From: Aaron Schulz Date: Thu, 22 Sep 2016 04:57:13 +0000 (-0700) Subject: Move MimeMagic code to libs/mime/MimeAnalyzer.php X-Git-Tag: 1.31.0-rc.0~5109^2 X-Git-Url: http://git.cyclocoop.org/%22%20.%20generer_url_ecrire%28%22suivi_revisions%22%29%20.%20%22?a=commitdiff_plain;h=0d87d67d61520d34064f2eaae0f20b8ef11d4c59;p=lhc%2Fweb%2Fwiklou.git Move MimeMagic code to libs/mime/MimeAnalyzer.php * The later resides in /libs with related files. * Explose MimeAnalyzer as a service. * Keep MimeMagic::singleton() as a b/c alias. * MimeMagic::applyDefaultConfig() will bootstrap the service with all of the old config, extension hook handler, and detector command shell-out behavior. Change-Id: Ie2695a52e7a3bcfda9f7fa83659a9ff31b372bc3 --- diff --git a/autoload.php b/autoload.php index 748d954e30..49a9bd4fc8 100644 --- a/autoload.php +++ b/autoload.php @@ -586,7 +586,7 @@ $wgAutoloadLocalClasses = [ 'IContextSource' => __DIR__ . '/includes/context/IContextSource.php', 'IDBAccessObject' => __DIR__ . '/includes/dao/IDBAccessObject.php', 'IDatabase' => __DIR__ . '/includes/libs/rdbms/database/IDatabase.php', - 'IEContentAnalyzer' => __DIR__ . '/includes/libs/IEContentAnalyzer.php', + 'IEContentAnalyzer' => __DIR__ . '/includes/libs/mime/IEContentAnalyzer.php', 'IEUrlExtension' => __DIR__ . '/includes/libs/IEUrlExtension.php', 'IExpiringStore' => __DIR__ . '/includes/libs/objectcache/IExpiringStore.php', 'IJobSpecification' => __DIR__ . '/includes/jobqueue/JobSpecification.php', @@ -943,6 +943,7 @@ $wgAutoloadLocalClasses = [ 'MessageSpecifier' => __DIR__ . '/includes/libs/MessageSpecifier.php', 'MigrateFileRepoLayout' => __DIR__ . '/maintenance/migrateFileRepoLayout.php', 'MigrateUserGroup' => __DIR__ . '/maintenance/migrateUserGroup.php', + 'MimeAnalyzer' => __DIR__ . '/includes/libs/mime/MimeAnalyzer.php', 'MimeMagic' => __DIR__ . '/includes/MimeMagic.php', 'MinifyScript' => __DIR__ . '/maintenance/minify.php', 'MostcategoriesPage' => __DIR__ . '/includes/specials/SpecialMostcategories.php', @@ -1580,7 +1581,7 @@ $wgAutoloadLocalClasses = [ 'XmlDumpWriter' => __DIR__ . '/includes/export/XmlDumpWriter.php', 'XmlJsCode' => __DIR__ . '/includes/Xml.php', 'XmlSelect' => __DIR__ . '/includes/XmlSelect.php', - 'XmlTypeCheck' => __DIR__ . '/includes/libs/XmlTypeCheck.php', + 'XmlTypeCheck' => __DIR__ . '/includes/libs/mime/XmlTypeCheck.php', 'ZhConverter' => __DIR__ . '/languages/classes/LanguageZh.php', 'ZipDirectoryReader' => __DIR__ . '/includes/utils/ZipDirectoryReader.php', 'ZipDirectoryReaderError' => __DIR__ . '/includes/utils/ZipDirectoryReader.php', diff --git a/includes/Defines.php b/includes/Defines.php index 02930ea815..06168980ce 100644 --- a/includes/Defines.php +++ b/includes/Defines.php @@ -97,32 +97,7 @@ define( 'CACHE_MEMCACHED', 2 ); // MemCached, must specify servers in $wgMemCac define( 'CACHE_ACCEL', 3 ); // APC, XCache or WinCache /**@}*/ -/**@{ - * Media types. - * This defines constants for the value returned by File::getMediaType() - */ -// unknown format -define( 'MEDIATYPE_UNKNOWN', 'UNKNOWN' ); -// some bitmap image or image source (like psd, etc). Can't scale up. -define( 'MEDIATYPE_BITMAP', 'BITMAP' ); -// some vector drawing (SVG, WMF, PS, ...) or image source (oo-draw, etc). Can scale up. -define( 'MEDIATYPE_DRAWING', 'DRAWING' ); -// simple audio file (ogg, mp3, wav, midi, whatever) -define( 'MEDIATYPE_AUDIO', 'AUDIO' ); -// simple video file (ogg, mpg, etc; -// no not include formats here that may contain executable sections or scripts!) -define( 'MEDIATYPE_VIDEO', 'VIDEO' ); -// Scriptable Multimedia (flash, advanced video container formats, etc) -define( 'MEDIATYPE_MULTIMEDIA', 'MULTIMEDIA' ); -// Office Documents, Spreadsheets (office formats possibly containing apples, scripts, etc) -define( 'MEDIATYPE_OFFICE', 'OFFICE' ); -// Plain text (possibly containing program code or scripts) -define( 'MEDIATYPE_TEXT', 'TEXT' ); -// binary executable -define( 'MEDIATYPE_EXECUTABLE', 'EXECUTABLE' ); -// archive file (zip, tar, etc) -define( 'MEDIATYPE_ARCHIVE', 'ARCHIVE' ); -/**@}*/ +require_once __DIR__ . '/libs/mime/defines.php'; /**@{ * Antivirus result codes, for use in $wgAntivirusSetup. diff --git a/includes/MediaWikiServices.php b/includes/MediaWikiServices.php index f91bbaed8a..7f94ced2c9 100644 --- a/includes/MediaWikiServices.php +++ b/includes/MediaWikiServices.php @@ -19,6 +19,7 @@ use MediaWiki\Services\SalvageableService; use MediaWiki\Services\ServiceContainer; use MediaWiki\Services\NoSuchServiceException; use MWException; +use MimeAnalyzer; use ObjectCache; use ProxyLookup; use SearchEngine; @@ -539,6 +540,14 @@ class MediaWikiServices extends ServiceContainer { return $this->getService( 'MediaHandlerFactory' ); } + /** + * @since 1.28 + * @return MimeAnalyzer + */ + public function getMimeAnalyzer() { + return $this->getService( 'MimeAnalyzer' ); + } + /** * @since 1.28 * @return ProxyLookup diff --git a/includes/MimeMagic.php b/includes/MimeMagic.php index 54d58d2d44..c03bce77f0 100644 --- a/includes/MimeMagic.php +++ b/includes/MimeMagic.php @@ -1,7 +1,5 @@ ext - * map. Each line contains a MIME type followed by a space separated list of - * extensions. If multiple extensions for a single MIME type exist or if - * multiple MIME types exist for a single extension then in most cases - * MediaWiki assumes that the first extension following the MIME type is the - * canonical extension, and the first time a MIME type appears for a certain - * extension is considered the canonical MIME type. - * - * (Note that appending $wgMimeTypeFile to the end of MM_WELL_KNOWN_MIME_TYPES - * sucks because you can't redefine canonical types. This could be fixed by - * appending MM_WELL_KNOWN_MIME_TYPES behind $wgMimeTypeFile, but who knows - * what will break? In practice this probably isn't a problem anyway -- Bryan) - */ -define( 'MM_WELL_KNOWN_MIME_TYPES', <<makeConfig( 'main' ); - } - $this->mConfig = $config; - - /** - * --- load mime.types --- - */ - - global $IP; - - # Allow media handling extensions adding MIME-types and MIME-info - Hooks::run( 'MimeMagicInit', [ $this ] ); - - $types = MM_WELL_KNOWN_MIME_TYPES; - - $mimeTypeFile = $this->mConfig->get( 'MimeTypeFile' ); - if ( $mimeTypeFile == 'includes/mime.types' ) { - $mimeTypeFile = "$IP/$mimeTypeFile"; - } - - if ( $mimeTypeFile ) { - if ( is_file( $mimeTypeFile ) && is_readable( $mimeTypeFile ) ) { - wfDebug( __METHOD__ . ": loading mime types from $mimeTypeFile\n" ); - $types .= "\n"; - $types .= file_get_contents( $mimeTypeFile ); - } else { - wfDebug( __METHOD__ . ": can't load mime types from $mimeTypeFile\n" ); - } - } else { - wfDebug( __METHOD__ . ": no mime types file defined, using built-ins only.\n" ); - } - - $types .= "\n" . $this->mExtraTypes; - - $types = str_replace( [ "\r\n", "\n\r", "\n\n", "\r\r", "\r" ], "\n", $types ); - $types = str_replace( "\t", " ", $types ); - - $this->mMimeToExt = []; - $this->mExtToMime = []; - - $lines = explode( "\n", $types ); - foreach ( $lines as $s ) { - $s = trim( $s ); - if ( empty( $s ) ) { - continue; - } - if ( strpos( $s, '#' ) === 0 ) { - continue; - } - - $s = strtolower( $s ); - $i = strpos( $s, ' ' ); - - if ( $i === false ) { - continue; - } - - $mime = substr( $s, 0, $i ); - $ext = trim( substr( $s, $i + 1 ) ); - - if ( empty( $ext ) ) { - continue; - } - - if ( !empty( $this->mMimeToExt[$mime] ) ) { - $this->mMimeToExt[$mime] .= ' ' . $ext; - } else { - $this->mMimeToExt[$mime] = $ext; - } - - $extensions = explode( ' ', $ext ); - - foreach ( $extensions as $e ) { - $e = trim( $e ); - if ( empty( $e ) ) { - continue; - } - - if ( !empty( $this->mExtToMime[$e] ) ) { - $this->mExtToMime[$e] .= ' ' . $mime; - } else { - $this->mExtToMime[$e] = $mime; - } - } - } - - /** - * --- load mime.info --- - */ - - $mimeInfoFile = $this->mConfig->get( 'MimeInfoFile' ); - if ( $mimeInfoFile == 'includes/mime.info' ) { - $mimeInfoFile = "$IP/$mimeInfoFile"; - } - - $info = MM_WELL_KNOWN_MIME_INFO; - - if ( $mimeInfoFile ) { - if ( is_file( $mimeInfoFile ) && is_readable( $mimeInfoFile ) ) { - wfDebug( __METHOD__ . ": loading mime info from $mimeInfoFile\n" ); - $info .= "\n"; - $info .= file_get_contents( $mimeInfoFile ); - } else { - wfDebug( __METHOD__ . ": can't load mime info from $mimeInfoFile\n" ); - } - } else { - wfDebug( __METHOD__ . ": no mime info file defined, using built-ins only.\n" ); - } - - $info .= "\n" . $this->mExtraInfo; - - $info = str_replace( [ "\r\n", "\n\r", "\n\n", "\r\r", "\r" ], "\n", $info ); - $info = str_replace( "\t", " ", $info ); - - $this->mMimeTypeAliases = []; - $this->mMediaTypes = []; - - $lines = explode( "\n", $info ); - foreach ( $lines as $s ) { - $s = trim( $s ); - if ( empty( $s ) ) { - continue; - } - if ( strpos( $s, '#' ) === 0 ) { - continue; - } - - $s = strtolower( $s ); - $i = strpos( $s, ' ' ); - - if ( $i === false ) { - continue; - } - - # print "processing MIME INFO line $s
"; - - $match = []; - if ( preg_match( '!\[\s*(\w+)\s*\]!', $s, $match ) ) { - $s = preg_replace( '!\[\s*(\w+)\s*\]!', '', $s ); - $mtype = trim( strtoupper( $match[1] ) ); - } else { - $mtype = MEDIATYPE_UNKNOWN; - } - - $m = explode( ' ', $s ); - - if ( !isset( $this->mMediaTypes[$mtype] ) ) { - $this->mMediaTypes[$mtype] = []; - } - - foreach ( $m as $mime ) { - $mime = trim( $mime ); - if ( empty( $mime ) ) { - continue; - } - - $this->mMediaTypes[$mtype][] = $mime; - } - - if ( count( $m ) > 1 ) { - $main = $m[0]; - $mCount = count( $m ); - for ( $i = 1; $i < $mCount; $i += 1 ) { - $mime = $m[$i]; - $this->mMimeTypeAliases[$mime] = $main; - } - } - } - } - +class MimeMagic extends MimeAnalyzer { /** * Get an instance of this class * @return MimeMagic + * @deprecated since 1.28 */ public static function singleton() { - if ( self::$instance === null ) { - self::$instance = new MimeMagic( - ConfigFactory::getDefaultInstance()->makeConfig( 'main' ) - ); - } - return self::$instance; - } - - /** - * Adds to the list mapping MIME to file extensions. - * As an extension author, you are encouraged to submit patches to - * MediaWiki's core to add new MIME types to mime.types. - * @param string $types - */ - public function addExtraTypes( $types ) { - $this->mExtraTypes .= "\n" . $types; - } - - /** - * Adds to the list mapping MIME to media type. - * As an extension author, you are encouraged to submit patches to - * MediaWiki's core to add new MIME info to mime.info. - * @param string $info - */ - public function addExtraInfo( $info ) { - $this->mExtraInfo .= "\n" . $info; - } - - /** - * Returns a list of file extensions for a given MIME type as a space - * separated string or null if the MIME type was unrecognized. Resolves - * MIME type aliases. - * - * @param string $mime - * @return string|null - */ - public function getExtensionsForType( $mime ) { - $mime = strtolower( $mime ); - - // Check the mime-to-ext map - if ( isset( $this->mMimeToExt[$mime] ) ) { - return $this->mMimeToExt[$mime]; - } - - // Resolve the MIME type to the canonical type - if ( isset( $this->mMimeTypeAliases[$mime] ) ) { - $mime = $this->mMimeTypeAliases[$mime]; - if ( isset( $this->mMimeToExt[$mime] ) ) { - return $this->mMimeToExt[$mime]; - } - } - - return null; + return MediaWikiServices::getInstance()->getMIMEAnalyzer(); } /** - * Returns a list of MIME types for a given file extension as a space - * separated string or null if the extension was unrecognized. - * - * @param string $ext - * @return string|null - */ - public function getTypesForExtension( $ext ) { - $ext = strtolower( $ext ); - - $r = isset( $this->mExtToMime[$ext] ) ? $this->mExtToMime[$ext] : null; - return $r; - } - - /** - * Returns a single MIME type for a given file extension or null if unknown. - * This is always the first type from the list returned by getTypesForExtension($ext). - * - * @param string $ext - * @return string|null - */ - public function guessTypesForExtension( $ext ) { - $m = $this->getTypesForExtension( $ext ); - if ( is_null( $m ) ) { - return null; - } - - // TODO: Check if this is needed; strtok( $m, ' ' ) should be sufficient - $m = trim( $m ); - $m = preg_replace( '/\s.*$/', '', $m ); - - return $m; - } - - /** - * Tests if the extension matches the given MIME type. Returns true if a - * match was found, null if the MIME type is unknown, and false if the - * MIME type is known but no matches where found. - * - * @param string $extension - * @param string $mime - * @return bool|null - */ - public function isMatchingExtension( $extension, $mime ) { - $ext = $this->getExtensionsForType( $mime ); - - if ( !$ext ) { - return null; // Unknown MIME type - } - - $ext = explode( ' ', $ext ); - - $extension = strtolower( $extension ); - return in_array( $extension, $ext ); - } - - /** - * Returns true if the MIME type is known to represent an image format - * supported by the PHP GD library. - * - * @param string $mime - * - * @return bool - */ - public function isPHPImageType( $mime ) { - // As defined by imagegetsize and image_type_to_mime - static $types = [ - 'image/gif', 'image/jpeg', 'image/png', - 'image/x-bmp', 'image/xbm', 'image/tiff', - 'image/jp2', 'image/jpeg2000', 'image/iff', - 'image/xbm', 'image/x-xbitmap', - 'image/vnd.wap.wbmp', 'image/vnd.xiff', - 'image/x-photoshop', - 'application/x-shockwave-flash', - ]; - - return in_array( $mime, $types ); - } - - /** - * Returns true if the extension represents a type which can - * be reliably detected from its content. Use this to determine - * whether strict content checks should be applied to reject - * invalid uploads; if we can't identify the type we won't - * be able to say if it's invalid. - * - * @todo Be more accurate when using fancy MIME detector plugins; - * right now this is the bare minimum getimagesize() list. - * @param string $extension - * @return bool - */ - function isRecognizableExtension( $extension ) { - static $types = [ - // Types recognized by getimagesize() - 'gif', 'jpeg', 'jpg', 'png', 'swf', 'psd', - 'bmp', 'tiff', 'tif', 'jpc', 'jp2', - 'jpx', 'jb2', 'swc', 'iff', 'wbmp', - 'xbm', - - // Formats we recognize magic numbers for - 'djvu', 'ogx', 'ogg', 'ogv', 'oga', 'spx', - 'mid', 'pdf', 'wmf', 'xcf', 'webm', 'mkv', 'mka', - 'webp', - - // XML formats we sure hope we recognize reliably - 'svg', - ]; - return in_array( strtolower( $extension ), $types ); - } - - /** - * Improves a MIME type using the file extension. Some file formats are very generic, - * so their MIME type is not very meaningful. A more useful MIME type can be derived - * by looking at the file extension. Typically, this method would be called on the - * result of guessMimeType(). - * - * @param string $mime The MIME type, typically guessed from a file's content. - * @param string $ext The file extension, as taken from the file name - * - * @return string The MIME type - */ - public function improveTypeFromExtension( $mime, $ext ) { - if ( $mime === 'unknown/unknown' ) { - if ( $this->isRecognizableExtension( $ext ) ) { - wfDebug( __METHOD__ . ': refusing to guess mime type for .' . - "$ext file, we should have recognized it\n" ); - } else { - // Not something we can detect, so simply - // trust the file extension - $mime = $this->guessTypesForExtension( $ext ); - } - } elseif ( $mime === 'application/x-opc+zip' ) { - if ( $this->isMatchingExtension( $ext, $mime ) ) { - // A known file extension for an OPC file, - // find the proper MIME type for that file extension - $mime = $this->guessTypesForExtension( $ext ); - } else { - wfDebug( __METHOD__ . ": refusing to guess better type for $mime file, " . - ".$ext is not a known OPC extension.\n" ); - $mime = 'application/zip'; - } - } elseif ( $mime === 'text/plain' && $this->findMediaType( ".$ext" ) === MEDIATYPE_TEXT ) { - // Textual types are sometimes not recognized properly. - // If detected as text/plain, and has an extension which is textual - // improve to the extension's type. For example, csv and json are often - // misdetected as text/plain. - $mime = $this->guessTypesForExtension( $ext ); - } - - # Media handling extensions can improve the MIME detected - Hooks::run( 'MimeMagicImproveFromExtension', [ $this, $ext, &$mime ] ); - - if ( isset( $this->mMimeTypeAliases[$mime] ) ) { - $mime = $this->mMimeTypeAliases[$mime]; - } - - wfDebug( __METHOD__ . ": improved mime type for .$ext: $mime\n" ); - return $mime; - } - - /** - * MIME type detection. This uses detectMimeType to detect the MIME type - * of the file, but applies additional checks to determine some well known - * file formats that may be missed or misinterpreted by the default MIME - * detection (namely XML based formats like XHTML or SVG, as well as ZIP - * based formats like OPC/ODF files). - * - * @param string $file The file to check - * @param string|bool $ext The file extension, or true (default) to extract it from the filename. - * Set it to false to ignore the extension. DEPRECATED! Set to false, use - * improveTypeFromExtension($mime, $ext) later to improve MIME type. - * - * @return string The MIME type of $file - */ - public function guessMimeType( $file, $ext = true ) { - if ( $ext ) { // TODO: make $ext default to false. Or better, remove it. - wfDebug( __METHOD__ . ": WARNING: use of the \$ext parameter is deprecated. " . - "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); - } - - $mime = $this->doGuessMimeType( $file, $ext ); - - if ( !