<?php
-/** Module defining helper functions for detecting and dealing with mime types.
+/**
+ * Module defining helper functions for detecting and dealing with mime types.
*
+ * @file
*/
- /** Defines a set of well known mime types
+/**
+ * Defines a set of well known mime types
* This is used as a fallback to mime.types files.
* An extensive list of well known mime types is provided by
* the file mime.types in the includes directory.
+ *
+ * This list concatenated with mime.types is used to create a mime <-> 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',<<<END_STRING
-application/ogg ogg ogm
+application/ogg ogx ogg ogm ogv oga spx
application/pdf pdf
+application/vnd.oasis.opendocument.chart odc
+application/vnd.oasis.opendocument.chart-template otc
+application/vnd.oasis.opendocument.formula odf
+application/vnd.oasis.opendocument.formula-template otf
+application/vnd.oasis.opendocument.graphics odg
+application/vnd.oasis.opendocument.graphics-template otg
+application/vnd.oasis.opendocument.image odi
+application/vnd.oasis.opendocument.image-template oti
+application/vnd.oasis.opendocument.presentation odp
+application/vnd.oasis.opendocument.presentation-template otp
+application/vnd.oasis.opendocument.spreadsheet ods
+application/vnd.oasis.opendocument.spreadsheet-template ots
+application/vnd.oasis.opendocument.text odt
+application/vnd.oasis.opendocument.text-template ott
+application/vnd.oasis.opendocument.text-master otm
+application/vnd.oasis.opendocument.text-web oth
application/x-javascript js
application/x-shockwave-flash swf
audio/midi mid midi kar
audio/mpeg mpga mpa mp2 mp3
audio/x-aiff aif aiff aifc
audio/x-wav wav
-audio/ogg ogg
+audio/ogg oga spx ogg
image/x-bmp bmp
image/gif gif
image/jpeg jpeg jpg jpe
image/png png
-image/svg+xml image/svg svg
+image/svg+xml svg
+image/svg svg
image/tiff tiff tif
-image/vnd.djvu image/x.djvu image/x-djvu djvu
+image/vnd.djvu djvu
+image/x.djvu djvu
+image/x-djvu djvu
image/x-portable-pixmap ppm
+image/x-xcf xcf
text/plain txt
text/html html htm
-video/ogg ogm ogg
+video/ogg ogv ogm ogg
video/mpeg mpg mpeg
END_STRING
);
- /** Defines a set of well known mime info entries
+/**
+ * Defines a set of well known mime info entries
* This is used as a fallback to mime.info files.
* An extensive list of well known mime types is provided by
* the file mime.info in the includes directory.
*/
define('MM_WELL_KNOWN_MIME_INFO', <<<END_STRING
application/pdf [OFFICE]
+application/vnd.oasis.opendocument.chart [OFFICE]
+application/vnd.oasis.opendocument.chart-template [OFFICE]
+application/vnd.oasis.opendocument.formula [OFFICE]
+application/vnd.oasis.opendocument.formula-template [OFFICE]
+application/vnd.oasis.opendocument.graphics [OFFICE]
+application/vnd.oasis.opendocument.graphics-template [OFFICE]
+application/vnd.oasis.opendocument.image [OFFICE]
+application/vnd.oasis.opendocument.image-template [OFFICE]
+application/vnd.oasis.opendocument.presentation [OFFICE]
+application/vnd.oasis.opendocument.presentation-template [OFFICE]
+application/vnd.oasis.opendocument.spreadsheet [OFFICE]
+application/vnd.oasis.opendocument.spreadsheet-template [OFFICE]
+application/vnd.oasis.opendocument.text [OFFICE]
+application/vnd.oasis.opendocument.text-template [OFFICE]
+application/vnd.oasis.opendocument.text-master [OFFICE]
+application/vnd.oasis.opendocument.text-web [OFFICE]
text/javascript application/x-javascript [EXECUTABLE]
application/x-shockwave-flash [MULTIMEDIA]
audio/midi [AUDIO]
audio/x-wav [AUDIO]
audio/mp3 audio/mpeg [AUDIO]
application/ogg audio/ogg video/ogg [MULTIMEDIA]
-image/x-bmp image/bmp [BITMAP]
+image/x-bmp image/x-ms-bmp image/bmp [BITMAP]
image/gif [BITMAP]
image/jpeg [BITMAP]
image/png [BITMAP]
image/svg+xml [DRAWING]
image/tiff [BITMAP]
image/vnd.djvu [BITMAP]
+image/x-xcf [BITMAP]
image/x-portable-pixmap [BITMAP]
text/plain [TEXT]
text/html [TEXT]
global $wgLoadFileinfoExtension;
if ($wgLoadFileinfoExtension) {
- if(!extension_loaded('fileinfo')) dl('fileinfo.' . PHP_SHLIB_SUFFIX);
+ wfDl( 'fileinfo' );
}
-/**
+/**
* Implements functions related to mime types such as detection and mapping to
* file extension.
