sw XHTML export: properly write <li>...</li> around multiple paragraphs

Writer just has a list of text nodes, but ODF and HTML can have multiple
paragraphs inside <li>, in case the non-first text nodes have their
RES_PARATR_LIST_ISCOUNTED set to false.

Improve b6f7a4908d1c3fc610440a1cbbbc1673a53da8b6 (sw XHTML export:
properly write <li>...</li> around paragraphs, 2018-04-11) to make the
decision around when to write </li> based on not only the current but
also the next text node. This way we omit </li> for non-first paragraphs
inside <li>, but keep writing it in other cases.

Change-Id: I994855961aeea92050268dba5f19f125c7ad2b7f
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/94622
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>
Tested-by: Jenkins

4 files changed, 69 insertions, 16 deletions

diff --git a/sw/CppunitTest_sw_htmlexport.mk b/sw/CppunitTest_sw_htmlexport.mk
index f1e2342ad1ba..3315566bb087 100644
--- a/sw/CppunitTest_sw_htmlexport.mk
+++ b/sw/CppunitTest_sw_htmlexport.mk
@@ -27,6 +27,7 @@ $(eval $(call gb_CppunitTest_use_libraries,sw_htmlexport, \
     sfx \
     sot \
     sw \
+    svl \
     svt \
     test \
 	tl \
diff --git a/sw/qa/extras/htmlexport/htmlexport.cxx b/sw/qa/extras/htmlexport/htmlexport.cxx
index 7ea6e300f603..512fd4c070cb 100644
--- a/sw/qa/extras/htmlexport/htmlexport.cxx
+++ b/sw/qa/extras/htmlexport/htmlexport.cxx
@@ -22,6 +22,9 @@
 #include <swmodule.hxx>
 #include <swdll.hxx>
 #include <usrpref.hxx>
+#include <wrtsh.hxx>
+#include <ndtxt.hxx>
+#include <paratr.hxx>
 
 #include <test/htmltesttools.hxx>
 #include <tools/urlobj.hxx>
@@ -33,6 +36,7 @@
 #include <svtools/rtftoken.h>
 #include <filter/msfilter/rtfutil.hxx>
 #include <sot/storage.hxx>
+#include <svl/eitem.hxx>
 
 class HtmlExportTest : public SwModelTestBase, public HtmlTestTools
 {
@@ -1053,6 +1057,55 @@ CPPUNIT_TEST_FIXTURE(SwHtmlDomExportTest, testReqifOle1PDF)
     CPPUNIT_ASSERT_EQUAL(static_cast<sal_uInt32>(39409), pOleNative->GetSize());
 }
 
+CPPUNIT_TEST_FIXTURE(SwHtmlDomExportTest, testMultiParaListItem)
+{
+    // Create a document with 3 list items: A, B&C and D.
+    loadURL("private:factory/swriter", nullptr);
+    SwXTextDocument* pTextDoc = dynamic_cast<SwXTextDocument*>(mxComponent.get());
+    SwWrtShell* pWrtShell = pTextDoc->GetDocShell()->GetWrtShell();
+    pWrtShell->Insert("A");
+    SwDoc* pDoc = pWrtShell->GetDoc();
+    {
+        // Enable numbering.
+        sal_uInt16 nPos = pDoc->MakeNumRule(pDoc->GetUniqueNumRuleName());
+        SwNumRule* pNumRule = pDoc->GetNumRuleTable()[nPos];
+        SwNode& rNode = pWrtShell->GetCursor()->GetPoint()->nNode.GetNode();
+        SwTextNode& rTextNode = *rNode.GetTextNode();
+        rTextNode.SetAttr(SwNumRuleItem(pNumRule->GetName()));
+    }
+    pWrtShell->SplitNode();
+    pWrtShell->Insert("B");
+    pWrtShell->SplitNode();
+    pWrtShell->Insert("C");
+    {
+        // C is in the same list item as B.
+        SwNode& rNode = pWrtShell->GetCursor()->GetPoint()->nNode.GetNode();
+        SwTextNode& rTextNode = *rNode.GetTextNode();
+        rTextNode.SetCountedInList(false);
+    }
+    pWrtShell->SplitNode();
+    pWrtShell->Insert("D");
+
+    uno::Reference<frame::XStorable> xStorable(mxComponent, uno::UNO_QUERY);
+    uno::Sequence<beans::PropertyValue> aStoreProperties = {
+        comphelper::makePropertyValue("FilterName", OUString("HTML (StarWriter)")),
+        comphelper::makePropertyValue("FilterOptions", OUString("xhtmlns=reqif-xhtml")),
+    };
+    xStorable->storeToURL(maTempFile.GetURL(), aStoreProperties);
+
+    SvMemoryStream aStream;
+    HtmlExportTest::wrapFragment(maTempFile, aStream);
+    xmlDocUniquePtr pXmlDoc = parseXmlStream(&aStream);
+    CPPUNIT_ASSERT(pDoc);
+    assertXPathContent(pXmlDoc, "//reqif-xhtml:ol/reqif-xhtml:li[1]/reqif-xhtml:p", "A");
+    assertXPathContent(pXmlDoc, "//reqif-xhtml:ol/reqif-xhtml:li[2]/reqif-xhtml:p[1]", "B");
+    // Without the accompanying fix in place, this test would have failed with:
+    // XPath '//reqif-xhtml:ol/reqif-xhtml:li[2]/reqif-xhtml:p[2]' not found
+    // i.e. </li> was writen before "C", not after "C", so "C" was not in the 2nd list item.
+    assertXPathContent(pXmlDoc, "//reqif-xhtml:ol/reqif-xhtml:li[2]/reqif-xhtml:p[2]", "C");
+    assertXPathContent(pXmlDoc, "//reqif-xhtml:ol/reqif-xhtml:li[3]/reqif-xhtml:p", "D");
+}
+
 CPPUNIT_PLUGIN_IMPLEMENT();
 