$mime ) { - wfDebug( __METHOD__ . ": internal type detection failed for $file (.$ext)...\n" ); - $mime = $this->detectMimeType( $file, $ext ); - } - - if ( isset( $this->mMimeTypeAliases[$mime] ) ) { - $mime = $this->mMimeTypeAliases[$mime]; - } - - wfDebug( __METHOD__ . ": guessed mime type of $file: $mime\n" ); - return $mime; - } - - /** - * Guess the MIME type from the file contents. - * - * @todo Remove $ext param - * - * @param string $file - * @param mixed $ext - * @return bool|string - * @throws MWException + * @param array $params + * @param Config $mainConfig + * @return array */ - private function doGuessMimeType( $file, $ext ) { - // Read a chunk of the file - MediaWiki\suppressWarnings(); - $f = fopen( $file, 'rb' ); - MediaWiki\restoreWarnings(); - - if ( !$f ) { - return 'unknown/unknown'; - } - - $fsize = filesize( $file ); - if ( $fsize === false ) { - return 'unknown/unknown'; - } - - $head = fread( $f, 1024 ); - $tailLength = min( 65558, $fsize ); // 65558 = maximum size of a zip EOCDR - if ( fseek( $f, -1 * $tailLength, SEEK_END ) === -1 ) { - throw new MWException( - "Seeking $tailLength bytes from EOF failed in " . __METHOD__ ); - } - $tail = $tailLength ? fread( $f, $tailLength ) : ''; - fclose( $f ); - - wfDebug( __METHOD__ . ": analyzing head and tail of $file for magic numbers.\n" ); - - // Hardcode a few magic number checks... - $headers = [ - // Multimedia... - 'MThd' => 'audio/midi', - 'OggS' => 'application/ogg', - - // Image formats... - // Note that WMF may have a bare header, no magic number. - "\x01\x00\x09\x00" => 'application/x-msmetafile', // Possibly prone to false positives? - "\xd7\xcd\xc6\x9a" => 'application/x-msmetafile', - '%PDF' => 'application/pdf', - 'gimp xcf' => 'image/x-xcf', - - // Some forbidden fruit... - 'MZ' => 'application/octet-stream', // DOS/Windows executable - "\xca\xfe\xba\xbe" => 'application/octet-stream', // Mach-O binary - "\x7fELF" => 'application/octet-stream', // ELF binary - ]; - - foreach ( $headers as $magic => $candidate ) { - if ( strncmp( $head, $magic, strlen( $magic ) ) == 0 ) { - wfDebug( __METHOD__ . ": magic header in $file recognized as $candidate\n" ); - return $candidate; - } - } - - /* Look for WebM and Matroska files */ - if ( strncmp( $head, pack( "C4", 0x1a, 0x45, 0xdf, 0xa3 ), 4 ) == 0 ) { - $doctype = strpos( $head, "\x42\x82" ); - if ( $doctype ) { - // Next byte is datasize, then data (sizes larger than 1 byte are very stupid muxers) - $data = substr( $head, $doctype + 3, 8 ); - if ( strncmp( $data, "matroska", 8 ) == 0 ) { - wfDebug( __METHOD__ . ": recognized file as video/x-matroska\n" ); - return "video/x-matroska"; - } elseif ( strncmp( $data, "webm", 4 ) == 0 ) { - wfDebug( __METHOD__ . ": recognized file as video/webm\n" ); - return "video/webm"; - } - } - wfDebug( __METHOD__ . ": unknown EBML file\n" ); - return "unknown/unknown"; - } - - /* Look for WebP */ - if ( strncmp( $head, "RIFF", 4 ) == 0 && strncmp( substr( $head, 8, 7 ), "WEBPVP8", 7 ) == 0 ) { - wfDebug( __METHOD__ . ": recognized file as image/webp\n" ); - return "image/webp"; - } - - /** - * Look for PHP. Check for this before HTML/XML... Warning: this is a - * heuristic, and won't match a file with a lot of non-PHP before. It - * will also match text files which could be PHP. :) - * - * @todo FIXME: For this reason, the check is probably useless -- an attacker - * could almost certainly just pad the file with a lot of nonsense to - * circumvent the check in any case where it would be a security - * problem. On the other hand, it causes harmful false positives (bug - * 16583). The heuristic has been cut down to exclude three-character - * strings like "wellFormed ) { - $xmlMimeTypes = $this->mConfig->get( 'XMLMimeTypes' ); - if ( isset( $xmlMimeTypes[$xml->getRootElement()] ) ) { - return $xmlMimeTypes[$xml->getRootElement()]; - } else { - return 'application/xml'; - } - } - - /** - * look for shell scripts - */ - $script_type = null; - - # detect by shebang - if ( substr( $head, 0, 2 ) == "#!" ) { - $script_type = "ASCII"; - } elseif ( substr( $head, 0, 5 ) == "\xef\xbb\xbf#!" ) { - $script_type = "UTF-8"; - } elseif ( substr( $head, 0, 7 ) == "\xfe\xff\x00#\x00!" ) { - $script_type = "UTF-16BE"; - } elseif ( substr( $head, 0, 7 ) == "\xff\xfe#\x00!" ) { - $script_type = "UTF-16LE"; - } - - if ( $script_type ) { - if ( $script_type !== "UTF-8" && $script_type !== "ASCII" ) { - // Quick and dirty fold down to ASCII! - $pack = [ 'UTF-16BE' => 'n*', 'UTF-16LE' => 'v*' ]; - $chars = unpack( $pack[$script_type], substr( $head, 2 ) ); - $head = ''; - foreach ( $chars as $codepoint ) { - if ( $codepoint < 128 ) { - $head .= chr( $codepoint ); - } else { - $head .= '?'; + public static function applyDefaultParameters( array $params, Config $mainConfig ) { + $logger = LoggerFactory::getInstance( 'Mime' ); + $params += [ + 'typeFile' => $mainConfig->get( 'MimeTypeFile' ), + 'infoFile' => $mainConfig->get( 'MimeInfoFile' ), + 'xmlTypes' => $mainConfig->get( 'XMLMimeTypes' ), + 'guessCallback' => + function ( $mimeAnalyzer, &$head, &$tail, $file, &$mime ) use ( $logger ) { + // Also test DjVu + $deja = new DjVuImage( $file ); + if ( $deja->isValid() ) { + $logger->info( __METHOD__ . ": detected $file as image/vnd.djvu\n" ); + $mime = 'image/vnd.djvu'; + + return; } - } - } - - $match = []; - - if ( preg_match( '%/?([^\s]+/)(\w+)%', $head, $match ) ) { - $mime = "application/x-{$match[2]}"; - wfDebug( __METHOD__ . ": shell script recognized as $mime\n" ); - return $mime; - } - } - - // Check for ZIP variants (before getimagesize) - if ( strpos( $tail, "PK\x05\x06" ) !== false ) { - wfDebug( __METHOD__ . ": ZIP header present in $file\n" ); - return $this->detectZipType( $head, $tail, $ext ); - } - - MediaWiki\suppressWarnings(); - $gis = getimagesize( $file ); - MediaWiki\restoreWarnings(); - - if ( $gis && isset( $gis['mime'] ) ) { - $mime = $gis['mime']; - wfDebug( __METHOD__ . ": getimagesize detected $file as $mime\n" ); - return $mime; - } - - // Also test DjVu - $deja = new DjVuImage( $file ); - if ( $deja->isValid() ) { - wfDebug( __METHOD__ . ": detected $file as image/vnd.djvu\n" ); - return 'image/vnd.djvu'; - } - - # Media handling extensions can guess the MIME by content - # It's intentionally here so that if core is wrong about a type (false positive), - # people will hopefully nag and submit patches :) - $mime = false; - # Some strings by reference for performance - assuming well-behaved hooks - Hooks::run( - 'MimeMagicGuessFromContent', - [ $this, &$head, &$tail, $file, &$mime ] - ); - - return $mime; - } - - /** - * Detect application-specific file type of a given ZIP file from its - * header data. Currently works for OpenDocument and OpenXML types... - * If can't tell, returns 'application/zip'. - * - * @param string $header Some reasonably-sized chunk of file header - * @param string|null $tail The tail of the file - * @param string|bool $ext The file extension, or true to extract it from the filename. - * Set it to false (default) to ignore the extension. DEPRECATED! Set to false, - * use improveTypeFromExtension($mime, $ext) later to improve MIME type. - * - * @return string - */ - function detectZipType( $header, $tail = null, $ext = false ) { - if ( $ext ) { # TODO: remove $ext param - wfDebug( __METHOD__ . ": WARNING: use of the \$ext parameter is deprecated. " . - "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); - } - - $mime = 'application/zip'; - $opendocTypes = [ - 'chart-template', - 'chart', - 'formula-template', - 'formula', - 'graphics-template', - 'graphics', - 'image-template', - 'image', - 'presentation-template', - 'presentation', - 'spreadsheet-template', - 'spreadsheet', - 'text-template', - 'text-master', - 'text-web', - 'text' ]; - - // http://lists.oasis-open.org/archives/office/200505/msg00006.html - $types = '(?:' . implode( '|', $opendocTypes ) . ')'; - $opendocRegex = "/^mimetype(application\/vnd\.oasis\.opendocument\.$types)/"; - - $openxmlRegex = "/^\[Content_Types\].xml/"; - - if ( preg_match( $opendocRegex, substr( $header, 30 ), $matches ) ) { - $mime = $matches[1]; - wfDebug( __METHOD__ . ": detected $mime from ZIP archive\n" ); - } elseif ( preg_match( $openxmlRegex, substr( $header, 30 ) ) ) { - $mime = "application/x-opc+zip"; - # TODO: remove the block below, as soon as improveTypeFromExtension is used everywhere - if ( $ext !== true && $ext !== false ) { - // These MIME's are stored in the database, where we don't really want - // x-opc+zip, because we use it only for internal purposes - if ( $this->isMatchingExtension( $ext, $mime ) ) { - /* A known file extension for an OPC file, - * find the proper mime type for that file extension - */ - $mime = $this->guessTypesForExtension( $ext ); - } else { - $mime = "application/zip"; - } - } - wfDebug( __METHOD__ . ": detected an Open Packaging Conventions archive: $mime\n" ); - } elseif ( substr( $header, 0, 8 ) == "\xd0\xcf\x11\xe0\xa1\xb1\x1a\xe1" && - ( $headerpos = strpos( $tail, "PK\x03\x04" ) ) !== false && - preg_match( $openxmlRegex, substr( $tail, $headerpos + 30 ) ) ) { - if ( substr( $header, 512, 4 ) == "\xEC\xA5\xC1\x00" ) { - $mime = "application/msword"; - } - switch ( substr( $header, 512, 6 ) ) { - case "\xEC\xA5\xC1\x00\x0E\x00": - case "\xEC\xA5\xC1\x00\x1C\x00": - case "\xEC\xA5\xC1\x00\x43\x00": - $mime = "application/vnd.ms-powerpoint"; - break; - case "\xFD\xFF\xFF\xFF\x10\x00": - case "\xFD\xFF\xFF\xFF\x1F\x00": - case "\xFD\xFF\xFF\xFF\x22\x00": - case "\xFD\xFF\xFF\xFF\x23\x00": - case "\xFD\xFF\xFF\xFF\x28\x00": - case "\xFD\xFF\xFF\xFF\x29\x00": - case "\xFD\xFF\xFF\xFF\x10\x02": - case "\xFD\xFF\xFF\xFF\x1F\x02": - case "\xFD\xFF\xFF\xFF\x22\x02": - case "\xFD\xFF\xFF\xFF\x23\x02": - case "\xFD\xFF\xFF\xFF\x28\x02": - case "\xFD\xFF\xFF\xFF\x29\x02": - $mime = "application/vnd.msexcel"; - break; - } - - wfDebug( __METHOD__ . ": detected a MS Office document with OPC trailer\n" ); - } else { - wfDebug( __METHOD__ . ": unable to identify type of ZIP archive\n" ); - } - return $mime; - } - - /** - * Internal MIME type detection. Detection is done using an external - * program, if $wgMimeDetectorCommand is set. Otherwise, the fileinfo - * extension is tried if it is available. If detection fails and $ext - * is not false, the MIME type is guessed from the file extension, - * using guessTypesForExtension. - * - * If the MIME type is still unknown, getimagesize is used to detect the - * MIME type if the file is an image. If no MIME type can be determined, - * this function returns 'unknown/unknown'. - * - * @param string $file The file to check - * @param string|bool $ext The file extension, or true (default) to extract it from the filename. - * Set it to false to ignore the extension. DEPRECATED! Set to false, use - * improveTypeFromExtension($mime, $ext) later to improve MIME type. - * - * @return string The MIME type of $file - */ - private function detectMimeType( $file, $ext = true ) { - /** @todo Make $ext default to false. Or better, remove it. */ - if ( $ext ) { - wfDebug( __METHOD__ . ": WARNING: use of the \$ext parameter is deprecated. " - . "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); - } - - $mimeDetectorCommand = $this->mConfig->get( 'MimeDetectorCommand' ); - $m = null; - if ( $mimeDetectorCommand ) { - $args = wfEscapeShellArg( $file ); - $m = wfShellExec( "$mimeDetectorCommand $args" ); - } elseif ( function_exists( "finfo_open" ) && function_exists( "finfo_file" ) ) { - $mime_magic_resource = finfo_open( FILEINFO_MIME ); - - if ( $mime_magic_resource ) { - $m = finfo_file( $mime_magic_resource, $file ); - finfo_close( $mime_magic_resource ); - } else { - wfDebug( __METHOD__ . ": finfo_open failed on " . FILEINFO_MIME . "!\n" ); - } - } else { - wfDebug( __METHOD__ . ": no magic mime detector found!\n" ); - } - - if ( $m ) { - # normalize - $m = preg_replace( '![;, ].*$!', '', $m ); # strip charset, etc - $m = trim( $m ); - $m = strtolower( $m ); - - if ( strpos( $m, 'unknown' ) !== false ) { - $m = null; - } else { - wfDebug( __METHOD__ . ": magic mime type of $file: $m\n" ); - return $m; - } - } - - // If desired, look at extension as a fallback. - if ( $ext === true ) { - $i = strrpos( $file, '.' ); - $ext = strtolower( $i ? substr( $file, $i + 1 ) : '' ); - } - if ( $ext ) { - if ( $this->isRecognizableExtension( $ext ) ) { - wfDebug( __METHOD__ . ": refusing to guess mime type for .$ext file, " - . "we should have recognized it\n" ); - } else { - $m = $this->guessTypesForExtension( $ext ); - if ( $m ) { - wfDebug( __METHOD__ . ": extension mime type of $file: $m\n" ); - return $m; - } - } - } - - // Unknown type - wfDebug( __METHOD__ . ": failed to guess mime type for $file!\n" ); - return 'unknown/unknown'; - } - - /** - * Determine the media type code for a file, using its MIME type, name and - * possibly its contents. - * - * This function relies on the findMediaType(), mapping extensions and MIME - * types to media types. - * - * @todo analyse file if need be - * @todo look at multiple extension, separately and together. - * - * @param string $path Full path to the image file, in case we have to look at the contents - * (if null, only the MIME type is used to determine the media type code). - * @param string $mime MIME type. If null it will be guessed using guessMimeType. - * - * @return string A value to be used with the MEDIATYPE_xxx constants. - */ - function getMediaType( $path = null, $mime = null ) { - if ( !$mime && !$path ) { - return MEDIATYPE_UNKNOWN; - } - - // If MIME type is unknown, guess it - if ( !$mime ) { - $mime = $this->guessMimeType( $path, false ); - } - - // Special code for ogg - detect if it's video (theora), - // else label it as sound. - if ( $mime == 'application/ogg' && file_exists( $path ) ) { - - // Read a chunk of the file - $f = fopen( $path, "rt" ); - if ( !$f ) { - return MEDIATYPE_UNKNOWN; - } - $head = fread( $f, 256 ); - fclose( $f ); - - $head = str_replace( 'ffmpeg2theora', '', strtolower( $head ) ); - - // This is an UGLY HACK, file should be parsed correctly - if ( strpos( $head, 'theora' ) !== false ) { - return MEDIATYPE_VIDEO; - } elseif ( strpos( $head, 'vorbis' ) !== false ) { - return MEDIATYPE_AUDIO; - } elseif ( strpos( $head, 'flac' ) !== false ) { - return MEDIATYPE_AUDIO; - } elseif ( strpos( $head, 'speex' ) !== false ) { - return MEDIATYPE_AUDIO; - } else { - return MEDIATYPE_MULTIMEDIA; - } - } - - $type = null; - // Check for entry for full MIME type - if ( $mime ) { - $type = $this->findMediaType( $mime ); - if ( $type !== MEDIATYPE_UNKNOWN ) { - return $type; - } - } - - // Check for entry for file extension - if ( $path ) { - $i = strrpos( $path, '.' ); - $e = strtolower( $i ? substr( $path, $i + 1 ) : '' ); - - // TODO: look at multi-extension if this fails, parse from full path - $type = $this->findMediaType( '.' . $e ); - if ( $type !== MEDIATYPE_UNKNOWN ) { - return $type; - } - } - - // Check major MIME type - if ( $mime ) { - $i = strpos( $mime, '/' ); - if ( $i !== false ) { - $major = substr( $mime, 0, $i ); - $type = $this->findMediaType( $major ); - if ( $type !== MEDIATYPE_UNKNOWN ) { - return $type; - } - } - } + // Some strings by reference for performance - assuming well-behaved hooks + Hooks::run( + 'MimeMagicGuessFromContent', + [ $mimeAnalyzer, &$head, &$tail, $file, &$mime ] + ); + }, + 'extCallback' => function ( $mimeAnalyzer, $ext, &$mime ) { + // Media handling extensions can improve the MIME detected + Hooks::run( 'MimeMagicImproveFromExtension', [ $mimeAnalyzer, $ext, &$mime ] ); + }, + 'initCallback' => function ( $mimeAnalyzer ) { + // Allow media handling extensions adding MIME-types and MIME-info + Hooks::run( 'MimeMagicInit', [ $mimeAnalyzer ] ); + }, + 'logger' => $logger + ]; - if ( !$type ) { - $type = MEDIATYPE_UNKNOWN; + if ( $params['infoFile'] === 'includes/mime.info' ) { + $params['infoFile'] = __DIR__ . "/libs/mime/mime.info"; } - return $type; - } - - /** - * Returns a media code matching the given MIME type or file extension. - * File extensions are represented by a string starting with a dot (.) to - * distinguish them from MIME types. - * - * This function relies on the mapping defined by $this->mMediaTypes - * @access private - * @param string $extMime - * @return int|string - */ - function findMediaType( $extMime ) { - if ( strpos( $extMime, '.' ) === 0 ) { - // If it's an extension, look up the MIME types - $m = $this->getTypesForExtension( substr( $extMime, 1 ) ); - if ( !