UriString::parseAuthority() - Code Metrics - thephpleague/uri - Measure and Improve Code Quality continuously with Scrutinizer

UriString::parseAuthority() A
last analyzed 2020-08-13 17:22 UTC

↳ Parent: UriString

Complexity

Conditions	3
Paths	3

Size

Total Lines	19
Code Lines	11

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	12
CRAP Score	3

Importance

Changes	2
Bugs	0	Features	0

Metric	Value
cc	3
eloc	11
c	2
b	0
f	0
nc	3
nop	1
dl	0
loc	19
ccs	12
cts	12
cp	1
crap	3
rs	9.9

<?php

/**
 * League.Uri (https://uri.thephpleague.com)
 *
 * (c) Ignace Nyamagana Butera <[email protected]>
 *
 * For the full copyright and license information, please view the LICENSE
 * file that was distributed with this source code.
 */

declare(strict_types=1);

namespace League\Uri;

use League\Uri\Exceptions\IdnSupportMissing;
use League\Uri\Exceptions\SyntaxError;
use function array_merge;
use function defined;
use function explode;
use function filter_var;
use function function_exists;
use function gettype;
use function idn_to_ascii;
use function implode;
use function inet_pton;
use function is_object;
use function is_scalar;
use function method_exists;
use function preg_match;
use function rawurldecode;
use function sprintf;
use function strpos;
use function substr;
use const FILTER_FLAG_IPV6;
use const FILTER_VALIDATE_IP;
use const IDNA_ERROR_BIDI;
use const IDNA_ERROR_CONTEXTJ;
use const IDNA_ERROR_DISALLOWED;
use const IDNA_ERROR_DOMAIN_NAME_TOO_LONG;
use const IDNA_ERROR_EMPTY_LABEL;
use const IDNA_ERROR_HYPHEN_3_4;
use const IDNA_ERROR_INVALID_ACE_LABEL;
use const IDNA_ERROR_LABEL_HAS_DOT;
use const IDNA_ERROR_LABEL_TOO_LONG;
use const IDNA_ERROR_LEADING_COMBINING_MARK;
use const IDNA_ERROR_LEADING_HYPHEN;
use const IDNA_ERROR_PUNYCODE;
use const IDNA_ERROR_TRAILING_HYPHEN;
use const INTL_IDNA_VARIANT_UTS46;

/**
 * A class to parse a URI string according to RFC3986.
 *
 * @link    https://tools.ietf.org/html/rfc3986
 * @package League\Uri
 * @author  Ignace Nyamagana Butera <[email protected]>
 * @since   6.0.0
 */
final class UriString
{
    /**
     * Default URI component values.
     */
    private const URI_COMPONENTS = [
        'scheme' => null, 'user' => null, 'pass' => null, 'host' => null,
        'port' => null, 'path' => '', 'query' => null, 'fragment' => null,
    ];

    /**
     * Simple URI which do not need any parsing.
     */
    private const URI_SCHORTCUTS = [
        '' => [],
        '#' => ['fragment' => ''],
        '?' => ['query' => ''],
        '?#' => ['query' => '', 'fragment' => ''],
        '/' => ['path' => '/'],
        '//' => ['host' => ''],
    ];

    /**
     * Range of invalid characters in URI string.
     */
    private const REGEXP_INVALID_URI_CHARS = '/[\x00-\x1f\x7f]/';

    /**
     * RFC3986 regular expression URI splitter.
     *
     * @link https://tools.ietf.org/html/rfc3986#appendix-B
     */
    private const REGEXP_URI_PARTS = ',^
        (?<scheme>(?<scontent>[^:/?\#]+):)?    # URI scheme component
        (?<authority>//(?<acontent>[^/?\#]*))? # URI authority part
        (?<path>[^?\#]*)                       # URI path component
        (?<query>\?(?<qcontent>[^\#]*))?       # URI query component
        (?<fragment>\#(?<fcontent>.*))?        # URI fragment component
    ,x';

    /**
     * URI scheme regular expresssion.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.1
     */
    private const REGEXP_URI_SCHEME = '/^([a-z][a-z\d\+\.\-]*)?$/i';

    /**
     * IPvFuture regular expression.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     */
    private const REGEXP_IP_FUTURE = '/^
        v(?<version>[A-F0-9])+\.
        (?:
            (?<unreserved>[a-z0-9_~\-\.])|
            (?<sub_delims>[!$&\'()*+,;=:])  # also include the : character
        )+
    $/ix';

    /**
     * General registered name regular expression.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     */
    private const REGEXP_REGISTERED_NAME = '/(?(DEFINE)
        (?<unreserved>[a-z0-9_~\-])   # . is missing as it is used to separate labels
        (?<sub_delims>[!$&\'()*+,;=])
        (?<encoded>%[A-F0-9]{2})
        (?<reg_name>(?:(?&unreserved)|(?&sub_delims)|(?&encoded))*)
    )
    ^(?:(?&reg_name)\.)*(?&reg_name)\.?$/ix';

    /**
     * Invalid characters in host regular expression.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     */
    private const REGEXP_INVALID_HOST_CHARS = '/
        [:\/?#\[\]@ ]  # gen-delims characters as well as the space character
    /ix';