 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */
diff --git a/sw/source/filter/html/htmlatr.cxx b/sw/source/filter/html/htmlatr.cxx
index 9bee7fc22157..fa70b2767f09 100644
--- a/sw/source/filter/html/htmlatr.cxx
+++ b/sw/source/filter/html/htmlatr.cxx
@@ -207,7 +207,6 @@ struct SwHTMLTextCollOutputInfo
     bool bParaPossible;         // a </P> may be output additionally
     bool bOutPara;              // a </P> is supposed to be output
     bool bOutDiv;               // write a </DIV>
-    bool bOutLi = false;        // write a </li>
 
     SwHTMLTextCollOutputInfo() :
         bInNumberBulletList( false ),
@@ -764,14 +763,8 @@ static void OutHTML_SwFormat( Writer& rWrt, const SwFormat& rFormat,
         html.start(OOO_STRING_SVTOOLS_HTML_li);
         if( USHRT_MAX != nNumStart )
             html.attribute(OOO_STRING_SVTOOLS_HTML_O_value, OString::number(nNumStart));
-        if (rHWrt.mbXHTML)
-        {
-            rWrt.Strm().WriteCharPtr(">");
-            rInfo.bOutLi = true;
-        }
-        else
-            // Finish the opening element, but don't close it.
-            html.characters(OString());
+        // Finish the opening element, but don't close it.
+        html.characters(OString());
     }
 
     if( rHWrt.m_nDefListLvl > 0 && !bForceDL )
@@ -1022,11 +1015,7 @@ static void OutHTML_SwFormatOff( Writer& rWrt, const SwHTMLTextCollOutputInfo& r
         rHWrt.m_bLFPossible = true;
     }
 
-    if (rInfo.bOutLi)
-        HTMLOutFuncs::Out_AsciiTag(rWrt.Strm(), rHWrt.GetNamespace() + OOO_STRING_SVTOOLS_HTML_li,
-                                   false);
-
-    // if necessary, close a bulleted or numbered list
+    // if necessary, close the list item, then close a bulleted or numbered list
     if( rInfo.bInNumberBulletList )
     {
         rHWrt.FillNextNumInfo();
diff --git a/sw/source/filter/html/htmlnumwriter.cxx b/sw/source/filter/html/htmlnumwriter.cxx
index 8977e0dbf684..ea3c949c3801 100644
--- a/sw/source/filter/html/htmlnumwriter.cxx
+++ b/sw/source/filter/html/htmlnumwriter.cxx
@@ -279,8 +279,18 @@ Writer& OutHTML_NumberBulletListEnd( SwHTMLWriter& rWrt,
 {
     SwHTMLNumRuleInfo& rInfo = rWrt.GetNumInfo();
     bool bSameRule = rNextInfo.GetNumRule() == rInfo.GetNumRule();
-    if( bSameRule && rNextInfo.GetDepth() >= rInfo.GetDepth() &&
-        !rNextInfo.IsRestart() )
+    bool bListEnd = !bSameRule || rNextInfo.GetDepth() < rInfo.GetDepth() || rNextInfo.IsRestart();
+
+    if (rWrt.mbXHTML)
+    {
+        if (bListEnd || (!bListEnd && rNextInfo.IsNumbered()))
+        {
+            HTMLOutFuncs::Out_AsciiTag(rWrt.Strm(),
+                                       rWrt.GetNamespace() + OOO_STRING_SVTOOLS_HTML_li, false);
+        }
+    }
+
+    if (!bListEnd)
     {
         return rWrt;
     }