$m ) { - return MEDIATYPE_UNKNOWN; - } - - $m = explode( ' ', $m ); - } else { - // Normalize MIME type - if ( isset( $this->mMimeTypeAliases[$extMime] ) ) { - $extMime = $this->mMimeTypeAliases[$extMime]; - } - - $m = [ $extMime ]; + if ( $params['typeFile'] === 'includes/mime.types' ) { + $params['typeFile'] = __DIR__ . "/libs/mime/mime.types"; } - foreach ( $m as $mime ) { - foreach ( $this->mMediaTypes as $type => $codes ) { - if ( in_array( $mime, $codes, true ) ) { - return $type; - } - } + $detectorCmd = $mainConfig->get( 'MimeDetectorCommand' ); + if ( $detectorCmd ) { + $params['detectCallback'] = function ( $file ) use ( $detectorCmd ) { + return wfShellExec( "$detectorCmd " . wfEscapeShellArg( $file ) ); + }; } - return MEDIATYPE_UNKNOWN; - } - - /** - * Get the MIME types that various versions of Internet Explorer would - * detect from a chunk of the content. - * - * @param string $fileName The file name (unused at present) - * @param string $chunk The first 256 bytes of the file - * @param string $proposed The MIME type proposed by the server - * @return array - */ - public function getIEMimeTypes( $fileName, $chunk, $proposed ) { - $ca = $this->getIEContentAnalyzer(); - return $ca->getRealMimesFromData( $fileName, $chunk, $proposed ); - } - - /** - * Get a cached instance of IEContentAnalyzer - * - * @return IEContentAnalyzer - */ - protected function getIEContentAnalyzer() { - if ( is_null( $this->mIEAnalyzer ) ) { - $this->mIEAnalyzer = new IEContentAnalyzer; - } - return $this->mIEAnalyzer; + return $params; } } diff --git a/includes/ServiceWiring.php b/includes/ServiceWiring.php index 42b75f042b..49183e5705 100644 --- a/includes/ServiceWiring.php +++ b/includes/ServiceWiring.php @@ -190,6 +190,15 @@ return [ ); }, + 'MimeAnalyzer' => function( MediaWikiServices $services ) { + return new MimeMagic( + MimeMagic::applyDefaultParameters( + [], + $services->getMainConfig() + ) + ); + }, + 'ProxyLookup' => function( MediaWikiServices $services ) { $mainConfig = $services->getMainConfig(); return new ProxyLookup( diff --git a/includes/libs/IEContentAnalyzer.php b/includes/libs/IEContentAnalyzer.php deleted file mode 100644 index 0d1e527ba6..0000000000 --- a/includes/libs/IEContentAnalyzer.php +++ /dev/null @@ -1,851 +0,0 @@ - [ - 'text/plain', - 'application/octet-stream', - 'application/x-netcdf', // [sic] - ], - 'text' /*3*/ => [ - 'text/richtext', 'image/x-bitmap', 'application/postscript', 'application/base64', - 'application/macbinhex40', 'application/x-cdf', 'text/scriptlet' - ], - 'binary' /*4*/ => [ - 'application/pdf', 'audio/x-aiff', 'audio/basic', 'audio/wav', 'image/gif', - 'image/pjpeg', 'image/jpeg', 'image/tiff', 'image/x-png', 'image/png', 'image/bmp', - 'image/x-jg', 'image/x-art', 'image/x-emf', 'image/x-wmf', 'video/avi', - 'video/x-msvideo', 'video/mpeg', 'application/x-compressed', - 'application/x-zip-compressed', 'application/x-gzip-compressed', 'application/java', - 'application/x-msdownload' - ], - 'html' /*5*/ => [ 'text/html' ], - ]; - - /** - * Changes to the type table in later versions of IE - */ - protected $addedTypes = [ - 'ie07' => [ - 'text' => [ 'text/xml', 'application/xml' ] - ], - ]; - - /** - * An approximation of the "Content Type" values in HKEY_CLASSES_ROOT in a - * typical Windows installation. - * - * Used for extension to MIME type mapping if detection fails. - */ - protected $registry = [ - '.323' => 'text/h323', - '.3g2' => 'video/3gpp2', - '.3gp' => 'video/3gpp', - '.3gp2' => 'video/3gpp2', - '.3gpp' => 'video/3gpp', - '.aac' => 'audio/aac', - '.ac3' => 'audio/ac3', - '.accda' => 'application/msaccess', - '.accdb' => 'application/msaccess', - '.accdc' => 'application/msaccess', - '.accde' => 'application/msaccess', - '.accdr' => 'application/msaccess', - '.accdt' => 'application/msaccess', - '.ade' => 'application/msaccess', - '.adp' => 'application/msaccess', - '.adts' => 'audio/aac', - '.ai' => 'application/postscript', - '.aif' => 'audio/aiff', - '.aifc' => 'audio/aiff', - '.aiff' => 'audio/aiff', - '.amc' => 'application/x-mpeg', - '.application' => 'application/x-ms-application', - '.asf' => 'video/x-ms-asf', - '.asx' => 'video/x-ms-asf', - '.au' => 'audio/basic', - '.avi' => 'video/avi', - '.bmp' => 'image/bmp', - '.caf' => 'audio/x-caf', - '.cat' => 'application/vnd.ms-pki.seccat', - '.cbo' => 'application/sha', - '.cdda' => 'audio/aiff', - '.cer' => 'application/x-x509-ca-cert', - '.conf' => 'text/plain', - '.crl' => 'application/pkix-crl', - '.crt' => 'application/x-x509-ca-cert', - '.css' => 'text/css', - '.csv' => 'application/vnd.ms-excel', - '.der' => 'application/x-x509-ca-cert', - '.dib' => 'image/bmp', - '.dif' => 'video/x-dv', - '.dll' => 'application/x-msdownload', - '.doc' => 'application/msword', - '.docm' => 'application/vnd.ms-word.document.macroEnabled.12', - '.docx' => 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', - '.dot' => 'application/msword', - '.dotm' => 'application/vnd.ms-word.template.macroEnabled.12', - '.dotx' => 'application/vnd.openxmlformats-officedocument.wordprocessingml.template', - '.dv' => 'video/x-dv', - '.dwfx' => 'model/vnd.dwfx+xps', - '.edn' => 'application/vnd.adobe.edn', - '.eml' => 'message/rfc822', - '.eps' => 'application/postscript', - '.etd' => 'application/x-ebx', - '.exe' => 'application/x-msdownload', - '.fdf' => 'application/vnd.fdf', - '.fif' => 'application/fractals', - '.gif' => 'image/gif', - '.gsm' => 'audio/x-gsm', - '.hqx' => 'application/mac-binhex40', - '.hta' => 'application/hta', - '.htc' => 'text/x-component', - '.htm' => 'text/html', - '.html' => 'text/html', - '.htt' => 'text/webviewhtml', - '.hxa' => 'application/xml', - '.hxc' => 'application/xml', - '.hxd' => 'application/octet-stream', - '.hxe' => 'application/xml', - '.hxf' => 'application/xml', - '.hxh' => 'application/octet-stream', - '.hxi' => 'application/octet-stream', - '.hxk' => 'application/xml', - '.hxq' => 'application/octet-stream', - '.hxr' => 'application/octet-stream', - '.hxs' => 'application/octet-stream', - '.hxt' => 'application/xml', - '.hxv' => 'application/xml', - '.hxw' => 'application/octet-stream', - '.ico' => 'image/x-icon', - '.iii' => 'application/x-iphone', - '.ins' => 'application/x-internet-signup', - '.iqy' => 'text/x-ms-iqy', - '.isp' => 'application/x-internet-signup', - '.jfif' => 'image/jpeg', - '.jnlp' => 'application/x-java-jnlp-file', - '.jpe' => 'image/jpeg', - '.jpeg' => 'image/jpeg', - '.jpg' => 'image/jpeg', - '.jtx' => 'application/x-jtx+xps', - '.latex' => 'application/x-latex', - '.log' => 'text/plain', - '.m1v' => 'video/mpeg', - '.m2v' => 'video/mpeg', - '.m3u' => 'audio/x-mpegurl', - '.mac' => 'image/x-macpaint', - '.man' => 'application/x-troff-man', - '.mda' => 'application/msaccess', - '.mdb' => 'application/msaccess', - '.mde' => 'application/msaccess', - '.mfp' => 'application/x-shockwave-flash', - '.mht' => 'message/rfc822', - '.mhtml' => 'message/rfc822', - '.mid' => 'audio/mid', - '.midi' => 'audio/mid', - '.mod' => 'video/mpeg', - '.mov' => 'video/quicktime', - '.mp2' => 'video/mpeg', - '.mp2v' => 'video/mpeg', - '.mp3' => 'audio/mpeg', - '.mp4' => 'video/mp4', - '.mpa' => 'video/mpeg', - '.mpe' => 'video/mpeg', - '.mpeg' => 'video/mpeg', - '.mpf' => 'application/vnd.ms-mediapackage', - '.mpg' => 'video/mpeg', - '.mpv2' => 'video/mpeg', - '.mqv' => 'video/quicktime', - '.NMW' => 'application/nmwb', - '.nws' => 'message/rfc822', - '.odc' => 'text/x-ms-odc', - '.ols' => 'application/vnd.ms-publisher', - '.p10' => 'application/pkcs10', - '.p12' => 'application/x-pkcs12', - '.p7b' => 'application/x-pkcs7-certificates', - '.p7c' => 'application/pkcs7-mime', - '.p7m' => 'application/pkcs7-mime', - '.p7r' => 'application/x-pkcs7-certreqresp', - '.p7s' => 'application/pkcs7-signature', - '.pct' => 'image/pict', - '.pdf' => 'application/pdf', - '.pdx' => 'application/vnd.adobe.pdx', - '.pfx' => 'application/x-pkcs12', - '.pic' => 'image/pict', - '.pict' => 'image/pict', - '.pinstall' => 'application/x-picasa-detect', - '.pko' => 'application/vnd.ms-pki.pko', - '.png' => 'image/png', - '.pnt' => 'image/x-macpaint', - '.pntg' => 'image/x-macpaint', - '.pot' => 'application/vnd.ms-powerpoint', - '.potm' => 'application/vnd.ms-powerpoint.template.macroEnabled.12', - '.potx' => 'application/vnd.openxmlformats-officedocument.presentationml.template', - '.ppa' => 'application/vnd.ms-powerpoint', - '.ppam' => 'application/vnd.ms-powerpoint.addin.macroEnabled.12', - '.pps' => 'application/vnd.ms-powerpoint', - '.ppsm' => 'application/vnd.ms-powerpoint.slideshow.macroEnabled.12', - '.ppsx' => 'application/vnd.openxmlformats-officedocument.presentationml.slideshow', - '.ppt' => 'application/vnd.ms-powerpoint', - '.pptm' => 'application/vnd.ms-powerpoint.presentation.macroEnabled.12', - '.pptx' => 'application/vnd.openxmlformats-officedocument.presentationml.presentation', - '.prf' => 'application/pics-rules', - '.ps' => 'application/postscript', - '.pub' => 'application/vnd.ms-publisher', - '.pwz' => 'application/vnd.ms-powerpoint', - '.py' => 'text/plain', - '.pyw' => 'text/plain', - '.qht' => 'text/x-html-insertion', - '.qhtm' => 'text/x-html-insertion', - '.qt' => 'video/quicktime', - '.qti' => 'image/x-quicktime', - '.qtif' => 'image/x-quicktime', - '.qtl' => 'application/x-quicktimeplayer', - '.rat' => 'application/rat-file', - '.rmf' => 'application/vnd.adobe.rmf', - '.rmi' => 'audio/mid', - '.rqy' => 'text/x-ms-rqy', - '.rtf' => 'application/msword', - '.sct' => 'text/scriptlet', - '.sd2' => 'audio/x-sd2', - '.sdp' => 'application/sdp', - '.shtml' => 'text/html', - '.sit' => 'application/x-stuffit', - '.sldm' => 'application/vnd.ms-powerpoint.slide.macroEnabled.12', - '.sldx' => 'application/vnd.openxmlformats-officedocument.presentationml.slide', - '.slk' => 'application/vnd.ms-excel', - '.snd' => 'audio/basic', - '.so' => 'application/x-apachemodule', - '.sol' => 'text/plain', - '.sor' => 'text/plain', - '.spc' => 'application/x-pkcs7-certificates', - '.spl' => 'application/futuresplash', - '.sst' => 'application/vnd.ms-pki.certstore', - '.stl' => 'application/vnd.ms-pki.stl', - '.swf' => 'application/x-shockwave-flash', - '.thmx' => 'application/vnd.ms-officetheme', - '.tif' => 'image/tiff', - '.tiff' => 'image/tiff', - '.txt' => 'text/plain', - '.uls' => 'text/iuls', - '.vcf' => 'text/x-vcard', - '.vdx' => 'application/vnd.ms-visio.viewer', - '.vsd' => 'application/vnd.ms-visio.viewer', - '.vss' => 'application/vnd.ms-visio.viewer', - '.vst' => 'application/vnd.ms-visio.viewer', - '.vsx' => 'application/vnd.ms-visio.viewer', - '.vtx' => 'application/vnd.ms-visio.viewer', - '.wav' => 'audio/wav', - '.wax' => 'audio/x-ms-wax', - '.wbk' => 'application/msword', - '.wdp' => 'image/vnd.ms-photo', - '.wiz' => 'application/msword', - '.wm' => 'video/x-ms-wm', - '.wma' => 'audio/x-ms-wma', - '.wmd' => 'application/x-ms-wmd', - '.wmv' => 'video/x-ms-wmv', - '.wmx' => 'video/x-ms-wmx', - '.wmz' => 'application/x-ms-wmz', - '.wpl' => 'application/vnd.ms-wpl', - '.wsc' => 'text/scriptlet', - '.wvx' => 'video/x-ms-wvx', - '.xaml' => 'application/xaml+xml', - '.xbap' => 'application/x-ms-xbap', - '.xdp' => 'application/vnd.adobe.xdp+xml', - '.xfdf' => 'application/vnd.adobe.xfdf', - '.xht' => 'application/xhtml+xml', - '.xhtml' => 'application/xhtml+xml', - '.xla' => 'application/vnd.ms-excel', - '.xlam' => 'application/vnd.ms-excel.addin.macroEnabled.12', - '.xlk' => 'application/vnd.ms-excel', - '.xll' => 'application/vnd.ms-excel', - '.xlm' => 'application/vnd.ms-excel', - '.xls' => 'application/vnd.ms-excel', - '.xlsb' => 'application/vnd.ms-excel.sheet.binary.macroEnabled.12', - '.xlsm' => 'application/vnd.ms-excel.sheet.macroEnabled.12', - '.xlsx' => 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet', - '.xlt' => 'application/vnd.ms-excel', - '.xltm' => 'application/vnd.ms-excel.template.macroEnabled.12', - '.xltx' => 'application/vnd.openxmlformats-officedocument.spreadsheetml.template', - '.xlw' => 'application/vnd.ms-excel', - '.xml' => 'text/xml', - '.xps' => 'application/vnd.ms-xpsdocument', - '.xsl' => 'text/xml', - ]; - - /** - * IE versions which have been analysed to bring you this class, and for - * which some substantive difference exists. These will appear as keys - * in the return value of getRealMimesFromData(). The names are chosen to sort correctly. - */ - protected $versions = [ 'ie05', 'ie06', 'ie07', 'ie07.strict', 'ie07.nohtml' ]; - - /** - * Type table with versions expanded - */ - protected $typeTable = []; - - /** constructor */ - function __construct() { - // Construct versioned type arrays from the base type array plus additions - $types = $this->baseTypeTable; - foreach ( $this->versions as $version ) { - if ( isset( $this->addedTypes[$version] ) ) { - foreach ( $this->addedTypes[$version] as $format => $addedTypes ) { - $types[$format] = array_merge( $types[$format], $addedTypes ); - } - } - $this->typeTable[$version] = $types; - } - } - - /** - * Get the MIME types from getMimesFromData(), but convert the result from IE's - * idiosyncratic private types into something other apps will understand. - * - * @param string $fileName the file name (unused at present) - * @param string $chunk the first 256 bytes of the file - * @param string $proposed the MIME type proposed by the server - * - * @return Array: map of IE version to detected MIME type - */ - public function getRealMimesFromData( $fileName, $chunk, $proposed ) { - $types = $this->getMimesFromData( $fileName, $chunk, $proposed ); - $types = array_map( [ $this, 'translateMimeType' ], $types ); - return $types; - } - - /** - * Translate a MIME type from IE's idiosyncratic private types into - * more commonly understood type strings - * @param $type - * @return string - */ - public function translateMimeType( $type ) { - static $table = [ - 'image/pjpeg' => 'image/jpeg', - 'image/x-png' => 'image/png', - 'image/x-wmf' => 'application/x-msmetafile', - 'image/bmp' => 'image/x-bmp', - 'application/x-zip-compressed' => 'application/zip', - 'application/x-compressed' => 'application/x-compress', - 'application/x-gzip-compressed' => 'application/x-gzip', - 'audio/mid' => 'audio/midi', - ]; - if ( isset( $table[$type] ) ) { - $type = $table[$type]; - } - return $type; - } - - /** - * Get the untranslated MIME types for all known versions - * - * @param string $fileName the file name (unused at present) - * @param string $chunk the first 256 bytes of the file - * @param string $proposed the MIME type proposed by the server - * - * @return Array: map of IE version to detected MIME type - */ - public function getMimesFromData( $fileName, $chunk, $proposed ) { - $types = []; - foreach ( $this->versions as $version ) { - $types[$version] = $this->getMimeTypeForVersion( $version, $fileName, $chunk, $proposed ); - } - return $types; - } - - /** - * Get the MIME type for a given named version - * @param $version - * @param $fileName - * @param $chunk - * @param $proposed - * @return bool|string - */ - protected function getMimeTypeForVersion( $version, $fileName, $chunk, $proposed ) { - // Strip text after a semicolon - $semiPos = strpos( $proposed, ';' ); - if ( $semiPos !== false ) { - $proposed = substr( $proposed, 0, $semiPos ); - } - - $proposedFormat = $this->getDataFormat( $version, $proposed ); - if ( $proposedFormat == 'unknown' - && $proposed != 'multipart/mixed' - && $proposed != 'multipart/x-mixed-replace' ) - { - return $proposed; - } - if ( strval( $chunk ) === '' ) { - return $proposed; - } - - // Truncate chunk at 255 bytes - $chunk = substr( $chunk, 0, 255 ); - - // IE does the Check*Headers() calls last, and instead does the following image - // type checks by directly looking for the magic numbers. What I do here should - // have the same effect since the magic number checks are identical in both cases. - $result = $this->sampleData( $version, $chunk ); - $sampleFound = $result['found']; - $counters = $result['counters']; - $binaryType = $this->checkBinaryHeaders( $version, $chunk ); - $textType = $this->checkTextHeaders( $version, $chunk ); - - if ( $proposed == 'text/html' && isset( $sampleFound['html'] ) ) { - return 'text/html'; - } - if ( $proposed == 'image/gif' && $binaryType == 'image/gif' ) { - return 'image/gif'; - } - if ( ( $proposed == 'image/pjpeg' || $proposed == 'image/jpeg' ) - && $binaryType == 'image/pjpeg' ) - { - return $proposed; - } - // PNG check added in IE 7 - if ( $version >= 'ie07' - && ( $proposed == 'image/x-png' || $proposed == 'image/png' ) - && $binaryType == 'image/x-png' ) - { - return $proposed; - } - - // CDF was removed in IE 7 so it won't be in $sampleFound for later versions - if ( isset( $sampleFound['cdf'] ) ) { - return 'application/x-cdf'; - } - - // RSS and Atom were added in IE 7 so they won't be in $sampleFound for - // previous versions - if ( isset( $sampleFound['rss'] ) ) { - return 'application/rss+xml'; - } - if ( isset( $sampleFound['rdf-tag'] ) - && isset( $sampleFound['rdf-url'] ) - && isset( $sampleFound['rdf-purl'] ) ) - { - return 'application/rss+xml'; - } - if ( isset( $sampleFound['atom'] ) ) { - return 'application/atom+xml'; - } - - if ( isset( $sampleFound['xml'] ) ) { - // TODO: I'm not sure under what circumstances this flag is enabled - if ( strpos( $version, 'strict' ) !