    /**
     * Invalid path for URI without scheme and authority regular expression.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.3
     */
    private const REGEXP_INVALID_PATH = ',^(([^/]*):)(.*)?/,';

    /**
     * Host and Port splitter regular expression.
     */
    private const REGEXP_HOST_PORT = ',^(?<host>\[.*\]|[^:]*)(:(?<port>.*))?$,';

    /**
     * IDN Host detector regular expression.
     */
    private const REGEXP_IDN_PATTERN = '/[^\x20-\x7f]/';

    /**
     * Only the address block fe80::/10 can have a Zone ID attach to
     * let's detect the link local significant 10 bits.
     */
    private const ZONE_ID_ADDRESS_BLOCK = "\xfe\x80";

    /**
     * Generate an URI string representation from its parsed representation
     * returned by League\Uri\parse() or PHP's parse_url.
     *
     * If you supply your own array, you are responsible for providing
     * valid components without their URI delimiters.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-5.3
     * @link https://tools.ietf.org/html/rfc3986#section-7.5
     *
     * @param array{
     *  scheme:?string,
     *  user:?string,
     *  pass:?string,
     *  host:?string,
     *  port:?int,
     *  path:string,
     *  query:?string,
     *  fragment:?string
     * } $components
     */
    public static function build(array $components): string
    {
        $result = $components['path'] ?? '';
        if (isset($components['query'])) {
            $result .= '?'.$components['query'];
        }

        if (isset($components['fragment'])) {
            $result .= '#'.$components['fragment'];
        }

        $scheme = null;
        if (isset($components['scheme'])) {
            $scheme = $components['scheme'].':';
        }

        if (!isset($components['host'])) {
            return $scheme.$result;
        }

        $scheme .= '//';
        $authority = $components['host'];
        if (isset($components['port'])) {
            $authority .= ':'.$components['port'];
        }

        if (!isset($components['user'])) {
            return $scheme.$authority.$result;
        }

        $authority = '@'.$authority;
        if (!isset($components['pass'])) {
            return $scheme.$components['user'].$authority.$result;
        }

        return $scheme.$components['user'].':'.$components['pass'].$authority.$result;
    }

    /**
     * Parse an URI string into its components.
     *
     * This method parses a URI and returns an associative array containing any
     * of the various components of the URI that are present.
     *
     * <code>
     * $components = (new Parser())->parse('http://[email protected]:42?query#');
     * var_export($components);
     * //will display
     * array(
     *   'scheme' => 'http',           // the URI scheme component
     *   'user' => 'foo',              // the URI user component
     *   'pass' => null,               // the URI pass component
     *   'host' => 'test.example.com', // the URI host component
     *   'port' => 42,                 // the URI port component
     *   'path' => '',                 // the URI path component
     *   'query' => 'query',           // the URI query component
     *   'fragment' => '',             // the URI fragment component
     * );
     * </code>
     *
     * The returned array is similar to PHP's parse_url return value with the following
     * differences:
     *
     * <ul>
     * <li>All components are always present in the returned array</li>
     * <li>Empty and undefined component are treated differently. And empty component is
     *   set to the empty string while an undefined component is set to the `null` value.</li>
     * <li>The path component is never undefined</li>
     * <li>The method parses the URI following the RFC3986 rules but you are still
     *   required to validate the returned components against its related scheme specific rules.</li>
     * </ul>
     *
     * @link https://tools.ietf.org/html/rfc3986
     *
     * @param mixed $uri any scalar or stringable object
     *
     * @throws SyntaxError if the URI contains invalid characters
     * @throws SyntaxError if the URI contains an invalid scheme
     * @throws SyntaxError if the URI contains an invalid path
     *
     * @return array{

     *                scheme:?string,
     *                user:?string,
     *                pass:?string,
     *                host:?string,
     *                port:?int,
     *                path:string,
     *                query:?string,
     *                fragment:?string
     *                }
     */
    public static function parse($uri): array
    {
        if (is_object($uri) && method_exists($uri, '__toString')) {
            $uri = (string) $uri;
        }

        if (!is_scalar($uri)) {
            throw new \TypeError(sprintf('The uri must be a scalar or a stringable object `%s` given', gettype($uri)));
        }

        $uri = (string) $uri;

        if (isset(self::URI_SCHORTCUTS[$uri])) {
            /** @var array{scheme:?string, user:?string, pass:?string, host:?string, port:?int, path:string, query:?string, fragment:?string} $components */
            $components = array_merge(self::URI_COMPONENTS, self::URI_SCHORTCUTS[$uri]);

            return $components;
        }

        if (1 === preg_match(self::REGEXP_INVALID_URI_CHARS, $uri)) {
            throw new SyntaxError(sprintf('The uri `%s` contains invalid characters', $uri));
        }

        //if the first character is a known URI delimiter parsing can be simplified
        $first_char = $uri[0];

