[Libreoffice-commits] core.git: sax/source sw/qa

[Mark Hung]

 sax/source/tools/fastserializer.cxx           |   25 ++-----------------------
 sw/qa/extras/ooxmlexport/data/nonbmpchar.docx |binary
 sw/qa/extras/ooxmlexport/ooxmlexport5.cxx     |    9 +++++++++
 3 files changed, 11 insertions(+), 23 deletions(-)

New commits:
commit 7ec5ba47783ac8b2d0141109d0efe6b20b363ced
Author: Mark Hung <marklh9 at gmail.com>
Date:   Sun Jan 10 00:28:14 2016 +0800

    Fix FastSaxSerializer::write() for non-BMP unicode characters.
    
    Change-Id: I6ee9d028813c970b9ac5b9f0574af932e73111bd
    Reviewed-on: https://gerrit.libreoffice.org/21293
    Tested-by: Jenkins <ci at libreoffice.org>
    Reviewed-by: Michael Stahl <mstahl at redhat.com>

diff --git a/sax/source/tools/fastserializer.cxx b/sax/source/tools/fastserializer.cxx
index 37ed847..8437ff0 100644
--- a/sax/source/tools/fastserializer.cxx
+++ b/sax/source/tools/fastserializer.cxx
@@ -94,29 +94,8 @@ namespace sax_fastparser {
 
     void FastSaxSerializer::write( const OUString& sOutput, bool bEscape )
     {
-        const sal_Int32 nLength = sOutput.getLength();
-        for (sal_Int32 i = 0; i < nLength; ++i)
-        {
-            const sal_Unicode cUnicode = sOutput[ i ];
-            const char cChar = cUnicode;
-            if (cUnicode & 0xff80)
-            {
-                write( OString(&cUnicode, 1, RTL_TEXTENCODING_UTF8) );
-            }
-            else if(bEscape) switch( cChar )
-            {
-                case '<':   writeBytes( "<", 4 );     break;
-                case '>':   writeBytes( ">", 4 );     break;
-                case '&':   writeBytes( "&", 5 );    break;
-                case '\'':  writeBytes( "'", 6 );   break;
-                case '"':   writeBytes( """, 6 );   break;
-                case '\n':  writeBytes( "
", 5 );    break;
-                case '\r':  writeBytes( "
", 5 );    break;
-                default:    writeBytes( &cChar, 1 );     break;
-            }
-            else
-                writeBytes( &cChar, 1 );
-        }
+        write( sOutput.toUtf8(), bEscape );
+
     }
 
     void FastSaxSerializer::write( const OString& sOutput, bool bEscape )
diff --git a/sw/qa/extras/ooxmlexport/data/nonbmpchar.docx b/sw/qa/extras/ooxmlexport/data/nonbmpchar.docx
new file mode 100644
index 0000000..a6b5d41
Binary files /dev/null and b/sw/qa/extras/ooxmlexport/data/nonbmpchar.docx differ
diff --git a/sw/qa/extras/ooxmlexport/ooxmlexport5.cxx b/sw/qa/extras/ooxmlexport/ooxmlexport5.cxx
index 9deea3e..68c0e2d 100644
--- a/sw/qa/extras/ooxmlexport/ooxmlexport5.cxx
+++ b/sw/qa/extras/ooxmlexport/ooxmlexport5.cxx
@@ -1011,6 +1011,15 @@ DECLARE_OOXMLEXPORT_TEST(testOO106020, "ooo106020-1.odt")
         assertXPath(pXmlDoc, "//w:tbl", 1);
 }
 
+DECLARE_OOXMLEXPORT_TEST(testNonBMPChar, "nonbmpchar.docx")
+{
+    sal_uInt32  nCh = 0x24b62;
+    OUString aExpected( &nCh, 1);
+    // Assert that UTF8 encoded non-BMP Unicode character is correct
+    uno::Reference<text::XTextRange> xTextRange1 = getRun(getParagraph(1), 1);
+    CPPUNIT_ASSERT_EQUAL(aExpected, xTextRange1->getString());
+}
+
 CPPUNIT_PLUGIN_IMPLEMENT();
 
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */