sw: fix modification of ole obj native data during HTML import

It is expected that if you load an OLE2 storage (from reqif-xhtml) and
you save it as ODT, the OLE2 data is not modified. This was almost the
case, but we insisted on removing the OLE2 preview from the storage.

Add a new flag to OleEmbeddedObject, so import filters can opt in for
not modifying the OLE2 data.

[ The nice situation is that we already had a test file which had a
preview stream in the OLE2 storage, so we can easily assert there that
the size doesn't change. ]

Change-Id: I9b8b29f015dd4f2513e51a1066767218580cb5d8
Reviewed-on: https://gerrit.libreoffice.org/63381
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins

5 files changed, 50 insertions, 3 deletions

diff --git a/embeddedobj/source/inc/oleembobj.hxx b/embeddedobj/source/inc/oleembobj.hxx
index b1f44fbbd429..d9b75d2a9836 100644
--- a/embeddedobj/source/inc/oleembobj.hxx
+++ b/embeddedobj/source/inc/oleembobj.hxx
@@ -37,6 +37,7 @@
 #include <com/sun/star/util/XCloseable.hpp>
 #include <com/sun/star/util/XCloseListener.hpp>
 #include <com/sun/star/io/XActiveDataStreamer.hpp>
+#include <com/sun/star/lang/XInitialization.hpp>
 #include <cppuhelper/implbase.hxx>
 #include <rtl/ref.hxx>
 
@@ -117,7 +118,8 @@ class OleEmbeddedObject : public ::cppu::WeakImplHelper
                         , css::embed::XLinkageSupport
                         , css::embed::XInplaceObject
                         , css::container::XChild
-                        , css::io::XActiveDataStreamer >
+                        , css::io::XActiveDataStreamer
+                        , css::lang::XInitialization >
 {
     friend class OleComponent;
 
@@ -205,6 +207,9 @@ class OleEmbeddedObject : public ::cppu::WeakImplHelper
 
     css::uno::Reference< css::uno::XInterface > m_xParent;
 
+    /// If it is allowed to modify entires in the stream of the OLE storage.
+    bool m_bStreamReadOnly = false;
+
 protected:
     /// @throws css::uno::Exception
     css::uno::Reference< css::io::XStream > TryToGetAcceptableFormat_Impl(
@@ -441,6 +446,9 @@ public:
     // XActiveDataStreamer
     void SAL_CALL setStream(const css::uno::Reference<css::io::XStream>& xStream) override;
     css::uno::Reference<css::io::XStream> SAL_CALL getStream() override;
+
+    // XInitialization
+    void SAL_CALL initialize(const css::uno::Sequence<css::uno::Any>& rArguments) override;
 };
 
 #endif
diff --git a/embeddedobj/source/msole/olemisc.cxx b/embeddedobj/source/msole/olemisc.cxx
index 83485c89d768..2d757a594533 100644
--- a/embeddedobj/source/msole/olemisc.cxx
+++ b/embeddedobj/source/msole/olemisc.cxx
@@ -29,6 +29,7 @@
 #include <com/sun/star/lang/DisposedException.hpp>
 
 #include <cppuhelper/interfacecontainer.h>
+#include <comphelper/sequenceashashmap.hxx>
 
 #include <oleembobj.hxx>
 #include "olepersist.hxx"
@@ -676,4 +677,17 @@ css::uno::Reference<css::io::XStream> OleEmbeddedObject::getStream()
     return m_xObjectStream;
 }
 
+void OleEmbeddedObject::initialize(const uno::Sequence<uno::Any>& rArguments)
+{
+    if (!rArguments.hasElements())
+        return;
+
+    comphelper::SequenceAsHashMap aValues(rArguments[0]);
+    for (const auto& rValue : aValues)
+    {
+        if (rValue.first == "StreamReadOnly")
+            rValue.second >>= m_bStreamReadOnly;
+    }
+}
+
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/embeddedobj/source/msole/olepersist.cxx b/embeddedobj/source/msole/olepersist.cxx
index eedc8c4fe63b..df78e09291ce 100644
--- a/embeddedobj/source/msole/olepersist.cxx
+++ b/embeddedobj/source/msole/olepersist.cxx
@@ -1201,7 +1201,8 @@ void OleEmbeddedObject::StoreToLocation_Impl(
             if ( !xCachedVisualRepresentation.is() )
                 xCachedVisualRepresentation = TryToRetrieveCachedVisualRepresentation_Impl( xTargetStream );
 