== false ) { - if ( $proposed == 'text/html' || $proposed == 'text/xml' ) { - return 'text/xml'; - } - } else { - return 'text/xml'; - } - } - if ( isset( $sampleFound['html'] ) ) { - // TODO: I'm not sure under what circumstances this flag is enabled - if ( strpos( $version, 'nohtml' ) !== false ) { - if ( $proposed == 'text/plain' ) { - return 'text/html'; - } - } else { - return 'text/html'; - } - } - if ( isset( $sampleFound['xbm'] ) ) { - return 'image/x-bitmap'; - } - if ( isset( $sampleFound['binhex'] ) ) { - return 'application/macbinhex40'; - } - if ( isset( $sampleFound['scriptlet'] ) ) { - if ( strpos( $version, 'strict' ) !== false ) { - if ( $proposed == 'text/plain' || $proposed == 'text/scriptlet' ) { - return 'text/scriptlet'; - } - } else { - return 'text/scriptlet'; - } - } - - // Freaky heuristics to determine if the data is text or binary - // The heuristic is of course broken for non-ASCII text - if ( $counters['ctrl'] != 0 && ( $counters['ff'] + $counters['low'] ) - < ( $counters['ctrl'] + $counters['high'] ) * 16 ) - { - $kindOfBinary = true; - $type = $binaryType ? $binaryType : $textType; - if ( $type === false ) { - $type = 'application/octet-stream'; - } - } else { - $kindOfBinary = false; - $type = $textType ? $textType : $binaryType; - if ( $type === false ) { - $type = 'text/plain'; - } - } - - // Check if the output format is ambiguous - // This generally means that detection failed, real types aren't ambiguous - $detectedFormat = $this->getDataFormat( $version, $type ); - if ( $detectedFormat != 'ambiguous' ) { - return $type; - } - - if ( $proposedFormat != 'ambiguous' ) { - // FormatAgreesWithData() - if ( $proposedFormat == 'text' && !$kindOfBinary ) { - return $proposed; - } - if ( $proposedFormat == 'binary' && $kindOfBinary ) { - return $proposed; - } - if ( $proposedFormat == 'html' ) { - return $proposed; - } - } - - // Find a MIME type by searching the registry for the file extension. - $dotPos = strrpos( $fileName, '.' ); - if ( $dotPos === false ) { - return $type; - } - $ext = substr( $fileName, $dotPos ); - if ( isset( $this->registry[$ext] ) ) { - return $this->registry[$ext]; - } - - // TODO: If the extension has an application registered to it, IE will return - // application/octet-stream. We'll skip that, so we could erroneously - // return text/plain or application/x-netcdf where application/octet-stream - // would be correct. - - return $type; - } - - /** - * Check for text headers at the start of the chunk - * Confirmed same in 5 and 7. - * @param $version - * @param $chunk - * @return bool|string - */ - private function checkTextHeaders( $version, $chunk ) { - $chunk2 = substr( $chunk, 0, 2 ); - $chunk4 = substr( $chunk, 0, 4 ); - $chunk5 = substr( $chunk, 0, 5 ); - if ( $chunk4 == '%PDF' ) { - return 'application/pdf'; - } - if ( $chunk2 == '%!' ) { - return 'application/postscript'; - } - if ( $chunk5 == '{\\rtf' ) { - return 'text/richtext'; - } - if ( $chunk5 == 'begin' ) { - return 'application/base64'; - } - return false; - } - - /** - * Check for binary headers at the start of the chunk - * Confirmed same in 5 and 7. - * @param $version - * @param $chunk - * @return bool|string - */ - private function checkBinaryHeaders( $version, $chunk ) { - $chunk2 = substr( $chunk, 0, 2 ); - $chunk3 = substr( $chunk, 0, 3 ); - $chunk4 = substr( $chunk, 0, 4 ); - $chunk5 = substr( $chunk, 0, 5 ); - $chunk5uc = strtoupper( $chunk5 ); - $chunk8 = substr( $chunk, 0, 8 ); - if ( $chunk5uc == 'GIF87' || $chunk5uc == 'GIF89' ) { - return 'image/gif'; - } - if ( $chunk2 == "\xff\xd8" ) { - return 'image/pjpeg'; // actually plain JPEG but this is what IE returns - } - - if ( $chunk2 == 'BM' - && substr( $chunk, 6, 2 ) == "\000\000" - && substr( $chunk, 8, 2 ) == "\000\000" ) - { - return 'image/bmp'; // another non-standard MIME - } - if ( $chunk4 == 'RIFF' - && substr( $chunk, 8, 4 ) == 'WAVE' ) - { - return 'audio/wav'; - } - // These were integer literals in IE - // Perhaps the author was not sure what the target endianness was - if ( $chunk4 == ".sd\000" - || $chunk4 == ".snd" - || $chunk4 == "\000ds." - || $chunk4 == "dns." ) - { - return 'audio/basic'; - } - if ( $chunk3 == "MM\000" ) { - return 'image/tiff'; - } - if ( $chunk2 == 'MZ' ) { - return 'application/x-msdownload'; - } - if ( $chunk8 == "\x89PNG\x0d\x0a\x1a\x0a" ) { - return 'image/x-png'; // [sic] - } - if ( strlen( $chunk ) >= 5 ) { - $byte2 = ord( $chunk[2] ); - $byte4 = ord( $chunk[4] ); - if ( $byte2 >= 3 && $byte2 <= 31 && $byte4 == 0 && $chunk2 == 'JG' ) { - return 'image/x-jg'; - } - } - // More endian confusion? - if ( $chunk4 == 'MROF' ) { - return 'audio/x-aiff'; - } - $chunk4_8 = substr( $chunk, 8, 4 ); - if ( $chunk4 == 'FORM' && ( $chunk4_8 == 'AIFF' || $chunk4_8 == 'AIFC' ) ) { - return 'audio/x-aiff'; - } - if ( $chunk4 == 'RIFF' && $chunk4_8 == 'AVI ' ) { - return 'video/avi'; - } - if ( $chunk4 == "\x00\x00\x01\xb3" || $chunk4 == "\x00\x00\x01\xba" ) { - return 'video/mpeg'; - } - if ( $chunk4 == "\001\000\000\000" - && substr( $chunk, 40, 4 ) == ' EMF' ) - { - return 'image/x-emf'; - } - if ( $chunk4 == "\xd7\xcd\xc6\x9a" ) { - return 'image/x-wmf'; - } - if ( $chunk4 == "\xca\xfe\xba\xbe" ) { - return 'application/java'; - } - if ( $chunk2 == 'PK' ) { - return 'application/x-zip-compressed'; - } - if ( $chunk2 == "\x1f\x9d" ) { - return 'application/x-compressed'; - } - if ( $chunk2 == "\x1f\x8b" ) { - return 'application/x-gzip-compressed'; - } - // Skip redundant check for ZIP - if ( $chunk5 == "MThd\000" ) { - return 'audio/mid'; - } - if ( $chunk4 == '%PDF' ) { - return 'application/pdf'; - } - return false; - } - - /** - * Do heuristic checks on the bulk of the data sample. - * Search for HTML tags. - * @param $version - * @param $chunk - * @return array - */ - protected function sampleData( $version, $chunk ) { - $found = []; - $counters = [ - 'ctrl' => 0, - 'high' => 0, - 'low' => 0, - 'lf' => 0, - 'cr' => 0, - 'ff' => 0 - ]; - $htmlTags = [ - 'html', - 'head', - 'title', - 'body', - 'script', - 'a href', - 'pre', - 'img', - 'plaintext', - 'table' - ]; - $rdfUrl = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'; - $rdfPurl = 'http://purl.org/rss/1.0/'; - $xbmMagic1 = '#define'; - $xbmMagic2 = '_width'; - $xbmMagic3 = '_bits'; - $binhexMagic = 'converted with BinHex'; - $chunkLength = strlen( $chunk ); - - for ( $offset = 0; $offset < $chunkLength; $offset++ ) { - $curChar = $chunk[$offset]; - if ( $curChar == "\x0a" ) { - $counters['lf']++; - continue; - } elseif ( $curChar == "\x0d" ) { - $counters['cr']++; - continue; - } elseif ( $curChar == "\x0c" ) { - $counters['ff']++; - continue; - } elseif ( $curChar == "\t" ) { - $counters['low']++; - continue; - } elseif ( ord( $curChar ) < 32 ) { - $counters['ctrl']++; - continue; - } elseif ( ord( $curChar ) >= 128 ) { - $counters['high']++; - continue; - } - - $counters['low']++; - if ( $curChar == '<' ) { - // XML - $remainder = substr( $chunk, $offset + 1 ); - if ( !strncasecmp( $remainder, '?XML', 4 ) ) { - $nextChar = substr( $chunk, $offset + 5, 1 ); - if ( $nextChar == ':' || $nextChar == ' ' || $nextChar == "\t" ) { - $found['xml'] = true; - } - } - // Scriptlet (JSP) - if ( !strncasecmp( $remainder, 'SCRIPTLET', 9 ) ) { - $found['scriptlet'] = true; - break; - } - // HTML - foreach ( $htmlTags as $tag ) { - if ( !strncasecmp( $remainder, $tag, strlen( $tag ) ) ) { - $found['html'] = true; - } - } - // Skip broken check for additional tags (HR etc.) - - // CHANNEL replaced by RSS, RDF and FEED in IE 7 - if ( $version < 'ie07' ) { - if ( !strncasecmp( $remainder, 'CHANNEL', 7 ) ) { - $found['cdf'] = true; - } - } else { - // RSS - if ( !strncasecmp( $remainder, 'RSS', 3 ) ) { - $found['rss'] = true; - break; // return from SampleData - } - if ( !strncasecmp( $remainder, 'rdf:RDF', 7 ) ) { - $found['rdf-tag'] = true; - // no break - } - if ( !strncasecmp( $remainder, 'FEED', 4 ) ) { - $found['atom'] = true; - break; - } - } - continue; - } - // Skip broken check for --> - - // RSS URL checks - // For some reason both URLs must appear before it is recognised - $remainder = substr( $chunk, $offset ); - if ( !strncasecmp( $remainder, $rdfUrl, strlen( $rdfUrl ) ) ) { - $found['rdf-url'] = true; - if ( isset( $found['rdf-tag'] ) - && isset( $found['rdf-purl'] ) ) // [sic] - { - break; - } - continue; - } - - if ( !strncasecmp( $remainder, $rdfPurl, strlen( $rdfPurl ) ) ) { - if ( isset( $found['rdf-tag'] ) - && isset( $found['rdf-url'] ) ) // [sic] - { - break; - } - continue; - } - - // XBM checks - if ( !strncasecmp( $remainder, $xbmMagic1, strlen( $xbmMagic1 ) ) ) { - $found['xbm1'] = true; - continue; - } - if ( $curChar == '_' ) { - if ( isset( $found['xbm2'] ) ) { - if ( !strncasecmp( $remainder, $xbmMagic3, strlen( $xbmMagic3 ) ) ) { - $found['xbm'] = true; - break; - } - } elseif ( isset( $found['xbm1'] ) ) { - if ( !strncasecmp( $remainder, $xbmMagic2, strlen( $xbmMagic2 ) ) ) { - $found['xbm2'] = true; - } - } - } - - // BinHex - if ( !strncmp( $remainder, $binhexMagic, strlen( $binhexMagic ) ) ) { - $found['binhex'] = true; - } - } - return [ 'found' => $found, 'counters' => $counters ]; - } - - /** - * @param $version - * @param $type - * @return int|string - */ - protected function getDataFormat( $version, $type ) { - $types = $this->typeTable[$version]; - if ( $type == '(null)' || strval( $type ) === '' ) { - return 'ambiguous'; - } - foreach ( $types as $format => $list ) { - if ( in_array( $type, $list ) ) { - return $format; - } - } - return 'unknown'; - } -} diff --git a/includes/libs/XmlTypeCheck.php b/includes/libs/XmlTypeCheck.php deleted file mode 100644 index f057140a6b..0000000000 --- a/includes/libs/XmlTypeCheck.php +++ /dev/null @@ -1,347 +0,0 @@ - '', - ]; - - /** - * @param string $input a filename or string containing the XML element - * @param callable $filterCallback (optional) - * Function to call to do additional custom validity checks from the - * SAX element handler event. This gives you access to the element - * namespace, name, attributes, and text contents. - * Filter should return 'true' to toggle on $this->filterMatch - * @param bool $isFile (optional) indicates if the first parameter is a - * filename (default, true) or if it is a string (false) - * @param array $options list of additional parsing options: - * processing_instruction_handler: Callback for xml_set_processing_instruction_handler - */ - function __construct( $input, $filterCallback = null, $isFile = true, $options = [] ) { - $this->filterCallback = $filterCallback; - $this->parserOptions = array_merge( $this->parserOptions, $options ); - $this->validateFromInput( $input, $isFile ); - } - - /** - * Alternative constructor: from filename - * - * @param string $fname the filename of an XML document - * @param callable $filterCallback (optional) - * Function to call to do additional custom validity checks from the - * SAX element handler event. This gives you access to the element - * namespace, name, and attributes, but not to text contents. - * Filter should return 'true' to toggle on $this->filterMatch - * @return XmlTypeCheck - */ - public static function newFromFilename( $fname, $filterCallback = null ) { - return new self( $fname, $filterCallback, true ); - } - - /** - * Alternative constructor: from string - * - * @param string $string a string containing an XML element - * @param callable $filterCallback (optional) - * Function to call to do additional custom validity checks from the - * SAX element handler event. This gives you access to the element - * namespace, name, and attributes, but not to text contents. - * Filter should return 'true' to toggle on $this->filterMatch - * @return XmlTypeCheck - */ - public static function newFromString( $string, $filterCallback = null ) { - return new self( $string, $filterCallback, false ); - } - - /** - * Get the root element. Simple accessor to $rootElement - * - * @return string - */ - public function getRootElement() { - return $this->rootElement; - } - - /** - * @param string $fname the filename - */ - private function validateFromInput( $xml, $isFile ) { - $reader = new XMLReader(); - if ( $isFile ) { - $s = $reader->open( $xml, null, LIBXML_NOERROR | LIBXML_NOWARNING ); - } else { - $s = $reader->XML( $xml, null, LIBXML_NOERROR | LIBXML_NOWARNING ); - } - if ( $s !== true ) { - // Couldn't open the XML - $this->wellFormed = false; - } else { - $oldDisable = libxml_disable_entity_loader( true ); - $reader->setParserProperty( XMLReader::SUBST_ENTITIES, true ); - try { - $this->validate( $reader ); - } catch ( Exception $e ) { - // Calling this malformed, because we didn't parse the whole - // thing. Maybe just an external entity refernce. - $this->wellFormed = false; - $reader->close(); - libxml_disable_entity_loader( $oldDisable ); - throw $e; - } - $reader->close(); - libxml_disable_entity_loader( $oldDisable ); - } - } - - private function readNext( XMLReader $reader ) { - set_error_handler( [ $this, 'XmlErrorHandler' ] ); - $ret = $reader->read(); - restore_error_handler(); - return $ret; - } - - public function XmlErrorHandler( $errno, $errstr ) { - $this->wellFormed = false; - } - - private function validate( $reader ) { - - // First, move through anything that isn't an element, and - // handle any processing instructions with the callback - do { - if ( !$this->readNext( $reader ) ) { - // Hit the end of the document before any elements - $this->wellFormed = false; - return; - } - if ( $reader->nodeType === XMLReader::PI ) { - $this->processingInstructionHandler( $reader->name, $reader->value ); - } - } while ( $reader->nodeType != XMLReader::ELEMENT ); - - // Process the rest of the document - do { - switch ( $reader->nodeType ) { - case XMLReader::ELEMENT: - $name = $this->expandNS( - $reader->name, - $reader->namespaceURI - ); - if ( $this->rootElement === '' ) { - $this->rootElement = $name; - } - $empty = $reader->isEmptyElement; - $attrs = $this->getAttributesArray( $reader ); - $this->elementOpen( $name, $attrs ); - if ( $empty ) { - $this->elementClose(); - } - break; - - case XMLReader::END_ELEMENT: - $this->elementClose(); - break; - - case XMLReader::WHITESPACE: - case XMLReader::SIGNIFICANT_WHITESPACE: - case XMLReader::CDATA: - case XMLReader::TEXT: - $this->elementData( $reader->value ); - break; - - case XMLReader::ENTITY_REF: - // Unexpanded entity (maybe external?), - // don't send to the filter (xml_parse didn't) - break; - - case XMLReader::COMMENT: - // Don't send to the filter (xml_parse didn't) - break; - - case XMLReader::PI: - // Processing instructions can happen after the header too - $this->processingInstructionHandler( - $reader->name, - $reader->value - ); - break; - default: - // One of DOC, DOC_TYPE, ENTITY, END_ENTITY, - // NOTATION, or XML_DECLARATION - // xml_parse didn't send these to the filter, so we won't. - } - - } while ( $this->readNext( $reader ) ); - - if ( $this->stackDepth !== 0 ) { - $this->wellFormed = false; - } elseif ( $this->wellFormed === null ) { - $this->wellFormed = true; - } - - } - - /** - * Get all of the attributes for an XMLReader's current node - * @param $r XMLReader - * @return array of attributes - */ - private function getAttributesArray( XMLReader $r ) { - $attrs = []; - while ( $r->moveToNextAttribute() ) { - if ( $r->namespaceURI === 'http://www.w3.org/2000/xmlns/' ) { - // XMLReader treats xmlns attributes as normal - // attributes, while xml_parse doesn't - continue; - } - $name = $this->expandNS( $r->name, $r->namespaceURI ); - $attrs[$name] = $r->value; - } - return $attrs; - } - - /** - * @param $name element or attribute name, maybe with a full or short prefix - * @param $namespaceURI the namespaceURI - * @return string the name prefixed with namespaceURI - */ - private function expandNS( $name, $namespaceURI ) { - if ( $namespaceURI ) { - $parts = explode( ':', $name ); - $localname = array_pop( $parts ); - return "$namespaceURI:$localname"; - } - return $name; - } - - /** - * @param $name - * @param $attribs - */ - private function elementOpen( $name, $attribs ) { - $this->elementDataContext[] = [ $name, $attribs ]; - $this->elementData[] = ''; - $this->stackDepth++; - } - - /** - */ - private function elementClose() { - list( $name, $attribs ) = array_pop( $this->elementDataContext ); - $data = array_pop( $this->elementData ); - $this->stackDepth--; - $callbackReturn = false; - - if ( is_callable( $this->filterCallback ) ) { - $callbackReturn = call_user_func( - $this->filterCallback, - $name, - $attribs, - $data - ); - } - if ( $callbackReturn ) { - // Filter hit! - $this->filterMatch = true; - $this->filterMatchType = $callbackReturn; - } - } - - /** - * @param $data - */ - private function elementData( $data ) { - // Collect any data here, and we'll run the callback in elementClose - $this->elementData[ $this->stackDepth - 1 ] .