From 0d87d67d61520d34064f2eaae0f20b8ef11d4c59 Mon Sep 17 00:00:00 2001 From: Aaron Schulz Date: Wed, 21 Sep 2016 21:57:13 -0700 Subject: [PATCH] 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 --- autoload.php | 5 +- includes/Defines.php | 27 +- includes/MediaWikiServices.php | 9 + includes/MimeMagic.php | 1161 +--------------- includes/ServiceWiring.php | 9 + .../libs/{ => mime}/IEContentAnalyzer.php | 0 includes/libs/mime/MimeAnalyzer.php | 1166 +++++++++++++++++ includes/libs/{ => mime}/XmlTypeCheck.php | 0 includes/libs/mime/defines.php | 46 + includes/{ => libs/mime}/mime.info | 0 includes/{ => libs/mime}/mime.types | 0 maintenance/dev/includes/router.php | 3 +- .../includes/MediaWikiServicesTest.php | 1 + .../mime/MimeAnalyzerTest.php} | 25 +- 14 files changed, 1304 insertions(+), 1148 deletions(-) rename includes/libs/{ => mime}/IEContentAnalyzer.php (100%) create mode 100644 includes/libs/mime/MimeAnalyzer.php rename includes/libs/{ => mime}/XmlTypeCheck.php (100%) create mode 100644 includes/libs/mime/defines.php rename includes/{ => libs/mime}/mime.info (100%) rename includes/{ => libs/mime}/mime.types (100%) rename tests/phpunit/includes/{MimeMagicTest.php => libs/mime/MimeAnalyzerTest.php} (65%) 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 = $services->getMainConfig(); return new ProxyLookup( diff --git a/includes/libs/IEContentAnalyzer.php b/includes/libs/mime/IEContentAnalyzer.php similarity index 100% rename from includes/libs/IEContentAnalyzer.php rename to includes/libs/mime/IEContentAnalyzer.php diff --git a/includes/libs/mime/MimeAnalyzer.php b/includes/libs/mime/MimeAnalyzer.php new file mode 100644 index 0000000000..5f4d7c9ef4 --- /dev/null +++ b/includes/libs/mime/MimeAnalyzer.php @@ -0,0 +1,1166 @@ + ext + * map. Each line contains a MIME type followed by a space separated list of + * extensions. If multiple extensions for a single MIME type exist or if + * multiple MIME types exist for a single extension then in most cases + * MediaWiki assumes that the first extension following the MIME type is the + * canonical extension, and the first time a MIME type appears for a certain + * extension is considered the canonical MIME type. + * + * (Note that appending the type file list to the end of self::$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/XmlTypeCheck.php b/includes/libs/mime/XmlTypeCheck.php similarity index 100% rename from includes/libs/XmlTypeCheck.php rename to includes/libs/mime/XmlTypeCheck.php 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/libs/mime/MimeAnalyzerTest.php similarity index 65% rename from tests/phpunit/includes/MimeMagicTest.php rename to tests/phpunit/includes/libs/mime/MimeAnalyzerTest.php index e00cf0ce30..85927a393d 100644 --- a/tests/phpunit/includes/MimeMagicTest.php +++ b/tests/phpunit/includes/libs/mime/MimeAnalyzerTest.php @@ -1,11 +1,22 @@ mimeMagic = MimeMagic::singleton(); + global $IP; + + $this->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(); } @@ -16,7 +27,7 @@ class MimeMagicTest extends PHPUnit_Framework_TestCase { * @param string $expectedMime MIME type after taking extension into account */ function testImproveTypeFromExtension( $ext, $oldMime, $expectedMime ) { - $actualMime = $this->mimeMagic->improveTypeFromExtension( $oldMime, $ext ); + $actualMime = $this->mimeAnalyzer->improveTypeFromExtension( $oldMime, $ext ); $this->assertEquals( $expectedMime, $actualMime ); } @@ -44,8 +55,8 @@ class MimeMagicTest extends PHPUnit_Framework_TestCase { * MEDIATYPE_VIDEO (bug 63584) */ function testOggRecognize() { - $oggFile = __DIR__ . '/../data/media/say-test.ogg'; - $actualType = $this->mimeMagic->getMediaType( $oggFile, 'application/ogg' ); + $oggFile = __DIR__ . '/../../../data/media/say-test.ogg'; + $actualType = $this->mimeAnalyzer->getMediaType( $oggFile, 'application/ogg' ); $this->assertEquals( $actualType, MEDIATYPE_AUDIO ); } } -- 2.20.1