        //The URI is made of the fragment only
        if ('#' === $first_char) {
            [, $fragment] = explode('#', $uri, 2);
            $components = self::URI_COMPONENTS;
            $components['fragment'] = $fragment;

            return $components;
        }

        //The URI is made of the query and fragment
        if ('?' === $first_char) {
            [, $partial] = explode('?', $uri, 2);
            [$query, $fragment] = explode('#', $partial, 2) + [1 => null];
            $components = self::URI_COMPONENTS;
            $components['query'] = $query;
            $components['fragment'] = $fragment;

            return $components;
        }

        //use RFC3986 URI regexp to split the URI
        preg_match(self::REGEXP_URI_PARTS, $uri, $parts);
        $parts += ['query' => '', 'fragment' => ''];

        if (':' === $parts['scheme'] || 1 !== preg_match(self::REGEXP_URI_SCHEME, $parts['scontent'])) {
            throw new SyntaxError(sprintf('The uri `%s` contains an invalid scheme', $uri));
        }

        if ('' === $parts['scheme'].$parts['authority'] && 1 === preg_match(self::REGEXP_INVALID_PATH, $parts['path'])) {
            throw new SyntaxError(sprintf('The uri `%s` contains an invalid path.', $uri));
        }

        /** @var array{scheme:?string, user:?string, pass:?string, host:?string, port:?int, path:string, query:?string, fragment:?string} $components */
        $components = array_merge(
            self::URI_COMPONENTS,
            '' === $parts['authority'] ? [] : self::parseAuthority($parts['acontent']),
            [
                'path' => $parts['path'],
                'scheme' => '' === $parts['scheme'] ? null : $parts['scontent'],
                'query' => '' === $parts['query'] ? null : $parts['qcontent'],
                'fragment' => '' === $parts['fragment'] ? null : $parts['fcontent'],
            ]
        );

        return $components;
    }

    /**
     * Parses the URI authority part.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2
     *
     * @throws SyntaxError If the port component is invalid
     *
     * @return array{user:?string, pass:?string, host:?string, port:?int}
     */
    private static function parseAuthority(string $authority): array
    {
        $components = ['user' => null, 'pass' => null, 'host' => '', 'port' => null];
        if ('' === $authority) {
            return $components;
        }

        $parts = explode('@', $authority, 2);
        if (isset($parts[1])) {
            [$components['user'], $components['pass']] = explode(':', $parts[0], 2) + [1 => null];
        }

        preg_match(self::REGEXP_HOST_PORT, $parts[1] ?? $parts[0], $matches);
        $matches += ['port' => ''];

        $components['port'] = self::filterPort($matches['port']);
        $components['host'] = self::filterHost($matches['host']);

        return $components;
    }

    /**
     * Filter and format the port component.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     *
     * @throws SyntaxError if the registered name is invalid
     */
    private static function filterPort(string $port): ?int
    {
        if ('' === $port) {
            return null;
        }

        if (1 === preg_match('/^\d*$/', $port)) {
            return (int) $port;
        }

        throw new SyntaxError(sprintf('The port `%s` is invalid', $port));
    }

    /**
     * Returns whether a hostname is valid.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     *
     * @throws SyntaxError if the registered name is invalid
     */
    private static function filterHost(string $host): string
    {
        if ('' === $host) {
            return $host;
        }

        if ('[' !== $host[0] || ']' !== substr($host, -1)) {
            return self::filterRegisteredName($host);
        }

        if (!self::isIpHost(substr($host, 1, -1))) {
            throw new SyntaxError(sprintf('Host `%s` is invalid : the IP host is malformed', $host));
        }

        return $host;
    }

    /**
     * Returns whether the host is an IPv4 or a registered named.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     *
     * @throws SyntaxError       if the registered name is invalid
     * @throws IdnSupportMissing if IDN support or ICU requirement are not available or met.
     */
    private static function filterRegisteredName(string $host): string
    {
        // @codeCoverageIgnoreStart
        // added because it is not possible in travis to disabled the ext/intl extension
        // see travis issue https://github.com/travis-ci/travis-ci/issues/4701
        static $idn_support = null;
        $idn_support = $idn_support ?? function_exists('idn_to_ascii') && defined('INTL_IDNA_VARIANT_UTS46');
        // @codeCoverageIgnoreEnd

        $formatted_host = rawurldecode($host);
        if (1 === preg_match(self::REGEXP_REGISTERED_NAME, $formatted_host)) {
            if (false === strpos($formatted_host, 'xn--')) {
                return $host;
            }

            // @codeCoverageIgnoreStart
            if (!$idn_support) {
                throw new IdnSupportMissing(sprintf('the host `%s` could not be processed for IDN. Verify that ext/intl is installed for IDN support and that ICU is at least version 4.6.', $host));
            }
            // @codeCoverageIgnoreEnd

            $unicode = idn_to_utf8($host, 0, INTL_IDNA_VARIANT_UTS46, $arr);
            if (0 !== $arr['errors']) {
                throw new SyntaxError(sprintf('The host `%s` is invalid : %s', $host, self::getIDNAErrors($arr['errors'])));
            }

            // @codeCoverageIgnoreStart
            if (false === $unicode) {
                throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
            }
            // @codeCoverageIgnoreEnd

            return $host;
        }

        //to test IDN host non-ascii characters must be present in the host
        if (1 !== preg_match(self::REGEXP_IDN_PATTERN, $formatted_host)) {
            throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
        }

