tdf#108806: convert CRLF into space in OOXML text

Change-Id: I8e2e108a705ecdb55c096a589d83d51c48b0b83c
Reviewed-on: https://gerrit.libreoffice.org/39286Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

sw/qa/extras/ooxmlimport/ooxmlimport.cxx

@@ -1291,6 +1291,16 @@ DECLARE_OOXMLIMPORT_TEST(testTdf108408, "tdf108408.docx")
     CPPUNIT_ASSERT_EQUAL(double(20), getProperty<double>(xRun, "CharHeight"));
 }
 
+DECLARE_OOXMLIMPORT_TEST(testTdf108806, "tdf108806.docx")
+{
+    // tdf#108806:The CRLF in the text contents of XML must be converted to single spaces.
+    CPPUNIT_ASSERT_EQUAL(1, getParagraphs());
+    uno::Reference< text::XTextRange > paragraph = getParagraph(1);
+    CPPUNIT_ASSERT_EQUAL(
+        OUString("First part of a line (before CRLF). Second part of the same line (after CRLF)."),
+        paragraph->getString());
+}
+
 DECLARE_OOXMLIMPORT_TEST(testVmlAdjustments, "vml-adjustments.docx")
 {
     uno::Reference<beans::XPropertySet> xPropertySet(getShape(1), uno::UNO_QUERY);

writerfilter/source/ooxml/OOXMLFastContextHandler.cxx

@@ -608,9 +608,14 @@ void OOXMLFastContextHandler::endTxbxContent()
 void OOXMLFastContextHandler::text(const OUString & sText)
 {
     if (isForwardEvents())
+    {
+        // tdf#108806: CRLFs in XML were converted to \n before this point.
+        // These must be converted to spaces before further processing.
+        OUString sNormalizedText = sText.replaceAll("\n", " ");
         mpStream->utext(reinterpret_cast < const sal_uInt8 * >
-                        (sText.getStr()),
-                        sText.getLength());
+                        (sNormalizedText.getStr()),
+                        sNormalizedText.getLength());
+    }
 }
 
 void OOXMLFastContextHandler::positionOffset(const OUString& rText)