Related: tdf#108269 oox: allow recovering broken DOCM files

The content type inside an OOXML file differs for DOCX and DOCM. These
must be in sync with the file extension, otherwise MSO refuses to open
the file. We used to always write the DOCX content-type even for files
which had the DOCM extension.

Allow users to recover those broken files by detecting a "has docm
extension but docx content-type" file as docm, so re-saving it will
produce output that's accepted by MSO as well.

(cherry picked from commit 97fa7024ce608b7908aca369e8c643a5de9ebf78)

Change-Id: I7d60c6f6c1d0421e95b3dc9e8fff617f101919f5
Reviewed-on: https://gerrit.libreoffice.org/38356
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>

1 files changed, 12 insertions, 7 deletions

diff --git a/oox/source/core/filterdetect.cxx b/oox/source/core/filterdetect.cxx
index 1882572f12d5..226668dfc767 100644
--- a/oox/source/core/filterdetect.cxx
+++ b/oox/source/core/filterdetect.cxx
@@ -52,8 +52,9 @@ using utl::MediaDescriptor;
 using comphelper::IDocPasswordVerifier;
 using comphelper::DocPasswordVerifierResult;
 
-FilterDetectDocHandler::FilterDetectDocHandler( const  Reference< XComponentContext >& rxContext, OUString& rFilterName ) :
+FilterDetectDocHandler::FilterDetectDocHandler( const  Reference< XComponentContext >& rxContext, OUString& rFilterName, const OUString& rFileName ) :
     mrFilterName( rFilterName ),
+    maFileName(rFileName),
     mxContext( rxContext )
 {
     maContextStack.reserve( 2 );
@@ -160,12 +161,12 @@ void FilterDetectDocHandler::parseRelationship( const AttributeList& rAttribs )
     }
 }
 
-OUString FilterDetectDocHandler::getFilterNameFromContentType( const OUString& rContentType )
+OUString FilterDetectDocHandler::getFilterNameFromContentType( const OUString& rContentType, const OUString& rFileName )
 {
-    if( rContentType == "application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml" )
+    if( rContentType == "application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml" && !rFileName.endsWith("docm") )
         return OUString( "writer_MS_Word_2007" );
 
-    if( rContentType == "application/vnd.ms-word.document.macroEnabled.main+xml" )
+    if( rContentType == "application/vnd.ms-word.document.macroEnabled.main+xml" || rFileName.endsWith("docm") )
         return OUString( "writer_MS_Word_2007_VBA" );
 
     if( rContentType == "application/vnd.openxmlformats-officedocument.wordprocessingml.template.main+xml" ||
@@ -209,14 +210,14 @@ void FilterDetectDocHandler::parseContentTypesDefault( const AttributeList& rAtt
         OUString aExtension = rAttribs.getString( XML_Extension, OUString() );
         sal_Int32 nExtPos = maTargetPath.getLength() - aExtension.getLength();
         if( (nExtPos > 0) && (maTargetPath[ nExtPos - 1 ] == '.') && maTargetPath.match( aExtension, nExtPos ) )
-            mrFilterName = getFilterNameFromContentType( rAttribs.getString( XML_ContentType, OUString() ) );
+            mrFilterName = getFilterNameFromContentType( rAttribs.getString( XML_ContentType, OUString() ), maFileName );
     }
 }
 
 void FilterDetectDocHandler::parseContentTypesOverride( const AttributeList& rAttribs )
 {
     if( rAttribs.getString( XML_PartName, OUString() ).equals( maTargetPath ) )
-        mrFilterName = getFilterNameFromContentType( rAttribs.getString( XML_ContentType, OUString() ) );
+        mrFilterName = getFilterNameFromContentType( rAttribs.getString( XML_ContentType, OUString() ), maFileName );
 }
 
 /* Helper for XServiceInfo */
@@ -400,7 +401,11 @@ OUString SAL_CALL FilterDetect::detect( Sequence< PropertyValue >& rMediaDescSeq
             aParser.registerNamespace( NMSP_packageRel );
             aParser.registerNamespace( NMSP_officeRel );
             aParser.registerNamespace( NMSP_packageContentTypes );
-            aParser.setDocumentHandler( new FilterDetectDocHandler( mxContext, aFilterName ) );
+
+            OUString aFileName;
+            aMediaDescriptor[utl::MediaDescriptor::PROP_URL()] >>= aFileName;
+
+            aParser.setDocumentHandler( new FilterDetectDocHandler( mxContext, aFilterName, aFileName ) );
 
             /*  Parse '_rels/.rels' to get the target path and '[Content_Types].xml'
                 to determine the content type of the part at the target path. */