= trim( $data ); - } - - /** - * @param $target - * @param $data - */ - private function processingInstructionHandler( $target, $data ) { - $callbackReturn = false; - if ( $this->parserOptions['processing_instruction_handler'] ) { - $callbackReturn = call_user_func( - $this->parserOptions['processing_instruction_handler'], - $target, - $data - ); - } - if ( $callbackReturn ) { - // Filter hit! - $this->filterMatch = true; - $this->filterMatchType = $callbackReturn; - } - } -} diff --git a/includes/libs/mime/IEContentAnalyzer.php b/includes/libs/mime/IEContentAnalyzer.php new file mode 100644 index 0000000000..0d1e527ba6 --- /dev/null +++ b/includes/libs/mime/IEContentAnalyzer.php @@ -0,0 +1,851 @@ + [ + 'text/plain', + 'application/octet-stream', + 'application/x-netcdf', // [sic] + ], + 'text' /*3*/ => [ + 'text/richtext', 'image/x-bitmap', 'application/postscript', 'application/base64', + 'application/macbinhex40', 'application/x-cdf', 'text/scriptlet' + ], + 'binary' /*4*/ => [ + 'application/pdf', 'audio/x-aiff', 'audio/basic', 'audio/wav', 'image/gif', + 'image/pjpeg', 'image/jpeg', 'image/tiff', 'image/x-png', 'image/png', 'image/bmp', + 'image/x-jg', 'image/x-art', 'image/x-emf', 'image/x-wmf', 'video/avi', + 'video/x-msvideo', 'video/mpeg', 'application/x-compressed', + 'application/x-zip-compressed', 'application/x-gzip-compressed', 'application/java', + 'application/x-msdownload' + ], + 'html' /*5*/ => [ 'text/html' ], + ]; + + /** + * Changes to the type table in later versions of IE + */ + protected $addedTypes = [ + 'ie07' => [ + 'text' => [ 'text/xml', 'application/xml' ] + ], + ]; + + /** + * An approximation of the "Content Type" values in HKEY_CLASSES_ROOT in a + * typical Windows installation. + * + * Used for extension to MIME type mapping if detection fails. + */ + protected $registry = [ + '.323' => 'text/h323', + '.3g2' => 'video/3gpp2', + '.3gp' => 'video/3gpp', + '.3gp2' => 'video/3gpp2', + '.3gpp' => 'video/3gpp', + '.aac' => 'audio/aac', + '.ac3' => 'audio/ac3', + '.accda' => 'application/msaccess', + '.accdb' => 'application/msaccess', + '.accdc' => 'application/msaccess', + '.accde' => 'application/msaccess', + '.accdr' => 'application/msaccess', + '.accdt' => 'application/msaccess', + '.ade' => 'application/msaccess', + '.adp' => 'application/msaccess', + '.adts' => 'audio/aac', + '.ai' => 'application/postscript', + '.aif' => 'audio/aiff', + '.aifc' => 'audio/aiff', + '.aiff' => 'audio/aiff', + '.amc' => 'application/x-mpeg', + '.application' => 'application/x-ms-application', + '.asf' => 'video/x-ms-asf', + '.asx' => 'video/x-ms-asf', + '.au' => 'audio/basic', + '.avi' => 'video/avi', + '.bmp' => 'image/bmp', + '.caf' => 'audio/x-caf', + '.cat' => 'application/vnd.ms-pki.seccat', + '.cbo' => 'application/sha', + '.cdda' => 'audio/aiff', + '.cer' => 'application/x-x509-ca-cert', + '.conf' => 'text/plain', + '.crl' => 'application/pkix-crl', + '.crt' => 'application/x-x509-ca-cert', + '.css' => 'text/css', + '.csv' => 'application/vnd.ms-excel', + '.der' => 'application/x-x509-ca-cert', + '.dib' => 'image/bmp', + '.dif' => 'video/x-dv', + '.dll' => 'application/x-msdownload', + '.doc' => 'application/msword', + '.docm' => 'application/vnd.ms-word.document.macroEnabled.12', + '.docx' => 'application/vnd.openxmlformats-officedocument.wordprocessingml.document', + '.dot' => 'application/msword', + '.dotm' => 'application/vnd.ms-word.template.macroEnabled.12', + '.dotx' => 'application/vnd.openxmlformats-officedocument.wordprocessingml.template', + '.dv' => 'video/x-dv', + '.dwfx' => 'model/vnd.dwfx+xps', + '.edn' => 'application/vnd.adobe.edn', + '.eml' => 'message/rfc822', + '.eps' => 'application/postscript', + '.etd' => 'application/x-ebx', + '.exe' => 'application/x-msdownload', + '.fdf' => 'application/vnd.fdf', + '.fif' => 'application/fractals', + '.gif' => 'image/gif', + '.gsm' => 'audio/x-gsm', + '.hqx' => 'application/mac-binhex40', + '.hta' => 'application/hta', + '.htc' => 'text/x-component', + '.htm' => 'text/html', + '.html' => 'text/html', + '.htt' => 'text/webviewhtml', + '.hxa' => 'application/xml', + '.hxc' => 'application/xml', + '.hxd' => 'application/octet-stream', + '.hxe' => 'application/xml', + '.hxf' => 'application/xml', + '.hxh' => 'application/octet-stream', + '.hxi' => 'application/octet-stream', + '.hxk' => 'application/xml', + '.hxq' => 'application/octet-stream', + '.hxr' => 'application/octet-stream', + '.hxs' => 'application/octet-stream', + '.hxt' => 'application/xml', + '.hxv' => 'application/xml', + '.hxw' => 'application/octet-stream', + '.ico' => 'image/x-icon', + '.iii' => 'application/x-iphone', + '.ins' => 'application/x-internet-signup', + '.iqy' => 'text/x-ms-iqy', + '.isp' => 'application/x-internet-signup', + '.jfif' => 'image/jpeg', + '.jnlp' => 'application/x-java-jnlp-file', + '.jpe' => 'image/jpeg', + '.jpeg' => 'image/jpeg', + '.jpg' => 'image/jpeg', + '.jtx' => 'application/x-jtx+xps', + '.latex' => 'application/x-latex', + '.log' => 'text/plain', + '.m1v' => 'video/mpeg', + '.m2v' => 'video/mpeg', + '.m3u' => 'audio/x-mpegurl', + '.mac' => 'image/x-macpaint', + '.man' => 'application/x-troff-man', + '.mda' => 'application/msaccess', + '.mdb' => 'application/msaccess', + '.mde' => 'application/msaccess', + '.mfp' => 'application/x-shockwave-flash', + '.mht' => 'message/rfc822', + '.mhtml' => 'message/rfc822', + '.mid' => 'audio/mid', + '.midi' => 'audio/mid', + '.mod' => 'video/mpeg', + '.mov' => 'video/quicktime', + '.mp2' => 'video/mpeg', + '.mp2v' => 'video/mpeg', + '.mp3' => 'audio/mpeg', + '.mp4' => 'video/mp4', + '.mpa' => 'video/mpeg', + '.mpe' => 'video/mpeg', + '.mpeg' => 'video/mpeg', + '.mpf' => 'application/vnd.ms-mediapackage', + '.mpg' => 'video/mpeg', + '.mpv2' => 'video/mpeg', + '.mqv' => 'video/quicktime', + '.NMW' => 'application/nmwb', + '.nws' => 'message/rfc822', + '.odc' => 'text/x-ms-odc', + '.ols' => 'application/vnd.ms-publisher', + '.p10' => 'application/pkcs10', + '.p12' => 'application/x-pkcs12', + '.p7b' => 'application/x-pkcs7-certificates', + '.p7c' => 'application/pkcs7-mime', + '.p7m' => 'application/pkcs7-mime', + '.p7r' => 'application/x-pkcs7-certreqresp', + '.p7s' => 'application/pkcs7-signature', + '.pct' => 'image/pict', + '.pdf' => 'application/pdf', + '.pdx' => 'application/vnd.adobe.pdx', + '.pfx' => 'application/x-pkcs12', + '.pic' => 'image/pict', + '.pict' => 'image/pict', + '.pinstall' => 'application/x-picasa-detect', + '.pko' => 'application/vnd.ms-pki.pko', + '.png' => 'image/png', + '.pnt' => 'image/x-macpaint', + '.pntg' => 'image/x-macpaint', + '.pot' => 'application/vnd.ms-powerpoint', + '.potm' => 'application/vnd.ms-powerpoint.template.macroEnabled.12', + '.potx' => 'application/vnd.openxmlformats-officedocument.presentationml.template', + '.ppa' => 'application/vnd.ms-powerpoint', + '.ppam' => 'application/vnd.ms-powerpoint.addin.macroEnabled.12', + '.pps' => 'application/vnd.ms-powerpoint', + '.ppsm' => 'application/vnd.ms-powerpoint.slideshow.macroEnabled.12', + '.ppsx' => 'application/vnd.openxmlformats-officedocument.presentationml.slideshow', + '.ppt' => 'application/vnd.ms-powerpoint', + '.pptm' => 'application/vnd.ms-powerpoint.presentation.macroEnabled.12', + '.pptx' => 'application/vnd.openxmlformats-officedocument.presentationml.presentation', + '.prf' => 'application/pics-rules', + '.ps' => 'application/postscript', + '.pub' => 'application/vnd.ms-publisher', + '.pwz' => 'application/vnd.ms-powerpoint', + '.py' => 'text/plain', + '.pyw' => 'text/plain', + '.qht' => 'text/x-html-insertion', + '.qhtm' => 'text/x-html-insertion', + '.qt' => 'video/quicktime', + '.qti' => 'image/x-quicktime', + '.qtif' => 'image/x-quicktime', + '.qtl' => 'application/x-quicktimeplayer', + '.rat' => 'application/rat-file', + '.rmf' => 'application/vnd.adobe.rmf', + '.rmi' => 'audio/mid', + '.rqy' => 'text/x-ms-rqy', + '.rtf' => 'application/msword', + '.sct' => 'text/scriptlet', + '.sd2' => 'audio/x-sd2', + '.sdp' => 'application/sdp', + '.shtml' => 'text/html', + '.sit' => 'application/x-stuffit', + '.sldm' => 'application/vnd.ms-powerpoint.slide.macroEnabled.12', + '.sldx' => 'application/vnd.openxmlformats-officedocument.presentationml.slide', + '.slk' => 'application/vnd.ms-excel', + '.snd' => 'audio/basic', + '.so' => 'application/x-apachemodule', + '.sol' => 'text/plain', + '.sor' => 'text/plain', + '.spc' => 'application/x-pkcs7-certificates', + '.spl' => 'application/futuresplash', + '.sst' => 'application/vnd.ms-pki.certstore', + '.stl' => 'application/vnd.ms-pki.stl', + '.swf' => 'application/x-shockwave-flash', + '.thmx' => 'application/vnd.ms-officetheme', + '.tif' => 'image/tiff', + '.tiff' => 'image/tiff', + '.txt' => 'text/plain', + '.uls' => 'text/iuls', + '.vcf' => 'text/x-vcard', + '.vdx' => 'application/vnd.ms-visio.viewer', + '.vsd' => 'application/vnd.ms-visio.viewer', + '.vss' => 'application/vnd.ms-visio.viewer', + '.vst' => 'application/vnd.ms-visio.viewer', + '.vsx' => 'application/vnd.ms-visio.viewer', + '.vtx' => 'application/vnd.ms-visio.viewer', + '.wav' => 'audio/wav', + '.wax' => 'audio/x-ms-wax', + '.wbk' => 'application/msword', + '.wdp' => 'image/vnd.ms-photo', + '.wiz' => 'application/msword', + '.wm' => 'video/x-ms-wm', + '.wma' => 'audio/x-ms-wma', + '.wmd' => 'application/x-ms-wmd', + '.wmv' => 'video/x-ms-wmv', + '.wmx' => 'video/x-ms-wmx', + '.wmz' => 'application/x-ms-wmz', + '.wpl' => 'application/vnd.ms-wpl', + '.wsc' => 'text/scriptlet', + '.wvx' => 'video/x-ms-wvx', + '.xaml' => 'application/xaml+xml', + '.xbap' => 'application/x-ms-xbap', + '.xdp' => 'application/vnd.adobe.xdp+xml', + '.xfdf' => 'application/vnd.adobe.xfdf', + '.xht' => 'application/xhtml+xml', + '.xhtml' => 'application/xhtml+xml', + '.xla' => 'application/vnd.ms-excel', + '.xlam' => 'application/vnd.ms-excel.addin.macroEnabled.12', + '.xlk' => 'application/vnd.ms-excel', + '.xll' => 'application/vnd.ms-excel', + '.xlm' => 'application/vnd.ms-excel', + '.xls' => 'application/vnd.ms-excel', + '.xlsb' => 'application/vnd.ms-excel.sheet.binary.macroEnabled.12', + '.xlsm' => 'application/vnd.ms-excel.sheet.macroEnabled.12', + '.xlsx' => 'application/vnd.openxmlformats-officedocument.spreadsheetml.sheet', + '.xlt' => 'application/vnd.ms-excel', + '.xltm' => 'application/vnd.ms-excel.template.macroEnabled.12', + '.xltx' => 'application/vnd.openxmlformats-officedocument.spreadsheetml.template', + '.xlw' => 'application/vnd.ms-excel', + '.xml' => 'text/xml', + '.xps' => 'application/vnd.ms-xpsdocument', + '.xsl' => 'text/xml', + ]; + + /** + * IE versions which have been analysed to bring you this class, and for + * which some substantive difference exists. These will appear as keys + * in the return value of getRealMimesFromData(). The names are chosen to sort correctly. + */ + protected $versions = [ 'ie05', 'ie06', 'ie07', 'ie07.strict', 'ie07.nohtml' ]; + + /** + * Type table with versions expanded + */ + protected $typeTable = []; + + /** constructor */ + function __construct() { + // Construct versioned type arrays from the base type array plus additions + $types = $this->baseTypeTable; + foreach ( $this->versions as $version ) { + if ( isset( $this->addedTypes[$version] ) ) { + foreach ( $this->addedTypes[$version] as $format => $addedTypes ) { + $types[$format] = array_merge( $types[$format], $addedTypes ); + } + } + $this->typeTable[$version] = $types; + } + } + + /** + * Get the MIME types from getMimesFromData(), but convert the result from IE's + * idiosyncratic private types into something other apps will understand. + * + * @param string $fileName the file name (unused at present) + * @param string $chunk the first 256 bytes of the file + * @param string $proposed the MIME type proposed by the server + * + * @return Array: map of IE version to detected MIME type + */ + public function getRealMimesFromData( $fileName, $chunk, $proposed ) { + $types = $this->getMimesFromData( $fileName, $chunk, $proposed ); + $types = array_map( [ $this, 'translateMimeType' ], $types ); + return $types; + } + + /** + * Translate a MIME type from IE's idiosyncratic private types into + * more commonly understood type strings + * @param $type + * @return string + */ + public function translateMimeType( $type ) { + static $table = [ + 'image/pjpeg' => 'image/jpeg', + 'image/x-png' => 'image/png', + 'image/x-wmf' => 'application/x-msmetafile', + 'image/bmp' => 'image/x-bmp', + 'application/x-zip-compressed' => 'application/zip', + 'application/x-compressed' => 'application/x-compress', + 'application/x-gzip-compressed' => 'application/x-gzip', + 'audio/mid' => 'audio/midi', + ]; + if ( isset( $table[$type] ) ) { + $type = $table[$type]; + } + return $type; + } + + /** + * Get the untranslated MIME types for all known versions + * + * @param string $fileName the file name (unused at present) + * @param string $chunk the first 256 bytes of the file + * @param string $proposed the MIME type proposed by the server + * + * @return Array: map of IE version to detected MIME type + */ + public function getMimesFromData( $fileName, $chunk, $proposed ) { + $types = []; + foreach ( $this->versions as $version ) { + $types[$version] = $this->getMimeTypeForVersion( $version, $fileName, $chunk, $proposed ); + } + return $types; + } + + /** + * Get the MIME type for a given named version + * @param $version + * @param $fileName + * @param $chunk + * @param $proposed + * @return bool|string + */ + protected function getMimeTypeForVersion( $version, $fileName, $chunk, $proposed ) { + // Strip text after a semicolon + $semiPos = strpos( $proposed, ';' ); + if ( $semiPos !== false ) { + $proposed = substr( $proposed, 0, $semiPos ); + } + + $proposedFormat = $this->getDataFormat( $version, $proposed ); + if ( $proposedFormat == 'unknown' + && $proposed != 'multipart/mixed' + && $proposed != 'multipart/x-mixed-replace' ) + { + return $proposed; + } + if ( strval( $chunk ) === '' ) { + return $proposed; + } + + // Truncate chunk at 255 bytes + $chunk = substr( $chunk, 0, 255 ); + + // IE does the Check*Headers() calls last, and instead does the following image + // type checks by directly looking for the magic numbers. What I do here should + // have the same effect since the magic number checks are identical in both cases. + $result = $this->sampleData( $version, $chunk ); + $sampleFound = $result['found']; + $counters = $result['counters']; + $binaryType = $this->checkBinaryHeaders( $version, $chunk ); + $textType = $this->checkTextHeaders( $version, $chunk ); + + if ( $proposed == 'text/html' && isset( $sampleFound['html'] ) ) { + return 'text/html'; + } + if ( $proposed == 'image/gif' && $binaryType == 'image/gif' ) { + return 'image/gif'; + } + if ( ( $proposed == 'image/pjpeg' || $proposed == 'image/jpeg' ) + && $binaryType == 'image/pjpeg' ) + { + return $proposed; + } + // PNG check added in IE 7 + if ( $version >= 'ie07' + && ( $proposed == 'image/x-png' || $proposed == 'image/png' ) + && $binaryType == 'image/x-png' ) + { + return $proposed; + } + + // CDF was removed in IE 7 so it won't be in $sampleFound for later versions + if ( isset( $sampleFound['cdf'] ) ) { + return 'application/x-cdf'; + } + + // RSS and Atom were added in IE 7 so they won't be in $sampleFound for + // previous versions + if ( isset( $sampleFound['rss'] ) ) { + return 'application/rss+xml'; + } + if ( isset( $sampleFound['rdf-tag'] ) + && isset( $sampleFound['rdf-url'] ) + && isset( $sampleFound['rdf-purl'] ) ) + { + return 'application/rss+xml'; + } + if ( isset( $sampleFound['atom'] ) ) { + return 'application/atom+xml'; + } + + if ( isset( $sampleFound['xml'] ) ) { + // TODO: I'm not sure under what circumstances this flag is enabled + if ( strpos( $version, 'strict' ) !== false ) { + if ( $proposed == 'text/html' || $proposed == 'text/xml' ) { + return 'text/xml'; + } + } else { + return 'text/xml'; + } + } + if ( isset( $sampleFound['html'] ) ) { + // TODO: I'm not sure under what circumstances this flag is enabled + if ( strpos( $version, 'nohtml' ) !== false ) { + if ( $proposed == 'text/plain' ) { + return 'text/html'; + } + } else { + return 'text/html'; + } + } + if ( isset( $sampleFound['xbm'] ) ) { + return 'image/x-bitmap'; + } + if ( isset( $sampleFound['binhex'] ) ) { + return 'application/macbinhex40'; + } + if ( isset( $sampleFound['scriptlet'] ) ) { + if ( strpos( $version, 'strict' ) !