Adding new mwthods to UriString

components/Components/Path.php

@@ -19,19 +19,14 @@
 use League\Uri\Contracts\UriInterface;
 use League\Uri\Encoder;
 use League\Uri\Uri;
+use League\Uri\UriString;
 use Psr\Http\Message\UriInterface as Psr7UriInterface;
 use Stringable;
 
-use function array_pop;
-use function array_reduce;
-use function end;
-use function explode;
-use function implode;
 use function substr;
 
 final class Path extends Component implements PathInterface
 {
-    private const DOT_SEGMENTS = ['.' => 1, '..' => 1];
     private const SEPARATOR = '/';
 
     private readonly string $path;
@@ -116,35 +111,7 @@ public function withoutDotSegments(): PathInterface
             return $this;
         }
 
-        $input = explode(self::SEPARATOR, $current);
-        $new = implode(self::SEPARATOR, array_reduce($input, $this->filterDotSegments(...), []));
-        if (isset(self::DOT_SEGMENTS[end($input)])) {
-            $new .= self::SEPARATOR ;
-        }
-
-        return new self($new);
-    }
-
-    /**
-     * Filter Dot segment according to RFC3986.
-     *
-     * @see http://tools.ietf.org/html/rfc3986#section-5.2.4
-     *
-     * @return string[]
-     */
-    private function filterDotSegments(array $carry, string $segment): array
-    {
-        if ('..' === $segment) {
-            array_pop($carry);
-
-            return $carry;
-        }
-
-        if (!isset(self::DOT_SEGMENTS[$segment])) {
-            $carry[] = $segment;
-        }
-
-        return $carry;
+        return new self(UriString::removeDotSegments($current));
     }
 
     /**

composer.json

@@ -68,7 +68,7 @@
         "benchmark": "phpbench run --report=default",
         "phpcs": "PHP_CS_FIXER_IGNORE_ENV=1 php-cs-fixer fix -v --diff --dry-run --allow-risky=yes --ansi",
         "phpcs:fix": "php-cs-fixer fix -vvv --allow-risky=yes --ansi",
-        "phpstan": "phpstan analyse -l max -c phpstan.neon --ansi --memory-limit=256M",
+        "phpstan": "phpstan analyse -l max -c phpstan.neon --ansi --memory-limit=512M",
         "phpunit": "XDEBUG_MODE=coverage phpunit --coverage-text",
         "phpunit:min": "phpunit --no-coverage",
         "test": [

docs/interfaces/7.0/uri-parser-builder.md

@@ -15,6 +15,7 @@ The class act as a drop-in replacement for PHP's `parse_url` feature.
 ## URI parsing
 
 ~~~php
+UriString::resolve(string $uri, ?string $baseUri = null): array
 UriString::parse(string $uri): array
 UriString::parseAuthority(string $autority): array
 ~~~
@@ -67,6 +68,25 @@ var_export(UriString::parse('http:www.example.com'));
 <p class="message-warning">This invalid HTTP URI is successfully parsed.</p>
 <p class="message-notice">The class also exposes a <code>UriString::parseAuthority</code> you can use to parse an authority string.</p>
 
