XHTML import: <br></br> is one line break

But only for XHTML, HTML stays unchanged to be in sync with web
browsers.

Change-Id: I3a1cf6651dab565bafece68963acb112ae715cdd
Reviewed-on: https://gerrit.libreoffice.org/59727
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Jenkins

2 files changed, 24 insertions, 1 deletions

diff --git a/svtools/qa/unit/testHtmlReader.cxx b/svtools/qa/unit/testHtmlReader.cxx
index b09abaaabdd5..6df6b5541c2d 100644
--- a/svtools/qa/unit/testHtmlReader.cxx
+++ b/svtools/qa/unit/testHtmlReader.cxx
@@ -22,8 +22,11 @@ class TestHTMLParser : public HTMLParser
 public:
     TestHTMLParser(SvStream& rStream);
     virtual void NextToken(HtmlTokenId nToken) override;
+    /// Make this public for test purposes.
+    using HTMLParser::SetNamespace;
 
     OUString m_aDocument;
+    int m_nLineBreakCount = 0;
 };
 
 TestHTMLParser::TestHTMLParser(SvStream& rStream)
@@ -35,6 +38,8 @@ void TestHTMLParser::NextToken(HtmlTokenId nToken)
 {
     if (nToken == HtmlTokenId::TEXTTOKEN)
         m_aDocument += aToken;
+    else if (nToken == HtmlTokenId::LINEBREAK)
+        ++m_nLineBreakCount;
 }
 
 /// Tests HTMLParser.
@@ -42,9 +47,11 @@ class Test : public CppUnit::TestFixture
 {
 public:
     void testTdf114428();
+    void testLineBreak();
 
     CPPUNIT_TEST_SUITE(Test);
     CPPUNIT_TEST(testTdf114428);
+    CPPUNIT_TEST(testLineBreak);
     CPPUNIT_TEST_SUITE_END();
 };
 
@@ -63,6 +70,21 @@ void Test::testTdf114428()
     CPPUNIT_ASSERT_EQUAL(OUString("hello"), xParser->m_aDocument.trim());
 }
 
+void Test::testLineBreak()
+{
+    SvMemoryStream aStream;
+    OString aDocument("aaa<br></br>bbb");
+    aStream.WriteBytes(aDocument.getStr(), aDocument.getLength());
+    aStream.Seek(0);
+
+    tools::SvRef<TestHTMLParser> xParser = new TestHTMLParser(aStream);
+    xParser->SetNamespace("reqif-xhtml");
+    xParser->CallParser();
+
+    // This was 2, <br></br> was interpreted as 2 line breaks in XHTML mode.
+    CPPUNIT_ASSERT_EQUAL(1, xParser->m_nLineBreakCount);
+}
+
 CPPUNIT_TEST_SUITE_REGISTRATION(Test);
 }
 
diff --git a/svtools/source/svhtml/parhtml.cxx b/svtools/source/svhtml/parhtml.cxx
index 14ed546ca9fe..c80fd00d77ed 100644
--- a/svtools/source/svhtml/parhtml.cxx
+++ b/svtools/source/svhtml/parhtml.cxx
@@ -1120,10 +1120,11 @@ HtmlTokenId HTMLParser::GetNextToken_()
                             // and there is an off token, return off token instead
                             nRet = static_cast<HtmlTokenId>(static_cast<int>(nRet) + 1);
                          }
-                         else if( HtmlTokenId::LINEBREAK!=nRet )
+                         else if( HtmlTokenId::LINEBREAK!=nRet || !maNamespace.isEmpty())
                          {
                             // and there is no off token, return unknown token.
                             // (except for </BR>, that is treated like <BR>)
+                            // No exception for XHTML, though.
                             nRet = HtmlTokenId::UNKNOWNCONTROL_OFF;
                          }
                     }