        // @codeCoverageIgnoreStart
        if (!$idn_support) {
            throw new IdnSupportMissing(sprintf('the host `%s` could not be processed for IDN. Verify that ext/intl is installed for IDN support and that ICU is at least version 4.6.', $host));
        }
        // @codeCoverageIgnoreEnd

        $retval = idn_to_ascii($formatted_host, 0, INTL_IDNA_VARIANT_UTS46, $arr);

        if ([] === $arr) {
            throw new SyntaxError(sprintf('Host `%s` is not a valid IDN host', $host));
        }

        if (0 !== $arr['errors']) {
            throw new SyntaxError(sprintf('Host `%s` is not a valid IDN host : %s', $host, self::getIDNAErrors($arr['errors'])));
        }

        // @codeCoverageIgnoreStart
        if (false === $retval) {
            throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
        }
        // @codeCoverageIgnoreEnd

        if (false !== strpos($retval, '%')) {
            throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
        }

        return $host;
    }

    /**
     * Retrieves and format IDNA conversion error message.
     *
     * @link http://icu-project.org/apiref/icu4j/com/ibm/icu/text/IDNA.Error.html
     */
    private static function getIDNAErrors(int $error_byte): string
    {
        /**
         * IDNA errors.
         */
        static $idn_errors = [
            IDNA_ERROR_EMPTY_LABEL => 'a non-final domain name label (or the whole domain name) is empty',
            IDNA_ERROR_LABEL_TOO_LONG => 'a domain name label is longer than 63 bytes',
            IDNA_ERROR_DOMAIN_NAME_TOO_LONG => 'a domain name is longer than 255 bytes in its storage form',
            IDNA_ERROR_LEADING_HYPHEN => 'a label starts with a hyphen-minus ("-")',
            IDNA_ERROR_TRAILING_HYPHEN => 'a label ends with a hyphen-minus ("-")',
            IDNA_ERROR_HYPHEN_3_4 => 'a label contains hyphen-minus ("-") in the third and fourth positions',
            IDNA_ERROR_LEADING_COMBINING_MARK => 'a label starts with a combining mark',
            IDNA_ERROR_DISALLOWED => 'a label or domain name contains disallowed characters',
            IDNA_ERROR_PUNYCODE => 'a label starts with "xn--" but does not contain valid Punycode',
            IDNA_ERROR_LABEL_HAS_DOT => 'a label contains a dot=full stop',
            IDNA_ERROR_INVALID_ACE_LABEL => 'An ACE label does not contain a valid label string',
            IDNA_ERROR_BIDI => 'a label does not meet the IDNA BiDi requirements (for right-to-left characters)',
            IDNA_ERROR_CONTEXTJ => 'a label does not meet the IDNA CONTEXTJ requirements',
        ];

        $res = [];
        foreach ($idn_errors as $error => $reason) {
            if ($error === ($error_byte & $error)) {
                $res[] = $reason;
            }
        }

        return [] === $res ? 'Unknown IDNA conversion error.' : implode(', ', $res).'.';
    }

    /**
     * Validates a IPv6/IPvfuture host.
     *
     * @link https://tools.ietf.org/html/rfc3986#section-3.2.2
     * @link https://tools.ietf.org/html/rfc6874#section-2
     * @link https://tools.ietf.org/html/rfc6874#section-4
     */
    private static function isIpHost(string $ip_host): bool
    {
        if (false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)) {
            return true;
        }

        if (1 === preg_match(self::REGEXP_IP_FUTURE, $ip_host, $matches)) {
            return !in_array($matches['version'], ['4', '6'], true);
        }

        $pos = strpos($ip_host, '%');
        if (false === $pos || 1 === preg_match(
            self::REGEXP_INVALID_HOST_CHARS,
            rawurldecode(substr($ip_host, $pos))
        )) {
            return false;
        }

        $ip_host = substr($ip_host, 0, $pos);

        return false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)
            && 0 === strpos((string) inet_pton($ip_host), self::ZONE_ID_ADDRESS_BLOCK);
    }
}