*
* Mapping of media types to arrays of mime types.
* This is used by findMediaType and getMediaType, respectively
*/
- var $mMediaTypes= NULL;
+ var $mMediaTypes= null;
/** Map of mime type aliases
*/
- var $mMimeTypeAliases= NULL;
+ var $mMimeTypeAliases= null;
/** map of mime types to file extensions (as a space seprarated list)
*/
- var $mMimeToExt= NULL;
+ var $mMimeToExt= null;
/** map of file extensions types to mime types (as a space seprarated list)
*/
- var $mExtToMime= NULL;
+ var $mExtToMime= null;
+
+ /** IEContentAnalyzer instance
+ */
+ var $mIEAnalyzer;
/** The singleton instance
*/
if ( $wgMimeTypeFile == 'includes/mime.types' ) {
$wgMimeTypeFile = "$IP/$wgMimeTypeFile";
}
-
+
if ( $wgMimeTypeFile ) {
if ( is_file( $wgMimeTypeFile ) and is_readable( $wgMimeTypeFile ) ) {
wfDebug( __METHOD__.": loading mime types from $wgMimeTypeFile\n" );
*/
function guessTypesForExtension( $ext ) {
$m = $this->getTypesForExtension( $ext );
- if ( is_null( $m ) ) return NULL;
+ if ( is_null( $m ) ) return null;
$m = trim( $m );
$m = preg_replace( '/\s.*$/', '', $m );
$ext = $this->getExtensionsForType( $mime );
if ( !$ext ) {
- return NULL; //unknown
+ return null; //unknown
}
$ext = explode( ' ', $ext );
*/
function isRecognizableExtension( $extension ) {
static $types = array(
+ // Types recognized by getimagesize()
'gif', 'jpeg', 'jpg', 'png', 'swf', 'psd',
'bmp', 'tiff', 'tif', 'jpc', 'jp2',
'jpx', 'jb2', 'swc', 'iff', 'wbmp',
- 'xbm', 'djvu'
+ '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().
+ *
+ * Currently, this method does the following:
+ *
+ * If $mime is "unknown/unknown" and isRecognizableExtension( $ext ) returns false,
+ * return the result of guessTypesForExtension($ext).
+ *
+ * If $mime is "application/x-opc+zip" and isMatchingExtension( $ext, $mime )
+ * gives true, return the result of guessTypesForExtension($ext).
+ *
+ * @param $mime String: the mime type, typically guessed from a file's content.
+ * @param $ext String: the file extension, as taken from the file name
+ *
+ * @return string the mime type
+ */
+ 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 );
+ }
+ }
+ else if ( $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";
+ }
+ }
+
+ 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 misinterpreter by the default mime detection (namely xml based formats like XHTML or SVG).
+ * or misinterpreter 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 mixed $ext The file extension, or true to extract it from the filename.
- * Set it to false to ignore the extension.
+ * @param $file String: the file to check
+ * @param $ext Mixed: the file extension, or true (default) to extract it from the filename.
+ * Set it to false to ignore the extension. DEPRECATED! Set to false, use
+ * improveTypeFromExtension($mime, $ext) later to improve mime type.
*
* @return string the mime type of $file
*/
function guessMimeType( $file, $ext = true ) {
- $mime = $this->detectMimeType( $file, $ext );
+ 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;
+ }
+
+ private function doGuessMimeType( $file, $ext ) { # TODO: remove $ext param
// Read a chunk of the file
wfSuppressWarnings();
$f = fopen( $file, "rt" );
wfRestoreWarnings();
if( !$f ) return "unknown/unknown";
$head = fread( $f, 1024 );
+ fseek( $f, -65558, SEEK_END );
+ $tail = fread( $f, 65558 ); // 65558 = maximum size of a zip EOCDR
fclose( $f );
- $sub4 = substr( $head, 0, 4 );
- if ( $sub4 == "\x01\x00\x09\x00" || $sub4 == "\xd7\xcd\xc6\x9a" ) {
- // WMF kill kill kill
- // Note that WMF may have a bare header, no magic number.
