UriString::parse() - Code Metrics - Inspection of "Moving code from uri-schemes to uri package" - thephpleague/uri - Measure and Improve Code Quality continuously with Scrutinizer

Completed

Pull Request — master (#127)

by ignace nyamagana

created 2019-05-19 06:07 UTC

UriString::parse() C

↳ Parent: UriString

Complexity

Conditions	15
Paths	8

Size

Total Lines	59
Code Lines	33

Duplication

Lines	0
Ratio	0 %

Code Coverage

Tests	34
CRAP Score	15

Importance

Changes

Metric	Value
cc	15
eloc	33
nc	8
nop	1
dl	0
loc	59
ccs	34
cts	34
cp	1
crap	15
rs	5.9166
c	0
b	0
f	0

How to fix Long Method Complexity

<?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\Exception\IdnSupportMissing;
use League\Uri\Exception\SyntaxError;
use TypeError;
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_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.
 *
 * @see     https://tools.ietf.org/html/rfc3986
 * @package League\Uri
 * @author  Ignace Nyamagana Butera <[email protected]>
 * @since   0.1.0
 */
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]/';

    /**
     * @see https://tools.ietf.org/html/rfc3986#appendix-B
     *
     * RFC3986 regular expression URI splitter
     */
    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';

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

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

    /**
     * @see https://tools.ietf.org/html/rfc3986#section-3.2.2
     *
     * General registered name regular expression
     */
    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';

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

    /**
     * @see https://tools.ietf.org/html/rfc3986#section-3.3
     *
     * invalid path for URI without scheme and authority regular expression
     */
    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.
     *
     * @see https://tools.ietf.org/html/rfc3986#section-5.3
     * @see https://tools.ietf.org/html/rfc3986#section-7.5
     */
    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>
     *
     * @see https://tools.ietf.org/html/rfc3986
     *
     * @param mixed $uri any scalar or stringable object
     *
     * @throws SyntaxError if the URI contains invalid characters
     * @throws SyntaxError if the URI contains an invalid scheme
     * @throws SyntaxError if the URI contains an invalid path
     */
    public static function parse($uri): array
    {
        if (!is_scalar($uri) && !method_exists($uri, '__toString')) {
            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[$uri])) {
            return array_merge(self::URI_COMPONENTS, self::URI_SCHORTCUTS[$uri]);
        }

        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));
        }

        return 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'],
            ]
        );
    }

    /**
     * Parses the URI authority part.
     *
     * @see https://tools.ietf.org/html/rfc3986#section-3.2
     *
     * @throws SyntaxError If the port component is invalid
     */
    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.
     *
     * @see 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.
     *
     * @see 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.
     *
     * @see http://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