1		<?php
2
3		/**
4		* League.Uri (https://uri.thephpleague.com)
5		*
6		* (c) Ignace Nyamagana Butera <[email protected]>
7		*
8		* For the full copyright and license information, please view the LICENSE
9		* file that was distributed with this source code.
10		*/
11
12		declare(strict_types=1);
13
14		namespace League\Uri;
15
16		use League\Uri\Exceptions\IdnSupportMissing;
17		use League\Uri\Exceptions\SyntaxError;
18		use function array_merge;
19		use function defined;
20		use function explode;
21		use function filter_var;
22		use function function_exists;
23		use function gettype;
24		use function idn_to_ascii;
25		use function implode;
26		use function inet_pton;
27		use function is_object;
28		use function is_scalar;
29		use function method_exists;
30		use function preg_match;
31		use function rawurldecode;
32		use function sprintf;
33		use function strpos;
34		use function substr;
35		use const FILTER_FLAG_IPV6;
36		use const FILTER_VALIDATE_IP;
37		use const IDNA_ERROR_BIDI;
38		use const IDNA_ERROR_CONTEXTJ;
39		use const IDNA_ERROR_DISALLOWED;
40		use const IDNA_ERROR_DOMAIN_NAME_TOO_LONG;
41		use const IDNA_ERROR_EMPTY_LABEL;
42		use const IDNA_ERROR_HYPHEN_3_4;
43		use const IDNA_ERROR_INVALID_ACE_LABEL;
44		use const IDNA_ERROR_LABEL_HAS_DOT;
45		use const IDNA_ERROR_LABEL_TOO_LONG;
46		use const IDNA_ERROR_LEADING_COMBINING_MARK;
47		use const IDNA_ERROR_LEADING_HYPHEN;
48		use const IDNA_ERROR_PUNYCODE;
49		use const IDNA_ERROR_TRAILING_HYPHEN;
50		use const INTL_IDNA_VARIANT_UTS46;
51
52		/**
53		* A class to parse a URI string according to RFC3986.
54		*
55		* @link https://tools.ietf.org/html/rfc3986
56		* @package League\Uri
57		* @author Ignace Nyamagana Butera <[email protected]>
58		* @since 6.0.0
59		*/
60		final class UriString
61		{
62		/**
63		* Default URI component values.
64		*/
65		private const URI_COMPONENTS = [
66		'scheme' => null, 'user' => null, 'pass' => null, 'host' => null,
67		'port' => null, 'path' => '', 'query' => null, 'fragment' => null,
68		];
69
70		/**
71		* Simple URI which do not need any parsing.
72		*/
73		private const URI_SCHORTCUTS = [
74		'' => [],
75		'#' => ['fragment' => ''],
76		'?' => ['query' => ''],
77		'?#' => ['query' => '', 'fragment' => ''],
78		'/' => ['path' => '/'],
79		'//' => ['host' => ''],
80		];
81
82		/**
83		* Range of invalid characters in URI string.
84		*/
85		private const REGEXP_INVALID_URI_CHARS = '/[\x00-\x1f\x7f]/';
86
87		/**
88		* RFC3986 regular expression URI splitter.
89		*
90		* @link https://tools.ietf.org/html/rfc3986#appendix-B
91		*/
92		private const REGEXP_URI_PARTS = ',^
93		(?<scheme>(?<scontent>[^:/?\#]+):)? # URI scheme component
94		(?<authority>//(?<acontent>[^/?\#]*))? # URI authority part
95		(?<path>[^?\#]*) # URI path component
96		(?<query>\?(?<qcontent>[^\#]*))? # URI query component
97		(?<fragment>\#(?<fcontent>.*))? # URI fragment component
98		,x';
99
100		/**
101		* URI scheme regular expresssion.
102		*
103		* @link https://tools.ietf.org/html/rfc3986#section-3.1
104		*/
105		private const REGEXP_URI_SCHEME = '/^([a-z][a-z\d\+\.\-]*)?$/i';
106
107		/**
108		* IPvFuture regular expression.
109		*
110		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
111		*/
112		private const REGEXP_IP_FUTURE = '/^
113		v(?<version>[A-F0-9])+\.
114		(?:
115		(?<unreserved>[a-z0-9_~\-\.])\|
116		(?<sub_delims>[!$&\'()*+,;=:]) # also include the : character
117		)+
118		$/ix';
119
120		/**
121		* General registered name regular expression.
122		*
123		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
124		*/
125		private const REGEXP_REGISTERED_NAME = '/(?(DEFINE)
126		(?<unreserved>[a-z0-9_~\-]) # . is missing as it is used to separate labels
127		(?<sub_delims>[!$&\'()*+,;=])
128		(?<encoded>%[A-F0-9]{2})
129		(?<reg_name>(?:(?&unreserved)\|(?&sub_delims)\|(?&encoded))*)
130		)
131		^(?:(?&reg_name)\.)*(?&reg_name)\.?$/ix';
132
133		/**
134		* Invalid characters in host regular expression.
135		*
136		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
137		*/
138		private const REGEXP_INVALID_HOST_CHARS = '/
139		[:\/?#\[\]@ ] # gen-delims characters as well as the space character
140		/ix';
141
142		/**
143		* Invalid path for URI without scheme and authority regular expression.
144		*
145		* @link https://tools.ietf.org/html/rfc3986#section-3.3
146		*/
147		private const REGEXP_INVALID_PATH = ',^(([^/]):)(.)?/,';
148
149		/**
150		* Host and Port splitter regular expression.
151		*/
152		private const REGEXP_HOST_PORT = ',^(?<host>\[.\]\|[^:])(:(?<port>.*))?$,';
153
154		/**
155		* IDN Host detector regular expression.
156		*/
157		private const REGEXP_IDN_PATTERN = '/[^\x20-\x7f]/';
158
159		/**
160		* Only the address block fe80::/10 can have a Zone ID attach to
161		* let's detect the link local significant 10 bits.
162		*/
163		private const ZONE_ID_ADDRESS_BLOCK = "\xfe\x80";
164
