MediaWikiTitleCodec.php

Go to the documentation of this file.

<?php
use MediaWiki\Linker\LinkTarget;

class MediaWikiTitleCodec implements TitleFormatter, TitleParser {
    protected $language;

    protected $genderCache;

    protected $localInterwikis;

    public function __construct( Language $language, GenderCache $genderCache,
        $localInterwikis = []
    ) {
        $this->language = $language;
        $this->genderCache = $genderCache;
        $this->localInterwikis = (array)$localInterwikis;
    }

    public function getNamespaceName( $namespace, $text ) {
        if ( $this->language->needsGenderDistinction() &&
            MWNamespace::hasGenderDistinction( $namespace )
        ) {

            // NOTE: we are assuming here that the title text is a user name!
            $gender = $this->genderCache->getGenderOf( $text, __METHOD__ );
            $name = $this->language->getGenderNsText( $namespace, $gender );
        } else {
            $name = $this->language->getNsText( $namespace );
        }

        if ( $name === false ) {
            throw new InvalidArgumentException( 'Unknown namespace ID: ' . $namespace );
        }

        return $name;
    }

    public function formatTitle( $namespace, $text, $fragment = '', $interwiki = '' ) {
        if ( $namespace !== false ) {
            // Try to get a namespace name, but fallback
            // to empty string if it doesn't exist
            try {
                $nsName = $this->getNamespaceName( $namespace, $text );
            } catch ( InvalidArgumentException $e ) {
                $nsName = '';
            }

            if ( $namespace !== 0 ) {
                $text = $nsName . ':' . $text;
            }
        }

        if ( $fragment !== '' ) {
            $text = $text . '#' . $fragment;
        }

        if ( $interwiki !== '' ) {
            $text = $interwiki . ':' . $text;
        }

        $text = str_replace( '_', ' ', $text );

        return $text;
    }

    public function parseTitle( $text, $defaultNamespace ) {
        // NOTE: this is an ugly cludge that allows this class to share the
        // code for parsing with the old Title class. The parser code should
        // be refactored to avoid this.
        $parts = $this->splitTitleString( $text, $defaultNamespace );

        // Relative fragment links are not supported by TitleValue
        if ( $parts['dbkey'] === '' ) {
            throw new MalformedTitleException( 'title-invalid-empty', $text );
        }

        return new TitleValue(
            $parts['namespace'],
            $parts['dbkey'],
            $parts['fragment'],
            $parts['interwiki']
        );
    }

    public function getText( LinkTarget $title ) {
        return $this->formatTitle( false, $title->getText(), '' );
    }

    public function getPrefixedText( LinkTarget $title ) {
        return $this->formatTitle(
            $title->getNamespace(),
            $title->getText(),
            '',
            $title->getInterwiki()
        );
    }

    public function getPrefixedDBkey( LinkTarget $target ) {
        $key = '';
        if ( $target->isExternal() ) {
            $key .= $target->getInterwiki() . ':';
        }
        // Try to get a namespace name, but fallback
        // to empty string if it doesn't exist
        try {
            $nsName = $this->getNamespaceName(
                $target->getNamespace(),
                $target->getText()
            );
        } catch ( InvalidArgumentException $e ) {
            $nsName = '';
        }

        if ( $target->getNamespace() !== 0 ) {
            $key .= $nsName . ':';
        }

        $key .= $target->getText();

        return strtr( $key, ' ', '_' );
    }

    public function getFullText( LinkTarget $title ) {
        return $this->formatTitle(
            $title->getNamespace(),
            $title->getText(),
            $title->getFragment(),
            $title->getInterwiki()
        );
    }

    public function splitTitleString( $text, $defaultNamespace = NS_MAIN ) {
        $dbkey = str_replace( ' ', '_', $text );

        # Initialisation
        $parts = [
            'interwiki' => '',
            'local_interwiki' => false,
            'fragment' => '',
            'namespace' => $defaultNamespace,
            'dbkey' => $dbkey,
            'user_case_dbkey' => $dbkey,
        ];

        # Strip Unicode bidi override characters.
        # Sometimes they slip into cut-n-pasted page titles, where the
        # override chars get included in list displays.
        $dbkey = preg_replace( '/\xE2\x80[\x8E\x8F\xAA-\xAE]/S', '', $dbkey );