        $formatted_host = rawurldecode(strtolower($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 (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.
     *
     * @see 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.
     *
     * @see http://tools.ietf.org/html/rfc3986#section-3.2.2
     * @see http://tools.ietf.org/html/rfc6874#section-2
     * @see http://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\Exception\IdnSupportMissing;
17		use League\Uri\Exception\SyntaxError;
18		use TypeError;
19		use function array_merge;
20		use function defined;
21		use function explode;
22		use function filter_var;
23		use function function_exists;
24		use function gettype;
25		use function idn_to_ascii;
26		use function implode;
27		use function inet_pton;
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		* @see https://tools.ietf.org/html/rfc3986
56		* @package League\Uri
57		* @author Ignace Nyamagana Butera <[email protected]>
58		* @since 0.1.0
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		* @see https://tools.ietf.org/html/rfc3986#appendix-B
89		*
90		* RFC3986 regular expression URI splitter
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		* @see https://tools.ietf.org/html/rfc3986#section-3.1
102		*
103		* URI scheme regular expresssion
104		*/
105		private const REGEXP_URI_SCHEME = '/^([a-z][a-z\d\+\.\-]*)?$/i';
106
107		/**
108		* @see https://tools.ietf.org/html/rfc3986#section-3.2.2
109		*
110		* IPvFuture regular expression
111		*/
112		private const REGEXP_IP_FUTURE = '/^
113		v(?<version>[A-F0-9])+\.
114		(?:
115		(?<unreserved>[a-z0-9_~\-\.])\|
116		(?<sub_delims>[!$&\'()*+,;=:]) # also include the : character
117		)+
118		$/ix';
119
120		/**
121		* @see https://tools.ietf.org/html/rfc3986#section-3.2.2
122		*
123		* General registered name regular expression
124		*/
125		private const REGEXP_REGISTERED_NAME = '/(?(DEFINE)
126		(?<unreserved>[a-z0-9_~\-]) # . is missing as it is used to separate labels
127		(?<sub_delims>[!$&\'()*+,;=])
128		(?<encoded>%[A-F0-9]{2})
129		(?<reg_name>(?:(?&unreserved)\|(?&sub_delims)\|(?&encoded))*)
130		)
131		^(?:(?&reg_name)\.)*(?&reg_name)\.?$/ix';
132
133		/**
134		* @see https://tools.ietf.org/html/rfc3986#section-3.2.2
135		*
136		* invalid characters in host regular expression
137		*/
138		private const REGEXP_INVALID_HOST_CHARS = '/
139		[:\/?#\[\]@ ] # gen-delims characters as well as the space character
140		/ix';
141
142		/**
143		* @see https://tools.ietf.org/html/rfc3986#section-3.3
144		*
145		* invalid path for URI without scheme and authority regular expression
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		* @see https://tools.ietf.org/html/rfc3986#section-5.3
173		* @see https://tools.ietf.org/html/rfc3986#section-7.5
174		*/
175	94	public static function build(array $components): string
176		{
177	94	$result = $components['path'] ?? '';
178	94	if (isset($components['query'])) {
179	50	$result .= '?'.$components['query'];
180		}
181
182	94	if (isset($components['fragment'])) {
183	56	$result .= '#'.$components['fragment'];
184		}
185
186	94	$scheme = null;
187	94	if (isset($components['scheme'])) {
188	52	$scheme = $components['scheme'].':';
189		}
190
191	94	if (!isset($components['host'])) {
192	42	return $scheme.$result;
193		}
194
195	52	$scheme .= '//';
196	52	$authority = $components['host'];
197	52	if (isset($components['port'])) {
198	14	$authority .= ':'.$components['port'];
199		}
200
201	52	if (!isset($components['user'])) {
202	34	return $scheme.$authority.$result;
203		}
204
205	18	$authority = '@'.$authority;
206	18	if (!isset($components['pass'])) {
207	4	return $scheme.$components['user'].$authority.$result;
208		}
209
210	14	return $scheme.$components['user'].':'.$components['pass'].$authority.$result;
211		}
212
213		/**
214		* Parse an URI string into its components.
215		*