165		/**
166		* Generate an URI string representation from its parsed representation
167		* returned by League\Uri\parse() or PHP's parse_url.
168		*
169		* If you supply your own array, you are responsible for providing
170		* valid components without their URI delimiters.
171		*
172		* @link https://tools.ietf.org/html/rfc3986#section-5.3
173		* @link https://tools.ietf.org/html/rfc3986#section-7.5
174		*
175		* @param array{
176		* scheme:?string,
177		* user:?string,
178		* pass:?string,
179		* host:?string,
180		* port:?int,
181		* path:string,
182		* query:?string,
183		* fragment:?string
184		* } $components
185		*/
186	94	public static function build(array $components): string
187		{
188	94	$result = $components['path'] ?? '';
189	94	if (isset($components['query'])) {
190	50	$result .= '?'.$components['query'];
191		}
192
193	94	if (isset($components['fragment'])) {
194	56	$result .= '#'.$components['fragment'];
195		}
196
197	94	$scheme = null;
198	94	if (isset($components['scheme'])) {
199	52	$scheme = $components['scheme'].':';
200		}
201
202	94	if (!isset($components['host'])) {
203	42	return $scheme.$result;
204		}
205
206	52	$scheme .= '//';
207	52	$authority = $components['host'];
208	52	if (isset($components['port'])) {
209	14	$authority .= ':'.$components['port'];
210		}
211
212	52	if (!isset($components['user'])) {
213	34	return $scheme.$authority.$result;
214		}
215
216	18	$authority = '@'.$authority;
217	18	if (!isset($components['pass'])) {
218	4	return $scheme.$components['user'].$authority.$result;
219		}
220
221	14	return $scheme.$components['user'].':'.$components['pass'].$authority.$result;
222		}
223
224		/**
225		* Parse an URI string into its components.
226		*
227		* This method parses a URI and returns an associative array containing any
228		* of the various components of the URI that are present.
229		*
230		* <code>
231		* $components = (new Parser())->parse('http://[email protected]:42?query#');
232		* var_export($components);
233		* //will display
234		* array(
235		* 'scheme' => 'http', // the URI scheme component
236		* 'user' => 'foo', // the URI user component
237		* 'pass' => null, // the URI pass component
238		* 'host' => 'test.example.com', // the URI host component
239		* 'port' => 42, // the URI port component
240		* 'path' => '', // the URI path component
241		* 'query' => 'query', // the URI query component
242		* 'fragment' => '', // the URI fragment component
243		* );
244		* </code>
245		*
246		* The returned array is similar to PHP's parse_url return value with the following
247		* differences:
248		*
249		* <ul>
250		* <li>All components are always present in the returned array</li>
251		* <li>Empty and undefined component are treated differently. And empty component is
252		* set to the empty string while an undefined component is set to the `null` value.</li>
253		* <li>The path component is never undefined</li>
254		* <li>The method parses the URI following the RFC3986 rules but you are still
255		* required to validate the returned components against its related scheme specific rules.</li>
256		* </ul>
257		*
258		* @link https://tools.ietf.org/html/rfc3986
259		*
260		* @param mixed $uri any scalar or stringable object
261		*
262		* @throws SyntaxError if the URI contains invalid characters
263		* @throws SyntaxError if the URI contains an invalid scheme
264		* @throws SyntaxError if the URI contains an invalid path
265		*
266		* @return array{
		0 ignored issues – show Documentation Bug introduced 2020-01-29 15:21 UTC by Report Bug Copy Issue Report The doc comment `array{` at position `2` could not be parsed: the token is null at position 2. Loading history...
267		* scheme:?string,
268		* user:?string,
269		* pass:?string,
270		* host:?string,
271		* port:?int,
272		* path:string,
273		* query:?string,
274		* fragment:?string
275		* }
276		*/
277	536	public static function parse($uri): array
278		{
279	536	if (is_object($uri) && method_exists($uri, '__toString')) {
280	2	$uri = (string) $uri;
281		}
282
283	536	if (!is_scalar($uri)) {
284	2	throw new \TypeError(sprintf('The uri must be a scalar or a stringable object `%s` given', gettype($uri)));
285		}
286
287	534	$uri = (string) $uri;
288
289	534	if (isset(self::URI_SCHORTCUTS[$uri])) {
290		/** @var array{scheme:?string, user:?string, pass:?string, host:?string, port:?int, path:string, query:?string, fragment:?string} $components */
291	46	$components = array_merge(self::URI_COMPONENTS, self::URI_SCHORTCUTS[$uri]);
292
293	46	return $components;
294		}
295
296	504	if (1 === preg_match(self::REGEXP_INVALID_URI_CHARS, $uri)) {
297	2	throw new SyntaxError(sprintf('The uri `%s` contains invalid characters', $uri));
298		}
299
300		//if the first character is a known URI delimiter parsing can be simplified
301	502	$first_char = $uri[0];
302
303		//The URI is made of the fragment only
