[Finder] Ignore paths from .gitignore #26714

src/Symfony/Component/Finder/CHANGELOG.md

@@ -1,6 +1,11 @@
 CHANGELOG
 =========
 
+4.3.0
+-----
+
+ * added Finder::ignoreVCSIgnored() to ignore files based on rules listed in .gitignore
+
 4.2.0
 -----
 

src/Symfony/Component/Finder/Finder.php

@@ -39,6 +39,7 @@ class Finder implements \IteratorAggregate, \Countable
 {
     const IGNORE_VCS_FILES = 1;
     const IGNORE_DOT_FILES = 2;
+    const IGNORE_VCS_IGNORED_FILES = 4;
 
     private $mode = 0;
     private $names = [];
@@ -373,6 +374,24 @@ class Finder implements \IteratorAggregate, \Countable
         return $this;
     }
 
+    /**
+     * Forces Finder to obey .gitignore and ignore files based on rules listed there.
+     *
+     * This option is disabled by default.
+     *
+     * @return $this
+     */
+    public function ignoreVCSIgnored(bool $ignoreVCSIgnored)
+    {
+        if ($ignoreVCSIgnored) {
+            $this->ignore |= static::IGNORE_VCS_IGNORED_FILES;
+        } else {
+            $this->ignore &= ~static::IGNORE_VCS_IGNORED_FILES;
+        }
+
+        return $this;
+    }
+
     /**
      * Adds VCS patterns.
      *
@@ -685,6 +704,14 @@ class Finder implements \IteratorAggregate, \Countable
             $notPaths[] = '#(^|/)\..+(/|$)#';
         }
 
+        if (static::IGNORE_VCS_IGNORED_FILES === (static::IGNORE_VCS_IGNORED_FILES & $this->ignore)) {
+            $gitignoreFilePath = sprintf('%s/.gitignore', $dir);
+            if (!is_readable($gitignoreFilePath)) {
+                throw new \RuntimeException(sprintf('The "ignoreVCSIgnored" option cannot be used by the Finder as the "%s" file is not readable.', $gitignoreFilePath));
+            }
+            $notPaths = array_merge($notPaths, [Gitignore::toRegex(file_get_contents($gitignoreFilePath))]);
+        }
+
         $minDepth = 0;
         $maxDepth = PHP_INT_MAX;
 

src/Symfony/Component/Finder/Gitignore.php

@@ -0,0 +1,107 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Finder;
+
+/**
+ * Gitignore matches against text.
+ *
+ * @author Ahmed Abdou <mail@ahmd.io>
+ */
+class Gitignore
+{
+    /**
+     * Returns a regexp which is the equivalent of the gitignore pattern.
+     *
+     * @param string $gitignoreFileContent
+     *
+     * @return string The regexp
+     */
+    public static function toRegex(string $gitignoreFileContent): string
+    {
+        $gitignoreFileContent = preg_replace('/^[^\\\\]*#.*/', '', $gitignoreFileContent);
+        $gitignoreLines = preg_split('/\r\n|\r|\n/', $gitignoreFileContent);
+        $gitignoreLines = array_map('trim', $gitignoreLines);
+        $gitignoreLines = array_filter($gitignoreLines);
+
+        $ignoreLinesPositive = array_filter($gitignoreLines, function (string $line) {
+            return !preg_match('/^!/', $line);
+        });
+
+        $ignoreLinesNegative = array_filter($gitignoreLines, function (string $line) {
+            return preg_match('/^!/', $line);
+        });
+
+        $ignoreLinesNegative = array_map(function (string $line) {
+            return preg_replace('/^!(.*)/', '${1}', $line);
+        }, $ignoreLinesNegative);
+        $ignoreLinesNegative = array_map([__CLASS__, 'getRegexFromGitignore'], $ignoreLinesNegative);
+
+        $ignoreLinesPositive = array_map([__CLASS__, 'getRegexFromGitignore'], $ignoreLinesPositive);
+        if (empty($ignoreLinesPositive)) {
+            return '/^$/';
+        }
+
+        if (empty($ignoreLinesNegative)) {
+            return sprintf('/%s/', implode('|', $ignoreLinesPositive));
+        }
+
+        return sprintf('/(?=^(?:(?!(%s)).)*$)(%s)/', implode('|', $ignoreLinesNegative), implode('|', $ignoreLinesPositive));
+    }
+
+    private static function getRegexFromGitignore(string $gitignorePattern): string
+    {
+        $regex = '(';
+        if (0 === strpos($gitignorePattern, '/')) {
+            $gitignorePattern = substr($gitignorePattern, 1);
+            $regex .= '^';
+        } else {
+            $regex .= '(^|\/)';
+        }
+
+        if ('/' === $gitignorePattern[\strlen($gitignorePattern) - 1]) {
+            $gitignorePattern = substr($gitignorePattern, 0, -1);
+        }
+
+        $iMax = \strlen($gitignorePattern);
+        for ($i = 0; $i < $iMax; ++$i) {
+            $doubleChars = substr($gitignorePattern, $i, 2);
+            if ('**' === $doubleChars) {
+                $regex .= '.+';
+                ++$i;
+                continue;
+            }
+
+            $c = $gitignorePattern[$i];
+            switch ($c) {
+                case '*':
+                    $regex .= '[^\/]+';
+                    break;
+                case '/':
+                case '.':
+                case ':':
+                case '(':
+                case ')':
+                case '{':
+                case '}':
+                    $regex .= '\\'.$c;
+                    break;
+                default:
+                    $regex .= $c;
+            }
+        }
+
+        $regex .= '($|\/)';
+        $regex .= ')';
+
+        return $regex;
+    }
+}