+If you need to resolve your URI in the context of a Base URI the `resolve` public static method will let you
+do just that. The method expect either a full URI as its single parameter or a relative URI following by
+a base URI which must be absolute, the URI will then be resolved using the base URI.
+
+```php
+$components = UriString::resolve('"/foo", "https://example.com");
+//returns the following array
+//array(
+//  'scheme' => 'https',
+//  'user' => null,
+//  'pass' => null,
+//  'host' => 'example.com'',
+//  'port' => null,
+//  'path' => '/foo',
+//  'query' => null,
+//  'fragment' => null,
+//);
+```
+
 ## URI Building
 
 ~~~php
@@ -99,3 +119,39 @@ echo UriString::build($components); //displays http://hello:[email protected][email protected]
 The `build` method provides similar functionality to the `http_build_url()` function from v1.x of the [`pecl_http`](https://pecl.php.net/package/pecl_http) PECL extension.
 
 <p class="message-notice">The class also exposes a <code>UriString::buildAuthority</code> you can use to build an authority from its hash representation.</p>
+
+## URI Normalization
+
+It is possible to normalize a URI against the RFC3986 rules using the `UriString::normalize` method.
+The method expects a string and will return the same array as `UriString::parse` but each component will
+have been normalized.
+
+```php
+use League\Uri\UriString;
+
+$parsed = UriString::parse("https://EXAMPLE.COM/foo/../bar");
+//returns the following array
+//array(
+//  'scheme' => 'http',
+//  'user' => null,
+//  'pass' => null,
+//  'host' => 'EXAMPLE.COM',
+//  'port' => null,
+//  'path' => '/foo/../bar',
+//  'query' => null,
+//  'fragment' => null,
+//);
+
+$normalized = UriString::normalize("https://EXAMPLE.COM/foo/../bar");
+//returns the following array
+//array(
+//  'scheme' => 'http',
+//  'user' => null,
+//  'pass' => null,
+//  'host' => 'example.com',
+//  'port' => null,
+//  'path' => '/bar',
+//  'query' => null,
+//  'fragment' => null,
+//);
+```

docs/uri/7.0/index.md

@@ -49,9 +49,6 @@ install the `fileinfo` extension otherwise an exception will be thrown.
 To convert a URI into an HTML anchor tag you need to have the `ext-dom` extension
 installed in your system.
 
-To enable URI normalization, the `ext-mbstring` extension or a polyfill
-like `symfony/polyfill-mbstring` must be present in your system.
-
 Installation
 --------
 

interfaces/CHANGELOG.md

@@ -15,6 +15,9 @@ All Notable changes to `League\Uri\Interfaces` will be documented in this file
 - `UriInterface::toNormalizedString`
 - `UriInterface::getUser`
 - `League\Uri\IPv6\Converter::isIpv6`
+- `UriString::resolve`
+- `UriString::removeDotSegments`
+- `UriString::normalize`
 
 ### Fixed
 

interfaces/Encoder.php

@@ -41,6 +41,13 @@ final class Encoder
     private const REGEXP_PART_UNRESERVED = 'A-Za-z\d_\-.~';
     private const REGEXP_PART_ENCODED = '%(?![A-Fa-f\d]{2})';
 
+    /**
+     * Unreserved characters.
+     *
+     * @see https://www.rfc-editor.org/rfc/rfc3986.html#section-2.3
+     */
+    private const REGEXP_UNRESERVED_CHARACTERS = ',%(2[1-9A-Fa-f]|[3-7][0-9A-Fa-f]|61|62|64|65|66|7[AB]|5F),';
+
     /**
      * Encode User.
      *
@@ -173,4 +180,17 @@ private static function decode(Stringable|string|int|null $component, Closure $d
             default => $component,
         };
     }
+
+    public static function decodeUnreservedCharacters(?string $str): ?string
+    {
+        return match (true) {
+            null === $str,
+            '' === $str => $str,
+            default => preg_replace_callback(
+                self::REGEXP_UNRESERVED_CHARACTERS,
+                static fn (array $matches): string => rawurldecode($matches[0]),
+                $str
+            ) ?? '',
+        };
+    }
 }

interfaces/Idna/Converter.php

@@ -141,7 +141,7 @@ public static function toUnicode(Stringable|string $domain, Option|int|null $opt
         $domain = rawurldecode((string) $domain);
 
         if (false === stripos($domain, 'xn--')) {
-            return Result::fromIntl(['result' => $domain, 'isTransitionalDifferent' => false, 'errors' => Error::NONE->value]);
+            return Result::fromIntl(['result' => strtolower($domain), 'isTransitionalDifferent' => false, 'errors' => Error::NONE->value]);
         }
 
         FeatureDetection::supportsIdn();