304	502	if ('#' === $first_char) {
305	6	[, $fragment] = explode('#', $uri, 2);
306	6	$components = self::URI_COMPONENTS;
307	6	$components['fragment'] = $fragment;
308
309	6	return $components;
310		}
311
312		//The URI is made of the query and fragment
313	498	if ('?' === $first_char) {
314	6	[, $partial] = explode('?', $uri, 2);
315	6	[$query, $fragment] = explode('#', $partial, 2) + [1 => null];
316	6	$components = self::URI_COMPONENTS;
317	6	$components['query'] = $query;
318	6	$components['fragment'] = $fragment;
319
320	6	return $components;
321		}
322
323		//use RFC3986 URI regexp to split the URI
324	494	preg_match(self::REGEXP_URI_PARTS, $uri, $parts);
325	494	$parts += ['query' => '', 'fragment' => ''];
326
327	494	if (':' === $parts['scheme'] \|\| 1 !== preg_match(self::REGEXP_URI_SCHEME, $parts['scontent'])) {
328	6	throw new SyntaxError(sprintf('The uri `%s` contains an invalid scheme', $uri));
329		}
330
331	488	if ('' === $parts['scheme'].$parts['authority'] && 1 === preg_match(self::REGEXP_INVALID_PATH, $parts['path'])) {
332	2	throw new SyntaxError(sprintf('The uri `%s` contains an invalid path.', $uri));
333		}
334
335		/** @var array{scheme:?string, user:?string, pass:?string, host:?string, port:?int, path:string, query:?string, fragment:?string} $components */
336	486	$components = array_merge(
337	486	self::URI_COMPONENTS,
338	486	'' === $parts['authority'] ? [] : self::parseAuthority($parts['acontent']),
339		[
340	450	'path' => $parts['path'],
341	450	'scheme' => '' === $parts['scheme'] ? null : $parts['scontent'],
342	450	'query' => '' === $parts['query'] ? null : $parts['qcontent'],
343	450	'fragment' => '' === $parts['fragment'] ? null : $parts['fcontent'],
344		]
345		);
346
347	450	return $components;
348		}
349
350		/**
351		* Parses the URI authority part.
352		*
353		* @link https://tools.ietf.org/html/rfc3986#section-3.2
354		*
355		* @throws SyntaxError If the port component is invalid
356		*
357		* @return array{user:?string, pass:?string, host:?string, port:?int}
358		*/
359	386	private static function parseAuthority(string $authority): array
360		{
361	386	$components = ['user' => null, 'pass' => null, 'host' => '', 'port' => null];
362	386	if ('' === $authority) {
363	14	return $components;
364		}
365
366	374	$parts = explode('@', $authority, 2);
367	374	if (isset($parts[1])) {
368	94	[$components['user'], $components['pass']] = explode(':', $parts[0], 2) + [1 => null];
369		}
370
371	374	preg_match(self::REGEXP_HOST_PORT, $parts[1] ?? $parts[0], $matches);
372	374	$matches += ['port' => ''];
373
374	374	$components['port'] = self::filterPort($matches['port']);
375	362	$components['host'] = self::filterHost($matches['host']);
376
377	338	return $components;
378		}
379
380		/**
381		* Filter and format the port component.
382		*
383		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
384		*
385		* @throws SyntaxError if the registered name is invalid
386		*/
387	374	private static function filterPort(string $port): ?int
388		{
389	374	if ('' === $port) {
390	264	return null;
391		}
392
393	114	if (1 === preg_match('/^\d*$/', $port)) {
394	102	return (int) $port;
395		}
396
397	12	throw new SyntaxError(sprintf('The port `%s` is invalid', $port));
398		}
399
400		/**
401		* Returns whether a hostname is valid.
402		*
403		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
404		*
405		* @throws SyntaxError if the registered name is invalid
406		*/
407	362	private static function filterHost(string $host): string
408		{
409	362	if ('' === $host) {
410	4	return $host;
411		}
412
413	360	if ('[' !== $host[0] \|\| ']' !== substr($host, -1)) {
414	324	return self::filterRegisteredName($host);
415		}
416
417	36	if (!self::isIpHost(substr($host, 1, -1))) {
418	10	throw new SyntaxError(sprintf('Host `%s` is invalid : the IP host is malformed', $host));
419		}
420
421	26	return $host;
422		}
423
424		/**
425		* Returns whether the host is an IPv4 or a registered named.
426		*
427		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
428		*
429		* @throws SyntaxError if the registered name is invalid
430		* @throws IdnSupportMissing if IDN support or ICU requirement are not available or met.
431		*/
432	324	private static function filterRegisteredName(string $host): string
433		{
434		// @codeCoverageIgnoreStart
435		// added because it is not possible in travis to disabled the ext/intl extension
436		// see travis issue https://github.com/travis-ci/travis-ci/issues/4701
437		static $idn_support = null;
438		$idn_support = $idn_support ?? function_exists('idn_to_ascii') && defined('INTL_IDNA_VARIANT_UTS46');
439		// @codeCoverageIgnoreEnd
440
441	324	$formatted_host = rawurldecode($host);
442	324	if (1 === preg_match(self::REGEXP_REGISTERED_NAME, $formatted_host)) {
443	308	if (false === strpos($formatted_host, 'xn--')) {
444	304	return $host;
445		}
446
447		// @codeCoverageIgnoreStart
