tdf#83227 oox: reuse RelId in DML/VML export for the same graphic

So that large images are written only once to the ZIP container when
they are exported using both markups. This affects drawinglayer images,
the Writer ones are handled directly in sw and were already
deduplicated.

(cherry picked from commit b484e9814c66d8d51cea974390963a6944bc9d73)

Conflicts:
oox/source/export/drawingml.cxx

Change-Id: Iff7c769329b42939833056b727b070f6a60da5e3
Reviewed-on: https://gerrit.libreoffice.org/18581
Reviewed-by: Caolán McNamara <caolanm@redhat.com>
Tested-by: Caolán McNamara <caolanm@redhat.com>

8 files changed, 77 insertions, 5 deletions

diff --git a/include/oox/export/drawingml.hxx b/include/oox/export/drawingml.hxx
index 4bfaf3a6e572..246960c22fda 100644
--- a/include/oox/export/drawingml.hxx
+++ b/include/oox/export/drawingml.hxx
@@ -73,6 +73,10 @@ public:
     virtual void WriteOutliner(const OutlinerParaObject& rParaObj) = 0;
     /// Write the contents of the textbox that is associated to this shape.
     virtual void WriteTextBox(css::uno::Reference<css::drawing::XShape> xShape) = 0;
+    /// Look up the RelId of a graphic based on its checksum.
+    virtual OUString FindRelId(sal_uInt32 nChecksum) = 0;
+    /// Store the RelId of a graphic based on its checksum.
+    virtual void CacheRelId(sal_uInt32 nChecksum, const OUString& rRelId) = 0;
 protected:
     DMLTextExport() {}
     virtual ~DMLTextExport() {}
diff --git a/include/oox/export/vmlexport.hxx b/include/oox/export/vmlexport.hxx
index a87ed16eb90a..da274c1039cf 100644
--- a/include/oox/export/vmlexport.hxx
+++ b/include/oox/export/vmlexport.hxx
@@ -39,6 +39,10 @@ public:
     virtual oox::drawingml::DrawingML& GetDrawingML() = 0;
     /// Write the contents of the textbox that is associated to this shape in VML format.
     virtual void WriteVMLTextBox(css::uno::Reference<css::drawing::XShape> xShape) = 0;
+    /// Look up the RelId of a graphic based on its checksum.
+    virtual OUString FindRelId(sal_uInt32 nChecksum) = 0;
+    /// Store the RelId of a graphic based on its checksum.
+    virtual void CacheRelId(sal_uInt32 nChecksum, const OUString& rRelId) = 0;
 protected:
     VMLTextExport() {}
     virtual ~VMLTextExport() {}
diff --git a/oox/source/export/drawingml.cxx b/oox/source/export/drawingml.cxx
index edff9b933dd1..66dfb1ffcbf4 100644
--- a/oox/source/export/drawingml.cxx
+++ b/oox/source/export/drawingml.cxx
@@ -762,7 +762,7 @@ void DrawingML::WriteOutline( Reference<XPropertySet> rXPropSet )
     mpFS->endElementNS( XML_a, XML_ln );
 }
 
-OUString DrawingML::WriteImage( const OUString& rURL, bool bRelPathToMedia )
+bool lcl_URLToGraphic(const OUString& rURL, Graphic& rGraphic)
 {
     OString aURLBS(OUStringToOString(rURL, RTL_TEXTENCODING_UTF8));
 
@@ -771,9 +771,18 @@ OUString DrawingML::WriteImage( const OUString& rURL, bool bRelPathToMedia )
 
     if ( index != -1 )
     {
-        DBG(fprintf (stderr, "begin: %ld %s\n", long( sizeof( aURLBegin ) ), USS( rURL ) + RTL_CONSTASCII_LENGTH( aURLBegin ) ));
-        Graphic aGraphic = GraphicObject( aURLBS.copy(RTL_CONSTASCII_LENGTH(aURLBegin)) ).GetTransformedGraphic ();
+        rGraphic = GraphicObject(aURLBS.copy(RTL_CONSTASCII_LENGTH(aURLBegin))).GetTransformedGraphic();
+        return true;
+    }
+
+    return false;
+}
 