-            RemoveVisualCache_Impl( xTargetStream );
+            if (!m_bStreamReadOnly)
+                RemoveVisualCache_Impl(xTargetStream);
         }
     }
 
diff --git a/sw/qa/extras/htmlexport/htmlexport.cxx b/sw/qa/extras/htmlexport/htmlexport.cxx
index d1db894519b0..81d851a7f831 100644
--- a/sw/qa/extras/htmlexport/htmlexport.cxx
+++ b/sw/qa/extras/htmlexport/htmlexport.cxx
@@ -555,7 +555,12 @@ DECLARE_HTMLEXPORT_ROUNDTRIP_TEST(testReqIfOle2, "reqif-ole2.xhtml")
     uno::Reference<io::XSeekable> xStream(xEmbeddedObject->getStream(), uno::UNO_QUERY);
     // This was 80913, the RTF hexdump -> OLE1 binary -> OLE2 conversion was
     // missing.
-    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int64>(38912), xStream->getLength());
+    // Also, this was 38912 when we re-generated the OLE2 preview, which is
+    // wrong, the OLE2 data is 38375 bytes in the ole2.ole (referenced by
+    // reqif-ole2.xhtml). To see that this is the correct value, convert the
+    // hexdump in ole2.ole to binary, remove the ole1 header and check the byte
+    // size.
+    CPPUNIT_ASSERT_EQUAL(static_cast<sal_Int64>(38375), xStream->getLength());
     // Finally the export also failed as it tried to open the stream from the
     // document storage, but the embedded object already opened it, so an
     // exception of type com.sun.star.io.IOException was thrown.
diff --git a/sw/source/filter/html/htmlplug.cxx b/sw/source/filter/html/htmlplug.cxx
index a7725b8f5b59..f067d18be3b3 100644
--- a/sw/source/filter/html/htmlplug.cxx
+++ b/sw/source/filter/html/htmlplug.cxx
@@ -65,6 +65,7 @@
 #include <com/sun/star/io/XActiveDataStreamer.hpp>
 #include <com/sun/star/io/IOException.hpp>
 #include <com/sun/star/embed/XEmbedPersist2.hpp>
+#include <com/sun/star/lang/XInitialization.hpp>
 
 #include <comphelper/embeddedobjectcontainer.hxx>
 #include <comphelper/classids.hxx>
@@ -76,6 +77,7 @@
 #include <filter/msfilter/msoleexp.hxx>
 #include <comphelper/fileurl.hxx>
 #include <osl/file.hxx>
+#include <comphelper/propertyvalue.hxx>
 
 using namespace com::sun::star;
 
@@ -632,10 +634,27 @@ bool SwHTMLParser::InsertEmbed()
     SetSpace( aSpace, aItemSet, aPropInfo, aFrameSet );
 
     // and insert into the document
+    uno::Reference<lang::XInitialization> xObjInitialization(xObj, uno::UNO_QUERY);
+    if (xObjInitialization.is())
+    {
+        // Request that the native data of the embedded object is not modified
+        // during parsing.
+        uno::Sequence<beans::PropertyValue> aValues{ comphelper::makePropertyValue("StreamReadOnly",
+                                                                                   true) };
+        uno::Sequence<uno::Any> aArguments{ uno::makeAny(aValues) };
+        xObjInitialization->initialize(aArguments);
+    }
     SwFrameFormat* pFlyFormat =
         m_xDoc->getIDocumentContentOperations().InsertEmbObject(*m_pPam,
                 ::svt::EmbeddedObjectRef(xObj, embed::Aspects::MSOLE_CONTENT),
                 &aFrameSet);
+    if (xObjInitialization.is())
+    {
+        uno::Sequence<beans::PropertyValue> aValues{ comphelper::makePropertyValue("StreamReadOnly",
+                                                                                   false) };
+        uno::Sequence<uno::Any> aArguments{ uno::makeAny(aValues) };
+        xObjInitialization->initialize(aArguments);
+    }
 
     // set name at FrameFormat
     if( !aName.isEmpty() )