216		* This method parses a URI and returns an associative array containing any
217		* of the various components of the URI that are present.
218		*
219		* <code>
220		* $components = (new Parser())->parse('http://[email protected]:42?query#');
221		* var_export($components);
222		* //will display
223		* array(
224		* 'scheme' => 'http', // the URI scheme component
225		* 'user' => 'foo', // the URI user component
226		* 'pass' => null, // the URI pass component
227		* 'host' => 'test.example.com', // the URI host component
228		* 'port' => 42, // the URI port component
229		* 'path' => '', // the URI path component
230		* 'query' => 'query', // the URI query component
231		* 'fragment' => '', // the URI fragment component
232		* );
233		* </code>
234		*
235		* The returned array is similar to PHP's parse_url return value with the following
236		* differences:
237		*
238		* <ul>
239		* <li>All components are always present in the returned array</li>
240		* <li>Empty and undefined component are treated differently. And empty component is
241		* set to the empty string while an undefined component is set to the `null` value.</li>
242		* <li>The path component is never undefined</li>
243		* <li>The method parses the URI following the RFC3986 rules but you are still
244		* required to validate the returned components against its related scheme specific rules.</li>
245		* </ul>
246		*
247		* @see https://tools.ietf.org/html/rfc3986
248		*
249		* @param mixed $uri any scalar or stringable object
250		*
251		* @throws SyntaxError if the URI contains invalid characters
252		* @throws SyntaxError if the URI contains an invalid scheme
253		* @throws SyntaxError if the URI contains an invalid path
254		*/
255	526	public static function parse($uri): array
256		{
257	526	if (!is_scalar($uri) && !method_exists($uri, '__toString')) {
258	2	throw new TypeError(sprintf('The uri must be a scalar or a stringable object `%s` given', gettype($uri)));
259		}
260
261	524	$uri = (string) $uri;
262
263	524	if (isset(self::URI_SCHORTCUTS[$uri])) {
264	44	return array_merge(self::URI_COMPONENTS, self::URI_SCHORTCUTS[$uri]);
265		}
266
267	494	if (1 === preg_match(self::REGEXP_INVALID_URI_CHARS, $uri)) {
268	2	throw new SyntaxError(sprintf('The uri `%s` contains invalid characters', $uri));
269		}
270
271		//if the first character is a known URI delimiter parsing can be simplified
272	492	$first_char = $uri[0];
273
274		//The URI is made of the fragment only
275	492	if ('#' === $first_char) {
276	6	[, $fragment] = explode('#', $uri, 2);
277	6	$components = self::URI_COMPONENTS;
278	6	$components['fragment'] = $fragment;
279
280	6	return $components;
281		}
282
283		//The URI is made of the query and fragment
284	488	if ('?' === $first_char) {
285	6	[, $partial] = explode('?', $uri, 2);
286	6	[$query, $fragment] = explode('#', $partial, 2) + [1 => null];
287	6	$components = self::URI_COMPONENTS;
288	6	$components['query'] = $query;
289	6	$components['fragment'] = $fragment;
290
291	6	return $components;
292		}
293
294		//use RFC3986 URI regexp to split the URI
295	484	preg_match(self::REGEXP_URI_PARTS, $uri, $parts);
296	484	$parts += ['query' => '', 'fragment' => ''];
297
298	484	if (':' === $parts['scheme'] \|\| 1 !== preg_match(self::REGEXP_URI_SCHEME, $parts['scontent'])) {
299	6	throw new SyntaxError(sprintf('The uri `%s` contains an invalid scheme', $uri));
300		}
301
302	478	if ('' === $parts['scheme'].$parts['authority'] && 1 === preg_match(self::REGEXP_INVALID_PATH, $parts['path'])) {
303	2	throw new SyntaxError(sprintf('The uri `%s` contains an invalid path.', $uri));
304		}
305
306	476	return array_merge(
307	476	self::URI_COMPONENTS,
308	476	'' === $parts['authority'] ? [] : self::parseAuthority($parts['acontent']),
309		[
310	442	'path' => $parts['path'],
311	442	'scheme' => '' === $parts['scheme'] ? null : $parts['scontent'],
312	442	'query' => '' === $parts['query'] ? null : $parts['qcontent'],
