[Libreoffice-commits] core.git: include/xmloff xmloff/source

[Tomaž Vajngerl]

 include/xmloff/XMLFontAutoStylePool.hxx      |    5 +
 xmloff/source/style/XMLFontAutoStylePool.cxx |   80 +++++++++++++++++++++++----
 2 files changed, 75 insertions(+), 10 deletions(-)

New commits:
commit b262e6e5da0c2e66e00cd6d68d7c2df11a34ff5c
Author: Tomaž Vajngerl <tomaz.vajngerl at collabora.co.uk>
Date:   Tue Jun 26 20:06:40 2018 +0200

    xmloff: prevent embedding same font files with checksumming
    
    When embedding fonts to ODF documents we can get into a situation
    where the same font is embedded multiple times and this wastes
    space. To prevent this we calculate the checksum of the font first
    and track the fonts that are embedded. When the font has the same
    checksum, the path to already embedded font is returned.
    
    Change-Id: I087c42d9a63b1697b9b4ea985dbce2fda760a996
    Reviewed-on: https://gerrit.libreoffice.org/56507
    Tested-by: Jenkins
    Reviewed-by: Tomaž Vajngerl <quikee at gmail.com>

diff --git a/include/xmloff/XMLFontAutoStylePool.hxx b/include/xmloff/XMLFontAutoStylePool.hxx
index e747bb8b20d8..530d05b2762a 100644
--- a/include/xmloff/XMLFontAutoStylePool.hxx
+++ b/include/xmloff/XMLFontAutoStylePool.hxx
@@ -26,6 +26,7 @@
 #include <tools/fontenum.hxx>
 #include <salhelper/simplereferenceobject.hxx>
 #include <set>
+#include <unordered_map>
 #include <memory>
 
 class XMLFontAutoStylePool_Impl;
@@ -33,13 +34,15 @@ class SvXMLExport;
 
 class XMLOFF_DLLPUBLIC XMLFontAutoStylePool : public salhelper::SimpleReferenceObject
 {
+private:
     SvXMLExport& rExport;
 
     std::unique_ptr<XMLFontAutoStylePool_Impl> m_pFontAutoStylePool;
     std::set<OUString> m_aNames;
     bool m_bTryToEmbedFonts;
+    std::unordered_map<OString, OUString> m_aEmbeddedFontFiles;
 
-    OUString embedFontFile( const OUString& fontUrl );
+    OUString embedFontFile(OUString const & rFileUrl, OUString const & rFamilyName);
 
 protected:
 
diff --git a/xmloff/source/style/XMLFontAutoStylePool.cxx b/xmloff/source/style/XMLFontAutoStylePool.cxx
index 1c13ea75cff4..1fb06598f841 100644
--- a/xmloff/source/style/XMLFontAutoStylePool.cxx
+++ b/xmloff/source/style/XMLFontAutoStylePool.cxx
@@ -34,6 +34,7 @@
 #include <com/sun/star/ucb/SimpleFileAccess.hpp>
 
 #include <XMLBase64Export.hxx>
+#include <comphelper/hash.hxx>
 
 using namespace ::com::sun::star;
 using namespace ::com::sun::star::uno;
@@ -367,7 +368,7 @@ void XMLFontAutoStylePool::exportXML()
                 if (!fontFilesMap.count(sFileUrl))
                 {
                     const OUString docUrl = bExportFlat ?
-                                                lcl_checkFontFile(sFileUrl) : embedFontFile(sFileUrl);
+                                                lcl_checkFontFile(sFileUrl) : embedFontFile(sFileUrl, pEntry->GetFamilyName());
                     if (!docUrl.isEmpty())
                         fontFilesMap[sFileUrl] = docUrl;
                     else
@@ -429,10 +430,72 @@ void XMLFontAutoStylePool::exportXML()
     }
 }
 
-OUString XMLFontAutoStylePool::embedFontFile( const OUString& fileUrl )
+OUString getFreeFontName(uno::Reference<embed::XStorage> const & rxStorage, OUString const & rFamilyName)
+{
+    OUString sName;
+    int nIndex = 1;
+    do
+    {
+        sName = "Font_" +
+                rFamilyName.replaceAll(" ", "_") + "_" +
+                OUString::number(nIndex) + ".ttf";
+        nIndex++;
+    } while (rxStorage->hasByName(sName));
+
+    return sName;
+}
+
+OString convertToHashString(std::vector<unsigned char> const & rHash)
+{
+    std::stringstream aStringStream;
+    for (auto const & rByte : rHash)
+    {
+        aStringStream << std::setw(2) << std::setfill('0') << std::hex << int(rByte);
+    }
+
+    return OString(aStringStream.str().c_str());
+}
+
+OString getFileHash(OUString const & rFileUrl)
+{
+    OString aHash;
+    osl::File aFile(rFileUrl);
+    if (aFile.open(osl_File_OpenFlag_Read) != osl::File::E_None)
+        return aHash;
+
+    comphelper::Hash aHashEngine(comphelper::HashType::SHA512);
+    for (;;)
+    {
+        sal_Int8 aBuffer[4096];
+        sal_uInt64 nReadSize;
+        sal_Bool bEof;
+        if (aFile.isEndOfFile(&bEof) != osl::File::E_None)
+        {
+            SAL_WARN("xmloff", "Error reading font file " << rFileUrl);
+            return aHash;
+        }
+        if (bEof)
+            break;
+        if (aFile.read(aBuffer, 4096, nReadSize) != osl::File::E_None)
+        {
+            SAL_WARN("xmloff", "Error reading font file " << rFileUrl);
+            return aHash;
+        }
+        if (nReadSize == 0)
+            break;
+        aHashEngine.update(reinterpret_cast<unsigned char*>(aBuffer), nReadSize);
+    }
+    return convertToHashString(aHashEngine.finalize());
+}
+
+OUString XMLFontAutoStylePool::embedFontFile(OUString const & fileUrl, OUString const & rFamilyName)
 {
     try
     {
+        OString sHashString = getFileHash(fileUrl);
+        if (m_aEmbeddedFontFiles.find(sHashString) != m_aEmbeddedFontFiles.end())
+            return m_aEmbeddedFontFiles.at(sHashString);
+
         osl::File file( fileUrl );
         if( file.open( osl_File_OpenFlag_Read ) != osl::File::E_None )
             return OUString();
@@ -443,12 +506,9 @@ OUString XMLFontAutoStylePool::embedFontFile( const OUString& fileUrl )
         uno::Reference< embed::XStorage > storage;
         storage.set( GetExport().GetTargetStorage()->openStorageElement( "Fonts",
             ::embed::ElementModes::WRITE ), uno::UNO_QUERY_THROW );
-        int index = 0;
-        OUString name;
-        do
-        {
-            name = "font" + OUString::number( ++index ) + ".ttf";
-        } while( storage->hasByName( name ) );
+
+        OUString name = getFreeFontName(storage, rFamilyName);
+
         uno::Reference< io::XOutputStream > outputStream;
         outputStream.set( storage->openStreamElement( name, ::embed::ElementModes::WRITE ), UNO_QUERY_THROW );
         uno::Reference < beans::XPropertySet > propertySet( outputStream, uno::UNO_QUERY );
@@ -484,7 +544,9 @@ OUString XMLFontAutoStylePool::embedFontFile( const OUString& fileUrl )
             if( transaction.is())
             {
                 transaction->commit();
-                return "Fonts/" + name;
+                OUString sInternalName = "Fonts/" + name;
+                m_aEmbeddedFontFiles.emplace(sHashString, sInternalName);
+                return sInternalName;
             }
         }
     } catch( const Exception& e )