- // The former of the above two checks is theoretically prone to false positives
- $mime = "application/x-msmetafile";
- }
-
- if ( strpos( $mime, "text/" ) === 0 || $mime === "application/xml" ) {
-
- $xml_type = NULL;
- $script_type = NULL;
-
- /*
- * look for XML formats (XHTML and SVG)
- */
- if ($mime === "text/sgml" ||
- $mime === "text/plain" ||
- $mime === "text/html" ||
- $mime === "text/xml" ||
- $mime === "application/xml") {
-
- if ( substr( $head, 0, 5 ) == "<?xml" ) {
- $xml_type = "ASCII";
- } elseif ( substr( $head, 0, 8 ) == "\xef\xbb\xbf<?xml") {
- $xml_type = "UTF-8";
- } elseif ( substr( $head, 0, 10 ) == "\xfe\xff\x00<\x00?\x00x\x00m\x00l" ) {
- $xml_type = "UTF-16BE";
- } elseif ( substr( $head, 0, 10 ) == "\xff\xfe<\x00?\x00x\x00m\x00l\x00") {
- $xml_type = "UTF-16LE";
- }
-
- if ( $xml_type ) {
- if ( $xml_type !== "UTF-8" && $xml_type !== "ASCII" ) {
- $head = iconv( $xml_type, "ASCII//IGNORE", $head );
- }
+ wfDebug( __METHOD__ . ": analyzing head and tail of $file for magic numbers.\n" );
- $match = array();
- $doctype = "";
- $tag = "";
+ // Hardcode a few magic number checks...
+ $headers = array(
+ // Multimedia...
+ 'MThd' => 'audio/midi',
+ 'OggS' => 'application/ogg',
- if ( preg_match( '%<!DOCTYPE\s+[\w-]+\s+PUBLIC\s+["'."'".'"](.*?)["'."'".'"].*>%sim',
- $head, $match ) ) {
- $doctype = $match[1];
- }
- if ( preg_match( '%<(\w+).*>%sim', $head, $match ) ) {
- $tag = $match[1];
- }
+ // 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
+ );
- #print "<br>ANALYSING $file ($mime): doctype= $doctype; tag= $tag<br>";
+ foreach( $headers as $magic => $candidate ) {
+ if( strncmp( $head, $magic, strlen( $magic ) ) == 0 ) {
+ wfDebug( __METHOD__ . ": magic header in $file recognized as $candidate\n" );
+ return $candidate;
+ }
+ }
- if ( strpos( $doctype, "-//W3C//DTD SVG" ) === 0 ) {
- $mime = "image/svg+xml";
- } elseif ( $tag === "svg" ) {
- $mime = "image/svg+xml";
- } elseif ( strpos( $doctype, "-//W3C//DTD XHTML" ) === 0 ) {
- $mime = "text/html";
- } elseif ( $tag === "html" ) {
- $mime = "text/html";
- }
+ /* 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";
+ } else if ( strncmp( $data, "webm", 4 ) == 0 ) {
+ wfDebug( __METHOD__ . ": recognized file as video/webm\n" );
+ return "video/webm";
}
}
+ wfDebug( __METHOD__ . ": unknown EBML file\n" );
+ return "unknown/unknown";
+ }
- /*
- * look for shell scripts
- */
- if ( !$xml_type ) {
- $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";
- }
+ /* Look for WebP */
+ if( strncmp( $head, "RIFF", 4 ) == 0 && strncmp( substr( $head, 8, 8), "WEBPVP8 ", 8 ) == 0 ) {
+ wfDebug( __METHOD__ . ": recognized file as image/webp\n" );
+ return "image/webp";
+ }
- if ( $script_type ) {
- if ( $script_type !== "UTF-8" && $script_type !== "ASCII") {
- $head = iconv( $script_type, "ASCII//IGNORE", $head);
- }
+ /*
+ * 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. :)
+ *
+ * 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 "<? ", but should it be axed completely?
+ */
+ if( ( strpos( $head, '<?php' ) !== false ) ||
+
+ ( strpos( $head, "<\x00?\x00p\x00h\x00p" ) !== false ) ||
+ ( strpos( $head, "<\x00?\x00 " ) !== false ) ||
+ ( strpos( $head, "<\x00?\x00\n" ) !== false ) ||
+ ( strpos( $head, "<\x00?\x00\t" ) !== false ) ||
+ ( strpos( $head, "<\x00?\x00=" ) !== false ) ) {
+
+ wfDebug( __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 ) {
+ global $wgXMLMimeTypes;
+ if( isset( $wgXMLMimeTypes[$xml->getRootElement()] ) ) {
+ return $wgXMLMimeTypes[$xml->getRootElement()];
+ } else {
+ return 'application/xml';
+ }
+ }
- $match = array();
+ /*
+ * 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 ( preg_match( '%/?([^\s]+/)(\w+)%', $head, $match ) ) {
- $mime = "application/x-{$match[2]}";
+ if ( $script_type ) {
+ if ( $script_type !== "UTF-8" && $script_type !== "ASCII") {
+ // Quick and dirty fold down to ASCII!
