vcl PDF export, norefxobj: copy each object only once

Even if they are referenced multiple times. This is especially important
as objects can refer to each other, creating a cyclic graph. But it also
makes the output a tiny bit smaller.

Change-Id: I561ac319683a19a797282fe259cc68f3a4c50c3e
Reviewed-on: https://gerrit.libreoffice.org/35855
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins <ci@libreoffice.org>
(cherry picked from commit 92ddc0409c8d3276183afdee543d28e1c307c2c7)

2 files changed, 17 insertions, 8 deletions

diff --git a/vcl/source/gdi/pdfwriter_impl.cxx b/vcl/source/gdi/pdfwriter_impl.cxx
index d132b3ff887e..51eed7262bbe 100644
--- a/vcl/source/gdi/pdfwriter_impl.cxx
+++ b/vcl/source/gdi/pdfwriter_impl.cxx
@@ -11697,9 +11697,16 @@ void PDFWriterImpl::writeJPG( JPGEmit& rObject )
     writeReferenceXObject(rObject.m_aReferenceXObject);
 }
 
-sal_Int32 PDFWriterImpl::copyExternalResource(SvMemoryStream& rDocBuffer, filter::PDFObjectElement& rObject)
+sal_Int32 PDFWriterImpl::copyExternalResource(SvMemoryStream& rDocBuffer, filter::PDFObjectElement& rObject, std::map<sal_Int32, sal_Int32>& rCopiedResources)
 {
+    auto it = rCopiedResources.find(rObject.GetObjectValue());
+    if (it != rCopiedResources.end())
+        // This resource was already copied once, nothing to do.
+        return it->second;
+
     sal_Int32 nObject = createObject();
+    // Remember what is the ID of this object in our output.
+    rCopiedResources[rObject.GetObjectValue()] = nObject;
     SAL_INFO("vcl.pdfwriter", "PDFWriterImpl::copyExternalResource: " << rObject.GetObjectValue() << " -> " << nObject);
 
     OStringBuffer aLine;
@@ -11722,7 +11729,7 @@ sal_Int32 PDFWriterImpl::copyExternalResource(SvMemoryStream& rDocBuffer, filter
                 if (pReferenced)
                 {
                     // Copy the referenced object.
-                    sal_Int32 nRef = copyExternalResource(rDocBuffer, *pReferenced);
+                    sal_Int32 nRef = copyExternalResource(rDocBuffer, *pReferenced, rCopiedResources);
 
                     sal_uInt64 nReferenceStart = pDictionary->GetKeyOffset(rItem.first) + rItem.first.getLength();
                     sal_uInt64 nReferenceEnd = pDictionary->GetKeyOffset(rItem.first) + pDictionary->GetKeyValueLength(rItem.first);
@@ -11784,7 +11791,7 @@ sal_Int32 PDFWriterImpl::copyExternalResource(SvMemoryStream& rDocBuffer, filter
                 if (pReferenced)
                 {
                     // Copy the referenced object.
-                    sal_Int32 nRef = copyExternalResource(rDocBuffer, *pReferenced);
+                    sal_Int32 nRef = copyExternalResource(rDocBuffer, *pReferenced, rCopiedResources);
 
                     sal_uInt64 nReferenceStart = pReference->GetObjectElement().GetLocation();
                     sal_uInt64 nReferenceEnd = pReference->GetOffset();
@@ -11833,7 +11840,7 @@ sal_Int32 PDFWriterImpl::copyExternalResource(SvMemoryStream& rDocBuffer, filter
     return nObject;
 }
 
-OString PDFWriterImpl::copyExternalResources(filter::PDFObjectElement& rPage, const OString& rKind)
+OString PDFWriterImpl::copyExternalResources(filter::PDFObjectElement& rPage, const OString& rKind, std::map<sal_Int32, sal_Int32>& rCopiedResources)
 {
     // A name - object ID map, IDs as they appear in our output, not the
     // original ones.
@@ -11863,7 +11870,7 @@ OString PDFWriterImpl::copyExternalResources(filter::PDFObjectElement& rPage, co
             continue;
 
         // Then copying over an object copy its dictionary and its stream.
-        sal_Int32 nObject = copyExternalResource(rDocBuffer, *pValue);
+        sal_Int32 nObject = copyExternalResource(rDocBuffer, *pValue, rCopiedResources);
         aRet[rItem.first] = nObject;
     }
 
@@ -11947,8 +11954,10 @@ void PDFWriterImpl::writeReferenceXObject(ReferenceXObjectEmit& rEmit)
             "Font",
             "XObject"
         };
+        // Maps from source object id (PDF image) to target object id (export result).
+        std::map<sal_Int32, sal_Int32> aCopiedResources;
         for (const auto& rKey : aKeys)
-            aLine.append(copyExternalResources(*pPage, rKey));
+            aLine.append(copyExternalResources(*pPage, rKey, aCopiedResources));
         aLine.append(">>");
         aLine.append(" /BBox [ 0 0 ");
         aLine.append(aSize.Width());
diff --git a/vcl/source/gdi/pdfwriter_impl.hxx b/vcl/source/gdi/pdfwriter_impl.hxx
index 2329cbdec0bb..e1750ab8b73b 100644
--- a/vcl/source/gdi/pdfwriter_impl.hxx
+++ b/vcl/source/gdi/pdfwriter_impl.hxx
@@ -861,10 +861,10 @@ i12626
     void writeReferenceXObject(ReferenceXObjectEmit& rEmit);
     /// Copies resources of a given kind from an external page to the output,
     /// returning what has to be included in the new resource dictionary.
-    OString copyExternalResources(filter::PDFObjectElement& rPage, const OString& rKind);
+    OString copyExternalResources(filter::PDFObjectElement& rPage, const OString& rKind, std::map<sal_Int32, sal_Int32>& rCopiedResources);
     /// Copies a single resource from an external document, returns the new
     /// object ID in our document.
-    sal_Int32 copyExternalResource(SvMemoryStream& rDocBuffer, filter::PDFObjectElement& rObject);
+    sal_Int32 copyExternalResource(SvMemoryStream& rDocBuffer, filter::PDFObjectElement& rObject, std::map<sal_Int32, sal_Int32>& rCopiedResources);
 
     /* tries to find the bitmap by its id and returns its emit data if exists,
        else creates a new emit data block */