+OUString DrawingML::WriteImage( const OUString& rURL, bool bRelPathToMedia )
+{
+    Graphic aGraphic;
+    if (lcl_URLToGraphic(rURL, aGraphic))
+    {
         return WriteImage( aGraphic , bRelPathToMedia );
     }
     else
@@ -923,7 +932,23 @@ OUString DrawingML::WriteImage( const Graphic& rGraphic , bool bRelPathToMedia )
 
 OUString DrawingML::WriteBlip( Reference< XPropertySet > rXPropSet, const OUString& rURL, bool bRelPathToMedia, const Graphic *pGraphic )
 {
-    OUString sRelId = pGraphic ? WriteImage( *pGraphic, bRelPathToMedia ) : WriteImage( rURL, bRelPathToMedia );
+    OUString sRelId;
+    sal_uInt32 nChecksum = 0;
+    if (!rURL.isEmpty() && mpTextExport)
+    {
+        Graphic aGraphic;
+        if (lcl_URLToGraphic(rURL, aGraphic))
+        {
+            nChecksum = aGraphic.GetChecksum();
+            sRelId = mpTextExport->FindRelId(nChecksum);
+        }
+    }
+    if (sRelId.isEmpty())
+    {
+        sRelId = pGraphic ? WriteImage( *pGraphic, bRelPathToMedia ) : WriteImage( rURL, bRelPathToMedia );
+        if (!rURL.isEmpty() && mpTextExport)
+            mpTextExport->CacheRelId(nChecksum, sRelId);
+    }
     sal_Int16 nBright = 0;
     sal_Int32 nContrast = 0;
 
diff --git a/oox/source/export/vmlexport.cxx b/oox/source/export/vmlexport.cxx
index a0a1ac7677db..557bb8c18d9e 100644
--- a/oox/source/export/vmlexport.cxx
+++ b/oox/source/export/vmlexport.cxx
@@ -605,7 +605,14 @@ void VMLExport::Commit( EscherPropertyContainer& rProps, const Rectangle& rRect
                         aStream.Seek(0);
                         Graphic aGraphic;
                         GraphicConverter::Import(aStream, aGraphic);
-                        OUString aImageId = m_pTextExport->GetDrawingML().WriteImage( aGraphic );
+
+                        sal_uInt32 nChecksum = aGraphic.GetChecksum();
+                        OUString aImageId = m_pTextExport->FindRelId(nChecksum);
+                        if (aImageId.isEmpty())
+                        {
+                            aImageId = m_pTextExport->GetDrawingML().WriteImage( aGraphic );
+                            m_pTextExport->CacheRelId(nChecksum, aImageId);
+                        }
                         pAttrList->add(FSNS(XML_r, XML_id), OUStringToOString(aImageId, RTL_TEXTENCODING_UTF8));
                         imageData = true;
                     }
diff --git a/sw/qa/extras/ooxmlexport/data/tdf83227.docx b/sw/qa/extras/ooxmlexport/data/tdf83227.docx
new file mode 100644
index 000000000000..bca19a955ecf
--- /dev/null
+++ b/sw/qa/extras/ooxmlexport/data/tdf83227.docx
diff --git a/sw/qa/extras/ooxmlexport/ooxmlexport4.cxx b/sw/qa/extras/ooxmlexport/ooxmlexport4.cxx
index 9a495fd2a944..56905077b4a2 100644
--- a/sw/qa/extras/ooxmlexport/ooxmlexport4.cxx
+++ b/sw/qa/extras/ooxmlexport/ooxmlexport4.cxx
@@ -798,6 +798,18 @@ DECLARE_OOXMLEXPORT_TEST(testSimpleSdts, "simple-sdts.docx")
 
