Canonicalizer

.gitignore

@@ -1,2 +1,3 @@
+.idea
 vendor
 composer.lock

src/Helper/functions.php

@@ -2,6 +2,7 @@
 
 namespace BenTools\UriFactory\Helper;
 
+use BenTools\UriFactory\UriCanonicalizer;
 use BenTools\UriFactory\UriFactory;
 use BenTools\UriFactory\UriFactoryInterface;
 use Psr\Http\Message\UriInterface;
@@ -19,8 +20,19 @@ function uri(string $uri, UriFactoryInterface $factory = null): UriInterface
 /**
  * @param UriFactoryInterface|null $factory
  * @return UriInterface
+ * @throws \RuntimeException
  */
 function current_location(UriFactoryInterface $factory = null): UriInterface
 {
     return UriFactory::factory()->createUriFromCurrentLocation($factory);
 }
+
+/**
+ * @param UriInterface $uri
+ * @return UriInterface
+ * @throws \InvalidArgumentException
+ */
+function canonicalize(UriInterface $uri): UriInterface
+{
+    return UriCanonicalizer::canonicalize($uri);
+}

src/UriCanonicalizer.php

@@ -0,0 +1,165 @@
+<?php
+
+namespace BenTools\UriFactory;
+
+use Psr\Http\Message\UriInterface;
+
+final class UriCanonicalizer
+{
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function canonicalize(UriInterface $uri): UriInterface
+    {
+        $uri = self::ensureIsPercentUnescaped($uri);
+        $uri = self::ensureSchemeIsNotBlank($uri);
+        $uri = self::removeUnwantedChars($uri);
+        $uri = self::removeFragment($uri);
+        $uri = self::removeLeadingAndTrailingDots($uri);
+        $uri = self::replaceConsecutiveDotsWithASingleDot($uri);
+        $uri = self::normalizeHostname($uri);
+        $uri = self::normalizePath($uri);
+        return $uri;
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function ensureIsPercentUnescaped(UriInterface $uri): UriInterface
+    {
+        return $uri
+            ->withHost(self::percentUnescape($uri->getHost()))
+            ->withPath(self::percentUnescape($uri->getPath()))
+            ;
+    }
+
+    /**
+     * @param string $string
+     * @return string
+     */
+    public static function percentUnescape(string $string): string
+    {
+        while ($string !== ($decoded = urldecode($string))) {
+            $string = $decoded;
+        }
+        return $string;
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function ensureSchemeIsNotBlank(UriInterface $uri): UriInterface
+    {
+        return '' === $uri->getScheme() ? $uri->withScheme('http') : $uri;
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     */
+    public static function removeFragment(UriInterface $uri): UriInterface
+    {
+        return $uri->withFragment('');
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function removeUnwantedChars(UriInterface $uri): UriInterface
+    {
+        $removeUnwantedChars = function (?string $string) {
+            if (null === $string) {
+                return null;
+            }
+            return str_replace(["\x09", "\x0A", "\x0D", "\x0B", "\t", "\r", "\n"], '', $string);
+        };
+
+        return $uri
+            ->withUserInfo($removeUnwantedChars($uri->getUserInfo()))
+            ->withHost($removeUnwantedChars($uri->getHost()))
+            ->withPath($removeUnwantedChars($uri->getPath()))
+            ->withQuery($removeUnwantedChars($uri->getQuery()))
+            ->withFragment($removeUnwantedChars($uri->getFragment()));
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function removeLeadingAndTrailingDots(UriInterface $uri): UriInterface
+    {
+        return $uri->withHost(
+            trim($uri->getHost(), '.')
+        );
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function replaceConsecutiveDotsWithASingleDot(UriInterface $uri): UriInterface
+    {
+        return $uri->withHost(
+            preg_replace('/\.{2,}/', '.', $uri->getHost())
+        );
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function normalizeHostname(UriInterface $uri): UriInterface
+    {
+        $hostname = strtolower($uri->getHost());
+
+        $hostnameIP = is_numeric($hostname) ? ip2long(long2ip($hostname)) : ip2long($hostname);
+
+        if (false !== $hostnameIP) {
+            $hostname = long2ip($hostnameIP);
+        }
+
+        return $uri->withHost($hostname);
+    }
+
+    /**
+     * @param UriInterface $uri
+     * @return UriInterface
+     * @throws \InvalidArgumentException
+     */
+    public static function normalizePath(UriInterface $uri): UriInterface
+    {
+        $path = $uri->getPath();
+        $segments = explode('/', $path);
+        $parts = [];
+        foreach ($segments as $segment) {
+            switch ($segment) {
+                case '.':
+                    // Don't need to do anything here
+                    break;
+                case '..':
+                    array_pop($parts);
+                    break;
+                default:
+                    $parts[] = $segment;
+                    break;
+            }
+        }
+        $path = implode('/', $parts);
+        $path = preg_replace('#/{2,}#', '/', $path);
+        if (0 !== strpos($path, '/')) {
+            $path = '/' . $path;
+        }
+        return $uri->withPath($path);
+    }
+}

