writerfilter: implement RTF derived styles defaulting

It turns out that the situation fixed in commit
1be0a3fa also applies to the definition
of the styles themselves.

To implement the same style import as Word, the style definitions need
to be stored twice: once as read from the file, and another time with
attributes defaulted and deduplicated vs. the parent style; the second
representation is then sent to the domain mapper.

To make this easier, add a bool parameter to cloneAndDeduplicate()
to disable the implicit pPr dereferencing that happens when creating the
hard formatted paragraph properties (this could potentially be cleaned
up further if those paragraph properties would use pPr wrapper
themselves).

Also implement defaulting of line spacing in getDefaultSPRM().

Change-Id: I4810e917697b3af244e5dbdd7f5a45b4767c93fc
Reviewed-on: https://gerrit.libreoffice.org/70320
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.com>

sw/qa/extras/rtfexport/data/para-style-bottom-margin-2.rtf

+{\rtf1
+\ansi\ansicpg1252\deff0
+{\fonttbl{\f0\fbidi \froman\fcharset0\fprq2 Times New Roman;}}
+{\stylesheet
+{\ql \li0\ri0\sa200\sl276\slmult1\widctlpar\wrapdefault\aspalpha\aspnum\faauto\adjustright\rin0\lin0\itap0 \snext0 Normal;}
+{\s15\ql \li720\ri0\sa200\sl276\slmult1\widctlpar\wrapdefault\aspalpha\aspnum\faauto\adjustright\rin0\lin720\itap0\contextualspace \sbasedon0 \snext15 List Paragraph;}
+}
+\pard\plain \ltrpar\s15\ql \li720\ri0\sl276\slmult1\widctlpar\wrapdefault\aspalpha\aspnum\faauto\adjustright\rin0\lin720\itap0\contextualspace
+\par
+\pard\plain
+\par
+}

sw/qa/extras/rtfexport/rtfexport5.cxx

@@ -194,6 +194,33 @@ DECLARE_RTFEXPORT_TEST(testParaBottomMargin, "para-bottom-margin.rtf")
     CPPUNIT_ASSERT_EQUAL(sal_Int32(2), getProperty<sal_Int32>(getParagraph(1), "ParaTopMargin"));
 }
 
+DECLARE_RTFIMPORT_TEST(testParaStyleBottomMargin2, "para-style-bottom-margin-2.rtf")
+{
+    uno::Reference<beans::XPropertySet> xPropertySet(
+        getStyles("ParagraphStyles")->getByName("Standard"), uno::UNO_QUERY);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(353), getProperty<sal_Int32>(xPropertySet, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP,
+                         getProperty<style::LineSpacing>(xPropertySet, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(115),
+                         getProperty<style::LineSpacing>(xPropertySet, "ParaLineSpacing").Height);
+
+    // the derived style contains \sa200, as does its parent
+    uno::Reference<beans::XPropertySet> xPropertySet1(
+        getStyles("ParagraphStyles")->getByName("List Paragraph"), uno::UNO_QUERY);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(353), getProperty<sal_Int32>(xPropertySet1, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP,
+                         getProperty<style::LineSpacing>(xPropertySet1, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(115),
+                         getProperty<style::LineSpacing>(xPropertySet1, "ParaLineSpacing").Height);
+    // for the paragraph there is no \saN, so it should default to 0
+    auto const xPara(getParagraph(1));
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(0), getProperty<sal_Int32>(xPara, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP,
+                         getProperty<style::LineSpacing>(xPara, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(115),
+                         getProperty<style::LineSpacing>(xPara, "ParaLineSpacing").Height);
+}
+
 DECLARE_RTFEXPORT_TEST(testFdo66040, "fdo66040.rtf")
 {
     uno::Reference<drawing::XDrawPageSupplier> xDrawPageSupplier(mxComponent, uno::UNO_QUERY);

sw/qa/extras/rtfimport/data/para-style-bottom-margin.rtf

+{\rtf1
+\ansi\ansicpg1252\deff0
+{\fonttbl
+{\f0 Times New Roman;}
+}
+{\stylesheet
+{\sa200\sl276\slmult1 \snext0 Normal;}
+{\s19 \sbasedon0 \snext0 toc 1;}
+}
+\pard\plain \s19
+foo
+\par
+\par
+}

sw/qa/extras/rtfimport/rtfimport.cxx

@@ -19,6 +19,8 @@
 #include <com/sun/star/lang/IndexOutOfBoundsException.hpp>
 #include <com/sun/star/lang/XServiceInfo.hpp>
 #include <com/sun/star/style/BreakType.hpp>
+#include <com/sun/star/style/LineSpacing.hpp>
+#include <com/sun/star/style/LineSpacingMode.hpp>
 #include <com/sun/star/style/ParagraphAdjust.hpp>
 #include <com/sun/star/style/TabStop.hpp>
 #include <com/sun/star/table/BorderLine2.hpp>
@@ -1562,6 +1564,35 @@ DECLARE_RTFIMPORT_TEST(testDefaultValues, "default-values.rtf")
     CPPUNIT_ASSERT_EQUAL(sal_Int32(COL_AUTO), getProperty<sal_Int32>(run, "CharColor"));
 }
 
