http://www.mediawiki.org/wiki/Special:Code/MediaWiki/68279
Revision: 68279
Author: hartman
Date: 2010-06-19 18:47:34 +0000 (Sat, 19 Jun 2010)
Log Message:
-----------
Bug 23642. Best effort mime detection for OpenXML files
OpenXML files are Open Package Convention files. Internally, we use the custom
mime application/x-opc+zip for these files. In the database, we store the
'proper' mime, which we gu
ess from the file extension, or if not supported, application/zip. All OPC
files are blacklisted by $wgMimeTypeBlacklist by default, just as other zip
files.
Modified Paths:
--------------
trunk/phase3/RELEASE-NOTES
trunk/phase3/includes/DefaultSettings.php
trunk/phase3/includes/MimeMagic.php
trunk/phase3/includes/mime.info
trunk/phase3/includes/mime.types
Modified: trunk/phase3/RELEASE-NOTES
===================================================================
--- trunk/phase3/RELEASE-NOTES 2010-06-19 18:40:28 UTC (rev 68278)
+++ trunk/phase3/RELEASE-NOTES 2010-06-19 18:47:34 UTC (rev 68279)
@@ -203,6 +203,7 @@
* (bug 23465) Don't ignore the predefined destination filename on
Special:Upload after following a red link
* Correct the escaping of the autosummary URI fragments.
+* (bug 23642) Recognize mime types of MS OpenXML documents.
=== API changes in 1.17 ===
* (bug 22738) Allow filtering by action type on query=logevent.
Modified: trunk/phase3/includes/DefaultSettings.php
===================================================================
--- trunk/phase3/includes/DefaultSettings.php 2010-06-19 18:40:28 UTC (rev
68278)
+++ trunk/phase3/includes/DefaultSettings.php 2010-06-19 18:47:34 UTC (rev
68279)
@@ -520,6 +520,9 @@
# A ZIP file may be a valid Java archive containing an applet which
exploits the
# same-origin policy to steal cookies
'application/zip',
+ # MS Office OpenXML and other Open Package Conventions files are zip
files
+ # and thus blacklisted just as other zip files
+ 'application/x-opc+zip',
);
/** This is a flag to determine whether or not to check file extensions on
upload. */
Modified: trunk/phase3/includes/MimeMagic.php
===================================================================
--- trunk/phase3/includes/MimeMagic.php 2010-06-19 18:40:28 UTC (rev 68278)
+++ trunk/phase3/includes/MimeMagic.php 2010-06-19 18:47:34 UTC (rev 68279)
@@ -548,7 +548,7 @@
// Check for ZIP (before getimagesize)
if ( strpos( $tail, "PK\x05\x06" ) !== false ) {
wfDebug( __METHOD__.": ZIP header present at end of
$file\n" );
- return $this->detectZipType( $head );
+ return $this->detectZipType( $head, $ext );
}
wfSuppressWarnings();
@@ -577,9 +577,13 @@
* If can't tell, returns 'application/zip'.
*
* @param $header String: some reasonably-sized chunk of file header
+ * @param $ext Mixed: the file extension, or true to extract it from
the filename.
+ * Set it to false to ignore the extension.
+ *
* @return string
*/
- function detectZipType( $header ) {
+ function detectZipType( $header, $ext = false ) {
+ $mime = 'application/zip';
$opendocTypes = array(
'chart-template',
'chart',
@@ -602,15 +606,33 @@
$types = '(?:' . implode( '|', $opendocTypes ) . ')';
$opendocRegex =
"/^mimetype(application\/vnd\.oasis\.opendocument\.$types)/";
wfDebug( __METHOD__.": $opendocRegex\n" );
+
+ $openxmlRegex = "/^\[Content_Types\].xml/";
if( preg_match( $opendocRegex, substr( $header, 30 ), $matches
) ) {
$mime = $matches[1];
wfDebug( __METHOD__.": detected $mime from ZIP
archive\n" );
- return $mime;
+ } elseif( preg_match( $openxmlRegex, substr( $header, 30 ),
$matches ) ) {
+ $mime = "application/x-opc+zip";
+ 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 {
wfDebug( __METHOD__.": unable to identify type of ZIP
archive\n" );
- return 'application/zip';
}
+ return $mime;
}
/** Internal mime type detection, please use guessMimeType() for
application code instead.
Modified: trunk/phase3/includes/mime.info
===================================================================
--- trunk/phase3/includes/mime.info 2010-06-19 18:40:28 UTC (rev 68278)
+++ trunk/phase3/includes/mime.info 2010-06-19 18:47:34 UTC (rev 68279)
@@ -61,6 +61,7 @@
application/x-bzip [ARCHIVE]
application/x-tar [ARCHIVE]
application/x-stuffit [ARCHIVE]
+application/x-opc+zip [ARCHIVE]
text/javascript application/x-javascript application/x-ecmascript
text/ecmascript [EXECUTABLE]
@@ -78,3 +79,22 @@
application/vnd.ms-powerpoint [OFFICE]
application/x-director [OFFICE]
text/rtf [OFFICE]
+
+application/vnd.openxmlformats-officedocument.wordprocessingml.document
[OFFICE]
+application/vnd.openxmlformats-officedocument.wordprocessingml.template
[OFFICE]
+application/vnd.ms-word.document.macroEnabled.12
[OFFICE]
+application/vnd.ms-word.template.macroEnabled.12
[OFFICE]
+application/vnd.openxmlformats-officedocument.presentationml.template
[OFFICE]
+application/vnd.openxmlformats-officedocument.presentationml.slideshow
[OFFICE]
+application/vnd.openxmlformats-officedocument.presentationml.presentation
[OFFICE]
+application/vnd.ms-powerpoint.addin.macroEnabled.12
[OFFICE]
+application/vnd.ms-powerpoint.presentation.macroEnabled.12
[OFFICE]
+application/vnd.ms-powerpoint.presentation.macroEnabled.12
[OFFICE]
+application/vnd.ms-powerpoint.slideshow.macroEnabled.12
[OFFICE]
+application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
[OFFICE]
+application/vnd.openxmlformats-officedocument.spreadsheetml.template
[OFFICE]
+application/vnd.ms-excel.sheet.macroEnabled.12
[OFFICE]
+application/vnd.ms-excel.template.macroEnabled.12
[OFFICE]
+application/vnd.ms-excel.addin.macroEnabled.12
[OFFICE]
+application/vnd.ms-excel.sheet.binary.macroEnabled.12
[OFFICE]
+
Modified: trunk/phase3/includes/mime.types
===================================================================
--- trunk/phase3/includes/mime.types 2010-06-19 18:40:28 UTC (rev 68278)
+++ trunk/phase3/includes/mime.types 2010-06-19 18:47:34 UTC (rev 68279)
@@ -152,3 +152,7 @@
application/vnd.ms-excel.template.macroEnabled.12 xltm
application/vnd.ms-excel.addin.macroEnabled.12 xlam
application/vnd.ms-excel.sheet.binary.macroEnabled.12 xlsb
+application/epub+zip epub
+model/vnd.dwfx+xps dwfx
+application/vnd.ms-xpsdocument xps
+application/x-opc+zip docx dotx docm dotm potx ppsx pptx ppam pptm potm ppsm
xlsx xltx xlsm xltm xlam xlsb epub dwfx xps
_______________________________________________
MediaWiki-CVS mailing list
MediaWiki-CVS@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-cvs