+ $pack = array( '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 {
+ $head .= '?';
}
}
}
- /*
- * look for PHP
- */
- if( !$xml_type && !$script_type ) {
+ $match = array();
- if( ( strpos( $head, '<?php' ) !== false ) ||
- ( strpos( $head, '<? ' ) !== false ) ||
- ( strpos( $head, "<?\n" ) !== false ) ||
- ( strpos( $head, "<?\t" ) !== false ) ||
- ( strpos( $head, "<?=" ) !== false ) ||
+ if ( preg_match( '%/?([^\s]+/)(\w+)%', $head, $match ) ) {
+ $mime = "application/x-{$match[2]}";
+ wfDebug( __METHOD__.": shell script recognized as $mime\n" );
+ return $mime;
+ }
+ }
- ( strpos( $head, "<\x00?\x00p\x00h\x00p" ) !== false ) ||
- ( strpos( $head, "<\x00?\x00 " ) !== false ) ||
- ( strpos( $head, "<\x00?\x00\n" ) !== false ) ||
- ( strpos( $head, "<\x00?\x00\t" ) !== false ) ||
- ( strpos( $head, "<\x00?\x00=" ) !== false ) ) {
+ // 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 );
+ }
- $mime = "application/x-php";
- }
- }
+ wfSuppressWarnings();
+ $gis = getimagesize( $file );
+ wfRestoreWarnings();
+ if( $gis && isset( $gis['mime'] ) ) {
+ $mime = $gis['mime'];
+ wfDebug( __METHOD__.": getimagesize detected $file as $mime\n" );
+ return $mime;
}
- if ( isset( $this->mMimeTypeAliases[$mime] ) ) {
- $mime = $this->mMimeTypeAliases[$mime];
+ // Also test DjVu
+ $deja = new DjVuImage( $file );
+ if( $deja->isValid() ) {
+ wfDebug( __METHOD__.": detected $file as image/vnd.djvu\n" );
+ return 'image/vnd.djvu';
}
- wfDebug(__METHOD__.": final mime type of $file: $mime\n");
+ return false;
+ }
+
+ /**
+ * 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 $header String: some reasonably-sized chunk of file header
+ * @param $tail String: the tail of the file
+ * @param $ext Mixed: the file extension, or true to extract it from the filename.
+ * Set it to false (default) to ignore the extension. DEPRECATED! Set to false,
+ * use improveTypeFromExtension($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 = array(
+ '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 ) {
+ /** 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";
+ }
+ }
+ wfDebug( __METHOD__.": detected an Open Packaging Conventions archive: $mime\n" );
+ } else if( 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, please use guessMimeType() for application code instead.
* Detection is done using an external program, if $wgMimeDetectorCommand is set.
* Otherwise, the fileinfo extension and mime_content_type are tried (in this order), if they are available.
- * If the dections fails and $ext is not false, the mime type is guessed from the file extension, using
+ * If the dections 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 mixed $ext The file extension, or true to extract it from the filename.
- * Set it to false to ignore the extension.
+ * @param $file String: the file to check
+ * @param $ext Mixed: the file extension, or true (default) to extract it from the filename.
+ * Set it to false to ignore the extension. DEPRECATED! Set to false, use
+ * improveTypeFromExtension($mime, $ext) later to improve mime type.
*
* @return string the mime type of $file
* @access private
*/
- function detectMimeType( $file, $ext = true ) {
+ private function detectMimeType( $file, $ext = true ) {
global $wgMimeDetectorCommand;
- $m = NULL;
+ 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" );
+ }
+
+ $m = null;
if ( $wgMimeDetectorCommand ) {
$fn = wfEscapeShellArg( $file );
$m = `$wgMimeDetectorCommand $fn`;
# see http://www.php.net/manual/en/ref.mime-magic.php for details.
$m = mime_content_type($file);
-
- if ( $m == 'text/plain' ) {
- // mime_content_type sometimes considers DJVU files to be text/plain.
- $deja = new DjVuImage( $file );
- if( $deja->isValid() ) {
- wfDebug( __METHOD__.": (re)detected $file as image/vnd.djvu\n" );
- $m = 'image/vnd.djvu';
- }
- }
} else {
wfDebug( __METHOD__.": no magic mime detector found!\n" );
}
$m = strtolower( $m );
if ( strpos( $m, 'unknown' ) !== false ) {
- $m = NULL;
+ $m = null;
} else {
wfDebug( __METHOD__.": magic mime type of $file: $m\n" );
return $m;
}
}
- # if still not known, use getimagesize to find out the type of image
- # TODO: skip things that do not have a well-known image extension? Would that be safe?