+DECLARE_RTFIMPORT_TEST(testParaStyleBottomMargin, "para-style-bottom-margin.rtf")
+{
+    uno::Reference<beans::XPropertySet> xPropertySet(
+        getStyles("ParagraphStyles")->getByName("Standard"), uno::UNO_QUERY);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(353), getProperty<sal_Int32>(xPropertySet, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP,
+                         getProperty<style::LineSpacing>(xPropertySet, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(115),
+                         getProperty<style::LineSpacing>(xPropertySet, "ParaLineSpacing").Height);
+
+    // The reason why this is 0 despite the default style containing \sa200
+    // is that Word will actually interpret \basedonN
+    // as "set style N and for every attribute of that style,
+    // set an attribute with default value on the style"
+    uno::Reference<beans::XPropertySet> xPropertySet1(
+        getStyles("ParagraphStyles")->getByName("Contents 1"), uno::UNO_QUERY);
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(0), getProperty<sal_Int32>(xPropertySet1, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP,
+                         getProperty<style::LineSpacing>(xPropertySet1, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(100),
+                         getProperty<style::LineSpacing>(xPropertySet1, "ParaLineSpacing").Height);
+    auto const xPara(getParagraph(1));
+    CPPUNIT_ASSERT_EQUAL(sal_Int32(0), getProperty<sal_Int32>(xPara, "ParaBottomMargin"));
+    CPPUNIT_ASSERT_EQUAL(style::LineSpacingMode::PROP, // 0 or 3 ???
+                         getProperty<style::LineSpacing>(xPara, "ParaLineSpacing").Mode);
+    CPPUNIT_ASSERT_EQUAL(sal_Int16(100),
+                         getProperty<style::LineSpacing>(xPara, "ParaLineSpacing").Height);
+}
+
 // tests should only be added to rtfIMPORT *if* they fail round-tripping in rtfEXPORT
 
 CPPUNIT_PLUGIN_IMPLEMENT();

writerfilter/source/rtftok/rtfdocumentimpl.cxx

@@ -522,8 +522,9 @@ RTFDocumentImpl::getProperties(const RTFSprms& rAttributes, RTFSprms const& rSpr
         }
 
         // Get rid of direct formatting what is already in the style.
-        RTFSprms const sprms(aSprms.cloneAndDeduplicate(aStyleSprms, nStyleType));
-        RTFSprms const attributes(rAttributes.cloneAndDeduplicate(aStyleAttributes, nStyleType));
+        RTFSprms const sprms(aSprms.cloneAndDeduplicate(aStyleSprms, nStyleType, true));
+        RTFSprms const attributes(
+            rAttributes.cloneAndDeduplicate(aStyleAttributes, nStyleType, true));
         return new RTFReferenceProperties(attributes, sprms);
     }
 
@@ -2031,6 +2032,61 @@ writerfilter::Reference<Properties>::Pointer_t RTFDocumentImpl::createStylePrope
     return pProps;
 }
 