448		if (!$idn_support) {
449		throw new IdnSupportMissing(sprintf('the host `%s` could not be processed for IDN. Verify that ext/intl is installed for IDN support and that ICU is at least version 4.6.', $host));
450		}
451		// @codeCoverageIgnoreEnd
452
453	4	$unicode = idn_to_utf8($host, 0, INTL_IDNA_VARIANT_UTS46, $arr);
454	4	if (0 !== $arr['errors']) {
455	2	throw new SyntaxError(sprintf('The host `%s` is invalid : %s', $host, self::getIDNAErrors($arr['errors'])));
456		}
457
458		// @codeCoverageIgnoreStart
459		if (false === $unicode) {
460		throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
461		}
462		// @codeCoverageIgnoreEnd
463
464	2	return $host;
465		}
466
467		//to test IDN host non-ascii characters must be present in the host
468	16	if (1 !== preg_match(self::REGEXP_IDN_PATTERN, $formatted_host)) {
469	2	throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
470		}
471
472		// @codeCoverageIgnoreStart
473		if (!$idn_support) {
474		throw new IdnSupportMissing(sprintf('the host `%s` could not be processed for IDN. Verify that ext/intl is installed for IDN support and that ICU is at least version 4.6.', $host));
475		}
476		// @codeCoverageIgnoreEnd
477
478	14	$retval = idn_to_ascii($formatted_host, 0, INTL_IDNA_VARIANT_UTS46, $arr);
479
480	14	if ([] === $arr) {
481	2	throw new SyntaxError(sprintf('Host `%s` is not a valid IDN host', $host));
482		}
483
484	12	if (0 !== $arr['errors']) {
485	4	throw new SyntaxError(sprintf('Host `%s` is not a valid IDN host : %s', $host, self::getIDNAErrors($arr['errors'])));
486		}
487
488		// @codeCoverageIgnoreStart
489		if (false === $retval) {
490		throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
491		}
492		// @codeCoverageIgnoreEnd
493
494	8	if (false !== strpos($retval, '%')) {
495	4	throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
496		}
497
498	4	return $host;
499		}
500
501		/**
502		* Retrieves and format IDNA conversion error message.
503		*
504		* @link http://icu-project.org/apiref/icu4j/com/ibm/icu/text/IDNA.Error.html
505		*/
506	6	private static function getIDNAErrors(int $error_byte): string
507		{
508		/**
509		* IDNA errors.
510		*/
511	6	static $idn_errors = [
512		IDNA_ERROR_EMPTY_LABEL => 'a non-final domain name label (or the whole domain name) is empty',
513		IDNA_ERROR_LABEL_TOO_LONG => 'a domain name label is longer than 63 bytes',
514		IDNA_ERROR_DOMAIN_NAME_TOO_LONG => 'a domain name is longer than 255 bytes in its storage form',
515		IDNA_ERROR_LEADING_HYPHEN => 'a label starts with a hyphen-minus ("-")',
516		IDNA_ERROR_TRAILING_HYPHEN => 'a label ends with a hyphen-minus ("-")',
517		IDNA_ERROR_HYPHEN_3_4 => 'a label contains hyphen-minus ("-") in the third and fourth positions',
518		IDNA_ERROR_LEADING_COMBINING_MARK => 'a label starts with a combining mark',
519		IDNA_ERROR_DISALLOWED => 'a label or domain name contains disallowed characters',
520		IDNA_ERROR_PUNYCODE => 'a label starts with "xn--" but does not contain valid Punycode',
521		IDNA_ERROR_LABEL_HAS_DOT => 'a label contains a dot=full stop',
522		IDNA_ERROR_INVALID_ACE_LABEL => 'An ACE label does not contain a valid label string',
523		IDNA_ERROR_BIDI => 'a label does not meet the IDNA BiDi requirements (for right-to-left characters)',
524		IDNA_ERROR_CONTEXTJ => 'a label does not meet the IDNA CONTEXTJ requirements',
525		];
526
527	6	$res = [];
528	6	foreach ($idn_errors as $error => $reason) {
529	6	if ($error === ($error_byte & $error)) {
530	6	$res[] = $reason;
531		}
532		}
533
534	6	return [] === $res ? 'Unknown IDNA conversion error.' : implode(', ', $res).'.';
535		}
536
537		/**
538		* Validates a IPv6/IPvfuture host.
539		*
540		* @link https://tools.ietf.org/html/rfc3986#section-3.2.2
541		* @link https://tools.ietf.org/html/rfc6874#section-2
542		* @link https://tools.ietf.org/html/rfc6874#section-4
543		*/
544	36	private static function isIpHost(string $ip_host): bool
545		{
546	36	if (false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)) {
547	20	return true;
548		}
549
550	16	if (1 === preg_match(self::REGEXP_IP_FUTURE, $ip_host, $matches)) {
551	4	return !in_array($matches['version'], ['4', '6'], true);
552		}
553
554	12	$pos = strpos($ip_host, '%');
555	12	if (false === $pos \|\| 1 === preg_match(
556	8	self::REGEXP_INVALID_HOST_CHARS,
557	12	rawurldecode(substr($ip_host, $pos))
558		)) {
559	6	return false;
560		}
561
562	6	$ip_host = substr($ip_host, 0, $pos);
563
564	6	return false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)
565	6	&& 0 === strpos((string) inet_pton($ip_host), self::ZONE_ID_ADDRESS_BLOCK);
566		}
567		}
568

thephpleague / uri

UriString::parseAuthority() A last analyzed 2020-08-13 17:22 UTC

Complexity

Size

Duplication

Code Coverage

Importance

Duplication Side-by-Side

Filter issues like

UriString::parseAuthority() A
last analyzed 2020-08-13 17:22 UTC