Add support to more output formats

[llaville]

Hello,

As author of the SARIF PHP binding solution https://github.com/llaville/sarif-php-sdk I would like to propose a SARIF support.

For audience that don't know yet what is SARIF (Static Analysis Results Interchange Format), I suggest to read :

• https://docs.github.com/en/code-security/code-scanning/integrating-with-code-scanning/sarif-support-for-code-scanning

But my proposal is not limited to SARIF. With it we could add support to other formats not yet supported.

That will include, for example :

• checkstyle, see old proposal Add checkstyle printer #139
• GitHub Action, see request Support github action output format #148

And whatever else we will need !

Now that introduction is opened, here is in details my proposal !

A Povils\PHPMND\Printer\Custom class that implements the standard Povils\PHPMND\Printer\Printer contract
and inspired by https://github.com/phpmd/phpmd/blob/2.15.0/src/main/php/PHPMD/TextUI/CommandLineOptions.php#L780

CLI options will be only :

  -o, --output=OUTPUT                  Generate an output to the specified path [default: "php://stdout"]
      --format=FORMAT                  Format of the report [default: "console"]

And remove old --xml-ouptut option

Here are some use cases :

Console (default output)

No change because this is the default. (Option to write/redirect results to a file with --output option)

SARIF format

bin/phpmnd /path/to/source --format SarifPrinter

source code

<?php

require_once '/path/to/autoload.php';

class SarifPrinter extends \Bartlett\Sarif\Converter\PhpMndConverter
{
}

Note

no implemention here because I propose to add support to my project as I did since release 1.2.0 for at least 3 other PHP linters : PHPCS, PHPLint, PHPStan

Checkstyle format

bin/phpmnd /path/to/source --format CheckstylePrinter

Now a version of CheckStyle implementation we can have :

source code

<?php

use Povils\PHPMND\DetectionResult;
use Povils\PHPMND\HintList;
use Symfony\Component\Console\Output\OutputInterface;
use Symfony\Component\Console\Output\StreamOutput;

class CheckstylePrinter implements \Povils\PHPMND\Printer\Printer
{
    public function printData(OutputInterface $output, HintList $hintList, array $detections): void
    {
        $groupedList = $this->groupDetectionResultPerFile($detections);

        $document = new DOMDocument('1.0', 'UTF-8');
        $document->formatOutput = true;

        $rootElement = $document->createElement('checkstyle');
        $document->appendChild($rootElement);

        foreach ($groupedList as $path => $detectionResults) {
            $fileNode = $document->createElement('file');
            $fileNode->setAttribute('name', $path);

            $rootElement->appendChild($fileNode);

            foreach ($detectionResults as $detectionResult) {
                $snippet = $this->getSnippet(
                    $detectionResult->getFile()->getContents(),
                    $detectionResult->getLine(),
                    $detectionResult->getValue()
                );

                $errorNode = $document->createElement('error');
                $errorNode->setAttribute('line', (string) $detectionResult->getLine());
                $errorNode->setAttribute('column', (string) $snippet['col']);
                $errorNode->setAttribute('severity', 'error');
                $errorNode->setAttribute(
                    'message',
                    sprintf('Magic Number: %s', $detectionResult->getValue())
                );

                $fileNode->appendChild($errorNode);
            }

            $rootElement->appendChild($fileNode);
        }

        $xmlString = $document->saveXML();
        if (is_string($xmlString)) {
            $output->write($xmlString);
        }
    }

    /**
     * @param array<int, DetectionResult> $list
     *
     * @return array<int, DetectionResult[]>
     */
    private function groupDetectionResultPerFile(array $list): array
    {
        $result = [];

        foreach ($list as $detectionResult) {
            $result[$detectionResult->getFile()->getRelativePathname()][] = $detectionResult;
        }

        return $result;
    }

    /**
     * Get the snippet and information about it
     *
     * @param int|string $text
     *
     * @return array<string, string>
     */
    private function getSnippet(string $content, int $line, $text): array
    {
        $content = str_replace(["\r\n", "\r"], "\n", $content);
        $lines = explode("\n", $content);

        $lineContent = array_slice($lines, $line - 1, 1);
        $lineContent = reset($lineContent);
        $start = strpos($lineContent, $text . '');

        return [
            'snippet' => $lineContent,
            'line' => $line,
            'magic' => $text,
            'col' => $start,
        ];
    }
}

I'll propose a POC PR (not optimized, because we can do better, but essential are available) and link to this feature request just after.

[llaville]

My SARIF PHP SDK has an upcoming version 1.4.0 that will add PHPMD (PHP Mess Detector) to support SARIF.

Even if PHPMD has already a native support, my version gave more informations, and is not maintained by authors of PHPMD.

This is the goal of my proposal here. No need to maintained a SARIF output format, because I'll do if you accept this POC !

[llaville]

I've just published https://github.com/llaville/sarif-php-sdk/releases/tag/1.4.0 that support PHPMD !

Do you want to be the next one on list ?

[llaville]

Any feedback will be greatly appreciated ?