src/Symfony/Component/Finder/Tests/FinderTest.php

@@ -347,6 +347,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder = $this->buildFinder();
         $this->assertSame($finder, $finder->ignoreVCS(false)->ignoreDotFiles(false));
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             '.git',
             'foo',
             'foo/bar.tmp',
@@ -373,6 +374,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder = $this->buildFinder();
         $finder->ignoreVCS(false)->ignoreVCS(false)->ignoreDotFiles(false);
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             '.git',
             'foo',
             'foo/bar.tmp',
@@ -399,6 +401,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder = $this->buildFinder();
         $this->assertSame($finder, $finder->ignoreVCS(true)->ignoreDotFiles(false));
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             'foo',
             'foo/bar.tmp',
             'test.php',
@@ -421,6 +424,28 @@ class FinderTest extends Iterator\RealIteratorTestCase
         ]), $finder->in(self::$tmpDir)->getIterator());
     }
 
+    public function testIgnoreVCSIgnored()
+    {
+        $finder = $this->buildFinder();
+        $this->assertSame(
+            $finder,
+            $finder
+                ->ignoreVCS(true)
+                ->ignoreDotFiles(true)
+                ->ignoreVCSIgnored(true)
+        );
+        $this->assertIterator($this->toAbsolute([
+            'foo',
+            'foo/bar.tmp',
+            'test.py',
+            'toto',
+            'foo bar',
+            'qux',
+            'qux/baz_100_1.py',
+            'qux/baz_1_2.py',
+        ]), $finder->in(self::$tmpDir)->getIterator());
+    }
+
     public function testIgnoreVCSCanBeDisabledAfterFirstIteration()
     {
         $finder = $this->buildFinder();
@@ -428,6 +453,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder->ignoreDotFiles(false);
 
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             'foo',
             'foo/bar.tmp',
             'qux',
@@ -450,7 +476,9 @@ class FinderTest extends Iterator\RealIteratorTestCase
         ]), $finder->getIterator());
 
         $finder->ignoreVCS(false);
-        $this->assertIterator($this->toAbsolute(['.git',
+        $this->assertIterator($this->toAbsolute([
+            '.gitignore',
+            '.git',
             'foo',
             'foo/bar.tmp',
             'qux',
@@ -479,6 +507,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder = $this->buildFinder();
         $this->assertSame($finder, $finder->ignoreDotFiles(false)->ignoreVCS(false));
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             '.git',
             '.bar',
             '.foo',
@@ -505,6 +534,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
         $finder = $this->buildFinder();
         $finder->ignoreDotFiles(false)->ignoreDotFiles(false)->ignoreVCS(false);
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             '.git',
             '.bar',
             '.foo',
@@ -574,6 +604,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
 
         $finder->ignoreDotFiles(false);
         $this->assertIterator($this->toAbsolute([
+            '.gitignore',
             'foo',
             'foo/bar.tmp',
             'qux',
@@ -842,6 +873,7 @@ class FinderTest extends Iterator\RealIteratorTestCase
 
         $expected = [
             self::$tmpDir.\DIRECTORY_SEPARATOR.'test.php',
+            __DIR__.\DIRECTORY_SEPARATOR.'GitignoreTest.php',
             __DIR__.\DIRECTORY_SEPARATOR.'FinderTest.php',
             __DIR__.\DIRECTORY_SEPARATOR.'GlobTest.php',
             self::$tmpDir.\DIRECTORY_SEPARATOR.'qux_0_1.php',