== false ) { + if ( $proposed == 'text/plain' || $proposed == 'text/scriptlet' ) { + return 'text/scriptlet'; + } + } else { + return 'text/scriptlet'; + } + } + + // Freaky heuristics to determine if the data is text or binary + // The heuristic is of course broken for non-ASCII text + if ( $counters['ctrl'] != 0 && ( $counters['ff'] + $counters['low'] ) + < ( $counters['ctrl'] + $counters['high'] ) * 16 ) + { + $kindOfBinary = true; + $type = $binaryType ? $binaryType : $textType; + if ( $type === false ) { + $type = 'application/octet-stream'; + } + } else { + $kindOfBinary = false; + $type = $textType ? $textType : $binaryType; + if ( $type === false ) { + $type = 'text/plain'; + } + } + + // Check if the output format is ambiguous + // This generally means that detection failed, real types aren't ambiguous + $detectedFormat = $this->getDataFormat( $version, $type ); + if ( $detectedFormat != 'ambiguous' ) { + return $type; + } + + if ( $proposedFormat != 'ambiguous' ) { + // FormatAgreesWithData() + if ( $proposedFormat == 'text' && !$kindOfBinary ) { + return $proposed; + } + if ( $proposedFormat == 'binary' && $kindOfBinary ) { + return $proposed; + } + if ( $proposedFormat == 'html' ) { + return $proposed; + } + } + + // Find a MIME type by searching the registry for the file extension. + $dotPos = strrpos( $fileName, '.' ); + if ( $dotPos === false ) { + return $type; + } + $ext = substr( $fileName, $dotPos ); + if ( isset( $this->registry[$ext] ) ) { + return $this->registry[$ext]; + } + + // TODO: If the extension has an application registered to it, IE will return + // application/octet-stream. We'll skip that, so we could erroneously + // return text/plain or application/x-netcdf where application/octet-stream + // would be correct. + + return $type; + } + + /** + * Check for text headers at the start of the chunk + * Confirmed same in 5 and 7. + * @param $version + * @param $chunk + * @return bool|string + */ + private function checkTextHeaders( $version, $chunk ) { + $chunk2 = substr( $chunk, 0, 2 ); + $chunk4 = substr( $chunk, 0, 4 ); + $chunk5 = substr( $chunk, 0, 5 ); + if ( $chunk4 == '%PDF' ) { + return 'application/pdf'; + } + if ( $chunk2 == '%!' ) { + return 'application/postscript'; + } + if ( $chunk5 == '{\\rtf' ) { + return 'text/richtext'; + } + if ( $chunk5 == 'begin' ) { + return 'application/base64'; + } + return false; + } + + /** + * Check for binary headers at the start of the chunk + * Confirmed same in 5 and 7. + * @param $version + * @param $chunk + * @return bool|string + */ + private function checkBinaryHeaders( $version, $chunk ) { + $chunk2 = substr( $chunk, 0, 2 ); + $chunk3 = substr( $chunk, 0, 3 ); + $chunk4 = substr( $chunk, 0, 4 ); + $chunk5 = substr( $chunk, 0, 5 ); + $chunk5uc = strtoupper( $chunk5 ); + $chunk8 = substr( $chunk, 0, 8 ); + if ( $chunk5uc == 'GIF87' || $chunk5uc == 'GIF89' ) { + return 'image/gif'; + } + if ( $chunk2 == "\xff\xd8" ) { + return 'image/pjpeg'; // actually plain JPEG but this is what IE returns + } + + if ( $chunk2 == 'BM' + && substr( $chunk, 6, 2 ) == "\000\000" + && substr( $chunk, 8, 2 ) == "\000\000" ) + { + return 'image/bmp'; // another non-standard MIME + } + if ( $chunk4 == 'RIFF' + && substr( $chunk, 8, 4 ) == 'WAVE' ) + { + return 'audio/wav'; + } + // These were integer literals in IE + // Perhaps the author was not sure what the target endianness was + if ( $chunk4 == ".sd\000" + || $chunk4 == ".snd" + || $chunk4 == "\000ds." + || $chunk4 == "dns." ) + { + return 'audio/basic'; + } + if ( $chunk3 == "MM\000" ) { + return 'image/tiff'; + } + if ( $chunk2 == 'MZ' ) { + return 'application/x-msdownload'; + } + if ( $chunk8 == "\x89PNG\x0d\x0a\x1a\x0a" ) { + return 'image/x-png'; // [sic] + } + if ( strlen( $chunk ) >= 5 ) { + $byte2 = ord( $chunk[2] ); + $byte4 = ord( $chunk[4] ); + if ( $byte2 >= 3 && $byte2 <= 31 && $byte4 == 0 && $chunk2 == 'JG' ) { + return 'image/x-jg'; + } + } + // More endian confusion? + if ( $chunk4 == 'MROF' ) { + return 'audio/x-aiff'; + } + $chunk4_8 = substr( $chunk, 8, 4 ); + if ( $chunk4 == 'FORM' && ( $chunk4_8 == 'AIFF' || $chunk4_8 == 'AIFC' ) ) { + return 'audio/x-aiff'; + } + if ( $chunk4 == 'RIFF' && $chunk4_8 == 'AVI ' ) { + return 'video/avi'; + } + if ( $chunk4 == "\x00\x00\x01\xb3" || $chunk4 == "\x00\x00\x01\xba" ) { + return 'video/mpeg'; + } + if ( $chunk4 == "\001\000\000\000" + && substr( $chunk, 40, 4 ) == ' EMF' ) + { + return 'image/x-emf'; + } + if ( $chunk4 == "\xd7\xcd\xc6\x9a" ) { + return 'image/x-wmf'; + } + if ( $chunk4 == "\xca\xfe\xba\xbe" ) { + return 'application/java'; + } + if ( $chunk2 == 'PK' ) { + return 'application/x-zip-compressed'; + } + if ( $chunk2 == "\x1f\x9d" ) { + return 'application/x-compressed'; + } + if ( $chunk2 == "\x1f\x8b" ) { + return 'application/x-gzip-compressed'; + } + // Skip redundant check for ZIP + if ( $chunk5 == "MThd\000" ) { + return 'audio/mid'; + } + if ( $chunk4 == '%PDF' ) { + return 'application/pdf'; + } + return false; + } + + /** + * Do heuristic checks on the bulk of the data sample. + * Search for HTML tags. + * @param $version + * @param $chunk + * @return array + */ + protected function sampleData( $version, $chunk ) { + $found = []; + $counters = [ + 'ctrl' => 0, + 'high' => 0, + 'low' => 0, + 'lf' => 0, + 'cr' => 0, + 'ff' => 0 + ]; + $htmlTags = [ + 'html', + 'head', + 'title', + 'body', + 'script', + 'a href', + 'pre', + 'img', + 'plaintext', + 'table' + ]; + $rdfUrl = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'; + $rdfPurl = 'http://purl.org/rss/1.0/'; + $xbmMagic1 = '#define'; + $xbmMagic2 = '_width'; + $xbmMagic3 = '_bits'; + $binhexMagic = 'converted with BinHex'; + $chunkLength = strlen( $chunk ); + + for ( $offset = 0; $offset < $chunkLength; $offset++ ) { + $curChar = $chunk[$offset]; + if ( $curChar == "\x0a" ) { + $counters['lf']++; + continue; + } elseif ( $curChar == "\x0d" ) { + $counters['cr']++; + continue; + } elseif ( $curChar == "\x0c" ) { + $counters['ff']++; + continue; + } elseif ( $curChar == "\t" ) { + $counters['low']++; + continue; + } elseif ( ord( $curChar ) < 32 ) { + $counters['ctrl']++; + continue; + } elseif ( ord( $curChar ) >= 128 ) { + $counters['high']++; + continue; + } + + $counters['low']++; + if ( $curChar == '<' ) { + // XML + $remainder = substr( $chunk, $offset + 1 ); + if ( !strncasecmp( $remainder, '?XML', 4 ) ) { + $nextChar = substr( $chunk, $offset + 5, 1 ); + if ( $nextChar == ':' || $nextChar == ' ' || $nextChar == "\t" ) { + $found['xml'] = true; + } + } + // Scriptlet (JSP) + if ( !strncasecmp( $remainder, 'SCRIPTLET', 9 ) ) { + $found['scriptlet'] = true; + break; + } + // HTML + foreach ( $htmlTags as $tag ) { + if ( !strncasecmp( $remainder, $tag, strlen( $tag ) ) ) { + $found['html'] = true; + } + } + // Skip broken check for additional tags (HR etc.) + + // CHANNEL replaced by RSS, RDF and FEED in IE 7 + if ( $version < 'ie07' ) { + if ( !strncasecmp( $remainder, 'CHANNEL', 7 ) ) { + $found['cdf'] = true; + } + } else { + // RSS + if ( !strncasecmp( $remainder, 'RSS', 3 ) ) { + $found['rss'] = true; + break; // return from SampleData + } + if ( !strncasecmp( $remainder, 'rdf:RDF', 7 ) ) { + $found['rdf-tag'] = true; + // no break + } + if ( !strncasecmp( $remainder, 'FEED', 4 ) ) { + $found['atom'] = true; + break; + } + } + continue; + } + // Skip broken check for --> + + // RSS URL checks + // For some reason both URLs must appear before it is recognised + $remainder = substr( $chunk, $offset ); + if ( !strncasecmp( $remainder, $rdfUrl, strlen( $rdfUrl ) ) ) { + $found['rdf-url'] = true; + if ( isset( $found['rdf-tag'] ) + && isset( $found['rdf-purl'] ) ) // [sic] + { + break; + } + continue; + } + + if ( !strncasecmp( $remainder, $rdfPurl, strlen( $rdfPurl ) ) ) { + if ( isset( $found['rdf-tag'] ) + && isset( $found['rdf-url'] ) ) // [sic] + { + break; + } + continue; + } + + // XBM checks + if ( !strncasecmp( $remainder, $xbmMagic1, strlen( $xbmMagic1 ) ) ) { + $found['xbm1'] = true; + continue; + } + if ( $curChar == '_' ) { + if ( isset( $found['xbm2'] ) ) { + if ( !strncasecmp( $remainder, $xbmMagic3, strlen( $xbmMagic3 ) ) ) { + $found['xbm'] = true; + break; + } + } elseif ( isset( $found['xbm1'] ) ) { + if ( !strncasecmp( $remainder, $xbmMagic2, strlen( $xbmMagic2 ) ) ) { + $found['xbm2'] = true; + } + } + } + + // BinHex + if ( !strncmp( $remainder, $binhexMagic, strlen( $binhexMagic ) ) ) { + $found['binhex'] = true; + } + } + return [ 'found' => $found, 'counters' => $counters ]; + } + + /** + * @param $version + * @param $type + * @return int|string + */ + protected function getDataFormat( $version, $type ) { + $types = $this->typeTable[$version]; + if ( $type == '(null)' || strval( $type ) === '' ) { + return 'ambiguous'; + } + foreach ( $types as $format => $list ) { + if ( in_array( $type, $list ) ) { + return $format; + } + } + return 'unknown'; + } +} diff --git a/includes/libs/mime/MimeAnalyzer.php b/includes/libs/mime/MimeAnalyzer.php new file mode 100644 index 0000000000..5f4d7c9ef4 --- /dev/null +++ b/includes/libs/mime/MimeAnalyzer.php @@ -0,0 +1,1166 @@ + ext + * map. Each line contains a MIME type followed by a space separated list of + * extensions. If multiple extensions for a single MIME type exist or if + * multiple MIME types exist for a single extension then in most cases + * MediaWiki assumes that the first extension following the MIME type is the + * canonical extension, and the first time a MIME type appears for a certain + * extension is considered the canonical MIME type. + * + * (Note that appending the type file list to the end of self::$wellKnownTypes + * sucks because you can't redefine canonical types. This could be fixed by + * appending self::$wellKnownTypes behind type file list, but who knows + * what will break? In practice this probably isn't a problem anyway -- Bryan) + */ + protected static $wellKnownTypes = <<typeFile = $params['typeFile']; + $this->infoFile = $params['infoFile']; + $this->xmlTypes = $params['xmlTypes']; + $this->initCallback = isset( $params['initCallback'] ) + ? $params['initCallback'] + : null; + $this->detectCallback = isset( $params['detectCallback'] ) + ? $params['detectCallback'] + : null; + $this->guessCallback = isset( $params['guessCallback'] ) + ? $params['guessCallback'] + : null; + $this->extCallback = isset( $params['extCallback'] ) + ? $params['extCallback'] + : null; + $this->logger = isset( $params['logger'] ) + ? $params['logger'] + : new \Psr\Log\NullLogger(); + + $this->loadFiles(); + } + + protected function loadFiles() { + /** + * --- load mime.types --- + */ + + # Allow media handling extensions adding MIME-types and MIME-info + if ( $this->initCallback ) { + call_user_func( $this->initCallback, $this ); + } + + $types = self::$wellKnownTypes; + + $mimeTypeFile = $this->typeFile; + if ( $mimeTypeFile ) { + if ( is_file( $mimeTypeFile ) && is_readable( $mimeTypeFile ) ) { + $this->logger->info( __METHOD__ . ": loading mime types from $mimeTypeFile\n" ); + $types .= "\n"; + $types .= file_get_contents( $mimeTypeFile ); + } else { + $this->logger->info( __METHOD__ . ": can't load mime types from $mimeTypeFile\n" ); + } + } else { + $this->logger->info( __METHOD__ . + ": no mime types file defined, using built-ins only.\n" ); + } + + $types .= "\n" . $this->extraTypes; + + $types = str_replace( [ "\r\n", "\n\r", "\n\n", "\r\r", "\r" ], "\n", $types ); + $types = str_replace( "\t", " ", $types ); + + $this->mimetoExt = []; + $this->mExtToMime = []; + + $lines = explode( "\n", $types ); + foreach ( $lines as $s ) { + $s = trim( $s ); + if ( empty( $s ) ) { + continue; + } + if ( strpos( $s, '#' ) === 0 ) { + continue; + } + + $s = strtolower( $s ); + $i = strpos( $s, ' ' ); + + if ( $i === false ) { + continue; + } + + $mime = substr( $s, 0, $i ); + $ext = trim( substr( $s, $i + 1 ) ); + + if ( empty( $ext ) ) { + continue; + } + + if ( !empty( $this->mimetoExt[$mime] ) ) { + $this->mimetoExt[$mime] .= ' ' . $ext; + } else { + $this->mimetoExt[$mime] = $ext; + } + + $extensions = explode( ' ', $ext ); + + foreach ( $extensions as $e ) { + $e = trim( $e ); + if ( empty( $e ) ) { + continue; + } + + if ( !empty( $this->mExtToMime[$e] ) ) { + $this->mExtToMime[$e] .= ' ' . $mime; + } else { + $this->mExtToMime[$e] = $mime; + } + } + } + + /** + * --- load mime.info --- + */ + + $mimeInfoFile = $this->infoFile; + + $info = self::$wellKnownInfo; + + if ( $mimeInfoFile ) { + if ( is_file( $mimeInfoFile ) && is_readable( $mimeInfoFile ) ) { + $this->logger->info( __METHOD__ . ": loading mime info from $mimeInfoFile\n" ); + $info .= "\n"; + $info .= file_get_contents( $mimeInfoFile ); + } else { + $this->logger->info( __METHOD__ . ": can't load mime info from $mimeInfoFile\n" ); + } + } else { + $this->logger->info( __METHOD__ . + ": no mime info file defined, using built-ins only.\n" ); + } + + $info .= "\n" . $this->extraInfo; + + $info = str_replace( [ "\r\n", "\n\r", "\n\n", "\r\r", "\r" ], "\n", $info ); + $info = str_replace( "\t", " ", $info ); + + $this->mimeTypeAliases = []; + $this->mediaTypes = []; + + $lines = explode( "\n", $info ); + foreach ( $lines as $s ) { + $s = trim( $s ); + if ( empty( $s ) ) { + continue; + } + if ( strpos( $s, '#' ) === 0 ) { + continue; + } + + $s = strtolower( $s ); + $i = strpos( $s, ' ' ); + + if ( $i === false ) { + continue; + } + + # print "processing MIME INFO line $s
"; + + $match = []; + if ( preg_match( '!\[\s*(\w+)\s*\]!', $s, $match ) ) { + $s = preg_replace( '!\[\s*(\w+)\s*\]!', '', $s ); + $mtype = trim( strtoupper( $match[1] ) ); + } else { + $mtype = MEDIATYPE_UNKNOWN; + } + + $m = explode( ' ', $s ); + + if ( !isset( $this->mediaTypes[$mtype] ) ) { + $this->mediaTypes[$mtype] = []; + } + + foreach ( $m as $mime ) { + $mime = trim( $mime ); + if ( empty( $mime ) ) { + continue; + } + + $this->mediaTypes[$mtype][] = $mime; + } + + if ( count( $m ) > 1 ) { + $main = $m[0]; + $mCount = count( $m ); + for ( $i = 1; $i < $mCount; $i += 1 ) { + $mime = $m[$i]; + $this->mimeTypeAliases[$mime] = $main; + } + } + } + } + + public function setLogger( LoggerInterface $logger ) { + $this->logger = $logger; + } + + /** + * Adds to the list mapping MIME to file extensions. + * As an extension author, you are encouraged to submit patches to + * MediaWiki's core to add new MIME types to mime.types. + * @param string $types + */ + public function addExtraTypes( $types ) { + $this->extraTypes .= "\n" . $types; + } + + /** + * Adds to the list mapping MIME to media type. + * As an extension author, you are encouraged to submit patches to + * MediaWiki's core to add new MIME info to mime.info. + * @param string $info + */ + public function addExtraInfo( $info ) { + $this->extraInfo .= "\n" . $info; + } + + /** + * Returns a list of file extensions for a given MIME type as a space + * separated string or null if the MIME type was unrecognized. Resolves + * MIME type aliases. + * + * @param string $mime + * @return string|null + */ + public function getExtensionsForType( $mime ) { + $mime = strtolower( $mime ); + + // Check the mime-to-ext map + if ( isset( $this->mimetoExt[$mime] ) ) { + return $this->mimetoExt[$mime]; + } + + // Resolve the MIME type to the canonical type + if ( isset( $this->mimeTypeAliases[$mime] ) ) { + $mime = $this->mimeTypeAliases[$mime]; + if ( isset( $this->mimetoExt[$mime] ) ) { + return $this->mimetoExt[$mime]; + } + } + + return null; + } + + /** + * Returns a list of MIME types for a given file extension as a space + * separated string or null if the extension was unrecognized. + * + * @param string $ext + * @return string|null + */ + public function getTypesForExtension( $ext ) { + $ext = strtolower( $ext ); + + $r = isset( $this->mExtToMime[$ext] ) ? $this->mExtToMime[$ext] : null; + return $r; + } + + /** + * Returns a single MIME type for a given file extension or null if unknown. + * This is always the first type from the list returned by getTypesForExtension($ext). + * + * @param string $ext + * @return string|null + */ + public function guessTypesForExtension( $ext ) { + $m = $this->getTypesForExtension( $ext ); + if ( is_null( $m ) ) { + return null; + } + + // TODO: Check if this is needed; strtok( $m, ' ' ) should be sufficient + $m = trim( $m ); + $m = preg_replace( '/\s.*$/', '', $m ); + + return $m; + } + + /** + * Tests if the extension matches the given MIME type. Returns true if a + * match was found, null if the MIME type is unknown, and false if the + * MIME type is known but no matches where found. + * + * @param string $extension + * @param string $mime + * @return bool|null + */ + public function isMatchingExtension( $extension, $mime ) { + $ext = $this->getExtensionsForType( $mime ); + + if ( !