.
When using custom ruleset, CS starts throwing error. Within one major version sniffs shouldn't be renamed as it causes bc break.

No utf8 support for LineLengthFixer.
The length of the strings is not correctly calculated, because of this, an extra line break is set, where it is not necessary.

The solution would be \Symplify\CodingStandard\TokenRunner\Transformer\FixerTransformer\FirstLineLengthResolver class to use the mb_strlen function instead of strlen, or be able to configure this behavior

Example code:

<?php

declare(strict_types=1);

class A
{
    public function art(): mixed
    {
        $a = $aq->withRow($q)->name;

        $b = $aq->withRow($q)->getName($s);

        $c = $aq->withRow($q)->withOtherRow($s)->name;

        return $a + $b - $c;
    }
}

Expected result after running MethodChainingNewLineFixer:

<?php

declare(strict_types=1);

class A
{
    public function art(): mixed
    {
        $a = $aq->withRow($q)
            ->name;

        $b = $aq->withRow($q)
            ->getName($s);

        $c = $aq->withRow($q)
            ->withOtherRow($s)
            ->name;

        return $a + $b - $c;
    }
}

Reality:

<?php

declare(strict_types=1);

class A
{
    public function art(): mixed
    {
        $a = $aq->withRow($q)
->name;

        $b = $aq->withRow($q)
            ->getName($s);

        $c = $aq->withRow($q)
            ->withOtherRow($s)
->name;

        return $a + $b - $c;
    }
}

This may not be immediately actionable, however you may have this as a goal anyway.

Can we get support for the new PHP Evolving Recommendations Coding Standard https://www.php-fig.org/per/coding-style/? Version 1.0 is a copy of PSR12 and is implemented in PHPCS-Fixer. A 1.1 version was expected, however this was recently released as 2.0 and a PR is open to upgrade this in PHPCS-Fixer.

I haven't looked too far into PHPCS, however there is an open issue regarding PERCS 2.0.

see: squizlabs/PHP_CodeSniffer#3932