src/Symfony/Component/Finder/Tests/GitignoreTest.php

@@ -0,0 +1,118 @@
+<?php
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Finder\Tests;
+
+use PHPUnit\Framework\TestCase;
+use Symfony\Component\Finder\Gitignore;
+
+class GitignoreTest extends TestCase
+{
+    /**
+     * @dataProvider provider
+     *
+     * @param string $patterns
+     * @param array  $matchingCases
+     * @param array  $nonMatchingCases
+     */
+    public function testCases(string $patterns, array $matchingCases, array $nonMatchingCases)
+    {
+        $regex = Gitignore::toRegex($patterns);
+
+        foreach ($matchingCases as $matchingCase) {
+            $this->assertRegExp($regex, $matchingCase, sprintf('Failed asserting path [%s] matches gitignore patterns [%s] using regex [%s]', $matchingCase, $patterns, $regex));
+        }
+
+        foreach ($nonMatchingCases as $nonMatchingCase) {
+            $this->assertNotRegExp($regex, $nonMatchingCase, sprintf('Failed asserting path [%s] not matching gitignore patterns [%s] using regex [%s]', $nonMatchingCase, $patterns, $regex));
+        }
+    }
+
+    /**
+     * @return array return is array of
+     *               [
+     *               [
+     *               '', // Git-ignore Pattern
+     *               [], // array of file paths matching
+     *               [], // array of file paths not matching
+     *               ],
+     *               ]
+     */
+    public function provider()
+    {
+        return [
+            [
+                '
+                    *
+                    !/bin/bash
+                ',
+                ['bin/cat', 'abc/bin/cat'],
+                ['bin/bash'],
+            ],
+            [
+                'fi#le.txt',
+                [],
+                ['#file.txt'],
+            ],
+            [
+                '
+                /bin/
+                /usr/local/
+                !/bin/bash
+                !/usr/local/bin/bash
+                ',
+                ['bin/cat'],
+                ['bin/bash'],
+            ],
+            [
+                '*.py[co]',
+                ['file.pyc', 'file.pyc'],
+                ['filexpyc', 'file.pycx', 'file.py'],
+            ],
+            [
+                'dir1/**/dir2/',
+                ['dir1/dirA/dir2/', 'dir1/dirA/dirB/dir2/'],
+                [],
+            ],
+            [
+                'dir1/*/dir2/',
+                ['dir1/dirA/dir2/'],
+                ['dir1/dirA/dirB/dir2/'],
+            ],
+            [
+                '/*.php',
+                ['file.php'],
+                ['app/file.php'],
+            ],
+            [
+                '\#file.txt',
+                ['#file.txt'],
+                [],
+            ],
+            [
+                '*.php',
+                ['app/file.php', 'file.php'],
+                ['file.phps', 'file.phps', 'filephps'],
+            ],
+            [
+                'app/cache/',
+                ['app/cache/file.txt', 'app/cache/dir1/dir2/file.txt', 'a/app/cache/file.txt'],
+                [],
+            ],
+            [
+                '
+                #IamComment
+                /app/cache/',
+                ['app/cache/file.txt', 'app/cache/subdir/ile.txt'],
+                ['a/app/cache/file.txt'],
+            ],
+        ];
+    }
+}

src/Symfony/Component/Finder/Tests/Iterator/DepthRangeFilterIteratorTest.php

@@ -33,6 +33,7 @@ class DepthRangeFilterIteratorTest extends RealIteratorTestCase
     public function getAcceptData()
     {
         $lessThan1 = [
+            '.gitignore',
             '.git',
             'test.py',
             'foo',
@@ -51,6 +52,7 @@ class DepthRangeFilterIteratorTest extends RealIteratorTestCase
         ];
 
         $lessThanOrEqualTo1 = [
+            '.gitignore',
             '.git',
             'test.py',
             'foo',

src/Symfony/Component/Finder/Tests/Iterator/ExcludeDirectoryFilterIteratorTest.php

@@ -31,6 +31,7 @@ class ExcludeDirectoryFilterIteratorTest extends RealIteratorTestCase
     public function getAcceptData()
     {
         $foo = [
+            '.gitignore',
             '.bar',
             '.foo',
             '.foo/.bar',
@@ -53,6 +54,7 @@ class ExcludeDirectoryFilterIteratorTest extends RealIteratorTestCase
         ];
 
         $fo = [
+            '.gitignore',
             '.bar',
             '.foo',
             '.foo/.bar',
@@ -77,6 +79,7 @@ class ExcludeDirectoryFilterIteratorTest extends RealIteratorTestCase
         ];
 
         $toto = [
+            '.gitignore',
             '.bar',
             '.foo',
             '.foo/.bar',

src/Symfony/Component/Finder/Tests/Iterator/RealIteratorTestCase.php

@@ -63,6 +63,8 @@ abstract class RealIteratorTestCase extends IteratorTestCase
         file_put_contents(self::toAbsolute('test.php'), str_repeat(' ', 800));
         file_put_contents(self::toAbsolute('test.py'), str_repeat(' ', 2000));
 
+        file_put_contents(self::toAbsolute('.gitignore'), '*.php');
+
         touch(self::toAbsolute('foo/bar.tmp'), strtotime('2005-10-15'));
         touch(self::toAbsolute('test.php'), strtotime('2005-10-15'));
     }