$ext ) { + return null; // Unknown MIME type + } + + $ext = explode( ' ', $ext ); + + $extension = strtolower( $extension ); + return in_array( $extension, $ext ); + } + + /** + * Returns true if the MIME type is known to represent an image format + * supported by the PHP GD library. + * + * @param string $mime + * + * @return bool + */ + public function isPHPImageType( $mime ) { + // As defined by imagegetsize and image_type_to_mime + static $types = [ + 'image/gif', 'image/jpeg', 'image/png', + 'image/x-bmp', 'image/xbm', 'image/tiff', + 'image/jp2', 'image/jpeg2000', 'image/iff', + 'image/xbm', 'image/x-xbitmap', + 'image/vnd.wap.wbmp', 'image/vnd.xiff', + 'image/x-photoshop', + 'application/x-shockwave-flash', + ]; + + return in_array( $mime, $types ); + } + + /** + * Returns true if the extension represents a type which can + * be reliably detected from its content. Use this to determine + * whether strict content checks should be applied to reject + * invalid uploads; if we can't identify the type we won't + * be able to say if it's invalid. + * + * @todo Be more accurate when using fancy MIME detector plugins; + * right now this is the bare minimum getimagesize() list. + * @param string $extension + * @return bool + */ + function isRecognizableExtension( $extension ) { + static $types = [ + // Types recognized by getimagesize() + 'gif', 'jpeg', 'jpg', 'png', 'swf', 'psd', + 'bmp', 'tiff', 'tif', 'jpc', 'jp2', + 'jpx', 'jb2', 'swc', 'iff', 'wbmp', + 'xbm', + + // Formats we recognize magic numbers for + 'djvu', 'ogx', 'ogg', 'ogv', 'oga', 'spx', + 'mid', 'pdf', 'wmf', 'xcf', 'webm', 'mkv', 'mka', + 'webp', + + // XML formats we sure hope we recognize reliably + 'svg', + ]; + return in_array( strtolower( $extension ), $types ); + } + + /** + * Improves a MIME type using the file extension. Some file formats are very generic, + * so their MIME type is not very meaningful. A more useful MIME type can be derived + * by looking at the file extension. Typically, this method would be called on the + * result of guessMimeType(). + * + * @param string $mime The MIME type, typically guessed from a file's content. + * @param string $ext The file extension, as taken from the file name + * + * @return string The MIME type + */ + public function improveTypeFromExtension( $mime, $ext ) { + if ( $mime === 'unknown/unknown' ) { + if ( $this->isRecognizableExtension( $ext ) ) { + $this->logger->info( __METHOD__ . ': refusing to guess mime type for .' . + "$ext file, we should have recognized it\n" ); + } else { + // Not something we can detect, so simply + // trust the file extension + $mime = $this->guessTypesForExtension( $ext ); + } + } elseif ( $mime === 'application/x-opc+zip' ) { + if ( $this->isMatchingExtension( $ext, $mime ) ) { + // A known file extension for an OPC file, + // find the proper MIME type for that file extension + $mime = $this->guessTypesForExtension( $ext ); + } else { + $this->logger->info( __METHOD__ . + ": refusing to guess better type for $mime file, " . + ".$ext is not a known OPC extension.\n" ); + $mime = 'application/zip'; + } + } elseif ( $mime === 'text/plain' && $this->findMediaType( ".$ext" ) === MEDIATYPE_TEXT ) { + // Textual types are sometimes not recognized properly. + // If detected as text/plain, and has an extension which is textual + // improve to the extension's type. For example, csv and json are often + // misdetected as text/plain. + $mime = $this->guessTypesForExtension( $ext ); + } + + # Media handling extensions can improve the MIME detected + $callback = $this->extCallback; + if ( $callback ) { + $callback( $this, $ext, $mime /* by reference */ ); + } + + if ( isset( $this->mimeTypeAliases[$mime] ) ) { + $mime = $this->mimeTypeAliases[$mime]; + } + + $this->logger->info( __METHOD__ . ": improved mime type for .$ext: $mime\n" ); + return $mime; + } + + /** + * MIME type detection. This uses detectMimeType to detect the MIME type + * of the file, but applies additional checks to determine some well known + * file formats that may be missed or misinterpreted by the default MIME + * detection (namely XML based formats like XHTML or SVG, as well as ZIP + * based formats like OPC/ODF files). + * + * @param string $file The file to check + * @param string|bool $ext The file extension, or true (default) to extract + * it from the filename. Set it to false to ignore the extension. DEPRECATED! + * Set to false, use improveTypeFromExtension($mime, $ext) later to improve MIME type. + * + * @return string The MIME type of $file + */ + public function guessMimeType( $file, $ext = true ) { + if ( $ext ) { // TODO: make $ext default to false. Or better, remove it. + $this->logger->info( __METHOD__ . + ": WARNING: use of the \$ext parameter is deprecated. " . + "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); + } + + $mime = $this->doGuessMimeType( $file, $ext ); + + if ( !$mime ) { + $this->logger->info( __METHOD__ . + ": internal type detection failed for $file (.$ext)...\n" ); + $mime = $this->detectMimeType( $file, $ext ); + } + + if ( isset( $this->mimeTypeAliases[$mime] ) ) { + $mime = $this->mimeTypeAliases[$mime]; + } + + $this->logger->info( __METHOD__ . ": guessed mime type of $file: $mime\n" ); + return $mime; + } + + /** + * Guess the MIME type from the file contents. + * + * @todo Remove $ext param + * + * @param string $file + * @param mixed $ext + * @return bool|string + * @throws UnexpectedValueException + */ + private function doGuessMimeType( $file, $ext ) { + // Read a chunk of the file + MediaWiki\suppressWarnings(); + $f = fopen( $file, 'rb' ); + MediaWiki\restoreWarnings(); + + if ( !$f ) { + return 'unknown/unknown'; + } + + $fsize = filesize( $file ); + if ( $fsize === false ) { + return 'unknown/unknown'; + } + + $head = fread( $f, 1024 ); + $tailLength = min( 65558, $fsize ); // 65558 = maximum size of a zip EOCDR + if ( fseek( $f, -1 * $tailLength, SEEK_END ) === -1 ) { + throw new UnexpectedValueException( + "Seeking $tailLength bytes from EOF failed in " . __METHOD__ ); + } + $tail = $tailLength ? fread( $f, $tailLength ) : ''; + fclose( $f ); + + $this->logger->info( __METHOD__ . + ": analyzing head and tail of $file for magic numbers.\n" ); + + // Hardcode a few magic number checks... + $headers = [ + // Multimedia... + 'MThd' => 'audio/midi', + 'OggS' => 'application/ogg', + + // Image formats... + // Note that WMF may have a bare header, no magic number. + "\x01\x00\x09\x00" => 'application/x-msmetafile', // Possibly prone to false positives? + "\xd7\xcd\xc6\x9a" => 'application/x-msmetafile', + '%PDF' => 'application/pdf', + 'gimp xcf' => 'image/x-xcf', + + // Some forbidden fruit... + 'MZ' => 'application/octet-stream', // DOS/Windows executable + "\xca\xfe\xba\xbe" => 'application/octet-stream', // Mach-O binary + "\x7fELF" => 'application/octet-stream', // ELF binary + ]; + + foreach ( $headers as $magic => $candidate ) { + if ( strncmp( $head, $magic, strlen( $magic ) ) == 0 ) { + $this->logger->info( __METHOD__ . + ": magic header in $file recognized as $candidate\n" ); + return $candidate; + } + } + + /* Look for WebM and Matroska files */ + if ( strncmp( $head, pack( "C4", 0x1a, 0x45, 0xdf, 0xa3 ), 4 ) == 0 ) { + $doctype = strpos( $head, "\x42\x82" ); + if ( $doctype ) { + // Next byte is datasize, then data (sizes larger than 1 byte are stupid muxers) + $data = substr( $head, $doctype + 3, 8 ); + if ( strncmp( $data, "matroska", 8 ) == 0 ) { + $this->logger->info( __METHOD__ . ": recognized file as video/x-matroska\n" ); + return "video/x-matroska"; + } elseif ( strncmp( $data, "webm", 4 ) == 0 ) { + $this->logger->info( __METHOD__ . ": recognized file as video/webm\n" ); + return "video/webm"; + } + } + $this->logger->info( __METHOD__ . ": unknown EBML file\n" ); + return "unknown/unknown"; + } + + /* Look for WebP */ + if ( strncmp( $head, "RIFF", 4 ) == 0 && + strncmp( substr( $head, 8, 7 ), "WEBPVP8", 7 ) == 0 + ) { + $this->logger->info( __METHOD__ . ": recognized file as image/webp\n" ); + return "image/webp"; + } + + /** + * Look for PHP. Check for this before HTML/XML... Warning: this is a + * heuristic, and won't match a file with a lot of non-PHP before. It + * will also match text files which could be PHP. :) + * + * @todo FIXME: For this reason, the check is probably useless -- an attacker + * could almost certainly just pad the file with a lot of nonsense to + * circumvent the check in any case where it would be a security + * problem. On the other hand, it causes harmful false positives (bug + * 16583). The heuristic has been cut down to exclude three-character + * strings like "logger->info( __METHOD__ . ": recognized $file as application/x-php\n" ); + return 'application/x-php'; + } + + /** + * look for XML formats (XHTML and SVG) + */ + $xml = new XmlTypeCheck( $file ); + if ( $xml->wellFormed ) { + $xmlTypes = $this->xmlTypes; + if ( isset( $xmlTypes[$xml->getRootElement()] ) ) { + return $xmlTypes[$xml->getRootElement()]; + } else { + return 'application/xml'; + } + } + + /** + * look for shell scripts + */ + $script_type = null; + + # detect by shebang + if ( substr( $head, 0, 2 ) == "#!" ) { + $script_type = "ASCII"; + } elseif ( substr( $head, 0, 5 ) == "\xef\xbb\xbf#!" ) { + $script_type = "UTF-8"; + } elseif ( substr( $head, 0, 7 ) == "\xfe\xff\x00#\x00!" ) { + $script_type = "UTF-16BE"; + } elseif ( substr( $head, 0, 7 ) == "\xff\xfe#\x00!" ) { + $script_type = "UTF-16LE"; + } + + if ( $script_type ) { + if ( $script_type !== "UTF-8" && $script_type !== "ASCII" ) { + // Quick and dirty fold down to ASCII! + $pack = [ 'UTF-16BE' => 'n*', 'UTF-16LE' => 'v*' ]; + $chars = unpack( $pack[$script_type], substr( $head, 2 ) ); + $head = ''; + foreach ( $chars as $codepoint ) { + if ( $codepoint < 128 ) { + $head .= chr( $codepoint ); + } else { + $head .= '?'; + } + } + } + + $match = []; + + if ( preg_match( '%/?([^\s]+/)(\w+)%', $head, $match ) ) { + $mime = "application/x-{$match[2]}"; + $this->logger->info( __METHOD__ . ": shell script recognized as $mime\n" ); + return $mime; + } + } + + // Check for ZIP variants (before getimagesize) + if ( strpos( $tail, "PK\x05\x06" ) !== false ) { + $this->logger->info( __METHOD__ . ": ZIP header present in $file\n" ); + return $this->detectZipType( $head, $tail, $ext ); + } + + MediaWiki\suppressWarnings(); + $gis = getimagesize( $file ); + MediaWiki\restoreWarnings(); + + if ( $gis && isset( $gis['mime'] ) ) { + $mime = $gis['mime']; + $this->logger->info( __METHOD__ . ": getimagesize detected $file as $mime\n" ); + return $mime; + } + + # Media handling extensions can guess the MIME by content + # It's intentionally here so that if core is wrong about a type (false positive), + # people will hopefully nag and submit patches :) + $mime = false; + # Some strings by reference for performance - assuming well-behaved hooks + $callback = $this->guessCallback; + if ( $callback ) { + $callback( $this, $head, $tail, $file, $mime /* by reference */ ); + }; + + return $mime; + } + + /** + * Detect application-specific file type of a given ZIP file from its + * header data. Currently works for OpenDocument and OpenXML types... + * If can't tell, returns 'application/zip'. + * + * @param string $header Some reasonably-sized chunk of file header + * @param string|null $tail The tail of the file + * @param string|bool $ext The file extension, or true to extract it from the filename. + * Set it to false (default) to ignore the extension. DEPRECATED! Set to false, + * use improveTypeFromExtension($mime, $ext) later to improve MIME type. + * + * @return string + */ + function detectZipType( $header, $tail = null, $ext = false ) { + if ( $ext ) { # TODO: remove $ext param + $this->logger->info( __METHOD__ . + ": WARNING: use of the \$ext parameter is deprecated. " . + "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); + } + + $mime = 'application/zip'; + $opendocTypes = [ + 'chart-template', + 'chart', + 'formula-template', + 'formula', + 'graphics-template', + 'graphics', + 'image-template', + 'image', + 'presentation-template', + 'presentation', + 'spreadsheet-template', + 'spreadsheet', + 'text-template', + 'text-master', + 'text-web', + 'text' ]; + + // http://lists.oasis-open.org/archives/office/200505/msg00006.html + $types = '(?:' . implode( '|', $opendocTypes ) . ')'; + $opendocRegex = "/^mimetype(application\/vnd\.oasis\.opendocument\.$types)/"; + + $openxmlRegex = "/^\[Content_Types\].xml/"; + + if ( preg_match( $opendocRegex, substr( $header, 30 ), $matches ) ) { + $mime = $matches[1]; + $this->logger->info( __METHOD__ . ": detected $mime from ZIP archive\n" ); + } elseif ( preg_match( $openxmlRegex, substr( $header, 30 ) ) ) { + $mime = "application/x-opc+zip"; + # TODO: remove the block below, as soon as improveTypeFromExtension is used everywhere + if ( $ext !== true && $ext !== false ) { + /** This is the mode used by getPropsFromPath + * These MIME's are stored in the database, where we don't really want + * x-opc+zip, because we use it only for internal purposes + */ + if ( $this->isMatchingExtension( $ext, $mime ) ) { + /* A known file extension for an OPC file, + * find the proper mime type for that file extension + */ + $mime = $this->guessTypesForExtension( $ext ); + } else { + $mime = "application/zip"; + } + } + $this->logger->info( __METHOD__ . + ": detected an Open Packaging Conventions archive: $mime\n" ); + } elseif ( substr( $header, 0, 8 ) == "\xd0\xcf\x11\xe0\xa1\xb1\x1a\xe1" && + ( $headerpos = strpos( $tail, "PK\x03\x04" ) ) !== false && + preg_match( $openxmlRegex, substr( $tail, $headerpos + 30 ) ) ) { + if ( substr( $header, 512, 4 ) == "\xEC\xA5\xC1\x00" ) { + $mime = "application/msword"; + } + switch ( substr( $header, 512, 6 ) ) { + case "\xEC\xA5\xC1\x00\x0E\x00": + case "\xEC\xA5\xC1\x00\x1C\x00": + case "\xEC\xA5\xC1\x00\x43\x00": + $mime = "application/vnd.ms-powerpoint"; + break; + case "\xFD\xFF\xFF\xFF\x10\x00": + case "\xFD\xFF\xFF\xFF\x1F\x00": + case "\xFD\xFF\xFF\xFF\x22\x00": + case "\xFD\xFF\xFF\xFF\x23\x00": + case "\xFD\xFF\xFF\xFF\x28\x00": + case "\xFD\xFF\xFF\xFF\x29\x00": + case "\xFD\xFF\xFF\xFF\x10\x02": + case "\xFD\xFF\xFF\xFF\x1F\x02": + case "\xFD\xFF\xFF\xFF\x22\x02": + case "\xFD\xFF\xFF\xFF\x23\x02": + case "\xFD\xFF\xFF\xFF\x28\x02": + case "\xFD\xFF\xFF\xFF\x29\x02": + $mime = "application/vnd.msexcel"; + break; + } + + $this->logger->info( __METHOD__ . + ": detected a MS Office document with OPC trailer\n" ); + } else { + $this->logger->info( __METHOD__ . ": unable to identify type of ZIP archive\n" ); + } + return $mime; + } + + /** + * Internal MIME type detection. Detection is done using the fileinfo + * extension if it is available. It can be overriden by callback, which could + * use an external program, for example. If detection fails and $ext is not false, + * the MIME type is guessed from the file extension, using guessTypesForExtension. + * + * If the MIME type is still unknown, getimagesize is used to detect the + * MIME type if the file is an image. If no MIME type can be determined, + * this function returns 'unknown/unknown'. + * + * @param string $file The file to check + * @param string|bool $ext The file extension, or true (default) to extract it from the filename. + * Set it to false to ignore the extension. DEPRECATED! Set to false, use + * improveTypeFromExtension($mime, $ext) later to improve MIME type. + * + * @return string The MIME type of $file + */ + private function detectMimeType( $file, $ext = true ) { + /** @todo Make $ext default to false. Or better, remove it. */ + if ( $ext ) { + $this->logger->info( __METHOD__ . + ": WARNING: use of the \$ext parameter is deprecated. " + . "Use improveTypeFromExtension(\$mime, \$ext) instead.\n" ); + } + + $callback = $this->detectCallback; + $m = null; + if ( $callback ) { + $m = $callback( $file ); + } elseif ( function_exists( "finfo_open" ) && function_exists( "finfo_file" ) ) { + $mime_magic_resource = finfo_open( FILEINFO_MIME ); + + if ( $mime_magic_resource ) { + $m = finfo_file( $mime_magic_resource, $file ); + finfo_close( $mime_magic_resource ); + } else { + $this->logger->info( __METHOD__ . + ": finfo_open failed on " . FILEINFO_MIME . "!\n" ); + } + } else { + $this->logger->info( __METHOD__ . ": no magic mime detector found!\n" ); + } + + if ( $m ) { + # normalize + $m = preg_replace( '![;, ].*$!', '', $m ); # strip charset, etc + $m = trim( $m ); + $m = strtolower( $m ); + + if ( strpos( $m, 'unknown' ) !== false ) { + $m = null; + } else { + $this->logger->info( __METHOD__ . ": magic mime type of $file: $m\n" ); + return $m; + } + } + + // If desired, look at extension as a fallback. + if ( $ext === true ) { + $i = strrpos( $file, '.' ); + $ext = strtolower( $i ? substr( $file, $i + 1 ) : '' ); + } + if ( $ext ) { + if ( $this->isRecognizableExtension( $ext ) ) { + $this->logger->info( __METHOD__ . ": refusing to guess mime type for .$ext file, " + . "we should have recognized it\n" ); + } else { + $m = $this->guessTypesForExtension( $ext ); + if ( $m ) { + $this->logger->info( __METHOD__ . ": extension mime type of $file: $m\n" ); + return $m; + } + } + } + + // Unknown type + $this->logger->info( __METHOD__ . ": failed to guess mime type for $file!