        # Clean up whitespace
        # Note: use of the /u option on preg_replace here will cause
        # input with invalid UTF-8 sequences to be nullified out in PHP 5.2.x,
        # conveniently disabling them.
        $dbkey = preg_replace(
            '/[ _\xA0\x{1680}\x{180E}\x{2000}-\x{200A}\x{2028}\x{2029}\x{202F}\x{205F}\x{3000}]+/u',
            '_',
            $dbkey
        );
        $dbkey = trim( $dbkey, '_' );

        if ( strpos( $dbkey, UtfNormal\Constants::UTF8_REPLACEMENT ) !== false ) {
            # Contained illegal UTF-8 sequences or forbidden Unicode chars.
            throw new MalformedTitleException( 'title-invalid-utf8', $text );
        }

        $parts['dbkey'] = $dbkey;

        # Initial colon indicates main namespace rather than specified default
        # but should not create invalid {ns,title} pairs such as {0,Project:Foo}
        if ( $dbkey !== '' && ':' == $dbkey[0] ) {
            $parts['namespace'] = NS_MAIN;
            $dbkey = substr( $dbkey, 1 ); # remove the colon but continue processing
            $dbkey = trim( $dbkey, '_' ); # remove any subsequent whitespace
        }

        if ( $dbkey == '' ) {
            throw new MalformedTitleException( 'title-invalid-empty', $text );
        }

        # Namespace or interwiki prefix
        $prefixRegexp = "/^(.+?)_*:_*(.*)$/S";
        do {
            $m = [];
            if ( preg_match( $prefixRegexp, $dbkey, $m ) ) {
                $p = $m[1];
                $ns = $this->language->getNsIndex( $p );
                if ( $ns !== false ) {
                    # Ordinary namespace
                    $dbkey = $m[2];
                    $parts['namespace'] = $ns;
                    # For Talk:X pages, check if X has a "namespace" prefix
                    if ( $ns == NS_TALK && preg_match( $prefixRegexp, $dbkey, $x ) ) {
                        if ( $this->language->getNsIndex( $x[1] ) ) {
                            # Disallow Talk:File:x type titles...
                            throw new MalformedTitleException( 'title-invalid-talk-namespace', $text );
                        } elseif ( Interwiki::isValidInterwiki( $x[1] ) ) {
                            // TODO: get rid of global state!
                            # Disallow Talk:Interwiki:x type titles...
                            throw new MalformedTitleException( 'title-invalid-talk-namespace', $text );
                        }
                    }
                } elseif ( Interwiki::isValidInterwiki( $p ) ) {
                    # Interwiki link
                    $dbkey = $m[2];
                    $parts['interwiki'] = $this->language->lc( $p );

                    # Redundant interwiki prefix to the local wiki
                    foreach ( $this->localInterwikis as $localIW ) {
                        if ( 0 == strcasecmp( $parts['interwiki'], $localIW ) ) {
                            if ( $dbkey == '' ) {
                                # Empty self-links should point to the Main Page, to ensure
                                # compatibility with cross-wiki transclusions and the like.
                                $mainPage = Title::newMainPage();
                                return [
                                    'interwiki' => $mainPage->getInterwiki(),
                                    'local_interwiki' => true,
                                    'fragment' => $mainPage->getFragment(),
                                    'namespace' => $mainPage->getNamespace(),
                                    'dbkey' => $mainPage->getDBkey(),
                                    'user_case_dbkey' => $mainPage->getUserCaseDBKey()
                                ];
                            }
                            $parts['interwiki'] = '';
                            # local interwikis should behave like initial-colon links
                            $parts['local_interwiki'] = true;

                            # Do another namespace split...
                            continue 2;
                        }
                    }

                    # If there's an initial colon after the interwiki, that also
                    # resets the default namespace
                    if ( $dbkey !== '' && $dbkey[0] == ':' ) {
                        $parts['namespace'] = NS_MAIN;
                        $dbkey = substr( $dbkey, 1 );
                    }
                }
                # If there's no recognized interwiki or namespace,
                # then let the colon expression be part of the title.
            }
            break;
        } while ( true );

        $fragment = strstr( $dbkey, '#' );
        if ( false !== $fragment ) {
            $parts['fragment'] = str_replace( '_', ' ', substr( $fragment, 1 ) );
            $dbkey = substr( $dbkey, 0, strlen( $dbkey ) - strlen( $fragment ) );
            # remove whitespace again: prevents "Foo_bar_#"
            # becoming "Foo_bar_"
            $dbkey = preg_replace( '/_*$/', '', $dbkey );
        }