313	442	'fragment' => '' === $parts['fragment'] ? null : $parts['fcontent'],
314		]
315		);
316		}
317
318		/**
319		* Parses the URI authority part.
320		*
321		* @see https://tools.ietf.org/html/rfc3986#section-3.2
322		*
323		* @throws SyntaxError If the port component is invalid
324		*/
325	376	private static function parseAuthority(string $authority): array
326		{
327	376	$components = ['user' => null, 'pass' => null, 'host' => '', 'port' => null];
328	376	if ('' === $authority) {
329	14	return $components;
330		}
331
332	364	$parts = explode('@', $authority, 2);
333	364	if (isset($parts[1])) {
334	88	[$components['user'], $components['pass']] = explode(':', $parts[0], 2) + [1 => null];
335		}
336
337	364	preg_match(self::REGEXP_HOST_PORT, $parts[1] ?? $parts[0], $matches);
338	364	$matches += ['port' => ''];
339
340	364	$components['port'] = self::filterPort($matches['port']);
341	352	$components['host'] = self::filterHost($matches['host']);
342
343	330	return $components;
344		}
345
346		/**
347		* Filter and format the port component.
348		*
349		* @see https://tools.ietf.org/html/rfc3986#section-3.2.2
350		*
351		* @throws SyntaxError if the registered name is invalid
352		*
353		*/
354	364	private static function filterPort(string $port): ?int
355		{
356	364	if ('' === $port) {
357	260	return null;
358		}
359
360	108	if (1 === preg_match('/^\d*$/', $port)) {
361	96	return (int) $port;
362		}
363
364	12	throw new SyntaxError(sprintf('The port `%s` is invalid', $port));
365		}
366
367		/**
368		* Returns whether a hostname is valid.
369		*
370		* @see https://tools.ietf.org/html/rfc3986#section-3.2.2
371		*
372		* @throws SyntaxError if the registered name is invalid
373		*/
374	352	private static function filterHost(string $host): string
375		{
376	352	if ('' === $host) {
377	4	return $host;
378		}
379
380	350	if ('[' !== $host[0] \|\| ']' !== substr($host, -1)) {
381	314	return self::filterRegisteredName($host);
382		}
383
384	36	if (!self::isIpHost(substr($host, 1, -1))) {
385	10	throw new SyntaxError(sprintf('Host `%s` is invalid : the IP host is malformed', $host));
386		}
387
388	26	return $host;
389		}
390
391		/**
392		* Returns whether the host is an IPv4 or a registered named.
393		*
394		* @see http://tools.ietf.org/html/rfc3986#section-3.2.2
395		*
396		* @throws SyntaxError if the registered name is invalid
397		* @throws IdnSupportMissing if IDN support or ICU requirement are not available or met.
398		*/
399	314	private static function filterRegisteredName(string $host): string
400		{
401		// @codeCoverageIgnoreStart
402		// added because it is not possible in travis to disabled the ext/intl extension
403		// see travis issue https://github.com/travis-ci/travis-ci/issues/4701
404		static $idn_support = null;
405		$idn_support = $idn_support ?? function_exists('idn_to_ascii') && defined('INTL_IDNA_VARIANT_UTS46');
406		// @codeCoverageIgnoreEnd
407
408	314	$formatted_host = rawurldecode(strtolower($host));
409	314	if (1 === preg_match(self::REGEXP_REGISTERED_NAME, $formatted_host)) {
410	300	if (false === strpos($formatted_host, 'xn--')) {
411	296	return $host;
412		}
413
414		// @codeCoverageIgnoreStart
415		if (!$idn_support) {
416		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));
417		}
418		// @codeCoverageIgnoreEnd
419
420	4	$unicode = idn_to_utf8($host, 0, INTL_IDNA_VARIANT_UTS46, $arr);
421	4	if (0 !== $arr['errors']) {
422	2	throw new SyntaxError(sprintf('The host `%s` is invalid : %s', $host, self::getIDNAErrors($arr['errors'])));
423		}
424
425		// @codeCoverageIgnoreStart
426		if (false === $unicode) {
427		throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
428		}
429		// @codeCoverageIgnoreEnd
430
431	2	return $host;
432		}
433
434		//to test IDN host non-ascii characters must be present in the host
435	14	if (1 !== preg_match(self::REGEXP_IDN_PATTERN, $formatted_host)) {