tests/UriCanonicalizerTest.php

@@ -0,0 +1,95 @@
+<?php
+
+namespace BenTools\UriFactory\Tests;
+
+use BenTools\UriFactory\Adapter\GuzzleAdapter;
+use function BenTools\UriFactory\Helper\canonicalize;
+use function BenTools\UriFactory\Helper\uri;
+use BenTools\UriFactory\UriCanonicalizer;
+use PHPUnit\Framework\TestCase;
+
+class UriCanonicalizerTest extends TestCase
+{
+
+    public function testEnsureSchemeIsNotBlank()
+    {
+        $uri = uri('www.example.org');
+        $this->assertEquals('http', UriCanonicalizer::ensureSchemeIsNotBlank($uri)->getScheme());
+    }
+
+    public function testRemoveFragment()
+    {
+        $uri = uri('http://example.org/foo#bar');
+        $this->assertEquals('', UriCanonicalizer::removeFragment($uri)->getFragment());
+    }
+
+    public function testRemoveLeadingAndTrailingDots()
+    {
+        $uri = uri('http://...www.example.org.../foo');
+        $this->assertEquals('www.example.org', UriCanonicalizer::removeLeadingAndTrailingDots($uri)->getHost());
+    }
+
+    public function testReplaceConsecutiveDotsWithASingleDot()
+    {
+        $uri = uri('http://www..example...org/foo');
+        $this->assertEquals('www.example.org', UriCanonicalizer::replaceConsecutiveDotsWithASingleDot($uri)->getHost());
+    }
+
+    public function testNormalizeHostname()
+    {
+        $uri = uri('http://WWW.EXAMPLE.ORG');
+        $this->assertEquals('www.example.org', UriCanonicalizer::normalizeHostname($uri)->getHost());
+    }
+
+    public function testNormalizePath()
+    {
+        $uri = uri('http://example.org/apple/../../cherry/banana/strawberry/../../orange/./pear//pineapple/');
+        $this->assertEquals('/cherry/orange/pear/pineapple/', UriCanonicalizer::normalizePath($uri)->getPath());
+        $uri = uri('http://example.org//');
+        $this->assertEquals('/', UriCanonicalizer::normalizePath($uri)->getPath());
+        $uri = uri('http://example.org/../');
+        $this->assertEquals('/', UriCanonicalizer::normalizePath($uri)->getPath());
+        $uri = uri('http://example.org/');
+        $this->assertEquals('/', UriCanonicalizer::normalizePath($uri)->getPath());
+        $uri = uri('http://example.org');
+        $this->assertEquals('/', UriCanonicalizer::normalizePath($uri)->getPath());
+    }
+
+    /**
+     * @dataProvider dataProvider
+     */
+    public function testCanonicalize(string $testedUrl, string $expectedUrl)
+    {
+        $result = (string) canonicalize(uri($testedUrl, GuzzleAdapter::factory()));
+        $this->assertEquals($expectedUrl, $result);
+    }
+
+    public function dataProvider()
+    {
+        return [
+            ['http://host/%25%32%35', 'http://host/%25'],
+            ['http://host/%25%32%35%25%32%35', 'http://host/%25%25'],
+            ['http://host/%2525252525252525', 'http://host/%25'],
+            ['http://host/asdf%25%32%35asd', 'http://host/asdf%25asd'],
+            ['http://host/%%%25%32%35asd%%', 'http://host/%25%25%25asd%25%25'],
+            ['http://www.google.com/', 'http://www.google.com/'],
+            ['http://%31%36%38%2e%31%38%38%2e%39%39%2e%32%36/%2E%73%65%63%75%72%65/%77%77%77%2E%65%62%61%79%2E%63%6F%6D/', 'http://168.188.99.26/.secure/www.ebay.com/'],
+            ['http://195.127.0.11/uploads/%20%20%20%20/.verify/.eBaysecure=updateuserdataxplimnbqmn-xplmvalidateinfoswqpcmlx=hgplmcx/', 'http://195.127.0.11/uploads/%20%20%20%20/.verify/.eBaysecure=updateuserdataxplimnbqmn-xplmvalidateinfoswqpcmlx=hgplmcx/'],
+            ['http://host.com/%257Ea%2521b%2540c%2523d%2524e%25f%255E00%252611%252A22%252833%252944_55%252B', 'http://host.com/~a!b@c%23d$e%25f%5E00&11*22(33)44_55%20'],
+            ['http://3279880203/blah', 'http://195.127.0.11/blah'],
+            ['http://www.google.com/blah/..', 'http://www.google.com/'],
+            ['http://www.evil.com/blah#frag', 'http://www.evil.com/blah'],
+            ['http://www.GOOgle.com/', 'http://www.google.com/'],
+            ['http://www.google.com.../', 'http://www.google.com/'],
+            ['http://www.google.com/q?r?', 'http://www.google.com/q?r?'],
+            ['http://www.google.com/q?r?s', 'http://www.google.com/q?r?s'],
+            ['http://evil.com/foo#bar#baz', 'http://evil.com/foo'],
+            ['http://evil.com/foo;', 'http://evil.com/foo;'],
+            ['http://evil.com/foo?bar;', 'http://evil.com/foo?bar;'],
+            ['http://notrailingslash.com', 'http://notrailingslash.com/'],
+            ['https://www.securesite.com/', 'https://www.securesite.com/'],
+            ['http://host.com/ab%23cd', 'http://host.com/ab%23cd'],
+            ['http://host.com//twoslashes?more//slashes', 'http://host.com/twoslashes?more//slashes'],
+        ];
+    }
+}