        # Reject illegal characters.
        $rxTc = self::getTitleInvalidRegex();
        $matches = [];
        if ( preg_match( $rxTc, $dbkey, $matches ) ) {
            throw new MalformedTitleException( 'title-invalid-characters', $text, [ $matches[0] ] );
        }

        # Pages with "/./" or "/../" appearing in the URLs will often be un-
        # reachable due to the way web browsers deal with 'relative' URLs.
        # Also, they conflict with subpage syntax.  Forbid them explicitly.
        if (
            strpos( $dbkey, '.' ) !== false &&
            (
                $dbkey === '.' || $dbkey === '..' ||
                strpos( $dbkey, './' ) === 0 ||
                strpos( $dbkey, '../' ) === 0 ||
                strpos( $dbkey, '/./' ) !== false ||
                strpos( $dbkey, '/../' ) !== false ||
                substr( $dbkey, -2 ) == '/.' ||
                substr( $dbkey, -3 ) == '/..'
            )
        ) {
            throw new MalformedTitleException( 'title-invalid-relative', $text );
        }

        # Magic tilde sequences? Nu-uh!
        if ( strpos( $dbkey, '~~~' ) !== false ) {
            throw new MalformedTitleException( 'title-invalid-magic-tilde', $text );
        }

        # Limit the size of titles to 255 bytes. This is typically the size of the
        # underlying database field. We make an exception for special pages, which
        # don't need to be stored in the database, and may edge over 255 bytes due
        # to subpage syntax for long titles, e.g. [[Special:Block/Long name]]
        $maxLength = ( $parts['namespace'] != NS_SPECIAL ) ? 255 : 512;
        if ( strlen( $dbkey ) > $maxLength ) {
            throw new MalformedTitleException( 'title-invalid-too-long', $text,
                [ Message::numParam( $maxLength ) ] );
        }

        # Normally, all wiki links are forced to have an initial capital letter so [[foo]]
        # and [[Foo]] point to the same place.  Don't force it for interwikis, since the
        # other site might be case-sensitive.
        $parts['user_case_dbkey'] = $dbkey;
        if ( $parts['interwiki'] === '' ) {
            $dbkey = Title::capitalize( $dbkey, $parts['namespace'] );
        }

        # Can't make a link to a namespace alone... "empty" local links can only be
        # self-links with a fragment identifier.
        if ( $dbkey == '' && $parts['interwiki'] === '' ) {
            if ( $parts['namespace'] != NS_MAIN ) {
                throw new MalformedTitleException( 'title-invalid-empty', $text );
            }
        }

        // Allow IPv6 usernames to start with '::' by canonicalizing IPv6 titles.
        // IP names are not allowed for accounts, and can only be referring to
        // edits from the IP. Given '::' abbreviations and caps/lowercaps,
        // there are numerous ways to present the same IP. Having sp:contribs scan
        // them all is silly and having some show the edits and others not is
        // inconsistent. Same for talk/userpages. Keep them normalized instead.
        if ( $parts['namespace'] == NS_USER || $parts['namespace'] == NS_USER_TALK ) {
            $dbkey = IP::sanitizeIP( $dbkey );
        }

        // Any remaining initial :s are illegal.
        if ( $dbkey !== '' && ':' == $dbkey[0] ) {
            throw new MalformedTitleException( 'title-invalid-leading-colon', $text );
        }

        # Fill fields
        $parts['dbkey'] = $dbkey;

        return $parts;
    }

    public static function getTitleInvalidRegex() {
        static $rxTc = false;
        if ( !$rxTc ) {
            # Matching titles will be held as illegal.
            $rxTc = '/' .
                # Any character not allowed is forbidden...
                '[^' . Title::legalChars() . ']' .
                # URL percent encoding sequences interfere with the ability
                # to round-trip titles -- you can't link to them consistently.
                '|%[0-9A-Fa-f]{2}' .
                # XML/HTML character references produce similar issues.
                '|&[A-Za-z0-9\x80-\xff]+;' .
                '|&#[0-9]+;' .
                '|&#x[0-9A-Fa-f]+;' .
                '/S';
        }

        return $rxTc;
    }
}