436	2	throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
437		}
438
439		// @codeCoverageIgnoreStart
440		if (!$idn_support) {
441		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));
442		}
443		// @codeCoverageIgnoreEnd
444
445	12	$retval = idn_to_ascii($formatted_host, 0, INTL_IDNA_VARIANT_UTS46, $arr);
446	12	if (0 !== $arr['errors']) {
447	4	throw new SyntaxError(sprintf('Host `%s` is not a valid IDN host : %s', $host, self::getIDNAErrors($arr['errors'])));
448		}
449
450		// @codeCoverageIgnoreStart
451		if (false === $retval) {
452		throw new IdnSupportMissing(sprintf('The Intl extension is misconfigured for %s, please correct this issue before proceeding.', PHP_OS));
453		}
454		// @codeCoverageIgnoreEnd
455
456	8	if (false !== strpos($retval, '%')) {
457	4	throw new SyntaxError(sprintf('Host `%s` is invalid : the host is not a valid registered name', $host));
458		}
459
460	4	return $host;
461		}
462
463		/**
464		* Retrieves and format IDNA conversion error message.
465		*
466		* @see http://icu-project.org/apiref/icu4j/com/ibm/icu/text/IDNA.Error.html
467		*/
468	6	private static function getIDNAErrors(int $error_byte): string
469		{
470		/**
471		* IDNA errors.
472		*/
473	6	static $idn_errors = [
474		IDNA_ERROR_EMPTY_LABEL => 'a non-final domain name label (or the whole domain name) is empty',
475		IDNA_ERROR_LABEL_TOO_LONG => 'a domain name label is longer than 63 bytes',
476		IDNA_ERROR_DOMAIN_NAME_TOO_LONG => 'a domain name is longer than 255 bytes in its storage form',
477		IDNA_ERROR_LEADING_HYPHEN => 'a label starts with a hyphen-minus ("-")',
478		IDNA_ERROR_TRAILING_HYPHEN => 'a label ends with a hyphen-minus ("-")',
479		IDNA_ERROR_HYPHEN_3_4 => 'a label contains hyphen-minus ("-") in the third and fourth positions',
480		IDNA_ERROR_LEADING_COMBINING_MARK => 'a label starts with a combining mark',
481		IDNA_ERROR_DISALLOWED => 'a label or domain name contains disallowed characters',
482		IDNA_ERROR_PUNYCODE => 'a label starts with "xn--" but does not contain valid Punycode',
483		IDNA_ERROR_LABEL_HAS_DOT => 'a label contains a dot=full stop',
484		IDNA_ERROR_INVALID_ACE_LABEL => 'An ACE label does not contain a valid label string',
485		IDNA_ERROR_BIDI => 'a label does not meet the IDNA BiDi requirements (for right-to-left characters)',
486		IDNA_ERROR_CONTEXTJ => 'a label does not meet the IDNA CONTEXTJ requirements',
487		];
488
489	6	$res = [];
490	6	foreach ($idn_errors as $error => $reason) {
491	6	if ($error === ($error_byte & $error)) {
492	6	$res[] = $reason;
493		}
494		}
495
496	6	return [] === $res ? 'Unknown IDNA conversion error.' : implode(', ', $res).'.';
497		}
498
499		/**
500		* Validates a IPv6/IPvfuture host.
501		*
502		* @see http://tools.ietf.org/html/rfc3986#section-3.2.2
503		* @see http://tools.ietf.org/html/rfc6874#section-2
504		* @see http://tools.ietf.org/html/rfc6874#section-4
505		*/
506	36	private static function isIpHost(string $ip_host): bool
507		{
508	36	if (false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)) {
509	20	return true;
510		}
511
512	16	if (1 === preg_match(self::REGEXP_IP_FUTURE, $ip_host, $matches)) {
513	4	return !in_array($matches['version'], ['4', '6'], true);
514		}
515
516	12	$pos = strpos($ip_host, '%');
517	12	if (false === $pos \|\| 1 === preg_match(
518	8	self::REGEXP_INVALID_HOST_CHARS,
519	12	rawurldecode(substr($ip_host, $pos))
520		)) {
521	6	return false;
522		}
523
524	6	$ip_host = substr($ip_host, 0, $pos);
525
526	6	return false !== filter_var($ip_host, FILTER_VALIDATE_IP, FILTER_FLAG_IPV6)
527	6	&& 0 === strpos((string) inet_pton($ip_host), self::ZONE_ID_ADDRESS_BLOCK);
528		}
529		}
530

thephpleague / uri

Pull Request — master (#127)

UriString::parse() C

Complexity

Size

Duplication

Code Coverage

Importance

How to fix Long Method Complexity

Long Method

Duplication Side-by-Side

Filter issues like