 }
 
+DECLARE_OOXMLEXPORT_TEST(testTdf83227, "tdf83227.docx")
+{
+    // Bug document contains a rotated image, which is handled as a draw shape (not as a Writer image) on export.
+    if (!mbExported)
+        return;
+
+    uno::Reference<packages::zip::XZipFileAccess2> xNameAccess = packages::zip::ZipFileAccess::createWithURL(comphelper::getComponentContext(m_xSFactory), maTempFile.GetURL());
+    CPPUNIT_ASSERT_EQUAL(true, bool(xNameAccess->hasByName("word/media/image1.png")));
+    // This was also true, image was written twice.
+    CPPUNIT_ASSERT_EQUAL(false, bool(xNameAccess->hasByName("word/media/image2.png")));
+}
+
 CPPUNIT_PLUGIN_IMPLEMENT();
 
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/sw/source/filter/ww8/docxattributeoutput.cxx b/sw/source/filter/ww8/docxattributeoutput.cxx
index 8d378218646e..4cba42779c80 100644
--- a/sw/source/filter/ww8/docxattributeoutput.cxx
+++ b/sw/source/filter/ww8/docxattributeoutput.cxx
@@ -4046,6 +4046,22 @@ void DocxAttributeOutput::WriteSrcRect(const SdrObject* pSdrObj )
 void DocxAttributeOutput::ClearRelIdCache()
 {
     m_aRelIdCache.clear();
+    m_aSdrRelIdCache.clear();
+}
+
+OUString DocxAttributeOutput::FindRelId(sal_uInt32 nChecksum)
+{
+    OUString aRet;
+
+    if (m_aSdrRelIdCache.find(nChecksum) != m_aSdrRelIdCache.end())
+        aRet = m_aSdrRelIdCache[nChecksum];
+
+    return aRet;
+}
+
+void DocxAttributeOutput::CacheRelId(sal_uInt32 nChecksum, const OUString& rRelId)
+{
+     m_aSdrRelIdCache[nChecksum] = rRelId;
 }
 
 void DocxAttributeOutput::FlyFrameGraphic( const SwGrfNode* pGrfNode, const Size& rSize, const SwFlyFrmFmt* pOLEFrmFmt, SwOLENode* pOLENode, const SdrObject* pSdrObj )
diff --git a/sw/source/filter/ww8/docxattributeoutput.hxx b/sw/source/filter/ww8/docxattributeoutput.hxx
index a5e13d9d07f5..f8ae6267b4bc 100644
--- a/sw/source/filter/ww8/docxattributeoutput.hxx
+++ b/sw/source/filter/ww8/docxattributeoutput.hxx
@@ -901,6 +901,8 @@ private:
 
     /// RelId <-> Graphic* cache, so that in case of alternate content, the same graphic only gets written once.
     std::map<const Graphic*, OString> m_aRelIdCache;
+    /// RelId <-> sal_uInt32 (bitmap checksum) cache, similar to m_aRelIdCache, but used for non-Writer graphics, handled in oox.
+    std::map<sal_uInt32, OUString> m_aSdrRelIdCache;
 
     /// members to control the existence of grabbagged SDT properties in the paragraph
     sal_Int32 m_nParagraphSdtPrToken;
@@ -958,6 +960,8 @@ public:
     virtual void WriteVMLTextBox(css::uno::Reference<css::drawing::XShape> xShape) SAL_OVERRIDE;
     /// DMLTextExport
     virtual void WriteTextBox(css::uno::Reference<css::drawing::XShape> xShape) SAL_OVERRIDE;
+    virtual OUString FindRelId(sal_uInt32 nChecksum) SAL_OVERRIDE;
+    virtual void CacheRelId(sal_uInt32 nChecksum, const OUString& rRelId) SAL_OVERRIDE;
     virtual oox::drawingml::DrawingML& GetDrawingML() SAL_OVERRIDE;
 
     void BulletDefinition(int nId, const Graphic& rGraphic, Size aSize) SAL_OVERRIDE;