- wfSuppressWarnings();
- $gis = getimagesize( $file );
- wfRestoreWarnings();
-
- $notAnImage = false;
-
- if ( $gis && is_array($gis) && $gis[2] ) {
-
- switch ( $gis[2] ) {
- case IMAGETYPE_GIF: $m = "image/gif"; break;
- case IMAGETYPE_JPEG: $m = "image/jpeg"; break;
- case IMAGETYPE_PNG: $m = "image/png"; break;
- case IMAGETYPE_SWF: $m = "application/x-shockwave-flash"; break;
- case IMAGETYPE_PSD: $m = "application/photoshop"; break;
- case IMAGETYPE_BMP: $m = "image/bmp"; break;
- case IMAGETYPE_TIFF_II: $m = "image/tiff"; break;
- case IMAGETYPE_TIFF_MM: $m = "image/tiff"; break;
- case IMAGETYPE_JPC: $m = "image"; break;
- case IMAGETYPE_JP2: $m = "image/jpeg2000"; break;
- case IMAGETYPE_JPX: $m = "image/jpeg2000"; break;
- case IMAGETYPE_JB2: $m = "image"; break;
- case IMAGETYPE_SWC: $m = "application/x-shockwave-flash"; break;
- case IMAGETYPE_IFF: $m = "image/vnd.xiff"; break;
- case IMAGETYPE_WBMP: $m = "image/vnd.wap.wbmp"; break;
- case IMAGETYPE_XBM: $m = "image/x-xbitmap"; break;
- }
-
- if ( $m ) {
- wfDebug( __METHOD__.": image mime type of $file: $m\n" );
- return $m;
- }
- else {
- $notAnImage = true;
- }
- } else {
- // Also test DjVu
- $deja = new DjVuImage( $file );
- if( $deja->isValid() ) {
- wfDebug( __METHOD__.": detected $file as image/vnd.djvu\n" );
- return 'image/vnd.djvu';
- }
- }
-
# if desired, look at extension as a fallback.
if ( $ext === true ) {
$i = strrpos( $file, '.' );
$ext = strtolower( $i ? substr( $file, $i + 1 ) : '' );
}
if ( $ext ) {
- $m = $this->guessTypesForExtension( $ext );
-
- # TODO: if $notAnImage is set, do not trust the file extension if
- # the results is one of the image types that should have been recognized
- # by getimagesize
-
- if ( $m ) {
- wfDebug( __METHOD__.": extension mime type of $file: $m\n" );
- return $m;
+ 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;
+ }
}
}
* @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
+ * @param $path String: 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.
+ * @param $mime String: mime type. If null it will be guessed using guessMimeType.
*
* @return (int?string?) a value to be used with the MEDIATYPE_xxx constants.
*/
- function getMediaType( $path = NULL, $mime = NULL ) {
+ function getMediaType( $path = null, $mime = null ) {
if( !$mime && !$path ) return MEDIATYPE_UNKNOWN;
# If mime type is unknown, guess it
}
# Check for entry for file extension
- $e = NULL;
if ( $path ) {
$i = strrpos( $path, '.' );
$e = strtolower( $i ? substr( $path, $i + 1 ) : '' );
if ( !$m ) return MEDIATYPE_UNKNOWN;
$m = explode( ' ', $m );
- } else {
+ } else {
# Normalize mime type
if ( isset( $this->mMimeTypeAliases[$extMime] ) ) {
$extMime = $this->mMimeTypeAliases[$extMime];
return MEDIATYPE_UNKNOWN;
}
-}
+ /**
+ * Get the MIME types that various versions of Internet Explorer would
+ * detect from a chunk of the content.
+ *
+ * @param $fileName String: the file name (unused at present)
+ * @param $chunk String: the first 256 bytes of the file
+ * @param $proposed String: the MIME type proposed by the server
+ */
+ public function getIEMimeTypes( $fileName, $chunk, $proposed ) {
+ $ca = $this->getIEContentAnalyzer();
+ return $ca->getRealMimesFromData( $fileName, $chunk, $proposed );
+ }
+ /**
+ * Get a cached instance of IEContentAnalyzer
+ */
+ protected function getIEContentAnalyzer() {
+ if ( is_null( $this->mIEAnalyzer ) ) {
+ $this->mIEAnalyzer = new IEContentAnalyzer;
+ }
+ return $this->mIEAnalyzer;
+ }
+}
dépôts / lhc / web / wiklou.git / blobdiff