+/** 2 different representations of the styles are needed:
+
+    1) flat content, as read from the input file:
+       stored in m_aStyleTableEntries, used as reference input for
+       deduplication both here and for hard formatting in getProperties()
+
+    2) real content, with proper override of sprms/attributes where it differs
+       from parent style; this is produced here and sent to domain mapper
+ */
+RTFReferenceTable::Entries_t RTFDocumentImpl::deduplicateStyleTable()
+{
+    RTFReferenceTable::Entries_t ret;
+    for (auto const& it : m_aStyleTableEntries)
+    {
+        auto pStyle = it.second;
+        // ugly downcasts here, but can't easily replace the members with
+        // RTFReferenceProperties because dmapper wants SvRef<Properties> anyway
+        RTFValue::Pointer_t const pBasedOn(
+            static_cast<RTFReferenceProperties&>(*pStyle).getSprms().find(
+                NS_ooxml::LN_CT_Style_basedOn));
+        if (pBasedOn)
+        {
+            int const nBasedOn(pBasedOn->getInt());
+            auto const itParent(m_aStyleTableEntries.find(nBasedOn)); // definition as read!
+            if (itParent != m_aStyleTableEntries.end())
+            {
+                auto const pStyleType(
+                    static_cast<RTFReferenceProperties&>(*pStyle).getAttributes().find(
+                        NS_ooxml::LN_CT_Style_type));
+                assert(pStyleType);
+                int const nStyleType(pStyleType->getInt());
+                RTFSprms const sprms(
+                    static_cast<RTFReferenceProperties&>(*pStyle).getSprms().cloneAndDeduplicate(
+                        static_cast<RTFReferenceProperties&>(*itParent->second).getSprms(),
+                        nStyleType));
+                RTFSprms const attributes(
+                    static_cast<RTFReferenceProperties&>(*pStyle)
+                        .getAttributes()
+                        .cloneAndDeduplicate(
+                            static_cast<RTFReferenceProperties&>(*itParent->second).getAttributes(),
+                            nStyleType));
+
+                pStyle = new RTFReferenceProperties(attributes, sprms);
+            }
+            else
+            {
+                SAL_WARN("writerfilter.rtf", "parent style not found: " << nBasedOn);
+            }
+        }
+        ret[it.first] = pStyle;
+    }
+    assert(ret.size() == m_aStyleTableEntries.size());
+    return ret;
+}
+
 void RTFDocumentImpl::resetSprms()
 {
     m_aStates.top().aTableSprms.clear();
@@ -2094,8 +2150,9 @@ RTFError RTFDocumentImpl::popState()
         break;
         case Destination::STYLESHEET:
         {
+            RTFReferenceTable::Entries_t const pStyleTableDeduplicated(deduplicateStyleTable());
             writerfilter::Reference<Table>::Pointer_t const pTable(
-                new RTFReferenceTable(m_aStyleTableEntries));
+                new RTFReferenceTable(pStyleTableDeduplicated));
             Mapper().table(NS_ooxml::LN_STYLESHEET, pTable);
         }
         break;

writerfilter/source/rtftok/rtfdocumentimpl.hxx

@@ -633,6 +633,8 @@ public:
     /// Buffers properties to be sent later.
     void bufferProperties(RTFBuffer_t& rBuffer, const RTFValue::Pointer_t& pValue,
                           const tools::SvRef<TableRowBuffer>& pTableProperties);
+    /// implement non-obvious RTF specific style inheritance
+    RTFReferenceTable::Entries_t deduplicateStyleTable();
 
 private:
     SvStream& Strm();

writerfilter/source/rtftok/rtfsprm.cxx

@@ -160,6 +160,12 @@ static RTFValue::Pointer_t getDefaultSPRM(Id const id, Id nStyleType)
             case NS_ooxml::LN_CT_Ind_firstLine:
                 return new RTFValue(0);
 
+            case NS_ooxml::LN_CT_Spacing_lineRule:
+                return new RTFValue(NS_ooxml::LN_Value_doc_ST_LineSpacingRule_auto);
+            case NS_ooxml::LN_CT_Spacing_line:
+                // presumably this means 100%, cf. static const int nSingleLineSpacing = 240;
+                return new RTFValue(240);
+
             default:
                 break;
         }
@@ -338,7 +344,8 @@ void RTFSprms::duplicateList(const RTFValue::Pointer_t& pAbstract)
     }
 }
 
-RTFSprms RTFSprms::cloneAndDeduplicate(RTFSprms& rReference, Id nStyleType) const
+RTFSprms RTFSprms::cloneAndDeduplicate(RTFSprms& rReference, Id const nStyleType,
+                                       bool const bImplicitPPr) const
 {
     RTFSprms ret(*this);
     ret.ensureCopyBeforeWrite();
@@ -351,7 +358,7 @@ RTFSprms RTFSprms::cloneAndDeduplicate(RTFSprms& rReference, Id nStyleType) cons
         // paragraphs, but they are below NS_ooxml::LN_CT_Style_pPr in case of
         // styles. So handle those children directly, to avoid unexpected
         // addition of direct formatting sprms at the paragraph level.
-        if (rSprm.first == NS_ooxml::LN_CT_Style_pPr)
+        if (bImplicitPPr && rSprm.first == NS_ooxml::LN_CT_Style_pPr)
         {
             for (auto& i : rSprm.second->getSprms())
                 cloneAndDeduplicateSprm(i, ret, nStyleType);

writerfilter/source/rtftok/rtfsprm.hxx

@@ -62,7 +62,9 @@ public:
     /// Removes elements which are already in the reference set.
     /// Also insert default values to override attributes of style
     /// (yes, really; that's what Word does).
-    RTFSprms cloneAndDeduplicate(RTFSprms& rReference, Id nStyleType) const;
+    /// @param bImplicitPPr implicit dereference of top-level pPr SPRM
+    RTFSprms cloneAndDeduplicate(RTFSprms& rReference, Id nStyleType,
+                                 bool bImplicitPPr = false) const;
     /// Inserts default values to override attributes of pAbstract.
     void duplicateList(const RTFValue::Pointer_t& pAbstract);
     /// Removes duplicated values based on in-list properties.