\n" ); + return 'unknown/unknown'; + } + + /** + * Determine the media type code for a file, using its MIME type, name and + * possibly its contents. + * + * This function relies on the findMediaType(), mapping extensions and MIME + * types to media types. + * + * @todo analyse file if need be + * @todo look at multiple extension, separately and together. + * + * @param string $path Full path to the image file, in case we have to look at the contents + * (if null, only the MIME type is used to determine the media type code). + * @param string $mime MIME type. If null it will be guessed using guessMimeType. + * + * @return string A value to be used with the MEDIATYPE_xxx constants. + */ + function getMediaType( $path = null, $mime = null ) { + if ( !$mime && !$path ) { + return MEDIATYPE_UNKNOWN; + } + + // If MIME type is unknown, guess it + if ( !$mime ) { + $mime = $this->guessMimeType( $path, false ); + } + + // Special code for ogg - detect if it's video (theora), + // else label it as sound. + if ( $mime == 'application/ogg' && file_exists( $path ) ) { + + // Read a chunk of the file + $f = fopen( $path, "rt" ); + if ( !$f ) { + return MEDIATYPE_UNKNOWN; + } + $head = fread( $f, 256 ); + fclose( $f ); + + $head = str_replace( 'ffmpeg2theora', '', strtolower( $head ) ); + + // This is an UGLY HACK, file should be parsed correctly + if ( strpos( $head, 'theora' ) !== false ) { + return MEDIATYPE_VIDEO; + } elseif ( strpos( $head, 'vorbis' ) !== false ) { + return MEDIATYPE_AUDIO; + } elseif ( strpos( $head, 'flac' ) !== false ) { + return MEDIATYPE_AUDIO; + } elseif ( strpos( $head, 'speex' ) !== false ) { + return MEDIATYPE_AUDIO; + } else { + return MEDIATYPE_MULTIMEDIA; + } + } + + $type = null; + // Check for entry for full MIME type + if ( $mime ) { + $type = $this->findMediaType( $mime ); + if ( $type !== MEDIATYPE_UNKNOWN ) { + return $type; + } + } + + // Check for entry for file extension + if ( $path ) { + $i = strrpos( $path, '.' ); + $e = strtolower( $i ? substr( $path, $i + 1 ) : '' ); + + // TODO: look at multi-extension if this fails, parse from full path + $type = $this->findMediaType( '.' . $e ); + if ( $type !== MEDIATYPE_UNKNOWN ) { + return $type; + } + } + + // Check major MIME type + if ( $mime ) { + $i = strpos( $mime, '/' ); + if ( $i !== false ) { + $major = substr( $mime, 0, $i ); + $type = $this->findMediaType( $major ); + if ( $type !== MEDIATYPE_UNKNOWN ) { + return $type; + } + } + } + + if ( !$type ) { + $type = MEDIATYPE_UNKNOWN; + } + + return $type; + } + + /** + * Returns a media code matching the given MIME type or file extension. + * File extensions are represented by a string starting with a dot (.) to + * distinguish them from MIME types. + * + * This function relies on the mapping defined by $this->mMediaTypes + * @access private + * @param string $extMime + * @return int|string + */ + function findMediaType( $extMime ) { + if ( strpos( $extMime, '.' ) === 0 ) { + // If it's an extension, look up the MIME types + $m = $this->getTypesForExtension( substr( $extMime, 1 ) ); + if ( !$m ) { + return MEDIATYPE_UNKNOWN; + } + + $m = explode( ' ', $m ); + } else { + // Normalize MIME type + if ( isset( $this->mimeTypeAliases[$extMime] ) ) { + $extMime = $this->mimeTypeAliases[$extMime]; + } + + $m = [ $extMime ]; + } + + foreach ( $m as $mime ) { + foreach ( $this->mediaTypes as $type => $codes ) { + if ( in_array( $mime, $codes, true ) ) { + return $type; + } + } + } + + return MEDIATYPE_UNKNOWN; + } + + /** + * Get the MIME types that various versions of Internet Explorer would + * detect from a chunk of the content. + * + * @param string $fileName The file name (unused at present) + * @param string $chunk The first 256 bytes of the file + * @param string $proposed The MIME type proposed by the server + * @return array + */ + public function getIEMimeTypes( $fileName, $chunk, $proposed ) { + $ca = $this->getIEContentAnalyzer(); + return $ca->getRealMimesFromData( $fileName, $chunk, $proposed ); + } + + /** + * Get a cached instance of IEContentAnalyzer + * + * @return IEContentAnalyzer + */ + protected function getIEContentAnalyzer() { + if ( is_null( $this->IEAnalyzer ) ) { + $this->IEAnalyzer = new IEContentAnalyzer; + } + return $this->IEAnalyzer; + } +} diff --git a/includes/libs/mime/XmlTypeCheck.php b/includes/libs/mime/XmlTypeCheck.php new file mode 100644 index 0000000000..f057140a6b --- /dev/null +++ b/includes/libs/mime/XmlTypeCheck.php @@ -0,0 +1,347 @@ + '', + ]; + + /** + * @param string $input a filename or string containing the XML element + * @param callable $filterCallback (optional) + * Function to call to do additional custom validity checks from the + * SAX element handler event. This gives you access to the element + * namespace, name, attributes, and text contents. + * Filter should return 'true' to toggle on $this->filterMatch + * @param bool $isFile (optional) indicates if the first parameter is a + * filename (default, true) or if it is a string (false) + * @param array $options list of additional parsing options: + * processing_instruction_handler: Callback for xml_set_processing_instruction_handler + */ + function __construct( $input, $filterCallback = null, $isFile = true, $options = [] ) { + $this->filterCallback = $filterCallback; + $this->parserOptions = array_merge( $this->parserOptions, $options ); + $this->validateFromInput( $input, $isFile ); + } + + /** + * Alternative constructor: from filename + * + * @param string $fname the filename of an XML document + * @param callable $filterCallback (optional) + * Function to call to do additional custom validity checks from the + * SAX element handler event. This gives you access to the element + * namespace, name, and attributes, but not to text contents. + * Filter should return 'true' to toggle on $this->filterMatch + * @return XmlTypeCheck + */ + public static function newFromFilename( $fname, $filterCallback = null ) { + return new self( $fname, $filterCallback, true ); + } + + /** + * Alternative constructor: from string + * + * @param string $string a string containing an XML element + * @param callable $filterCallback (optional) + * Function to call to do additional custom validity checks from the + * SAX element handler event. This gives you access to the element + * namespace, name, and attributes, but not to text contents. + * Filter should return 'true' to toggle on $this->filterMatch + * @return XmlTypeCheck + */ + public static function newFromString( $string, $filterCallback = null ) { + return new self( $string, $filterCallback, false ); + } + + /** + * Get the root element. Simple accessor to $rootElement + * + * @return string + */ + public function getRootElement() { + return $this->rootElement; + } + + /** + * @param string $fname the filename + */ + private function validateFromInput( $xml, $isFile ) { + $reader = new XMLReader(); + if ( $isFile ) { + $s = $reader->open( $xml, null, LIBXML_NOERROR | LIBXML_NOWARNING ); + } else { + $s = $reader->XML( $xml, null, LIBXML_NOERROR | LIBXML_NOWARNING ); + } + if ( $s !== true ) { + // Couldn't open the XML + $this->wellFormed = false; + } else { + $oldDisable = libxml_disable_entity_loader( true ); + $reader->setParserProperty( XMLReader::SUBST_ENTITIES, true ); + try { + $this->validate( $reader ); + } catch ( Exception $e ) { + // Calling this malformed, because we didn't parse the whole + // thing. Maybe just an external entity refernce. + $this->wellFormed = false; + $reader->close(); + libxml_disable_entity_loader( $oldDisable ); + throw $e; + } + $reader->close(); + libxml_disable_entity_loader( $oldDisable ); + } + } + + private function readNext( XMLReader $reader ) { + set_error_handler( [ $this, 'XmlErrorHandler' ] ); + $ret = $reader->read(); + restore_error_handler(); + return $ret; + } + + public function XmlErrorHandler( $errno, $errstr ) { + $this->wellFormed = false; + } + + private function validate( $reader ) { + + // First, move through anything that isn't an element, and + // handle any processing instructions with the callback + do { + if ( !$this->readNext( $reader ) ) { + // Hit the end of the document before any elements + $this->wellFormed = false; + return; + } + if ( $reader->nodeType === XMLReader::PI ) { + $this->processingInstructionHandler( $reader->name, $reader->value ); + } + } while ( $reader->nodeType != XMLReader::ELEMENT ); + + // Process the rest of the document + do { + switch ( $reader->nodeType ) { + case XMLReader::ELEMENT: + $name = $this->expandNS( + $reader->name, + $reader->namespaceURI + ); + if ( $this->rootElement === '' ) { + $this->rootElement = $name; + } + $empty = $reader->isEmptyElement; + $attrs = $this->getAttributesArray( $reader ); + $this->elementOpen( $name, $attrs ); + if ( $empty ) { + $this->elementClose(); + } + break; + + case XMLReader::END_ELEMENT: + $this->elementClose(); + break; + + case XMLReader::WHITESPACE: + case XMLReader::SIGNIFICANT_WHITESPACE: + case XMLReader::CDATA: + case XMLReader::TEXT: + $this->elementData( $reader->value ); + break; + + case XMLReader::ENTITY_REF: + // Unexpanded entity (maybe external?), + // don't send to the filter (xml_parse didn't) + break; + + case XMLReader::COMMENT: + // Don't send to the filter (xml_parse didn't) + break; + + case XMLReader::PI: + // Processing instructions can happen after the header too + $this->processingInstructionHandler( + $reader->name, + $reader->value + ); + break; + default: + // One of DOC, DOC_TYPE, ENTITY, END_ENTITY, + // NOTATION, or XML_DECLARATION + // xml_parse didn't send these to the filter, so we won't. + } + + } while ( $this->readNext( $reader ) ); + + if ( $this->stackDepth !== 0 ) { + $this->wellFormed = false; + } elseif ( $this->wellFormed === null ) { + $this->wellFormed = true; + } + + } + + /** + * Get all of the attributes for an XMLReader's current node + * @param $r XMLReader + * @return array of attributes + */ + private function getAttributesArray( XMLReader $r ) { + $attrs = []; + while ( $r->moveToNextAttribute() ) { + if ( $r->namespaceURI === 'http://www.w3.org/2000/xmlns/' ) { + // XMLReader treats xmlns attributes as normal + // attributes, while xml_parse doesn't + continue; + } + $name = $this->expandNS( $r->name, $r->namespaceURI ); + $attrs[$name] = $r->value; + } + return $attrs; + } + + /** + * @param $name element or attribute name, maybe with a full or short prefix + * @param $namespaceURI the namespaceURI + * @return string the name prefixed with namespaceURI + */ + private function expandNS( $name, $namespaceURI ) { + if ( $namespaceURI ) { + $parts = explode( ':', $name ); + $localname = array_pop( $parts ); + return "$namespaceURI:$localname"; + } + return $name; + } + + /** + * @param $name + * @param $attribs + */ + private function elementOpen( $name, $attribs ) { + $this->elementDataContext[] = [ $name, $attribs ]; + $this->elementData[] = ''; + $this->stackDepth++; + } + + /** + */ + private function elementClose() { + list( $name, $attribs ) = array_pop( $this->elementDataContext ); + $data = array_pop( $this->elementData ); + $this->stackDepth--; + $callbackReturn = false; + + if ( is_callable( $this->filterCallback ) ) { + $callbackReturn = call_user_func( + $this->filterCallback, + $name, + $attribs, + $data + ); + } + if ( $callbackReturn ) { + // Filter hit! + $this->filterMatch = true; + $this->filterMatchType = $callbackReturn; + } + } + + /** + * @param $data + */ + private function elementData( $data ) { + // Collect any data here, and we'll run the callback in elementClose + $this->elementData[ $this->stackDepth - 1 ] .= trim( $data ); + } + + /** + * @param $target + * @param $data + */ + private function processingInstructionHandler( $target, $data ) { + $callbackReturn = false; + if ( $this->parserOptions['processing_instruction_handler'] ) { + $callbackReturn = call_user_func( + $this->parserOptions['processing_instruction_handler'], + $target, + $data + ); + } + if ( $callbackReturn ) { + // Filter hit! + $this->filterMatch = true; + $this->filterMatchType = $callbackReturn; + } + } +} diff --git a/includes/libs/mime/defines.php b/includes/libs/mime/defines.php new file mode 100644 index 0000000000..ae0b5f8b61 --- /dev/null +++ b/includes/libs/mime/defines.php @@ -0,0 +1,46 @@ + [ 'LinkRenderer', LinkRenderer::class ], 'LinkRendererFactory' => [ 'LinkRendererFactory', LinkRendererFactory::class ], '_MediaWikiTitleCodec' => [ '_MediaWikiTitleCodec', MediaWikiTitleCodec::class ], + 'MimeAnalyzer' => [ 'MimeAnalyzer', MimeAnalyzer::class ], 'TitleFormatter' => [ 'TitleFormatter', TitleFormatter::class ], 'TitleParser' => [ 'TitleParser', TitleParser::class ], 'ProxyLookup' => [ 'ProxyLookup', ProxyLookup::class ], diff --git a/tests/phpunit/includes/MimeMagicTest.php b/tests/phpunit/includes/MimeMagicTest.php deleted file mode 100644 index e00cf0ce30..0000000000 --- a/tests/phpunit/includes/MimeMagicTest.php +++ /dev/null @@ -1,51 +0,0 @@ -mimeMagic = MimeMagic::singleton(); - parent::setUp(); - } - - /** - * @dataProvider providerImproveTypeFromExtension - * @param string $ext File extension (no leading dot) - * @param string $oldMime Initially detected MIME - * @param string $expectedMime MIME type after taking extension into account - */ - function testImproveTypeFromExtension( $ext, $oldMime, $expectedMime ) { - $actualMime = $this->mimeMagic->improveTypeFromExtension( $oldMime, $ext ); - $this->assertEquals( $expectedMime, $actualMime ); - } - - function providerImproveTypeFromExtension() { - return [ - [ 'gif', 'image/gif', 'image/gif' ], - [ 'gif', 'unknown/unknown', 'unknown/unknown' ], - [ 'wrl', 'unknown/unknown', 'model/vrml' ], - [ 'txt', 'text/plain', 'text/plain' ], - [ 'csv', 'text/plain', 'text/csv' ], - [ 'tsv', 'text/plain', 'text/tab-separated-values' ], - [ 'js', 'text/javascript', 'application/javascript' ], - [ 'js', 'application/x-javascript', 'application/javascript' ], - [ 'json', 'text/plain', 'application/json' ], - [ 'foo', 'application/x-opc+zip', 'application/zip' ], - [ 'docx', 'application/x-opc+zip', - 'application/vnd.openxmlformats-officedocument.wordprocessingml.document' ], - [ 'djvu', 'image/x-djvu', 'image/vnd.djvu' ], - [ 'wav', 'audio/wav', 'audio/wav' ], - ]; - } - - /** - * Test to make sure that encoder=ffmpeg2theora doesn't trigger - * MEDIATYPE_VIDEO (bug 63584) - */ - function testOggRecognize() { - $oggFile = __DIR__ . '/../data/media/say-test.ogg'; - $actualType = $this->mimeMagic->getMediaType( $oggFile, 'application/ogg' ); - $this->assertEquals( $actualType, MEDIATYPE_AUDIO ); - } -} diff --git a/tests/phpunit/includes/libs/mime/MimeAnalyzerTest.php b/tests/phpunit/includes/libs/mime/MimeAnalyzerTest.php new file mode 100644 index 0000000000..85927a393d --- /dev/null +++ b/tests/phpunit/includes/libs/mime/MimeAnalyzerTest.php @@ -0,0 +1,62 @@ +mimeAnalyzer = new MimeAnalyzer( [ + 'infoFile' => $IP . "/includes/libs/mime/mime.info", + 'typeFile' => $IP . "/includes/libs/mime/mime.types", + 'xmlTypes' => [ + 'http://www.w3.org/2000/svg:svg' => 'image/svg+xml', + 'svg' => 'image/svg+xml', + 'http://www.lysator.liu.se/~alla/dia/:diagram' => 'application/x-dia-diagram', + 'http://www.w3.org/1999/xhtml:html' => 'text/html', // application/xhtml+xml? + 'html' => 'text/html', // application/xhtml+xml? + ] + ] ); + parent::setUp(); + } + + /** + * @dataProvider providerImproveTypeFromExtension + * @param string $ext File extension (no leading dot) + * @param string $oldMime Initially detected MIME + * @param string $expectedMime MIME type after taking extension into account + */ + function testImproveTypeFromExtension( $ext, $oldMime, $expectedMime ) { + $actualMime = $this->mimeAnalyzer->improveTypeFromExtension( $oldMime, $ext ); + $this->assertEquals( $expectedMime, $actualMime ); + } + + function providerImproveTypeFromExtension() { + return [ + [ 'gif', 'image/gif', 'image/gif' ], + [ 'gif', 'unknown/unknown', 'unknown/unknown' ], + [ 'wrl', 'unknown/unknown', 'model/vrml' ], + [ 'txt', 'text/plain', 'text/plain' ], + [ 'csv', 'text/plain', 'text/csv' ], + [ 'tsv', 'text/plain', 'text/tab-separated-values' ], + [ 'js', 'text/javascript', 'application/javascript' ], + [ 'js', 'application/x-javascript', 'application/javascript' ], + [ 'json', 'text/plain', 'application/json' ], + [ 'foo', 'application/x-opc+zip', 'application/zip' ], + [ 'docx', 'application/x-opc+zip', + 'application/vnd.openxmlformats-officedocument.wordprocessingml.document' ], + [ 'djvu', 'image/x-djvu', 'image/vnd.djvu' ], + [ 'wav', 'audio/wav', 'audio/wav' ], + ]; + } + + /** + * Test to make sure that encoder=ffmpeg2theora doesn't trigger + * MEDIATYPE_VIDEO (bug 63584) + */ + function testOggRecognize() { + $oggFile = __DIR__ . '/../../../data/media/say-test.ogg'; + $actualType = $this->mimeAnalyzer->getMediaType( $oggFile, 'application/ogg' ); + $this->assertEquals( $actualType, MEDIATYPE_AUDIO ); + } +}